Ophiolites in earth history

Ophiolites in Earth History

Geological Society Special Publications Society Book Editors R. J. PANKHURST (CHIEF EDITOR) P. DOYLE F. J. GREGORY J. S. GRIFFITHS A. J. HARTLEY R. E. HOLDSWORTH

J. A. HOWE P. T. LEAT A. C. MORTON N. S. ROBINS J. P. TURNER

Special Publication reviewing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society's Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society has a team of Book Editors (listed above) who ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees' forms and comments must be available to the Society's Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. Geological Society Special Publications are included in the ISI Index of Scientific Book Contents, but they do not have an impact factor, the latter being applicable only to journals. More information about submitting a proposal and producing a Special Publication can be found on the Society's web site: www.geolsoc.org.uk.

It is recommended that reference to all or part of this book should be made in one of the following ways: DILEK, Y. & ROBINSON, P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218. BAZYLEV, B. A., KARAMATA, S. & ZAKARIADZE, G. S. (2003). Petrology and evolution of the Brezovica ultramafic massif, Serbia. In: DILEK, Y. & ROBINSON, P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 91-108.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 218

Ophiolites in Earth History EDITED BY Y. DILEK Miami University, USA and

P. T. ROBINSON Dalhousie University, Canada

2003 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society's fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society's international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the American Geological Institute (AGI), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists' Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies' publications at a discount. The Society's online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J OBG: Tel. +44 (0)20 7434 9944; Fax +44 (0)20 7439 8975; E-mail: [email protected]. For information about the Society's meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society's Corporate Affiliates Scheme, write to [email protected]. Published by The Geological Society from: The Geological Society Publishing House Unit 7, Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN, UK (Orders: Tel. +44 (0)1225 445046 Fax+44 (0)1225 442836) Online bookshop: http://bookshop.geolsoc.org.uk The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. © The Geological Society of London 2003. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 03058719/03/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 1-86239-145-9 Typeset by Keytec Typesetting Ltd, Bridport, UK Printed by Cromwell Press, Trowbridge, UK.

Distributors USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA Orders: Tel. + 1 918 584-2555 Fax+1 918560-2652 E-mail [email protected] India Affiliated East-West Press PVT Ltd G-l/16 Ansari Road, Daryaganj, New Delhi 110002 India Orders: Tel. +91 11 2327-9113 Fax+91 11 2326-0538 E-mail [email protected] Japan Kanda Book Trading Company Cityhouse Tama 204 Tsurumaki 1-3-10 Tama-shi, Tokyo 206-0034 Japan Orders'. Tel. +81 (0)423 57-7650 Fax+81(0)423 57-7651 Email [email protected]

Contents Preface

ix

Introduction DlLEK, Y. & ROBINSON, P. T. Ophiolites in Earth history: introduction

1

DILEK, Y. Ophiolite pulses, mantle plumes and orogeny

9

Tethyan ophiolites in the Alpine-Himalayan orogenic system FLOWER, M. F. J. & DILEK, Y. Arc-trench rollback and forearc accretion: 1. A collisioninduced mantle flow model for Tethyan ophiolites

21

DILEK, Y. & FLOWER, M. F. J. Arc-trench rollback and forearc accretion: 2. A model template for ophiolites in Albania, Cyprus, and Oman

43

MUNTENER, O. & PICCARDO, G. B. Melt migration in ophiolitic peridotites: the message from Alpine-Apennine peridotites and implications for embryonic ocean basins

69

BAZYLEV, B. A., KARAMATA, S. & ZAKARIADZE, G. S. Petrology and evolution of the Brezovica ultramafic massif, Serbia

91

SACCANI, E., PADOA, E. & PHOTIADES, A. Triassic mid-ocean ridge basalts from the Argolis Peninsula (Greece): new constraints for the early oceanization phases of the NeoTethyan Pindos basin

109

SARKARINEJAD, K. Structural and microstructural analysis of a palaeo-transform fault zone in the Neyriz ophiolite, Iran

129

AITCHISON, J. C, DAVIS, A. M., ABRAJEVITCH, A. V, An, J. R., BADENGZHU, Liu, J., Luo, H., McDERMlD, I. R. C. & ZIABREV, S. V Stratigraphic and sedimentological constraints on the age and tectonic evolution of the Neotethyan ophiolites along the Yarlung Tsangpo suture Zone, Tibet

147

HEBERT, R., HUOT, F, WANG, C. & Liu, Z. Yarlung Zangbo ophiolites (Southern Tibet) revisited: geodynamic implications from the mineral record

165

MALPAS, I, ZHOU, M.-F, ROBINSON, P. T. & REYNOLDS, P. H. Geochemical and geochronological constraints on the origin and emplacement of the Yarlung Zangbo ophiolites, Southern Tibet

191

Magmatic, metamorphic and tectonic processes in ophiolite genesis HARPER, G. D. Tectonic implications of boninite, arc tholeiite, and MORB magma types in the Josephine Ophiolite, California-Oregon

207

SCHROETTER, J. M., PAGE, P., BEDARD, J. H., TREMBLAY, A. & BECU, V Forearc extension and sea-floor spreading in the Thetford Mines Ophiolite Complex

231

RAYMOND, L. A., SWANSON, S. E., LOVE A. B. & ALLAN, J. F Cr-spinel compositions, metadunite petrology, and the petrotectonic history of Blue Ridge ophiolites, Southern Appalachian Orogen, USA

253

HIRANO, N., OGAWA, Y, SAITO, K., YOSHIDA, T, SATO, H. & TANIGUCHI, H. Multi-stage evolution of the Tertiary Mineoka ophiolite, Japan: new geochemical and age constraints

279

TAKAHASHI, A., OGAWA, Y, OHTA, Y & HIRANO, N. The nature of faulting and deformation in the Mineoka ophiolite, NW Pacific Rim

299

vi

CONTENTS

STAKES, D. S. & TAYLOR, H. P. Jr Oxygen isotope and chemical studies on the origin of large plagiogranite bodies in northern Oman, and their relationship to the overlying massive sulphide deposits

315

Hydrothermal and biogenic alteration of oceanic crust as recorded in ophiolites GREGORY, R. T. Ophiolites and global geochemical cycles: implications for the isotopic evolution of seawater

353

GIGUERE, E., HEBERT, R., BEAUDOIN, G., BEDARD, J. H. & BERCLAZ, A. Hydrothermal circulation and metamorphism in crustal gabbroic rocks of the Bay of Islands ophiolite complex, Newfoundland, Canada: evidence from mineral and oxygen isotope geochemistry

369

MUEHLENBACHS, K., FURNES, H., FONNELAND, H. C. & HELLEVANG, B. Ophiolites as

401

faithful records of the oxygen isotope ratio of ancient seawater: the Solund-Stavfjord Ophiolite Complex as a Late Ordovician example FURNES, H. & MUEHLENBACHS, K. Bioalteration recorded in ophiolitic pillow lavas

415

Ophiolite emplacement: mechanisms and processes WAKABAYASHI, J. & DILEK, Y. What constitutes 'emplacement' of an ophiolite?: Mechanisms and relationship to subduction initiation and formation of metamorphic soles

427

GRAY, D. R. & GREGORY R. T. Ophiolite obduction and the Samail Ophiolite: the behaviour of the underlying margin

449

SEARLE, M. P., WARREN, C. I, WATERS, D. J. & PARRISH, R. R. Subduction zone polarity in the Oman Mountains: implications for ophiolite emplacement

467

Regional occurrence of ophiolites and geodynamics HARRIS, R. Geodynamic patterns of ophiolites and marginal basins in the Indonesian and New Guinea regions

481

MlLSOM, J. Forearc ophiolites: a view from the western Pacific

507

SPAGGIARI, C. V, GRAY, D. R. & FOSTER, D. A. Tethyan- and Cordilleran-type ophiolites of eastern Australia: implications for the evolution of the Tasmanides

517

ZHANG, Q., WANG, Y., ZHOU, G. Q., QIAN, Q. & ROBINSON P. T. Ophiolites in China: their distribution, ages and tectonic settings

541

SPADEA, P., ZANETTI, A. & VANNUCCI, R. Mineral chemistry of ultramafic massifs in the Southern Uralides orogenic belt (Russia) and the petrogenesis of the Lower Palaeozoic ophiolites of the Uralian Ocean

567

ISHIWATARI, A., SOKOLOV, S. D. & VYSOTSKIY S. V Petrological diversity and origin of ophiolites in Japan and Far East Russia with emphasis on depleted harzburgite

597

SOKOLOV, S. D., LUCHITSKAYA, M. V, SILANTYEV, S. A., MOROZOV, O. L., GANELIN, A. V, BAZYLEV, B. A., OSIPENKO, A. B., PALANDZHYAN, S. A. & KRAVCHENKO-BEREZHNOY, I. R. Ophiolites in accretionary complexes along the Early Cretaceous margin of NE Asia: age, composition, and geodynamic diversity

619

STERN, C. R. & DE WIT, M. J. Rocas Verdes ophiolites, southernmost South America: remnants of progressive stages of development of oceanic-type crust in a continental margin back-arc basin

665

DILEK, Y. & AHMED, Z. Proterozoic ophiolites of the Arabian Shield and their significance in Precambrian tectonics

685

Preface

This book is derived from the interdisciplinary, contemporary work of the international ophiolite community in a most up-to-date treatment of process-oriented problems and questions on the generation and evolution of ophiolites. It is a large collection of research papers from a wide range of international contributors. Some of these papers were presented in thematic ophiolite sessions at the 2001 Annual Meeting of the Geological Society of America (Boston) and the 2001 Fall Meeting of the American Geophysical Union (San Francisco). The 32 papers here examine the mode and nature of igneous, metamorphic, tectonic, sedimentological, and biological processes associated with the evolution of oceanic crust in different tectonic settings in Earth history as revealed in various ophiolites and ophiolite belts around the world, and the geodynamic significance of these ophiolites in the evolution of different orogenic systems. Divided into six thematic sections, the book presents a wealth of new data and syntheses from mainly Phanerozoic ophiolites around the world. We would like to express our thanks to the contributors to this book for their time and effort. We also would like to extend our sincere appreciation and gratitude to Angharad Hills (Staff Editor) and Andy Morton (Book Series Editor) for their help and advice at review stages, and to the Geological Society Publishing House staff for their support in the publication process. Diligent work by Senior Production Editor Sarah Gibbs at all stages throughout the preparation and reproduction of this book contributed to its success. Cathy Edwards in the Geology Department at Miami University helped with manuscript preparation and proofreading of the chapters. The Office of Advancement of Research and Scholarship, the College of Arts and Science, and the Department of Geology at Miami University provided partial financial support for the preparation of the book that we gratefully acknowledge. We wish to thank the following colleagues for their timely and thorough reviews of the manuscripts that helped us maintain the high scientific standards for which we have striven: James Allan (Appalachian State University, USA); Jeffrey C. Alt (University of Michigan, USA); Shoji Arai (Kanazawa University, Japan); Neil Banerjee (University of Alberta, Canada); Asish Basu (University of Rochester, New York, USA); Jean Bebien (Universite de Paris-Sud, Orsay, France); Manuel Berberian (New Jersey, USA); Sherman Bloomer (Oregon State University, USA); Craig Buchan (Curtin University of Technology, Australia); Sun-Lin Chung (National Taiwan University, Taiwan); Ian WD. Dalziel (University of Texas at Austin, USA); Hugh Davies (University of Papua New Guinea);

Yildirim Dilek (Miami University, USA); Jaroslav Dostal (St. Mary's University, Canada); Grenville Draper (Florida International University, USA); Stephen Edwards (University of Greenwich, England); Don Elthon (University of Houston, USA); John Encarnacion (St. Louis University, USA); Martin Fisk (Oregon State University, USA); Martin F.J. Flower (University of Illinois at Chicago, USA); Gretchen Frueh-Green (ETH-Zentrum, Switzerland); Ulrich Glasmacher (Germany); David Gray (University of Melbourne, Australia); Ron Harris (Brigham Young University, USA); Kendall Hauer (Miami University, USA); James W. Hawkins (Scripps Institution of Oceanography, California, USA); Rejean Hebert (Universite Laval, Quebec, Canada); Rod Holcombe (University of Queensland, Australia); Paul Holm (Earlham College, USA); Francois Huot (Universite Laval, Quebec, Canada); Akira Ishiwatari (Kanazawa University, Japan); Barbara John (University of Wyoming, USA); Thierry Juteau (IUEM, Plouzane, France); Ade Kadarusman (Tokyo Institute of Technology, Japan); Andrew Kerr (Cardiff University, UK); Elena Konstantinovskaia (Russian Academy of Sciences, Moscow-Russia); John Malpas (University of Hong Kong, China); Catherine Mevel (Institute de Physique du Globe, Paris-France); Calvin Miller (Vanderbilt University, USA); John Milsom (University College London, UK); Eldridge Moores (University of California at Davis, USA); Kula Misra (University of Tennessee, USA); Karlis Muehlenbachs (University of Alberta, Canada); Christopher Parkinson (University of New Orleans, USA); Gene Perry (Northern Illinois University, USA); Tjerk Peters (Universitat Bern, Switzerland); Ali Polat (University of Windsor, Canada); Elisabetta Rampone (University of Genova, Italy); Paul T. Robinson (Dalhousie University, Canada); Andrew Saunders (University of Leicester, UK); Peter Schiffman (University of California at Davis, USA); Richard Sedlock (San Jose State University, California, USA); John Shervais (Utah State University, USA); Eli Silver (University of California at Santa Cruz, USA); Ian Smith (The University of Auckland, New Zealand); Piera Spadea (Universita di Udine, Italy); Catherine Spaggiari (Monash University, Australia); Debra Stakes (MBARI, California, USA); Hubert Staudigel (Scripps Institution of Oceanography, California, USA); Charles Stern (University of Colorado, USA); Mohamed Sultan (University at Buffalo, New York, USA); Damon A. H. Teagle (University of Southampton, UK); David Vanko (Towson University, Maryland); John Wakabayashi (Hayword, USA); and Steve Wojtal (Oberlin College, USA). Yildirim Dilek Oxford, USA, October 2003

This page intentionally left blank

Ophiolites in Earth history: introduction YILDIRIM DILEK 1 & PAUL T. ROBINSON 2 Department of Geology, Miami University, Oxford, OH 45056, USA (e-mail: [email protected]) 2 Department of Earth Sciences, Dalhousie University, Halifax, N.S. B3H 3J5, Canada 1

Ophiolites record significant evidence for tectonic and magmatic processes from rift-drift through accretionary and collisional stages of continental margin evolution in various tectonic settings. Structural, petrological and geochemical features of Ophiolites and associated rock units provide essential information on mantle flow field effects, including plume activities, collision-induced aesthenospheric extrusion, crustal growth via magmatism and tectonic accretion in subductionaccretion cycles, changes in the structure and composition of the crust and mantle reservoirs through time, and evolution of global geochemical cycles and seawater compositions. Ophiolite studies over the years have played a major role in better understanding of mid-ocean ridge and subduction zone processes, mantle dynamics and heterogeneity, magma chamber processes, fluid flow mechanisms and fluid-rock interactions in oceanic lithosphere, the evolution of deep biosphere, the role of plate tectonics and plume tectonics in crustal evolution during the Precambrian and the Phanerozoic, and mechanisms of continental growth in accretionary and collisional mountain belts. Through multi-disciplinary investigations and comparative studies of Ophiolites and modern oceanic crust and using advanced instrumentation and computational facilities, the international ophiolite community has gathered a wealth of new data and syntheses from Ophiolites around the world during the last 10 years. The purpose of this book is to present the most recent data, observations and ideas on different aspects of 'ophiolite science' through case studies and to document the mode and nature of igneous, metamorphic, tectonic, sedimentological and/or biological processes associated with the evolution of oceanic crust in different tectonic settings in Earth's history. It comprises 32 papers collected in six sections on temporal relations amongst ophiolite genesis, mantle plume events and orogeny in Earth history; Tethyan ophiolites in the Alpine Himalayan orogenic system; magmatic, metamorphic and tectonic processes in ophiolite genesis; hydrothermal and biogenic alteration of oceanic crust; mechanisms of ophiolite emplace-

ment; and regional occurrences of ophiolites and their geodynamic implications.

Ophiolites, mantle plumes and orogeny Ophiolite occurrences around the world are not a random geological phenomenon. Ophiolites with certain age groups in different orogenic belts characterize distinct ophiolite pulses, which mark times of enhanced ophiolite genesis and emplacement. Examining the geological record of mountain-building episodes and related events, Dilek shows that ophiolite pulses overlap significantly with the timing of major collisional events during the assembly of supercontinents, their break-up and increased mantle plume activities that developed extensive large igneous provinces (LIPs). These global events have been involved in the Wilson cycle evolution of ancient ocean basins that in turn contributed to ophiolite genesis in diverse tectonic settings. Suprasubduction zone ophiolites represent anomalous oceanic crust generation in subduction rollback cycles during the closing stages of basins prior to terminal continental collisions. Accelerated LIP formation associated with superplume activities may have facilitated both the generation and tectonic emplacement of ophiolites at global scales. These spatial and temporal relations suggest that ophiolite pulses, mantle plume activities and orogenic events have been closely linked through complex mantle dynamics in Earth history.

Tethyan ophiolites in the AlpineHimalayan orogenic system Papers in this section present diverse data from Tethyan ophiolites and provide refined geodynamic models for their evolution. Flower & Dilek examine the processes of arc-trench rollback and forearc accretion, and present an 'actualistic' model for ophiolites based on recent observations of forearc evolution in western Pacific and Mediterranean marginal basins. Collision-induced mantle flow and 'slab-pull' forces may result in rapid

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 1-8. 0305-8719/03/$15 © The Geological Society of London 2003.

2

Y. DILEK & P. T. ROBINSON

arc-trench rollback pulses and associated extensional episodes (splitting of nascent volcanic proto-arcs), producing proto-ophiolites in arc-forearc settings. These ophiolites commonly include hightemperature metamorphic soles, boninitic rocks, juxtaposed refractory peridotites and high-temperature epidosites that are generally absent in mid-ocean ridge, normal arc and back-arc basin environments. As subduction rollback continues, arc-forearc complexes become increasingly heterogeneous, displaying significant internal age and structural discrepancies, a common feature both in the SW Pacific subduction zone environments and Tethyan ophiolites. When an arc-trench rollback cycle is terminated by a collision, heterogeneous forearc lithosphere is accreted as ophiolites in the initial stages of the evolution of collisional orogenic belts. This model demonstrates the apparent correspondence of subduction nucleation and mantle flow to plate collisions at regional and global scales. In a companion paper, Dilek & Flower explore the application of the arc-trench rollback and forearc accretion model to Neo-Tethyan ophiolites, specifically to the Mirdita (Albania), Troodos (Cyprus) and Semail (Oman) ophiolites. NeoTethyan oceans evolved as east-west-oriented basins separated by discrete continental fragments, which were rifted off from the northern edge of Gondwana beginning in the Triassic. Triassic rift assemblages containing within-plate-type alkaline basalt to transitional (T-MORB) and mid-ocean ridge basalt (MORE) are spatially associated with ophiolites in the eastern Mediterranean region and may represent the precursor of Late Triassic oceanic crust, which was subsequently consumed to produce the suprasubduction zone ophiolites. The three ophiolites examined here include a basement of typical 'oceanic' lithosphere intruded and overlain by boninitic (ultra-refractory) to calcalkaline series rocks that formed in a proto-arcforearc setting. This progression was a result of upper plate extension and further melting of previously depleted asthenosphere that occurred in response to successive stages of slab rollback. This igneous evolution of the ophiolites involved subduction initiation and one or more episodes of proto-arc splitting before the termination of slab rollback cycles as a result of trench-continent collisions. Miintener & Piccardo examine the Lanzo and Corsica ophiolitic peridotites in the Alpine-Apennine mountain system that are interpreted as remnants of the Ligurian Tethys. The texture, geochemistry and petrology of these peridotites suggest that they represent exhumed subcontinental lithospheric mantle, which was modified and refertilized by migrating melts during opening of

the embryonic Piedmont-Ligurian Ocean. Pervasive melt infiltration and melt-rock reaction produced gabbroic intrusions with a wide range of compositions characteristic of the melting column beneath mid-ocean ridges. These observations are critical to better understand the effects of melt percolation and impregnation in development of plagioclase-enriched peridotites. The Ligurian ophiolites clearly do not represent a typical Penrose-type, idealized oceanic crust. Bazylev et al. present mineral and bulk-rock chemistry data from the Jurassic Brezovica ultramafic massif (Serbia) in the Dinarides and show that its petrogenetic evolution involved two distinct magmatic stages. A suite of spinel harzburgites was produced during the first stage as a result of partial melting of the mantle and segregation of tholeiitic melts. Percolation of melt through these spinel harzburgites and melt-rock reaction produced dunites and refractory harzburgites during the second stage and generated highCa boninitic melt. The authors conclude that the second magmatic stage had to occur in a suprasubduction zone setting. Saccani et al. present new field and geochemical constraints from the Western Hellenides in Greece, documenting that initial stages of seafloor spreading and oceanic crust formation in the Pindos basin probably occurred in the Mid- to Late Triassic, earlier than previously thought. Pillow lavas from the Argolis Peninsula have MORB trace element characteristics and are divided into T-MORB and normal MORB (NMORB). These are the oldest unequivocally dated oceanic crust in the Hellenide sector of the Pindos Basin. Early Triassic rifting produced shoshonitic and calc-alkaline lavas derived from a mantle source that was previously contaminated by subduction components. Associated alkaline basalts were derived from ocean island basalt-type (OIB) mantle source. Mixing of mantle sources produced enriched MORB (E-MORB) and T-MORB, and then N-MORB lavas were erupted in Mid(?)- to Late Triassic, suggesting that sea-floor spreading had reached a steady state. The authors cite the Red Sea as a modern analogue with along-strike chemical variations for the Pindos Basin. Sarkarinejad describes the internal structure of the Cretaceous Neyriz ophiolite in southern Iran, and presents structural and microstructural observations for the existence of a NW-trending palaeotransform fault zone within this Neo-Tethyan ophiolite. Fabric analysis of mylonitic rocks (including hornblende and plagioclase textures and chemistry) suggests that the plastic deformation of mafic-ultramafic rocks occurred at amphibolitefacies conditions within a dextrally slipping oceanic transform fault zone. The author infers that

INTRODUCTION the Neyriz transform fault separated ENE-trending spreading centre segments within a Neo-Tethyan basin. The last three papers in this section present diverse stratigraphic, petrological, geochemical and geochronological data from the YarlungTsangpo suture zone ophiolites in southern Tibet. Aitchison et al. define several discrete terranes along the suture zone and use their sedimentological and biostratigraphic data to constrain the timing of ophiolite formation and terrane accretion within this segment of the HimalayanTibetan orogenic belt. Different ages of ophiolitic assemblages from Xigaze, Jungwa and Zedong indicate that the suture zone may contain remnants of multiple (two?) island arc complexes that had evolved within the same branch of Neo-Tethys. Hebert et al. report mineral chemistry data and petrological findings from mafic-ultramafic rocks of the Yarlung Tsangpo ophiolites. Mantle peridotites were exhumed from depths of more than 50 km and underwent 10-40% partial melting and melt percolation within a suprasubduction zone wedge. The Yarlung Tsangpo ophiolites represent a heterogeneous collage of arc, forearc and backarc oceanic lithosphere developed in a NeoTethyan basin south of the active continental margin of Eurasia. Malpas et al. present new geochronological data from the Yarlung-Tsangpo ophiolites and a refined geodynamic model for their evolution. The new sensitive high-resolution ion microprobe date of 126 Ma for the Dazhuqu massif indicates that the Xigaze ophiolite is significantly younger than the Loubusa ophiolite and Zedong island arc complex (c. 175 Ma). These findings are consistent with the geochemical interpretations of Hebert et al. Basaltic rocks from all ophiolites are composed of island arc tholeiites, and the peridotites show textural and chemical evidence for percolation of boninitic melts through the upper mantle at later stages of magmatism. The Yarlung-Tsangpo ophiolites may have formed at different times in suprasubduction zone environments and were subsequently juxtaposed during the collision of the Indian continental margin with the arc-trench system around 90-80 Ma.

Magmatic, metamorphic and tectonic processes in ophiolite genesis The six papers in this section present processoriented case studies of oceanic crust evolution from the Appalachian, Cordilleran, Tethyan and Japanese ophiolites. Harper demonstrates that the extrusive sequence and sheeted dyke complex in the Jurassic Josephine ophiolite in CaliforniaOregon (USA) display chemical evidence for a

3

wide range in magma types and degree of fractionation. New discoveries of Fe-Ti-rich and Ti-poor (boninitic) magmas in the Josephine ophiolite illustrate its compositional complexity and provide new constraints on its tectonic environment of formation. The Fe-Ti lavas imply formation along a propagating rift, whereas the low-Ti lavas suggest a forearc environment of their origin. The Lau Basin is cited as a likely modern analogue because the available geochemical data from several environments within this modern back-arc basin are consistent with the new chemical data and interpretations from the Josephine ophiolite. Northern Tonga and the Andaman Sea may also be plausible analogues for the Josephine ophiolite. Schroetter et al. examine the internal structure and stratigraphy of the Ordovician Thetford Mines ophiolite in Quebec (Canada). The discovery of a locally well-developed sheeted dyke complex, combined with other structural data, indicates that the Ordovician oceanic crust was developed at a slow-spreading centre, where faulting and magmatism were coeval, keeping pace with crustal extension. The boninitic affinity of cumulate rocks and lavas suggests that the Thetford Mines ophiolite probably formed in a forearc setting. This is one of the best-documented cases of well-established pre-collisional extensional tectonics in a palaeoforearc environment. Raymond et al. investigate the occurrence and petrogenesis of ultramafic rock bodies in the Southern Appalachian (USA) orogenic belt. These ultramafic rocks are part of dismembered Ordovician ophiolites, which probably formed in a slowspreading centre within a subduction zone setting. A suprasubduction zone environment of origin is supported by the existence of metadunites representing sublithospheric melt channels and zones of high melt flux. The authors suggest that the Taconic subduction zone that was responsible for the formation of the Southern Appalachian ophiolites may have been west-directed, rather than east-directed as previous models have inferred. Hirano et al. show that the Tertiary Mineoka ophiolite in central Japan had a multi-stage tectonic evolution prior to its emplacement onto the Japanese continental margin. It occurs near a trench-trench-trench triple junction and contains tholeiitic pillow basalts and dolerites, calc-alkaline plutonic rocks and alkali-basaltic sheet flows. The sea-floor spreading stage of the ophiolite probably occurred during the generation of an oceanic Mineoka Plate in the Eocene. Subduction of the Pacific Plate beneath the Mineoka Plate produced island arc volcanism during 40-25 Ma (second stage). Eruption of the within-plate-type alkali basalts (WPB) during the third stage occurred around 20 Ma, shortly before the emplacement of

4

Y. DILEK & P. T. ROBINSON

the polygenetic Mineoka ophiolite onto the continental margin. The ophiolite was derived from the Mineoka Plate, not from the Philippine Sea or Pacific Plates as previous models suggest. The companion paper by Takahashi et aL examines the internal structure of the Mineoka ophiolite and reports three main phases of deformation recorded by ophiolitic rocks. The first deformation phase was manifested in oblique normal faults and associated vein systems, and was associated with extensional tectonics at a palaeo-spreading centre. The second phase of deformation, characterized by thrust faults and strike-slip shear zones, was related to the emplacement of the ophiolite. The third phase of deformation is represented by transpressional dextral faults, manifestation of the modern tectonic regime in a trench-trench-trench triple junction. The last paper in this section, by Stakes & Taylor, documents the occurrence of large plagiogranite intrusions in the northern part of the Semail ophiolite (Oman) and their spatial and temporal association with the formation of massive sulphide deposits. Chemical, isotopic and field relations indicate that plagiogranite bodies near the overlying sheeted dykes formed through a complex process of combined assimilation and fractional crystallization, and recharge by injection of basaltic magma in open-system magma chambers. These plagiogranites were clearly late-stage magmatic products postdating the formation of the main ophiolitic crust and acted as shallow point sources of heat and metals for development of the overlying economic massive sulphide deposits.

Hydrothermal and biogenic alteration of oceanic crust as recorded in ophiolites The four papers in this section examine the nature, mechanisms and products of hydrothermal and biogenic alteration of oceanic crust and their implications for geochemical cycles in Earth history. Gregory demonstrates that the hydrothermal alteration history of ophiolites has major implications for the isotopic evolution of seawater. Isotopic profiles through ophiolites (e.g. Semail) show completely different characteristics depending on the element involved (Nd, Sr and O) and its residence time in the ocean. Oxygen isotopes are perhaps the most useful indicators of geochemical cycles and seawater-rock interaction. The mean value of altered oceanic crust is close to its primary 18O/16O ratio, which means that there must be complementary reservoirs of 18Odepleted and -enriched rocks in the altered ocean crust. Ophiolites are particularly useful because they are pieces of oceanic lithosphere that have

escaped recycling. Ophiolite studies show that oxygen isotopic composition of seawater resides at near steady-state conditions over Earth history. Giguere et aL present mineral and oxygen isotope geochemistry data from gabbroic rocks of the North Arm Mountain massif in the Bay of Islands ophiolite in Newfoundland (Canada) to constrain the chronology and temperature conditions of fluid circulation with respect to the timing and nature of deformation as recorded in these lower-crustal rocks. With continued cooling of gabbroic rocks, amphibole compositions changed as temperatures of amphibole formation fell steadily. Early amphiboles show near igneous oxygen isotope compositions typical of MORB or backarc basin basalt (BABB). Seawater infiltration into the lower crust occurred along listric shear zones under low fluid/rock ratios during the initial stages of deformation and metamorphism. Further cooling facilitated brittle deformation and greater seawater penetration at depth with increased fluid/ rock ratios, as suggested by very low 618O values. Field relations suggest that late-stage trondhjemitic intrusions may have provided heat and convective circulation of hydrothermal fluids causing high- T alteration superimposed on earlier stage of lower-T alteration. These relations clearly show that successive episodes of hydrothermal alteration of fossil lower crust in the Bay of Islands ophiolite were entirely intra-oceanic in origin. Muehlenbachs et aL use the hydrothermal alteration history of the Ordovician SolundStavfjord Ophiolite Complex (SSOC) in western Norway to examine the oxygen isotope ratio of ancient seawater. Similar to most ophiolites, the SSOC shows enrichment of 18O in the lavas altered at low temperatures and depletion in the dykes and gabbros altered at higher temperatures; this is also compatible with the alteration profile of 5.9 Ma in situ oceanic crust drilled in Ocean Drilling Program Hole 504B south of the Costa Rica Rift. Ophiolites can reflect the isotopic composition of ancient seawater. There is no observable secular trend in the 618O of seawater, and hence the mode and scale of seawatersea-floor interaction has not changed with time. The 618O of sediments and fossils may not record true values but rather owe their compositions to isotopic resetting, warmer oceans or biased sampling of restricted basins. Thus, models of ancient climates and ocean volumes determined from such data may be incorrect. Furnes & Muehlenbachs examine the nature and extent of bioalteration in fossil oceanic crust with different ages. Bioalteration of volcanic glass has been demonstrated in in situ oceanic crust but is not yet well documented from ophiolites. The authors have looked for evidence of bioalteration

INTRODUCTION

5

in glassy pillow lavas from four major ophiolites: Cretaceous Troodos (Cyprus), Jurassic Mirdita (Albania), Ordovician Solund-Stavfjord (western Norway) and early Proterozoic Jormua (Finland). Bioalteration may be recognized from textural evidence, organic remains, chemical fingerprints (C, N, S and P) and carbon isotopic signatures. Textural evidence in the form of coalesced spheres and tubes is present only in Troodos and Mirdita, the youngest of the ophiolites investigated. Some textural features in the SSOC resemble biogenerated textures, but rocks metamorphosed to amphibolite facies grade lack any evidence of bioalteration. Organic remains, in the form of twisted filaments, have been found only in Troodos. Probable organic carbon has been found in rocks from Troodos and the SSOC. Carbon isotope data in glassy samples are shifted to lower values and have a pattern very similar to that for in situ oceanic lavas. Evidence of bio-alteration appears to survive low-grade greenschist-facies metamorphism but is generally destroyed at higher grades of metamorphism.

structural data from rocks beneath the ophiolite nappe suggesting that there was an earlier period of underthrusting-subduction beneath the Arabian continental margin prior to its formation and obduction. Therefore, emplacement of the Semail nappe cannot simply be linked to a single subduction zone dipping away from the continent during the evolution of the ophiolite. The age of eclogite metamorphism in the lower-plate rocks beneath the ophiolite nappe (Saih Hatat Window) is crucial in testing this and other existing models. Searle et al. dispute this model by Gray & Gregory and discuss whether all structures and metamorphism observed in northern Oman are related to a single, prolonged episode of ophiolite emplacement, lasted for c. 27 million years and associated with a subduction zone dipping away from the Arabian continent. Suprasubduction zone origin of the ophiolite, metamorphic sole generation and eclogite formation were all linked to this subduction zone. Clearly, more precise age dating of the highpressure rocks beneath the ophiolite is needed to resolve the current debate.

Ophiolite emplacement: mechanisms and processes

Regional occurrence of ophiolites and geodynamic implications

Emplacement of ophiolites into continental margins is a first-order tectonic problem in plate tectonics and a significant phase in the evolution of orogenic belts. Ever since their recognition as on-land fragments of ancient oceanic lithosphere, mechanisms and processes involved in incorporation of ophiolites into continents have been a subject of discussion amongst researchers. The three papers in this section evaluate the existing models and ideas on ophiolite emplacement mechanisms with a focus on the Cretaceous Semail ophiolite in Oman. Wakabayashi & Dilek discuss the mechanisms and significance of subduction initiation and metamorphic sole formation in ophiolite emplacement and define four prototype ophiolites based on their emplacement mechanisms. Tethyan ophiolites are collisional-type emplaced over passive continental margins, whereas Cordilleran ophiolites are emplaced over subduction complexes through accretionary processes. Emplacement of ridge-trench intersection (RTI) ophiolites occurs through complex processes resulting from interaction of a spreading ridge and a subduction zone. Macquarie Island-type ophiolite represents oceanic crust exposed as a result of shifts in plate boundary configurations (i.e. spreading ridge segments converting into a diffuse transpressional plate boundary). Gray & Gregory review emplacement models for the Semail ophiolite in Oman and present

The papers in this section involve the regional occurrence of ophiolite belts on different continents and provide new petrological, geochemical and geochronological data and syntheses to better constrain their geodynamic evolution. Harris explores the spatial, temporal, geological and geochemical patterns of ophiolites in the Indonesian and New Guinea region (ING) in the first paper. ING is a repository of island arcs, marginal basins, continental fragments and ophiolites amalgamated by repeated plate boundary reorganizations. Major plate boundary reorganizations in the ING region coincide with global plate motions and there is a strong correlation in space and time between ophiolite genesis and collisional events. Opening of basins and suprasubduction zone generation of ophiolites are likely to have been 'enhanced' by extrusion of aesthenosphere escaping collisional zones in the region. Ophiolites forming in these suprasubduction zone environments display age and compositional heterogeneity, indicating their composite nature. Milsom examines the New Caledonia region in the SW Pacific to determine the spatial relations between forearc ophiolites and their volcanic arc systems. Repeated episodes of collisional events, postcollisional faulting and magmatism, and sea-floor spreading appear to have displaced and separated forearc tectonic assemblages from their respective volcanic arc systems in the New Caledonia-New

6

Y. DILEK & P. T. ROBINSON

Guinea region. This complex history may be responsible for the apparent lack of volcanic arc edifices associated with other forearc ophiolites (e.g. Troodos in Cyprus) around the world. Spaggiari et al. provide an overview of the Neoproterozoic to Cambrian ophiolites of the Tasmanides in eastern Australia, and examine the differences in their emplacement styles and tectonic settings. Eastern Australian ophiolites fall into Tethyan- and Cordilleran-type categories depending on their relationship to 'continental basement', and they appear to have developed in various suprasubduction zone environments (arc, forearc, back-arc) along the eastern Gondwana margin. Their age progression and geochemistry, combined with regional structural and tectonic constraints, suggest that they evolved in a complex rifted arc-back-arc system during 530-485 Ma, and that the collapse of this system into the continental margin of East Gondwana resulted in their emplacement. This event might have been related to far-field stresses associated with the collisional assembly of greater Gondwana in the early Palaeozoic. Zhang et al. summarize the regional distribution, ages and inferred tectonic settings of ophiolites in China. The Chinese ophiolites fall into four major age groups, Proterozoic, early Palaeozoic, late Palaeozoic and Mesozoic-Cenozoic, and they mainly occur along suture zones separating different tectonic blocks. They have a melange character in general and display structural and metamorphic evidence for multiple episodes of collisional events. The majority of the Chinese ophiolites are compositionally heterogeneous, containing mixtures of island arc tholeiite and boninite with lesser amounts of MORE and OIB. Palaeo-Tethyan ophiolites mostly have MORBtype rocks and may have formed in small intracontinental basins. Spadea et al. investigate the pyroxene and amphibole compositions of various mantle peridotites, particularly the Nurali and Mindyak massifs in the Southern Uralides in Russia. The Ural Mountains are a fold mountain system that records a Late Paleozoic arc-continent collision along the eastern European palaeomargin of Baltica. The Main Uralian Fault marks the related suture zone that consists of a melange composed of arc fragments and dismembered ophiolites. The Nurali and Mindyak peridotites have several anomalous features for abyssal peridotites: fertile composition; internal zoning from Iherzolite to dunite to harzburgite; anomalous crust-mantle transition with amphibole-bearing, plagioclase-free, ultramafic cumulates; lack of associated crustal section; and intrusion of late (400 Ma) gabbro-diorite plutons. These peridotite bodies underwent multi-

stage igneous events including porous flow, and rock-melt interaction involving pyroxene dissolution and plagioclase precipitation. They thus show some similarities to peridotites of subcontinental mantle and/or continent-ocean transition zone mantle. The authors present two explanations for the origin of these peridotite massifs in the Southern Uralides: (1) the anomalous features (for abyssal peridotites) reflect modification of normal MORE peridotites formed beneath a spreading axis by large volumes of island arc melts; or (2) the peridotites were originally part of subcontinental mantle, which underwent modification by dominantly tholeiitic melts causing plagioclase precipitation. Ishiwatari et al. discuss the petrological diversity and origin of ophiolites in Japan and Far East Russia, and distinguish highly depleted mantle harzburgite (DH) massifs in them. These ophiolites range in age from Early Palaeozoic to Cenozoic and are tectonically underlain by blueschist-bearing rocks and accretionary complexes that are generally younger in age. The majority of the ophiolites probably formed intra-oceanic island arc systems, as their petrological and geochemical characteristics suggest, and were incorporated into the Eurasian continental margin through repeated episodes of Mariana-type nonaccretionary subduction zone processes over time. There is little in the English literature on the ophiolite complexes of NE Asia. Sokolov et al. present new data on the age, structure and composition of ophiolites in the West Koryak fold belt in Far East Russia. The region consists chiefly of a variety of accreted terranes of different age and character. The ophiolites fall into two main categories. Palaeozoic ophiolites are primarily oceanic (MORB) in character and are viewed as fragments of the Panthalassa Ocean. Mesozoic ophiolites typically have an SSZ signature. In general, the ophiolites become younger towards the Pacific Ocean in the east. Accretionary prisms contain terrigeneous melanges similar to those of the Shimanto Belt of SW Japan. Stern & De Wit describe the geology and geochemistry of the Mesozoic Rocas Verdes ophiolites in the southernmost Andes (South America) and show that these ophiolites evolved in a Late Jurassic-Early Cretaceous intra-arc basin along the southern edge of Gondwana. Primary crosscutting relations of ophiolitic dyke swarms with the surrounding crystalline basement rocks of the Andean magmatic arc indicate that Rocas Verdes basin was an ensialic small ocean that opened up by 'unzipping' from the south to the north, synchronously with the onset of seafloor spreading in the South Atlantic at c. 132 Ma. Thus the Rocas Verdes ophiolites provide a unique

INTRODUCTION opportunity to investigate the mode and nature of igneous, metamorphic and tectonic processes associated with continental rifting, sea-floor spreading and tectonic collapse of a back-arc basin in an Andean-type active continental margin. Finally, Dilek & Ahmed present an overview of the Proterozoic ophiolites in the Arabian Shield and discuss their significance in Precambrian tectonics. The Arabian Shield ophiolites range in age from c. 870 Ma to c. 627 Ma and display a record of rift-drift, sea-floor spreading and collision tectonics during the evolution of the East African Orogen in the aftermath of the break-up of Rodinia. Ophiolites in the western part of the shield were part of ensimatic are terranes, which were sutured through a series of collisional events. Younger ophiolites in the eastern Arabian Shield were incorporated into accretionary complexes through offscraping and collisional events during continued subduction, similar to the accretionary history of those Phanerozoic ophiolites in NE Asia as reported by Sokolov et al. The youngest ophiolites in the shield (Nabitah-Hamdah fault zone ophiolites) are post-collisional in origin and they represent Ligurian-type oceanic crust developed in an intracontinental para-rift basin. The Arabian shield ophiolites are clearly diverse in origin and provide a great opportunity to investigate oceanic and juvenile crust evolution in the latest Precambrian.

Concluding remarks Ophiolites are critical windows into Earth history to examine the mode and nature of and the interplay between various igneous, metamorphic, sedimentological, hydrothermal and tectonic processes during generation of oceanic lithosphere. They also provide essential information on the mechanics and kinematics of mountain building episodes, as their incorporation into continental margins involved major tectonic events in orogenesis. New data and observations presented in different papers in this book clearly show that there is not a single, unique tectonic environment of ophiolite formation, and that ophiolites are diverse in origin, representing fragments of fossil oceanic lithosphere formed in various tectonic settings and in different stages of Wilson cycle evolution of ancient ocean basins. Most ophiolites are heterogeneous in lithological make-up, internal architecture and alteration history, indicating that their formation involved complex and multiple phases of magmatism, metamorphism and tectonism. Precise radiometric, isotopic and biostratigraphic age dating is needed to better constrain the timing of different evolutionary phases in ophiolite generation.

7

Some ophiolites contain peridotites that may represent exhumed subcontinental lithospheric mantle. It is particularly interesting that this appears to be the case for those ophiolitic assemblages in the Alps and Apennines, where the ophiolite concept was born and first developed through keen observations by influential researchers such as Alexandre Brogniart (1740-1847) and Gustav Steinmann (1856-1929). The existence of these subcontinental lithospheric mantle peridotites suggests that some ophiolites may record the initial stages of rift-drift evolution of small ocean basins in Earth history. Detailed petrological studies of some of the peridotite massifs (i.e. Miintener & Piccardo; Spadea et al.) indicate that pervasive melt migration through these ultramafic rocks resulted in extensive melt-rock reaction, precipitation of plagioclase-enriched peridotites and generation of gabbroic intrusions during the early stages of oceanic lithosphere formation. Late-stage and off-axis(?) magmatism that produced large plagiogranite-trondhjemite intrusions into the pre-existing oceanic crust was responsible for extensive hdyrothermal alteration and mineralization in some ophiolites (Semail, Oman, Stakes & Taylor; Bay of Islands, Newfoundland, Giguere et al.). These intrusive bodies provided the local heat source that set up convective circulation of high-temperature fluids reacting with the host rocks and precipitating in due course epidosites and economic massive sulphide deposits. These spatial and temporal links between late plagiogranite intrusions and alteration-mineralization indicate that magmatism in oceanic crust generation is commonly episodic and multi-stage. Mantle dynamics and heterogeneity at regional and global scales appear to have played a critical role in the evolution of small ocean basins (mostly back-arc and/or marginal basins) and their lithosphere. Collision-induced mantle extrusion and flow strongly affected arc-trench rollback mechanisms, melt flow patterns and thermal state in subduction environments that collectively controlled ophiolite-forming processes (Dilek & Flower; Flower & Dilek). Some ophiolites and related tectonic units (i.e. rift assemblages as precursors to ophiolite generation) display geochemical evidence for mantle source(s), which were contaminated by previous subduction events in the region (e.g. Saccani et al.). These observations and interpretations from ophiolites, coupled with isotopic signatures of oceanic basalts, suggest that the mantle is heterogeneous at all scales mainly as a result of subduction of sediments, hydrothermal alteration of oceanic crust and melting-induced differentiation. Emplacement of ophiolites in collisional orogenic belts involves underplating of less-dense

8

Y. DILEK & P. T. ROBINSON

crustal material beneath displaced oceanic lithosphere in subduction zone environments. The arrival of, and attempted partial subduction of, passive continental margins and/or island arc complexes at oceanic trenches provides the necessary physical conditions for this type of ophiolite emplacement. In accretionary-type orogenic belts (such as in Japan, Far East Asia and late Mesozoic-Cenozoic western North American Cordillera), continued consumption of ocean floor at active continental margins facilitates progressive ophiolite emplacement through tectonic incorporation of stranded slabs of oceanic crust, abyssal peridotites and seamounts into the subduction-accretion complexes. These kinds of ophiolites (defined as 'Cordilleran' by Wakabayashi & Dilek) are commonly spatially associated with blueschist-bearing tectonostratigraphic units and subduction melanges. 'Ophiolite science' is a dynamic, evolving and interdisciplinary enterprise that is at its best

through international collaboration. Future international ophiolite studies, focusing on: (1) careful and systematic documentation of primary (seafloor spreading and/or igneous accretion stage) and secondary (emplacement and post-emplacement) structures within different ophiolitic subunits and of contact relations between them; (2) precise and systematic radiometric and isotopic dating of igneous and metamorphic rocks in ophiolites, and biostratigraphic dating of overlying sedimentary cover and underlying melange units; (3) isotopic analysis of ophiolite peridotites to delineate the mantle composition and signatures of their melt source, and mantle domains; and (4) combined geochemical, petrological and structural studies of ophiolites and associated tectonic units to differentiate tectonic settings of their origin and evolution, will help us better understand the Earth history and the processes involved in its evolution through time.

Ophiolite pulses, mantle plumes and orogeny YILDIRIM DILEK Department of Geology, Miami University, Oxford, OH 45056, USA (e-mail: [email protected]) Abstract: Ophiolites show a wide range of internal structure, pseudostratigraphy and chemical fingerprints suggesting various tectonic settings of their origin. In general, they are characterized as mafic-ultramafic assemblages and associated sedimentary and metamorphic rock units that formed during different stages of the Wilson cycle evolution of ancient oceans, and that were subsequently incorporated into continental margins through collisional and/or accretionary erogenic events. Distributions of ophiolites with certain age groups in different erogenic belts define distinct ophiolite pulses, times of enhanced ophiolite genesis and emplacement, in Earth history. These pulses coincide with the timing of major collisional events during the assembly of supercontinents (i.e. Rodinia, Gondwana and Pangaea), dismantling of these supercontinents, and increased mantle plume activities that formed widespread large igneous provinces (LIPs). Suprasubduction zone ophiolites in orogenic belts signify oceanic crust generation in subduction rollback cycles during the closing stages of basins prior to terminal continental collisions. Both collision-driven assembly of supercontinents and deep penetration of subducted slabs into the lower mantle may produce plumes that in turn facilitate continental rifting, sea-floor spreading and oceanic plateau generation, all of which seem to have contributed to ophiolite genesis. Accelerated LIP formation and seafloor spreading that are associated with superplume events are likely to have caused widespread collisions and tectonic accretion of ophiolites at global scales. Together, these spatial and temporal relations suggest close links between ophiolite pulses, mantle plumes and orogenic events in Earth history.

The traditional definition of ophiolites as on-land fragments of fossil oceanic lithosphere developed at palaeo-spreading centres (Gass 1990) has played an important role on the formulation and advancement of the plate tectonic theory (Coleman 1977, and references therein), and ophiolites have been used extensively to make palinspastic reconstructions of ancient ocean basins and mountain belts (i.e. Dewey et al 1973; Dercourt et al 1986; Lemoine et al. 1986). Exposures of ophiolite complexes along curvilinear fault zones in orogenic belts have been interpreted to represent suture zones, where plate collisions (commonly involving continents and island arcs) occurred in the past (Burke et al 1977). The 1972 Penrose definition of an ophiolite suite having a layer-cake pseudostratigraphy, complete with a sheeted dyke complex, resulting from sea-floor spreading has been central to the ophiolite studies and palaeogeographical reconstructions (Anonymous 1972). Although this ophiolite-oceanic crust analogy and the mid-ocean ridge origin of ophiolites were challenged early on, mainly by geochemists (e.g. Miyashiro 1973), it has been assumed that, in general, ophiolites represent the beginning stages (rift-drift and seafloor spreading) of Wilson cycles.

This traditional view of ophiolites was modified in the mid- to late 1970s, when researchers recovered ophiolitic rocks from the Lau and Mariana back-arc basins, the inner trench walls of the Yap and Mariana trenches, and the Mariana forearc (Hawkins 1977). A new paradigm for ophiolite genesis emerged in the early 1980s, asserting that most ophiolites had developed in suprasubduction zone (SSZ) environments at convergent plate boundaries (Pearce et al. 1984). The widely accepted association of ophiolite genesis with subduction zone settings has shifted the temporal position of ophiolites within Wilson cycles from the beginning to the closing stages. This inference suggests that many ophiolites were produced in the closing stages of ocean basins prior to continental collisions. Dilek & Flower (2003) and Flower & Dilek (2003) have shown, based on actualistic models from the Western Pacific and Eastern Asia and their Tethyan examples, that mantle flow and slab rollback may have played a major role in the formation of SSZ ophiolites in the Alpine-Himalayan orogenic system during the final stages of the evolution of Tethyan basins. Slab rollback is driven by trenchward mantle flow and slab buoyancy forces and

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 9-19. 0305-8719/037$ 15 © The Geological Society of London 2003.

10

Y. DILEK

results in lithospheric-scale extension and associated magmatism in the upper plate that collectively play a major role in the formation of proto-arc, arc and back-arc 'oceanic crust'. The Penrose definition of an ophiolite suite does not prescribe a specific tectonic setting of its genesis and states that 'the use of the term should be independent of its supposed origin' (Anonymous 1972). Extensive research by the international scientific community during the last 30 years has shown that individual ophiolites differ significantly in terms of their structural architecture, chemical fingerprints and evolutionary paths, indicating different tectonic environments of origin, even within the same orogenic belt (Nicolas 1989; Dilek et al 2000). Ophiolites in the Alpine-Himalayan orogenic belt, for example, range from relics of intracontinental rift basins and embryonic normal oceanic crust with mid-ocean ridge basalt (MORE) affinity (Ligurian-type) to protoarc-forearc-back-arc assemblages with SSZ affinities (Mediterranean-type) (Dilek 2003). The peri-Caribbean ophiolites include tectonically emplaced fragments of oceanic crust, which in part represent a large igneous province (LIP), whereas some of the Pacific Rim ophiolites may have had protracted and polygenetic igneous histories that involved the evolution of ensimatic arc terranes through multiple episodes of magmatism, rifting and tectonic accretion, as documented from the Mesozoic ophiolites in the Philippines and the Sierra Nevada foothills (Sierran-type) (Dilek 2003). Ophiolites situated within the accretionary complexes of ancient active margins are commonly associated with melanges and high-pressure metamorphic rocks and may represent fragments of abyssal peridotites and ocean island basalts (OIB), seamounts, island arcs and/or mid-ocean ridge crust scraped off from downgoing plates. These kinds of ophiolites (Franciscan-type) in ancient accretionary complexes do not show genetic and temporal relations (i.e. no melt-residua relationship or chronostratigraphic order) and commonly display diverse chemical affinities and metamorphic grades. Thus, ophiolites are highly diverse in terms of their tectonic origin and emplacement mechanisms (Wakabayashi & Dilek 2003). Despite significant differences in their origin and emplacement mechanisms, ophiolites around the world appear to show distinct patterns of distribution through time and space, suggesting that their evolution may have been linked to some first-order global tectonic events. In this paper I present an overview of the spatial and temporal occurrences of major ophiolite belts in the Earth's history and discuss the possible causes of this 'ophiolite pattern' in a global tectonic framework.

However, examples and the discussion in this paper are constrained to the Neoproterozoic and Phanerozoic occurrences of ophiolites because our knowledge of the Archaean ophiolites, specifically their igneous and emplacement ages and tectonic environment of origin, is still limited.

Distribution of ophiolite belts in space and time Pan-African and Brasiliano ophiolites Figure 1 shows the distribution of major ophiolites with certain age groups in semi-continuous, curvilinear belts around the world. The Late Proterozoic (<860 Ma) ophiolites appear to concentrate mainly in South America, Africa and Arabia, with minor occurrences in central and eastern Europe (Cadomian belts in NW France and Rhodope massif, respectively), the Lesser Caucasus (Transcaucasian massif), central Asia (i.e. Agardagh Tes-Chem, Songshugou and Jiangxi ophiolites) and northwestern India. The widespread existence of Neoproterozoic ophiolites in Afro-Arabia and South America is associated with the evolution of several Pan-African-Brazilide ocean basins (e.g. Mozambique Ocean) in the aftermath of the break-up of the supercontinent Rodinia and during the assembly of West Gondwana (Stern 1994; Dalziel 1997). These Pan-African-Brasiliano ophiolites are fragments of Proterozoic oceanic crust, juvenile island arcs and oceanic plateaux that were amalgamated during the evolution of both collisional- and accretionary-type orogens (Windley 1992). Ophiolites in the ArabianNubian Shield are diverse in origin and include Ligurian-type ophiolites associated with continental rifting and incipient sea-floor spreading (e.g. Nabitah, Hamdah ophiolites), Mediterranean-type SSZ ophiolites developed in forearc-infant arc settings, and Sierran-type, polygenetic ophiolites that evolved in mature island arcs (Dilek & Ahmed 2003). Neoproterozoic ophiolites in AfroArabia and South America thus represent a record of the Wilson cycle opening, narrowing and closing of ocean basins, indicating that modernstyle plate tectonic processes were in operation by 1 Ga. Proterozoic ophiolites in Europe and the Caucasus are isolated fragments of the Pan-African oceanic crust separated from Afro-Arabia as a result of the opening of Palaeozoic and Mesozoic Tethyan basins (Zakariadze et al. 2002).

Eastern Australian ophiolites Cambrian ophiolites in eastern Australia are part of the Tasmanides, which developed during the evolution of East Gondwana in the early Palaeo-

OPHIOLITE PULSES, MANTLE PLUMES AND OROGENY zoic. These ophiolites show diverse chemical affinities (MORE, SSZ and OIB) and have complex origin and emplacement histories (Spaggiari et al. 2003). The age progression, internal structure and chemical fingerprints of the ophiolites in the eastern Tasmanides suggest that they developed in a complex rifted arc-back-arc system, fringing the eastern margin of East Gondwana during 530-485 Ma. The collapse of this fringing arc-back-arc system into the East Gondwana continental margin and the emplacement of the Eastern Australian ophiolites might have been related to far-field stresses associated with the collisional assembly of Greater Gondwana in the early Palaeozoic (Spaggiari et al. 2003).

11

Appalachian, Caledonian, Hercynian and Uralian ophiolites Palaeozoic ophiolites occur mainly in the Appalachian, Caledonian, Hercynian, Uralian and Altaid orogenic belts, extending from eastern North America through Scotland, northern Europe (including the Iberian Peninsula) and Scandinavia, to polar-central Russia and central Asia (Fig. 1). Isolated occurrences of Early Palaeozoic ophiolites also exist in the crystalline basement of the Central Andes in South America (Ramos et al. 2000) and the Trinity terrane of the eastern Klamath Mountains in California (Metcalf et al. 2000). The Early Palaeozoic ophiolites in the Appalachian-Caledo-

Fig. 1. Global distribution of Proterozoic and Phanerozoic ophiolite belts and modern mid-ocean ridge systems (fine double lines) on a North polar projection map (base map from Coleman 1977). Bold black lines represent trenches of modern subduction zones. Ophiolite data sources: Coleman (1977), IGCP (1979), Lippard et al. (1986), Nicolas (1989), Helmstaedt & Scott (1992), Windley (1995), Ramos et al (2000), Spadea & Scarrow (2000), Giunta et al. (2002) and Spaggiari et al. (2003).

12

Y. DILEK

nian erogenic belts, such as the Bay of Islands in Newfoundland, Canada (Dewey & Bird 1971; Irvine & Findlay 1972; Casey et al 1983), Ballantrae ophiolite in Scotland (Oliver & McAlpine 1998), and Karm0y and Solund-Stavfjord ophiolites in western Norway (Pedersen & Furnes 1991) are of SSZ origin, and they commonly show a complete Penrose pseudostratigraphy typical of Mediterranean-type ophiolites (Dilek 2003). The Caledonian-Appalachian erogenic belt developed as the Eastern lapetus Ocean and its seaways between North America and Baltica-Avalonia closed during the Palaeozoic through a series of arc-continent and continent-continent collisions (Dalziel 1997). The Early Palaeozoic ophiolites in the Central Andes have MORB to SSZ affinities (Ramos et al. 2000) and are likely to have developed during the Early Cambrian-Ordovician evolution of the Western lapetus Ocean between Laurentia and South America (Dalziel 1997). The Hercynian ophiolites in Iberia, western and central Europe, and northwestern Africa formed during the closing stages of the Rheic Ocean, which had evolved between the Baltica-Avalonia and Gondwana continental masses (Condie & Sloan 1998). A series of collisions between Baltica and south Europe and between Africa and south Europe resulted in the closure of the Rheic Ocean and its seaways during the Late DevonianCarboniferous, culminating in the Hercynian orogeny. Timing of the formation of the Hercynian orogenic belt coincides with the development of the Alleghany-Ouachita belt in southern North America as a result of the collision of Africa with Laurentia (Dalziel 1997). The Palaeozoic ophiolites in the Uralides and their extension in north-central Asia (Fig. 1) were derived from the Pleionic Ocean and its seaways. The collapse of marginal basins and island-arccontinent collisions caused the emplacement of Uralian ophiolites prior to the terminal closure of the Pleionic Ocean as a result of the collision of Baltica-Eastern Europe with Kazakhstan-Siberia during the Late Permian (Brown et al. 1998; Condie & Sloan 1998). Recent studies have shown that the Uralian ophiolites are diverse in origin, ranging from Ligurian-type, continental rifting and initial sea-floor spreading-related peridotite massifs (Nurali and Mindyak massifs of Spadea et al. 2003) to Mediterranean-type ophiolites of forearc-island-arc origin (Magnitogorsk arc ofSpadea & Scarrow 2000). Thus, the Uralian ophiolites represent different stages of the Wilson cycle evolution of the Pleionic Ocean. The distribution, origin and geodynamics of the Palaeozoic ophiolites in central Asia, specifically those in China and Inner Mongolia, have been discussed by Zhang et al. (2003).

Tethyan - Caribbean ophiolites Jurassic-Cretaceous Tethyan ophiolites occur in the Betic-Rif and Pyrenees (c. 157-145 Ma), Alpine-Apennine mountain systems (Corsica, Sardinia, Eastern Alps, Internal and External Ligurides; c. 200-145 Ma), Carpathians (Apusseni, Bodva; 160-140 Ma), Dinaride-AlbanideHellenide mountain belt (190-150 Ma), Lesser Caucasus (c. 230-220 Ma), Intra-Pontide and Izmir-Ankara-Erzincan suture zones in northern Turkey (Triassic-Late Jurassic), Tauride and Zagros mountain belts (98-90 Ma), Himalaya-Tibet orogenic system (128-70 Ma) and Andaman SeaIndonesian region (c. 140-110 Ma; Fig. 1). These ophiolites developed in Palaeo- and Neo-Tethyan oceans and their seaways that had evolved between Gondwana and Eurasia (Stampfli 2000). Neo-Tethyan ophiolites west of the Aegean Sea are Jurassic in age and contain fertile Iherzolite peridotites that, in part, represent fragments of exhumed subcontinental mantle lithosphere situated in ocean-continent transitions (see Dilek & Flower 2003; Miintener & Piccardo 2003). Typical examples of these ophiolites occur in the Ligurides (hence the designation of Ligurian-type ophiolites; Rampone & Piccardo 2000). Tethyan ophiolites east of the Aegean Sea are Cretaceous in age, progressively younging eastwards in the Himalaya-Tibet orogenic system. The existence of depleted harzburgites, calc-alkaline extrusive rocks and boninites in most of these ophiolites indicates the involvement of subduction-zone processes in their development, although mature island-arc complexes are rare to absent. Wellpreserved Mediterranean ophiolites (i.e. Troodos, Kizildag, Semail, Neyriz) have sheeted dyke complexes and a relatively complete, layer-cake Penrose pseudostratigraphy. Most Neo-Tethyan ophiolites were incorporated into their respective mountain belts by collisions of arc-trench systems with trailing passive continental margins in downgoing plates. Such collisions effectively terminated the subduction rollback cycles in which the ophiolites had developed (Dilek & Flower 2003; Flower & Dilek 2003). The Jurassic-Cretaceous peri-Caribbean ophiolites (Fig. 1) display a complex record of magmatism associated with continental rifting, sea-floor spreading, oceanic plateau accretion and islandarc development. According to Giunta et al. (2002), opening of the Central Atlantic Ocean in the Jurassic initiated continental rifting and seafloor spreading in the Caribbean region, producing the MORB-type proto-Caribbean oceanic lithosphere, fragments of which are found in Costa Rica, Guatemala, Hispaniola and Cuba. Mantle plume magmatism in the Early Cretaceous caused

OPHIOLITE PULSES, MANTLE PLUMES AND OROGENY thickening of this proto-Caribbean oceanic crust in the west and development of an oceanic plateau (Caribbean-Colombian Cretaceous Igneous Province; Kerr et al. 1998, and references therein). Burke (1988) suggested that the Caribbean-Colombian oceanic plateau originally formed in the Pacific and was pushed into the Caribbean basin as a result of the eastward movement of the Farallon plate in the Late Cretaceous-Early Tertiary. Opening of the South Atlantic in the Early Cretaceous caused the northwestward encroachment of South America on the Caribbean basin, resulting in regional compression and development of intra-basinal subduction zones and associated arc magmatism. Oblique convergence and wrench tectonics associated with the motion of South America facilitated the emplacement of ophiolites and oceanic plateau scrapings along the periphery of the basin. Unlike the coeval Tethyan ophiolites, the Caribbean ophiolites thus include fragments of LIP-generated oceanic crust (Caribbean-type ophiolites of Dilek 2003).

Western Pacific and Cordilleran ophiolites The Western Pacific and Cordilleran ophiolites (Fig. 1) range in age from Palaeozoic to Cenozoic and are commonly associated with subductionaccretion complexes. Typical examples occur in the Indonesia-New Guinea region Philippines, Japan, Koryak-Kamchatka orogenic belt, Verkhoyansk-Chukotka fold belt, Alaska and North American Cordillera (Dilek et al. 1990; Saleeby 1992; Encarnacion et al. 1993; Ishiwatari 1994; Harper 2003; Harris 2003; Hirano et al. 2003; Ishiwatari et al. 2003; Sokolov et al. 2003). Ophiolites in the Indonesia-New Guinea region are diverse in age and composition and commonly show SSZ chemical affinities. Some Indonesian ophiolites may include exhumed subcontinental mantle fragments, reminiscent of Ligurian-type ophiolites (i.e. Iherzolite bodies of the Banda arc in Timor). Emplacement of the Indonesian-New Guinea ophiolites was facilitated by the collision of the northern edge of the Australian continent with the subduction-accretion systems of the Pacific Plate during the Cenozoic. Ophiolites in the Philippines constitute the oceanic basement of volcanic arc complexes that underwent magmatic and tectonic extension through multiple phases of back-arc basin opening (Encarnacion et al. 1993), and thus they have a polygenetic evolutionary history typical of Sierran-type ophiolites (Dilek 2003). The ophiolites in Japan, Kamchatka-Koryak and Verkhoyansk-Chukotka commonly occur as imbricated thrust sheets in blueschist-bearing accretionary complexes, with older ophiolitic units generally occupying structurally higher positions

13

(Ishiwatari 1994; Ishiwatari et al. 2003). They show a wide range of geochemical diversity (MORB, OIB, SSZ and LIP) and metamorphic gradients, and they represent tectonic slices of oceanic rocks scraped from downgoing plates at active continental margins. Dilek (2003) classified these types of ophiolites in subduction-accretion complexes as Franciscan-type to distinguish them from Mediterranean-type SSZ ophiolites that are generally underlain by collided passive continental margins. The Jurassic Cordilleran ophiolites in the western USA commonly represent forearc-arcback-arc assemblages that locally display tectonic and geochemical evidence for intra-arc or backarc spreading, episodic and polygenetic magmatism, and mature arc volcanism. Some of these Cordilleran ophiolites might have evolved adjacent to the North American continental margin (e.g. Josephine ophiolite; Harper 2003); some others may represent fragments of an intra-oceanic island arc terrane that were accreted to the North American continental margin during an arc-continent collision (e.g. Smartville arc terrane of Dilek et al. 1990; Guerrero island arc terrane of Dickinson & Lawton 2001).

Ophiolites, plumes and orogeny The Precambrian ophiolite record is still poorly constrained in part because of intense deformation and reworking of continental crust through multiple episodes of orogenic events over time. However, it is also possible that some Precambrian ophiolites may have gone unrecognized because of the search for an ideal, Penrose-type ophiolite sequence showing a layer-cake pseudostratigraphy, complete with a sheeted dyke complex. Some Archaean greenstone belts, for example, may represent dismembered ophiolites (Helmstaedt & Scott 1992). Additionally, it is likely that the Archaean oceanic crust was fundamentally different from its Phanerozoic counterpart in terms of its thickness and internal architecture (Moores 2002). A survey of the reasonably well-documented pre-1 Ga ophiolites shows that their igneous ages appear to cluster at times of 1.0-1.5Ga, 1.8-2.3 Ga, c. 2.5-2.7 Ga and c. 3.4 Ga (Moores 2002). It is difficult to relate these 'ophiolite pulses' in the early Precambrian to any global tectonic events because of our limited knowledge of the early history of the Earth. When we consider the occurrence of Late Proterozoic and Phanerozoic ophiolites, we see discrete pulses of ophiolite generation and emplacement in Earth history (Fig. 2). The most prominent ophiolite pulse was during 180140 Ma when the Tethyan, Caribbean and some of the Circum-Pacific (Western Pacific and North

Fig. 2. Histogram showing the occurrence of major ophiolites and ophiolite pulses through time. Also shown are the life spans of supercontinents and major collisional-orogenic events that led to their assembly, and formation of large igneous provinces (LIPs) and Giant Dyke Swarms. The change in time scale about the Phanerozoic—Proterozoic boundary should be noted. Abbreviations for orogenic events (from youngest to oldest): Ar-Eu, Arabia-Eurasia collision; In-Eu, India-Eurasia collision; Al-Ur, Altaid-Uralian orogenies of central Asia; Ap-Hy, Appalachian-Hercynian orogenies; Cld, Caledonian orogeny; Fmt, Famatinian orogeny; P-Af-Br, Pan-African-Brasiliano orogenies; Grn, Grenville and related orogenies. Period of 'No Magnetic Reversals' between 120 and 80 Ma coincides with the mid-Cretaceous 'superplume' event (Larson 1991). Ophiolite data sources: Coleman (1977), Abbate et al. (1985), Lippard et al. (1986), Nicolas (1989), Ishiwatari (1994), Yakubchuk et al. (1994), Ramos et al. (2000), Giunta et al. (2002) and Spaggiari et al. (2003). Data sources for supercontinental cycles and orogenic events: Stern (1994), Windley (1995), Rodgers (1996), Dalziel (1997), Condie & Sloan (1998) and Moores et al. (2000). Data sources for LIPs and Giant Dyke Swarms: Yale & Carpenter (1998) and Coffin & Eldholm (2001). (See text for discussion.)

OPHIOLITE PULSES, MANTLE PLUMES AND OROGENY American Cordilleran) ophiolites were forming. The second important peak in ophiolite production was during the Late Cretaceous (mostly Tethyan ophiolites), following the mid-Cretaceous 'superplume' event (Fig. 2; Larson 1991). Then, we see second-order ophiolite pulses in the latest Permian-Early Triassic (c. 250-230 Ma), early Devonian-Silurian (c. 400-440 Ma) and Late Cambrian-Early Ordovician (460-500 Ma). The existing record of well-studied Neoproterozoic ophiolites shows ophiolite generation at times of 700 Ma, 740-780 Ma and 820-860 Ma (Fig. 2). A distinct ophiolite pulse around 1 Ga is also apparent. Correlation of these ophiolite pulses with specific global tectonic events reveals apparent patterns and possible links between them. To a first approximation, the observed ophiolite pulses overlap with major orogenic events that led to the assembly of supercontinents, particularly during the Proterozoic (Meso- and Neoproterozoic) and Palaeozoic (Fig. 2). We see this temporal relation during the build-up of Rodinia around 1 Ga, the collision of East and West Gondwana and the construction of Pannotia (c. 700 and 600 Ma), Pan-African-Brasiliano orogenies (520-500 Ma), Caledonian-Famatinian orogenies (460-440 Ma), Appalachian-Hercynian orogenies (c. 300-270 Ma), Altaid-Uralian orogenies in central Asia (c. 240 Ma), and smaller-scale continental collisions within the Alpine-Himalayan system throughout Mesozoic-early Cenozoic time that are not depicted in Figure 2. Most of the ophiolites that originated and were emplaced within narrow time spans during these orogenic events probably represent subduction rollback cycles. As such, they constitute parts of subduction-accretion systems that were accreted onto continental margins prior to terminal closures of ocean basins. Rapid opening of small basins in the upper plates of subduction zones operating along irregular continental margins, reminiscent of the modern Tyrrhenian Sea, can also produce oceanic crust or ophiolites during the late stages of orogenesis. In addition to SSZ ophiolites generated during the late stages of Wilson cycle evolution of ocean basins, some ophiolites are clearly rift-related mafic-ultramafic assemblages (i.e. exhumed subcontinental mantle fragments) and/or fragments of embryonic ocean floor, as documented from the Arabian-Nubian Shield, Uralides, Central and NE Asia, Precordillera of South America and Tethysides. The most prominent ophiolite pulse during the Mesozoic coincides with the break-up of Pangaea (Fig. 2) through discrete episodes of continental rifting during the Late Triassic and Jurassic (Dilek 2001). The Mesozoic Neo-Tethyan Ocean developed, for example, as a result of

15

continental rifting along the northern periphery of Gondwana that began in the Late Triassic. Upper Triassic rift assemblages consisting of basaltic extrusive and intrusive rocks that are intercalated with pelagic to hemipelagic sedimentary sequences are common throughout the Mediterranean region (Dilek & Flower 2003) and may represent the precursor and/or the upper-crustal units of Late Triassic oceanic crust. There are no Late Triassic ophiolites preserved in the region, but it is thought that this inferred Late Triassic oceanic crust within the Neo-Tethyan realm was consumed entirely to produce Jurassic to Cretaceous SSZ ophiolites (Dilek & Flower 2003, and references therein). This interpretation is compatible with the buoyancy analysis of subducting lithosphere that suggests that 10 Ma and older oceanic lithosphere is inherently susceptible to deep subduction (Cloos 1993). Indeed, this is why mid-ocean ridge generated normal oceanic lithosphere has rarely been preserved in the geological record; almost all of it has been consumed during the Phanerozoic (Coleman 1977). Major ophiolite pulses also appear to overlap with the production of plume-related LIPs and giant dyke swarms (Fig. 2). Of particular interest are the LIPs generated as oceanic plateaux, rifted volcanic margins and ocean-basin flood basalts because of their potential contribution to ophiolite-forming processes (Coffin & Eldholm 2001). Although some LIP production has been recognized in the early history of the Earth (Yale & Carpenter 1998), major LIP pulses occur between 250 Ma (Siberian Traps) and 65 Ma (Deccan Plateau), coinciding with the generation of the Tethyan, Caribbean, Japanese and some SW Pacific ophiolites. The enhanced LIP formation and peak ophiolite generation in the Cretaceous are particularly striking. The Cretaceous 'superplume', marked by a long period of no magnetic reversals between 120 and 80 Ma (Fig. 2), is interpreted to be responsible for the development of oceanic plateaux in the Pacific and Indian Oceans, global high sea levels, and increased seafloor spreading rates (Larson 1991). This is also when the majority of the Phanerozoic ophiolites formed in different ocean basins and their seaways. The abrupt ending of LIP formation and ophiolite generation in the Cenozoic is conspicuous (Fig. 2) and further suggests probable spatial and temporal links between plume activity and oceanic crust generation. The temporal relations between supercontinent cycles and LIPs are also revealing and may be inherently linked to ophiolite genesis. LIPs and giant dyke swarms appear to overlap in time with the late stages of supercontinent existence throughout Earth history. The most important case

16

Y. DILEK

is the coupling of Pangaea's existence with extensive LIP formation. Some of the most prominent LIPs, such as the Siberian Traps (c. 250 Ma), Central Atlantic magmatic province (c. 225 Ma), Weddell Sea volcanic margin (184 Ma), Shatsky Rise (147 Ma) and Ontong Java Plateau (12190 Ma; Coffin & Eldholm 2001) formed when the Pangaean supercontinent was still in existence. Some researchers have suggested that construction of a supercontinent insulates the underlying mantle, which initiates major convective activities and elevated fluxes that in turn result in the formation of plumes (Gurnis 1988; Yale & Carpenter 1998). These plumes probably played a major role in the subsequent break-up of the supercontinents (Storey 1995; Dalziel et al 2000). Some ophiolites associated with rifted volcanic margins and oceanic plateaux (i.e. peri-Caribbean ophiolites) are clearly related to LIP production. Moores et al. (2000) suggested spatial and temporal links between LIP formation and regional collisional events. They proposed that major orogenic events driven by slab-pull forces might have supplied cold subducted slabs that penetrated into and depressed the core-mantle boundary, causing large upwellings in the lower mantle (mantle push-ups) at some distance from subduction zones. Their model suggests that these upwelling zones fed mantle plumes that led into continental rifting, sea-floor spreading and LIP formation. Thus, in this model, the production of LIPs is likely to follow times of rapid subduction and major orogenic events. It is also possible that increased LIP production during certain time periods, such as the mid-Cretaceous superplume activity, and increased sea-floor spreading rates induced widespread compression at convergent margins (Vaughan 1995) by carrying more buoyant geological material (i.e. seamounts, thick oceanic plateaux, microcontinents, island arcs) into trenches and thus causing collisions, subduction zone arrests and ophiolite emplacement.

gely coincide with discrete orogenic events leading to the assembly of supercontinents, break-up of supercontinents, and formation of plume-generated LIPs. Spatial and temporal relations between these global tectonic events suggest possible links through complex mantle dynamics. Collision-induced mantle flow results in subduction rollback and one or more episodes of arc splitting and basin opening, producing a collage of 'forearc oceanic lithosphere' (future ophiolites). The collision of these arc-trench systems with continents terminates subduction rollback cycles and facilitates ophiolite emplacement. Assembly of supercontinents may lead to insulation of the mantle that in turn sets up thermal instabilities and plumes. Plume activities weaken the continental lithosphere and cause the break-up of supercontinents, followed by continental rifting, sea-floor spreading and oceanic crust formation. The onset of plumes may also be induced by deep penetration of subducted slabs into the lower mantle that causes mantle push-ups to form; thus, major orogenic events driven by rapid subduction may trigger the formation of plumes and hence the production of LIPs. Global ophiolite emplacement events overlap with superplume activities in certain periods of time (e.g. the mid-Cretaceous). These possible links and the feedback mechanisms between major global tectonic events need to be further investigated by systematic structural, geochemical and geochronological studies of ophiolites, LIPs and ancient continental margins.

Conclusions

References

Ophiolites in orogenic belts occur as curvilinear zones of mafic-ultramafic rock assemblages (with associated metamorphic and sedimentary rocks) within certain age groups, and they represent relics of different stages of the Wilson cycle evolution of ancient ocean basins. There is no single 'blueprint' model of ophiolite formation; ophiolites within even the same orogenic belt may have developed in different tectonic settings before being incorporated into continental margins during the closing stages of ocean basins. Ophiolite pulses in Earth history, times of enhanced ophiolite genesis and emplacement, lar-

ABB ATE, E., BORTOLOTTI, V., PASSERINI, P. & PRINCIPI, G. 1985. The rhythm of Phanerozoic ophiolites. Ofioliti, 10, 109-138. ANONYMOUS 1972. Penrose field conference on ophiolites. Geotimes, 17, 24-25. BROWN, D., JUHLIN, C., ALVAREZ-MARRON, J., PEREZESTAUN, A. & OSLINANSKI, A. 1998. Crustal-scale structure and evolution of an arc-continent collision zone in the southern Urals, Russia. Tectonophysics, 17, 158-171. BURKE, K.C. 1988. Tectonic evolution of the Caribbean. Annual Review of Earth and Planetary Sciences, 16,201-230. BURKE, K.C., DEWEY, J.F. & KIDD, W.S.F. 1977. World

Discussions with J. Encarnacion, M. Flower, H. Furnes, E. Moores, P. Robinson, P. Thy and J. Wakabayashi on various aspects of ophiolite geology were most helpful. I thank K. Hauer, P. Holm and P. Robinson for thorough and constructive reviews of the manuscript. Financial support from the US National Science Foundation (EAR-9796011), National Geographic Society, NATO (CRG-970263 and ESTCLG-97617), and Miami University Hampton Research Funds is gratefully acknowledged.

OPHIOLITE PULSES, MANTLE PLUMES AND OROGENY distribution of sutures: the sites of former oceans. Tectonophysics, 40, 69-99. CASEY, J.F., KARSON, J.A., ELTHON, D., ROSENCRANTZ, E. & TITUS, M. 1983. Deconstruction of the geometry of accretion during formation of the Bay of Islands ophiolite complex. Tectonics, 2, 509-528. CLOOS, M. 1993. Lithospheric buoyancy and collisional orogenesis: subduction of oceanic plateaus, continental margins, island arcs, spreading ridges, and seamounts. Geological Society of America Bulletin, 105, 715-737. COFFIN, M.F. & ELDHOLM, O. 2001. Large igneous provinces: progenitors of some ophiolites? In: ERNST, R.E. & BUCHAN, K.L. (eds) Mantle Plumes: Their Identification Through Time. Geological Society of America, Special Papers, 352, 59-70. COLEMAN, R.G. 1977. Ophiolites. Springer, New York. CONDIE, K.C. & SLOAN, R.E. 1998. Origin and Evolution of the Earth, Principles of Historical Geology. Prentice Hall, Englewood Cliffs, NJ. DALZIEL, I.W.D. 1997. Overview: NeoproterozoicPaleozoic geography and tectonics: review, hypothesis, environmental speculation. Geological Society of America Bulletin, 109, 16-42. DALZIEL, I.W.D., LAWVER, L.A. & MURPHY, J.B. 2000. Plumes, orogenesis, and supercontinent fragmentation. Earth and Planetary Science Letters, 178, 1-11. DERCOURT, J., ZONENSHAIN, L.P. & Ricou, L.E. ET AL. 1986. Geological evolution of the Tethys belt from the Atlantic to the Pamirs since the Lias. Tectonophysics, 123, 241 -315. DEWEY, J.F. & BIRD, J.M. 1971. Origin and emplacement of the ophiolite suite: Appalachian ophiolites in Newfoundland. Journal of Geophysical Research, 76,3179-3206. DEWEY, J.F., PITMAN, W.C. Ill, RYAN, W.B.F. & BONNIN, J. 1973. Plate tectonics and the evolution of the Alpine system. Geological Society of America Bulletin, 84, 3137-3180. DICKINSON, W.R. & LAWTON, T.F. 2001. Carboniferous to Cretaceous assembly and fragmentation of Mexico. Geological Society of America Bulletin, 113, 1142-1160. DILEK, Y. 2001. Causes and effects of Gondwana breakup and implications for the evolution of Neo-Tethys (abstract). 4th International Turkish Geology Symposium, Work in Progress on the Geology of Turkey and Its Surroundings, Cukurova University, Adana, Turkey, 24-28 September, 2001, 87. DILEK, Y. 2003. Ophiolite concept and its evolution. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 1-16. DILEK, Y. & AHMED, Z. 2003. Proterozoic ophiolites of the Arabian Shield and their significance in Precambrian tectonics. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 685-700. DILEK, Y. & FLOWER, M.FJ. 2003. Arc-trench rollback and forearc accretion: 2. A model template for ophiolites in Albania, Cyprus and Oman. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Pub-

17

lications, 218, 43-68. DILEK, Y., THY, P., MOORES, E.M. & GRUNDVIG, S. 1990. Late Paleozoic-early Mesozoic oceanic basement of a Jurassic arc terrane in the northwestern Sierra Nevada, California. In: HARWOOD, D. (ed.) Paleozoic and Early Mesozoic Paleogeographic Relations; Sierra Nevada, Klamath Mountains, and Related Terranes. Geological Society of America, Special Papers, 255, 351-370. DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (EDS) 2000. Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349. ENCARNACION, J.P., MUKASA, S.B. & OBILLE, B.C. 1993. Zircon U-Pb geochronology of the Zambales and Angat ophiolites, Luzon, Philippines—evidence for an Eocene arc-backarc pair. Journal of Geophysical Research, 98, 19991-20004. FLOWER, M.FJ. & DILEK, Y. 2003. Arc-trench rollback and forearc accretion: 1. A collision-induced mantle flow model for Tethyan ophiolites. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218,21-41. GASS, I.G. 1990. Ophiolites and oceanic lithosphe. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues, Proceedings of the Symposium 'TROODOS 1987'. Geological Survey Department, Nicosia, 1-10. GIUNTA, G., BECCALUVA, L., COLTORTI, M., MORTELLARO, D., SIENA, F. & CUTRUPIA, D. 2002. The peri-Caribbean ophiolites: structure, tectono-magmatic significance and geodynamic implications. Caribbean Journal of Earth Sciences, 36, 1 -20. GURNIS, M. 1988. Large-scale mantle convection and the aggregation and dispersal of supercontinents. Nature, 332, 695-699. HARPER, G.D. 2003. Tectonic implications of boninite, arc tholeiite, and MORB magma types in the Josephine Ophiolite, California-Oregon. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 207-230. HARRIS, R. 2003. Geodynamic patterns of ophiolites and marginal basins in the Indonesian and New Guinea regions. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 481-505. HAWKINS, J.W. 1977. Petrological and geochemical characteristics of marginal basin basalts. In: TALWANI, M. & PITMAN, W.C. II (eds) Island Arcs, Deep Sea Trenches and Back Arc Basins. American Geophysical Union, Washington, DC, 1, 355-365. HELMSTAEDT, H.H. & SCOTT, DJ. 1992. The Proterozoic ophiolite problem. In: CONDIE, K.C. (ed.) Proterozoic Crustal Evolution. Developments in Precambrian Geology, 10, 55-95. HIRANO, N., OGAWA, Y., SAITO, K., YOSHIDA, T., SATO, H. & TANIGUCHI, H. 2003. Multi-stage evolution of the Tertiary Mineoka ophiolite, Japan: new geochemical and age constraints. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History.

18

Y. DILEK

Geological Society, London, Special Publications, 218, 279-298. IGCP, 1979. International Atlas of Ophiolites. IGCP Project 'Ophiolites'. Geological Society of America Map and Chart Series, MC-33. IRVINE, T.N. & FINDLAY, T.C. 1972. Alpine-type peridotite with particular reference to the Bay of Islands igneous complex. In: IRVING E. (ed.) The Ancient Oceanic Lithosphere. Canadian Department of Energy, Mines, and Resources, Publication of the Earth Sciences Branch, 42, 97-128. ISHIWATARI, A. 1994. Circum-Pacific Phanerozoic multiple ophiolite belts. In: ISHIWATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Proceedings of the 29th International Geological Congress, Kyoto, Japan, Part D. VSP, Utrecht, The Netherlands, 7-28. ISHIWATARI, A., SOKOLOV, S.D. & VYSOTSKIY, S.V. 2003. Petrological diversity and origin of Ophiolites in Japan and Far East Russia with emphasis on depleted harzburgite. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 597617. KERR, A.C., TARNEY, J., NIVIA, A., MARRINER, G.F. & SAUNDERS, A.D. 1998. The internal structure of oceanic plateaus: inferences from obducted Cretaceous terranes in western Colombia and the Caribbean. Tectonophysics, 292, 173-188. LARSON, R.L. 1991. Latest pulse of Earth. Evidence for a mid-Cretaceous superplume. Geology, 19, 547-550. LEMOINE, M., BAS, T. & ARNAUD-VANNEAU, A. ET AL. 1986. The continental margin of the Mesozoic Tethys in the Western Alps. Marine and Petroleum Geology, 3, 179-199. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Geological Society, London, Memoirs, 11. METCALF, R.V., WALLIN, E.T., WILLSE, K.R. & MULLER, E.R. 2000. Geology and geochemistry of the ophiolitic Trinity terrane, California: evidence of middle Paleozoic depleted supra-subduction zone magmatism in a proto-arc setting. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349,403-418. MIYASHIRO, A. 1973. The Troodos complex was probably formed in an island arc. Earth and Planetary Science Letters, 19, 218-224. MOORES, E.M. 2002. Pre-1 Ga (pre-Rodinian) ophiolites: their tectonic and environmental implications. Geological Society of America Bulletin, 114, 80-95. MOORES, E.M., KELLOGG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency': A resolution of the 'ophiolite conundrum'. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights From Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 3-12. MUNTENER, O. & PICCARDO, G.B. 2003. Melt migration in ophiolites: the message from Alpine-Apennine peridotites and implications for embryonic ocean

basins. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 69-89. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht. OLIVER, GJ.H. & MCALPINE, R.R. 1998. Occurrence of a sheeted dolerite dyke complex in the Ballantrae ophiolite, Scotland. Geological Magazine, 135, 509-517. PEARCE, J.A., LIPPARD, SJ. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PEDERSEN, R. & FURNES, H. 1991. Geology, magmatic affinity and geotectonic environment of some Caledonian ophiolites in Norway. Journal of Geodynamics, 13, 183-203. RAMOS, V.A., ESCAYOLA, M., MUTTI, D.I. & VUJOVICH, G.I. 2000. Proterozoic-early Paleozoic ophiolites of the Andean basement of southern South America. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 331-349. RAMPONE, E. & PICCARDO, G.B. 2000. The ophioliteoceanic lithosphere analogue: new insights from the Northern Apennines (Italy). In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights From Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 21—34. RODGERS, J.J.W. 1996. A history of continents in the past three billion years. Journal of Geology, 104, 91-107. SALEEBY, J.B. 1992. Petrotectonic and paleogeographic settings of U.S. Cordilleran ophiolites. In: BURCHFIEL, B.C., LlPMAN, P.W. & ZOBACK, M.L. (eds)

The Cordilleran Orogen: Conterminous U.S. The Geology of North America. Geological Society of America, Boulder, CO, G-3, 653-682. SOKOLOV, S.D., LUCHITSKAYA, M.V. & SILANTYEV, S.A. ET AL. 2003. Ophiolites in accretionary complexes along the Early Cretaceous margin of NE Asia: age, composition, and geodynamic diversity. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 619-664. SPADEA, P. & SCARROW, J.H. 2000. Early Devonian boninites from the Magnitogorsk arc, southern Urals (Russia): implications for early development of a collisional orogen. In: DILEK, Y., MOORES, E.M., ELTHON, E. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 461-472. SPADEA, P., ZANETTI, A. & VANNUCCI, R. 2003. Mineral chemistry of ultramafic massifs in the Southern Uralides orogenic belt (Russia) and the petrogenesis of the Lower Palaeozoic ophiolites of the Uralian Ocean. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological So-

OPHIOLITE PULSES, MANTLE PLUMES AND OROGENY ciety, London, Special Publications, 218, 567-596. SPAGGIARI, C.V., GRAY, D.R. & FOSTER, D.A. 2003. Tethyan- and Cordilleran-type ophiolites of eastern Australia: implications for the evolution of the Tasmanides. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 517-539. STAMPFLI, G.M. 2000. Tethyan oceans. In: BOZKURT, E., WINCHESTER, J.A. & PIPER, J.D.A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 1-23. STERN, RJ. 1994. Arc assembly and continental collision in the Neoproterozoic East African Orogen: implications for the consolidation of Gondwanaland. Annual Review of Earth and Planetary Sciences, 22, 319-351. STOREY, B.C. 1995. The role of mantle plumes in continental breakup: case histories from Gondwanaland. Nature, 377, 301-308. VAUGHAN, A.P.M. 1995. Circum-Pacific mid-Cretaceous deformation and uplift: a superplume-related event? Geology, 23, 491-494. WAKABAYASHI, J. & DILEK, Y. 2003. What constitutes 'emplacement' of an ophiolite? Mechanisms and relationship to subduction initiation and formation of metamorphic soles. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 427-447. WINDLEY, B.F. 1992. Proterozoic collisional and accretionary orogens. In: CONDIE, K.C. (ed.) Proterozoic

19

Crustal Evolution. Developments in Precambrian Geology, 10, 419-446. WINDLEY, B.F. 1995. The Evolving Continents, 3rd. Wiley, Chichester. YAKUBCHUK, A.S., NIKISHIN, A.M. & ISHIWATARI, A. 1994. A Late Proterozoic ophiolite pulse. In: ISHIWATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Proceedings of the 29th International Geological Congress, Kyoto, Japan, Part D. VSP Utrecht, The Netherlands, 273-286. YALE, L.B. & CARPENTER, SJ. 1998. Large igneous province and giant dike swarms: proxies for supercontinent cyclicity and mantle convection. Earth and Planetary Science Letters, 163, 109-122. ZAKARIADZE, G.S., DILEK, Y., ADAMIA, Sn.A. & OBERHANSLI, R.E. 2002. Geodynamics of the Transcaucasian Massif and the ophiolites of the Lesser Caucasus and its implications for the evolution of Palaeo-Tethys in the NE Mediterranean region (abstract. First International Symposium of the Faculty of Mines, Istanbul Technical University (ITU), Eastern Mediterranean Ophiolites and the Tethyan Geodynamics, 16-18 May 2002. Istanbul Technical University, Siileyman Demirel Cultural Centre, Istanbul, Turkey, 130. ZHANG, Q., WANG, Y., ZHOU, Q., QIAN, G.Q. & ROBINSON, P.T. 2003. Ophiolites in China: their distribution, ages and tectonic settings. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 541-566.

This page intentionally left blank

Arc-trench rollback and forearc accretion: 1. A collision-induced mantle flow model for Tethyan ophiolites M. F. J. FLOWER 1 & Y. DILEK 2 1

Department of Earth and Environmental Sciences, University of Illinois at Chicago (m/c 186), 845 W. Taylor Street, Chicago, IL 60607-7059, USA (e-mail: [email protected]) 2 Department of Geology, Miami University, 114 Shideler Hall, Oxford, OH 45056, USA Abstract: Tectonically active remnants of Neo-Tethys represented by Mediterranean and western Pacific marginal seas are characterized by rapidly propagating backarc extension episodes. These appear to be triggered by random subduction nucleation events, commonly signalled by the appearance of refractory boninites in volcanic 'proto-arcs'. As backarc basins evolve, active arcs separate from their 'proto-arc' remnants and may split again if more than one basin-opening episode occurs. Accreting arc-forearc terranes are therefore likely to incorporate proto-arc, backarc, and (in some cases) inherited continental fragments, as evidenced by their structural complexity and lithological diversity. Forearc complexes typically show positive Bouguer gravity anomalies and significant age discrepancies within and between their crustal and mantle components. Where exposed, their lower stratigraphic horizons may include boninite-bearing assemblages along with tectonized fragments of mid-ocean ridge basalt (MORB) basement and hydrated refractory peridotite. These are typically intruded by sodic plagiogranite (adakite) and high-temperature Mn-, Fe-rich hydrothermal veins ('epidosites'), further indications of subduction nucleation at, or close to a pre-existing spreading axis. Where the arc-trench rollback process is terminated by collision with an approaching continent, or with another retreating forearc complex, MORB-like backarc lithosphere is rapidly reconsumed, in some cases following a change in subduction polarity. In contrast, given their preponderance of ultra-refractory serpentinized peridotite, forearc complexes are relatively buoyant, resist subduction, and are prone to entrapment during early stages of an orogeny. The associated interplay of extension and compression offers a compelling scenario for resolving the so-called ophiolite 'conundrum' and explaining the near-ubiquity of ophiolites in orogenic belts. We propose that rapid arc-trench rollback pulses are driven largely by collision-induced mantle flow in addition to commonly cited 'slab pull' effects. This is supported by the evidence of isotopic mantle flow tracers, seismic tomography, and the coupled kinematics of marginal basins and continental escape. Model applications to some well-known Tethyan ophiolites are developed in a companion paper.

Enigmatic rock assemblages known as ophiolites are characteristic features of most orogenic mountain belts (Dilek et al., 2000). For several decades ophiolitic rock assemblages have been regarded as 'obducted' fragments of the oceanic lithosphere, generated at the global mid-ocean ridge system (Gass 1968, 1989; Moores et al. 1968; Coleman 1977) or above intra-oceanic subduction zones (Miyashiro 1973, 1975, 1977; Pearce et al. 1981, 1984). The implicit paradox of these interpretations has been reconciled to some extent by ascribing ophiolites to subduction-related marginal basins (Miyashiro 1973, 1975, 1977; Pearce et al. 1981, 1984; Beccaluva et al. 1994; Shervais 2001). However, because modern marginal basins per se, along with mid-ocean ridges and supra-subduction volcanic arcs, lack several key features of orogenic ophiolites, the ophiolite conundrum persists (Dilek

et al., 2000: Moores et al. 2000). Are some ophiolites exclusively oceanic in character and others dominantly arc-like, and are refractory boninites and high-Mg andesites (HMAs), absent from 'normal' ocean ridges, volcanic arcs and backarc basins, ubiquitous in ophiolites? Also, why are their high-temperature metamorphic 'soles' invariably mid-ocean ridge basalt (MORB)-like and near equivalent in age, and what causes their inflected metamorphic pressure-temperature-time (P-T-f) histories? Finally, what is the significance of the apparent correspondence observed between ophiolite genesis and distal plate tectonic events, and why is the time lag (less than c. 10 Ma) between the inception of ophiolite formation and their emplacement in continental margins so short? As part of a solution to the ophiolite conundrum, these questions may be largely resolved if

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 21-41. 0305-8719/037$ 15 © The Geological Society of London 2003.

22

M. F. J. FLOWER & Y. DILEK

ophiolites are considered to represent entrapped intra-oceanic forearc assemblages, as proposed by, for example, Dewey & Casey (1979), Casey & Dewey (1984), and Stern & Bloomer (1992). This notion has been resisted on the grounds that any plausible, general explanation for such a linkage is lacking. If, on the other hand, the ophiolite forearc analogue is valid, and possibly unique, questions concerning the genesis of forearcs and how they may be incorporated into 'classic' erogenic assemblages still pose an important problem. There is, in fact, substantial evidence in support of the ophiolite-forearc analogy. Both features are uniquely characterized by the presence of MORB-like and calc-alkaline sequences, boninites, HMA (e.g. Figs 1 and 2), sodic granitoids ('adakite'), and Fe-, Mn-rich hydrothermal deposits, and share structural attributes that record a progression from sea-floor spreading to subduc-

tion-related tectonic processes (e.g. Shervais 2001). Moreover, a preliminary synthesis of geochronological data for Tethyan ophiolites (Flower et aL in prep.) suggests that their inception may be connected to distal plate tectonic effects. It is reasonable, therefore, to suggest that the sum of shared structural, stratigraphic, and petrological features in ophiolites and forearcs is an essential basis for resolving the ophiolite conundrum. The Tethyan tectonic belt is the Earth's most active locus of continental plate collisions and offers a unique opportunity to pursue this goal. Remnant Neo-Tethyan basins in the Mediterranean and western Pacific, for example, show examples of forearc accretion in response to repeated episodes of subduction nucleation and backarc basin opening (Royden 1993a, 1993b; Bloomer et aL 1995; Jolivet & Faccenna 2000; Wortel & Spakman 2000) (Fig. 3). Moreover, the correlation of

Fig. 1. Plots of TiC>2 vs. FeO* (in wt%) for eruptive and intrusive lithologies sampled from typical Izu-BoninMariana intra-oceanic forearcs and the Troodos (Cyprus) and Semail (Oman) ophiolites. (a) Bonin Islands and Sumisu Rift (Izu-Bonin-Mariana system) (Pearce et al. 1992a; Taylor et al. 1994). (b) Mariana Islands (Reagan & Meijer 1984; Stern et al. 1989) and Mariana Trough (Gribble et al. 1998). (c) Troodos ophiolite, Cyprus (Malpas et al. 1984; Flower & Levine 1987; Gibson et al. 1987; Rogers et al. 1989; Taylor et al. 1992; Bednarz & Schmincke 1994; Portnyagin et al. 1996, 1997). (d) Semail ophiolite, Oman (Umino et al. 1990; Lachize et al. 1996; Pezard et al. 2000; Ishikawa et al. 2002).

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES

23

Fig. 2. MORB-normalized incompatible element distributions for eruptive lithologies sampled from typical forearcs and ophiolites. (a) Troodos ophiolite, Cyprus: Upper Pillow Lavas (series 1 and 3), and Lower Pillow Lavas (Flower & Levine 1987; Gibson et al. 1987; Rogers et al. 1989; Bednarz & Schmincke 1994; Portnyagin et al 1996, 1997). (b) Semail ophiolite, Oman (Umino et al 1990; Lachize et al 1996; Pezard et al 2000; Ishikawa et al 2002). Alley Series volcanic rocks (calc-alkaline and boninitic), Geotimes Series volcanic rocks (MORB-type volcanic rocks), (c) Izu-Bonin arc-forearc, and backarc Sumisu Rift (Izu-Bonin-Mariana system) (Pearce et al 1992a; Taylor et al 1994). (d) Mariana arc-forearc (Reagan & Meijer 1984; Stern et al 1989) and backarc Mariana Trough (Gribble et al 1998). such processes with distal plate kinematic changes suggests a fundamental connection between ophiolite genesis and global-scale plate tectonics (Flower et al. 2001; Flower 2003). The causes of arctrench rollback may therefore be crucial to our understanding of ophiolites and their geodynamic significance in Earth history. Here, following the 'Tectonic Facies' approach of Hsu (1994) and Hsu (1997), we present an 'actualistic' model for ophiolites based on processes of forearc evolution in western Pacific and Mediterranean marginal basins. In a companion paper (Dilek & Flower this volume) the model is adopted as a template for interpreting three well-studied Tethyan ophiolites.

A brief history of Tethys Tethyan orogens mark a succession of continental plate collisions that followed breakup of the Gondwana continent and repeated cycles of ocean basin opening and closure. Although the relevant plate kinematic reconstructions are controversial,

there is a general agreement that the northward drift of Gondwana fragments involved three or more such cycles of opening (e.g. Dercourt et al. 1986; Audley-Charles & Harris 1990; Ustaomer & Robertson 1993; Metcalfe 1996; Stampfli & Borel 2002). These cycles were commenced with diachronous 'unzipping' of the northern margin of Gondwana, and produced Tethyan basins evolving as triangular inlets that propagated westward from the proto-Pacific Ocean. According to the majority of views, Palaeo-Tethys was initiated in the Late Devonian with the separation of continental blocks that later amalgamated as the North China, South China, Iran, Kazakhstan, Indochina, Qaidam, Tarim, and Hainan blocks (Audley-Charles & Harris 1990; Metcalfe et al, 1999). Meso-Tethys probably began to open in the Early Permian with detachment of the Cimmerian and other microcontinents, and Neo-Tethys began opening between the Late Triassic and Late Jurassic with the separation of what later became the Pelagonia, Tauride-Anatolide, Lhasa, West Burma, and Woyla blocks (Dilek et al 1999; Metcalfe et a/.,1999).

24

M. F. J. FLOWER & Y. DILEK

Fig. 3. Plate boundary evolution and arc—trench rollback in active Tethyan domains, (a) The circum-Mediterranean region (after Wortel & Spakman 2000). Arrows indicate directions of probable slab tearing beneath the ApennineCalabrian, Hellenic, and Carpathian arcs. Adr, Adriatic Sea; Aeg, Aegean Sea; Alb, Alboran Sea; Ap, Apennines; AP, Algero-Provencal Basin; Cal, Calabria; Car, Carpathians; Co, Corsica; Cr, Crete; Cyp, Cyprus; Hel, Hellenic arctrench; Ion, Ionian Sea; Lev, Levantine Basin; Mag, Maghrebides (from the Rif to Sicily); NAF, North Anatolian Fault; Sa, Sardinia; Si, Sicily; Tyr, Tyrrhenian Sea. Barbs indicate subduction or thrusting vergence, black suggesting a continuous slab, and white, possible post-collision breaker!, (b) Western Pacific region (after Flower et al. 1998, 2001). Arrows indicate directions of probable slab tearing beneath the Himalayas, Sunda-Banda arcs, and Northern Luzon-Taiwan. TS, Tien Shan; ATF, Altyn Tagh Fault; And-Nic, Andaman-Nicobar Islands; 1C, Indochina; Ma, Malay Peninsula; Sm, Sumatra; Bo, Borneo; Ja, Java; Su, Sunda; Ba, Banda; Sw, Sulawesi; Phil, Philippines; Tw, Taiwan; Izu-Bon, Izu-Bonin Islands; Ma, Mariana Islands; Wma, West Mariana arc; PKR, Palau-Kyushu ridge; Ru, Ryukyu Islands; SCS, South China Sea; SS, Sulu Sea; CS, Celebes Sea; MS, Molucca Sea; WPSB, West Philippine Sea Basin; BS, Banda Sea; SB, Shikoku Basin; SR, Sumisu Rift; PVB, Parece Vela Basin.

Palaeo-Tethys The closure of Palaeo-Tethys and the corresponding inception of Meso-Tethys were marked by the accretion of Kunlun, Qaidam and Ala Shan Ter-

ranes to Kazakhstan-Siberia in the Early Permian, followed in the Late Permian to Early Triassic by suturing of Sibumasu and Qiangtang to Cathaysialand as Palaeo-Tethys was finally consumed by subduction (Metcalfe et al., 1999). Meso-Tethys

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES closure between the Late Triassic and Late Jurassic was accompanied in the east by diachronous accretion of the Lhasa, West Burma, and Woyla micro-continents (Metcalfe et al., 1999), and in the west, Cimmeria, Iran, Pelagonia, and others (Dercourt et al. 1986; Ustaomer & Robertson 1993; Stampfli & Borel 2002) to Eurasia. Finally, as the African, Arabian, and Indian plates collided with Eurasia, and Australia collided with newly accreted Sunda, remnants of the Neo- and MesoTethyan lithosphere were being progressively consumed by subduction (Dercourt et al. 1986; Audley-Charles & Harris 1990; Metcalfe et al., 1999; Stampfli & Borel 2002). Although successive Tethyan basins were more or less separated by micro-continents throughout much of the Triassic and Jurassic, they may have remained connected at their western extremities, between the Mediterranean and Caucasus. This interpretation is supported by evidence suggesting that Mid-Jurassic remnant basins were being consumed by subduction as collisions between retreating arc-forearc complexes and continents prevented further extension (Stampfli & Borel 2002). For example, Paleocene closure of the Liguria-Piedmont basin was coeval with the Betic-Rif, western-northern Alpine, and Carpathian orogenies and, as younger basins collapsed, it was followed by the NeogenePleistocene Apennine, Maghrebe, Dinaride, and Hellenide orogenies (Faccenna et al. 1997; Jolivet et al. 1999; Jolivet & Faccenna 2000; Stampfli & Borel 2002).

Neo-Tethys The closure of Neo-Tethys coincided with opening of the North Atlantic Ocean that began at c. 180 Ma and was followed by the separation of East and West Gondwana at c. 158 Ma. By c. 130 Ma East Gondwana (Africa-India-SeychellesMadagascar-Australia-Antarctica- South America) had also begun to sunder as opening of the South Atlantic commenced at c. 110-100 Ma and the North Atlantic opening continued. By the Mid-Cretaceous, a block comprising Africa and India-Seychelles-Madagascar began to detach from Australia-Antarctica, followed shortly by the separation of Australia and initiation of seafloor spreading at the Southeast Indian Ridge (Metcalfe 1996). At c. 98 Ma, the India-Seychelles block separated from Madagascar and by the Late Cretaceous, along with Africa-Arabia and Australia, was moving rapidly northwards towards accreting Eurasia. Finally, following separation from the Seychelles at c. 65 Ma (Gnos et al. 1997), India collided with Eurasia between c. 50 and 45 Ma (Lee & Lawver 1994). After

25

separating at c. 40 Ma, Arabia and Africa collided with accreting Eurasia at c. 30 Ma and 25 Ma, respectively (Dewey et al. 1989; Jolivet & Faccenna 2000). The record of Gondwana disaggregation and (partial) reassembly offers a potential rationale for the timing and location of 'spontaneous' subduction nucleation and, in turn, the processes giving rise to ophiolites. However, the causes of subduction nucleation remain unclear. Are such events determined by global-scale plate kinematics (as suggested by Gnos et al. 1997) or do they reflect viscous mantle instabilities caused by density and thermal heterogeneities (e.g. Toth & Gurnis 1998; Faccenna et al. 1997)? These questions bear, in turn, on whether ophiolites represent a global-scale plate tectonic phenomenon (e.g. determining where they are initiated) or local phenomena related to an imminent plate collision (determining both where they are initiated and where they are emplaced). Although this latter question is beyond the scope of the present paper, we will attempt to provide a basis for its future consideration.

Towards an actualistic model Today, Tethyan tectonic and magmatic activity is dominated by effects of the African, Arabian, Indian, and Australian collisions concomitant with continued basin opening in parts of the Mediterranean Sea and western Pacific (e.g. Fig. 3). On a global scale, subduction zones may remain static for lengthy periods and evolve as simple linear orogens. Others, notably in the regions discussed here, are observed to migrate oceanward at rates exceeding 100m ma"1, often developing into spectacular oroclines (Fig. 3). Such rapid 'arctrench rollback' processes are expected to continue indefinitely unless they are terminated by collisions of their retreating arc-forearc complexes with mid-ocean ridges, continental plates, or other migrating subduction systems.

Forearc complexes as lithological 'high-tide marks' In the Mediterranean, rollback cycles have mostly been interrupted by forearc collisions and the ensuing consumption (or 'collapse') by subduction of short-lived backarc basins (Dalziel 1989; Clift & Dixon 1998; Jolivet & Faccenna 2000; Robertson 2000). In contrast, arc-trench rollback in the western Pacific has been relatively unconstrained with continuing eastward propagation of forearcs, free from the 'jaws' of an impending plate collision (Karig 1971; Hussong et al. 1981; Karig et al. 1986; Jolivet et al. 1991b; Tamaki & Honza

26

M. F. J. FLOWER & Y. DILEK

1991). On the other hand, these regions show strong similarities, arc-trench rollback cycles in both cases being initiated by splitting of nascent volcanic 'proto-arcs' into active and remnant segments, the associated refractory magmas indicating unusual thermal conditions (Pe-Piper & Piper 1989, 1994; Stern & Bloomer 1992; Hawkins & Castillo 1998; Insergueix-Filippi et al. 1998, 2000). The net effect of this type of process, sometimes compounded by additional arc splitting events (Hussong et al. 1981; Ishii et al 1995; Fassoulas 1999), is to produce an evolving forearc terrane that potentially includes the igneous and metamorphic products of successive 'proto-arc', arc, and backarc episodes (Bloomer 1983; HickeyVargas 1989; Johnson et al. 1991, 1992; Giaramita et al. 1992; Ishii et al. 1992, 1995; Marlow et al. 1992) along with characteristically high-temperature hydrothermal deposits (Banerjee et al. 2000; Fryer et al. 2000; Gillis & Banerjee 2000; Banerjee & Gillis 2001; Gillis 2002). The presence of allochthonous continental and oceanic lithosphere fragments may represent crustal features prior to the inception of rollback (Johnson et al. 1991; Giaramita et al. 1992; Parkinson et al. 1998; Parkinson & Arculus 1999). As they evolve, therefore, arc-forearc complexes progressively resemble lithological 'high tide marks' (HTMs), increasingly heterogeneous, accreted assemblages of proto-arc, arc, and backarc crust exhibiting significant internal age and structural discrepancies (Flower et al. 1998, 2001; Flower 2003). Subduction nucleation At least two lines of evidence highlight the anomalous thermal character of asthenospheric mantle associated with subduction nucleation events that appear to trigger rollback cycles: the presence of boninite and high-magnesium andesites (HMA) in proto-arcs (Casey & Dewey 1984; Stern et al. 1989; Stern & Bloomer 1992; Clift & Dixon 1998; Wallin & Metcalf 1998) and the inflected P-T-t histories of sub-ophiolitic metamorphic 'soles' interpreted from thermobarometric studies (Wakabayashi & Dilek 1988, 2000; Gjata et al. 1992; Insergueix-Filippi et al. 1998, 2000; Searle & Cox 1999; Bebien et al 2000; Dimo-Lahitte et al. 2001). Boninite-HMA volcanism is relatively rare in modern settings. A notable exception, however, is the Hunter Ridge, between the southernmost New Hebrides and Fiji islands, where boninite magmatism marks a locus of incipient subduction along an active transform fracture zone (Falloon & Crawford 1991; Danyushevsky et al., 1995; Crawford et al. 1997). Here, the eastern extremity of the New Hebrides subduction system appears to

be nucleating along a transform fault that is linked to the southward-propagating North Fiji spreading centre (Monzier et al. 1993a, 1993b, 1997). Examples of coeval, if now-extinct, boninite and HMA volcanism also characterize forearc-remnant arc pairs in the Mediterranean and western Pacific regions, confirming the unusual thermal character of subduction nucleation events. The best-documented example of subduction nucleation occurs in the Izu-Bonin-Mariana (IBM) 'subduction factory' (MARGINS) where Mid-Eocene subduction nucleation was followed by rapid backarc basin opening and rollback of the IBM arc-forearc terrane (Karig 1971; Hussong et al. 1981; Stern & Bloomer 1992; Bloomer et al. 1995; Hawkins & Castillo 1998). The locus of subduction inception is marked by the Palau-Kyushu ridge, a boninite-bearing 'proto-arc' remnant that dissects the Philippine Sea Plate. Subduction probably began at c. 50 Ma, shortly before the India-Asia collision (c. 45-40 Ma) and reorientation of Pacific Plate motion (c. 43 Ma) (Bloomer et al. 1995; Hawkins & Castillo 1998) (Fig. 4) with underthrusting along a transform fracture zone of the West Philippine Sea Basin Plate, either by the Pacific Plate or a younger, hypothetical, North New Guinea Plate (Stern & Bloomer 1992). As subduction continued, splitting of the Palau-Kyushu proto-arc led to opening of the Parece Vela Basin (c. 40-25 Ma), with concomitant rollback of the newly active West Mariana arc. Splitting of the latter gave way to further rollback of the active Mariana arc with opening of the still-active Mariana Trough (Karig 1971; Hussong et al. 1981). Mariana Trough opening continues today accompanied by northward 'unzipping' of the West Mariana-Mariana arc, offering a diachronous, actualistic model for protoophiolite genesis. According to such a model, the present-day Mariana forearc includes the accumulated products of West Philippine Sea sea-floor spreading, 'proto-arc' boninitic and calc-alkaline activity, and subsequent (West Mariana, Mariana) arc volcanism. Boninites dredged from the Palau-Kyushu ridge match those in lower horizons of the Mariana forearc, the latter feature conforming in these and other respects to an in situ 'protoophiolite' (Ishii et al. 1988; Ogawa & Taniguchi 1989). Mid-Miocene HMA volcanic rocks (c. 13 Ma) are likewise preserved in central and southern parts of the Ryukyu forearc and are matched by analogous activity in the FujianTaiwan region (Shinjo 1999), suggesting subduction nucleation, triggered by collision of the Luzon arc with Eurasia, prior to opening of the Okinawa Trough (Shinjo 1999). Other examples include Mid-Miocene (c. 14-15 Ma) HMA vol-

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES

27

Fig. 4. Arc-trench rollback model, from Bloomer et al. (1995), based on evolution of the Izu-Bonin-Mariana forearc and eastern part of the Philippine Sea Plate (following Karig 1971; Hussong et al. 1981; Stern & Bloomer 1992). (a) 50-40 Ma. Subduction nucleation beneath the West Philippine Sea Basin Plate either by the Pacific Plate or (hypothetical) young North New Guinea Plate along a transform fracture zone in proto- West Philippine Sea Basin spreading centre. Boninite melt genesis accompanies early forearc development with inception of the calc-alkaline arc forming the Palau-Kyushu ridge, (b) 40-25 Ma. Continued subduction with slab steepening of the Pacific Plate, splitting of the Palau-Kyushu arc, Parece Vela Basin opening, and rollback of the active West Mariana arc. (c) 250 Ma. Continued subduction, Mariana Trough opening by splitting of the West Mariana arc, and rollback of the active Mariana arc. (d) 0 Ma. Subduction beneath the modern Mariana arc-trench system with continued Mariana Trough opening by 'unzipping' of the West Mariana-Mariana arc to the north (Iwo Jima). PKR, Palau-Kyushu ridge; PVB, Parece Vela Basin; WMA, West Mariana arc; MT, Mariana Trough; MA, Mariana arc; IBM, Izu-Bonin-Mariana forearc; PAC, Pacific Plate; WPSB, West Philippine Sea Basin; WPAC, Western Pacific; NNG, 'North New Guinea' Plate.

canism recorded from islands between the Hellenic forearc and Cycladean metamorphic core complexes (Smith & Spray 1984; Pe-Piper 1994; Forster & Lister 1999; Migiros et al. 2000) that corresponds to HMA-bearing ophiolite fragments in the Hellenic forearc (Fortuin et al. 1997; Clift 1998), and Oligo-Miocene HMA (c. 18 Ma) in Sardinia and Corsica, and in Calabrian forearc ophiolite fragments (Beccaluva 1982; Delaloye et al. 1984; Compagnoni et al. 1989; Beccaluva et al. 1994), which pre-date opening of the Tyrrhenian Sea (Morra et al. 1991 \ Padoa 1999). Experimental studies of boninite melts suggest that they result from the combined effects of mantle decompression and slab-derived H2O-rich fluid (Umino & Kushiro 1989; van der Laan

et al. 1989; Falloon & Danyushevsky 2000), interpreted by some as an indication of ocean ridge spreading conditions (Gjata et al. 1992; Stern & Bloomer 1992; Peacock 1994; Peacock et al. 1995). However, comparison of experimental data for basalts, boninites, and variably fertile peridotites suggests that three additional conditions are required for boninites to form: (1) anomalous asthenospheric potential temperatures (Tp > c. 1400°C); (2) significant lithospheric extension (stretching factors, /?, > c. 3) (McKenzie & Sickle 1988; Latin & White 1990); (3) refractory (previously melt-depleted) peridotite sources (e.g. van der Laan et al. 1989; Hirose & Kawamoto 1995; Hirose 1997; Falloon & Danyushevsky 2000) (Fig. 5). In other words, boni-

28

M. F. J. FLOWER & Y. DILEK

Fig. 5. Partial melting models for (a) fertile and (b) refractory peridotite in the presence of H2O and CO2 between pressures of 0 and 5 GPa, interpolated from published experimental data. Geotherms (bold dashed curves) are calculated for asthenospheric potential temperatures (7J,) of 1280 °C and 1440 °C, respectively, for stretching factors (/?) of c. >3 and >5, assuming 'pure shear' lithosphere extension (McKenzie & Sickle 1988; Latin & White 1990). The solidus curve (bold line) for 'fertile' peridotite (pyrolite)- C-H-O is taken from Wyllie (1990) for the condition X = CO2/(CO2 + H2O) = 0.8 along the bold grey curve marking equilibrium between amphibole, carbonate, peridotite, and vapor; based on sources given by Wyllie et al., (1990). The anhydrous solidus for fertile peridotite (HK-66) is from Hirose & Kushiro (1993), with a hypothetical H2O-undersaturated (1 wt% H2O) solidus interpolated from hydrous experiments on refractory Iherzolite (KLB-1) (Hirose 1997). Anhydrous, f^O-saturated, and H2Oundersaturated (1 wt% H2O) solidi for the refractory Iherzolite (KLB-1) are from Hirose & Kawamoto (1995). Predicted melt segregation conditions agree with those determined experimentally for primitive MORB and high-Ca boninite (HCB) (e.g. van der Laan et al. 1989; Falloon & Danyushevsky 2000). MBL, mechanical boundary layer; BON, boninite; LVZ, low-velocity zone; PIC, picrite; NE, nephelenite; BAS, basanite.

nite liquidus temperatures appear to exceed those of 'normal' mid-ocean ridge magmas by c. 150200 °C and require a source that is significantly less fertile than that producing normal MORE (van der Laan et al. 1989; Falloon & Danyushevsky 2000).

These observations concur with predictions from 2D numerical models (Gjata et al. 1992; Insergueix-Filippi et al. 1998, 2000) and are also supported by anomalous P-T-t histories recorded from sub-ophiolitic metamorphic soles (Fig. 6). These appear to record patterns of cold thrusting

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES

29

Dilek et al 1997; Hacker & Gnos 1997; Searle & Cox 1999; Dimo-Lahitte et al 2001), suggesting that metamorphic 'soles' may be best interpreted as MORB-like relics of incipient subduction (e.g. Fig. 6; Wababayashi & Dilek, this volume). Although some ophiolites slightly post-date their high-temperature soles, most are coeval or slightly older, consistent with interpretations that sundered forearc and remnant 'proto-arc' components were single entities prior to splitting and the onset of rollback. Incipient subduction relics are thus preserved a priori in forearc 'proto-ophiolites' (Ishii 1989; Ogawa 1995).

Arc-trench rollback: endogenous vs. exogenous causes? Plate kinematic

Fig. 6. Interpolated P-T-t histories for sub-ophiolitic metamorphic soles and other metamorphic lithologies. (a) Mirdita ophiolite, Albania (Gjata et al. 1992; Robertson & Shallo 2000; Dimo-Lahitte et al 2001). (b) Troodos-Mamonia complexes, Cyprus (Malpas et al. 1992; Bailey et al. 2000). (c) Semail ophiolite, Oman. 1 and 2, Saih Hatat; 3, Saih Hatat-Wadi Tayin; 4, Asimah-Bani Hamid; 5, Hacker & Gnos (1997); 6, As Sitah eclogites -> Hulw blueschist (Searle et al. 1994; Searle & Cox 1999). Pressure-temperature equilibration conditions for ophiolitic blueschist-bearing units from the Shuksan, and Franciscan terranes and Cyclades (Aegean) remnant arc are also shown. Clockwise (CW) and counter-clockwise (CCW) P-T-t paths appear to distinguish 'high temperature' from 'high pressure' sole types.

to at least c. 60 km depth with temperatures rising to well over 700 °C, a stage of near-isobaric cooling, and (eventual) exhumation (Wakabayashi & Dilek 1988, 2000; Hacker 1991; Gjata et al 1992; Beccaluva et al 1994; Searle et al 1994; Wakabayashi & Unruh 1995; Dilek & Whitney 1997;

effects

Although reliable age data for ophiolites are sometimes difficult to acquire and have proved contradictory, a clearer spatial-temporal picture is emerging of relations between active Tethyan forearc complexes and their respective remnant proto-arcs. For example, boninite-bearing 'protoarcs' (c. 49-45 Ma), preserved in the IBM forearc and its remnants (Hawkins & Castillo 1998) and the Zambales (Philippines) ophiolite (Encarnacion 1997), are nearly coeval with the initiation of Celebes Sea opening (c. 48 Ma) (Beiersdorf et al. 1997), whereas they slightly predate the 'hard' collision of India with Eurasia (c. 45-40 Ma) and the corresponding change in Pacific Plate motion from NNW to NW. The Palawan and Mindoro ophiolites (c. 34 Ma) also appear to mark a change from sea-floor spreading to convergence, coeval with the inception of Alao Shan-Red River leftlateral shearing (c. 33 Ma) and sea-floor spreading in the South China Sea (c. 32 Ma). Both were terminated by a collision between the Sulu Ridge arc remnant and the West Philippine arc (c. 17 Ma) (Rangin et al. 1995), which triggered the initiation of Sulu Sea opening (c. 17 Ma) (Rangin et al. 1995; Yumul et al. 1998, 2001). New subduction that produced the Ryukyu proto-arc probably occurred between c. 21 and 18 Ma, preceding collisions of Taiwan (c. 15-12 Ma) (Chung et al. 1994), and other micro-continents with the Luzon arc, at c. 17-16 Ma (Rangin et al. 1985; Pubellier & Cobbold 1996; Pubellier et al. 1996) and 12-6 Ma (Sibuet & Hsu 1997). In western Tethys, the Late Cretaceous Semail and Troodos ophiolites (c. 98-75 Ma) (Urquhart & Banner 1994; Hacker & Mosenfelder 1996; Hacker et al. 1996) and those exposed in the Zagros and Tauride belts (c. 98-75 Ma) (Dilek et al. 1999; Parlak & Delaloye 1999; Ghazi & Hassanipak 2000; Parlak et al. 2000; Babaie et al.

30

M. F. J. FLOWER & Y. DILEK

2001a, 2001b; Ghasemi et al. 2002) and Hellenic arc (c. 98-75 Ma) (Langosch et al. 1999, 2000), were formed shortly before collisions of the Iranian (c. 75-70 Ma), Apulian (c. 70-65 Ma), Pelagonian (c. 72 Ma), and other micro-continents (Robertson & Shallo 2000; Stampfli et al. 2001; Stampfli & Borel 2002) with accreting Eurasia. However, younger ophiolitic remnants (c. 12— 10 Ma) in Crete and the Aegean Cyclades are near-coeval with both Aegean continental collapse (Lee et al. 1990) and the initiation of westward escape by Anatolia (c. 13-5 Ma) (Le Pichon 1982; Le Pichon et al. 1995). Thus, although rollback of Hellenic subduction may have continued since the Paleocene, as inferred from seismic tomography (Spakman et al. 1992; Spakman & Bijwaard 1998), it was probably interrupted by the effects of regional microplate collisions in the Late Cretaceous and Pliocene (Le Pichon et al. 1995).

'Slab pull' and extrusion tectonics Encarnacion et al. (2001) proposed a linkage between genesis of the South Palawan (Philippines) ophiolite and the coupled inception of South China Sea spreading and left-lateral motion on the Red River fault (Lee et al. 2000; Wang et al. 2000). According to the classic extrusion tectonics model (Tapponnier et al. 1982, 1986), marginal basin opening and arc-trench rollback are linked responses to collision-induced lithosphere 'escape', as interpreted for opening of the Aegean and South China Sea basins and seaward escape, respectively, of Anatolia and Indochina (Briais et al. 1993; Le Pichon et al. 1995; Lundgren et al. 1996). However, as a general explanation of marginal basin opening, extrusion tectonics seems unable, at least in these cases, to account for the observation that basin opening commenced prior to, and proceeded at a faster rate than, the escape of their respective conjugate blocks (Chung et al. 1997; Lee et al. 2000; Wang et al. 2000; Le Pichon et al. 2002). Given the apparent linkage of continental escape and marginal basin opening, we need to consider the extent to which backarc basin opening is intrinsic to subduction and what, if any, exogenous factors play a role. As already noted, arc-trench rollback has usually been ascribed to slab buoyancy forces, assuming these to exceed those of the convecting asthenosphere (Isacks & Molnar 1971; Uyeda & Kanamori 1979). Most studies of the mechanical interactions between subducting and overriding plates suggest that where backarc basin opening is passive, the seismicity associated with subduction shows 'down-dip compression' (e.g. Fig. 7a). If horizon-

Fig. 7. Hypothetical effects of 'slab pull' and 'mantle extrusion', (a) Slab force model, showing interaction between subducting slab and overriding plate (from Seno & Yamanaka 1998). Where backarc basin opening is a passive response to slab compression, slabs show 'downdip compression'. However, down-dip tension accompanies backarc opening in some cases (e.g. the Mariana, Kyushu and Hellenic arcs) (Seno & Yamanaka 1998), suggesting slab retreat is driven by trenchward mantle flow. The dip of the subducting plate is q, AB is the trench axis and CE the aseismic front. PS' is the effective ridge push, FS the slab pull, and PC the collision force. FS' is the horizontal component of the traction on CD', t is the shear stress at the thrust zone (see Seno & Yamanaka 1998). Where rollback is slow, FS' is negative, in which case Fc could be positive or negative depending on whether Ps' is smaller or larger than FS'. If FS' is negative and opposed to FP', FC is likely to be very small, resulting in back-arc extension, (b) Two-dimensional mantle flow model (after van Keken et al. 2002); the slab is assumed to be subducting at constant speed. Two flow components are shown: (1) 'endogenous' (slab-induced) flow; (2) exogeneous (e.g. collision-induced) flow.

tal traction is negative, backarc extension will be relatively modest (Seno & Yamanaka 1998). In some cases (e.g. the Mariana, Kyushu and Hellenic arcs), however, down-dip tension accompanies backarc opening, suggesting that slab retreat is more rapid, and driven by trenchward mantle flow (Seno & Yamanaka 1998). These relationships are shown in Figure 7b, indicating the possibility of two potential flow field components: 'endogenous' (slab-induced) and 'exogeneous' (mantle-driven).

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES

Collision-induced mantle extrusion Thus although slab pull can explain the dynamics of many subduction zones, it is probably unable to account for those cases, such as the Hellenic and Mariana arc systems (Seno & Yamanaka 1998), where slab steepening, arc bending, and accelerated basin opening coincide (McCabe & Uyeda 1983; Hynes & Mott 1985; Dvorkin et al 1993; Bevis et al. 1995). Slab pull is probably subsidiary to exogenous mantle flow (Seno & Yamanaka 1998), a mechanism that can potentially reconcile marginal basin and continental escape kinematics with the accretionary build-up of forearc complexes. During early stages of typical Wilson Cycles, plate motions are probably driven by a combination of mantle upwelling ('ridge push'), downwelling ('slab pull'), and the effects of lateral impingement (on continental cratonic keels) (Forsyth & Uyeda 1975; Russo & Silver 1996). For example, the correspondence of arc-trench rollback in the Caribbean and South Scotia Sea regions (Dalziel et al. 2001) to accelerated westward motion of South America (Russo et al. 1993; Russo & Silver 1996) may be a far-field mantle flow response to the 30-25 Ma Africa-Eurasia collision (Silver & Russo 1996). In response to the latter, westward migration of the Mid-Atlantic Ridge and the corresponding eastward offset of major Mid-Atlantic hotspots (Iceland, St. Helena, Tristan da Cunha, the Azores, and Bouvet) would have been immediately translated to South American plate motion. At later stages, asthenosphere flow is likely to be displaced by thick continental plates as they approach each other and, eventually, collide. For example, Tamaki (1995) suggested that lateral displacement of asthenosphere prior to and following the India-Asia collision led to rapid eastward propagation of Western Pacific marginal basins. Such a process, broadly consistent with the timing and kinematics of basin opening (Hall 2002), also offers a plausible explanation for widespread intra-plate volcanism that characterizes much of east and SE Asia, contamination of the upper mantle beneath eastern Eurasia and western Pacific basins (attributed to delamination of the Sino-Korean craton), and the sharp boundary separating DUPAL-like (contaminated) from N-MORB Pacific mantle. Accordingly, if HTM ('high-tide mark') forearc assemblages (the accreted igneous and metamorphic products of arc-trench rollback) are a valid analogue for ophiolite, such features can be taken to represent distal mantle flow boundaries (Flower et al. 1998, 2001; Flower 2003). This is not to say that arc-trench rollback is exclusively triggered by lateral mantle flow produced by plate

31

collisions. As already noted, mantle flow fields giving rise to arc-trench rollback and protoophiolite genesis may be contingent on other modes of differential plate motion. However, the notion of collision-induced mantle extrusion as the driver of Tethyan ophiolite genesis appears able to reconcile coeval continental escape (Armijo et al. 1989; Jolivet et al. 1991a, 1991b), postcollision lithosphere stretching (England & Molnar 1997a, 1997b; Ren et al. 2002), and arc-trench rollback (Hussong et al. 1981; Tamaki & Honza 1991) (Fig. 7), along with regional mantle attributes cited by Flower et al. (1998, 2001) and Flower (2003).

From proto-ophiolite to ophiolite: a preliminary Tethyan verdict Studies of metamorphic soles record an unambiguous pattern of cold thrusting to at least c. 60 km depth, temperatures rising to well over 700 °C, a stage of near-isobaric cooling, and (eventual) exhumation (Wakabayashi & Dilek 1988, 2000, this volume; Hacker 1991; Gjata et al. 1992; Beccaluva et al. 1994; Searle et al. 1994; Wakabayashi & Unruh 1995; Dilek & Whitney 1997; Dilek et al. 1997; Hacker & Gnos 1997; Searle & Cox 1999; Dimo-Lahitte et al. 2001) (e.g. Fig. 5). The only serious alternative to subduction nucleation as a trigger for ophiolite formation is the proposal that ophiolites result from the consumption of recently active spreading centres at preexisting subduction zones (e.g. Hacker & Mosenfelder 1996; Hacker et al. 1996). This rests on the assumption that newly formed (<10Ma) oceanic lithosphere is too buoyant to be subducted (Cloos et al. 1998) and that high-temperature metamorphic soles represent mid-ocean ridge rather than 'normal' subduction conditions (Hacker & Mosenfelder 1996; Hacker et al. 1996). Although the two interpretations are not mutually exclusive, the question of which process dominates appears to hinge on: (1) the correspondence or otherwise of metamorphic sole and ophiolite compositions; (2) validity of the ophiolite-forearc analogue; (3) possible spatial-temporal correlations between ophiolites and regional plate tectonics. In general, 'ridge subduction' is the less appealing, given its absence from recent or active rollback cycles, whereas 'subduction nucleation', on the other hand, appears to be the rule rather than the exception (Dewey & Casey 1979; Claesson et al. 1984; Gjata et al. 1992; Stern et al. 1992; Monzier et al. 1993a; Beccaluva et al. 1994; Bloomer et al. 1995; Crawford et al. 1997; Clift & Dixon 1998; Insergueix-Filippi et al. 1998, 2000; Wakabayashi & Dilek, this volume).

32

M. F. J. FLOWER & Y. DILEK

Although some ophiolites slightly post-date their high-temperature soles, most are coeval or slightly older, consistent with interpretations that sundered forearc and remnant 'proto-arc' components were single entities prior to splitting and the onset of rollback. Incipient subduction relics are thus preserved a priori in forearc 'proto-ophiolites' (Ishii 1989; Ogawa 1995). This model is supported by thermochronological data for metamorphic soles from the Mirdita (Albania), Troodos (Cyprus), Tauride (Turkey), and Semail (Oman) ophiolites (e.g. Searle & Malpas 1980; Gnos 1998; Malpas et al. 1992; Searle et al 1994; Hacker & Gnos 1997; Dilek et al. 1999; Searle & Cox 1999). The MORB-like compositional character, inflected thermal gradients, and counterclockwise P-T-t trajectories of many metamorphic soles (Searle & Malpas 1980; Ghent & Stout 1981; Malpas et al. 1992; Shallo 1992; Encarnacion & Mukasa 1995; Hacker & Mosenfelder 1996; Hacker et al. 1996; Gnos et al. 1997; Bebien et al. 1998; Searle & Cox 1999; Bebien et al. 2000; Dimo-Lahitte et al. 2001) are all consistent if these features represent relict slab fragments from a subduction nucleation event (e.g. Fig. 5). The association of boninites with fossil transform faults such as the Arakapas fault zone (Simonian & Gass 1978; Flower & Levine 1987; MacLeod & Murton 1993), and analogous features in the Semail ophiolite (Smewing et al. 1977; Boudier et al. 1988; MacLeod & Rothery 1992), reinforces the likelihood that 'hot' subduction nucleation is largely confined to near-ridge loci. The available age data for Neo-Tethyan ophiolites allow for three broad conclusions that, we contend, are consistent with exogenous mantle flow as the driving force in their genesis. First, many, if not most, Tethyan ophiolites were emplaced after relatively short time intervals following subduction nucleation at oceanic spreading axes (Hacker & Mosenfelder 1996; Hacker et al. 1996; Dilek et al. 1999). This strongly suggests that 'oceanic' lithosphere consumed during processes of ophiolite development (although not necessarily their inception) was formed in marginal basin rather than major ocean basin settings (Pearce et al. 1984). Second, the significance of subduction nucleation, as a precursor to ophiolite genesis, in relation to distal plate tectonic events, needs further study (Gnos et al. 1997; Moores et al. 2000). Finally, given the near-ubiquitous presence of ophiolites in orogens, their formation is necessarily restricted to late stages of Wilson Cycles, immediately preceding continent-continent plate collisions. In summary, our model for Tethyan ophiolite genesis is depicted as six hypothetical stages (illustrated in Fig. 8): (1) continental plates sepa-

rate as Palaeo-Tethys begins opening; passive asthenosphere upwells beneath spreading axis to produce MORB-like oceanic lithosphere (Fig. 8a); (2) Laurasia blocks continued Palaeo-Tethys opening; new subduction is initiated at a weak (e.g. transform) zone, with boninite magmatism forming a 'proto-arc' on the overriding MORB plate, followed by 'normal' calc-alkaline arc volcanism; Neo-Tethyan rifting is initiated in Gondwana (Fig. 8b); (3) the 'Cimmeria' micro-continent detaches from Gondwana as Palaeo-Tethys continues subducting; MORB-like backarc extension and arcforearc rollback occur in response to the compression of Tethyan asthenosphere beneath Laurasia and Cimmeria (Fig. 8c); (4) continued 'Cimmerian' micro-continent migration leads to oblique collision and diachronous breakoff of the PalaeoTethyan ocean slab, accompanied by enhanced continental sediment subduction; deflected asthenosphere flow field leads to shoshonite and potassic granite magmatism derived from continentcontaminated asthenosphere (Fig. 8d); (5) NeoTethys continues to open until relict MORB-like backarc lithosphere is completely subducted and detached from the overriding continental plate (Fig. 8e); (6) Neo-Tethys ceases opening as Palaeo-Tethyan basin collapse is completed, and new subduction is initiated; assemblage of relict arc-forearc-backarc (ophiolite) is entrapped in the ensuing continent-continent orogeny (Fig. 8f). Petrological, structural, and stratigraphic data from Tethyan ophiolites consistently support the role of subduction nucleation and arc-trench rollback in their development and are reviewed by Dilek & Flower (this volume).

Conclusions (1) A unique combination of features, rare or absent from mid-ocean ridge spreading systems and subduction zones (ancient or modern), characterize Tethyan ophiolites: MORB-like 'oceanic' basement and near-coeval high-temperature metamorphic soles, succeeded by boninitic 'proto-arcs', juxtaposed refractory peridotites, and anomalous high-temperature ' epidosites'. (2) These features appear to preclude mid-ocean ridge, and 'normal' arc or backarc basin provenance, and are uniquely analogous to 'protoophiolites' currently forming in modern forearcs. (3) Modern forearcs are generated following subduction nucleation (commonly at mid-ocean ridge transforms, signalled by sequences of boninitic 'proto-arc', 'normal' arc, and backarc activity, accompanied by high-temperature hydrothermal flow) and evolve in response to one or more episodes of arc splitting and basin opening produced in response to arc-trench rollback.

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES

33

Fig. 8. Mantle-driven Tethyan 'sub-cycle' during late (pre-collision) stages of a Wilson Cycle, (a) Continental plates separate as Palaeo-Tethys begins to open; passive asthenosphere upwells beneath spreading axis to produce MORBlike oceanic lithosphere. (b) Laurasia blocks continued Palaeo-Tethys opening; new subduction is initiated at weak (e.g. transform) zone, with boninite magmatism forming a 'proto-arc' on the overriding MORE plate, followed by 'normal' calc-alkaline arc volcanism; Neo-Tethyan rifting is initiated in Gondwana. (c) 'Cimmerian' microcontinent detaches from Gondwana as Palaeo-Tethys continues subducting; MORB-like backarc extension and arc-forearc rollback occur in response to the compression of Tethyan asthenosphere beneath Laurasia and Cimmeria. (d) Continued 'Cimmerian' microcontinent migration leads to oblique collision and diachronous breakoff of the PalaeoTethyan ocean slab, accompanied by enhanced continental sediment subduction; deflected asthenosphere flow field leads to shoshonite and potassic granite magmatism derived from continent-contaminated asthenosphere. (e) NeoTethys continues to open until relict MORB-like backarc lithosphere is completely subducted and progressively detached from the overriding continental plate, (f) Neo-Tethys ceases opening as Palaeo-Tethyan basin collapse is completed, and new subduction is initiated; assemblage of relict arc-forearc-backarc (ophiolite) is entrapped in the ensuing continent-continent orogeny; asthenospheric flow contaminated by delaminated or subducted continental crust (not shown) provides sources for shoshonite, lamproite, and kamafugite (± carbonatite) magmatism.

(4) Subduction nucleation and arc-trench rollback cycles in the Mediterranean and western Pacific appear to be triggered by the effects of subhorizontal mantle flow resulting either directly from collision-induced asthenosphere extrusion, or indirectly via collision-induced plate kinematic adjustments. (5) If a rollback episode successfully evades orogenic entrapment, forearc accretion continues indefinitely, as appears to be the case in the western Pacific. On the other hand, if entrapped by a collision, as in the Mediterranean, forearc lithosphere tends to resist subduction (in contrast to backarc basin lithosphere) and is readily preserved as ophiolites.

(6) The apparent correspondence of modern and recent subduction nucleation events to 'hard' plate collisions and their respective plate kinematic responses may be discerned from Tethyan ophiolite age data. (7) The ophiolite 'conundrum' is a false dichotomy if, as seems to be the case, traditional 'ocean ridge' and 'supra-subduction' models are 'correct' only when considered together in a unified arctrench rollback model. (8) Refractory mantle sources of the type yielding calc-alkaline and boninitic magmas at newly forming subduction systems are almost certainly not generic to mid-ocean ridges. Moreover, nearubiquitous ophiolite components such as boninite

34

M. F. J. FLOWER & Y. DILEK

and HMA are exclusive to the initial 'hot subduction' stages of a rollback cycle.

We thank H. Furnes, J. Deng, V Mocanu, E. Moores, P. Robinson, R. Russo, P. Thy, J. Wakabayashi and G. Zakariadze for discussions of the ideas presented here, and J. Encarnacion, P. Robinson and an anonymous referee for thorough and constructive reviews of the paper. Support from the US National Science Foundation (EAR- 0238416 and INT-0129492 to M.F.J.F; and EAR9796011 to Y.D.), UNESCO Earth Sciences Division, the National Geographical Society (to Y.D.), the International Union of Geological Sciences, and NATO (CRG970263 and EST.CLG-97617 to YD.) is gratefully acknowledged.

References ARMIJO, R., TAPPONNIER, P. & HAN, T. 1989. Late Cenozoic right-lateral strike-slip faulting in southern Tibet, scale 1333000. American Geophysical Union, Washington, DC. AUDLEY-CHARLES, M.G. & HARRIS, R.A. 1990. Allochthonous terranes of the southwest Pacific and Indonesia. Philosophical Transactions of the Royal Society of London, Series A, 331, 571-587. BABAIE, A., ARVIN, M. & BABAIE, HA. 200la. An oblique convergence and rotation model for the emplacement of the Baft ophiolitic melange in Iran. Ofioliti, 26, 401-407. BABAIE, H.A., GHAZI, A.M., BABAEI, A., LA TOUR, T.E. & HASSANIPAK, A.A. 200 Ib. Geochemistry of arc volcanic rocks of the Zagros crush zone, Neyriz, Iran. Journal of Asian Earth Sciences, 19, 61-76. BAILEY, W.R., HOLDSWORTH, R.E. & SWARBRICK, R.E. 2000. Kinematic history of a reactivated oceanic suture; the Mamonia Complex suture zone, SW Cyprus. Journal of the Geological Society, London, 157, 1107-1126. BANERJEE, N.R. & GILLIS, K.M. 2001. Hydrothermal alteration in a modern suprasubduction zone, the Tonga forearc crust. Journal of Geophysical Research, B-106, 21737-21750. BANERJEE, N.R., GILLIS, K.M. & MUEHLENBACHS, K. 2000. Discovery of epidosites in a modern oceanic setting, the Tonga forearc. Geology, 28, 151-154. BEBIEN, J., SHALLO, M., MANIKA, K. & GEGA, D. 1998. The Shebenik massif (Albania), a link between MOR- and SSZ-type ophiolites? Ofioliti, 23, 7-15. BEBIEN, J., DIMO-LAHITTE, A., VERGELY, P., INSERGUEix-FiLiPPi, D. & DUPEYRAT, L. 2000. Albanian ophiolites: magmatic and metamorphic processes associated with the initiation of subduction. Ofioliti, 25, 39-45. BECCALUVA, L. 1982. Ophiolites and actualism, petrological constraints. Ofioliti, 7, 109-112. BECCALUVA, L., COLTORTI, M., DOSTAL, J., SACCANI, E., SERRI, G., SIENA, F. & VACCARO, C. 1994. High-Ti and low-Ti magmatic affinities from Tethyan ophiolites. In: BORTOLOTTI, V. (ed.) The Northern Apennines, the Tuscan Nappe, the Ophiolitic Sequences, the Turbiditic Successions. Societa Geologica Italiana, Rome, 149-150.

BEDNARZ, U. & SCHMINCKE, H.-U. 1994. Petrological and chemical evolution of the northeastern Troodos Extrusive Series, Cyprus. Journal of Petrology, 35, 489-523. BEIERSDORF, H., BACH, W. & DELISLE, G. ET AL. 1997. Age and possible modes of formation of the Celebes Sea basement, and thermal regimes within the accretionary complexes off SW Mindanao and N Sulawesi. In: HINTHONG, P.C. (ed.) Stratigraphy and Tectonic Evolution of Southeast Asia and the South Pacific. Department of Mineral Resources, Bangkok, 369-387. BEVIS, M., TAYLOR, F.W. & SCHUTZ, B.E. ET AL. 1995. Geodetic observations of very rapid convergence and back-arc extension at the Tonga arc. Nature, 374,249-251. BLOOMER, S.H. 1983. Distribution and origin of igneous rocks from the landward slopes of the Mariana Trench; implications for its structure and evolution. Journal of Geophysical Research, B-88, 7411-7428. BLOOMER, S.H., TAYLOR, B., MACLEOD, C.J., STERN, R.J., FRYER, P., HAWKINS, J.W. & JOHNSON, L. 1995. Early arc volcanism and the ophiolite problem; a perspective from drilling in the western Pacific. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 1-30. BOUDIER, F., CEULENEER, G. & NICOLAS, A. 1988. Shear zones, thrusts and related magmatism in the Oman Ophiolite; initiation of thrusting on an oceanic ridge. Tectonophysics, 151, 275-296. BRIAIS, A., PATRIAT, P. & TAPPONNIER, P. 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea; implications for the Tertiary tectonics of Southeast Asia. Journal of Geophysical Research, B-98, 6299-6328. CASEY, J.F. & DEWEY, J.F. 1984. Initiation of subduction zones along transform and accreting plate boundaries, triple-junction evolution, and forearc spreading centres; implications for ophiolitic geology and obduction. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 269-290. CHUNG, S.-L., SUN, S.-s., Tu, K., CHEN, C.-H. & LEE, C-.Y. 1994. Late Cenozoic basaltic volcanism around the Taiwan Strait, SE China; product of lithosphere-asthenosphere interaction during continental extension. Chemical Geology, 112, 1-20. CHUNG, S.-L., LEE, T.-Y., Lo, C.-H., WANG, P.-L., CHEN, C.-Y., NGUYEN TRONG, Y., TRAN TRONG, H. & Wu, G. 1997. Intraplate extension prior to continental extrusion along the Ailao Shan-Red River shear zone. Geology, 25, 311-314. CLAESSON, S., PALLISTER, J.S. & TATSUMOTO, M. 1984. Samarium-neodymium data on two late Proterozoic ophiolites of Saudi Arabia and implications for crustal and mantle evolution. Contributions to Mineralogy and Petrology, 85, 244-252. CLIFT, P.D. 1998. On the origin of some Eohellenic ophiolites—Reply to Commentary on Clift & Dixon 1998, Jurassic ridge collapse, subduction nucleation

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES and ophiolite obduction in the southern Greek Tethys. Eclogae Geologicae Helvetiae, 91, 432-434. CLIFT, P.D. & DIXON, I.E. 1998. Jurassic ridge collapse, subduction nucleation and ophiolite obduction in the southern Greek Tethys. Eclogae Geologicae Helvetica, 91, 123-138. CLOOS, M., SAPHE, B. & WEILAND, R.J. 1998. Continental margin subduction, collision, and lithospheric delamination in New Guinea. Geological Society of America, Abstracts with Programs, 30, 208. COLEMAN, R.G. 1977. Emplacement and metamorphism of ophiolites. Ofioliti, 2, 41-74. COMPAGNONI, R., MORLOTTI, E. & TORELLI, L.

1989.

Crystalline and sedimentary rocks from the scarps of the Sicily-Sardinia Trough and Cornaglia Terrace (southwestern Tyrrhenian Sea, Italy); paleogeographic and geodynamic implications. Chemical Geology, 77, 375-398. CRAWFORD, A.J., VERBEETEN, A. & EGGINS, S. 1997. Hot subduction; magmatism along the Hunter Ridge, SW Pacific. In: EDITOR, A. (ed.) State of the Arc '97, Island Arc Magma Genesis Workshop, Abstracts with Program. Geological Society of Australia, 45, 9-12. DALZIEL, I. 1989. Inversion of circum-Pacific marginal basins. In: COOPER, M.A. & WILLIAMS, G.D. (eds) Inversion Tectonics. Geological Society, London, Special Publications, 44, 351. DANYUSHEVSKY, L.V., SOBOLEV, A.V. & FALLOON, T.J. 1995. North Tongan high-Ca boninite petrogenesis—role of the Samoan plume and subduction zone-transform fault transition. Journal of Geodynamics, 20, 219-241. DELALOYE, M., FONTIGNIE, D. & SPADEA, P. 1984. Radiometric dates of Calabria-Lucania ophiolites (southern Italy) and associated continental rocks. Ofioliti, 9, 687. DERCOURT, J., ZONENSHAIN, L.P. & Ricou, L.E. ET AL. 1986. Geological evolution of the Tethys belt from the Atlantic to the Pamirs since the Lias, scale 1:20000000. In: AUBOUIN, J., LEPICHON, X. & MONIN, A.S. (eds) Evolution of the Tethys. Elsevier, Amsterdam, 241 - 315. DEWEY, J.F. & CASEY, J.F. 1979. Nucleation of subduction zones along transform faults and ophiolite emplacement. Geological Society of America, Abstracts with Program, 11, 413. DEWEY, J.F., HELMAN, M.L., TURCO, E., HUTTON, D.H.W. & KNOT, S.D. 1989. Kinematics of the western Mediterranean. In: COWARD, M.P., DIETRICH, D. & PARK, R.G. (eds) Alpine Tectonics. Geological Society, London, Special Publications, 45, 265-283. DILEK, Y. & WHITNEY, D.L. 1997. Counterclockwise P-T-t trajectory from the metamorphic sole of a Neo-Tethyan ophiolite (Turkey). Tectonophysics, 280,295-310. DILEK, Y., CARVER, J.I. & WHITNEY, D.L. 1997. Extensional exhumation, uplift, and crustal cooling in a collision orogen and the geomorphic response, central Anatolia, Turkey. Geological Society of America, Abstracts with Programs, 29, 477. DILEK, Y., THY, P., HACKER, B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and

35

mafic dike intrusions (Turkey): implications for the Neo-Tethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) 2000. Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 552pp. DIMO-LAHITTE, A., MONIE, P. & VERGELY, P. 2001. Metamorphic soles from the Albanian ophiolites; petrology, 40Ar/39Ar geochronology, and geodynamic evolution. Tectonics, 20, 78-96. DVORKIN, J., NUR, A., MAVKO, G. & BEN-AVRAHAM, Z. 1993. Narrow subducting slabs and the origin of backarc basins. Tectonophysics, 227, 63-79. ENCARNACION, J. 1997. Multiple ophiolite generation preserved in the northern Philippines and the growth of island arc complexes. EOS Transactions, American Geophysical Union, 78, 697. ENCARNACION, J.P. & MUKASA, S.B. 1995. Post-spreading crustal melting in the South China Sea area; zircon and monazite U-Pb systematics and geochemistry of the Capoas Granite, Palawan, Philippines. EOS Transactions, American Geophysical Union, 76, 286. ENCARNACION, J., FERNANDEZ, D. & MATTINSON, J. 2001. Subduction nucleation by extrusion tectonics? Evidence from the Palawan Ophiolite, Philippines. EOS Transactions, American Geophysical Union, 82, 47. ENGLAND, P. & MOLNAR, P. 1997a. Active deformation of Asia; from kinematics to dynamics. Science, 278, 647-650. ENGLAND, P. & MOLNAR, P. 1997b. The field of crustal velocity in Asia calculated from Quaternary rates of slip on faults. Geophysical Journal International, 130, 551-582. FACCENNA, C, MATTEI, M., FUNICIELLO, R. & JOLIVET, L. 1997. Styles of back-arc extension in the central Mediterranean. Terra Nova, 9, 126-130. FALLOON, T.J. & CRAWFORD, A.J. 1991. The petrogenesis of high-calcium boninite lavas dredged from the northern Tonga Ridge. Earth and Planetary Science Letters, 102, 375-394. FALLOON, T.J. & DANYUSHEVSKY, L.V. 2000. Melting of refractory mantle at 1.5, 2 and 2.5 GPa under anhydrous and H2O-undersaturated conditions: implications for the petrogenesis of high-Ca boninites and the influence of subduction components on mantle melting. Journal of Petrology, 41, 257-283. FASSOULAS, C. 1999. The structural evolution of central Crete; insight into the tectonic evolution of the southern Aegean (Greece). Journal ofGeodynamics, 27, 23-43. FLOWER, M.FJ. 2003. Ophiolites, historical contingency, and the Wilson cycle. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolites and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 125-150. FLOWER, M.FJ. & LEVINE, H.M. 1987. Petrogenesis of a tholeiite-boninite sequence from Ayios Mamas, Troodos ophiolite; evidence for splitting of a volcanic arc? Contributions to Mineralogy and Petrology, 97, 509-524.

36

M. F. J. FLOWER & Y. DILEK

FLOWER, M.F.J., TAMAKI, K. & HOANG, N. 1998. Mantle extrusion: a model for dispersed volcanism and DUPAL-like asthenosphere in East Asia and the western Pacific. In: FLOWER, M.F.J., CHUNG, S.-L., Lo, C.-H. & LEE, T.-Y. (eds) Mantle dynamics and Plate Interactions in East Asia. American Geophysical Union, Geodynamics Series, 27, 67-88. FLOWER, M.F.J., Russo, R.M., TAMAKI, K. & HOANG, N. 2001. Mantle contamination and the IzuBonin-Mariana (IBM) 'high-tide mark': evidence for mantle extrusion caused by Tethyan closure. Tectonophysics, 333, 9-34. FORSTER, M.A. & LISTER, G.S. 1999. Separate episodes of eclogite and blueschist facies metamorphism in the Aegean metamorphic core complex of los, Cyclades, Greece. In: MAC NIOCAILL, C. & RYAN, P.O. (ed.) Continental Tectonics. Geological Society, London, Special Publications, 164, 157-177. FORSYTH, D.W. & UYEDA, S. 1975. On the relative importance of the driving forces of plate motion. Geophysical Journal of the Royal Astronomical Society, 43, 163-200. FORTUIN, A.R., VANDERWERFF, W. & WENSINK, H. 1997. Neogene basin history and paleomagnetism of a rifted and inverted forearc region, on- and offshore Sumba, Eastern Indonesia. Journal of Asian Earth Sciences, 15, 61—88. FRYER, P., LOCKWOOD, J.P., BECKER, N., PHIPPS, S. & TODD, C.S. 2000. Significance of serpentine mud volcanism in convergent margins. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349,35-51. GASS, I.G. 1968. Is the Troodos massif of Cyprus a fragment of Mesozoic ocean floor? Nature, 220, 39-42. GASS, I.G. 1989. Magmatic processes at and near constructive plate margins as deduced from the Troodos (Cyprus) and Semail Nappe (N Oman) ophiolites. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 1-15. GHASEMI, H., JUTEAU, T., BELLON, H., SABZEHEI, M., WHITECHURCH, H. & Ricou, L.E. 2002. The mafic-ultramafic complex of Sikhoran (central Iran), a polygenetic ophiolite complex. Comptes Rendus de I'Academic des Sciences (Geosciences), 334,431-438. GHAZI, A.M. & HASSANIPAK, A.A. 2000. Petrology and geochemistry of the Shahr-Babak Ophiolite, central Iran. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 485-497. GHENT, E.D. & STOUT, M.Z. 1981. Metamorphism at the base of the Samail ophiolite, southeastern Oman Mountains. Journal of Geophysical Research, B-86, 2557-2571. GIARAMITA, M.J., MACPHERSON, GJ. & PHIPPS, S.P. 1992. Forearc geochemical signatures in volcanic rocks, Llanada and Black Mountain remnants, Coast

Range Ophiolite, California. Geological Society of America, Abstracts with Programs, 24, 42. GIBSON, I.L., McGiLL, G.C. & ROBINSON, P.T. 1987. The geochemistry and nature of unusual glassy lavas from CCSP holes CY-2 and CY-2a, Troodos Ophiolite, Cyprus. In: ROBINSON, P.T., GIBSON, I.L. & PANAYIOTOU, A. (eds) Cyprus Crustal Study Project; Initial Report, Holes CY-2 and 2a. Geological Survey of Canada, Special Papers, 85-29, 79-85. GILLIS, K.M. 2002. The rootzone of an ancient hydrothermal system exposed in the Troodos ophiolite, Cyprus. Journal of Geology, 110, 57-74. GILLIS, K.M. & BANERJEE, N.R. 2000. Hydrothermal alteration patterns in supra-subduction zone ophiolites. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 283-297. GJATA, K., KORNPROBST, J.K.A., BRIOT, D. & PINEAU, F. 1992. Hot subduction close to a ridge—example of the garnet pyroxenite enclaves in the serpentinite breccia at Derveni, Albania. Bulletin de la Societe Geologique de France, 163, 469-476. GNOS E. 1998. Peak metamorphic conditions of garnet amphibolites beneath the Semail Ophiolite; implications for an inverted pressure gradient. International Geology Review, 40, 281-304. GNOS, E. & PETERS, T. 1993. K-Ar ages of the metamorphic sole of the Semail Ophiolite; implications for ophiolite cooling history. Contributions to Mineralogy and Petrology, 113, 325-332. GNOS, E., IMMENHAUSER, A. & PETERS, T. 1997. Late Cretaceous/early Tertiary convergence between the Indian and Arabian plates recorded in ophiolites and related sediments. Tectonophysics, III, 1-19. GRIBBLE, R.F., STERN, R.J., NEWMAN, S., BLOOMER, S.H. & O'HEARN, T. 1998. Chemical and isotopic composition of lavas from the northern Mariana Trough; implications for magmagenesis in back-arc basins. Journal of Petrology, 39, 125-154. HACKER, B.R. 1991. The role of deformation in the formation of metamorphic gradients; ridge subduction beneath the Oman Ophiolite. Tectonics, 10, 455-473. HACKER, B.R. & GNOS, E. 1997. The conundrum of Samail; explaining the metamorphic history. Tectonophysics, 279, 215-226. HACKER, B.R. & MOSENFELDER, J.L. 1996. Metamorphism and deformation along the emplacement thrust of the Samail Ophiolite, Oman. Earth and Planetary Science Letters, 144, 435-451. HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman Ophiolite; thermal and geochronologic constraints. Tectonics, 15, 1230-1247. HALL, R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific, computerbased reconstructions, model and animations. Journal of Asian Earth Sciences, 20, 353-431. HAWKINS, J.W. & CASTILLO, P.R. 1998. Early history of the Izu-Bonin-Mariana arc system, evidence from Belau and the Palau Trench. Island Arc., 1,

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES 559-578. MICKEY-VARGAS, R. 1989. Boninites and tholeiites from DSDP Site 458, Mariana forearc. In: CRAWFORD, A.J. (ed.) Boninites. Allen & Unwin, London, 339-356. HIROSE, K. 1997. Melting experiments on Iherzolite KLB-1 under hydrous conditions and generation of high-magnesian andesitic melts. Geology, 25, 42-44. HIROSE, K. & KUSHIRO, L. 1993. Partial melting of dry peridotites at high pressures: determination of composition of melts segregated from peridotite using aggregate of diamond. Earth and Planetary Science Letters, 114, 477-489. HIROSE, K. & KAWAMOTO, T. 1995. Hydrous partial melting of Iherzolite at 1 GPa; the effect of H2O on the genesis of basaltic magmas. Earth and Planetary Science Letters, 133, 463-473. Hsu, KJ. 1994. Tectonic facies in an archipelago model of intra-plate orogenesis. GSA Today, 4, 289-290. Hsu, K.J. 1997. An application of the tectonic facies concept and the archipelago model of orogenesis to interpret the Phanerozoic tectonic history of China. Journal of the Geological Society of China, 40, 469-480. HUSSONG, D.M., UYEDA, S. & NATLAND, J.H. 1981. Tectonic processes and history of the Mariana arc system. EOS Transactions, American Geophysical Union, 62, 1090. HYNES, A. & MOTT, J. 1985. On the causes of back-arc spreading. Geology, 13, 387-389. iNSERGUEIX-FlLIPPI, D., DUPEYRAT, L., DlMO, A., VER-

GELY, P. & BEBIEN, J. 1998. Model of subduction zone infancy at a mid-ocean ridge; magmatic and metamorphic implications. In: HAUSKI, E. & VUOLLO, J. (ed.) International Ophiolite Symposium and Field Excursion; Generation and Emplacement of Ophiolites through Time; Abstracts, Excursion Guide. Geological Survey of Finland, Special Paper, 26, 28. INSERGUEIX-FlLIPPI, D., DUPEYRAT, L., DlMO-LAHITTE,

A., VERGELY, P. & BEBIEN, J. 2000. Albanian ophiolites, II—Model of subduction zone infancy at a mid-ocean ridge. Ofioliti, 25, 47-53. ISACKS, B. & MOLNAR, P. 1971. Distribution of stresses in the descending lithosphere from a global survey of focal-mechanism solutions of mantle earthquakes. Reviews of Geophysics and Space Physics, 9, 103-174. ISHII, T. 1989. Origin of proto-ophiolites in IzuOgasawara—Mariana fore-arc and their resemblance to the Troodos Ophiolite. International Geological Congress, Abstracts with Program. Tokyo, 28, 2.99. ISHII, T., ROBINSON, P.T., TOKUYAMA, H. & KOBAYASHI, K. 1988. Relations of fore-arc proto-ophiolites, mantle and serpentine diapirs. Geological Society of America, Abstracts with Programs, 20, 158. ISHII, T., ROBINSON, P.T. & MAEKAWA, H. ET AL. 1992. Petrological studies of peridotites from diapiric serpentinite seamounts in the Izu-OgasawaraMariana forearc. In: DEARMONT, L.H. (ed.) Leg 125, Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 445-485.

37

ISHII, T., HARAGUCHI, S. & TAIRA, A. ET AL. 1995. Dredged igneous rocks from the two remnant arcs; the West Mariana (KT95-9 cruise) and KyushuPalau (KT94-10) ridges. EOS Transactions, American Geophysical Union, 76, 682. ISHIKAWA, T., NAGAISHI, K. & UMINO, S. 2002. Boninitic volcanism in the Oman Ophiolite; implications for thermal condition during transition from spreading ridge to arc. Geology, 30, 899-902. JOHNSON, L.E., FRYER, P. & TAYLOR, B. ET AL. 1991. New evidence for crustal accretion in the outer Mariana fore arc; Cretaceous radiolarian cherts and mid-ocean ridge basalt-like lavas. Geology, 19, 811-814. JOHNSON, L.E., FRYER, P. & PEARCE, J.A. ET AL. 1992. Mafic clasts in serpentine seamounts; petrology and geochemistry of a diverse crustal suite from the outer Mariana forearc. In: DEARMONT, L.H. (ed.) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, 401-413. JOLIVET, L. & FACCENNA, C. 2000. Mediterranean extension and the Africa-Eurasia collision. Tectonics, 19, 1095-1106. JOLIVET, L., BRUN, M.P. & GAUTIER, P. 199la. Threedimensional geometry of extension in the Aegean. Sixth Biennial Meeting, European Union of Geosciences. Blackwell Scientific, Oxford, 271. JOLIVET, L., HUCHON, P., BRUN, J.P., LE PICHON, X., CHAMOT-ROOKE, N. & THOMAS, J.C. 1991b. Arc deformation and marginal basin opening; Japan Sea as a case study. Journal of Geophysical Research, B-96, 4367-4384. JOLIVET, L., FRIZON DE LAMOTTE, D., MASCLE, A. & SERANNE, M. ET AL. 1999. The Mediterranean basins; Tertiary extension within the Alpine Orogen; an introduction. In: DURAND, B. (ed.) The Mediterranean Basins; Tertiary Extension within the Alpine Orogen. Geological Society, London, Special Publications, 156, 1-14. KARIG, D.E. 1971. Origin and development of marginal basins in the western Pacific. Journal of Geophysical Research, B-76, 2542-2561. KARIG, D.E., SAREWITZ, D.R. & HAECK, G.D. 1986. Role of strike-slip faulting in the evolution of allochthonous terranes in the Philippines. Geology, 14, 852-855. LACHIZE, M., LORAND, J.P. & JUTEAU, T. 1996. Calcalkaline differentiation trend in the plutonic sequence of the Wadi Haymiliyah section, Haylayn Massif, Semail Ophiolite, Oman. Lithos, 38, 207-232. LANGOSCH, A., STOSCH, H.-G., SEIDEL, E. & OKRUSCH, M. 1999. Late Cretaceous plutonic rocks within the ophiolitic melange of Crete; constraints on their origin and evolution. 10th Biennial Meeting, European Union ofGeosciences. Cambridge University Press, Cambridge, 776. LANGOSCH, A., SEIDEL, E., STOSCH, H.-G. & OKRUSCH, M. 2000. Intrusive rocks in the ophiolitic melange of Crete; witnesses to a Late Cretaceous thermal event of enigmatic geological position. Contributions to Mineralogy and Petrology, 139, 339-355. LATIN, D. & WHITE, N. 1990. Generating melt during

38

M. F. J. FLOWER & Y. DILEK

lithospheric extension; pure shear vs. simple shear. Geology, 18, 327-331. LE PICHON, X. 1982. Land-locked oceanic basins and continental collision; the eastern Mediterranean as a case example. In: Hsu, K.J. (ed.) Mountain Building Processes. Academic Press, London, 201—211. LE PICHON, X., HENRY, P., GOFFE, B. & CHOPIN, C. 1995. From crustal collision to crustal subduction; a model of evolution from the Alps to the Himalayas. European Union of Geosciences 8; Oral and Poster Presentations. Blackwell Scientific, Oxford, 8, 114. LE PICHON, X., LALLEMANT, S.J., CHAMOT-ROOKE, N., LEMEUR, D. & PASCAL, G. 2002. The Mediterranean Ridge backstop and the Hellenic nappes. Marine Geology, 186, 111-125. LEE, C.W., ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. 1990. A review of platform sedimentation in the Early and Late Permian of Oman, with particular reference to the Oman Mountains. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 39-47. LEE, T.-Y. & LAWYER, L.A. 1994. Cenozoic plate reconstruction of the South China Sea region, geology and geophysics of the South China Sea and environs. Tectonophysics, 235, 149-180. LEE, T.-Y., Lo, C.-H., CHUNG, S.-L., LAN, C.-Y., WANG, P.-L. & LAWVER, L.A. 2000. Extrusion tectonics in Southeast Asia; constraints from South China continental margin. AAPG International Conference and Exhibition; Abstracts. American Association of Petroleum Geologists, Tulsa, OK, 1454. LUNDGREN, P.R.,

RUSSO, R.M.,

D'AGOSTINO, N.

&

GIARDINI, D. 1996. Mediterranean crustal motions; numerical models constrained by geologic and space geodetic data. EOS Transactions, American Geophysical Union, 77, 149. MACLEOD, CJ. & MURTON, B.J. 1993. Structure and tectonic evolution of the southern Troodos transform fault zone, Cyprus. In: PRICHARD, H.M., ALABASTER, T., HARRIS, N.B.W. & NEARY, C.R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 141-176. MACLEOD, CJ. & ROTHERY, D.A. 1992. Ridge axial segmentation in the Oman Ophiolite; evidence from along-strike variations in the sheeted dyke complex. In: PARSON, L.M., MURTON, B.J. & BROWNING, P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 39-63. MALPAS, J., SPOERLI, K.B., BLACK, P.M. & SMITH, LE.M. 1992. Northland Ophiolite, New Zealand, and implications for plate-tectonic evolution of the Southwest Pacific. Geology, 20, 149-152. MARLOW, M.S., JOHNSON, L.E., PEARCE, J.A., FRYER, P.B., PICKTHORN, L.G. & MURTON, B.J. 1992. Upper Cenozoic volcanic rocks in the Mariana forearc recovered from drilling at Ocean Drilling Program Site 781; implications for forearc magmatism. Journal of Geophysical Research, B-97, 15085-15097. McCABE, R. & UYEDA, S. 1983. Hypothetical model for

the bending of the Mariana Arc. In: HAYES, D. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands; Part 2. Geophysical Monograph, American Geophysical Union, 27, 281-293. MCKENZIE, D. & BICKLE, MJ. 1988. The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology, 29, 625-679. METCALFE, I. 1996. Gondwanaland dispersion, Asian accretion and evolution of eastern Tethys, breakup of Rodinia and Gondwanaland and assembly of Asia. Australian Journal of Earth Sciences, 43, 605-623. METCALFE, I. ET AL. 1999. Late Palaeozoic-early Mesozoic evolution and palaeogeography of eastern Tethys. In: YIN, H. (ed.) Proceedings of the International Conference on Pangea and the Paleozoic-Mesozoic Transition. China Universities Geosciences Press, Wuhan, 131-133. MIGIROS, G., HATZIPANAGIOTOU, K., GARTZOS, E., SERELIS, K. & TSIKOURAS, B. 2000. Petrogenetic evolution of ultramafic rocks from Lesvos Island (NE Aegean, Greece). Chemie der Erde, 60, 27-46. MIYASHIRO, A. 1973. The Troodos ophiolitic complex was probably formed in an island arc. Earth and Planetary Science Letters, 19, 218-224. MIYASHIRO, A. 1975. Classification, characteristics, and origin of ophiolites. Journal of Geology, 83, 249-281. MIYASHIRO, A. 1977. Subduction-zone ophiolites and island-arc ophiolites. In: SAXENA, S.K. & BHATTACHARJI, S. (eds) Energetics of Geological Processes. Springer, Heidelberg, 188-213. MONZIER, M., DANYUSHEVSKY, L.V., CRAWFORD, A.J., BELLON, H. & GOTTEN, J. 1993a. High-Mg andesites from the southern termination of the New Hebrides island arc (SW Pacific). Journal of Volcanology and Geothermal Research, 57, 193-217. MONZIER, M., ROBIN, C., EISSEN, J.P., DUGGAN, M.B.C. & KNUTSON, J.C. 1993b. Seismotectonics-geochemistry relationships along the New Hebrides volcanic arc. IAVCEI Symposium, Abstracts with Program, Canberra, 15. MONZIER, M., ROBIN, C., EISSEN, J.-P. & GOTTEN, J. 1997. Geochemistry vs. seismo-tectonics along the volcanic New Hebrides central chain (Southwest Pacific). Journal of Volcanology and Geothermal Research, 78, 1-29. MOORES, E.M., SCOTT, R.B. & LUMSDEN, W.W. 1968. Tertiary tectonics of the White Pine-Grant Range region, east-central Nevada, and some regional implications. Geological Society of America Bulletin, 19, 1703-1726. MOORES, E.M., KELLOGG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency'; a resolution of the 'Ophiolite conundrum'. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 3-12. MORRA, V., SECCHI, F.A.G., MELLUSO, L. & FRANCIOSI, L. 1997. High-Mg subduction-related Tertiary basalts in Sardinia, Italy. Lithos, 40, 69-91.

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES OGAWA, Y. 1995. Some problems on emplacement of ophiolite rocks around plate subduction boundaries. Dynamics of Oceanic Crust: Chigaku Zasshi (J. Geogr.), 104, 468-472. OGAWA, Y. & TANIGUCHI, H. 1989. Origin and emplacement of basalt rocks in the accretionary complexes in SW Japan. Ofioliti, 14, 177-193. PADOA, E. 1999. Les Ophiolites du Massif de 1'Inzecca (Corse alpine); lithostratigraphie, structure geologique et evolution geodynamique. Geologic de la France, 3, 37-48. PARKINSON, I.J. & ARCULUS, RJ. 1999. The redox state of subduction zones, insights from arc peridotites. Chemical Geology, 160, 409-423. PARKINSON, I.J., HAWKESWORTH, CJ. & COHEN, A.S. 1998. Ancient mantle in a modern arc: osmium isotopes in Izu-Bonin-Mariana forearc peridotites. Science, 281, 2011-2013. PARLAK, O. & DELALOYE, M. 1999. Precise 40Ar/39Ar ages from the metamorphic sole of the Mersin ophiolite (southern Turkey). Tectonophysics, 301, 145-158. PARLAK, O., HOCK, V. & DELALOYE, M. 2000. Suprasubduction zone origin of the Pozanti-Karsanti Ophiolite (southern Turkey) deduced from wholerock and mineral chemistry of the gabbroic cumulates. In: BOZKURT, E., WINCHESTER, J.A. & PIPER, J.D.A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 219-234. PEACOCK, S.M. 1994. The thermal and petrologic structure of subduction zones; implications for seismic tomography. EOS Transactions, American Geophysical Union, 75, 641. PEACOCK, S.M., PONKO, S.C. & SANDERS, C.O. 1995. On the petrologic, thermal, and seismic structure of the subduction-zone mantle wedge. IUGG XXI, Abstracts with Program, 21, 470. PEARCE, J.A., ALABASTER, T., SHELTON, A.W. & SEARLE, M.P. 1981. The Oman ophiolite as a Cretaceous arc-basin complex: evidence and implications. Proceedings of the Geological Society, London, 300, 299-317. PEARCE, J.A., LIPPARD, SJ. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites, marginal basin geology; volcanic and associated sedimentary and tectonic processes in modern and ancient marginal basins. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 74-94. PEARCE, J.A, THIRLWALL, M.F & INGRAM, G. ET AL. \992a. Isotopic evidence for the origin of boninites and related rocks drilled in the Izu-Bonin (Ogasawara) forearc, Leg 125. Proceedings of the Ocean Drilling Program, Scientific Results, 125, 237-261. PEARCE, J.A, THIRLWALL, M.F & INGRAM, G. ET AL. 1992fl. Isotopic evidence for the origin of boninites and related rocks drilled in the Izu-Bonin (Ogasawara) forearc, Leg 125. Proceedings of the Ocean Drilling Program, Scientific Results, 125, 237-261. PE-PIPER, G. 1994. Lead isotopic compositions of Neogene volcanic rocks from the Aegean extensional area. Chemical Geology, 118, 27-41.

39

PE-PIPER, G. & PIPER, DJ.W. 1989. Spatial and temporal variation in late Cenozoic back-arc volcanic rocks, Aegean Sea region. Tectonophysics, 169, 113-134. PE-PIPER, G. & PIPER, DJ.W. 1994. Miocene magnesian andesites and dacites, Evia, Greece; adakites associated with subducting slab detachment and extension. Lithos, 31, 125-140. PEZARD, P.A., EINAUDI, F., HERMITTE, D., COCHEME, J.-J., COULON, C. & LAVERNE, C. 2000. MORB emplacement and structure; insights from the Semail Ophiolite, Oman. Geophysical Research Letters, 21, 3933-3936. PORTNYAGIN, M.V.,

MAGAKYAN, R.

& SCHMINCKE,

H.U. 1996. Geochemical variability of boninite magmas: evidence from magmatic inclusions in highly magnesian olivine from lavas of southwestern Cyprus. Petrology, 4, 231-246. PORTNYAGIN, M.V., DANUSHEVSKY, L.V. & KAMANETSKY, V.S. 1997. Coexistence of two distinct mantle sources during formation of ophiolites: a case study of primitive pillow-lavas from the lowest part of the volcanic section of the Troodos Ophiolite, Cyprus. Contributions to Mineralogy and Petrology, 128, 287-301. PUBELLIER, M. & COBBOLD, P.R. 1996. Analogue models for the transpressional docking of volcanic arcs in the western Pacific. Tectonophysics, 253, 33-52. PUBELLIER, M., QUEBRAL, R., AURELIO, M. & RANGIN, C. 1996. Docking and post-docking escape tectonics in the southern Philippines. In: HALL, R. & BLUNDELL, D.J. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 511-523. RANGIN, C., STEPHAN, J.F. & MUELLER, C. 1985. Middle Oligocene oceanic crust of South China Sea jammed into Mindoro collision zone (Philippines). Geology, 13, 425-428. RANGIN, C., SILVER, E.A. & TAMAKI, K. 1995. Closure of western Pacific marginal basins; rupture of the oceanic crust and the emplacement of ophiolites. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 405-417. REAGAN, M.K. & MEIJER, A. 1984. Geology and geochemistry of early arc-volcanic rocks from Guam. Geological Society of America Bulletin, 95, 701-713. REN, J., TAMAKI, K., Li, S. & ZHANG, J. 2002. Late Mesozoic and Cenozoic rifting and its dynamic setting in eastern China and adjacent areas. Tectonophysics, 344, 175-205. ROBERTSON, A. & SHALLO, M. 2000. Mesozoic-Tertiary tectonic evolution of Albania in its regional eastern Mediterranean context. Tectonophysics, 316, 197-254. ROBERTSON, A.H.F. 2000. Mesozoic-Tertiary tectonicsedimentary evolution of a South Tethyan oceanic basin and its margins in southern Turkey. In: BOZKURT, E., WINCHESTER, J.A. & PIPER, J.D.A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London,

40

M. F. J. FLOWER & Y. DILEK

Special Publications, 173, 97-138. ROGERS, N.W., MACLEOD, CJ. & MURTON, BJ. 1989. Petrogenesis of boninitic lavas from the Limassol Forest Complex, Cyprus. In: CRAWFORD, A.J. (ed.) Boninites. Unwin Hyman, London, 288-313. ROYDEN, L.H. 1993a. The tectonic expression of slab pull at continental convergent boundaries. Tectonics, 12, 303-325. ROYDEN, L.H. 1993b. Evolution of retreating subduction boundaries formed during continental collision. Tectonics, 12, 629-638. Russo, R.M. & SILVER, P.G. 1996. Cordillera formation, mantle dynamics, and the Wilson cycle. Geology, 24,511-514. Russo, R.M., SPEED, R.C., OKAL, E.A., SHEPHERD, J.B. & ROWLEY, K.C. 1993. Seismicity and tectonics of the southeastern Caribbean. Journal of Geophysical Research, B-98, 14299-14319. SEARLE, M. & Cox, J. 1999. Tectonic setting, origin, and obduction of the Oman ophiolite. Geological Society of America Bulletin, I I I , 104-122. SEARLE, M.P. & MALPAS, J. 1980. Structure and metamorphism of rocks beneath the Semail Ophiolite of Oman and their significance in ophiolite obduction. Transactions of the Royal Society of Edinburgh, Part 4, 71, 247-262. SEARLE, M.P., WATERS, D.J., MARTIN, H.N. & REX, D.C. 1994. Structure and metamorphism of blueschist-eclogite facies rocks from the northeastern Oman Mountains. Journal of the Geological Society, London, 151, 555-576. SENO, T. & YAMANAKA, Y. 1998. Arc stresses determined by slabs: implications for mechanisms of back-arc spreading. Geophysical Research Letters, 25, 3227-3230. SHALLO, M. 1992. Geological evolution of the Albanian ophiolites and their platform periphery. Geologische Rundschau, 81, 681-694. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, U1-U45. SHINJO, R. 1999. Geochemistry of high-Mg andesites and the tectonic evolution of the Okinawa TroughRyukyu Arc system. Chemical Geology, 157, 69-88. SILVER, P.G. & Russo, R.M. 1996. Coupling of plate motion and deformation between the African and South American plates. EOS Transactions, American Geophysical Union, 733-734. SIMONIAN, K.O. & GASS, I.G. 1978. Arakapas fault belt, Cyprus; a fossil transform fault. Geological Society of America Bulletin, 89, 1220-1230. SMEWING, J.D., SIMONIAN, K.O., ELBOUSHI, I.M. & GASS, I.G. 1977. Mineralized fault zone parallel to the Oman ophiolite spreading axis. Geology, 5, 534-538. SMITH, A.G. & SPRAY, J.G. 1984. A half-ridge transform model for the Hellenic-Dinaric ophiolites. In: DIXON, I.E. & ROBERTSON, A.H.F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 629-644. SPAKMAN, W. & BIJWAARD, H. 1998. Lithosphere and mantle structure below the African and Eurasian

plates. 23rd General Assembly, European Geophysical Society:!. Solid Earth, Geophysics, Geodesy, Abstracts with Program, 15. SPAKMAN, W., VAN DER LEE, S. & VAN DER HILST, R. 1992. P-wave velocity anomalies in the Aegean mantle to a depth of 1400km from delay time tomography. EOS Transactions, American Geophysical Union, 202. STAMPFLI, G.M. & BOREL, G.D. 2002. A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrons. Earth and Planetary Science Letters, 196, 17-33. STAMPFLI, G.M., MOSAR, J., FAVRE, P., PILLEVUIT, A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethys realm; the Neo-Tethys East Mediterranean Basin connection. In: ZIEGLER, P.A., CAVAZZA, W., ROBERTSON, A.H.F & CRASQUINSOLEAU, S. (eds) Peri-Tethyan Rift/Wrench Basins and Passive Margins. Peri-Tethys Memoir 6. Memoires de la Museum Nationale d'Histoire Naturelle, 186,51-108. STERN, R.J. & BLOOMER, S.H. 1992. Subduction zone infancy; examples from the Eocene Izu—BoninMariana and Jurassic California arcs. Geological Society of America Bulletin, 104, 1621-1636. STERN, R.J., BLOOMER, S.H., PING-NAN, L. & SMOOT, N.C. 1989. Submarine arc volcanism in the southern Mariana Arc as an ophiolite analogue. Tectonophysics, 168, 151-170. STERN, R.J., SULTAN, M. & ABDEL-SALAM, M.G. 1992. Pacific margins of Laurentia and East AntarcticaAustralia as a conjugate rift pair: evidence and implications for an Eocambrian supercontinent. Geology, 20, 190. TAMAKI, K. 1995. Upper mantle extrusion tectonics of southeast Asia and formation of western Pacific backarc basins. In: LE PICHON, X., RANGIN, C. & TAPPONNIER, P. (eds) International Workshop, Cenozoic Evolution of the Indochina Peninsula, HanoiDo Son, Abstracts with Program, Hanoi, 89. TAMAKI, K. & HONZA, E. 1991. Global tectonics and formation of marginal basins: role of the western Pacific. Episodes, 14, 224-230. TAPPONNIER, R., PELTZER, G., LE DAIN, A.Y., ARMIJO, R. & COBBOLD, P. 1982. Propagating extrusion tectonics in Asia; new insights from simple experiments with plasticine. Geology, 10, 611-616. TAPPONNIER, P., PELTZER, G. & ARMIJO, R. ET AL. 1986. On the mechanics of the collision between India and Asia. In: COWARD, M.P. (ed.) Collision Tectonics. Geological Society, London, Special Publications, 19, 115-157. TAYLOR, R.N., NESBITT, R.W. & VIDAL, P., ET AL. 1994. Mineralogy, chemistry, and genesis of the boninite series volcanics, Chichijima, Bonin Islands, Japan. Journal of Petrology, 35, 577-617. TOTH, J. & GURNIS, M. 1998. Dynamics of subduction initiation at preexisting fault zones. Journal of Geophysical Research, B-103, 18053-18067. UMINO, S. & KUSHIRO, I. 1989. Experimental studies on boninite petrogenesis. In: CRAWFORD, A.J. (ed.) Boninites. Allen & Unwin, London, 89- 111. UMINO, S., YANAI, S., JAMAN, A.R., NAKAMURA, Y. &

MANTLE FLOW MODEL FOR TETHYAN OPHIOLITES IIYAMA, J.T. 1990. The transition from spreading to subduction; evidence from the Semail Ophiolite, northern Oman Mountains. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C, (eds) Ophiolites; Oceanic Crustal Analogues; Proceedings of the Symposium Troodos 1987. Ministry of Agriculture and Natural Resources, Nicosia, 1987, 375-384. URQUHART, E. & BANNER, F.T. 1994. Biostratigraphy of the Supra-Ophiolite Sediments of the Troodos Massif, Cyprus—the Cretaceous Peraphedhi, Kannaviou, Moni and Kathikas Formations. Geological Magazine, 131,499-518. USTAOMER, T. & ROBERTSON, A.H.F. 1993. A late Paleozoic—early Mesozoic marginal basin along the active southern continental margin of Eurasia; evidence from the central Pontides (Turkey) and adjacent regions. In: KELLING, G., PARK, R.G. & ROWBOTHAM, G. (eds) Aspects of the Geology of Turkey. Wiley, Chichester, 219-238. UYEDA, S. ET AL. 1977. Some basic problems in the trench-arc-back arc system. In: TALWANI, M. (ed.) Island Arcs, Deep Sea Trenches and Back-Arc Basins. Geophysical Monograph, American Geophysical Union, 1, 1-14. UYEDA, S. & KANAMORI, H. 1979. Back-arc opening and the mode of subduction. Journal of Geophysical Research, B-84, 1049-1061. VAN DER LAAN, S.R., FLOWER, M.F.J. & KOSTER VAN GROOS, A.F. 1989. Experimental evidence for the origin of boninites; near-liquidus phase relations to 7.5 kbar. In: CRAWFORD, AJ. (ed.) Boninites. Allen & Unwin, London, 112-147. VAN KEKEN, P.E., KIEFER, B. & PEACOCK, S.M. 2002. High-resolution models of subduction zones; implications for mineral dehydration reactions and the transport of water into the deep mantle. Geochemistry, Geophysics, Geosystems (G3), 3(10), 22. WAKABAYASHI, J. & DILEK, Y. 1988. Ophiolites, subophiolitic soles, and the problem of fore-arc Ophio-

41

lites. EOS Transactions, American Geophysical Union, 69, 1450. WAKABAYASHI, J. & DILEK, Y. 2000. Spatial and temporal relationships between ophiolites and their metamorphic soles: a test of models of forearc ophiolite genesis. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 53-64. WAKABAYASHI, J. & UNRUH, J.R. 1995. Tectonic wedging, blueschist metamorphism, and exposure of blueschists; are they compatible? Geology, 23, 85-88. WALLIN, E.T. & METCALF, R.V. 1998. Supra-subduction zone ophiolite formed in an extensional forearc, Trinity terrane, Klamath Mountains, California. Journal of Geology, 106, 591-608. WANG, P.-L., Lo, C.-H., CHUNG, S.-L., LEE, T.-Y., LAN, C.-Y. & TRANG VAN, T. 2000. Onset timing of left-lateral movement along the Ailao Shan-Red River shear zone; 40Ar/39Ar dating constraint from the Nam Dinh area, northeastern Vietnam. Journal of Asian Earth Sciences, 18, 281-292. WORTEL, MJ.R. & SPAKMAN, W. 2000. Subduction and slab detachment in the Mediterranean-Carpathian region. Science, 290, 1910-1917. WYLLIE, P.J., BAKER, M.B. & WHITE, B.S. 1990. Experimental boundaries for the origin and evolution of carbonatites. Lithos, 26, 3-19. YUMUL, G.P. JR, DlMALANTA, C.B., FAUSTINO, D.V. &

DE JESUS, J.V. 1998. Translation and docking of an arc terrane; geological and geochemical evidence from the southern Zambales ophiolite complex, Philippines. Tectonophysics, 293, 255-272. YUMUL, G.P., ZHOU, M.F., TAMAYO, R.A., FAUSTINO, D.V. & GOTTEN, J. 2001. On-ramping of cold oceanic lithosphere in a forearc setting, the Southeast Bohol ophiolite complex, Central Philippines. International Geological Review, 43, 850-866.

This page intentionally left blank

Arc-trench rollback and forearc accretion: 2. A model template for ophiolites in Albania, Cyprus, and Oman YILDIRIM DILEK 1 & MARTIN F. J. FLOWER 2 1

Department of'Geology, Miami University, Oxford, OH 45056, USA (e-mail: [email protected]) ^Department of Earth and Environmental Sciences, University of Illinois at Chicago, Chicago, IL 60607, USA Abstract: Ophiolite assemblages record structural, magmatic, and metamorphic processes that preceded their entrapment in orogenic belts by continental plate collisions. Ophiolite genetic models appealing to 'oceanic' or 'suprasubduction' provenance are still unable to reconcile several basic problems, including: (1) the association of boninites with oceanic ridge-type structural settings; (2) the diachronous 'patch-like' distribution of ophiolites in orogenic belts; (3) disparate ages between and within their mantle and crustal sections; (4) the lack of evidence for 'obduction' at modern passive margins. In contrast, the proposal that Ophiolite genesis is exclusive to intra-oceanic forearc settings is compelling, given their uniquely shared structural, lithological, and stratigraphic attributes. Forearcs are interpreted to record discrete stages of subduction 'rollback' cycles, examples of which begin with subduction nucleation and the formation of boninitic 'proto-arcs', followed by arc splitting and concomitant retreat of the evolving arc-forearc complex. Forearc assemblages are likely to resist subduction to become entrapped in orogens, in contrast to denser, recently formed back-arc basin lithosphere, which is reconsumed by subduction following collision of the retreating forearc. As a model for NeoTethyan ophiolite genesis, this is predicated on the notion that rollback cycles are driven by ductile asthenosphere mobilized prior to and during collisions of Gondwana fragments with accreting Eurasia. It is also consistent with the apparent correlation of ophiolite ages with collisional events and their conjugate plate kinematic adjustments. Here, we use the slab rollback model as a template for interpreting the structural, magmatic, and metamorphic characteristics of well-studied Tethyan ophiolites, in Albania (Mirdita), Cyprus (Troodos), and Oman (Semail).

A longstanding debate persists concerning the roles of sea-floor spreading, subduction, and backarc basin opening in the genesis of ophiolites (Dilek et al. 2000). In the case of Tethyan ophiolites (Fig. 1) it is increasingly clear that dynamic processes triggering their formation are linked to the pattern of Gondwana break-up, evolution of subduction-accretion systems within and around Neo-Tethys, and the subsequent collisions of detached Gondwana fragments with accreting Eurasia (Dilek & Moores 1990; Gnos et al. 1997; Dilek 1999; Dilek et al 2000, and references therein). In this light, we believe that it is now possible to explain petrogenetic and mantle dynamic implications of these processes by reference to the continuing evolution of Western Pacific and Mediterranean marginal basins, which represent technically active remnants of NeoTethys (Flower & Dilek 2003). We contend that intra-oceanic subduction-zone environments provide a compelling analogy for most of the world's

ophiolites and that these features are formed in response to collision-related cycles of subduction nucleation, boninite-bearing 'proto-arc' formation, and back-arc basin opening that appear to be characteristic of late (pre-collision) stages of typical Wilson Cycles (Flower & Dilek 2003). Modern examples of subduction rollback appear to be associated with, and generally terminated by, collisions of their retreating slabs with approaching continental plates or other forearc complexes (e.g. Burchfiel & Royden 1991; Royden 1993a, 1993b; Jolivet & Patriat 1999; Jolivet et al. 1999). Following these collisions, recently formed, and relatively dense, back-arc lithosphere is reconsumed by subduction (back-arc basin collapse) whereas the forearcs themselves (because of their preponderant contents of hydrated, strongly refractory peridotite) resist subduction, to become entrapped in the 'jaws' of ensuing plate collisions, 'Subduction nucleation-rollback' cycles thus show numerous consistent features, shared by

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 43-68. 0305-8719/037$ 15 © The Geological Society of London 2003.

Fig. 1. (a) Distribution of major Neo-Tethyan ophiolites and suture zones in the Alpine-Himalayan orogenic system (data from Lippard et al. 1986; Dilek & Moores 1990). Numbers 1, 2 and 3 correspond to the Mirdita (Albania), Troodos (Cyprus), and Semail (Oman) ophiolites, respectively.

Fig. 1. (b) Distribution of the Neo-Tethyan ophiolites and major tectonic features in the eastern Mediterranean region (data from Dilek & Moores 1990; Robertson 2002). AC, Antalya Complex (Turkey); AO, Aladag (or Pozanti-Karsanti) ophiolite (Turkey); BHN, Beysehir-Hoyran Nappes (Turkey); IAESZ, Izmir-Ankara-Erzincan Suture Zone (Turkey); IPO, Intra-Pontide ophiolites; MO, Mersin ophiolite (Turkey); PO, Pindos ophiolite (Greece); VO, Vourinos ophiolite (Greece). —, transform fault plate boundaries (dashed where inferred); A A , Bitlis - Zagros continent-continent collision zone; A A A, oceanic trenches.

46

Y. DILEK & M. F. J. FLOWER

modern forearcs and remnant arcs (on the one hand), and ophiolites and their respective metamorphic 'soles' (on the other). The chief of these is the near-ubiquitous presence of refractory magma types such as boninites, high-Mg andesites (HMA), Na-rich granitoids (adakites), and thermally anomalous hydrothermal products, all of which serve as petrological signals of 'hot' subduction nucleation. In seeking a fundamental model of ophiolite genesis in the Tethyan collisional belt, we invoke the hypothesis that subduction rollback is triggered and propagated by lateral mantle flow ('extrusion') resulting from the approach of, and collision between, continental plates (Flower et al 1998, 2001; Flower & Dilek 2003). In this paper, we explore the implications of mantle-driven subduction rollback with respect to three well-studied Neo-Tethyan ophiolite associations (Fig. 1), Albania (Mirdita), in Cyprus (Troodos-Mamonia), and Oman (Semail), focusing on two specific aspects: (1) application of the model as a 'uniformitarian' template for interpreting generic features of ophiolite development; (2) extending it as a basis for reconciling ophiolite genesis and Neo-Tethyan plate kinematics. To better understand the petrotectonic evolution of the Mirdita, Semail, and Troodos ophiolites within the regional tectonic framework of Neo-Tethys, we first summarize the salient features of Tethyan ophiolite geology, and then discuss the NeoTethyan subduction rollback mechanisms in the late Cenozoic and the case for mantle-driven rollback and its implications for ophiolite genesis.

Overview of Tethyan ophiolite geology Initial rifting and precursory oceanic crust formation Tethyan oceans developed in response to successive episodes of the break-up of Gondwana and evolved as latitudinal, east-west-oriented basins separated by discrete continental fragments. The timing of their opening and their size and palaeogeography are still hotly debated (see Stampfli 2000 for a recent overview), although there is a general agreement that Jurassic to Cretaceous Alpine-Himalayan ophiolites and associated metamorphic rocks represent fragments of oceanic lithosphere generated during the closing stages of these Tethyan basins. The progression of ophiolite ages from Jurassic in the Alps, Apennines, and Dinarides-Hellenides to Cretaceous in the Taurides (Turkey), Zagros-Oman Mountains (Iran and Oman), and the Himalayas clearly reflects diachronous, zipper-like closure of the latitudinal Tethyan ocean basins from west to east. Whereas

ophiolites in the west contain island-arc tholeiite and boninitic rocks of incipient subduction-zone origin, the ophiolites in eastern Turkey and the Himalayas are spatially associated with calc-alkaline rocks of mature island-arc to Andean-type magmatic arc assemblages (i.e. the Baskil arc in SE Turkey and the Kohistan arc and the Lhasa terrane in the Himalayas; Fig. 1) suggesting prolonged periods of ocean-floor consumption and possibly larger ocean basins within the eastern Tethyan sectors. Most eastern Mediterranean ophiolites are spatially associated with extensional Triassic volcanic rocks that rest tectonically on passive margin sequences. These extrusive rocks display withinplate alkaline basalt (WPB) to transitional and mid-ocean ridge basalt (MORE) chemistry and are intercalated with hemipelagic to pelagic sedimentary rocks. They appear to represent riftrelated magmatic pulses along the northern periphery of Gondwana (Dilek & Rowland 1993, and references therein). Examples include the Triassic rift assemblages adjacent to the Mirdita ophiolites in Albania (Shallo et al. 1990), Triassic extrusive rocks associated with the western Hellenic ophiolites in Greece (Jones & Robertson 1991; Pe-Piper 1998; Saccani et al. 2003), Triassic volcanic rocks within the Antalya Complex in SW Turkey (Juteau 1980), and Upper Triassic volcanic and sedimentary rocks of the Dhiarizos Group in the Mamonia Complex south of the Troodos ophiolite (Malpas et al. 1993). The absence of hypabyssal oceanic rocks in these Triassic rift assemblages precludes a definitive assessment of whether rifting and associated magmatism in the Triassic was followed by fully developed sea-floor spreading and oceanic crust formation. Recent studies of these rift assemblages suggest, however, that Triassic MORE-type crust indeed characterize the early stages of the evolution of Neo-Tethyan basin development (Malpas et al. 1993; Saccani et al. 2003), the latest subduction of which might have been associated with the generation of suprasubduction-zone ophiolites in the eastern Mediterranean region.

Metamorphic soles and ophiolitic melanges Supporting evidence for the existence of Triassic oceanic crust as a precursor of the JurassicCretaceous Tethyan ophiolites comes from metamorphic soles commonly found structurally beneath these ophiolites. Up to several hundred metres in thickness, these metamorphic soles comprise imbricated thrust sheets showing metamorphic field gradients ranging from granuliteto upper amphibolite-facies rocks immediately beneath the ophiolitic peridotites to greenschist-

SLAB ROLLBACK AND OPHIOLITE GENESIS facies and lower-grade rocks structurally downsection and close to the passive margin sequences in the tectonic basement of the ophiolite complexes (Hacker & Mosenfelder 1996). The metamorphic protoliths are mainly of WPB- to MORB-type oceanic rocks at structurally higher levels, and sandstone, limestone and chert of proximal continental margin setting in structurally lower levels (Robertson 2002, and references therein). Although 40Ar/39Ar hornblende and mica cooling ages of the sole rocks indicate that metamorphism was coeval with and/or slightly younger than the igneous ages of the ophiolites in the upper plate (see Flower & Dilek 2003; Wakabayashi & Dilek 2003), it is likely that the sole rock protoliths are indeed Triassic in age. Melange units composed of blocks of oceanic rocks, platform carbonates, and metamorphic rocks occur beneath ophiolite complexes and their metamorphic soles throughout the eastern Mediterranean region and are part of the emplacement history of the Neo-Tethyan ophiolites in subduction-accretion systems. These melanges are commonly made of imbricate thrust sheets whose direction of tectonic transport is synthetic to the ophiolite emplacement direction. Oceanic rocks within the melanges include ophiolitic material, fragments of seamounts, and metamorphosed lavas (mainly sole rocks) commonly enveloped by a serpentinite matrix displaying ductile to cataclastic deformation features. Detrital material made of platform carbonate rocks ranges in size from centimetre-scale clasts to kilometre-scale blocks and commonly consists of neritic, distal deepwater limestone locally intercalated with sheared radiolarian chert, black metalliferous chert, and litharenites. These sedimentary rocks occur in a clay-rich argillaceous matrix showing a locally well-developed scaly fabric and represent distal parts of passive margin sequences, on which the Neo-Tethyan ophiolites were tectonically emplaced. The sedimentary melanges locally show a chaotic, olistostromal character, although locally well-preserved layering and stratigraphy are not uncommon. These sedimentary components of the ophiolitic melanges appear to have developed within flexural basins in front of advancing ophiolite nappes that received material from both the displaced ocean floor in the upper plate and the platform carbonates in the lower plate. Examples of emplacement-related melanges include the Avdella melange beneath the Jurassic Pindos ophiolite in Greece (Jones & Robertson 1991), the Lycian melange beneath the Cretaceous Lycian ophiolite nappes in SW Turkey (Collins & Robertson 1997), the Aladag melange beneath the Pozanti-Karsanti ophiolite and the melange unit underneath the Cretaceous Mersin ophiolite in the

47

central Tauride Mountains in Turkey (Polat et al 1996; Dilek et al. 1999), the coloured melange unit beneath the Cretaceous Neyriz ophiolite in southern Iran (Sarkarinejad 2003), and the melange unit underneath the Cretaceous Muslim Bagh ophiolite within the Zhob Valley Ophiolite Belt in the Pakistani Himalayas (Moores et al. 1980). These all represent subduction-accretion complexes developed during the collision of rollbackgenerated arc-forearc complexes with passive continental margins. Thus ophiolitic melanges signal the end of subduction rollback cycles, thereby corresponding to the death and resurrection phases of Shervais' ophiolitic life cycle (Shervais 2001).

Association of mature arc complexes and its implications Most Neo-Tethyan ophiolites show contemporaneous sedimentary covers that contain pelagic to hemipelagic sediments, terrigenous mud, and ironrich manganiferous deposits, and that lack volcaniclastic and tuffaceous sediments of mature arc type; commonly, they are not spatially and temporally associated with volcanic and plutonic rocks of mature island-arc complexes. These features have led some researchers to conclude that the Tethyan ophiolites probably formed in broad oceans and/or back-arc basins far from continental terrigenous input, and that the associated subduction of pre-existing ocean floor was either shortlived (e.g. Robertson et al. 1991) or highly oblique (e.g. Moores et al. 1984). Although such models favour a 'pre-arc' origin of suprasubduction-zone ophiolites within the Tethyan realm, there are several exceptions, where some Tethyan ophiolites are juxtaposed against coeval volcanoplutonic arc assemblages. The Eastern Hellenic ophiolites of the Vardar Zone origin (east of the Korabi-Pelagonian platform; Fig. Ib) are locally in tectonic contact with calc-alkaline volcanic and plutonic rocks of the coeval Late Jurassic Paikon island arc complex (Bebien et al. 1994; Robertson 2002). Similar intrusive and extrusive rocks of an arc origin (Chortiatis arc) occur along the western edge of the Serbo-Macedonian continental block (Eurasian affinity) farther to the east. The Jurassic Vardar Zone ophiolites (i.e. Guevgueli) and the volcanic arc complexes (Paikon and Chortiatis) are interpreted to have formed above a subduction zone dipping NNE (in present co-ordinate system), beneath the Eurasian continental margin. Extrusive rocks in these Vardar Zone ophiolites display chemical affinities ranging from MORB in the

48

Y. DILEK & M. F. J. FLOWER

east to island-arc tholeiite (IAT) towards the west close to the Paikon arc (Bebien et al. 1986). A similar occurrence of volcanic arc-ophiolite association is found in the central Pontide belt in northern Turkey farther east, where the Jurassic Kure ophiolite and the coeval Cangaldag volcanic arc are technically imbricated with the tectonostratigraphic units of the Eurasian continental margin (Fig. Ib; Yilmaz et al. 1991 \ Ustaomer & Robertson 1999). The Kike ophiolite shows a complete, although highly deformed and dismembered, pseudostratigraphy and MORE chemical affinity, whereas the £angaldag arc complex, built on an ophiolitic basement containing IAT to boninitic extrusive rocks, consists of tuffaceous sedimentary rocks and a c. 5 km thick volcanosedimentary unit consisting of porphyritic andesite, basaltic andesite, rhyolite, rhyodacite and dacite crosscut by small plagiogranite intrusions (Ustaomer & Robertson 1999). In both the Eastern Hellenic and central Pontide examples the contemporaneous evolution of the Jurassic ophiolites and mature volcanic arc complexes appears to be related to robust magmatism associated with subduction zones dipping north beneath Eurasia. Whether these inferred subduction zones were continuous (linked) in space and time from the Balkans to north-central Anatolia and whether the Jurassic ophiolites are of a Palaeo-Tethyan, rather than Neo-Tethyan, origin remain to be documented unequivocally through systematic structural field studies, isotopic investigations, and geochronological analyses. Mature arc volcanism as part of the NeoTethyan ocean basin evolution is well established in the SE Anatolian orogenic system in Turkey (Yilmaz 1993, and references therein). Here, the Cretaceous Ispendere, Guleman, and Komurhan ophiolites occur between the eastern extension of the Tauride carbonate platform (represented by the metamorphosed rocks of the Keban platform) and the Bitlis-Putiirge metamorphic massifs (a series of continental blocks) and represent the alongstrike continuation of the Troodos and Kizildag ophiolites that originated from the same NeoTethyan ocean basin (Fig. Ib). The IspendereKomurhan ophiolites consist of mafic-ultramafic cumulates, layered and isotropic gabbros, sheeted dykes, and extrusive rocks overlain by radiolarite and micritic limestone (Yazgan 1984; Yazgan & Chessex 1991); they are unconformably overlain by a flysch deposit with Upper Campanian-Lower Maastrichtian fossils (Michard et al. 1984; Yazgan 1984). These ophiolites include differentiated, felsic volcanic rocks (dacite, rhyolite) in their upper extrusive sequences and are cut by Coniacian-Campanian plutons composed of granitoids, quartz monzonites, monzodiorites, and gabbros.

Basaltic andesite, dacite and pyroclastic rocks constitute the extrusive equivalents of these Late Cretaceous intrusions and together they form the Baskil arc (Yazgan 1984). Plutonic rocks of the Baskil arc are observed to intrude the northern edge of the metamorphosed Keban platform rocks to the north (Yazgan & Chessex 1991), indicating that the construction of this Late Cretaceous arc complex occurred on an oceanic-ophiolitic basement (Ispendere-Komurhan-Guleman ophiolites) adjacent to the southern edge of the eastern Tauride platform (Dilek & Moores 1990). This arc complex was subsequently emplaced southwards onto the northern edge of a continental block (Bitlis-Puturge massif) via continent-trench collision in the early Tertiary. The lack of such a mature arc complex associated with the coeval ophiolites to the west (i.e. Troodos, Kizildag) and to the east (Neyriz, Kermanshah in Iran) of the Baskil arc and its ophiolitic basement suggests that there might have been a regional embayment within the southern Neo-Tethyan ocean basin, analogous to the modern Tyhrrenian Sea, the floor of which was consumed to produce the ophiolitearc duo by rapid arc-trench rollback and associated subduction magmatism. Paired ophiolite belts The Jurassic ophiolites to the west of Turkey occur in paired belts displaying distinct mantle sequences with different mineral assemblages, chemical compositions, and rare earth element (REE) and platinum group element patterns, reflecting a range of primary magmas and mantle source regions between MORB-like and refractory boninitic associations. Good examples include the Penninic zone ophiolites in the Alps and the Dinaride-Albanide-Hellenide ophiolites in the Balkans. Within the Mesozoic Neo-Tethyan remnants in the Eastern Alps, serpentinized harzburgites occur in association with metagabbro, metabasalt, radiolarite and ophicalcites in the lower Engandine, Tauern, and Rechnitz tectonic windows, whereas serpentinized Iherzolites are exposed in the Matrei Zone, and the Nauders and Reckner Complexes along the southern edge of the Lower Engadine (Melcher et al. 2002). The harzburgites record up to 20%, high-temperature partial melting and are interpreted as remnants of a depleted residual mantle, whereas the Iherzolites show low degrees of decompressional partial melting (0-10%) at lower temperatures and are interpreted to represent remnants of subcontinental (fertile) mantle exhumed during initial continental rifting (Melcher et al. 2002). Farther east in the Balkans, the two subparallel ophiolite belts of the Dinarides and the Vardar

SLAB ROLLBACK AND OPHIOLITE GENESIS Zone show a clear range in their mantle peridotite compositions and fertility. The mantle tectonites exposed in the western and central segments of the Dinaride and Vardar Zone ophiolites are made of fertile spinel Iherzolite, whereas those in the southern and eastern segments are composed of depleted harzburgites (Pamic et al. 2002). Similar to the ophiolites in the Eastern Alps, the Dinaric and Vardar Zone ophiolites appear to show a transition from earlier rifting-related oceanic crust towards subduction zone generated oceanic crust to the east. The range of peridotite fertility in the Dinarides and Vardar Zone ophiolites continues southwards into the ophiolite belts exposehd in the Albanides (Albania) and the Hellenides (Greece) where MORE-type, thin fossil oceanic crust in the west is juxtaposed against suprasubduction-zone (SSZ) type, much thicker fossil oceanic crust in the east without a major tectonic discontinuity (see the section on Mirdita ophiolites below). In the Southern Aegean a significant change in ophiolite age and mantle compositions occurs, whereby the NW-trending Jurassic ophiolite belts show a diachronous progression into the Cretaceous ophiolites in Turkey. Jurassic (c. 160 Ma) ophiolitic assemblages in Crete contain refractory Iherzolites intruded by calc-alkaline gabbros and plagiogranites (Koepke et al. 2002). Cretaceous (c. 90 Ma) ophiolites on the islands of Karpathos and Rhodes to the NE show strongly refractory serpentinized harzburgites that are intruded by calc-alkaline doleritic dykes (Koepke et al. 2002), resembling those Neo-Tethyan ophiolites within the Tauride belt in southern Turkey (Dilek et al. 1999). The zonation of Iherzolitic and harzburgitic peridotites in paired Alpine, Dinaride, Albanide, and Hellenide ophiolite belts in the Alps, Dinarides, and Hellenides dissipates in the South Aegean, where the ophiolites show a marked progression in age from Jurassic to Cretaceous. Instead, however, the Cretaceous eastern Mediterranean ophiolites show a similar transition in space and time from MORB-type to calc-alkaline affinities. According to our template model, this regional pattern may be reconciled with mantle flow effects interpreted to characterize the latest pre-collisional stages of Tethyan closure.

Neo-Tethyan subduction rollback: a precollision 'last gasp'? As the African plate approached Eurasia during the Mid-Tertiary, the inception of new subduction led to a proliferation of marginal basins, some intra-oceanic, others splitting young mountain belts asunder in the eastern Mediterranean region (Jolivet et al. 1999; Dilek et al. 2000). For

49

example, in apparent conjunction with the westward escape of Anatolia, the Hellenide-Tauride segment of the Alpine orogenic system was sundered by opening of the Aegean Sea and southward retreat of the Hellenic arc (Le Pichon et al. 2002). The Alboran basin probably opened in response to subduction rollback and counterclockwise rotation of the Corsica-Sardinia crustal blocks (Royden 1993a; Lonergan & White 1997; Morra et al. 1997; Padoa 1999) whereas the 'proto-Italian' Appeninic subduction system was rolled back during Tyrrhenian Sea opening, prior to its collision with the Apulian microplate (Lavecchia & Stoppa 1996; Doglioni et al. 1999; Wortel & Spakman 2000). Assuming that these basins opened in response to subduction rollback and associated crustal block rotations, protoophiolite crust developed accordingly in the Early Miocene-Present oceanic Liguria-Provengal Basin west of Corsica and Sardinia, and PliocenePresent Southern Tyrrhenian Sea (Jolivet et al. 1999). In eastern Tethys, the West Philippine, Celebes, and proto-South China Sea basins commenced opening during and after the collision of India and Asia (c. 45 Ma), and following the detachment of east Eurasian orogen slivers, the latter forming pieces of what are now Borneo, Sulawesi, and Sumatra (Hutchison 1990; Rangin et al. 1995, 1999; Pubellier et al. 1996). Oceanic lithosphere generated during successive basin opening episodes now forms the floor of most of the Philippine Sea and Sunda plates (Uyeda 1992; Bloomer et al. 1995), continued subduction-trench rollback being accommodated by sea-floor spreading in the Mariana Trough, Sumizu Rift, Okinawa Trough, and Marinduque Basin (Sarewitz & Lewis 1991; Taylor et al. 1991; Fryer 1996; Clift & Lee 1998; Sibuet et al. 1998). In each of these settings the sea-floor spreading produced new lithosphere within pre-existing oceanic domains by the initiation of new subduction, signalled by the formation of ultra-refractory 'proto-arc'. This model is uniquely able to explain the juxtaposition of oceanic lithologies of diverse ages and chemical affinity. According to this reconsumption (or 'collapse') model, newly formed back-arc basins play a key role in the emplacement of ophiolites, as evidenced by the presence of ophiolites in Borneo, Sulawesi, and the Philippines (Silver et al. 1983; Mitchell et al. 1986; Geary & Kay 1989; Encarnacion et al. 1993; Monnier et al. 1995; Leybourne et al. 1999; Yumul et al. 2001; Villeneuve et al. 2002), and tomographic images of detached oceanic slab fragments beneath these regions (Rangin et al. 1995, 1999). In accord with the tectonic facies hypothesis (Hsu 1994a, 1994b), we contend that modern Neo-Tethyan subduction rollback

50

Y. DILEK & M. F. J. FLOWER

episodes offer critical insights into the early stages of orogen development and, in turn, the geodynamic significance of ophiolites entrapped in older (Meso-, and Palaeo-) Tethyan collision sutures.

The case for mantle-driven rollback Flower et al. (1998, 2001) proposed that collisioninduced mantle extrusion successfully reconciles the timing of continental plate collisions, 'alkaline' (ultrapotassic) and 'intra-plate' (basaltic) magmatic activity, continental lithospheric 'escape', and the opening of marginal basins in the broader Tethyan system (see Briais et al. 1993; Jolivet & Faccenna 2000; Le Pichon et al. 2002). Mantle extrusion is supported by spatial-temporal correlations of the Africa-Eurasia, Arabia-Eurasia, India-Eurasia, and Australia-Sunda plate collisions with post-collisional lithospheric 'escape' kinematics, associated magmatic activity, and tomographic indications of perturbed uppermantle thermal structure (Flower et al. 1998, 2001; Jolivet & Faccenna 2000; Wortel & Spakman 2000). The model provides a mechanism for the lateral dispersion of thermally anomalous asthenosphere, evidence for which includes lobeshaped upper-mantle low-velocity regions beneath east and SE Asia, western Pacific basins, and the Mediterranean-Pannonian region (Spakman 1991; Flower et al. 1998; Rangin et al 1999; Wortel & Spakman 2000), the abundance of 'dispersed' post-collision basaltic volcanism (Hoang & Flower 1998), and flow-related mantle isotopic provinciality (Flower et al. 1998). Of particular importance in this model is the generation of 'protoophiolitic' crust in small ocean basins in the upper plate of retreating (rolling) subduction zones adjacent to collisional indentors. In the western Pacific and Mediterranean marginal basins, lithological products of subduction rollback processes include fragments of proto-arc, back-arc, and (in some cases) continental lithosphere fragments, juxtaposed together with remnants of MORE basement in which subduction was initiated. These lithospheric fragments are typically tectonized and show large age discrepancies within and between their crustal and mantle components. As these disparate fragments are progressively incorporated within evolving forearcs, the forearc lithosphere becomes become increasingly heterogeneous in its lithological makeup, whose lowest structural components match those of their respective remnant arcs (Bloomer et al. 1995; Hawkins & Castillo 1998). Where slab rollback is terminated by a collision as a result of the arrival of a continental fragment at the trench, newly formed back-arc lithosphere may be entirely reconsumed by intra-basinal subduction (a phe-

nomenon known as back-arc basin collapse), whereas accreted lithospheric fragments within forearcs become technically emplaced onto the collided margin as ophiolites. Here we suggest that subduction rollback is responsible for rapid evolution of suprasubduction-zone ophiolites in extended arc-forearc settings, and that the subduction rollback phenomenon itself is likely to be a geodynamic response to mantle extrusion driven by regional plate collisions. We believe that this interpretation can be tested against the Neo-Tethyan geological record, in particular that of the widespread and commonly well-exposed ophiolites. Many of the latter appear to record stratigraphic progressions from successions of typical 'oceanic' lithosphere (having MORB affinity) to a range of 'non-oceanic' lithologies including boninite and calc-alkaline series types. Here, as a preliminary attempt to validate the actualistic model, we review petrological, structural, and geodynamic evidence for three of the best preserved Neo-Tethyan ophiolite complexes, in Albania, Cyprus, and Oman.

A template for interpreting Tethyan ophiolites Albania (Mirdita) The Jurassic Neo-Tethyan ophiolites in Albania are collinear with those in the Dinarides and Hellenides (Fig. Ib), and occur in two subparallel zones (Shallo 1990a; Bortolotti et al. 1996; Dilek et al. 2001). The 'Western' and 'Eastern' ophiolite zones represent a northward continuation of the Pindos and Vourinos ophiolites (Greece), which are separated from the nearly coeval Vardar Zone ophiolites to the east by the Korabi-Pelagonian microcontinent (Robertson & Shallo 2000; Fig. 2). The Western ophiolites consist mainly of uppermantle peridotites overlain by plutonic rocks and massive to pillow lavas. Peridotites are composed of plagioclase Iherzolite, Iherzolite-harzburgite and rare dunite; the plutonic sequence includes troctolite, olivine gabbro, ferrogabbro, gabbro, and minor plagiogranite (Shallo et al. 1985). Volcanic rocks range in composition from basalt to basaltic andesite with a MORB affinity (Fig. 2; Bebien et al. 1997, 1998; Cortesogno et al. 1998; Bortolotti et al. 2002). Bajocian-Bathonian chert and radiolarites overlie these volcanic rocks. The Eastern ophiolites, up to 10 km in thickness, display a relatively complete Penrose-type ophiolite pseudostratigraphy, consisting of refractory harzburgite, harzburgite-dunite, overlain by layered gabbronorite, sheeted dykes, and volcanic rocks. The extrusive rocks include, in ascending stratigraphic order, basalt, basaltic-andesite, andesite, dacite,

SLAB ROLLBACK AND OPHIOLITE GENESIS

51

Fig. 2. (a) Distribution of Neo-Tethyan ophiolites in the Balkan Peninsula. The Mirdita ophiolites in Albania occur at a sharp northeasterly bend along the NW-trending Dinarides-Hellenide mountain belt. The Korabi-Pelagonian platform continues in a northwesterly direction between the Dinaride-Hellenide and Vardar Zone ophiolites. (b) Simplified geological map of the Albanian ophiolites showing the distribution of peridotite massifs and upper-crustal rocks (extrusive sequence and dyke intrusions only) (modified from the Geological Map of Albania, ISPGJ-IGIN 1983). Amphibolite and greenschist rocks of 'metamorphic soles' along the western and eastern peripheries of the ophiolite belt are also displayed. The dashed line depicts the inferred boundary between the Western (MORB) and Eastern (SSZ) ophiolites in the Mirdita Zone, (c) Spatial-temporal evolution of the Mirdita ophiolites inferred from the litho- and chemo-stratigraphy of three type sections. Approximate locations of Sections 1 (Kalur-Kushnen), 2 (Kimez-Tuc), 3 (near Shemri) are shown in (b). The crystallization order of major mineral phases and the chemical composition of the rocks show a progression from MORB-like oceanic basement in the west (Kalur-Kushnen) to calc-alkaline, boninitic proto-arc with an ultra-refractory mantle restite in the east (Shemri) through time. (See Fig. 5a, below, for tectonic interpretation of this progression.) Data and relevant interpretations are from Shallo (1992), Hoxha et al. (1993), Hoxha & Boullier (1995), Carosi et al. (1996), Manika et al (1997), Bebien et al (1997, 1998, 2000), Beccaluva et al. (1998), Insergueix-Filippi et al (2000), and Dimo-Lahitte et al. (2001).

52

Y. DILEK & M. F. J. FLOWER

and rhyodacite-rhyolite displaying an IAT chemistry (Fig. 2c; Shallo et al 1985; Shallo 1990a; Bortolotti et aL 2002). Locally, boninitic late-stage dykes crosscut the older dyke swarms in the sheeted dyke complex and boninitic lavas occur in the extrusive sequence (Fig. 2c; Shallo 1990b). The spatial and temporal relations between the Western and Eastern ophiolites are complex and significantly modified by crustal shortening associated with the oblique collision of Apulia with the Korabi-Pelagonian microcontinent in the early Cenozoic. However, original igneous relations are locally well preserved. Quartz diorites of the Eastern ophiolites are observed, for example, to have been intrusive into the Western peridotites in several locations within the Mirdita Zone (Shallo et al. 1996). Plagioclase peridotites and troctolites of the Western ophiolites (with MORE affinity) are locally in primary igneous contact with the harzburgite-dunite mantle sequences of the Eastern ophiolites (with IAT affinity), and both ophiolite types are locally overlain by IAT and boninitic extrusives (Bortolotti et aL 1996; Shallo et al. 1987; Bebien et al. 1998). These relations suggest, in general, that there is a gradational change from MORB-type Western ophiolites to subduction zone related Eastern ophiolites from west to east and structurally upwards within the Mirdita ophiolite belt in northern Albania (Dilek et al. 2001). The spatial-temporal distribution of Albanian ophiolite lithologies is consistent with predictions of the arc-trench rollback model (Flower & Dilek 2003). This is illustrated in Figure 2c, by reference to sections at (1) Kalur-Kushnen, (2) Kimez-Tuc,, and (3) Shemri, which we interpret to represent a progression from MORB-like (olivineand plagioclase-phyric) oceanic basement (section 1) to a boninitic to calc-alkaline proto-arc, following a subduction initiation event. In accord with the rollback model, the proto-arc sections (2 and 3) consist respectively of eruptive and hypabyssal crustal sequences, and sheeted dykes, boninitic to calc-alkaline plutonic rocks, and ultra-refractory mantle restite. The Greek ophiolites farther south in the western Hellenides show similar lithological, structural, and geochemical features (Spray et al. 1984). The Pindos ophiolite west of the Meso-Hellenic Trough (Fig. 2a) includes harzburgitic peridotites cut by pyroxenite and gabbro dykes and overlain by layered cumulates (plagioclase dunite, troctolite, and gabbro), which display MORB chemistry (Capedri et al. 1980). These cumulate rocks are locally intruded by boninitic dykes (Rassios & Smith 2000). In the Asropotamos submassif within the Pindos ophiolite the lavas and sheeted dykes exhibit an evolution from N-MORB (normal MORB) to MORB-IAT, then IAT, and finally to

boninite-series rock units that crosscut pre-existing ophiolitic lithologies (Beccaluva et al. 1984). The Vourinos ophiolite east of the Meso-Hellenic Trough (Fig. 2a) consists mainly of refractory harzburgite with pods and lenses of dunite and chromite, overlain by mafic-ultramafic cumulate rocks, a plutonic sequence composed of gabbro, diorite and plagiogranite (c. 1 km thick), a sheeted dyke complex (c. 1.5km thick), and extrusive rocks made of pillow lava screens and massive lava flows interlayered with metalliferous sedimentary deposits (Rassios & Smith 2000, and references therein). Both the refractory harzburgite and crustal units (gabbros, dykes, lavas) display chemical signatures of an IAT affinity; the extrusive rocks are locally cut by boninite-type dykes (Beccaluva et al. 1984). Thus, similar to the Mirdita ophiolites in Albania, the Greek ophiolites appear to show a spatial and temporal transition from MORB to IAT affinities from west to east across the western Hellenides. Regional geological constraints suggest that the Mirdita and western Hellenic ophiolites were derived from the Pindos-Mirdita ocean basin located west of the Korabi-Pelagonian microcontinent (Shallo & Dilek 2003). The role of 'hot' subduction initiation is further supported by the thermal and petrological character of metamorphic soles, which underlie both Eastern and Western ophiolite assemblages, separating serpentinized upper-mantle rocks from underlying volcano-sedimentary sequences (DimoLahitte et al. 2001). The upper sole contacts are sharp and display metamorphic fabrics parallel to the foliation in the overlying peridotites, whereas contacts with the underlying volcanosedimentary series are more diffuse. Several genetic models have been advanced to explain observed relationships between the Eastern and Western ophiolites and the significance of their respective metamorphic soles. One view contends that the two groups reflect contrasting mid-ocean ridge and suprasubduction-zone environments (Shallo 1992; Beccaluva et al. 1994). Other models ascribe their differences to the 'rhythmic' evolution of a slowspreading ocean ridge axis (Nicolas et al. 1999) or to eastward transgression of the latter to a fasterspreading regime (Kodra et al. 1993, 2002). However, none of these models successfully accommodates the presence of boninite and its significance regarding the ophiolite generation. More recent studies have attempted to link boninites to the influence of transform faulting (Tashko 1996; Vergely et al. 1997), in particular regarding possible subduction nucleation within the Pindos-Mirdita basin. Dimo-Lahitte et al. (2001) argued that the MORB-like Eastern metamorphic sole represents

SLAB ROLLBACK AND OPHIOLITE GENESIS the upper part of newly subducted lithosphere, metamorphosed at granulite- and eclogite-facies conditions between c. 30 and 45 km depth (Gjata et al. 1992), and concluded that west-dipping subduction was terminated with the arrival of a colder, denser lithosphere (Beaumont et al. 1996). Western metamorphic soles show MORB-like trace element and isotopic affinities (Gjata et al. 1992) whereas 40Ar/39Ar age data suggest that metamorphic soles and their contiguous ophiolites young northwards from c. 174 to 160 Ma (Vergely et al. 1997; Dimo-Lahitte et al. 2001). Together, thermobarometric and age data from metamorphic soles record a rapid 'counter-clockwise' P-T-t path (c. 1200 °C, 15 kbar to 960 °C, 8 kbar) at c. 166 ±2 Ma or later, suggesting that prior to their exhumation, oceanic lithosphere had been subducted to depths of at least c. 40-45 km (Gjata et al. 1992; Dimo-Lahitte et al. 2001). Sea-floor spreading within the Pindos-Mirdita basin was interrupted by new subduction initiation possibly at a ridge transform, leading to its collapse. According to this scenario, the MORBlike Western ophiolites represent the preserved remnants of the rifting-related Pindos-Mirdita ocean floor remaining after the inception of westdipping subduction and the ensuing development of the near-contemporaneous (c. 162-174 Ma) Eastern ophiolites during a brief interval of 'proto-arc'-forearc splitting as a result of slab rollback (see Fig. 5a, below). The collision of a boninitic 'proto-arc' (or retreating arc-forearc complex) with the Korabi-Pelagonian microcontinent must have occurred shortly thereafter, resulting in the eastward emplacement of the Mirdita ophiolites. Uplift, fragmentation, and partial burial of the proto-ophiolitic forearc beneath flysch (Shallo 1990c, 1992) were succeeded in the Eocene-Oligocene by an oblique collision between Korabi-Pelagonia and Apulia. Cyprus (Troodos, Mamonia) On Cyprus, three major ophiolitic terranes that are intimately related in space and time regarding their evolution are tectonically juxtaposed: the Troodos ophiolite sensu stricto north of the east-west-trending Arakapas fault zone, the Limassol Forest Complex to the south, and the Mamonia Complex to the SW of Troodos (Fig. 3a). The c. 1 km wide Arakapas fault zone contains coarse-grained clastic sedimentary rocks, fine-grained hydrothermally derived sedimentary rocks, and submarine lava flows resting on cataclastically deformed oceanic rocks; it has been interpreted as an oceanic transform fault that offset the Troodos spreading axis during the mid-Cretaceous (Simonian & Gass 1978; Ma-

53

cLeod & Murton 1993). The Limassol Forest Complex south of the Arakapas fault zone includes serpentinized peridotites in tectonic contact with gabbros, ultramafic cumulates, sheeted dykes, and volcanic rocks, and displays gneissic fabric and vertical shear zones (Fig. 3a; Murton 1990; MacLeod & Murton 1993). Locally in the Western Limassol Forest Complex late-stage small gabbroic and wehrlitic plutons and picritic, basaltic dykes crosscut the deformation fabric in the serpentinized peridotites (Murton 1990). Low-angle normal faults and detachment surfaces are common in the Eastern Limassol Forest, suggesting widespread crustal extension in this fossil oceanic lithosphere (MacLeod & Murton 1993). The Troodos ophiolite sensu stricto north of the Arakapas fault zone displays a dome structure with a central core of serpentinized peridotites overlain by layered to isotropic gabbros, sheeted dykes, and extrusive rocks (Fig. 3a). Both sheeted dykes and volcanic rocks are cut by numerous normal faults, some of which contain veins of epidote, quartz, and pyrite of hydrothermal origin. The sheeted dyke complex includes several structural grabens (Fig. 3a) defined by blocks of dykes and extensional faults dipping towards each other, and locally (i.e. in the Solea graben) highly rotated sheeted dykes are separated from the underlying gabbros by low-angle detachment faults. These structures collectively indicate a seafloor spreading origin of the Troodos ophiolite sensu stricto. The Mamonia Complex to the SW is separated from the Troodos ophiolite sensu stricto by thin slivers of amphibolite and greenschist along complex thrust and strike-slip faults (Fig. 3a; Malpas et al. 1993). Locally, the L-S mylonitic fabric in serpentinized peridotites is parallel to the foliation and mineral lineation in the adjacent amphibolite blocks (Malpas et al. 1993), reminiscent of metamorphic soles in other Neo-Tethyan ophiolites. Some of these amphibolites are dated at 9390 Ma (Spray & Roddick 1981), nearly identical to the zircon ages (92-90 Ma) from plagiogranites in the Troodos ophiolite sensu stricto (Mukasa & Ludden 1987). The Mamonia Complex includes blocks and thrust sheets of Upper Triassic volcanic rocks displaying a dominant MORB affinity with subordinate WPB-type volcanic and Upper Triassic-Lower Cretaceous sedimentary rocks characteristic of passive margin sequences (Malpas et al. 1993, and references therein). Based on the chemostratigraphy of the volcanic rocks (tholeiitic Dhiarizos Group) and the nature of complex tectonic contact relations between the Troodos ophiolite sensu stricto and the Mamonia Complex, Malpas et al. (1993) suggested that a large tract of

54

Y. DILEK & M. F. J. FLOWER

Fig. 3. (a) Simplified geological map of Cyprus showing the Troodos ophiolite, Limassol Forest Complex, Mamonia Complex, Moni Melange, Arakapas Fault, and Kyrenia Range (undifferentiated). Data are from Murton (1990), Dilek & Eddy (1992), and Robertson (2002). Extrusive rocks in the Troodos ophiolite include the Lower and Upper Pillow Lavas. CY1/1A and CY2/2A refer to the drill sites of the Cyprus Crustal Study Project. Numbers 1, 2, and 3 mark the type sections (Akaki Canyon, Ayios Onouphrios-Potamos, and Margi-Kataliondas) within the extrusive sequence along the northern edge of the Troodos ophiolite where the composite lava chemical stratigraphy has been examined, (b) Spatial-temporal evolution of the Troodos extrusive sequence inferred from the litho- and chemo-stratigraphy of three type sections. LPL, high-Ti Lower Pillow Lava series (both primitive and evolved types); UPL, low-Ti Upper Pillow Lava series. (See text for discussion and refer to Flower & Dilek (2003, fig. 1 and related text) for further explanation.) Data and relevant interpretations are from Malpas & Langdon (1984), Moores et al. (1984), Rautenschlein et al (1985), Bednarz et al. (1987a, 1987b), Bednarz & Schmincke (1994), Gibson et al. (1987), Murton (1989), Rogers et al. (1989), Taylor (1990), Allerton & Vine (1991), Taylor et al (1992), Portnyagin et al. (1996), and Schouten & Kelemen (2002).

SLAB ROLLBACK AND OPHIOLITE GENESIS Triassic ocean floor was subducted beneath the leading edge of the 'Troodos microplate'. Field mapping in and petrological studies of drill cores from the plutonic complex in the Troodos ophiolite sensu stricto have shown the existence of two suites of gabbroic rocks (Malpas et al 1989; Thy et al 1989). The lower suite is compositionally and mineralogically similar to the underlying ultramafic cumulates and is intrusive into the upper suite, which is compositionally similar to the sheeted dyke complex. Thy & Esbensen (1993) suggested that this phenomenon could be explained by a two-stage evolution of the ophiolite. The sea-floor spreading generated sequence of the Troodos ophiolite includes the upper plutonic suite, sheeted dykes and the lower pillow lavas (LPL), and may have formed at an oceanic spreading centre. This oceanic lithosphere was subsequently underplated by magma chamber(s) of basaltic andesite compositions that produced the mafic-ultramafic cumulates and depleted upper pillow lavas (UPL). These higher-level extrusive rocks also include high-magnesian andesites (boninitic-type) (Robinson et al. 1983). Dykes that feed into these upper lavas have steep dips, and they crosscut the earlier-formed sheeted dykes (Thy & Esbensen 1993). Extensive epidosite formation and hydrothermal alteration within the sheeted dyke complex were probably related to this second-stage magmatism underplating the already extended, pre-existing Troodos oceanic crust. The dominantly harzburgitic mantle rocks in the Troodos ophiolite are the residuum after the formation of the highly refractory magmas of this late magmatic event, whereas the residuum after the first melting event was probably Iherzolitic (Thy & Esbensen 1993). This interpretation was substantiated by recent findings of Batanova & Sobolev (2000), documenting that the Troodos mantle sequences include two distinct peridotite types. The first type comprises spinel Iherzolite, with spinels showing Cr numbers (Cr/(Cr -f Al)) of 0.22-0.28, containing veins of dunite and clinopyroxene-bearing harzburgite, and the second group consisting mainly of refractory clinopyroxene-poor harzburgite, with spinel Cr numbers of 0.51-0.70. The spinel Iherzolites were interpreted as residua to the extraction of MORE melts and, although not involved in the production of calcalkaline or boninitic magmas, were probably modified by the latter, as evidenced by veins of dunite and harzburgite (Batanova & Sobolev 2000). The oldest exposures in the Troodos-Mamonia duo are probably relict fragments of the 'proto-' Cyprus basin, bounded to the north by a carbonate platform, now preserved in the Kyrenia Range (Fig. 3a), and to the west and SW by tectonized

55

MORB-like crust and pelagic sedimentary rocks, part of the present-day Mamonia Complex (Robertson et al. 1991; Eaton & Robertson 1993). Whereas the latter has been interpreted as a deformed passive margin sequence sutured to Troodos during the Late Cretaceous (Bailey et al. 2000), the MORB-like compositional character and thermal history of the Mamonia Complex and underlying Ayia Varvara Formation resemble those of metamorphic soles typically seen beneath the Neo-Tethyan ophiolites. Detailed mapping and microstructural studies in Mamonia suggest that an episode of prograde metamorphism (c. 9088 Ma) was followed by coeval retrograde hydration and dextral transtension, and serpentinite diapirism (c. 83-73 Ma), prior to final exhumation of the complex at c. 65 Ma (Malpas et al. 1992). Contacts between the Troodos and Mamonia complexes, and the subjacent Moni melange, are dominated by thrust faulting and associated shearing (Robertson et al. 1991; Eaton & Robertson 1993) while both bodies underwent coeval counter-clockwise rotation (Morris et al. 1998). Taken together, spatial-temporal relationships of the Troodos and Mamonia complexes also support the rollback model (Flower & Dilek 2003). As shown in Figure 3b, Troodos eruptive sections may be interpreted to reflect the diachronous inception of calc-alkaline to boninitic protoarc magmatism, as sea-floor spreading gives way to the inferred tectonic and petrological responses to subduction initiation. In the Akaki Canyon and Ayios Onouphrios Potamos sections, olivine- and olivine -f plagioclase-phyric, relatively high-Ti lavas (with MORB-like abundances of primitive to evolved melts) of the Lower Pillow Lava series (LPL-HTS: high-Ti series) give way upwards to aphyric and olivine-phyric, low-Ti calc-alkaline lavas of the Upper Pillow Lava series, whereas the Margi-Kataliondas section reflects a relatively sharp transition from the primitive to evolved Lower Pillow Lava series to ultra low-Ti and ultra-depleted, high-Ca boninitic lavas. Our interpretation is strongly reinforced by the association of these ultra-depleted high-Ca boninitic extrusive rocks with transform faulting in the Limassol Forest area to the south, and by the transgression from Iherzolitic to ultra-refractory harzburgitic mantle sequences (Batanova & Sobolev 2000). Evidence for subduction nucleation in southern and southeastern Cyprus includes the accretionary wedge lithologies in the Limassol Forest area (Eaton & Robertson 1993), and high-Ca boninites associated with the Arakapas fault (e.g. Simonian & Gass 1978; Flower & Levine 1987; Rogers et al. 1989; MacLeod & Murton 1993) and uppermost Mamonia complex, and the underplating of subduction-generated, second-stage magmas be-

56

Y. DILEK & M. F. J. FLOWER

neath the Troodos 'proto oceanic crust'. Cann et al. (2001) observed similarities between extensional detachment faults in the Limassol Forest area, immediately south of the Arakapas Fault, and those in the 'inside corners' of modern oceanridge-transform intersections. At the Mid-Atlantic Ridge, exposed detachment surfaces are corrugated in the direction of spreading and include peridotite and gabbro core complexes. Analogous complexes in the Limassol Forest show the effects of exposure to seawater, whereas associated dyke swarm orientations are consistent with the effects of shearing in newly formed oceanic crust (Borradaile 2001). Given that ophiolitic lithosphere is less extended north of the Arakapas Fault (i.e. representing a section of 'outside corner' crust), Cann et al. (2001) concluded that the Troodos ophiolite (in its current orientation) formed to the east of a 'ridge-transform-ridge' intersection, consisting of spreading segments linked by a dextral transform. According to this interpretation, the Limassol Forest block would have been formed in the western 'inside corner' of the ridge-transform intersect, having spread eastward until it passed a second spreading axis. At this point magmatic crust north of the Arakapas Fault would have become welded to the Limassol Forest block as the fault evolved as a transform fracture zone. Although this scenario is supported by dyke orientations, and magmatic and mantle flow fields inferred from rock fabric studies (Abelson et al. 2001; Borradaile 2001; Borradaile and Lagroix 2001), it does not take account of the boninite activity, evident both north and south of the Arakapas Fault, which suggests that a subduction nucleation event was in progress. Palaeomagnetic studies indicate that Troodos underwent significant differential rotation, consistent with pre-emplacement clockwise rollback effects (e.g. juxtaposing Mamonia and the Arakapas fracture zone along arcuate, strike-slip faults) and those associated with the cessation of subduction (Morris et al.

1998; compare McCabe & Uyeda 1983). Thus, structural, kinematic, and petrological attributes of the Troodos and Mamonia complexes concur closely with those predicted for mantle-driven subduction rollback. Further slab rollback was prevented by a collision between Cyprus and the Eratosthenes Seamount, a north-moving rifted fragment of continental Africa (Robertson 2000). Oman (Semail) The Semail ophiolite is commonly considered to be the most typical of 'oceanic' ophiolites. However, although oceanic lithological and structural attributes do indeed dominate, there is clear evidence here, as in most other Neo-Tethyan ophiolites, that the exposed lithologies record the demise of sea-floor spreading and the advent of a new, 'hot' subduction episode. The Semail ophiolite occurs in a c. 600 km long, up to 150 km wide thrust sheet in the Oman Mountains in the SE part of the Arabian Peninsula (Fig. 4a) and rests tectonically on a discontinuously exposed melange consisting of blocks of amphibolite and greenschist rocks in a serpentinite matrix (Searle & Cox 1999). Locally, the metamorphic sole and/or the serpentinite matrix melange tectonically overlie the Haybi sedimentary melange, which includes Triassic volcanic rocks (seamounts of ocean-island basalt) and reefal limestone blocks. These tectonostratigraphic units are thrust upon the Mesozoic passive margin sequences of the Arabian continent. The U-Pb zircon ages from plagiogranite rocks in the ophiolite range from 97.3 to 93.5 ± 0.25 Ma (Tilton et al. 1981) representing the crystallization ages of the late-stage differentiates in the Semail fossil oceanic crust. The 40Ar/ 39 Ar hornblende ages from the metamorphic sole beneath the ophiolite range from 95.7 to 92.6 ± 0.6 Ma (Hacker et al. 1996), indicating that the displacement of the Cretaceous oceanic crust from its igneous spreading environment occurred within several million years of its formation.

Fig. 4. (a) Distribution of the Neo-Tethyan ophiolites and the location of the Semail ophiolite in the Gulf region east of the Arabian Peninsula, (b) Simplified geological map showing the various massifs in the Semail ophiolite (Oman) and the occurrence of the metamorphic sole rocks beneath the ophiolite (data from Lippard et al. 1986; Searle & Cox 1999). (c) Spatial-temporal evolution of the Semail ophiolite inferred from the litho- and chemo-stratigraphy of three type sections, depicted in (b). Crystallization order of mineral phases in the extrusive rocks changes from those typical of MORE (olivine + plagioclase) to calc-alkaline (clinopyroxene + olivine) affinity upsection and towards the NW within the ophiolite. More detailed information on the geochemistry of the Semail extrusive sequence has been given by Umino et al. (1990) and Ishikawa et al. (2002). (See Fig. 5c, below, for a tectonic interpretation of this geochemical evolution.) Data and relevant interpretations are from Searle & Malpas (1980, 1982), Pallister (1981), Pallister & Hopson (1981), Pearce et al (1981), Alabaster et al. (1982), Browning (1984), Benn et al. (1988), Ernewein et al. (1988), Juteau et al. (1988a, 1988b), Nicolas et al. (1988a, 1988b, 1994, 1996), Umino et al. (1990), Nicolas & Boudier (1991, 1995), Umino (1995); Lachize et al. (1996), Schiano et al. (1997), Jousselin & Mainprice (1998), Jousselin et al (1998), Searle & Cox (1999), and Ishikawa et al (2002).

SLAB ROLLBACK AND OPHIOLITE GENESIS

57

58

Y. DILEK & M. F. J. FLOWER

Petrological and structural features of the Semail ophiolite resemble those of lithosphere formed at fast-spreading ridges (Pallister 1981; Pallister & Hopson 1981; Pearce et al 1981; Alabaster et al. 1982; Umino et al. 1990; Nicolas et al. 1994, 1996; Nicolas & Boudier 1995; Umino 1995) as exemplified by MORB-like compositions of some of its volcanic rocks (Fig. 4c; Pearce et al. 1981; Alabaster et al. 1982; Ernewein et al. 1988; Umino et al. 1990), sheeted dykes (Nicolas & Boudier 1991), and layered plagioclase-rich gabbros (Pallister & Hopson 1981; Browning 1984; Benn et al. 1988; Ernewein et al. 1988; Juteau et al. 1988a, 1988b; Nicolas et al. 1988a, 1988b). Jousselin & Mainprice (1998), Jousselin et al. (1998), and Nicolas & Boudier (2000) have demonstrated the presence of fossil sub-axial mantle diapirs beneath a palaeo-spreading centre that are marked by a thick Moho transition zone and abundant wehrlitic intrusions in the lower crust. The Geotimes volcanic unit in the stratigraphically lower levels in the extrusive sequence is made of low-K tholeiitic basalts with slightly depleted chondrite-normalized REE patterns that are enriched in the large ion lithophile elements, Sr, Rb, K, Ba and Th, and depleted in the compatible elements (Alabaster et al. 1982; Umino et al. 1990). The majority of the sheeted dykes are associated with these extrusive rocks, and most of the layered cumulate gabbros, display mineral assemblages similar to those in the Geotimes volcanic unit, suggesting their common parental magmas. Intrusions of ultramafic and calc-alkaline magmas into MORB-like basement plutonic rocks (Benn et al. 1988; Ernewein et al. 1988; Juteau et al. 1988a, 1988b), the eruption of boninites and calc-alkaline tholeiites (Alabaster et al. 1982; Ernewein et al. 1988; Umino et al. 1990), and subsequent intrusion of calc-alkaline plutons, as in the Haylayn Massif and Sumail nappe regions, mark a second magmatic stage in the development of the Semail ophiolite (Pallister & Hopson 1981; Pallister 1984; Benn et al. 1988; Juteau et al. 1988a, 1988b; Lachize et al. 1996; Schiano et al. 1997). These intrusive rocks form stocks and dykes and are mostly (although not generally) discordant with older MORB-like layered rocks (Benn et al. 1988; Juteau et al. 1988a, 1988b; Umino et al. 1990; Umino 1995). They consist of a suite of dunite, wehrlite, pyroxenite, troctolite, olivine gabbro, and diorite cumulates, and a series of Iherzolite, gabbro-norite, two-pyroxene diorite, and trondhjemite (Umino et al. 1990). Previously interpreted as remobilized melt-cumulate emulsions (Pallister & Hopson 1981), these lithologies appear to represent refractory magma series that

originated from boninitic parental magmas (Umino et al. 1990; Schiano et al. 1997). The intrusive rocks of this stage include a Cpx series of dunite, wehrlite, olivine gabbro, and diorite, and an Opx series of Iherzolite, gabbronorite, two-pyroxene diorite, and trondhjemite (Shervais 2001). These plutonic rocks correlate with olivine + cpx-phyric lavas of the Lasail and opx-phyric lavas of the Alley units (Alabaster et al. 1982; Umino et al. 1990; Shervais 2001). More evolved calc-alkaline rocks (dacite, rhyolite) in the Alley volcanic unit overlie these Lasail volcanic rocks (Fig. 4c; Pearce et al. 1981; Alabaster et al. 1982; Ernewein et al. 1988; Umino et al. 1990). Boninites also appear as lavas and dykes in the Alley volcanic sequence (Umino et al 1990; Ishikawa et al. 2002) unconformably overlying and in some cases cross-cutting the Geotimes unit (Fig. 4c). Plutonic equivalents of these evolved volcanic rocks include dykes, stocks and plutons of gabbro, diorite and trondhjemite rocks intruding into the older cumulate gabbros and the sheeted dyke complex. Collectively, these late-stage volcanic and plutonic rocks represent calc-alkaline island-arc magmatism. The most recent Semail activity produced alkali basalts of the Salahi volcanic unit between c. 85 and 90 Ma (Umino et al. 1990; Lachize et al. 1996). These alkali basalts are separated from the Alley volcanic unit by a thin horizon of pelagic sedimentary rocks (Ernewein et al. 1988), indicating a period of quiescence between the last eruption of suprasubduction-zone lavas and the Salahi volcanic rocks. Diabasic dykes with similar compositions crosscut the ophiolitic units and the underlying metamorphic sole, analogous to the late-stage dyke intrusions in the Tauride ophiolites in Turkey (Dilek et al. 1999), and together with the Salahi volcanic rocks they represent off-axis magmatism fed by melts that possibly originated from an asthenospheric window beneath the displaced oceanic lithosphere in the upper plate. The above relationships, summarized in Figure 4c, are virtually identical to those observed in the Albanian and Cyprus ophiolites, although it should be emphasized that the relative proportion of oceanic lithosphere is substantially greater in Semail. Typical 'fast-spreading' oceanic lithosphere is seen in the oldest, southeasternmost section (the Nakhl-Rustaq-Maqsad Massifs). However, to the NW, younger sections increasingly reflect the presence of boninitic and calcalkaline lithologies and ultra-refractory harzburgitic mantle. As commonly observed elsewhere, the Semail metamorphic sole is MORB-like in character (Searle & Malpas 1980, 1982; Searle & Cox 1999) and post-dates the ophiolite by <2 Ma (Hacker & Gnos 1997). Metamorphic rocks in the

SLAB ROLLBACK AND OPHIOLITE GENESIS sole include amphibolite, garnet amphibolite, and granulite, displaying high geothermal gradients with counter-clockwise P-T-t paths (Gnos 1998). The inferred P-T-t path of the sole rocks shows peak temperatures of 775-875 °C and pressures of c. 1.1 GPa, implying rapid, although shortlived, subduction. The inferred P-T-t paths for sub-ophiolitic continental rocks suggest that the Arabian passive margin, a potential source for two-mica, garnet-, tourmaline-, cordierite- and andalusite-bearing granites, was also subducted to >70 km depth (Searle & Cox 1999), following the collision of the Arabian passive margin with the trench around 85 Ma (Yanai et al 1990).

Discussion Brief reviews of three relatively well-preserved Neo-Tethyan ophiolites provide a basis for evaluating the rollback model in terms of the forearc analogue, and predicted features such as an oceanic basement component, evidence for subduction nucleation, and spatial-temporal correlations with far-field plate tectonic effects. At the outset, we suggest that the array of petrological and structural features in these ophiolites is remarkably consistent with those associated with modern forearcs. Even more significantly, the ophiolites lack features that preclude the forearc analogy upon which the slab rollback model is crucially dependent. In each of our examples, MORB-like basement lithologies (along with some 'transitional' and calc-alkaline compositions) are characterized by the structural attributes of sea-floor spreading: sheeted dykes with 'one-way' chilling, 'fossil' transforms, and imbricated plutonic and pillow lava sequences, albeit in differing proportions. The Semail ophiolite, moreover, offers a spectacular 'window' to sub-axial magma supply. In the Mirdita and Troodos ophiolites, it is hard, if not impossible, to distinguish 'true' oceanic from marginal basin provenance on the basis of petrological and geochemical criteria, although regional constraints suggest that the Pindos-Mirdita and Cyprus basins were initiated at continental margins rather than as exclusively intraoceanic basins. In contrast, given its overall geotectonic context, the bulk of the Semail ophiolite may be reasonably ascribed to a 'major' spreading axis (e.g. Godard et al. 2000) although an intra-oceanic basin provenance cannot be excluded (Pearce et al. 1981). Field relationships in each of the ophiolites show that MORB-like lithologies invariably predate the emplacement of boninitic and calc-alkaline magmas. The evidence for subduction nucleation (boninite melts segregated at temperatures of at least 1260°C at depths of less than c. 30km,

59

and the inflected P-T-t trajectories of MORBlike metamorphic soles) is compelling, as is the association of these features with axial transform faulting (e.g. Cann et al. 2001). Although the evidence for subduction initiation in Semail has been downplayed (probably considering the preponderance of 'oceanic' lithosphere) and even explained by anomalous, ad hoc hydrothermal effects (Benoit et al. 1999), the presence of boninite is now established beyond doubt in the Alley volcanics and dyke swarms, unconformably post-dating MORB-like Geotimes lavas and dykes (Umino et al. 1990; Ishikawa et al. 2002). As in the Troodos and Mirdita ophiolites, the Semail metamorphic sole is a near-contemporaneous feature whose inflected P-T-t trajectory is consistent with rapid, 'hot' subduction beneath very young crust (Hacker & Gnos 1997). Based on the examples considered here, structural and petrological relationships, especially those concerning boninites and metamorphic sole P-T-t histories, accord in every respect with those observed and interpreted for the inception of modern slab rollback cycles (Fig. 5). It is clear that modern rollback cycles are initiated in both intra-oceanic and intra-continental settings and show variably unconstrained (e.g. Mediterranean compared with western Pacific) life histories. In any case, further careful interpretation of spatialtemporal relationships in lava, dyke, and plutonic sequences in these ophiolites is needed to interpret the complexity, or otherwise, of their evolutionary progress. Finally, the model is consistent with internal and regional-scale age relations of the three ophiolites. Given the available database for NeoTethys, the ophiolites belong to discrete temporal ophiolite 'clusters', Mirdita to an 'Outer Hellenic' Pindos-Mirdita group (c. 180-145 Ma) that includes paired ophiolite belts, and TroodosMamonia and Semail to a 'Tauride-Zagros' group (c. 98-65 Ma), which we tentatively interpret to represent discrete pre-collision subduction rollback cycles. These cycles appear to correlate with global-scale sea-floor spreading adjustments (Gnos et al. 1997; Flower & Dilek, in prep.).

Conclusions On the basis of observed structural, dynamic, and petrological relationships of western Pacific and Mediterranean marginal basins, we argue that the genesis, entrapment, and eventual collapse of Tethyan marginal basins reflect mantle flow deflected by imminent collisions with Eurasia of detached Gondwana continental fragments (Flower & Dilek 2003).

60

Y. DILEK & M. F. J. FLOWER

Fig. 5. Generalized tectonic diagrams depicting late-stage (prior to trench-passive margin collision and subsequent ophiolite emplacement) petrogenetic evolution of the Mirdita (Albania), Troodos (Cyprus), and Semail (Oman) ophiolites in slab rollback cycles within the Neo-Tethyan realm. Mantle flow in the upper plate of subduction zones comprises corner flow plus mantle extrusion (see Flower & Dilek 2003, for explanation). In all three cases the earlier formed oceanic crust is intruded and overlain by calc-alkaline and boninitic rocks that formed in a proto-arc-forearc setting of a subduction zone. Slab rollback and mantle flow collectively result in upper-plate extension and further melting of previously depleted asthenosphere producing the boninitic proto-arc material. In Semail, the late-stage alkali basalts and dykes of the Salahi unit are probably the result of off-axis magmatism fed by melts that originated within an asthenospheric window, shortly before the emplacement of the ophiolite onto the Arabian continental margin. Subduction rollback cycles within the Pindos—Mirdita and Troodos Basins might have been initiated adjacent to continental margins, but in intra-oceanic conditions within the Southern Neo-Tethys in the case of the Semail ophiolite.

The three ophiolite complexes reviewed here share (with each other and, we believe, with most other Neo-Tethyan ophiolites) three fundamental attributes predicted by this model: the presence of

'normal' oceanic basement, the evidence of refractory melts and high-temperature metamorphic soles for 'hot' subduction initiation, and temporalkinematic links to distal plate tectonic effects. For

SLAB ROLLBACK AND OPHIOLITE GENESIS each of these ophiolites, their lithological diversity and structural relationships match those associated with modern forearcs, demonstrably produced by progressive incorporation of the igneous and metamorphic products of proto-arc, back-arc spreading, and rejuvenescent arc magmatic activity. The igneous evolution of these Tethyan ophiolites thus involved subduction initiation and one or more episodes of arc splitting and basin opening. Tectonic emplacement of the Tethyan ophiolites was facilitated by trench-continent collisions that effectively terminated subduction rollback cycles. An entrapped collage of 'forearc oceanic lithosphere' and associated tectonic units overrode the partially subducted passive continental margins in the case of the Mirdita and Semail ophiolites. The Troodos ophiolite is at present in the process of being emplaced via the collision of the Eratosthenes Seamount with the Cyprus Trench. The tectonic facies hypothesis (Hsu 1994a, 1994b) is strongly validated. Because mantledriven slab rollback is able to explain the ubiquity of subparallel oceanic, suprasubduction, and exotic continental fragments in mountain belts, magmatic and tectonic relationships observed in remnant Tethys provide a graphic blueprint for early stages of an orogeny. We thank H. Furnes, E. Moores, P. Robinson, M. Shallo, P. Thy, and Y. Yilmaz for discussions on the geodynamics of the Tethyan ophiolites, and J. Encarnacion, P. Robinson, and an anonymous referee for thorough and constructive reviews of the paper. Support from the US National Science Foundation (EAR-9796011 to Y.D.; INT-0129492 to M.F.J.F.), UNESCO Earth Sciences Division, the National Geographic Society (to YD.), the International Union of Geological Sciences, and NATO (CRG-970263 and ESTCLG-97617 to YD.) is gratefully acknowledged. Logistical support by the University of Ankara, Middle East Technical University (METU-Ankara), and the Suleyman Demirel University (Isparta) during fieldwork in Turkey and by the Tirana Polytechnic University and the Albanian Geological Survey during our field investigations in Albania aided our studies greatly, and we extend our sincere gratitude to our colleagues and friends in these institutions for their help and support.

References ABELSON, M., BAER, G. & AGNON, A. 2001. Evidence from gabbro of the Troodos Ophiolite for lateral magma transport along a slow-spreading ocean ridge. Nature, 409, 72-75. ALABASTER, T., PEARCE, J.A. & MALPAS, J. 1982. The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex. Contributions to Mineralogy and Petrology, 81, 168-183. ALLERTON, S. & VINE, FJ. 1991. Spreading evolution of the Troodos Ophiolite, Cyprus. Geology, 19,

61

637-640. BAILEY, W.R., HOLDSWORTH, R.E. & SWARBRICK, R.E. 2000. Kinematic history of a reactivated oceanic suture, the Mamonia Complex suture zone, SW Cyprus. Journal of the Geological Society, London, 157, 1107-1126. BATANOVA, V.G. & SOBOLEV, A.V. 2000. Compositional heterogeneity in subduction-related mantle peridotites, Troodos massif, Cyprus. Geology, 28, 55-58. BEAUMONT, C., ELLIS, S., HAMILTON, J. & FULLSACK, P. 1996. Mechanical model for subduction-collision tectonics of Alpine-type compressional orogens. Geology, 24, 675-678. BEBIEN, J., DUBOIS, R. & GAUTHIER, A. 1986. Example of ensialic ophiolites emplaced in a wrench zone: innermost Hellenic ophiolite belt (Greek Macedonia). Geology, 14, 1016-1019. BEBIEN, J., PLATEVOET, B. & MERCIER, J. 1994. Geodynamic significance of the Paikon Massif in the Hellenides: contribution of the volcanic rocks studies. Bulletin of the Geological Society of Greece, XXX(l), 63-67. BEBIEN, J., DAUTAJ, N., SHALLO, M., TURKU, I. & BARBARIN, B. 1997. Diversity of ophiolitic plagiogranites, the Albanian example. Comptes Rendus de I'Academic des Sciences, Serie II-A, 324, 875-882. BEBIEN, J., SHALLO, M., MANIKA, K. & GEGA, D. 1998. The Shebenik massif (Albania), a link between MOR- and SSZ-type ophiolites? Ofioliti, 23, 7-15. BEBIEN, J., DIMO-LAHITTE, A., VERGELY, P., INSERGUEix-FiLiPPi, D. & DUPEYRAT, L. 2000. Albanian ophiolites. I—Magmatic and metamorphic processes associated with the initiation of a subduction. Ofioliti, 25, 39-45. BECCALUVA, L., OHNENSTETTER, D., OHNENSTETTER, M. & PAUPY, A. 1984. Two magmatic series with island arc affinities within the Vourinos ophioliote. Contributions to Mineralogy and Petrology, 85, 253-271. BECCALUVA, L., COLTORTI, M., PREMTI, L, SACCANI, E., SERRI, G. & ZEDA, O. 1994. Mid-ocean and suprasubduction affinities in ophiolitic belts from Albania. Ofioliti, 19, 77-96. BECCALUVA, L., COLTORTI, M., FERRINI, V., SIENA, F. & ZEDA, O. 1998. Petrological modelling of Albanian ophiolites with particular regard to the Bulqiza chromite ore deposits. Periodico di Mineralogia, 67, 7-23. BEDNARZ, U. & SCHMINCKE, H.-U. 1994. Petrological and chemical evolution of the northeastern Troodos Extrusive Series, Cyprus. Journal of Petrology, 35, 489-523. BEDNARZ, U., BRACE, T., CANN, J.R., HERZIG, P.M., JAMIESON, H.E., LYDON, J.W. & MALPAS, J. 1987a. Major element and trace element analytical data, Cyprus Crustal Study Project Hole CY2-2a. In: ROBINSON, P.T., GIBSON, I.L. & PANAYIOTOU, A. (eds) Cyprus Crustal Study Project, Initial Report, CY-2 and 2a. Geological Survey of Canada, Papers, 85-29, 353-381. BEDNARZ, U., SUNKEL, G. & SCHMINCKE, H.-U. 1987b. The basaltic andesite-andesite and the andesitedacite series from the ICRDG drill holes CY-2a and CY-2a. In: ROBINSON, P.T., GIBSON, I.L. & PA-

62

Y. DILEK & M. F. J. FLOWER

NAYIOTOU, A. (eds) Cyprus Crustal Study Project, Initial Report, CY-2 and 2a. Geological Survey of Canada, Papers, 85-29, 183-204. BENN, K., NICOLAS, A. & REUBER, I. 1988. Mantle-crust transition zone and origin of wehrlitic magmas, evidence from the Oman Ophiolite. Tectonophysics, 151, 75-85. BENOIT, M., CEULENEER, G. & POLVE, M. 1999. The remelting of hydrothermally altered peridotite at ocean ridges by intruding mantle diapirs. Nature, 402,514-518. BLOOMER, S.H., TAYLOR, B., MACLEOD, C.J., STERN, R.J., FRYER, P., HAWKINS, J.W. & JOHNSON, L. 1995. Early arc volcanism and the ophiolite problem, a perspective from drilling in the western Pacific. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 1-30. BORRADAILE, GJ. 2001. Paleomagnetic vectors and tilted dikes. Tectonophysics, 333, 417-426. BORRADAILE, G.J. & LAGROIX, F. 2001. Magnetic fabrics reveal upper mantle flow fabrics in the Troodos ophiolite complex, Cyprus. Journal of Structural Geology,23, 1299-1317. BORTOLOTTI, V., KODRA, A., MARRONI, M., MUSTAFA,

F., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 1996. Geology and petrology of ophiolitic sequences in the Mirdita region (northern Albania). Ofioliti, 21, 3-20. BORTOLOTTI, V., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2002. Interaction between midocean ridge and subduction magmatism in Albanian ophiolites. Journal of Geology, 110, 561-576. BRIAIS, A., PATRIAT, P. & TAPONNIER, P. 1993. Updated interpretation of magnetic anomalies and seafloor spreading rates in the South China Sea; implications for the Tertiary tectonics of Southeast Asia. Journal of Geophysical Research, 98, 6299-6328. BROWNING, P. 1984. Cryptic variation within the Cumulate Sequence of the Oman ophiolite, magma chamber depth and petrological implications. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 71-82. BURCHFIEL, B.C. & ROYDEN, L.H. 1991. Antler orogeny, a Mediterranean-type orogeny. Geology, 19, 66-69. CANN, J.R., PRICHARD, H.M., MALPAS, J.G. & XENOPHONTOS, C. 2001. Oceanic inside corner detachments of the Limassol Forest area, Troodos Ophiolite, Cyprus. Journal of the Geological Society, London, 158, 757-767. CAPEDRI, S., VENTURELLI, E., BOCCHI, E., DOSTAL, J., GARUTI, G. & Rossi, A. 1980. The geochemistry and petrogenesis of an ophiolite sequence from Pindos, Greece. Contributions to Mineralogy and Petrology, 74, 189-200. CAROSI, R., CORTESOGNO, L., GAGGERO, L. & MARRONI, M. 1996. Geological and petrological features of the metamorphic sole from the Mirdita ophiolites, northern Albania. Ofioliti, 21, 21-40. CLIFT, P.O. & LEE, J. 1998. Temporal evolution of the Mariana arc during rifting of the Mariana Trough traced through the volcaniclastic record. Island Arc,

7,496-512. COLLINS, A.S. & ROBERTSON, A.H.F. 1997. Lycian melange, southwestern Turkey: an emplaced Late Cretaceous accretionary complex. Geology, 25, 255-258. CORTESOGNO, L., GAGGERO, L., JAHO, E., MARRONI, M., PANDOLFI, L. & SHTJEFANAKU, D. 1998. The gabbroic complex of the western ophiolitic belt, northern Albania, an example of multilayered sequence in an intermediate-spreading oceanic ridge. Ofioliti, 23, 49-64. DILEK, Y. 1999. Non-volcanic rifted margins and continental margin ophiolites of the Mesozoic NeoTethys, and dispersal of Gondwanaland. Non-Volcanic Rifting of Continental Margins: a Comparison of Evidence from Land and Sea. Abstracts of Talks & Posters. Geological Society, London, 17. DILEK, Y. & EDDY, C.A. 1992. The Troodos (Cyprus) and Kizildag (S. Turkey) ophiolites as structural models for slow-spreading ridge segments. Journal of Geology, 100, 305-322. DILEK, Y. & MOORES, E.M. 1990. Regional tectonics of the Eastern Mediterranean ophiolites. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) 295-309.Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. DILEK, Y. & ROWLAND, J.C. 1993. Evolution of a conjugate passive margin pair in Mesozoic southern Turkey. Tectonics, 12, 954-970. DILEK, Y., THY, P., HACKER, B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and mafic dike intrusions (Turkey): implications for the Neo-Tethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) 2000. Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349. DILEK, Y., FURNES, H., SHALLO, M. & MORTENSEN, J.K. 2001. Structure, petrology, and geochronology of the Albanian ophiolites and their tectonic evolution within the Neotethyan orogenic belt. Journal of Conference Abstracts, 6(1), 321. DIMO-LAHITTE, A., MONIE, P. & VERGELY, P. 2001. Metamorphic soles from the Albanian ophiolites, petrology, 40Ar/39Ar geochronology, and geodynamic evolution. Tectonics, 20, 78-96. DOGLIONI, C., HARABAGLIA, P., MERLINI, S., MONGELLI, F., PECCERILLO, A. & PIROMALLO, C. 1999. Orogens and slabs versus their direction of subduction. Earth-Science Reviews, 45, 167-208. EATON, S. & ROBERTSON, A. 1993. The Miocene Pakhna Formation, Southern Cyprus and its relationship to the Neogene tectonic evolution of the eastern Mediterranean. Sedimentary Geology, 86, 273-296. ENCARNACION, J.P., MUKASA, S.B. & OBILE, B.C. JR 1993. Origin of the Zambales and Angat ophiolites, Philippines, new constraints from zircon U—Pb geochronology. EOS Transactions, American Geophysical Union, 74, 286. ERNEWEIN, M., PFLUMIO, C. & WHITECHURCH, H. 1988. The death of an accretion zone as evidenced

SLAB ROLLBACK AND OPHIOLITE GENESIS by the magmatic history of the Sumail Ophiolite (Oman). Tectonophysics, 151, 247-274. FLOWER, M.FJ. & DILEK, Y. 2003. Arc-trench rollback and forearc accretion: 1. A collision-induced mantle flow model for Tethyan ophiolites. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218,21-41. FLOWER, M.FJ. & LEVINE, H.M. 1987. Petrogenesis of a tholeiite-boninite sequence from Ayios Mamas, Troodos ophiolite, evidence for splitting of a volcanic arc? Contributions to Mineralogy and Petrology, 97, 509-524. FLOWER, M.FJ., TAMAKI, K. & HOANG, N. 1998. Mantle extrusion, a model for dispersed volcanism and DUPAL-like asthenosphere in East Asia and the western Pacific. In: FLOWER, M.FJ., CHUNG, S.-L., Lo, C.-H. & LEE, T.-Y. (eds) Mantle Dynamics and Plate Interactions in East Asia. American Geophysical Union, Geodynamics Series, 27, 67-88. FLOWER, M.FJ., Russo, R.M., TAMAKI, K. & HOANG, N. 2001. Mantle contamination and the IzuBonin-Mariana (IBM) 'high-tide mark', evidence for mantle extrusion caused by Tethyan closure. Tectonophysics, 333, 9-34. FRYER, P. 1996. Evolution of the Mariana convergent plate margin system. Reviews of Geophysics, 34, 89-125. GEARY, E.E. & KAY, R.W. 1989. Identification of an Early Cretaceous ophiolite in the Camarines NorteCalaguas Islands basement complex, eastern Luzon, Philippines. Tectonophysics, 168, 109-126. GIBSON, I.L., McGiLL, G.C. & ROBINSON, P.T. 1987. The geochemistry and nature of unusual glassy lavas from CCSP holes CY-2 and CY-2a, Troodos Ophiolite, Cyprus. In: ROBINSON, P.T., GIBSON, I.L. & PANAYIOTOU, A. (eds) Cyprus Crustal Study Project, Initial Report, CY-2 and 2a,. Geological Survey of Canada, Papers, 85-29, 79-85. GJATA, K., KORNPROBST, J., KONDRA, A., BRIOT, D. & PINEAU, F. 1992. Hot subduction close to a ridge— example of the garnet pyroxenite enclaves in the serpentinite breccia at Derveni, Albania. Bulletin de la Societe Geologique de France, 163, 469-476. GNOS, E. 1998. Peak metamorphic conditions of garnet amphibolites beneath the Semail Ophiolite, implications for an inverted pressure gradient. International Geology Review, 40, 281-304. GNOS, E., IMMENHAUSER, A. & PETERS, T. 1997. Late Cretaceous/early Tertiary convergence between the Indian and Arabian plates recorded in ophiolites and related sediments. Tectonophysics, 271, 1-19. GODARD, M, JOUSSELIN, D. & BODINIER, J.-L. 2000.

Relationships between geochemistry and structure beneath a palaeo-spreading centre, a study of the mantle section in the Oman Ophiolite. Earth and Planetary Science Letters, 180, 133-148. HACKER, B.R. & GNOS, E. 1997. The conundrum of Samail, explaining the metamorphic history. Tectonophysics, 279, 215-226. HACKER, B.R. & MOSENFELDER, J.L. 1996. Metamorphism and deformation along the emplacement thrust of the Semail Ophiolite, Oman. Earth and Planetary Science Letters, 144, 435-451.

63

HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman ophiolite, thermal and geochronologic constraints. Tectonics, 15, 1230-1247. HAWKINS, J.W. & CASTILLO, P.R. 1998. Early history of the Izu-Bonin-Mariana arc system, evidence from Belau and the Palau Trench. Island Arc, 1, 559-578. HOANG, N.T. & FLOWER, M.FJ. 1998. Petrogenesis of Cenozoic basalts from Vietnam: implication for origins of a 'diffuse igneous province'. Journal of Petrology, 39, 34-50. HOXHA, M. & BOULLIER, A.M. 1995. The Peridotites of the Kukes Ophiolite (Albania)—structure and kinematics. Tectonophysics, 249, 217-231. HOXHA, M., BOULLIER, A.M. & ZHUPI, S. 1993. Structure of the peridotites from the Kukes Ophiolite (Albania). Bulletin de la Societe Geologique de France, 164, 449-458. Hsu, KJ. 1994a. Tectonic facies in an archipelago model of intra-plate orogenesis. GSA Today, 4, 289-290. Hsu, KJ. 1994b. Orogeny of back-arc collision and its tectonic facies. Nanjing Daxue Xuebao-Diqiu Kexue (Journal of Nanjing University—Earth Sciences), 6, 1-12. HUTCHISON, C.S. 1990. The Sulu Sea, a Southeast Asian Neogene marginal basin with outcropping extensions in Sabah. Geological Society of Malaysia Annual Meeting (Persidangan Tahunan Geology '90), Program with Abstracts, 16, 134. iNSERGUEIX-FlLIPPI, D., DUPEYRAT, L., DlMO-LAHITTE,

A., VERGELY, P. & BEBIEN, J. 2000. Albanian ophiolites: model of subduction zone infancy at an ocean ridge. Ofioliti, 25, 47-53. ISHIKAWA, T., NAGAISHI, K. & UMINO, S. 2002. Boninitic volcanism in the Oman ophiolite: implications for thermal condition during transition from spreading ridge to arc. Geology, 30, 899-902. ISPGJ-IGJN 1983. Geological Map of Albania. NMMS 'Hamid Shijaku', Scale 1:200000. ISPGJ-IGJN, Tirana. JOLIVET, L. & PATRIAT, M. 1999. Ductile extension and the formation of the Aegean Sea. In: DURAND, B., JOLIVET, L., HORATH, F. & SERANNE, M (eds) Mediterranean Basins, Tertiary Extension within the Alpine Orogen. Geological Society, London, Special Publications, 156, 427-456. JOLIVET, L. & FACCENNA, C. 2000. Mediterranean extension and the Africa-Eurasia collision. Tectonics, 19, 1095-1106. JOLIVET, L., FACCENNA, C., D'AGOSTINO, N., FOURNIER, M. & WORRALL, D. 1999. The kinematics of backarc basins, examples from the Tyrrhenian, Aegean, and Japan seas. In: MAC NIOCAILL, C. & RYAN, P.D. (eds) Continental Tectonics. Geological Society, London, Special Publications, 164, 21-53. JONES, G. & ROBERTSON, A.H.F. 1991. Tectono-stratigraphy and evolution of the Mesozoic Pindos Ophiolite and related units, Northwestern Greece. Journal of the Geological Society, London, 148, 267-288. JOUSSELIN, D. & MAINPRICE, D. 1998. Melt topology and seismic anisotropy in mantle peridotites of the Oman Ophiolite. Earth and Planetary Science

64

Y. DILEK & M. F. J. FLOWER

Letters, 164, 553-568. JOUSSELIN, D., NICOLAS, A. & BOUDIER, F. 1998. Detailed mapping of a mantle diapir below a paleospreading center in the Oman Ophiolite. Journal of Geophysical Research, 103, 18153-18170. JUTEAU, T. 1980. Ophiolites of Turkey. Ofioliti, 2, 199-238. JUTEAU, T., BEURRIER, M., DAHL, R. & NEHLIG, P. 1988a. Segmentation at a fossil spreading axis, the plutonic sequence of the Wadi Haymiliyah area (Haylayn Block, Sumail Nappe, Oman). Tectonophysics, 151, 167-197. JUTEAU, T., ERNEWEIN, M., REUBER, L, WHITECHURCH, H. & DAHL, R. 1988b. Duality of magmatism in the plutonic sequence of the Sumail Nappe, Oman. Tectonophysics, 151, 107-135. KODRA, A., VERGELY, P., GJATA, K., BAKALLI, F. & GODROLI, M. 1993. La formation volcano-sedimentaire du Jurassique superieur: temoin de 1'ouverture du domaine ophiolitique dans les Albanides internes. Bulletin de la Societe Geologique de France, 164, 61067. KODRA, A., GJATA, K. & ROBERTSON, A.H.F. 2002. Mesozoic—Tertiary tectonic evolution of Albania in its regional eastern Mediterranean context. Discussion and Reply. Tectonophysics, 344, 297-306. KOEPKE, J., SEIDEL, E. & KREUZER, H. 2002. Ophiolites on the Southern Aegean islands Crete, Karpathos and Rhodes: composition, geochronology and position within the ophiolite belts of the eastern Mediterranean. Lithos, 65, 183-203. LACHIZE, M., LORAND, J.P. & JUTEAU, T. 1996. Calcalkaline differentiation trend in the plutonic sequence of the Wadi Haymiliyah section, Haylayn Massif, Semail Ophiolite, Oman. Lithos, 38, 207-232. LAVECCHIA, G. & STOPPA, F. 1996. The tectonic significance of Italian magmatism, an alternative view to the popular interpretation. Terra Nova, 8, 435-446. LE PICHON, X., LALLEMANT, S.J., CHAMOT-ROOKE, N., LEMEUR, D. & PASCAL, G. 2002. The Mediterranean Ridge backstop and the Hellenic nappes. Marine Geology, 186, 111-125. LEYBOURNE, M.I., VAN WAGONER, N. & AYRES, L.D. 1999. Partial melting of a refractory subducted slab in a Paleoproterozoic island arc, implications for global chemical cycles. Geology, 27, 731—734. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Geological Society, London, Memoirs, 11. LONERGAN, L. & WHITE, N. 1997. Origin of the BeticRif mountain belt. Tectonics, 16, 504-522. MACLEOD, C.J. & MURTON, B.J. 1993. Structure and tectonic evolution of the southern Troodos transform fault zone, Cyprus. In: PRICHARD, H.M., ALABASTER, T., HARRIS, N.B.W. & NEARY, C.R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 141-176. MALPAS, J. & LANGDON, I.G. 1984. Petrology of the Upper Pillow Lava suite, Troodos Ophiolite, Cyprus. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere.

Geological Society, London, Special Publications, 13, 155-167. MALPAS, J., BRACE, T. & DUNSWORTH, S.M. 1989. Structural and petrological relationships of the CY-4 drill hole of the Cyprus Crustal Study Project. In: GIBSON, I.L., MALPAS, J., ROBINSON, P.T. & XENOPHONTOS, C. (eds) Initial Report, Hole CY-4. Geological Survey of Canada, Papers, 88-9, 39-67. MALPAS, J., XENOPHONTOS, C. & WILLIAMS, D. 1992. The Ayia Varvara Formation of SW Cyprus—a product of complex collisional tectonics. Tectonophysics, 212, 193-211. MALPAS, J., CALON, T. & SQUIRES, G. 1993. The development of a Late Cretaceous microplate suture zone in SW Cyprus. In: PRICHARD, H.M., ALABASTER, T., HARRIS, N.B.W. & NEARY, C.R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 177-195. MANIKA, K., SHALLO, M. & GREGA, D. 1997. The plutonic sequence of Shebenik ophiolitic complex, Albania: evidence for dual magmatism. Ofioliti, 22, 93-99. McCABE, R. & UYEDA, S. 1983. Hypothetical model for the bending of the Mariana Arc. In: HAYES, D.E. (ed.) The Tectonic and Geological Evolution of Southeast Asian Seas and Islands, Part 2. Geophysical Monograph, American Geophysical Union, 27, 281-293. MELCHER, F., MEISEL, T., PUHL, J. & ROLLER, F. 2002. Petrogenesis and geotectonic setting of ultramafic rocks in the Eastern Alps: constraints from geochemistry. Lithos, 65, 69-112. MICHARD, A., WHITECHURCH, H., Ricou, L.E., MONTIGNY, R. & YAZGAN, E. 1984. Tauric subduction (Malatya-Elazig provinces) and its bearing on tectonites of the Tethyan realm in Turkey. In: DIXON, J.E. & ROBERTSON, A.H.F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 349-360. MITCHELL, A.H.G., HERNANDEZ, F. & DELA CRUZ, A.P. 1986. Cenozoic evolution of the Philippine Archipelago. Journal of SE Asian Earth Sciences, 1, 3-22. MONNIER, C., GlRARDEAU, J., MAURY, R.C. & GOTTEN,

J. 1995. Back-arc basin origin for the East Sulawesi Ophiolite (eastern Indonesia). Geology, 23, 851-854. MOORES, E.M., ROEDER, D.H., ABBAS, S.G. & AHMED, Z. 1980. Geology and emplacement of the Muslim Bagh ophiolite complex. In: PANAYIOTOU, A. (ed.) Ophiolites, Proceedings International Ophiolite Symposium Cyprus 1979. Geological Survey Department, Nicosia, 424-429. MOORES, E.M., ROBINSON, P.T., MALPAS, J. & XENOPHONTOS, C. 1984. Model for the origin of the Troodos Massif, Cyprus, and other Mideast Ophiolites. Geology, 12, 500-503. MORRA, V., SECCHI, F.A.G., MELLUSO, L. & FRANCIOSI, L. 1997. High-Mg subduction-related Tertiary basalts in Sardinia, Italy. Lithos, 40, 69-91. MORRIS, A., ANDERSON, M.W. & ROBERTSON, A.H.F. 1998. Multiple tectonic rotations and transform

SLAB ROLLBACK AND OPHIOLITE GENESIS tectonism in an intraoceanic suture zone, SW Cyprus. Tectonophysics, 299, 229-253. MUKASA, S.B. & LUDDEN, J.N. 1987. Uranium-lead ages of plagiogranites from the Troodos ophiolite, Cyprus, and their tectonic significance. Geology, 15, 825-828. MURTON, BJ. 1989. Tectonic controls on boninite genesis. In: SAUNDERS, A.D. & NORRY, MJ. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 347-377. MURTON, BJ. 1990. Was the Southern Troodos Transform Fault a victim of microplate rotation? In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 87-98. NICOLAS, A. & BOUDIER, F. 1991. Rooting of the sheeted dike complex in the Oman Ophiolite. /W.PETERS, TJ., NICOLAS, A. & COLEMAN, R.G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Petrology and Structural Geology. Kluwer, Dordrecht, 5, 39-54. NICOLAS, A. & BOUDIER, F. 1995. Mapping oceanic ridge segments in Oman Ophiolite. Journal of Geophysical Research, 100, 6179-6197. NICOLAS, A. & BOUDIER, F. 2000. Large mantle upwellings and related variations in crustal thickness in the Oman ophiolite. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 67-73. NICOLAS, A., BOUDIER, F. & CEULENEER, G. 1988a. Mantle flow patterns and magma chambers at ocean ridges, evidence from the Oman Ophiolite. Marine Geophysical Research, 9, 293-310. NICOLAS, A., REUBER, I. & BENN, K. 1988b. A new magma chamber model based on structural studies in the Oman Ophiolite. Tectonophysics, 151, 87-105. NICOLAS, A., BOUDIER, F. & ILDEFONSE, B. 1994. Evidence from the Oman ophiolite for active mantle upwelling beneath a fast-spreading ridge. Nature, 370,51-53. NICOLAS, A., BOUDIER, F. & ILDEFONSE, B. 1996. Variable crustal thickness in the Oman Ophiolite, implication for oceanic crust. Journal of Geophysical Research, 101, 17941-17950. NICOLAS, A., BOUDIER, F. & MESHI, A. 1999. Slow spreading accretion and mantle denudation in the Mirdita Ophiolite (Albania). Journal of Geophysical Research, 104, 15155-15167. PADOA, E. 1999. Les ophiolites du Massif de 1'Inzecca (Corse Alpine), lithostratigraphie, structure geologique et evolution geodynamique. Geologie de la France, 1999, 37-48. PALLISTER, J.S. 1981. Structure of the sheeted dike complex of the Samail ophiolite near Ibra, Oman. Journal of Geophysical Research, 86, 2661-2672. PALLISTER, J.S. 1984. Parent magmas of the Semail ophiolite, Oman. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special

65

Publications, 13, 63-70. PALLISTER, J.S. & HOPSON, C.A. 1981. Samail ophiolite plutonic suite, field relations, phase variation, cryptic variation and layering, and a model of a spreading ridge magma chamber. Journal of Geophysical Research, 86, 2593-2644. PAMIC, J., TOMLJENOVIC, B. & BALEN, D. 2002. Geodynamic and petrogenetic evolution of Alpine ophiolites from the central and NW Dinarides: an overview. Lithos, 65, 113-142. PEARCE, J.A., ALABASTER, T., SHELTON, A.W., SEARLE, M.P., VINE, F.J. & SMITH, A.G. 1981. The Oman ophiolite as a Cretaceous arc-basin complex, evidence and implications. Philosophical Transactions of the Royal Society of London, Series A, 300, 299-317. PE-PIPER, G. 1998. The nature of Triassic extensionrelated magmatism in Greece: evidence from Nd and Pb isotope geochemistry. Geological Magazine, 135,331-348. POLAT, A., CASEY, J.F. & KERRICH, R. 1996. Geochemical characteristics of accreted material beneath the Pozanti-Karsanti ophiolite, Turkey: intra-oceanic detachment, assembly and obduction. Tectonophysics, 263, 249-276. PORTNYAGIN, M.V., MAGAKYAN, R. & SCHMINCKE, H.-

U. 1996. Geochemical variability of boninite magmas: evidence from magmatic inclusions in highly magnesian olivine from lavas of southwestern Cyprus. Petrology, 4, 231-246. PUBELLIER, M., QUEBRAL, R., AURELIO, M. & RANGIN,

C. 1996. Docking and post-docking escape tectonics in the southern Philippines. In: HALL, R. & BLUNDELL, D. J. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 511-523. RANGIN, C., SILVER, E.A. & TAMAKI, K. 1995. Closure of western Pacific marginal basins, rupture of the oceanic crust and the emplacement of ophiolites. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 405-417. RANGIN, C., SPAKMAN, W., PUBELLIER, M. & BuWAARD, H. 1999. Tomographic and geological constraints on subduction along the eastern Sundaland continental margin (South-East Asia). Bulletin de la Societe Geologique de France, 170, 775-788. RASSIOS, A. & SMITH, A.G. 2000. Constraints on the formation and emplacement age of western Greek ophiolites (VourinosPindos, and Othris) inferred from deformation structures in peridotites. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights From Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 473-483. RAUTENSCHLEIN, M., JENNER, G.A., HERTOGEN, J., HOFMANN, A.W., KERRICH, R., SCHMINCKE, H.-U. & WHITE, W.M. 1985. Isotopic and trace element composition of volcanic glasses from the Akaki Canyon, Cyprus, implications for the origin of the Troodos Ophiolite. Earth and Planetary Science Letters, 75, 369-383.

66

Y. DILEK & M. F. J. FLOWER

ROBERTSON, A. & SHALLO, M. 2000. Mesozoic-Tertiary tectonic evolution of Albania in its regional eastern Mediterranean context. Tectonophysics, 316, 197-254. ROBERTSON, A.H.F. 2000. Tectonic evolution of Cyprus in its easternmost Mediterranean setting. In: PANAYIDES, I., XENOPHONTON, C. & MALPAS, J. (eds) Proceedings Third International Conference on the Geology of the Eastern Mediterranean. Geological Survey Department, Nicosia, 11 -44. ROBERTSON, A.H.F. 2002. Overview of the genesis and emplacement of Mesozoic ophiolites in the Eastern Mediterranean Tethyan region. Lithos, 65, 1-67. ROBERTSON, A.H.F., CLIFT, P.O., DEGNAN, PJ. & JONES, G. 1991. Palaeogeographic and palaeotectonic evolution of the eastern Mediterranean NeoTethys. Paleoceanography, Palaeoclimatology, Palaeoecology, 87, 289-343. ROBINSON, P.T., MELSON, W.G., O'HEARN, T. & SCHMINCKE, H.-U. 1983. Volcanic glass compositions of the Troodos ophiolite, Cyprus. Geology, 11, 400-404. ROGERS, N.W., MACLEOD, C.J. & MURTON, B.J. 1989. Petrogenesis of boninitic lavas from the Limassol Forest Complex Cyprus. In: CRAWFORD, A.J. (ed.) Boninites. Allen & Unwin, London, 288-313. ROYDEN, L.H. 1993a. The tectonic expression of slab pull at continental convergent boundaries. Tectonics, 12, 303-325. ROYDEN, L.H. 1993b. Evolution of retreating subduction boundaries formed during continental collision. Tectonics, 12, 629-638. SACCANI, E., PADOA, E. & PHOTIADES, A. 2003. Triassic mid-ocean ridge basalts from the Argolis Peninsula (Greece): new constraints for the early oceanization phases of the Neo-Tethyan Pindos basin. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 109-127. SAREWITZ, D.R. & LEWIS, S.D. 1991. The Marinduque intra-arc basin, Philippines, basin genesis and in situ ophiolite development in a strike-slip setting. Geological Society of America Bulletin, 103, 597-614. SARKARINEJAD, K. 2003. Structural and microstructural analysis of a palaeo-transform fault zone in the Neyriz ophiolite, Iran. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 129145. SCHIANO, P., CLOCCHIATTI, R., LORAND, J.-P., MASSARE, D., DELOULE, E. & CHAUSSIDON, M. 1997. Primitive basaltic melts included in podiform chromites from the Oman Ophiolite. Earth and Planetary Science Letters, 146, 489-497. SCHOUTEN, H. & KELEMEN, P.B. 2002. Melt viscosity, temperature and transport processes, Troodos Ophiolite, Cyprus. Earth and Planetary Science Letters, 201, 337-352. SEARLE, M. & Cox, J. 1999. Tectonic setting, origin, and obduction of the Oman Ophiolite. Geological Society of A merica Bulletin ,111, 104-122. SEARLE, M.P. & MALPAS, J. 1980. Structure and metamorphism of rocks beneath the Semail Ophio-

lite of Oman and their significance in ophiolite obduction. Transactions of the Royal Society of Edinburgh, Part, 71, 247-262. SEARLE, M.P. & MALPAS, J. 1982. Petrochemistry and origin of sub-ophiolitic metamorphic and related rocks in the Oman Mountains. Journal of the Geological Society, London, 139, 235-248. SHALLO, M. 1990a. Volcanic glasses of the Albanian ophiolite belt. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 271-278. SHALLO, M. 1990b. Boninites within eastern ophiolites of Albania (in Albanian with English abstract). Buletini i Shencave Gjeologjike, 4, 19-37. SHALLO, M. 1990c. Ophiolitic melange and flyschoidal sediments of the Tithonian-Lower Cretaceous in Albania. Terra Nova, 2, 476-482. SHALLO, M. 1992. Geological evolution of the Albanian ophiolites and their platform periphery. Geologische Rundschau, 81, 681-694. SHALLO, M. & DILEK, Y. 2003. Development of the ideas on the origin of Albanian ophiolites. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 351-364. SHALLO, M., KOTE, DH., VRANAJ, A. & PREMTI, I. 1985. Ophiolitic Magmatism of Albania (in Albanian). Scientific Reports, Fakulteti i Gjeologjise dhe Minierave, Universiteti Politeknik, Tirana. SHALLO, M., KOTE, D. & VRANAJ, A. 1987. Geochemistry of the volcanics from ophiolitic belts of Albanides. Ofioliti, 12, 125-136. SHALLO, M., KODRA, A. & GJATA, K. 1990. Geotectonics of the Albanian ophiolites. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 165-170. SHALLO, M., TURKU, L, MANIKA, K. & MARTO, A. 1996. Genesis and Geotectonic Setting of the Albanian Ophiolites. Scientific Reports, Fakulteti i Gjeologjise dhe Minierave, Universiteti Politeknik, Tirana. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, Paper Number 2000GC000080. SIBUET, J.C., DEFFONTAINES, B., Hsu, S.K., THAREAU, N., LE FORMAL, J.P. & Liu, C.S. 1998. Okinawa trough backarc basin, early tectonic and magmatic evolution. Journal of Geophysical Research, 103, 30245-30267. SILVER, E.A., MCCAFFREY, R., JOYODIWIRYO, Y. & STEVENS, S. 1983. Ophiolite emplacement by collision between the Sula Platform and the Sulawesi island arc, Indonesia. Journal of Geophysical Research, 88, 9419-9435. SIMONIAN, K.O. & GASS, I.G. 1978. Arakapas fault belt, Cyprus, a fossil transform fault. Geological Society of America Bulletin, 89, 1220-1230. SPAKMAN, W. 1991. Delay-time tomography of the upper

67

SLAB ROLLBACK AND OPHIOLITE GENESIS mantle below Europe, the Mediterranean, and Asia Minor. Geophysical Journal International, 107, 309-332. SPRAY, J.G. & RODDICK, J.C. 1981. Evidence of Upper Cretaceous transform metamorphism in West Cyprus. Earth and Planetary Science Letters, 55, 72-75. SPRAY, J.G., BEBIEN, I, REX, D.C. & RODDICK, J.C. 1984. Age constraints on the igneous and metamorphic evolution of the Hellenic-Dinaric ophiolites. In: DIXON, I.E. & ROBERTSON, A.H.F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 619-627. STAMPFLI, G.M. 2000. Tethyan oceans. In: BOZKURT, E., WINCHESTER, J.A. & PIPER, J.D.A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 1-23. TASHKO, A. 1996. Diversites geochimiques, petrologiques et structurales des ophiolites des Albanides. Bulletin de la Societe Geologique de France, 167, 335-343. TAYLOR, B., KLAUS, A., BROWN, G.R., MOORE, G.F., OKAMURA, Y. & MURAKAMI, F. 1991. Structural development of Sumisu Rift, Izu-Bonin Arc. Journal of Geophysical Research, 96, 16113 -16129. TAYLOR, R.N. 1990. Geochemical stratigraphy of the Troodos extrusive sequence, temporal developments of a spreading centre magma chamber. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 173-183. TAYLOR, R.N., MURTON, B.J. & NESBITT, R.W. 1992. Chemical transects across intra-oceanic arcs, implications for the tectonic setting of ophiolites. In: PARSON, L.M., MURTON, B.J. & BROWNING, P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 117-132. THY, P. & ESBENSEN, K.H. 1993. Seafloor spreading and the ophiolitic sequences of the Troodos Complex: a principal component analysis of lava and dike compositions. Journal of Geophysical Research, 98, 11799-11805. THY, P., SCHIFFMAN, P. & MOORES, E.M. 1989. Igneous mineral stratigraphy and chemistry of the Cyprus Crustal Study Project drill core in the plutonic sequences of the Troodos ophiolite. In: GIBSON, I.L., MALPAS, J., ROBINSON, P.T. & XENOPHONTOS, C. (eds) Cyprus Crustal Study Project: Initial Report, Hole CY-4. Geological Survey of Canada, Papers, 88-9, 147-185. TILTON, G.R., HOPSON, C.A. & WRIGHT, J.E. 1981. Uranium-lead isotopic ages of the Samail ophiolite, Oman, with applications to Tethyan ocean ridge tectonics. Journal of Geophysical Research, 86, 2763-2775. UMINO, S. 1995. Geology of the Semail Ophiolite, northern Oman Mountains. Journal of Geology, 104, 321-349. UMINO, S., YANAI, S., JAMAN, A.R., NAKAMURA, Y. & IIYAMA, J.T. 1990. The transition from spreading to

subduction, evidence from the Semail Ophiolite, northern Oman Mountains. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues, Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 375-384. USTAOMER, T. & ROBERTSON, A.H.F. 1999. Geochemical models used to test alternative plate tectonic models for pre-Upper Jurassic (Palaeotethyan) units in the Central Pontides, N. Turkey. Geological Journal, 34, 25-53. UYEDA, S. ET AL. 1982. Subduction zones: an introduction to comparative subductology. Tectonophysics, 81, 133-159. VERGELY, P., DIMO, A. & MONIE, P. 1997. Datation des semelles metamorphiques ophiolitiques de 1'Albanie par la methode Ar40/Ar39: consequences sur le mecanisme de 1'obduction. Comptes Rendus de I'Academic des Sciences, 326, 713-723. VlLLENEUVE, M.,

GUNAWAN, W.,

CORNEE, JJ.

&

VIDAL, O. 2002. Geology of the central Sulawesi belt (eastern Indonesia), constraints for geodynamic models. International Journal of Earth Sciences, 91, 524-537. WAKABAYASHI, J. & DILEK, Y. 2003. What constitutes 'emplacement' of an ophiolite? Mechanisms and relationship to subduction initiation and formation of metamorphic soles. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications. Geological Society, London, Special Publications, 218, 427-447. WORTEL, MJ.R. & SPAKMAN, W. 2000. Subduction and slab detachment in the Mediterranean-Carpathian region. Science, 290, 1910-1917. YANAI, S., UMINO, S., IBRAHIM, S.O., NAKAMURA, Y. & IIYAMA, J.T. 1990. Subduction- and collision-related emplacement of the Semail Ophiolite, northern Oman Mountains. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of the Symposium Troodos 1987. Geological Survey Department, Nicosia, 385-395. YAZGAN, E. 1984. Geodynamic evolution of the Eastern Taurus region. In: TEKELI, O. & GONCUOGLU, M.C. (eds) Geology of the Taurus Belt, Proceedings of the International Symposium, Ankara. Mineral Research and Exploration Institute (MTA) of Turkey, Ankara, 199-208. YAZGAN, E. & CHESSEX, R. 1991. Geology and tectonic evolution of Southeastern Taurus in the region of Malatya. Bulletin of the Turkish Association of Petroleum Geologists, 3, 1 -42. YILMAZ, Y. 1993. New evidence and model on the evolution of the southeast Anatolian orogen. Geological Society of America Bulletin, 105, 251-271. YILMAZ, Y., TUYSUZ, O., YIGITBAS, E., GENC, S.C. & SENGOR, A.M.C. 1997. Geology and tectonic evolution of the Pontides. In: ROBINSON, A.G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists, Memoirs, 68, 183-226. YUMUL, G.P., ZHOU, M.F., TAMAYO, R.A., MAURY, R.C., FAUSTINO, D.V., OLAGUERA, P.O. & GOTTEN,

68

Y. DILEK & M. F. J. FLOWER J. 2001. Onramping of cold oceanic lithosphere in a forearc settings the Southeast Bohol Ophiolite

Complex, Central Philippines. International Geology Review, 43, 850-866.

Melt migration in ophiolitic peridotites: the message from AlpineApennine peridotites and implications for embryonic ocean basins OTHMAR MUNTENER 1 & GIOVANNI B. PICCARDO 2 1

Institute of Geology, University of Neuchdtel, 11 rue Emile Argand, CH-2007 Neuchdtel, Switzerland (e-mail: [email protected]) 2 Dipteris, Universita di Genova, Corso Europa 26, 1-16132 Genoa, Italy Abstract: Results of a field study as well as petrological and geochemical data demonstrate that substantial portions of the lithospheric mantle, exhumed during opening of the Jurassic Piedmont Ligurian ocean, were infiltrated by and reacted with migrating melts. Intergranular flow of ascending liquids produced by the underlying hot asthenosphere dissolved clinopyroxene ± spinel and precipitated orthopyroxene + plagioclase ± olivine, forming orthopyroxene + plagioclase-rich peridotite. Migrating liquids became progressively saturated in clinopyroxene, and then precipitated microgranular aggregates of clinopyroxene-bearing gabbronorite. Later, diffuse porous melt flow was replaced by focused porous flow, producing a system of discordant dunite bodies. Upon cooling, liquids migrating in dunite channels became progressively saturated in clinopyroxene and plagioclase, forming interstitial clinopyroxene at olivine triple points followed by clinopyroxene ± plagioclase megacrysts and gabbro veinlets within the dunite, and gabbro dykelets within plagioclase peridotites. Subsequent cooling during continued exhumation was accompanied by intrusion of kilometre-scale gabbroic dykes evolving from troctolite to Mg-Al and Fe-Ti gabbros. Migrating liquids, which infiltrated peridotite and formed gabbroic rocks, span a wide range of compositions from silica-rich single melt fractions to T- and N-MORB (mid-ocean ridge basalt), characteristic of the melting column beneath midocean ridges. Explanations for the progressive evolution of an igneous system from diffuse to focused porous flow and finally dyking include the competing effects of heating of the lithospheric mantle by ascending magmas from the underlying hot asthenosphere and conductive cooling by exhumation. Whether or not rift-related melt infiltration and heating is recorded by exhumed subcontinental lithospheric mantle along ocean-continent transitions and/ or oceanic lithospheric mantle along slow-spreading ridges depends on the relative position to the underlying up welling asthenosphere.

Exhumation of mantle rocks and formation of (slow) spreading ridges involve two principal geological processes, tectonism and magmatism, which reflect the strain rate and temperaturedependent processes of deformation and adiabatic decompression melting within the Earth. Important extensional environments with direct exposures of mantle rocks are embryonic ocean basins (e.g. the Red Sea) and ocean-continent transition zones (e.g. Iberia margin). Slow-spreading midocean ridges constitute an important extensional environment where mantle rocks are exposed and magma supply is limited (Dick et al. 1984; Cannat 1993, 1996). There, direct outcrop of peridotite on the ocean floor via denudation of the mantle occurs in different environments such as fracture zones, magma-starved rift segments, and oceanic core complexes (e.g. Blackman et al. 1998; Tucholke et al. 1998). Models of continental break-up (rifting) followed by sea-floor spreading (drifting) conventionally separate continental

and oceanic crustal types. Studies from the Red Sea (Bonatti et al. 1983; Piccardo et al. 1988, 1993), fieldwork in the Alps (Lagabrielle et al. 1984; Hermann & Miintener 1996; Manatschal & Nievergelt 1997; Desmurs et al. 2001) and Ocean Drilling Program results from the Iberian margin (Boillot et al. 1995) challenged this view, and it has been proposed (Whitmarsh et al. 2001), that the concept of an abrupt boundary should be replaced by that of a several tens-of-kilometres wide ocean-continent transition zone (OCTZ) of mainly exhumed continental mantle rocks and subordinate mafic intrusions separating thinned continental crust from oceanic crust. Both slowspreading lithosphere and the thinned lithosphere in ocean-continent transitions are different from ophiolites as defined by the Penrose conference (Anonymous 1972) but can better account for many recent observations on ophiolites from the Alps and the Apennines (e.g. Rampone & Piccardo 2000).

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 69-89. 0305-8719/037$ 15 © The Geological Society of London 2003.

70

O. MUNTENER & G.B. PICCARDO

The peculiar stratigraphy of the Alpine -Apennine ophiolites, first described by Steinmann (1927) in the Central Eastern Alps and by Decandia & Elter (1972) in the Ligurian Alps, has led researchers to propose three classes of genetic models: (1) the slow-spreading ridge model (Lagabrielle & Cannat 1990; Lagabrielle & Lemoine 1997); (2) the transform fault model (Gianelli & Principi 1977; Weissert & Bernoulli 1985); (3) the low-angle detachment faulting model (Lemoine et al. 1987; Froitzheim & Eberli 1990; Piccardo et al. 1990; Froitzheim & Manatschal 1996). All these models have some widely accepted analogues in modern oceans. However, Dick et al. (2000) questioned whether on-land analogues of slow-spreading crust such as the peridotites in the Alps can directly be compared with the results obtained from the slow-spreading Southwest Indian Ridge (Leg 176), where more than 1500m of continuous igneous crust have been drilled and which is probably the best known slow-spreading oceanic crust from modern oceanic basins. In any case, the formation of a slowspreading ridge is necessarily preceded by a period of continental rifting and at some point there must be a transition from (largely) amagmatic passive rifting to the formation of igneous crust and finally the establishment of a slowspreading system. As for the Ligurian ophiolites, the subcontinental origin of mantle peridotites was proposed earlier by some workers (Piccardo 1976), who outlined the diversity of the Alpine-Apennine ophiolites in comparison with mature oceanic lithosphere formed at modern mid-ocean ridges. Based on the atypical association of fertile subcontinental-type mantle and mid-ocean ridge basalt (MORB) magmatism, Piccardo (1977) suggested that the Ligurian ophiolites were formed during the early stages of opening of the oceanic basin, following continental rifting, thinning and break-up of the continental crust, and that they were therefore located in a marginal, pericontinental position to the Jurassic oceanic basin. The temperature field in all these 'thin crust regions' is determined by the rate of magma supply. Recent results from deep-sea drilling and dredging indicate that some of the liquids generated in the asthenosphere crystallize on a conductive geotherm in the mantle, and thus the igneous crust is significantly reduced or even absent in OCTZs and slow-spreading ridges (Cannat 1996; Cannat et al. 1997; Bonatti et al 2001; Desmurs et al. 2001). A fundamental remaining question is whether mantle peridotites from marginal settings of ancient and modern oceanic basins provide some systematic information that could shed light on the enigmatic evolution between rifting and the

formation of a slow-spreading system. Although it is well established that many of the Alpine peridotites underwent non-adiabatic uplift and subsolidus evolution from upper-mantle levels to the sea floor (Piccardo 1976; Hoogerduijn-Strating et al. 1993; Rampone et al. 1993; Miintener et al. 2000), followed by the intrusion of small volumes of gabbros and the extrusion of MORB, little is known about the interaction of peridotite with migrating liquids, which must pass through the overlying mantle to form oceanic crust. In this paper we present observations from the Lanzo and Corsica ophiolitic peridotites, summarized below, which show that porous flow of melt and melt-rock reaction are widespread in exhumed ex-subcontinental and oceanic peridotites and are probably related to incipient opening of embryonic ocean basins. In contrast to previous treatments, we consider the effects of rising temperature during melt percolation and impregnation and the development of plagioclase-enriched peridotites. We evaluate the mode and nature of viable mechanisms for plagioclase formation and examine interactions of subcontinental mantle with rising MORB melt fractions. We also discuss the implications of our findings for the highly variable pressure-temperature paths of exhumed peridotites.

General features of the Piedmont Ligurian ophiolites The Ligurian Tethys is believed to have developed by progressive divergence of the European and Adriatic continents, in connection with the Early to mid-Jurassic rifting and the Cretaceous opening of the Northern Atlantic (Dewey et al. 1973). Palaeotectonic reconstructions of the Ligurian Tethys suggest that the Piedmont Ligurian ocean was not wider than 400-500 km (Stampfli 1993) and that the Late Cretaceous to Tertiary plate convergence led to complete closure of the Ligurian Tethys in the Early Tertiary, resulting in the emplacement of fragments of the oceanic lithosphere as west-vergent thrust units in the Alps and east-vergent thrust units in the Apennines. Depending on their stratigraphic, structural and metamorphic characteristics, the ophiolitic sequences have been related to different palaeogeographical settings in the Jurassic-Cretaceous Ligurian Tethys. The Voltri Massif, the Piemontese and Saas Zermatt ophiolites (Fig. 1), which were subducted and recrystallized at eclogite facies conditions, were located west of the subduction zone, whereas some of the Northern Apennine ophiolites (External Ligurides) as well as the Eastern Central Alpine ophiolites (Davos-PlattaMalenco), which underwent low-grade oceanic

A L P I N E - A P E N N I N E PERIDOTITES

71

Fig. 1. Generalized tectonic overview of the Alpine and Northern Apennine ophiolites. Modified from Schaltegger et al. (2002). D, Totalp peridotite; PI, Platta ophiolite; M, Malenco peridotite; ZE, Saas Zermatt ophiolite; Ch, Chenaillet ophiolite; ET, Erro Tobbio peridotite; EL, External Ligurides; IL, Internal Ligurides; T, Tuscany peridotites; CO, Corsica ophiolite; GE, Nappe des gets ophiolites.

and erogenic metamorphism, were located east of the subduction zone, close to the Adria margin.

Field relations and petrography of serpentinites and peridotites The ophiolites of the Piedmont Ligurian ocean show a predominance of a largely serpentinized peridotite basement intruded and/or covered by small or moderate volumes of mafic rocks, and the lack of a 'complete' ophiolite stratigraphy (Fig. 2). Serpentinized peridotites are commonly overlain by ophicalcites, which represent tectonosedimentary breccias related to mantle exhumation (Lemoine et al. 1987; Desmurs et al 2001). In the Eastern Central Alps extensional allochthons of

continental basement rocks and their pre- and synrift sedimentary cover locally overlie the exhumed mantle rocks (Froitzheim & Manatschal 1996; Manatschal & Nievergelt 1997). More importantly, the serpentinized mantle rocks are in places stratigraphically overlain by Jurassic radiolarites, indicating that they must have been uplifted from mantle depth to the sea floor in Mesozoic times. From field and geochemical evidence it appears that in many areas of the Alps and the Apennines (Piccardo 1976; Piccardo et al. 1990; Trommsdorff et al. 1993; Rampone et al. 1995, 1998; Miintener & Hermann 1996; Rampone & Piccardo 2000), the serpentinized peridotite basement represents former subcontinental mantle; however, the chemical composition and the conditions of equilibration of these peri-

72

O. MUNTENER & G.B. PICCARDO

Fig. 2. Generalized structural and stratigraphic relationships of an ancient ocean-continent transition in the Alps (Platta-Totalp-Malenco area in Eastern Switzerland and Northern Italy; after Manatschal & Nievergelt 1997 and Desmurs et al. 2001). This area is characterized by an incomplete ophiolite sequence typical for many Alpine ophiolites, with basalts and post-rift sediments covering the top of exhumed subcontinental mantle. In the proximal (continentward) part of the ocean-continent transition, the exhumed subcontinental mantle rocks are locally overlain by extensional allochthons, stranded klippen of continental basement, pre-rift sediments and synrift marine breccias, emplaced along extensional detachment faults. In the distal (oceanward) part of the ocean-continent transition the mantle rocks are intruded by gabbros and stratigraphically overlain by pillow lavas and breccias. JU, Jurassic; TR, Trias: JL, lower Jurassic; JM, middle Jurassic.

dotites are different from place to place. Many of these ophiolitic peridotite massifs show areas where plagioclase is present and, frequently, rather abundant. Two microtextural settings might be established: (1) granular to porphyroclastic spinel peridotites with only incipient recrystallization of plagioclase; (2) granular to porphyroclastic peridotites with abundant, granoblastic aggregates of plagioclase, interstitial to deformed spinel-facies mantle minerals. The majority of the mantle peridotites of the first group (e.g. Davos, Malenco, Upper Platta, some of the External Ligurides, Erro Tobbio, Tuscany) are fertile, clinopyroxene-rich Iherzolites, whereas depleted, clinopyroxene-poor peridotites are subordinate. Most of these massifs are composed of amphibole-bearing spinel peridotite with abundant (garnet) pyroxenite layers and locally phlogopite-hornblendite veins (Peters 1963; Piccardo 1976; Rampone et al. 1995; Miintener & Hermann 1996; Desmurs 2001). They display a static equilibrium recrystallization under spinel-facies conditions, and the presence of light rare earth element (LREE)-depleted titanian pargasite, in structural and chemical equilibrium with the spinel-bearing assemblage (Vannucci et al. 1995). Thermobarometric estimates on the spinelfacies assemblages yield temperatures in the range of 900-1100 °C and equilibration pressures of 1.0-1.5GPa. Plagioclase recrystallization is rare or non-existent and restricted to porphyroclastic to mylonitic peridotites. There, olivine-plagioclase

symplectites form as subsolidus reaction between pyroxenes and spinel. The second group of peridotites (e.g. Lanzo, Corsica, Lower Platta, Internal Ligurides, Chenaillet) are in general made of clinopyroxene-poor Iherzolites similar to abyssal peridotites; however, fertile peridotites are common. Most of these peridotites contain little or no Ti-amphibole and pyroxenite layers are locally transformed into olivine gabbro. They are strongly enriched in plagioclase within the granoblastic aggregates. Thermobarometric estimates yield high temperatures in the range of 1100-1280 °C and equilibration pressures of less than about 1 GPa. Plagioclase is abundant and forms granoblastic aggregates, interstitial to deformed spinelfacies mineral assemblages.

Plutonic rocks Gabbroic rocks occur mainly as small intrusive bodies and dykes in the peridotites and are usually discordant to the high-temperature mantle structures. Their compositions range from ultramafic cumulates to highly differentiated plagiogranites and represent the crystallization products of a typical low-pressure, tholeiitic fractionation of MORB-type parental magmas (Serri 1980; Hebert et al. 1989; Tiepolo et al. 1997; Tribuzio et al. 2000; Desmurs et al. 2002). The gabbros show the crystallization sequence olivine —» plagioclase —> clinopyroxene, and covariations of forsterite (fo) content in olivine, anorthite (an) content in plagi-

ALPINE-APENNINE PERIDOTITES oclase and Mg-number in clinopyroxene, which are typical of low-pressure crystallization of olivine tholeiites. Clinopyroxenes of primitive cumulates have rather flat heavy REE (HREE) to middle REE (MREE) patterns, at about (910) X Cl, and LREE depletion (CeN/SmN = 0.21-0.29). Calculated liquid compositions from the most primitive samples indicate a clear MORE affinity, in agreement with the Sr and Nd isotope ratios of some ol-gabbros and their clinopyroxenes (Rampone et al 1998; Bill et al 2000). U-Pb ages of zircons from highly differentiated Fe-Ti gabbros exhibit a surprisingly narrow window of crystallization ages from 166 to 160 Ma (Schaltegger et al. 2002), whereas some plagiogranites of the Western Alps and Apennines are distinctly younger (153-148 Ma, Borsi et al. 1996; Costa & Caby 2001). This suggests that regional-scale upwelling and partial melting of a MORB-type asthenospheric source started in the Mid-Jurassic.

Extrusive rocks Basaltic rocks are common in Alpine ophiolites and occur as pillows or massive flows and as discrete dykes intruding deformed gabbros and mantle peridotites (Fig. 2). Petrological and geochemical studies have provided evidence of their overall tholeiitic composition and MORB affinity, ranging from T-MORB to N-MORB (Venturelli et al. 1981; Beccaluva et al. 1984; Ottonello et al. 1984; Rampone et al. 1998; Bill et al. 2000; Desmurs et al. 2002). The most primitive basalts show either moderate LREE fractionation (CeN/SmN = 0.6) or almost flat to slightly LREEenriched REE spectra, and HREE abundances at about 10 X Cl. They have fairly homogeneous Nd isotopic ratios, consistent with their MORB affinity, but variable Sr isotopic ratios (up to 0.7085), which are related to oceanic sea-water alteration (Rampone et al. 1998; Bill et al. 2000; Schaltegger et al. 2002). Geochemical modelling indicates that the most primitive T-MORB and NMORB-type basalts are consistent with melts generated by low to moderate degrees of fractional melting of a MORB-type asthenospheric mantle source (Vannucci et al. 1993a); however, the source of some basalt is enriched in incompatible elements (Desmurs et al. 2002). This compositional variation seems to correlate with the spatial distribution of the mafic rocks within the oceancontinent transition whereby mafic rocks with TMORB signatures occur close to the continental margin whereas N-MORB signatures are predominantly found oceanwards (Desmurs et al. 2002).

73

Observations from the Lanzo and Corsica peridotites Field observations The Lanzo and Monte Maggiore (Corsica) peridotites comprise the mantle section of partially dismembered ophiolites exposed in the Western Alps and the Northern Apennines (Fig. 1). The peridotites are composed mainly of massive plagioclase peridotites, and minor spinel peridotites, harzburgites and dunites. At Lanzo, plagioclase peridotites were thought to be formed either as the residuum of low degrees of partial melting (Bodinier 1988) or as a result of melt formation and incomplete melt extraction and crystallization in an upwelling diapir (Boudier & Nicolas 1972; Boudier 1978; Nicolas 1986). The peridotite is rich in plagioclase and clinopyroxene, and spinels commonly have high Cr/(Cr + Al) ratios. These characteristics are similar to some plagioclase peridotites dredged from slow-spreading ridges and along fracture zones (Dick 1989; Cannat et al. 1997; Seyler & Bonatti 1997; Tartarotti et al. 2002). Isotopic studies of the Lanzo peridotites have pointed out important differences between the Northern and Southern bodies (Bodinier et al. 1991). The Southern body of the Lanzo Massif has been interpreted as an asthenospheric diapir that rose from the garnet stability field and was emplaced in the early Mesozoic, during the opening of the Piedmont Ligurian basin. The Northern body has been considered a fragment of the subcontinental lithosphere that became isolated by the convective mantle at 400-700 Ma (Bodinier et al. 1991). In Lanzo, the peridotites are cut by an older suite of spinel websterites and a younger suite of discordant dunite, followed by various sets of gabbroic veins and dykes. Dunite cuts and locally replaces earlier spinel websterites (Fig. 3a), as indicated by trains of Cr-spinel that are continuous with the surrounding spinel websterites (see also Boudier & Nicolas 1972; Boudier 1978). New field observations show that some discordant dunites locally contain small interstitial clinopyroxene (Fig. 3b), and large clinopyroxene megacrysts (crystals of more than 1 cm in diameter, Fig. 3d) sometimes associated with plagioclase (Fig. 3c). In places, large, euhedral clinopyroxenes form aggregates a few millimetres wide or (deformed) gabbroic veinlets (Fig. 3e and f), similar to the 'indigenous' dykes described by Boudier & Nicolas (1972) and Boudier (1978). Locally, medial pyroxenite dykes in dunite have also been observed. The gabbroic dykes can be separated into two groups (Boudier & Nicolas 1972; Boudier 1978). Type 1 is an older 'indigenous'

74

O. MUNTENER & G.B. PICCARDO

Fig. 3. Field aspects of plagioclase peridotites and dunites from Lanzo South (Northern Italy), showing plagioclase peridotites, cut by spinel dunites. (a) Harzburgite-dunite contact (southern flank of Monte Musine'). It should be noted that the foliation is discordant to the dunite-harzburgite contact, (b) Interstitial clinopyroxene (green Crdiopside) within dunite. (c) Clinopyroxene (cpx) + plagioclase (pig) cluster in dunite (Monte Musine'). (d) Clinopyroxene megacryst in dunite (Monte Musine'). (e) 'Indigenous' microgabbroic vein and a clinopyroxene megacryst within dunite (Monte Arpone). (f) Weakly deformed 'indigenous' Mg-gabbro dykelet cutting dunite (Mt Arpone). At the lower right, the gabbro is discordant to the peridotite-dunite contact. Subsequent high-temperature ductile deformation formed dunite mylonite. du, dunite; hz, harzburgite; pig 1hz, plagioclase Iherzolite.

group of olivine gabbronorite, frequently occurring in en echelon fuzzy contacts with the surrounding Iherzolites and dunites. This type is restricted to the southwestern part of the massif

(Compagnoni et al 1984; Pognante et al. 1985). Type 2 is an intrusive group of troctolites, (olivine) gabbros, gabbronorites to oxide gabbros, with sharp contacts and chilled margins towards

A L P I N E - A P E N N I N E PERIDOTITES

the peridotite. These dykes can be found all over the massif (Compagnoni et al. 1984; Pognante et al. 1985), cut across dunites and plagioclase peridotites, and are generally undeformed. Thus it seems that penetrative high-temperature deformation of the peridotite ceased between the formation of type 1 and 2; however, both are locally mylonitized and partially hydrated under upper amphibolite- to granulite-facies conditions (Compagnoni et al. 1984; Pognante et al. 1985). The geochemistry of the mafic rocks reveals that most gabbros represent cumulates with little or no trapped liquid, indicating efficient extraction of derivative liquids (Bodinier et al. 1986). In addition, late porphyritic basaltic dykes of N-MORB affinity (Pognante et al. 1985) cut the peridotites and gabbros. The extracted melts had a T-MORB and a T- to N-MORB composition in the north and south, respectively, (Bodinier 1988), similar to basalts from the Ligurian Alps (Beccaluva et al. 1984). In Corsica, the Monte Maggiore peridotites are strongly depleted, with a spinel-facies granular assemblage: they are clinopyroxene-poor, refractory spinel Iherzolites, which are interpreted as mantle residua after MORB-type partial melting processes (Jackson & Ohnenstetter 1981; Rampone et al. 1997). Preliminary Sm/Nd isotope data provide a mid-Jurassic (165 Ma) DM (depleted mantle) model age of depletion (Rampone 2002). In places these peridotites contain plagioclase and show evidence for trapped melt crystallization (Rampone et al. 1997), In the Monte Maggiore region, peridotites with oriented and diffuse impregnation (Fig. 4a and b) are cut by discordant dunites (Fig. 4e), followed by the intrusion of gabbroic veins (Fig. 4c) and metrescale pockets of mafic-ultramafic cumulates, composed mainly of olivine gabbronorites (Fig. 4d). The cumulates cut dunite-peridotite contacts and the existing oriented impregnation. Locally these pockets dominate volumetrically and the former peridotite is completely replaced by gabbronorite mineral assemblages consisting of euhedral ortho- and clinopyroxene and interstitial plagioclase (Fig. 4d). Another common feature at Monte Maggiore is the formation of coarsegrained and undeformed late gabbroic dykes (with crystals exceeding 5 cm in diameter), which crosscut deformed peridotites and gabbronorite cumulates (Fig. 4f).

Residual mantle mineral assemblages The studied spinel- and plagioclase-bearing Iherzolites are mainly porphyroclastic peridotites comprising a deformed mantle assemblage and less deformed or undeformed interstitial igneous as-

75

semblages. Rare samples are nearly plagioclasefree spinel Iherzolites and show textures typical of common spinel peridotites. Olivine occurs in large grains (up to 1 cm) and pyroxenes form millimetre- to centimetre-scale porphyroclasts. Spinel is brown and Al-rich, and commonly shows hollyleaf shapes (Fig. 5a). Deformation-induced undulatory extinction and gliding in olivine is common. Pyroxenes are commonly deformed and show fine exsolution lamellae of the complementary pyroxene. In samples not affected by melt impregnation, Al-rich spinel is locally intergrown with orthopyroxene, producing a texture similar to that of garnet breakdown (Vannucci et al. 1993b). In the same sample, a small rim of olivine + plagioclase is locally developed between orthopyroxene and spinel, according to the reaction orthopyroxene + spinel —> olivine + plagioclase (Fig. 5b) This microstructure probably formed during decompression from the spinel peridotite to the plagioclase peridotite field before melt impregnation and melt-rock reaction, and provides rare evidence for a metamorphic origin of the plagioclase in the Lanzo peridotite.

Impregnation textures The sequence of igneous microstructures in the spinel peridotite is well established. Early meltrock reaction dissolved clinopyroxene along grain boundaries and precipitated orthopyroxene + plagioclase around and within clinopyroxene (Fig. 5c and d). Textural relationships indicate cotectic crystallization of plagioclase + orthopyroxene (Fig. 5d). These intergrowths are not deformed, contrary to the original mantle clinopyroxene. A similar structure can be observed in spinel websterites. These features indicate that the migrating liquid crystallized clinopyroxene-free, orthopyroxene-rich gabbronoritic microgranular aggregates. Orthopyroxene partially replaced mantle minerals, showing concave contacts against the peridotite clinopyroxene (Fig. 5e). However, in many samples crystallization of two pyroxenes and plagioclase is also common. This is illustrated in Figure 5e and f, where undeformed and interstitial clinoand orthopyroxene crystallized between large mantle minerals. Large kinked mantle olivine is recrystallized close to the interstitial orthopyroxene (Fig. 5f), supporting the general observation that the impregnating assemblages are less deformed than the precursor mantle assemblage. In places the igneous domains consist of plagioclase patches, replacing spinel, together with granular orthopyroxene + olivine ± clinopyroxene (Fig. 5g). These microgabbroic aggregates form xenomorphic granoblastic mosaic textures between mantle minerals generally several millimetres to

76

O. MUNTENER & G.B. P I C C A R D O

Fig. 4. Field aspects of plagioclase peridotite and gabbronorite from the Monte Maggiore peridotite (Corsica), (a) Impregnated peridotites with interstitial plagioclase surrounding mantle minerals. Locally plagioclase (pig) + orthopyroxene (opx) coalesce, forming gabbronorite veinlets. (b) Oriented plagioclase-rich impregnation in peridotite discordantly cut by a cpx-rich gabbronorite dykelet, related to the cumulate pods. It should be noted that in the upper portion of the outcrop the gabbronorite dyke ends in a millimetre-scale veinlet. (c) Impregnated peridotite intruded by irregular gabbroic veins or pods related to the cumulate suite. Euhedral green cpx and interstitial pig in the gabbroic veins or pods should be noted, (d) Euhedral olivine + opx + cpx and anhedral plagioclase in a gabbronorite cumulate, (e) Discordant dunite-plagioclase peridotite contact cut by a gabbro dykelet. The euhedral clinopyroxene megacrysts in the dykelet should be noted. The foliation in both peridotite and dunite runs approximately perpendicular to the contact. Locally, millimetre-scale plagioclase seams of the gabbro intrude the surrounding dunite and plagioclase peridotite. (f) Coarse-grained Mg-gabbro dyke with chilled margins cuts spinel peridotite.

A L P I N E - A P E N N I N E PERIDOTITES

77

Fig. 5. Selected thin section specimens from the Lanzo South area, (a) Intergrowth of orthopyroxene (opx) and spinel (spl) in a spinel peridotite from Mt Arpone. Of particular note is the overall shape of this intergrowth, recalling former garnet. It should be noted also that spl-opx contacts are fresh without any sign of reaction, (b) Opx-spl contacts separated by a small rim of olivine + plagioclase (completely altered). This indicates that orthopyroxene and spinel are unstable and transform into a (metamorphic) assemblage of olivine (ol) + plagioclase (pig), (c) Corrosion of exsolved and deformed mantle clinopyroxene by opx + pig intergrowths. (d) Intergrowth of opx + pig indicating cotectic crystallization of the two phases, (e) Undeformed interstitial opx + cpx separating large, kinked mantle olivine. (f) Rim of interstitial opx replacing mantle cpx. Of particular note are the concave opx-cpx contacts. It should be noted also that kinked olivine is replaced by undeformed olivine in the left part of the photomicrograph. This demonstrates that cpx + liquid reacted to form ol + opx. (g) Undeformed micro-gabbronorite aggregate between deformed mantle minerals. The anhedral shape of pig and opx between granular euhedral olivine should be noted. Cpx forms small interstitial grains. These textural relationships demonstrate the crystallization sequence ol —> opx + plag —> cpx and indicate that migrating liquids were saturated in opx before cpx. (h) anhedral cpx along triple point of olivine in discordant dunite.

78

O. MUNTENER & G.B. PICCARDO

centimetres wide. The granoblasts do not show a preferred orientation and form troctolitic to gabbronoritic assemblages. Interstitial pyroxenes do not show exsolution of the complementary pyroxene, are virtually undeformed and show concave contacts to the host peridotite minerals (Fig. 5e and f). In other samples olivine, clinopyroxene and orthopyroxene form euhedral crystals surrounded by interstitial plagioclase. Mantle clinopyroxene is seemingly unreacted. Microstructures in the dunites are characterized by coarse-grained olivine (up to 2 cm in size) and more or less rounded spinels, as described previously (Boudier & Nicolas 1972; Boudier 1978; Nicolas 1986). In addition, many dunites contain interstitial clinopyroxene, which is exclusively found along olivine triple junctions (Fig. 5h).

Geochemical data We analysed crystals of peridotite and gabbroic samples by electron microprobe and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). Preliminary results are shown in Tables 1 and 2 and illustrated in Figures 6 and 7. In terms of major element compositions, the main variation in clinopyroxenes in peridotites is reflected in the Al and Ti contents, with high Al and low Ti in spinel peridotites, and low Al and higher Ti in plagioclase peridotite. Mg numbers (molar Mg/(Mg + Fetot)) range from 0.89 to 0.92 (Table 1). Na2O contents are invariably low (<0.5 wt%). Plagioclase in impregnated peridotites is generally calcic (Angg-94) for Monte Maggiore peridotites (Rampone et al. 1997), whereas it is much more variable in the Lanzo peridotite, ranging from An9p to An67. Figure 5 illustrates that Monte Maggiore minerals have similar REE slopes to those of Lanzo South, but generally lower concentrations. There is a first-order difference between spinel and plagioclase peridotite in both clinopyroxene and orthopyroxene compositions. Spinel peridotite clinopyroxene has a slightly to strongly LREEdepleted, chondrite-normalized pattern, with essentially flat MREE to HREE and no negative Eu anomaly. These characteristics are similar to those of some spinel peridotites in other Alpine peridotites (Rampone et al. 1996; Miintener et al. 2002) and of abyssal peridotites (Johnson et al. 1990). Plagioclase peridotite clinopyroxene REE patterns (both porphyroclastic relics and igneous grains in granular aggregates) are generally convex upward with a significant MREE enrichment (up to 20 X Cl), GdN/YbN >1 and a weak to significant negative Eu anomaly (Fig. 6), which increases from core to rim, indicating equilibration with plagioclase. In addition, most of the trace ele-

ments (i.e. Ti, Sc, V, Zr, Y) in clinopyroxene from plagioclase peridotites are enriched with respect to the precursor clinopyroxene in spinel Iherzolites (Table 1). Orthopyroxene follows the trends given by clinopyroxene. Plagioclase from both Lanzo and Monte Maggiore peridotites show similar REE chondrite-normalized patterns with significant LREE fractionation (CeN/NdN < 0.5) and very low Sr (<10ppm) and Na (Na2O <1.00wt%) contents (see also Rampone et al. 1997); however, samples with nearly flat or slightly enriched LREE, and relatively higher Sr (up to 150ppm) and Na (Na2O up to 4.00 wt%) contents are found in Lanzo. Within-sample variations and within-group variations are much smaller than variations between spinel peridotites and plagioclase peridotites. Simple modelling indicates that about 6% of polybaric fractional melting is sufficient to produce the depleted REE signatures of the Lanzo spinel peridotites, whereas slightly higher degrees (about 8%) are necessary for the Monte Maggiore peridotites. The composition of clinopyroxene thus reflects near-fractional melting processes in the spinel peridotite field before impregnation and melt-rock reaction at lower pressure. It is more difficult to evaluate the liquid composition, which modified the peridotite mineral compositions and formed the igneous plagioclase-pyroxene-bearing assemblages. Rampone et al. (1997) assumed that the trace element signatures of plagioclase peridotites from Monte Maggiore resulted from variable proportions of trapped liquid within the peridotites. Modelling of the liquids that impregnate the peridotites suggests that they most probably consisted of unmixed single melt increments that originated from deeper levels of the mantle by a near-fractional melting process (Rampone et al. 1997). In striking contrast to the clinopyroxenes from plagioclase peridotite, interstitial clinopyroxene in dunite from Lanzo shows no Eu anomaly, is much less depleted in LREE (Fig. 6), and shows lower MREE and HREE contents (at <10 X Cl). It also has higher incompatible element contents (Table 1). Calculated liquids in equilibrium with clinopyroxenes have REE slopes and concentrations similar to MORB crystallized from low percentage aggregate liquids (<5%). In addition, spinels in Lanzo dunite (Boudier & Nicolas 1972; Boudier 1978; Nicolas 1986) are similar to spinels from MORB (Dick & Bullen 1984). Thus both spinel and interstitial clinopyroxene compositions suggest that Lanzo dunite formed in equilibrium with liquids similar to MORB, whereas the plagioclase peridotites formed from single melt increments that were trapped within the peridotites.

Table 1. Representative analyses of minerals from Lanzo South and Monte Maggiore peridotites Monte Maggiore spinel peridotites Sample: Mineral:

CC1 CC1 Cpx core Opx core

CC1 Sp

wt% SiO2 Ti02 Cr2O3 A12O3 Fe203 FeO MnO MgO NiO CaO Na20 K2O

49.53 0.28 1.18 6.12 0.00 3.31 0.15 15.31 0.00 23.02 n.d. n.d.

54.43 0.13 0.79 4.91 0.00 6.54 0.15 31.32 0.00 1.38 n.d. n.d.

0.00 0.11 20.71 47.53 1.98 12.03 0.00 18.27 0.06 0.00 n.d. n.d.

99.65

100.69

98.9 Total REE + Y(ppm) 0.005 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Y Er Tm Yb Lu

0.071 0.047 0.740 0.684 0.311 1.240 0.260 1.923 0.418 12.450 1.228 0.187 1.223 0.170

n.d., not determined.

0.001 0.012 0.012 0.008 0.040 0.015 0.200 0.059 1.642 0.226 0.043 0.250 0.059

Plagioclase peridotites CC2 Cpx c

Lanzo South spinel peridotites

CC2 Opx c

CC2 Pig

CC2 Spl

49.72 0.27 1.31 6.36 0.00 3.75 0.03 16.71 0.00 21.24 n.d. n.d.

55.09 0.28 0.98 3.67 0.00 6.38 0.25 33.59 0.00 0.78 n.d. n.d.

45.56 0.00 0.00 35.01 0.15 0.00 0.00 0.00 0.00 18.12 1.12 0.05

0.00 0.55 34.15 31.29 4.06 15.31 0.29 14.18 0.00 0.00 n.d. n.d.

50.52 0.17 1.27 6.79 0.00 2.82 0.10 16.27 0.01 22.02 0.01 0.10

99.36

101.02

100.01

99.88

100.08

0.006 0.149 0.095 1.056 0.927 0.355 1.791 0.389 2.915 0.675 19.368 1.900 0.288 1.682 0.257

0.002 0.001 0.002 0.025 0.039 0.015 0.158 0.047 0.412 0.110 2.970 0.415 0.060 0.380 0.062

0.005 0.074 0.156 0.046 0.265 0.061 0.008

0.150

LSI LSI Cpx pc c Opx pc c

0.010 0.213 0.120 1.093 0.967 0.397 0.853 0.340 2.387 14.853 0.550 1.620 0.233 1.490 0.193

Plagioclase peridotites

LSI Spl

53.61 0.24 0.81 4.86 0.00 5.93 0.10 32.76 0.05 1.51 0.00 0.11

40.99 0.07 0.01 0.00 0.00 9.48 0.14 48.77 0.09 0.20 0.00 0.06

0.00 0.19 16.58 52.05 0.86 10.91 0.00 18.96 0.11 0.13 0.00 0.07

49.71 0.55 1.30 6.49 0.00 3.38 0.10 15.30 0.06 22.29 0.55 0.22

50.83 0.62 1.36 4.39 0.00 3.44 0.12 17.13 0.00 21.59 0.73 0.17

51.28 0.56 1.44 3.35 0.00 3.31 0.18 16.98 0.18 22.14 0.00 0.19

54.14 0.35 0.93 3.52 0.00 6.67 0.18 32.67 0.03 1.61 0.00 0.10

54.95 0.41 0.71 2.43 0.00 6.46 0.15 33.36 0.16 1.37 0.00 0.15

55.06 0.34 0.74 2.32 0.00 6.47 0.14 33.74 0.02 1.26 0.00 0.11

99.98

99.81

99.86

99.94

100.38

99.61

100.20

100.15

100.20

0.010 0.020 0.020 0.060 0.030 0.200 0.070 1.950 0.250 0.040 0.320 0.050

LS2 LS2 Cpx pc c Cpx pc r

0.021 0.526 0.272 2.617 1.883 0.711 2.993 0.640 4.167 1.016 26.940 2.873 0.403 2.733 0.346

0.023 0.535 0.310 3.105 1.940 0.625 3.650 0.705 4.400 1.060 27.030 2.700 0.365 2.220 0.295

LS2 Cpx gb

LS2 LS2 LS2 Opx pc c Opx pc r Opx grain

LSI Ol

0.020 0.487 0.274 2.503 1.910 0.700 3.383 0.643 4.700 1.060 25.600 2.720 0.360 2.370 0.297

0.019 0.003 0.045 0.083 0.014 0.240 0.054 0.570 0.150 3.900 0.610 0.079 0.740 0.120

0.002 0.005 0.033 0.083 0.021 0.190 0.043 0.470 0.150 3.700 0.480 0.076 0.660 0.110

0.012 0.005 0.100 0.016 0.160 0.041 0.460 0.152 3.500 0.500 0.092 0.690 0.105

LS2 PI

LS2 Ol

LS2 Sp

45.20 0.14 0.08 34.58 0.23 0.00 0.00 0.00 0.00 17.64 1.55 0.22

40.66 0.09 0.06 0.00 0.00 10.23 0.18 48.85 0.10 0.17 0.00 0.05

0.00 0.97 36.20 27.87 5.50 16.76 0.00 13.19 0.13 0.12 0.00 0.11

99.64

100.39

100.85

0.011 0.170 0.045 0.330 0.100 0.280 0.089 0.012 0.066 0.012 0.210 0.021 0.012

O. MUNTENER & G.B. PICCARDO

80

Table 2. Representative analyses of minerals from Lanzo South replacive dunite, gabbro veinlets and Monte Maggiore gabbronorite cumulates

Sample: Mineral:

Lanzo gabbro veinlets

Monte Maggiore cumulates

Lanzo dunites La70 Cpx

La71 Cpx core

La71 Cpx rim

La71 Pig

CR5/2 Cpx core

CR5/2 Cpx rim

CR5/2 Opx core

CR5/2 Pig

wt% SiO2 Ti02 Cr203 A1203 Fe2O3 FeO MnO MgO NiO CaO Na2O K2O

51.88 0.34 1.64 4.64

50.41 0.37 1.56 4.53

51.66 0.45 1.66 3.10

52.58 0.11 0.14 29.44

50.21 0.31 1.48 5.91

52.3 0.28 1.23 2.93

54.67 0.18 0.91 4.01

3.11 0.15 16.07 0 22.3 0.58 0.12

3.41 0.02 16.59 0.00 22.19 0.80 0.12

3.34 0.12 17.30 0.06 22.20 0.00 0.10

0.20 0.00 0.00 0.23 12.70 3.62 0.19

3.25 0.11 15.96 0.06 22.79 n.d. n.d.

2.89 0.19 16.98 0.21 23.3 n.d. n.d.

6.73 0.16 32.89 0.05 1.13 n.d. n.d.

45.41 0 0 34.65 0.25 0 0 0 0 17.75 0.81 0.08

Total

100.83

100.00

99.99

99.21

100.08

100.31

100.73

98.95

REE + Y (ppm) La Ce Pr Nd Sm Eu Gd Tb Dy Ho Y Er Tm Yb Lu

0.232 1.383 0.334 2.100 1.058 0.434 1.573 0.303 2.127 0.481 12.643 1.273 0.202 1.266 0.172

0.220 1.407 0.384 2.237 1.140 0.555 1.877 0.367 2.324 0.550 14.560 1.570 0.215 1.405 0.202

0.224 1.313 0.335 2.272 1.276 0.494 1.665 0.359 2.362 0.504 13.650 1.367 0.215 1.324 0.179

0.317 0.796 0.114 0.403 0.075 0.380 0.069 0.012 0.054 0.009 0.201 0.019 0.000 0.017 0.003

0.020 0.190 0.080 0.780 0.720 0.330 1.310 0.290 2.070 0.510 12.700 1.320 0.200 1.340 0.190

0.020 0.240 0.090 0.980 0.860 0.410 1.590 0.320 2.530 0.540 15.000 1.550 0.240 1.490 0.200

0.020 0.010 0.070 0.065 0.030 0.170 0.050 0.390 0.090 2.590 0.290 0.050 0.450 0.070

0.020 0.090 0.020 0.110 0.020 0.090 0.020 0.005

0.130 0.013

n.d., not determined

The clinopyroxene and plagioclase data of megacrysts and indigenous gabbroic dykelets from Lanzo are compared with those of the Monte Maggiore gabbroic veinlets and cumulates in Figure 7 and listed in Table 2. Pyroxene and olivine Mg number ranges from 0.89 to 0.91, indicating that the intruding liquids were rather primitive. Plagioclase in indigenous gabbronorite veins and dykelets in Lanzo is relatively sodic (An54-66) and enriched in Sr (500-750 ppm). Plagioclase in the Monte Maggiore mafic ultramafic cumulates and gabbronorite dykelets is in turn extremely poor in Sr (20-30 ppm) and Carich (An88-96). Lanzo clinopyroxenes have REE concentrations and slopes strikingly different from those for the Monte Maggiore area (Fig. 7). Liquids in equilibrium with clinopyroxene from these dykelets and cumulate pods indicate that: (1) the Lanzo dykelets have REE slopes and concentrations corresponding to low-degree (2-

3%), single melt increments after fractional melting of an asthenospheric mantle source; (2) the primary liquids of the Monte Maggiore dykelets and mafic-ultramafic cumulates are compatible with crystallization from single melt fractions produced by moderate degrees (6-7%) of fractional melting. Thus, whereas the Lanzo data are generally compatible with crystallization from low percentage (less than 5%) aggregate MORE, the data from Monte Maggiore suggest that single melt fractions remained isolated during transport and crystallized strongly depleted gabbroic cumulates, similar to those from the Mid-Atlantic Ridge, Deep Sea Drilling Project Site 334 (Ross & Elthon 1993).

Discussion and conclusions Our field observations and geochemical data suggest that ex-subcontinental and oceanic litho-

ALPINE-APENNINE PERIDOTITES

Fig. 6. Representative REE concentrations in minerals from the plagioclase-free spinel Iherzolites and plagioclase-enriched impregnated Iherzolites from (a) Monte Maggiore, Corsica and (b) Lanzo South. Data are normalized to Cl chondrite of Anders & Ebihara (1982). (a) Monte Maggiore (Corsica), (i) Spinel Iherzolite: 1, cpx porphyroclast core; 2, opx porphyroclast core. (2) Impregnated Iherzolite: 3, cpx porphyroclast core; 4, opx porphyroclast core; 5, opx interstitial grain; 6, cpx interstitial grain; 7 and 8, interstitial pig. (b) Lanzo South, (i) Spinel Iherzolite: 1, cpx porphyroclast core; 2, opx porphyroclast core. (2) Impregnated Iherzolite: 3, cpx porphyroclast core; 4, cpx porphyroclast rim; 5, cpx interstitial grain; 6, opx porphyroclast core; 7, opx porphyroclast rim; 8, opx interstitial grain; 9, interstitial pig-

spheric mantle may be substantially modified by migrating magmas during opening of embryonic ocean basins. The occurrence of melt-rock reaction and trapped liquids within peridotite is well known from xenoliths (Menzies et al. 1987), peridotite massifs (Van Der Wai & Bodinier 1996) and studies of present-day oceanic settings (Dick 1989; Elthon 1992; Cannat & Casey 1995; Cannat et al. 1997; Tartarotti et al. 2002), and it has been suggested that upwelling hot asthenosphere causes

81

Fig. 7. Representative REE concentrations in minerals from (a) gabbroic cumulates of Monte Maggiore and (b) replacive dunites and gabbro veinlets in Lanzo South.

chemical and thermal modifications of the overlying lithosphere. However, in the past, the importance of melt migration processes has been underestimated in explaining the evolution of the Piedmont Ligurian ophiolites, and only a few studies have addressed this topic (Rampone et al. 1997; Dijkstra et al 2001). The reason was that the origin of plagioclase was explained by subsolidus formation during non-adiabatic decompression in a closed system (Piccardo 1976; Hoogerduijn-Strating et al. 1993; Rampone et al. 1993, 1995). This might be the case in some places; however, lithospheric extension was accompanied by almost adiabatic upwelling of the underlying asthenosphere, which underwent decompression melting and produced MORB-type liquids of mid-Jurassic age (e.g. Schaltegger et al. 2002). As shown in the previous sections, liquids produced by the upwelling asthenosphere migrated into the overlying peridotite, reacted with, and impregnated the lithospheric mantle. Melt migration evolved from diffuse porous flow to focused porous flow and finally to dyking. This is schema-

82

O. M U N T E N E R & G.B. P I C C A R D O

tically illustrated in Figure 8. Below, we discuss the compositions of the primary liquids and present some explanations for the variable melt migration mechanisms. These explanations are presented as a group of alternative hypotheses, which are not mutually exclusive.

Porous flow of melt during incipient opening of the Piedmont Ligurian ocean Field and microstructural relationships demonstrate that migrating melts produced nearly pervasive impregnation of the lithospheric mantle in the form of small irregular veins and microgranular aggregates of undeformed igneous minerals between deformed mantle minerals. The length scale over which the magmas continued to ascend by porous flow might be several kilometres, as suggested by the ubiquitous occurrence of plagio-

Fig. 8. Generalized evolution of subcontinental peridotites from embryonic ocean basins above an upwelling asthenosphere, as evidenced from the Piedmont Ligurian ophiolitic peridotites. (a) Summary of the situation before exhumation. Peridotite shows high-temperature foliation and several generations of pyroxenite dykes. Temperatures are well below the solidus of peridotites and correspond to a conductively cooled mantle. Model ages of such peridotites range from Proterozoic to Permian (Bodinier et al. 1991; Rampone et al. 1995, 1996). (b) Pervasive melt infiltration and melt-rock reaction at rising temperatures. Ascending liquid from below reacted extensively with its surrounding peridotite, producing orthopyroxene-saturated, cpx-undersaturated compositions, followed by saturation of both pyroxenes. Crystallization products are millimetre- to centimetrescale depleted olivine gabbronorites. (c) Focused porous melt flow in dunite channels at a thermal maximum close to the peridotite solidus. Dunites are discordant and replace plagioclase peridotites and spinel websterites. The transition from divergent to focused flow of melt might indicate that the competing effects between heating of the mantle by ascending magmas from the underlying hot asthenosphere and cooling by exhumation are dominated by the rising isotherms, (d) Beginning of cooling would gradually fill the dunite conduits with interstitial cpx, cpx and plagioclase megacrysts, and eventually change the melt migration mechanism from focused porous flow to cracks. These cracks are probably represented by the small 'indigenous' gabbroic veinlets. It should be noted that the cpx crystallizing in dunite is close to equilibrium with MORB. (e) Continued cooling and exhumation leads to the formation of kilometre-scale gabbroic dykes, discordant to the previous igneous features. These gabbroic dykes range from primitive troctolite to Fe-Tigabbros, indicating the increasing contribution of crystal fractionation on the evolution of magmas at falling temperatures.

clase-bearing peridotite in the Lanzo massif (e.g. Boudier 1978). The impregnating melts either: (1) reacted with mantle clinopyroxenes, forming symplectitic orthopyroxene + plagioclase reaction rims, partially replacing the deformed mantle clinopyroxenes, and crystallized as clinopyroxenefree, orthopyroxene-rich gabbroic micro-granular aggregates, or (2) did not react with mantle clinopyroxene and crystallized as orthopyroxenerich, clinopyroxene-bearing gabbroic microgranular aggregates.

ALPINE-APENNINE PERIDOTITES

The porphyroclastic mantle pyroxenes in the impregnated Monte Maggiore, Lanzo and Internal Ligurides peridotites have trace element compositions unlike residues of fractional melting and are enriched in many trace elements (i.e. REE, Ti, Sc, V, Zr, Y) with respect to porphyroclastic pyroxenes from spinel Iherzolites. Clinopyroxene shows convex-upward REE spectra, with a significant MREE enrichment, and both pyroxenes generally show a negative EUN anomaly. Core-rim analyses of deformed and reacted porphyroclastic clinopyroxene and the interstitial undeformed, granular igneous pyroxene have very similar trace element compositions. This indicates that (1) pyroxenes in the impregnated peridotites attained trace element equilibrium with the migrating liquid; (2) the impregnating liquids were significantly enriched in many trace elements. Early crystallization and abundance of orthopyroxene in the interstitial magmatic micro-gabbroic aggregates suggest that the impregnating melts were silica-saturated. Thus, the primary liquids must have interacted with the ambient peridotite during upward migration in the mantle column, where they dissolved mantle pyroxenes and crystallized olivine, attaining orthopyroxene saturation. Magmatic plagioclase and clinopyroxene in the interstitial granular aggregates have high Mg number (Mg/(Mg + Fe)), are strongly LREE depleted, and have very low Sr contents, significantly different relative to aggregated MORE. These features indicate that the parental melts were strongly depleted in the most incompatible elements. This interpretation suggests that the primary liquids represent depleted single melt increments, produced during fractional melting of an asthenospheric mantle source (Rampone et al. 1997). A likely alternative is that the depleted compositions of the crystallizing phases record their open-system provenance. The trace element composition of both plagioclase and clinopyroxene might be attained by liquid rising adiabatically through the mantle and reacting with the surrounding host rocks. Such liquids would dissolve pyroxenes and crystallize olivine. As discussed by Kelemen et al. (1995), the crystallization sequence of liquids that reacted with mantle peridotite depends on the relative effects of reaction with the surrounding host rocks and cooling. For relatively rapid reactions and slow cooling, liquids might quickly become saturated in orthopyroxene. More rapid cooling, or slower pyroxene dissolution would produce less silica-rich liquid compositions. Continued melt-rock reaction and/or cooling finally led to saturation in two pyroxenes. Thermometric estimates based on trace element (Sc, V) distribution between coexisting ortho- and clinopyroxene in impregnated peridotites (Seitz

83

et al. 1999) indicate high temperatures (c. 12001300°C), close to the peridotite dry solidus, during melt percolation and impregnation. The transition from diffuse porous flow to focused porous flow is accompanied by the formation of discordant dunite (Fig. 8c). Textural evidence for a replacive origin of dunite in the Lanzo peridotite was first given by Boudier & Nicolas (1972). Dunite cuts and locally replaces earlier spinel websterites, as indicated by trains of Crspinel that are continuous with the surrounding spinel websterites. Melt percolating in the dunite channels sporadically crystallized small interstitial clinopyroxenes and, later, trails of clinopyroxene megacrysts (up to 2 cm in size), which are precursors to the early gabbroic veins and dykelets. Calculated liquids in equilibrium with clinopyroxene are similar to normal MORB. Both field evidence and geochemical data indicate that dunite formed from liquids that were significantly different from those that impregnated the peridotites.

Extraction of melt in dykes The development of interstitial clinopyroxene to megacrysts and finally the formation of centimetre-scale gabbroic veins and dykelets, which are characterized by euhedral clinopyroxene and interstitial plagioclase, indicate that the interconnected melt network in dunite channels becomes progressively clogged with crystallization products (Fig. 8d). This suggests that cooling was important and melt migration changed from focused porous flow in dunite channels to intrusion into narrow conduits. Cooling and crystallization might produce substantial hydrostatic overpressure, which might initiate the formation of cracks, and liquid might be expelled in dykes (Nicolas 1986; Kelemen et al. 1997; Kelemen & Aharonov 1998). In both peridotite massifs, the gabbroic veins and dykelets, and the mafic-ultramafic cumulate pods are rather primitive, with Mg number of olivine and pyroxenes of about 90. However, the trace element composition of liquids in equilibrium with clinopyroxene from early dykelets and cumulate pods at Monte Maggiore indicates that: (1) the primary melts of the Lanzo dykelets most probably correspond to low-degree (2-3%) melts, after fractional melting of an asthenospheric mantle source; (2) the primary melts of the Monte Maggiore dykelets and maficultramafic cumulates correspond to higher-degree (6-7%), depleted single melt increments after fractional melting. This indicates that the earliest dyke intrusions allowed both depleted and enriched single melt increments to migrate in the lithospheric mantle without undergoing significant

84

O. MUNTENER & G.B. PICCARDO

melt-peridotite reaction within the mantle column. The Western Alpine-Northern Apennine ophiolitic peridotites are intruded by metre-scale gabbroic dykes and kilometre-scale gabbroic bodies, which have sharp contacts with the country peridotite and which cut across all the previous mantle and magmatic structures (Fig. 8e). They vary in composition from rather primitive troctolite to Mg-gabbros to Fe-Ti-gabbros, and rare plagiogranites. In addition, they show chilled

margins against their country rocks, indicating that the surrounding peridotites were substantially colder than the intruding liquids. Calculated liquids in equilibrium with clinopyroxenes from the most primitive olivine gabbros from all peridotite massifs have almost flat to slightly LREEdepleted, chondrite-normalized patterns: they are closely similar to those of average aggregated normal MORE (Beccaluva et al 1984; Bodinier 1988). The intrusion of MORB-type fractionated Mg-rich and Fe-Ti-rich magmas most probably

Fig. 9. Possible scenario of mantle exhumation and melt-rock reaction in the framework of the tectonic evolution of the Piedmont Ligurian ocean (modified from Whitmarsh et al. 2001). Whether or not rift-related melt infiltration and heating are recorded by exhumed lithospheric mantle along ocean-continent transitions and/or slow-spreading ridges depends on the position of each peridotite relative to the underlying up welling asthenosphere. Peridotites in the eastern Central Alps (Malenco, Upper Platta) are still associated with the lower continental crust and show a 'cold' exhumation history during opening of the Piedmont Ligurian ocean (Miintener et al. 2000; Miintener & Hermann 2001), indicating a considerable distance to the upwelling asthenosphere. Lanzo, Corsica and the Ligurides may illustrate the other extreme of a 'hot' exhumation, where melt infiltration and melt-rock reaction by asthenospheretype liquids substantially modified the mantle peridotites.

A L P I N E - A P E N N I N E PERIDOTITES

occurred when the lithospheric mantle became more brittle at shallower levels in the conductive lithosphere.

Chemical refertilization and thermal erosion of the lithosphere An important effect of melt impregnation was significant heating of the impregnated peridotites; estimates indicate raised temperatures in the impregnated lithospheric mantle from the 10001100°C of the spinel-facies annealing recrystallization, attained during lithosphere accretion and subsequent cooling, to > 1250 °C, reached during impregnation, close to the dry peridotite solidus. Impregnation added basaltic components (e.g. the gabbroic microgranular aggregates) to the lithospheric peridotites, whereas mantle minerals, and particularly clinopyroxenes, were significantly enriched in most of the trace elements. Accordingly, the lithospheric mantle was significantly enriched and refertilized by the impregnating melts (Miintener et aL 2002). The heating combined with the chemical modification of the mantle rocks is similar to that proposed for asthenosphere-lithosphere interaction during early continental rifting (Menzies et al. 1987; Bedini et aL 1997) and along slow-spreading ridges (Elthon 1992; Cannat & Casey 1995; Cannat et al. 1997). Such a model of asthenosphere-lithosphere interaction could reconcile the highly contrasting interpretations with respect to the Lanzo peridotites. Whereas Nicolas, Boudier and coworkers interpreted the various igneous rocks as a consequence of dynamic melting of a rising mantle diapir during formation of the Piedmont Ligurian ocean, Pognante et al. (1985) stated that the Lanzo ultramafic rocks might be a section of subcontinental lithosphere with a polyphase history of partial melting and decompression during rifting that was later intruded in the Jurassic by N-MORB type liquids. Bodinier et al. (1991) provided some convergence of ideas in that they showed a transition from continental to oceanic mantle, based on Sr and Nd isotopes. Our interpretation of the Lanzo and Corsica peridotites as a product of refertilized and thermally modified lithospheric peridotite (Fig. 8), which was exhumed during the formation of the Piedmont Ligurian ocean, might serve as a general model to explain the highly contrasting evolution of different peridotite bodies along ocean-continent transitions and (ultra-)slow-spreading ridges. Whether or not rift-related melt infiltration and heating are recorded by exhumed lithospheric mantle along ocean-continent transitions and/or slow-spreading ridges depends on the relative

85

position to the underlying upwelling asthenosphere (Fig. 9). For example, peridotites in the eastern Central Alps are still associated with the lower continental crust and show a 'cold' exhumation history during opening of the Piedmont Ligurian ocean (Miintener et al. 2000; Muntener & Hermann 2001), indicating a considerable distance from the upwelling asthenosphere. Lanzo, Corsica and the Ligurides might illustrate the other extreme of a 'hot' exhumation, where modification of lithospheric peridotite by asthenospheric-type liquids was dominant. During the late rifting stage of embryonic oceans, the thermochemical erosion of mantle lithosphere above the upwelling asthenosphere could have played a fundamental role in the dynamics of the rifting system and in the transition from passive lithospheric extension to active oceanic spreading. Melt infiltration into mantle peridotite may be one of the reasons for the absence of well-developed layers of oceanic crust in the Alpine-Apennine system, which caused problems in interpreting these ophiolites in the sense of the 1972 Penrose conference definition. We gratefully acknowledge the financial support by the Swiss National Science Foundation (Grant 21-66923.01) and Italian MURST and the University of Genova. We thank A. Zanetti for LA-ICP-MS analyses, A. Romairone and S. Bruzzone for assistance in the field and in the laboratory, G. Manatschal for comments, and Y. Dilek and P. T. Robinson for constructive reviews.

References ANDERS, E. & EBIHARA, M. 1982. Solar System abundances of the elements. Geochimica et Cosmochimica Acta, 46, 2363-2380. ANONYMOUS 1972. Penrose field conference on ophiolites. Geotimes, 17, 24-25. BECCALUVA, L., MACCIOTTA, G., PICCARDO, G.B. & ZEDA, O. 1984. Petrology of Iherzolitic rocks from the Northern Appennine ophiolites. Lithos, 17, 299-316. BILL, M., NAGLER, T. & MASSON, H. 2000. Major, minor and trace element, Sm-Nd and Sr isotope composition of mafic rocks from the earliest oceanic crust of the Alpine Tethys. Schweizerische Mineralogische und Petrographische Mitteilungen, 80, 131-146. BLACKMAN, D.K., CANN, J.R., JANSSEN, B. & SMITH, D.K. 1998. Origin of extensional core complexes: evidence from the Mid-Atlantic Ridge at Atlantis fracture zone. Journal of Geophysical Research, 103,21315-21333. BODINIER, J.L. 1988. Geochemistry and petrogenesis of the Lanzo peridotite body, Western Alps. Tectonophysics, 149, 67-88. BODINIER, J.L., GUIRAUD, M., DUPUY, C. & DOSTAL, J. 1986. Geochemistry of basic dikes in the Lanzo massif (Western Alps): petrogenetic and geody-

86

O. MUNTENER & G.B. PICCARDO

namic implications. Tectonophysics, 128, 75—95. BODINIER, J.L., MENZIES, M.A. & THIRLWALL, M.F. 1991. Continental to oceanic mantle transition: REE and Sr-Nd isotopic geochemistry of the Lanzo Lherzolite Massif. Journal of Petrology (Special Lherzolite Issue), 191-210. BOILLOT, G.M., BESLIER, M.O., KRAWCZYK, C.M., RAPPIN, D. & RESTON, TJ. 1995. The formation of passive margins: constraints from the crustal structure and segmentation of the deep Galicia margin. In: SCRUTTON, R.A. (ed.) The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region. Geological Society, London, Special Publications, 90, 71-91. BONATTI, E., CLOCCHIATTI, R. & COLANTONI, P. ET AL. 1983. Zabargad (St. John's) Island: an uplifted fragment of sub-Red Sea lithosphere. Journal of the Geological Society, London, 140, 677-690. BONATTI, E., BRUNELLI, D., FABRETTI, P., LIGI, M., PORTARO, R.A. & SEYLER, M. 2001. Steady-state creation of crust-free lithosphere at cold spots in mid-ocean ridges. Geology, 29, 979-982. BORSI, L., SCHARER, U., GAGGERO, L. & CRISPINI, L.

1996. Age, origin and geodynamic significance of plagiogranites in Iherzolites and gabbros of the Piedmont-Ligurian ocean basin. Earth and Planetary Science Letters, 140, 227-241. BOUDIER, F. 1978. Structure and petrology of the Lanzo peridotite massif (Piedmont Alps). Geological Society of America Bulletin, 89, 1574-1591. BOUDIER, F. & NICOLAS, A. 1972. Fusion partielle gabbroi'que dans la Iherzolite de Lanzo (Alpes piemontaises). Schweizerische Mineralogische und Petrographische Mitteilungen, 52, 39—56. CANNAT, M. 1993. Emplacement of mantle rocks in the seafloor at mid-ocean ridges. Journal of Geophysical Research, 98, 4163-4172. CANNAT, M. 1996. How thick is the magmatic crust at slow spreading oceanic ridges? Journal of Geophysical Research, 101, 2847-2857. CANNAT, M. & CASEY, J.F. 1995. An ultramafic lift at the Mid-Atlantic Ridge: successive stages of magmatism in serpentinized peridotites from the 15°N region. In: VISSERS, R.L.M. & NICOLAS, A. (eds) Mantle and Lower Crust Exposed in Oceanic Ridges and in Ophiolites. Kluwer Academic, Dordrecht, 5-34. CANNAT, M., CHATIN, F., WHITECHURCH, H. & CEULENEER, G. 1997. Gabbroic rocks trapped in the upper mantle at the Mid-Atlantic Ridge. In: KARSON, J.A., CANNAT, M., MILLER, D.J. & ELTHON, D. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 153. Ocean Drilling Program, College Station, TX, 153, 243-264. COMPAGNONI, R., RADICATI DI BRONZOLO, F. & SANDRONE, R. 1984. Kaersutite-bearing mylonitic gabbro from the Lanzo peridotite (Western Italian Alps). Geologie en Mijnbouw, 63, 189-196. COSTA, S. & CABY, R. 2001. Evolution of the Ligurian Tethys in the Western Alps: Sm/Nd and U/Pb geochronology and rare-earth element geochemistry of the Montgenevre ophiolite (France). Chemical Geology, 175, 449-466. DECANDIA, F.A. & ELTER, P. 1972. La 'zona' ofiolitifera

del Bracco nel settore compreso fra Levanto e la Val Graveglia (Appennino ligure). Memorie della Societd Geologica Italiana, 11, 503—530. DESMURS, L. 2001. Mantle evolution and magmatism in an evolving ocean-continent transition: the Platta nappe, eastern Switzerland. PhD thesis, ETH Zurich. DESMURS, L., MANATSCHAL, G. & BERNOULLI, D. 2001. The Steinmann trinity revisited: exhumation and magmatism along an ocean-continent transition: the Platta nappe, eastern Switzerland. In: WILSON, R.C.L., WHITMARSH, R.B., TAYLOR, B. & FROITZHEIM, N. (eds) Non-Volcanic Rifting of Continental Margins: a Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 235-266. DESMURS, L., MUNTENER, O. & MANATSCHAL, G. 2002. Onset of magmatic accretion within magma-poor passive margin: a case study from the Err-Platta ocean-continent transition, Eastern Switzerland. Contributions to Mineralogy and Petrology, 144, 365-382. DEWEY, J.F., PITTMAN, W.C., RYAN, W.B.F. & BONNIN, J. 1973. Plate tectonics and the evolution of the alpine system. Geological Society of America Bulletin, 84, 3137-3180. DICK, HJ.B. 1989. Abyssal peridotites, very slow spreading ridges and ocean ridge magmatism. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 71-105. DICK, HJ.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DICK, HJ.B., FISHER, R.L. & BRYAN, W.B. 1984. Mineralogic variability of the uppermost mantle along mid-ocean ridges. Earth and Planetary Science Letters, 69, 88-106. DICK, HJ.B., NATLAND, J.H. & ALT, J.C. ET AL. 2000. A long in situ section of the lower ocean crust: results of ODP Leg 176 drilling at the Southwest Indian ridge. Earth and Planetary Science Letters, 179,31-51. DIJKSTRA, A.H., DRURY, M.R. & VISSERS, R.L.M. 2001. Structural petrology of plagioclase peridotites in the West Othris Mountains (Greece): melt impregnation in mantle lithosphere. Journal of Petrology, 42(1), 5-24. ELTHON, D. 1992. Chemical trends in abyssal peridotites: refertilization of depleted suboceanic mantle. Journal of Geophysical Research, 97, 9015-9025. FROITZHEIM, N. & EBERLI, G.P. 1990. Extensional detachment faulting in the evolution of a Tethys passive continental margin, Eastern Alps, Switzerland. Geological Society of America Bulletin, 102, 1297-1308. FROITZHEIM, N. & MANATSCHAL, G. 1996. Kinematics of Jurassic rifting, mantle exhumation, and passivemargin formation in the Austroalpine and Penninic nappes (eastern Switzerland). Geological Society of America Bulletin, 108, 1120-1133. GIANELLI, G. & PRINCIPI, G. 1977. Northern Apennine ophiolite: an ancient transcurrent fault zone. Bolle-

ALPINE-APENNINE PERIDOTITES tino della Societa Geologica Italiana, 96, 53-58. HEBERT, R., SERRI, G. & HEKINIAN, R. 1989. Mineral chemistry of ultramafic tectonites and ultramafic to gabbroic cumulates from the major oceanic basins and Northern Apennine ophiolites (Italy)—a comparison. Chemical Geology, 77, 183-207. HERMANN, J. & MUNTENER, O. 1996. Exhumationrelated structures in the Malenco-Margna system: implications for paleogeography and its consequences for rifting and Alpine tectonics. Schweizerische Mineralogische und Petrographische Mitteilungen, 76, 501-520. HOOGERDUIJN-SXRATING, E.H., RAMPONE, E., PlCCARDO,

G.B., DRURY, M.R. & VISSERS, R.L.M. 1993. Subsolidus emplacement of mantle peridotites during incipient oceanic rifting and opening of the Mesozoic Tethys (Voltri Massif, NW Italy). Journal of Petrology, 34, 901-927. JACKSON, M.D. & OHNENSTETTER, M. 1981. Peridotite and gabbroic structures in the Monte Maggiore Massif, Alpine Corsica. Journal of Geology, 89, 703-719. JOHNSON, K.T.M., DICK, H.J.B. & SHIMIZU, N. 1990. Melting in the oceanic upper mantle: an ion microprobe study of diopsides in abyssal peridotites. Journal of Geophysical Research, 95(B3), 2661-2678. KELEMEN, P.B. & AHARONOV, E. 1998. Periodic formation of magma fractures and generation of layered gabbros in the lower crust beneath spreading centers. In: BUCK, R., DELANEY, P.T., KARSON, J.A. & LAGABRIELLE, Y. (eds) Faulting and Magmatism at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union, 106, 267-289. KELEMEN, P.B., WHITEHEAD, J.A., AHARONOV, E. & JORDAHL, K.A. 1995. Experiments on flow focusing in soluble porous media, with applications to melt extraction from the mantle. Journal of Geophysical Research, 100, 475-496. KELEMEN, P.B., KOGA, K. & SHIMIZU, N. 1997. Geochemistry of gabbro sills in the crust-mantle transition zone of the Oman ophiolite: implications for the origin of the oceanic lower crust. Earth and Planetary Science Letters, 146, 475-488. LAGABRIELLE, Y. & CANNAT, M. 1990. Alpine Jurassic ophiolites resemble the modern central Atlantic basement. Geology, 18, 319-322. LAGABRIELLE, Y. & LEMOINE, M. 1997. Alpine, Corsican and Apennine ophiolites: the slow spreading ridge model. Comptes Rendus de I'Academie des Sciences, 325, 909-920. LAGABRIELLE, Y., POLINO, R. & AUZENDE, J.M. 1984. Les temoins d'une tectonique intra-oceanique dans le domaine tethysien: analyse des rapports entre les ophiolites et leur couverture metasedimentaire dans la zone piemontaise des Alpes franco-italiennes. Ofioliti, 9, 67-88. LEMOINE, M., TRICART, P. & BOILLOT, G. 1987. Ultramafic and gabbroic ocean floor of the Ligurian Tethys (Alps, Corsica, Apennines). In search of a genetic model. Geology, 15, 622-625. MANATSCHAL, G. & NIEVERGELT, P. 1997. A continentocean transition recorded in the Err and Platta nappes (eastern Switzerland). Eclogae Geologicae

87

Helvetiae, 90, 3-27. MENZIES, M.A., ROGERS, N., TINDLE, A. & HAWKESWORTH, CJ. 1987. Metasomatic and enrichment processes in lithospheric peridotites, an effect of asthenosphere-lithosphere interaction. In: MENZIES, M.A. & HAWKESWORTH, CJ. (eds) Mantle Metasomatism. Academic Press, London, 313-361. MUNTENER, O. & HERMANN, J. 1996. The Val Malenco lower crust-upper mantle complex and its field relations (Italian Alps). Schweizerische Mineralogische und Petrographische Mitteilungen, 76, 475500. MUNTENER, O. & HERMANN, J. 2001. The role of lower crust and continental upper mantle during formation of non-volcanic passive margins: evidence from the Alps. In: WILSON, R.C.L., WHITMARSH, R.B., TAYLOR, B. & FROITZHEIM, N. (eds) Non-Volcanic Rifting of Continental Margins: a Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 267-288. MUNTENER, O., HERMANN, J. & TROMMSDORFF, V. 2000. Cooling history and exhumation of lower crustal granulites and upper mantle (Malenco, Eastern Central Alps). Journal of Petrology, 41, 175200. MUNTENER, O., DESMURS, L., PETTKE, T., MEIER, M. & SCHALTEGGER, U. 2002. Melting and melt/rock reaction in extending mantle lithosphere: trace element and isotopic constraints from passive margin peridotites. Geochimica et Cosmochimica Acta, 66, A536. NICOLAS, A. 1986. A melt extraction model based on structural studies in mantle peridotites. Journal of Petrology, 27(4), 999-1022. OTTONELLO, G., JORON, J.L. & PICCARDO, G.B. 1984. Rare earth and 3d transition element geochemistry of peridotitic rocks: II. Ligurian peridotites and associated basalts. Journal of Petrology, 25, 373393. PETERS, T. 1963. Mineralogie und Petrographie des Totalpserpentins bei Davos. Schweizerische Mineralogische und Petrographische Mitteilungen, 43, 527-686. PICCARDO, G.B. 1976. Petrologia del massiccio Iherzolitico di Suvero (La Spezia). Ofioliti, 1, 279-317. PICCARDO, G.B. 1977. Le ofioliti dell'areale ligure: petrologia ed ambiente geodinamico di formazione. Rendiconti Societa Italiana di Mineralogia e Petrologia, 33,221 -252. PICCARDO, G.B., RAMPONE, E. & VANNUCCI, R. 1990. Upper mantle evolution during continental rifting and ocean formation: evidence from peridotite bodies of the Western Alpine-Northern Apennine system. Memoires de la Societe Geologique de France, 156, 323-333. PICCARDO, G.B., RAMPONE, E., VANNUCCI, R., SHIMIZU, N., OTTOLINI, L. & BOTTAZZI, P. 1993. Mantle processes in the subcontinental lithosphere: the case study of the rifted sp-lherzolites from Zabargad (Red Sea). European Journal of Mineralogy, 5, 1039-1056. POGNANTE, U., ROSLI, U. & ToscANi, L. 1985. Petrology of ultramafic and mafic rocks from the Lanzo peridotite body (Western Alps). Lithos, 18, 201-214.

88

O. MUNTENER & G.B. PICCARDO

RAMPONE, E. 2002. Mantle dynamics during passive rifting vs. slow-spreading ocean formation. In: BUCK, R., DELANEY, P.T., KARSON, J.A. & LAGABRIELLE, Y. (eds) Birth and Early Evolution of Alpine Ocean Basins. Annual Meeting of the Swiss Geological Society and the Swiss Society of Mineralogy and Petrography, Davos. Schweizerische, Geologische, 38. RAMPONE, E. & PICCARDO, G.B. 2000. The ophiolite oceanic lithosphere analogue: new insights from the Northern Apennines (Italy). In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 21-34. RAMPONE, E., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L. 1993. Subsolidus reactions monitored by trace element partitioning: the spinelto plagioclase-facies transition in mantle peridotites. Contributions to Mineralogy and Petrology, 115, 1-17. RAMPONE, E., HOFMANN, A.W., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L. 1995. Petrology, mineral and isotope geochemistry of the External Liguride Peridotites (Northern Apennines, Italy). Journal of Petrology, 36, 81-105. RAMPONE, E., HOFMANN, A.W., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L. 1996. Trace element and isotope geochemistry of depleted peridotites from an N-MORB type ophiolite (Internal Liguride, N-Italy). Contributions to Mineralogy and Petrology, 123, 61-76. RAMPONE, E., PICCARDO, G.B., VANNUCCI, R. & BOTTAZZI, P. 1997. Chemistry and origin of trapped melts in ophiolitic peridotites. Geochimica et Cosmochimica Acta, 61, 4557-4569. RAMPONE, E., HOFMANN, A.W. & RACZEK, I. 1998. Isotopic contrasts within the Internal Liguride ophiolite (N. Italy): the lack of a genetic mantlecrust link. Earth and Planetary Science Letters, 163, 175-189. Ross, K. & ELTHON, D. 1993. Cumulates from strongly depleted mid-ocean-ridge basalt. Nature, 365, 826829. SCHALTEGGER, U.,

DfiSMURS, L., MANATSCHAL,

G.,

MUNTENER, O., MEIER, M., FRANK, M. & BERNOULLI, D. 2002. The transition from rifting to seafloor spreading within a magma-poor rifted margin: field and isotopic constraints. Terra Nova, 14, 156-162. SEITZ, H.M., ALTHERR, R. & LUDWIG, T. 1999. Partitioning of transition elements between orthopyroxene and clinopyroxene in peridotitic and websterititc xenoliths: new empirical geothermometers. Geochimica et Cosmochimica Acta, 63, 3967-3982. SERRI, G. 1980. Chemistry and petrology of gabbroic complexes from the northern Apennine ophiolites. In: PANAYIOTOU, A. (ed.) Ophiolites. Cyprus Geological Survey Department, Nicosia, 296-313. SEYLER, M. & BONATTI, E. 1997. Regional-scale meltrock interaction in Iherzolitic mantle in the Romanche fracture zone (Atlantic Ocean). Earth and Planetary Science Letters, 146, 273-287.

STAMPFLI, G. 1993. Le Briansonnais, terrain exotique dans les Alpes? Eclogae Geologicae Helvetiae, 86, 1-45. STEINMANN, G. 1927. Die ophiolitischen Zonen in den mediterranen Kettengebirgen. In: 14th International Geological Congress. Compote-rendu, XlVe Congres Geologique International, 1926, Madrid, Graficas Reunidas, 2, 637-667. TARTAROTTI, P., SUSINI, S., NIMIS, P. & OTTOLINI, L. 2002. Melt migration in the upper mantle along the Romanche fracture zone (Equatorial Atlantic). Lithos,63, 125-149. TIEPOLO, M., TRIBUZIO, R. & VANNUCCI, R. 1997. Mgand Fe-gabbroids from Northern Apennine ophiolites: parental liquids and igneous differentiation processes. Ofioliti, 22, 57-69. TRIBUZIO, R., TIEPOLO, M. & VANNUCCI, R. 2000. Evolution of gabbroic rocks from the Northern Apennine ophiolites (Italy): comparison with the lower oceanic cust from modern slow spreading ridges. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and Ocean Drilling Program. Geological Society of America, Special Papers, 349, 129-138. TROMMSDORFF, V., PICCARDO, G.B. & MONTRASIO, A. 1993. From magmatism, through metamorphism to sea-floor emplacement of subcontinental Adria lithosphere during pre-Alpine rifting (Malenco, Italy). Schweizerische Mineralogische und Petrographische Mitteilungen, 73, 191-203. TUCHOLKE, B.E., LIN, J. & KLEINROCK, M.C. 1998. Megamullions and mullion structure denning oceanic metamorphic core complexes on the MidAtlantic Ridge. Journal of Geophysical Research, 103, 9857-9866. VAN DER WAL, D. & BODINIER, J.-L. 1996. Origin of the recrystallisation front in the Ronda peridotite by km-scale pervasive porous melt flow. Contributions to Mineralogy and Petrology, 122, 387-405. VANNUCCI, R., RAMPONE, E., PICCARDO, G.B., OTTOLINI, L. & BOTTAZZI, P. 1993a. Ophiolitic magmatism in the Ligurian Tethys: an ion microprobe study of basaltic clinopyroxenes. Contributions to Mineralogy and Petrology, 115, 123-137. VANNUCCI, R., SHIMIZU, N., PICCARDO, G.B., OTTOLINI, L. & BOTTAZZI, P. 1993b. Distribution of trace elements during breakdown of mantle garnet: an example from Zabargad. Contributions to Mineralogy and Petrology, 113, 437-449. VANNUCCI, R., PICCARDO, G.B. & RIVALENTI, G. ET AL. 1995. Origin of LREE-depleted amphiboles in the subcontinental mantle. Geochimica et Cosmochimica Acta, 59, 1763-1771. VENTURELLI, G., THORPE, R.S. & POTTS, PJ. 1981. Rare earth and trace element characteristics of ophiolitic metabasalts from the Alpine-Apennine belt. Earth and Planetary Science Letters, 53, 109123. WEISSERT, H. & BERNOULLI, D. 1985. A transform margin in the Mesozoic Tethys: evidence from the Swiss Alps. Geologische Rundschau, 14, 665-679.

A L P I N E - A P E N N I N E PERIDOTITES WHITMARSH, R.B., MANATSCHAL, G. & MINSHULL, T.A. 2001. Evolution of magma-poor continental

89

margins from rifting to sea-floor spreading. Nature, 413, 150-154.

This page intentionally left blank

Petrology and evolution of the Brezovica ultramafic massif, Serbia B. A. BAZYLEV 1 , S. KARAMATA 2 & G. S. ZAKARIADZE 1 1

Vernadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences, Kosygina 19, 119991, Moscow, Russia (e-mail: [email protected]) 2 Serbian Academy of Science and Arts, Knez Mihajlova 35, Belgrade, Yugoslavia Abstract: The tectonomagmatic history of ultramafic rocks in the Brezovica massif (Serbia) involved two separate magmatic stages, as inferred from the mineral and bulk-rock chemistry data and thermal history of the peridotites. In the first stage, a suite of spinel harzburgites was formed during partial melting of the mantle and segregation of tholeiitic melts. During the second stage, these spinel harzburgites were repeatedly heated and affected by percolating melt. This process formed dunites and refractory spinel harzburgites during melt-harzburgite interaction. The melt that segregated from these rocks during the second magmatic stage was of high-Ca boninite affinity. Both magmatic stages occurred in a suprasubduction geodynamic setting at a relatively deep level (25-28 km). In its present position the Brezovica massif has been interpreted as a relic of a suprasubduction-type oceanic lithosphere derived from the Central Dinaridic-Mirdita ocean basin. During eastward emplacement of the Brezovica massif over the underlying olistostrome, the ultramafic rocks were cooled to temperatures around 735 ± 20 °C.

In the Dinarides, peridotite massifs occur in two distinct, NW-SB-trending subparallel belts (Fig. 1), both interpreted as remnants of pre-existing basins with an oceanic lithosphere. The western belt, also called the Central Dinaridic ophiolite belt, is dominated in its northern part by Iherzolite massifs with a poorly developed hypabyssalextrusive component (Lugovic et aL 1991). This belt continues southward through Albania (Albanian peridotite massifs) to Greece (Pindos, Vourinos, Othris complexes), where a significant part of the peridotite massifs is composed of spinel harzburgites (Beccaluva et aL 1994). The basin that existed in this region during the Mesozoic is called the Mirdita Ocean in the Albanides and the Pindos Ocean in the Hellenides (Robertson & Shallo 2000). Although many investigators apply the term 'ocean' to this basin (Aliaj & Mego 1994; Robertson & Karamata 1994; Karamata & Krstic 1996), it was rather an ephemeral (in the Dinaridic part from the Mid- or Late Triassic to Late Jurassic) marginal basin in the long-lived Palaeotethys Ocean (Karamata et al. 2000). A suprasubduction zone origin has been established for the majority of the Albanian and Greek ultramafic massifs (Beccaluva et al. 1994; Bebien et al. 1998, 2000; Bizimis et al. 2000; Robertson & Shallo 2000). The available data on the mafic rock geochemistry for the northern part of this basin in the Dinarides are consistent with either a back-arc basin or mid-ocean ridge origin (Lugovic

et al. 1991). The eastern belt, called the Vardar zone, is characterized by harzburgite peridotites (Karamata et al. 1980; Pamic 1983) and also continues southward through Macedonia to Greece. The Vardar zone represents a collisional suture zone marking the closure of the long-lived Palaeotethyan Ocean basin during the Late Jurassic-Late Cretaceous (Karamata et al. 2000). These two belts are separated by a microplate called the Drina-Ivanjica terrane in the Dinarides, the Korabi zone in the Albanides, and the Pelagonian zone in the Hellenides (Shallo 1996; Robertson & Shallo 2000; Fig. 1). Just south of the Peg-Srbica line, the two belts are juxtaposed along a tectonic zone, but the relationship between them is still unclear (Karamata & Krstic 1996). The Brezovica ultramafic massif is situated within this zone, about 50 km east of the DjakovicaMirdita (Kukes) and Tropoja and Lores (Lura) ultramafic massifs in Albania. Based on the available data we cannot confidently assign the Brezovica massif to the Central Dinaridic or to Vardar zone ophiolite belt. Recently, a number of papers devoted to the peridotite petrology of the Eastern Mediterranean ultramafic massifs have led to the significant revision of existing scenarios for the geological evolution of this region (Lugovic et al. 1991; Bazylev et al. 1993; Beccaluva et al. 1994; Carosi et al. 1996; Bebien et al. 1998; Bizimis et al. 2000). However, the peridotites in Serbia and

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 91-108. 0305-8719/037$ 15 © The Geological Society of London 2003.

92

B. A. BAZYLEV ETAL.

Fig. 1. Location of the Brezovica ultramafic massif (modified after Karamata & Krstic 1996). DOB, Dinaride ophiolite belt, showing the western branch (WB) and the eastern branch (EB) at the southern part. Ultramafic bodies are shown in grey: Bo, Borja massif; Od, Ozren of Doboj massif; KK, Krivaja-Konjuh massif; Zl, Zlatibor massif; Os, Ozren of Sjenica massif; Or, Orahovac ultramafic rocks; DjM, Djacovica-Mirdita massif; Br, Brezovica ultramafic rocks. Neighbouring terranes: DHCT, Dalmatian-Hercegovinian composite terrane; SUT, Sana-Una terrane; CBMT, Central Bosnian mountains terrane; EBDT, SE Bosnian-Durmitor terrane; DIE, Drina-Ivanjica terrane; KPTs, Korab and Pelagonian terranes; VZWB, western belt of the Vardar zone; JB, Jadar block terrane; KBR, Kopaonik block and ridge; MVZ, main belt of the Vardar zone; SMCT, Serbo-Macedonian composite terrane; KBT, terranes of the Carpathian-Balkan arc.

Bosnia remain the least investigated. In this paper, we present new data on the Brezovica ultramafic massif in Serbia. A peculiar feature of the Brezovica peridotite massif is a metamorphic aureole (sole) with a continuous decrease of the metamorphic grade from the ultramafic massif downwards (Karamata 1968a, 1968b). This massif is also of special interest because of its tectonic position and the existence of a well-preserved accretionary complex. The aim of this paper is to document and to discuss some aspects of the magmatic and metamorphic history of the Brezovica massif peridotites based on the available mineralogical, petrochemical and petrological data.

Geological setting The area of Brezovica (Fig. 2) consists mainly of a Middle to Upper Jurassic olistostromal melange composed mainly of sandstone (greywackes and subordinate arkosic rocks) and basaltic lavas (sheet flows or pillow lavas), and locally contains chert, limestone, manganiferous sedimentary rocks, and gabbro olistoliths in a silty-shaly matrix. Limestone lenses with intercalated basalt and claystone occur as blocks up to a kilometre long and tens of metres thick that slid into this assemblage from the adjacent continental mass (probably the Drina-Ivanjica terrane). This melange was subsequently metamorphosed up to the

B R E Z O V I C A U L T R A M A F I C MASSIF, SERBIA

93

Fig. 2. Geological map and cross-section of the Brezovica area with sites of rock sampling.

stilpnomelane-lawsonite-prehnite-pumpellyite association, indicating a depth of burial of about 25 km (Karamata et al. 1991). An ophiolitic thrust sheet, composed of tectonite and cumulate ultramafic rocks, was emplaced over or into this melange unit. Other members of a typical ophiolite assemblage, such as massive and layered gabbros, sheeted dykes, extrusive rocks, and a sedimentary cover, are generally lacking. Emplacement of the peridotite body resulted in shearing and formation of a cleavage system in the underlying olistostromal melange unit, and caused their metamorphism. During this process, a narrow contact zone, up to 100 m on the east and up to 400 m on the west (Fig. 2) was produced underlying the ultramafic body (Karamata 1985). The maximum temperature in the olistostromal contact metamorphic aureole is

estimated to be about 700 °C in the western part of the Brezovica massif (Karamata & Milovanovic 1990) and about 600 °C or lower in the eastern part (Karamata et al. 1991; Ciric & Eric 1996). The ultramafic body seems to have been emplaced from west to east, as evidenced by a fall in the calculated maximum temperature of contact metamorphism and by a decrease in the thickness of both the contact zone and the tectonized peridotite in this direction (Fig. 2). Following emplacement of the peridotite slice, the olistostromal rocks and the peridotite were uplifted and eroded, leaving behind small, discontinuous ultramafic masses forming topographic highs along the Tromedja-JezerinaLivad ridge (Fig. 2). These erosional remnants are collectively called the Brezovica ultramafic body or massif.

94

B. A. BAZYLEV ETAL.

Field occurrence of the Brezovica ultramafic rocks Several ultramafic rock types are recognizable in outcrops in the Brezovica massif. The stratigraphically lowest level of ultramafic rocks consists of homogeneous tectonites composed of spinel harzburgite. This level is up to 300 m in thickness in the west (Ostrovica) and thins eastwards (Mali Krst and Livad; Fig. 2). These rocks exhibit a typical tectonite fabric with mineral deformation indicative of solid-state flow. In the highest stratigraphic levels in the western masses and at all levels in the eastern ones, the harzburgites are intruded by spinel dunites, including clinopyroxene-bearing varieties. Dunite appears first as amoebiform areas within the harzburgite, but the amount of dunite increases up-section in the sequence, forming a grid of dyke-like bodies with isolated harzburgite blocks between them; structurally higher still, dunites become dominant, grading into the dunitic basal parts of the cumulate sequence. Dunites within the harzburgites are characterized by a linear to planar arrangement of chromite-spinel grains, but there is no significant solid-state flow in the olivine and clinopyroxene grains. The chromite in dunite shows banding, folding and pull-apart structures. The harzburgite-dunite unit grades upward into a layered cumulate rock sequence consisting of dunite, feldspar-bearing dunite, feldspar-bearing Iherzolite, pyroxenite, poikilitic Iherzolite, wehrlite and troctolite, with abundant cross-cutting pyroxenite and gabbro dykes (Batocanin & Memovic 1996). This cumulate sequence is preserved only in the highest structural levels of the easternmost masses (Livad and Budanac; Fig. 2).

Analytical procedures Bulk-rock compositions were analysed through routine X-ray fluorescence (XRF) methods by I. A. Roshchina and T. V Romashova with a Phillips PW-1600 apparatus in the Vernadsky Institute of Geochemistry and Analytical Chemistry, Moscow, Russia. The mineral chemistry was determined by N. N. Kononkova with a CAMECA CAMEBAX electron microprobe at the Vernadsky Institute using an accelerating voltage of 15 kV and beam current of 35 nA. Natural and synthetic minerals were used as standards.

Primary mineralogy Spinel harzburgites display notable differences in mineral compositions, which are correlated partly with distance from dunite bodies, and partly with

their structural position in the section (i.e. with their distance from the upper layered cumulate sequence and from the tectonic contact with the underlying olistostromal rocks). Three groups are distinguished among the investigated harzburgites based on their mineral compositions. The first group is represented by samples that are most distant from the dunites, and that are located in the stratigraphically lower levels of ultramafic masses (samples BR-2, -13, -14; Fig. 2). Some petrographic features of high-temperature plastic deformation (kink bands of orthopyroxene porphyroblasts, mosaic extinction of olivine grains) are evident in these rocks. Spinels have elevated Cr number (Cr/(Cr + Al)) of 0.45-0.49 (Table 1), are low in titanium (<0.08 wt% TiO2), and are weakly oxidized. The Fe number (Fe3+/ (Cr + Al + Fe3+)) varies from 0.008 to 0.019. All of these features are typical of spinels in residual peridotites. Olivine and orthopyroxene are highly magnesian (Mg number 90.6-91.1) (Tables 2 and 3), and olivine has high nickel (0.36-0.43 wt% NiO) and low calcium (<0.03 wt% CaO) contents. Orthopyroxene is low in aluminium (2.3-2.7 wt% A12O3) and mostly has low calcium (0.6-0.7 wt% CaO), indicating subsolidus re-equilibration. Clinopyroxene in these rocks is low in aluminium (1.5-2.0wt% A12O3), titanium and sodium (0.02-0.09 wt% TiO2 and Na2O). Harzburgites of the second group (samples BR1, -4, -7, -8) are petrographically similar to those of the first group, but differ in having a more refractory mineralogy. Spinel is significantly more chromian (Cr number 0.54-0.64) (Fig. 3), olivine and orthopyroxene are more magnesian (Mg number 90.9-92.3) (Fig. 4), olivine is slightly richer in nickel (0.38-0.45 wt% NiO), and both pyroxenes are poorer in aluminium (1.0-2.1 wt% Al2Os). In outcrop, these rocks generally are located nearer to dunites than harzburgites of the first group. Sample BR-3 is similar to rocks of this group in its Cr number and titanium content of spinel, and in the aluminium and sodium contents of pyroxenes. It differs in the degree of iron oxidation in the spinels (Fe number 0.058) (Fig. 5) and in having somewhat more iron-rich olivine and orthopyroxene. The third group of harzburgites is represented by samples that show anomalous mineral compositions (BR-5, -6). These rocks are located in the upper parts of the Malo Borce and Borov Vrh remnants, as well as locally in Livad (Fig. 2), and are closely associated with dunites. Plastic deformation textures are preserved only in the central parts of some large orthopyroxene porphyroblasts (kink bands). Olivine grains generally lack mosaic extinction, and areas with 'cumulate' textural features (strong xenomorphism of pyroxenes rela-

Table 1. Spinel compositions Sample: Rock: Generation: Points: Ti02 A12O3 FeO MgO Cr2O3 Total Crno. Mg no. Feno.

2

BR-12 D C 4

BR-12 D R 2

BR-12 D /Hbl 2

BR-l 3 HI Av. 4

BR-l 4 HI Av. 6

0.19 6.21 30.24 6.01 58.24 100.89 0.863 0.305 0.084

0.30 14.95 25.91 9.13 50.10 100.38 0.692 0.437 0.068

0.33 14.09 28.23 7.97 48.84 99.46 0.699 0.389 0.082

0.46 15.72 31.73 6.91 46.24 101.62 0.664 0.334 0.098

0.08 30.99 16.22 14.02 38.63 99.94 0.455 0.616 0.008

0.03 29.27 17.30 13.49 40.06 100.15 0.479 0.598 0.014

BR-l H2 Av. 3

BR-2 HI Av. 3

BR-3 H2-3 Av. 3

BR-4 H2 Av. 2

BR-5 H3 Av. 4

BR-6 H3 Av. 3

BR-7 H2 Av. 3

BR-8 H2 Av. 5

BR-9 D Av. 3

BR-10 D Av. 4

BR-ll D M 11

BR-ll D

0.02 19.36 18.66 12.03 50.30 100.37 0.635 0.559 0.023

0.03 28.56 17.76 13.25 40.16 99.77 0.485 0.592 0.019

0.05 19.06 22.95 10.72 46.46 99.24 0.621 0.507 0.058

0.05 24.37 18.90 13.10 44.21 100.63 0.549 0.591 0.035

0.14 13.30 25.17 9.20 51.63 99.44 0.722 0.449 0.069

0.10 13.39 24.33 9.12 53.29 100.24 0.727 0.443 0.053

0.03 25.15 16.25 14.16 44.01 99.60 0.540 0.638 0.025

0.03 21.27 19.47 11.80 47.51 100.09 0.600 0.546 0.026

0.23 11.24 25.66 9.60 53.58 100.32 0.762 0.467 0.084

0.18 9.45 24.80 9.35 57.03 100.81 0.802 0.457 0.070

0.17 10.38 25.26 8.15 55.95 99.91 0.783 0.404 0.053

s

Table 2. Olivine compositions Sample: Rock: Points:

BR-l

BR-2

H2 9

HI 8

SiO2

41.38 8.64 0.13 50.05 0.03 0.38 100.60 91.2

41.76 9.03 0.11 49.90 0.03 0.40 101.22 90.8

FeO MnO MgO CaO NiO Total Mg no.

BR-4

BR-5

BR-6

BR-7

BR-8

BR-9

BR-10

6

H3 4

H2 4

H2 9

D 5

D 8

BR-12 D 9

BR-l 4

H3 5

BR-ll D 8

BR-l 3

H2 4

HI 10

HI 6

41.06 9.60 0.15 49.02 0.02 0.36 100.23 90.1

41.72 8.90 0.14 49.85 0.02 0.45 101.07 90.9

40.65 9.10 0.14 49.40 0.02 0.34 99.66 90.6

41.66 9.08 0.14 49.90 0.03 0.33 101.14 90.7

41.49 7.61 0.14 51.25 0.02 0.38 100.89 92.3

41.38 8.44 0.12 50.34 0.02 0.42 100.73 91.4

41.83 8.21 0.12 50.72 0.06 0.38 101.31 91.7

40.64 8.25 0.12 50.94 0.06 0.37 100.38 91.7

41.76 7.81 0.12 50.68 0.01 0.35 100.73 92.0

41.55 9.45 0.13 50.06 0.03 0.31 101.52 90.4

41.40 8.72 0.11 49.95 0.02 0.36 100.57 91.1

41.47 9.03 0.15 50.32 0.02 0.43 101.41 90.9

BR-3 H2-3

96

B. A. BAZYLEV ETAL.

Fig. 3. Compositions of spinels from ultramafic rocks of Brezovica. Hzl, Hz2, Hz3, first, second and third group of spinel harzburgites, respectively; D, dunites. A, field of spinels from mid-ocean ridge peridotites after Dick & Bullen (1984); B, field of spinels in harzburgites and dunites from the Mariana forearc, Conical Seamount (Ishiietal. 1992).

Fig. 4. Variations in Mg number of olivine and orthopyroxene with Cr number of spinels in the Brezovica ultramafic rocks. Symbols are the same as in Figure 3. The field corresponds to mineral compositions of the most refractory spinel harzburgites from midocean ridges, not influenced by melt-rock interaction (Shibata & Thompson 1986; Dick & Natland 1996; Bazylev 2000; Bazylev et al. 2001, and unpublished data of the authors); arrow indicates the average trend of the mantle partial melting.

97

BREZOVICA ULTRAMAFIC MASSIF, SERBIA

Fig. 5. Spinel titanium content and iron oxidation in Brezovica ultramafic rocks. Symbols are the same as in Figure 3.

tively to olivine, small inclusions of silicate minerals in spinels) are widespread in the rocks. Spinels in these rocks are more chromian (Cr number 0.62-0.72) but have elevated titanium contents (0.05-0.14 wt% TiO2) and elevated iron oxidation states (Fe number 0.053-0.066) (Fig. 5). Distinct from 'normal' harzburgites of the first two groups, the increase in Cr number of spinel is coupled here with a decrease in magnesium number of olivine and orthopyroxene (Mg number 90.1-91.0) (Fig. 4) and in the nickel content of olivine (0.33-0.36 wt% NiO). The aluminium contents of orthopyroxenes are relatively low (1.0-1.2wt% A12O3). A slightly elevated sodium content in clinopyroxenes is notable (0.110.13wt%Na 2 O). In dunites, a sharp idiomorphism of spinels and xenomorphism of clinopyroxene relative to olivine

are typical. Some banding in the rocks is expressed by chains of Cr-rich spinel or clinopyroxene grains, but there is no pronounced deformation (such as mosaic extinction or granulation of large olivine grains). Spinels from these rocks are rich in chromium (Cr number 0.690.80) and titanium (0.17-0.30 wt% TiO2). Some spinel grains are zoned with an increase in Cr number, Fe number and titanium content and a decrease in Mg number from core to rim. The magnesium number of olivines is variable (90.492.0), as are the calcium (0.05-0.11% CaO) and nickel (0.31-0.38 wt% NiO) contents. Clinopyroxenes have low titanium (0.04-0.09 wt% TiO2) and aluminium (0.4-0.7 wt% A12O3) contents, but elevated sodium contents (0.14-0.42% Na2O). Pargasite hornblende (Mg number 90.4, TiO2 0.63 wt%) is spatially associated with clinopyroxene in one of the dunite samples (Table 4). Both minerals occur as small (up to 0.3 mm) xenomorphic grains in interstices between relatively large (up to 10 mm) rounded olivine grains, which are generally serpentinized.

Secondary mineralogy Signs of medium-temperature metamorphism (corresponding to greenschist- and lower amphibolitefacies conditions) are scarce in the investigated samples. In the first group of harzburgites, small grains of tremolitic hornblende are present (Table 4). These grains are round or elongate and are developed along clinopyroxene grain margins, or along two-pyroxene grain boundaries. Because no other associated metamorphic minerals were found, the crystallization temperature of the tremolitic hornblende can only be estimated from its

Table 4. Compositions of other minerals Sample: Rock: Mineral: Points:

BR-1 H2 Tre 1

BR-1 H2 Serp 2

BR-1 H2 Bast 1

BR-9 D Serp 2

BR-9 D Bruc 2

BR-11 D Tre 5

BR-11 D Serp 2

BR-1 2 D Hbl 6

BR-1 3 HI TreH 5

BR-14 HI TreH 5

Si02 TiO2 A1203 FeO MnO MgO CaO Na2O K20 Cr203 Total Cl Mg no.

57.35 0.02 2.13 1.91 0.03 22.73 13.54 0.08 0.33 98.13 95.5

42.78 0.08 0.86 5.92 0.19 37.98 0.39 0.02 0.02 88.24 92.0

42.20 0.00 1.27 7.33 0.21 37.49 0.32 0.09 0.61 89.52 90.1

37.25 0.00 0.13 4.22 0.03 40.29 0.04 0.01 0.22 82.18 94.5

2.89 0.02 0.02 22.47 0.57 59.66 0.26 0.01 0.03 85.94 82.6

58.05 0.04 0.30 1.40 23.55 13.18 0.10 0.09 96.70 96.8

39.85 0.03 0.03 5.17 39.77 0.05 0.00 0.09 85.00 93.2

44.76 0.63 13.39 3.54 0.05 18.71 13.46 3.48 1.95 99.97 90.4

52.39 0.24 7.03 2.49 0.15 21.24 13.18 1.00 0.04 1.25 99.01 0.02 93.8

55.76 0.05 4.51 2.45 0.08 21.85 12.81 0.25 0.03 0.80 98.60 0.01 94.1

98

B. A. BAZYLEV ETAL.

aluminium content (Bazylev & Silantyev 2000a) as 600-650 °C. In harzburgites of the second group, calcium amphibole is represented by tremolite (Table 4), formed at c. 550 °C. Tremolite is developed in the rims of clinopyroxene grains (obviously replacing it) as aggregates of subparallel elongate or prismatic grains. No amphiboles were found in harzburgites of the third group, and in dunites amphibole is represented by low-aluminium tremolite, which is closely associated with metamorphic diopside. Here, tremolite and diopside evidently replace grains of the primary clinopyroxene. The association of diopside and tremolite in ultramafic rocks is stable at 450500 °C at low to moderate pressures (Evans 1977). Low-temperature metamorphism of peridotites (serpentinization) is present in all of the studied ultramafic rocks. We note the presence of brucite associated with serpentine (Table 4) and discuss it below.

Petrochemistry Although the investigated ultramafic rocks display varying degrees of serpentinization, there are no evident changes in their bulk-rock compositions (Table 5) as a result of this process. For example, CaO/Al2O3 ratios are similar in most fresh and serpentinized harzburgite, indicating no loss of calcium (e.g. Bazylev et al 1993). MgO/SiO2 ratios in dunites are similar and high, being distinctly higher than those in the harzburgites. This observation argues against loss of magnesium during serpentinization. The magnesium number of bulk-rock samples correlates well with the Mg number of olivine and orthopyroxene, indicating no loss or gain of iron. Any compositional peculiarities of peridotites, therefore, should re- ' fleet conditions of their origin, and not of their hydration. Generally, the bulk-rock compositions of residual spinel peridotites vary in limited ranges along a trend, which reflects depletion of the primitive mantle in a basaltic component. To characterize this trend, a database composed of about 300 compositions of abyssal, alpine-type and xenolithic spinel peridotites (both original and from the literature) was used. The database was tightly filtered to exclude both non-isochemically metamorphosed rocks and rocks with apparent signs of melt-rock interaction (i.e. elevated titanium content in spinel, lowered Mg number of silicate minerals, large inhomogeneity in mineral compositions, the presence of trace hornblende, mica, plagioclase, etc.). Some chemical characteristics of this trend appear to be similar for mantle spinel peridotites formed in various geodynamic settings (suprasubduction, within-plate, mid-oceanic

BREZOVICAULTRAMAFIC MASSIF, SERBIA ridge). The FeO/SiOi ratio in these rocks varies from 0.170 to 0.202 (the oxide contents are in wt%), the Cr2O3/SiO2 ratio varies from 0.0065 to 0.0119, and the Cr numbers of both spinel and bulk-rock samples correlate using the following equation (Bazylev et a/., 1999):

Abyssal spinel peridotites also lie in a restricted field when bulk-rock Ti/Al ratios are plotted against the parameter -ln[Cr/(Cr + Al)]Spi. The nature of the variations is related to the origin of all mantle peridotites from a single source (McDonough & Sun 1995), and also to a relatively small influence of the degree of partial melting, the mechanism of melting (i.e. batch or fractional), and pressure on the Fe, Si and Cr contents in the residual minerals. The rock-spinel correlation given above is in agreement with the correlation between Cr numbers of spinels and pyroxenes from spinel peridotites and also with the limited variation of the spinel/pyroxene modal ratio in these rocks (e.g. Dick et al 1984; Dick 1989). Harzburgites of the first two groups (excluding sample BR-3) show a good correlation between the Cr number of spinels and the Cr number of the bulk rock and plot along the residual trend (Fig. 6). Dunites plot in this diagram near the ideal spinel-dunite line (where Cr number of spinels and bulk rocks are equal, which should be the case for olivine-spinel modal rock compositions). Harzburgite BR-3 plots above the residual trend, showing an enrichment of modal pyroxene. Harzburgites of the third group plot between the residual spinel peridotite trend and the dunite line. This feature indicates that these rocks are poorer in pyroxene (orthopyroxene) than typical residual

Fig. 6. Correspondence of Cr number of spinels and rocks for Brezovica ultramafics. A, trend of residual spinel peridotites; B, the line of spinel dunites. Symbols are the same as in Figure 3.

99

harzburgites showing the same Cr number of spinel. It should be mentioned that olivine crystallization cannot provide a shift of the points in this diagram. The Cr2O3/SiO2 ratio in all harzburgites and in most dunites (excluding sample BR-11) varies in the range typical for mantle residual rocks (Fig. 7). This observation indicates that crystallization of spinel from melt took place only for sample BR-11, and implies that the rest of the dunites were formed by dissolution of orthopyroxene during harzburgite-melt interaction. Variations of the FeO/SiC>2 ratio in bulk-rock compositions demonstrate a dual distribution of the investigated rocks (Fig. 8). Harzburgites of the first two groups (excluding sample BR-3) show relatively low FeO/SiO2 ratios, which fall mainly within the range typical for residual mantle, and which tend to decrease slightly with an increase of spinel Cr number. Harzburgite BR-3, harzburgites of the third group, and dunites have elevated FeO/ SiC>2 ratios. This compositional shift may reflect

Fig. 7. Variations of bulk-rock Cr2O3/SiO2 ratios for Brezovica ultramafic rocks. Symbols are the same as in Fig. 3.

Fig. 8. Variations of bulk-rock FeO/SiO2 ratios for Brezovica ultramafic rocks. Symbols are the same as in Figure 3.

100

B. A. BAZYLEV ETAL.

olivine crystallization, dissolution of orthopyroxene, or influx of iron during re-equilibration of the peridotite with a fractionated melt. Anomalous harzburgites of the third group also differ from abyssal residual spinel peridotites in their elevated Ti/Al ratios (Fig. 9), which can reflect interaction of these rocks with a fractionated melt.

Thermal history as a reflection of magmatic processes The geothermometry of spinel peridotites provides important information about the cooling history of these rocks, as the equilibrium temperatures, estimated by mineral geothermometers, may be treated as the closure temperatures of the corresponding exchange reactions. It was recognized that abyssal spinel peridotites have higher equilibration temperatures than alpine-type (suprasubduction) peridotites, which argue for a faster cooling of the mantle rocks in a mid-ocean ridge (MOR) geodynamic setting (Dick & Fisher 1984; Parkinson & Pearce 1998). The closure temperatures of both two-pyroxene and olivine-spinel reactions in mantle spinel peridotite were modelled by Bazylev & Silantyev (1999), assuming a fixed value for the average size of the spinel grains (0.3 mm) and for the distance of the analysed points from the orthopyroxene-clinopyroxene contact (20 urn). It was established that continuous cooling of the rocks after melt segre-

Fig. 9. Variations of chondrite-normalized bulk-rock Ti/ Al ratios with Cr number of spinels. Chondrite composition is after Anders & Grevesse (1989). Field of spinel peridotites from Mid-Atlantic Ridge is after Bazylev et al. (1999). PM is the composition of the primitive mantle (McDonough & Sun 1995); arrow indicates the compositional changes during mantle partial melting. Symbols are the same as in Figure 3.

gation in a MOR setting should produce a relatively large difference (150-200 °C) in the closure temperatures of the two-pyroxene and olivine spinel reactions as a result of different diffusion parameters. Based on the model of adiabatic melting below mid-ocean ridges (Langmuir et al. 1992), we infer that the spinel peridotites in a single area should be heated to similar maximum temperatures during their ascent, resulting in a similar degree of partial melting and a similar cooling rate after melt segregation. Thus, rocks from a restricted MOR locality should demonstrate similar values for both the spinel Cr number and the exchange reaction closure temperatures. This was found to be the case for all MOR spinel peridotites for which representative analytical data are available (Bazylev & Silantyev 2000a), including areas where spinel dunites compose a significant part of the mantle, such as at Ocean Drilling Program Site 895 in the Hess Deep (Arai & Matsukage 1996; Dick & Natland 1996) or at the southern inner corner high of the 15°20'N Fracture Zone in the Atlantic Ocean (Bazylev & Silantyev 2000a). The three harzburgite groups within the Brezovica massif, which are different both in their mineral compositions and in their structural position within the ultramafic body, are characterized by different average closure temperatures of twopyroxene and olivine-spinel exchange reactions. Generally, these temperatures are 894 °C and 714 °C for the first group, 831 °C and 755 °C for the second group, and 916 °C and 726 °C for the third group. Calibrations are from papers by Wells (1977) and Ballhaus et al. (1991). Generally, significant variations in the two-pyroxene and olivine-spinel temperatures within the spinel peridotites from a restricted area are not consistent with continuous rock cooling after melt segregation and with a MOR setting (Bazylev & Silantyev 2000b). Both the mineral chemistry and the geothermometry of these rocks are consistent with a model of locally repeated heating of the rocks to solidus temperatures (i.e. a second magmatic episode for a part of these rocks; Bazylev & Silantyev 2000b). Because the process of mantle melting is manifested in an increase of the primary spinel Cr number (Dick & Bullen 1984; Hellebrand et al., 2001), the second magmatic episode should evidently be related to the formation of rocks with the most chromian spinels that are represented in the case of the Brezovica by dunites and the third group harzburgites. The difference between the two-pyroxene and olivine-spinel temperatures varies for the Brezovica rock groups generally because of variations in the two-pyroxene temperatures; the olivine-

BREZOVICAULTRAMAFIC MASSIF, SERBIA spinel temperatures are rather similar for all the rock groups (735 ± 20 °C). This observation is consistent with a variant of modelled spatial distribution of the mineral exchange closure temperatures in peridotites located at an isothermal level within the lithosphere after repeated heating by an infiltrated (percolated) melt, as presented in Figure 10. The values of various parameters (e.g. the temperature of the percolating melt, the duration of percolation, the initial temperature of the peridotites and initial two-pyroxene and olivinespinel temperatures in these rocks before percolation) were chosen to explain the phenomena of spatially associated spinel peridotites with different closure temperatures for exchange reactions (Bazylev & Silantyev 2000b); the values for the Brezovica rocks may have been somewhat different. Nevertheless, this model reproduces well the main features of the Brezovica peridotite petrologyThe large difference (190°C) between the Opx-Cpx and Ol-Spl temperatures for the third group of harzburgites indicates continuous cooling after melt segregation (or crystallization) to a temperature lower than about 726 °C (which is the

101

estimated average Ol-Spl temperature for these rocks). These rocks demonstrate the highest twopyroxene temperatures and the highest spinel Cr number, both indicating their initial position near a zone of magmatic channels (Fig. 10) that is marked now by dunite dykes and bodies (Kelemen et al. 1995). The second group of harzburgites, located adjacent to the third group, was heated during the second magmatic event to temperatures lower than the solidus (and lower than the temperatures to which the third group harzburgites were heated). As a result of this event, their cooling after the cessation of the melt percolation was slower, resulting in lower closure temperature of the two-pyroxene reaction (Fig. 10). This process provided the small difference between the average values of the two-pyroxene and olivinespinel temperatures shown by these rocks (c. 80 °C). The harzburgites of the first group appear to have been even farther from the heat source (Fig. 10). They were heated (if heated at all) to temperatures lower than 830°C, and the time of heating was too short to provide pyroxene reequilibration at this temperature. Therefore, the

Fig. 10. An example of the calculated distribution of closure temperatures of exchange reactions in mantle peridotites as a result of their secondary heating by the percolated melt (Bazylev & Silantyev 2000b). Tr, maximum temperature of the rocks; 7pp, final closure temperature of the two-pyroxene reaction; 7^s, final closure temperature of the olivine-spinel reaction. The parameters for this model are: percolating melt temperature is assumed to be 1250 °C, the duration of percolation is assumed to be 75 ka, the assumed initial temperature of the peridotites is 800 °C, the adopted initial closure temperature in peridotites (before the melt percolation) are 950 °C and 820 °C for the two-pyroxene exchange reaction and for the olivine-spinel reaction, respectively. 1, outer zone, where rpp is not affected by heating (corresponds to the location of the Brezovica first group of harzburgites); 2, intermediate zone (corresponds to the location of the Brezovica second group of harzburgites); 3, inner zone adjacent to the magmatic channel; the temperature during melt percolation exceeds 1200 °C in its innermost part, causing peridotite melting or melt-rock interaction. This zone corresponds to the location of Brezovica dunites and third group of harzburgites.

102

B. A. BAZYLEV ETAL.

two-pyroxene temperatures in these rocks are relict and rather reflect their cooling conditions after the previous magmatic episode. Based on the model, the final olivine-spinel temperatures for all the Brezovica ultramafic rocks should be similar and equal to the initial temperature of the rocks at the level where melt percolation took place; this temperature is estimated to be 735 ± 20 °C. Thus, continuous cooling of the Brezovica ultramafic rocks after segregation of a melt produced during the first magmatic stage is reflected in the two-pyroxene reaction closure temperatures in the first group of harzburgites. Continuous cooling after the second heating event and segregation of melt during the second magmatic stage are reflected in the closure temperatures of the third group of harzburgites. During continuous cooling, the closure temperatures of both the two-pyroxene and olivine-spinel reactions depend on cooling rate. It was established by a physical modelling that the main factor influencing the mantle rock cooling rate is the thickness of the lower lithospheric level where conductive heat transfer occurs (Bazylev & Silantyev 1999, 2000a). The thickness of this level (a09 km) can be estimated directly from the calculated average temperatures of the two-pyroxene (7J,p, °C) and olivine-spinel (7J>S, °C) equilibria using the equations of Bazylev & Silantyev (2000a):

and

where T^ (°C) is the maximum temperature of the hydrothermal fluid penetrating the rocks (i.e. the temperature at the upper surface of the zone of conductive heat transfer in the lithosphere), which can be estimated from the data on the metamorphic mineralogy. An average value of 550 °C can be adopted if mineral data are lacking. It is also possible to evaluate the pressure at which the last melt fractions were segregated from the mantle peridotites from the thickness of the zones of conductive and convective heat transfer in the lithosphere by using the equation (Bazylev & Silantyev 2000a)

where a0 is an average of estimates from both the two-pyroxene and olivine-spinel geothermometry, 40 (deg km~ l ) is an appropriate average geothermal gradient below an axial part of a mid-ocean ridge, 3.2 is a coefficient reflecting the average density of the lithosphere, and 0.3 is a correction for the average water depth.

The calculations performed for the Brezovica ultramafic rocks allow us to conclude that in the first magmatic stage, the last melt segregation occurred at 29 km, or at 9.3 kbar. During the second magmatic stage, the last melt segregation occurred at about 25 km, a depth roughly equivalent to a pressure of 8.2 kbar. A conductive layer of 12 km thickness is estimated for this magmatic event. Because lateral heat transfer may have occurred during cooling after cessation of melt percolation, the reported depth and pressure estimates for the second magmatic stage should be considered minimum values. The time necessary for the closure of the twopyroxene reaction in the first group of harzburgites at 900 °C indicates the minimum amount of time between the two melt stages during the evolution of the Brezovica ultramafic rocks; this time is estimated to be c. 0.5 Ma (Bazylev & Silantyev 2000b).

Discussion Magmatic evolution of the Brezovica ultramafic rocks The mineral chemistry, bulk-rock chemistry and thermal history of the Brezovica ultramafic rocks provide evidence for two separate stages of melting with a cooling phase between them. The spinel harzburgites of the first group and probably the harzburgites of the second group were formed during the first melting stage. Spinel compositions of these rocks lie outside the midocean peridotite field (Fig. 3), but they are comparable with those for suprasubduction zone peridotites (Ishii et al. 1992). Relatively large variations in spinel Cr number in these rocks (0.46-0.64) indicate large local variations in the degree of partial melting (Hellebrand et al. 2001). Such variations cannot be produced by decompression melting in a MOR setting (e.g. Langmuir et al. 1992) and are not exhibited by MOR spinel harzburgites (Dick & Natland 1996). However, they can be produced by melting induced by fluid or melt input in a hot mantle in a suprasubduction zone setting (Ishii et al. 1992; Parkinson & Pearce 1998). Nevertheless, the obtained mineral and bulk-rock chemistry does not seem to have been influenced significantly by the influx of such subduction zone component. In the harzburgites of the first and the second groups, the increase in spinel Cr number is accompanied by an increase in the Mg number of olivine and orthopyroxene (Fig. 4), an increase in the olivine nickel content, an increase in the bulk-rock Cr number (Fig. 6), and a decrease in the bulk-rock FeO/SiO2 (Fig. 8) and Ti/Al (Fig. 9) ratios. These

BREZOVICAULTRAMAFIC MASSIF, SERBIA features suggest simple partial melting of a mantle that was not affected by melt-rock interaction. Therefore, the harzburgites of these two groups represent different stages of mantle partial melting, with the second group harzburgites corresponding to higher degrees of melting. Additional evidence for a suprasubduction setting of the first magmatic stage is provided by the estimated pressure at which the last melt fraction was segregated. This value (9.3 kbar), coupled with the spinel Cr number in these rocks, indicates deeper conditions of melting than those established for MOR settings (Bazylev & Silantyev 2000b). Knowing the pressure at which the last melt fraction was segregated from the mantle and the spinel Cr number in these rocks, we can estimate the aluminium content of the melt using the expression of Bazylev (1996):

This expression yields 14.3 wt% A12O3 for the melt segregated from the first group of harzburgites during the first magmatic stage. This value is consistent with the tholeiitic nature of the melt. The oxygen fugacity during the first magmatic stage, calculated using the method of Ballhaus et al. (1991), is on average 1.7 log units below the fayalite-magnetite-quartz buffer (FMQ) for the first group of harzburgites and 0.3 log units below FMQ for the second group. These values are generally within the ranges typical for both MOR and suprasubduction zone spinel peridotites (Parkinson & Pearce 1998). Thus we suggest that the first magmatic stage in the Brezovica ultramafic rocks occurred in a suprasubduction zone setting and resulted in the segregation of melts with a tholeiitic affinity. During the second magmatic stage, the harzburgites of the third group and the dunites were formed in the Brezovica massif. As inferred from the thermal history of these rocks, this stage occurred after they cooled to temperatures of about 900 °C as a result of local heating to solidus temperatures by the percolating melt. The ranges of spinel Cr number in the third group of harzburgites (0.62-0.72) and dunites (0.69-0.80) indicate a suprasubduction setting of this magmatic stage (Dick & Bullen 1984; Arai 1994). The aluminium content in the melt segregated from harzburgites of the third group during the second magmatic stage is estimated to be 11.1 wt% A12O3. Such a low aluminium content is indicative of melts that are intermediate in composition between tholeiite and high-Ca boninite, although it is more typical of boninitic melts.

103

The oxygen fugacity during the second magmatic stage is calculated to be 0.7 log units above FMQ. This value lies outside the range typical for MOR spinel peridotites but is consistent with the values for suprasubduction zone peridotites (Parkinson & Pearce 1998). However, the compositional features of the primary minerals and rocks formed at this stage are not consistent with a simple partial melting process. In particular, the Mg number of olivines and pyroxenes in harzburgites of the third group and in harzburgite BR-3 is significantly lowered compared with the trend of mantle partial melting (Fig. 4), the nickel content in olivines from these rocks is lower than for the first and second group of harzburgites, the titanium contents in the Crrich spinels are elevated (Fig. 5), the sodium content in the clinopyroxenes is higher than for the first and second group harzburgites, and the bulk-rock FeO/SiO2 and Ti/Al ratios are significantly higher than those of the first and second group harzburgites (Figs 8 and 9). All these features are typical for melt-rock interaction. This process is related to silica undersaturation of deep mantle melts during their ascent and results in dissolution of orthopyroxene from the wall rocks and crystallization of olivine (and sometimes of Cr-rich spinel), accompanied by partial chemical re-equilibration between the melt and wall rocks (Kelemen et al 1992, 1995). The dissolution of orthopyroxene in the Brezovica harzburgites is manifested by an elevated bulk-rock Cr number (Fig. 6) and FeO/SiO2 ratio (Fig. 8), whereas crystallization of olivine is indicated by its low Mg number (Fig. 4) and nickel content. Chemical re-equilibration with the melt is manifested by a lower orthopyroxene Mg number (Fig. 4), an elevated sodium content in clinopyroxene, and an elevated titanium content in both spinel and bulkrock compositions. Dunite formation indicates a prograde stage of melt-rock interaction manifested by the complete dissolution of orthopyroxene (Kelemen et al. 1995). The melt evolution accompanying this process results in clinopyroxene crystallization (Dick & Natland 1996) and locally in the crystallization of hornblende in dunites (Arai & Matsukage 1996). These features are found in the dunites of the Brezovica massif. A coupled increase in the titanium content and degree of iron oxidation in the Cr-rich spinel within the rock suite and within a single rock (dunite BR-12) (Fig. 5) is also indicative of melt-rock interaction (Allan & Dick 1996). In most of the Brezovica dunites, olivine has high Mg number and high nickel contents, which are also indicative of melt-rock interaction rather than crystallization of dunite from a melt (Dick & Natland 1996). Finally, most of the

104

B. A. BAZYLEV ETAL.

investigated Brezovica dunites have bulk-rock Cr2O3/SiO2 ratios similar to those of the residual harzburgites (Fig. 7). This observation is easily explained by orthopyroxene dissolution in the initial harzburgite, but would not be compatible with direct crystallization of the dunites from a melt. The titanium contents in the magmatic hornblende from the dunite BR-12 (0.63 wt% TiO2) and in associated spinel (0.46 wt% TiOa) are significantly lower than the titanium contents in these minerals from most refractory abyssal spinel peridotites formed in MOR segments (Arai & Matsukage 1996; Bazylev et al. 2001). Because water and titanium are both incompatible during melt fractionation, the titanium content of the first hornblende to crystallize from a melt should reflect the titanium content of the melt at the time of its water saturation. For melts with similar major element compositions, which are differentiated in similar conditions, the titanium content in hornblende reflects the water content (water/titanium ratio) in an initial melt: the higher the water content in the primary melt, the earlier melt saturation in water occurs, and the lower the titanium contents in the hornblende and related melt. Hence, we infer that the melt that percolated through the Brezovica ultramafic rocks during the second magmatic stage was significantly richer in water than primary MOR melts and was characterized by a very high water/titanium ratio. This is consistent with the boninite affinity of this melt and its suprasubduction zone affinity. Therefore, we infer that the second magmatic stage recorded in the Brezovica ultramafic rocks occurred in a suprasubduction geodynamic setting and that it resulted in the segregation of melts with a high-Ca boninite affinity Interaction between spinel harzburgites and percolating melt, rather than partial melting, took place at this stage. As mentioned above, basalts are not preserved in the Brezovica massif possibly because of erosion. Rare gabbros (fresh and rodingitized) from the upper part of the Brezovica massif in the Livad area show extremely low titanium contents, below 0.15wt% TiOi (Batocanin & Memovic 1996), and from this parameter we infer that the gabbro may be cogenetic with melts of boninite affinity segregated from the Brezovica harzburgites.

Metamorphic evolution of the Brezovica ultramafic massif It is surprising that the Brezovica harzburgites show no signs of a thermal influence from the

overlying cumulate rocks. This can be explained either by a genetic link between the dunite bodies in the mantle harzburgites and the dunites and peridotites in the cumulate sequence, or by a tectonic contact between the mantle and the crustal (cumulate) parts of the ultramafic sheet. In the first case, repeated heating of the mantle harzburgites is related not only to melt penetration, but also to the crystallization of the cumulate rocks that should have taken place at c. 25 km depth. We do not yet have any direct evidence for such deep crystallization. On the other hand, the nature of the contact between the mantle and the cumulate rock sequences in Brezovica is obscured by intensive serpentinization, obscuring a possible tectonic contact. Another unexpected result of our investigation is the absence of any influence of the metamorphism related to the metamorphic sole below the ultramafic rocks on the closure temperatures of exchange reactions between the primary minerals in peridotites. Such an influence should be evident if the ultramafic rocks were as hot at the time of their tectonic juxtaposition with cold olistostromal rocks as was proposed earlier, 900 °C or even higher (Karamata 1968a, 1968b; Djordjevic et al. 1987). As is indicated by numerical physical and thermodynamic modelling (Bazylev & Silantyev 2000b), the difference between the two-pyroxene and olivine-spinel temperatures for all the Brezovica harzburgites should be similar and large (c. 150-200 °C), and the calculated temperatures in the harzburgites should fall regularly away from the contact zone. However, the observed distribution of the closure temperatures of exchange reactions is consistent with a local secondary heating process, as demonstrated above. Preservation of this distribution requires that the initial temperature of the Brezovica ultramafic rocks at the time of their tectonic juxtaposition with the olistostromal melange should not exceed the minimal value of the olivine-spinel, i.e. about 735 ± 20 °C. This temperature is close to the maximum metamorphic temperature of the olistostromal rocks in the underlying sole (600-700 °C). This problem disappears if the hot ultramafic rocks were not instantaneously juxtaposed with the cold olistostromal rocks (where the maximal contact temperature should be an average of the two initial temperatures (Jaeger 1968)). Beginning with development of the thrust fault, the olistostromal rocks were progressively heated by every new part of the moving ultramafic slice coming into contact with the structurally lower rocks. The time required to heat the olistostromal rocks is estimated to be 0.5-5 Ma assuming an initial depth of thrusting of 25 km, a thrust rate of 1-10 cm a"1

BREZOVICA ULTRAMAFIC MASSIF, SERBIA and a fault zone dip of 30°. This time seems to be sufficient to provide the necessary heat for the olistostromal rocks. Therefore, it can be expected that the maximum contact temperature would be only c. 50-100 °C lower than the initial temperature of the ultramafic rocks that is consistent with the data from the metamorphic sole. The metamorphic mineral chemistry and the estimated temperatures of the amphibolite- to greenschist-facies metamorphism of the ultramafic rocks indicate that the temperature of the peridotite metamorphism falls from the bottom to the top of the ultramafic slice. This raises the question of whether the metamorphism was related to fluid circulation in the lithosphere or to a high-gradient metamorphic aureole similar to that described in the olistostromal rocks at the sole of the ultramafic body. Although the data presented in this paper are not sufficient to resolve this problem, we believe that they will stimulate continuing investigations. The last metamorphic stage recognizable in the Brezovica ultramafic rocks was serpentinization. Both the isochemical character of the serpentinization and the presence of brucite associated with serpentine in these rocks argue against a MOR setting for this process. In fact, the serpentinization of abyssal peridotites is not an isochemical process, and brucite is generally absent in these rocks (Snow & Dick 1995). Both the isochemical character of the serpentinization and the presence of brucite are typical of alpine-type peridotites (Wicks & Plant 1979), which seem to be serpentinized in the upper crust by meteoric water (Wenner & Thaylor 1974), or in a frontal part of the mantle wedge in a suprasubduction zone setting by fluid derived from subducted sediments (Dmitriev et al. 1999). In any case, serpentinization of the Brezovica ultramafic rocks appears to have occurred after they were technically juxtaposed with the olistostromal rocks.

Initial geological setting of the Brezovica massif To identify the Brezovica massif as part of the Central Dinaridic ophiolite belt or the Vardar zone we need to compare the Brezovica ultramafic rocks with possible analogues in the neighbouring areas. The Brezovica peridotites are most similar geologically and petrographically to the harzburgite massifs of the eastern belt of Albania (eastern-type massifs: Kukes, Bulqiza, Shebenic). Spinel compositions in harzburgites and dunites from all these massifs (Bebien et al. 1998) are close to those of the Brezovica body, with Cr number 0.4-0.8. In the Bulqiza massif, harzbur-

105

gites with relatively low-chromian spinels (0.55) occur in the lowest part, harzburgites with subordinate dunites with high-chromian spinels (0.76-0.80) in the middle part, dunites with subordinate harzburgites (Spl Cr number 0.810.835) in the upper part, and cumulates with basal dunites (Spl Cr number 0.82-0.835) in the uppermost part (Alliu et al. 1994). The possible initial presence of very low-Ti basalts and boninites in the Brezovica massif is inferred from the mineralogy and chemistry of the ultramafic and gabbroic rocks, and similar volcanic rocks are widespread in the Albanian massifs (Bebien et al. 2000). Similar to the Brezovica massif, the Albanian harzburgite massifs are allochthonous slices underlain by metamorphic soles that originated under similar conditions. The thickness of the mantle rock sequence in the Djakovica area (which is immediately adjacent to Albania; Fig. 1) is >7 km (Roksandic 1974), similar to that of the Albanian ultramafic massifs (Alliu et al. 1994). Considering these data, as well as the existence of the numerous peridotite outcrops between Djakovica and Brezovica, the decrease in the thickness of the ultramafic rocks and of the metamorphic sole eastward, and the higher temperature of metamorphism beneath the ultramafic slices westward, the Brezovica ultramafic massif can be interpreted as the frontal, easternmost relic of suprasubduction lithosphere emplaced during and after the closure of the Central Dinaridic basin. Therefore, the hypothesis of Nicolas & Boudier (1999) that the eastern-type Albanian massifs formed at a fast-spreading mid-ocean ridge is not supported by the data presented in this paper.

Conclusions (1) The Brezovica ultramafic rocks in southern Serbia underwent two separate magmatic stages, as is inferred from the mineral and bulk-rock chemistry and their thermal history. In the first stage, a suite of spinel harzburgites was formed as a result of partial melting to produce tholeiitic melts. During the second stage, these spinel harzburgites were repeatedly heated and affected by percolating melt. This process resulted in the formation of dunites and anomalous spinel harzburgites by melt-harzburgite interaction. The melt that segregated from these rocks during the second magmatic stage was high-Ca boninite. Both magmatic stages occurred in a suprasubduction geodynamic setting at a relatively deep level (2528km). (2) At the time of tectonic juxtaposition of the Brezovica massif with underlying olistostromal rocks that produced a metamorphic sole beneath

106

B. A. BAZYLEV ETAL.

the massif, the Brezovica ultramafic rocks were cooled to temperatures not exceeding 735 ± 20 °C. (3) The initial geological setting of the Brezovica ultramafic rocks is related to the Central Dinaridic-Mirdita basin. The Brezovica ultramafic massif in its present position is interpreted as the easternmost part of the eastward-emplaced Mirdita-Djakovica-Orahovac-Brezovica 'nappe' (wedge), which represents the obducted fragments of a suprasubduction oceanic-type lithosphere formed in the Central Dinaridic-Mirdita basin. The authors thank N. N. Kononkova (Vernadsky Institute, Moscow) for providing the electron microprobe data, and I. A. Roshchina and T. V Romashova (Vernadsky Institute, Moscow) for providing bulk-rock analyses. J. Allan, E. Rampone, J. Bebien, and Y. Dilek are thanked for their constructive criticism and suggestions for improvements of the manuscript. The work was supported by Russian Foundation of Basic Research grant 01-05-64288.

References ALIAJ, SH. & MECO, S. 1994. Mirdita oceanic basin was located westwards of Korab zone (=Pelagonian zone). Ofioliti, 19, 97-103. ALLAN, J.F. & DICK, H.J.B. 1996. Cr-rich spinel as a tracer for melt migration and melt-rock interaction in the mantle: Hess Deep, Leg 147. In: MEVEL, C., GILLIS, K.M., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, College Station, TX, 147, 157-172. ALLIU, I., BECCALUVA, L. & f INA, A. ET AL.1994. The Skanderbeg and Bulqiza mafic-ultramafic ophiolitic complexes and their relationships with chromititic ore deposits (Working Group Meeting of IGCP Project 256—Field Trip B). Ofioliti, 19, 27-55. ANDERS, E. & GREVESSE, N. 1989. Abundances of the elements: meteoritic and solar. Geochimica et Cosmochimica Acta, 53, 197-214. ARAI, S. 1994. Characterization of spinel peridotites by olivine-spinel compositional relationships: review and interpretation. Chemical Geology, 113, 191 — 204. ARAI, S. & MATSUKAGE, K. 1996. Petrology of gabbrotroctolite-peridotite complex from Hess Deep, equatorial Pacific: implications for mantle-melt interaction within the oceanic lithosphere. In: MEVEL, C., GILLIS, K.M., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, College Station, TX, 147, 135-156. BALLHAUS, C., BERRY, R.F. & GREEN, D.H. 1991. High pressure experimental calibration of the olivineorthopyroxene-spinel oxygen geobarometer: implications for the oxidation state of the upper mantle. Contributions to Mineralogy and Petrology, 107, 27-40. BATOCANIN, D. & MEMOVIC, E. 1996. Two members of a dismembered oceanic crust, the deepest and the

highest, at Budanac, Brezovica. In: KNEZEVIC, V. & KRSTIC, B. (eds) Terranes of Serbia. Barex, Belgrade, 265-268. BAZYLEV, B.A. 1996. Compositions of clinopyroxene and spinel in restite hyperbasites as indicators of the generation conditions and compositions of coupled primary mantle magmas. Geochemistry International, 33, 115-124. BAZYLEV, B.A. 2000. Appearance of the awaruitebearing mineral assemblage in peridotites from 15°20'N FZ (Atlantic Ocean) as one of manifestations of the oceanic metamorphism. Russian Journal of Earth Sciences, 2(3/4), 279-293 (in Russian). BAZYLEV, B.A. & SILANTYEV, S.A. 1999. Geodynamic interpretation of subsolidus metamorphism of mantle spinel peridotites. Ofioliti, 24, 61-62. BAZYLEV, B.A. & SILANTYEV, S.A. 2000a. Geodynamic interpretation of subsolidus recrystallization of mantle spinel peridotites: 1. Mid-oceanic ridges. Petrology, 8, 201-213. BAZYLEV, B.A. & SILANTYEV, S.A. 2000b. Geodynamic interpretation of subsolidus recrystallization of mantle spinel peridotites: 2. Ophiolites and xenoliths. Petrology, 8, 311-331. BAZYLEV, B.A., MAGAKYAN, R., SILANT'EV, S.A., IGNATENKO, K.I., ROMASHOVA, T.V. & XENOPHONTOS, C. 1993. Petrology of peridotites from Mamonia complex, southwestern Cyprus. Petrology, 1, 302-330. BAZYLEV, B.A., ZAKARIADZE, G.S., ZHELYAZKOVAPANAYOTOVA, M.D., KOLCHEVA, K., OBERHAENSLI, R.E. & SOLOV'EVA, N.V. 1999. Petrology of ultramafic rocks from the ophiolite association in the crystalline basement of the Rhodope massif. Petrology,!, 192-211. BAZYLEV, B.A., SILANTYEV, S.A., DICK, H.J.B. & KONONKOVA, N.N. 2001. Magmatic amphiboles and micas in oceanic peridotites and some specific features of the related magmas: 15°20'N MAR Fracture Zone. Russian Journal of Earth Sciences, 3(3), 219-234. BEBIEN, J., SHALLO, M., MANIK, K. & GEGA, D. 1998. The Shebenik massif (Albania): a link between MOR- and SSZ-type ophiolites. Ofioliti, 23, 7-16. BEBIEN, J., DIMO-LAHITTE, A., VERGELY, P., INSERGUEix-FiLippi, D. & DUPEYRAT, L. 2000. Albanian ophiolites. I—Magmatic and metamorphic processes associated with the initiation of a subduction. Ofioliti, 25, 39-46. BECCALUVA, L., COLTORTI, M., PREMTI, L, SACCANI, E., SIENA, F. & ZEDA, O. 1994. Mid-ocean ridge and suprasubduction affinities in the ophiolitic belts of Albania. Ofioliti, 19, 77-96. BIZIMIS, M., SALTERS, V.J.M. & BONATTI, E. 2000. Trace and REE content of clinopyroxenes from supra-subduction zone peridotites. Implications for melting and enrichment processes in island arcs. Chemical Geology, 165, 67-85. CAROSI, R., CORTESOGNO, L., GAGGERO, L. & MARRONI, M. 1996. Geological and petrological features of the metamorphic sole from the Mirdita nappe, northern Albania. Ofioliti, 21, 21-40. CIRIC, A. & ERIC, V. 1996. Contact metamorphism

B R E Z O V I C A U L T R A M A F I C MASSIF, SERBIA beneath the peridotite of Borov Vrh-Brezovica (Serbia). In: KNEZEVIC, V. & KRSTIC, B. (eds) Terranes of Serbia. Barex, Belgrade, 259-264. DICK, H.J.B. 1989. Abyssal peridotites, very slow spreading ridges and ocean ridge magmatism. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 71-105. DICK, H.J.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DICK, H.J.B. & FISHER, R.L. 1984. Mineralogic studies of the residues of mantle melting: abyssal and alpine-type peridotites. In: KORNPROBST, J. (ed.) Kimberlites II. The Mantle and Crust-Mantle Relationships. Elsevier, Amsterdam, 295-308. DICK, H.J.B. & NATLAND, J.H. 1996. Late-stage melt evolution and transport in the shallow mantle beneath the East Pacific Rise. In: MEVEL, C., GILLIS, K.M., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, College Station, TX, 147, 103-134. DICK, H.J.B., FISHER, R.L. & BRYAN, W.B. 1984. Mineralogical variability of the uppermost mantle along mid-ocean ridges. Earth and Planetary Science Letters, 73, 91-104. DJORDJEVIC, P., JOVANOVIC, L., KARAMATA, S., KURAT, G., MAJER, V. & NTAFLOS, TH. 1987. Thermal and tectonic history of some ultramafic rocks of Brezovica—the Livad mass, Yugoslavia. Ofioliti, 12, 245. DMITRIEV, L.V., BAZYLEV, B.A., SILANTYEV, S.A., BORISOV, M.V., SOKOLOV, S.Yu. & BOUGAULT, H. 1999. Hydrogen and methane formation during serpentinization of mantle ultramafic rocks of ocean and oil origin. Russian Journal of Earth Sciences, 1(6), 511-519 (in Russian). EVANS, B.W. 1977. Metamorphism of alpine peridotite and serpentinite. Annual Review of Earth and Planetary Sciences, 5, 397-447. HELLEBRAND, E., SNOW, J.E., DICK, H.J.B. & HOFMANN, A. 2001. Coupled major and trace elements as indicators of the extent of melting in mid-oceanridge peridotites. Nature, 410, 677-681. ISHII, T., ROBINSON, P.T., MAEKAWA, H. & FISKE, R. 1992. Petrological studies of peridotites from diapiric serpentinite seamounts in the Izu-Ogasawara— Mariana forearcs. In: FRYER, P., PEARCE, J.A., STOKKING, L.R. ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results. Ocean Drilling Program, College Station, TX, 125, 445486. JAEGER, J.C. 1968. Cooling and solidification of igneous rocks. In: HESS, H.H. & POLDERVAART, A. (eds) Basalts. Interscience, New York, 2, 503-536. KARAMATA, S. 1968a. Ultrabasic (ultramafic) intrusions of Dinarides. Zapiski Vsesoyuznogo Mineralogicheskogo Obschestva, XCVII, 49-54 (in Russian). KARAMATA, S. 1968b. Zonality in contact metamorphic rocks around the ultramafic mass of Brezovica (Serbia, Yugoslavia). In: Proceedings of the XXIIIth International Geological Congress, Prague. 1, 197207.

107

KARAMATA, S. 1985. Metamorphism in the contact aureole of Brezovica (Serbia, Yugoslavia) as a model of metamorphism beneath obducted hot ultramafic bodies. Bulletin XCX de I'Academic Serbe des Sciences et des Arts, Classe des Sciences Naturelles et Mathematique, 26, 51-68. KARAMATA, S. & KRSTIC, B. 1996. Terranes of Serbia and neighbouring areas. In: KNEZEVIC, V. & KRSTIC, B. (eds) Terranes of Serbia. Barex, Belgrade, 25-40. KARAMATA, S. & MILOVANOVIC, D. 1990. Metamorphism beneath the Brezovica ultramafics in the Ostrovica area. In: XII Kongres geologia Jugoslavije, II, Ohrid. 290-300 (in Serb.-Croat.). KARAMATA, S., MAJER, V. & PAMIC, J. 1980. Ophiolites of Yugoslavia. Ofioliti, Tethian ophiolites (special issue), 1, 105-125. KARAMATA, S., KORIKOVSKIY, S.P., BORONIHI, V.A. & ZINOVYEVA, N.G. 1991. Equilibria in metapsammites and mica gneisses from a contact aureole in a basement of ophiolitie ultramafic massif Brezovica (south Serbia). Izvestiya Akademii Nauk SSSR, Seriya Geologicheskaya, 10, 58-70 (in Russian). KARAMATA, S., DIMITRIJEVIC, M.D. & DIMITRIJEVIC, M.N. 2000. A correlation of ophiolitie belts and oceanic realms of the Vardar zone and the Dinarides. In: Proceedings of the International Symposium 'Geology and Metallogeny of the Dinarides and the Vardar Zone'. Academy of Sciences and Arts of the Republic of Srpska, Collections and Monographs, Department of Natural, Mathematical and Technical Sciences, 1, 191-194. KELEMEN, P.B., DICK, H.J.B. & QUICK, J.E. 1992. Formation of harzburgite by pervasive melt/rock reaction in the upper mantle. Nature, 358, 635-641. KELEMEN, P.B., SHIMIZU, N. & SALTERS, VJ.M. 1995. Extraction of mid-ocean ridge basalts from the upwelling mantle by focused flow of melt in dunite channels. Nature, 375, 747-753. LANGMUIR, C.H., KLEIN, E. & PLANK, T. 1992. Petrological systematics of mid-ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: PHIPPS MORGAN, J., BLACKMAN, D.K. & SINTON, V.M. (eds) Mantle Flow and Melt Generation at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union, 71, 183-280. LUGOVIC, B., ALTHERR, R., RACZEK, I., HOFMANN, A.W. & MAJER, V. 1991. Geochemistry of peridotites and mafic igneous rocks from the Central Dinaric Ophiolite Belt, Yugoslavia. Contributions to Mineralogy and Petrology, 106, 201-216. McDoNOUGH, W.F. & SUN, S.S. 1995. The composition of the Earth. Chemical Geology, 120, 223-253. NICOLAS, A. & BOUDIER, F. 1999. Slow spreading accretion and mantle denudation in the Mirdita ophiolite (Albania). Journal of Geophysical Research, 104, 15155-15167. PAMIC, J. 1983. Considerations on the boundary between Iherzolite and harzburgite subprovinces in the Dinarides and northern Hellenides. Ofioliti, 8, 153-164. PARKINSON, I.J. & PEARCE, J.A. 1998. Peridotites from the Izu-Bonin-Mariana Forearc (ODP Leg 125): evidence for mantle melting and melt-mantle interaction in a supra-subduction zone setting. Journal

108

B. A. BAZYLEV ETAL.

of Petrology, 39, 1577-1618. ROBERTSON, A.H.F. & KARAMATA, S. 1994. The role of subduction-accretion processes in the tectonic evolution of the Mesozoic Tethys in Serbia. Tectonophysics, 234, 73-94. ROBERTSON, A. & SHALLO, M. 2000. Mesozoic-Tertiary tectonic evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197— 254. ROKSANDIC, M. 1974. Application of results of geophysical explorations in studying of structural features and evolution of Dinarides. Glas SANU CCXCII, Prirodny i Matematicny Nauke, Beograd, 38, 61-95 (in Serb.-Croat.). SHALLO, M. 1996. Relationships between Albanian ophiolites and Korab-Pelagonian microplate. In: KNEZEVIC, V. & KRSTIC, B. (eds) Terranes of

Serbia. Barex, Belgrade, 231-236. SHIBATA, T. & THOMPSON, G. 1986. Peridotites from the Mid-Atlantic Ridge at 43°N and their petrogenetic relations to abyssal tholeiites. Contributions to Mineralogy and Petrology, 93, 144-159. SNOW, I.E. & DICK, H.J.B. 1995. Pervasive magnesium loss by marine weathering of peridotite. Geochimica et Cosmochimica Ada, 59, 4219-4235. WELLS, P.R.A. 1977. Pyroxene thermometry in simple and complex systems. Contributions to Mineralogy and Petrology, 62, 129-139. WENNER, D.B. & THAYLOR, H.P. JR 1974. D/H and O18/ O16 studies of serpentization of ultramafic rocks. Geochimica et Cosmochimica Acta, 38, 1255-1286. WICKS, F.J. & PLANT, A.G. 1979. Electron-microprobe and X-ray microbeam studies of serpentine textures. Canadian Mineralogist, 17, 785-830.

Triassic mid-ocean ridge basalts from the Argolis Peninsula (Greece): new constraints for the early oceanization phases of the Neo-Tethyan Pindos basin EMILIO SACCANI 1 , ELISA PADOA 2 & ADONIS PHOTIADES 3 1

Dipartimento di Scienze della Terra, Universita di Ferrara, C.so Ercole I d'Este 32, 44100 Ferrara, Italy (e-mail: [email protected]) 2 Dipartimento di Scienze della Terra, Universita di Firenze, Via G. La Pira 4, 50121 Florence, Italy 3 'Institute of Geology and Mineral Exploration (IGME), 70 Messoghion Str., 11527 Athens, Greece Abstract: The Middle Unit of the central-northern Argolis Peninsula, in NE Peloponnesus (Greece), is composed of several tectonic slices, locally including intact sequences of mafic volcanic rocks topped by radiolarian cherts. Although some of these sequences are Jurassic in age, many of them display a Triassic age based on biostratigraphical evidence. The petrological studies presented in this paper indicate that the Triassic volcanic rocks were generated in a mid-ocean ridge setting, and that they represent the oldest remnants of the Pindos oceanic crust so far recognized in the Subpelagonian zone. On the basis of immobile trace element analyses, two chemically distinct groups of Triassic lavas can be recognized in the various volcanic sequences. One group is represented by transitional-type mid-ocean ridge basalts (T-MORBs) displaying moderate light rare earth element (LREE) enrichment, and incompatible element abundances very similar to those observed in present-day T-MORBs. The other group exhibits a range of characteristics typical of many normal-type MORBs: that is, variable LREE depletion and flat N-MORB normalized patterns of incompatible element abundance. Moreover, many geochemical characteristics indicate that the various N-MORB type volcanic sequences originated from chemically distinct (heterogeneous) sub-oceanic mantle sources. Analogous to similar basalts from ophiolitic melanges of the Dinaride-Hellenide belt, the T-MORBs from the Argolis Middle Unit are interpreted as having originated from a primitive mantle source variably enriched by an ocean-island basalt (OIB)-type component. In contrast, the contemporaneous occurrence of N-MORBs implies that, during the Mid-Late Triassic, oceanic spreading of the Pindos basin had already reached, at least in some sectors, a quasi-steady state involving only sub-oceanic mantle sources and their partial melt derivatives. Our model for the Triassic opening of the Pindos oceanic basin and its related tectonomagmatic evolution is largely supported by comparison with the Red Sea embryonic ocean, a modern analogous setting.

The Pindos ocean is one of the Neo-Tethyan basins in the Eastern Mediterranean that developed during the early Mesozoic along the northern margin of Gondwanaland (Robertson et al 1991). Remnants of the Pindos basin are widely preserved in late Mesozoic-Cenozoic accretionary complexes in the Mirdita-Subpelagonian zone of the Dinaride-Hellenide belt and are mainly represented by complex tectonosedimentary associations of Triassic rift-related volcanic rocks (Pe-Piper 1998, and references therein), Middle Jurassic mid-ocean ridge (MOR) and suprasubduction-zone (SSZ) ophiolitic sequences (Jones & Robertson 1991; Capedri et al. 1996), Permian-Jurassic marginal and platform-related sedimentary rocks and Juras-

sic-Tertiary trench-type sedimentary rocks (Clift & Robertson 1989; Robertson 1994; Degnan & Robertson 1998). The nature and composition of these various rock types suggest a general geodynamic evolution of the Pindos basin characterized by Late Permian-Triassic rifting phases between the Apulian and Pelagonian microplates, followed by Jurassic spreading of the Pindos Neo-Tethyan oceanic basin and subsequent development of intra-oceanic convergent zones, as well as MidLate Jurassic generation of oceanic lithosphere in a suprasubduction setting (Jones & Robertson 1991; Doutsos et al 1993). A Triassic age for the beginning of spreading in the Hellenide sector of the Pindos ocean has been

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 109-127. 0305-8719/037$ 15 © The Geological Society of London 2003.

110

E. SACCANI ETAL.

proposed by some workers (Ferriere 1982; Robertson et al 1991; Robertson et al 1996; PePiper 1998). However, this conclusion is generally based on limited data regarding Triassic radiolarites, which are commonly tentatively associated with basaltic sequences mainly showing withinplate (alkaline) affinity or, subordinately, ranging from transitional within-plate to transitional midocean ridge basalt (T-MORB) compositions. Moreover, a number of available biostratigraphical and radiometric ages (Spray et al. 1984; Bebien et al. 2000) indicate that the normal mid-ocean ridge magmatism largely developed during the Jurassic. Consequently, the possible Triassic beginning of oceanic development within the Hellenide sector of the Neo-Tethys is still poorly constrained. The Middle Unit of the central-northern Argolis (eastern Peloponnesus, Greece) is represented by a composite tectonic association of various types of thrust sheets, some of which include coherent sequences of basalts topped by radiolarian cherts. Cherts indicate two age ranges: MidLate Triassic (Bortolotti et al. 2001, 2002) and Mid-Late Jurassic (Baumgartner 1985). The Jurassic volcanic rocks are represented by MORB and are related to the magmatic activities that developed during the evolution of the Jurassic Pindos oceanic lithosphere (Baumgartner 1985; Dostal et al. 1991). By contrast, the nature of the Triassic basalts is still unknown. The main purpose of this paper is to present the petrological and geochemical characteristics of the Triassic basalts from the Middle Unit of the central-northern Argolis Peninsula to constrain their tectonomagmatic implications during the early stages of oceanic spreading in the Pindos basin.

Geology of the central-northern Argolis Peninsula The Argolis Peninsula (Fig. 1), in the southern sector of the Dinaride-Hellenide belt, is represented by a composite tectonic complex that was assembled during the closure of the Pindos basin, mainly from the Late Jurassic to the Miocene. Three main tectonostratigraphic units can be recognized in the central-northern Argolis Peninsula. The Lower Unit is represented by Mesozoic continental sequences (mainly made up of Middle Triassic-Early Jurassic carbonate successions), which represent both the subsiding continental platform of Apulia (Pantokrator Unit: Baumgartner 1985; Clift & Robertson 1990) and riftrelated intra-platform basins (Asklipion Unit: Clift & Robertson 1990). These sequences are overlain

by Oxfordian-Kimmeridgian siliceous mudstones and radiolarian cherts (Angelokastron and Koliaki Cherts: Baumgartner 1980, 1985, 1987), which, in turn, are stratigraphically followed by highly sheared ophiolite-derived sedimentary rocks (Lower Ophiolitic Unit or 'volcano-sedimentary ophiolitic succession': Photiades 1986, 1989). These rocks consist of coarsening upward turbidites, as well as ophiolitic sandstones and breccias (Late Oxfordian-Early Kimmeridgian Dhimaina Fm: Baumgartner 1985), followed by disorganized ophiolitic olistostromes (Potami Fm: Baumgartner 1985). The olistostromes include rounded fragments of boninitic lavas and coarse-grained boninitic-type rocks, set in an arenitic matrix consisting of various ophiolitic clasts, as well as fragments originating from the underlying limestones and cherts. The boninitic lavas and boninitic-type rocks originated in a subduction-related environment, possibly in an intra-oceanic islandarc setting (Dostal et al. 1991; Capedri et al. 1996). The Middle Unit consists of several imbricated tectonic slices, which can be roughly subdivided into two main types: (1) tectonic sheets mainly consisting of basic volcanic rocks, which locally have serpentinite slivers at their base and are associated with radiolarian cherts (Jurassic Migdhalitsa ophiolite unit: Baumgartner 1985); (2) tectonic sheets preserving well-developed stratigraphic successions, which consist of Albian-Cenomanian neritic limestones, talus-breccias rich in clasts of basalt and chert cemented by Campanian-Maastrichtian hemipelagic limestones, PaleoceneMiddle Eocene pelagic to reefal limestones ('mesoautochthonous sequence': Baumgartner 1985; Photiades & Skourtsis-Coroneou 1994), and finally a post-Ypresian flysch. The Upper Unit tectonically overlies the Middle Unit, and is almost always found in thrust contact over the post-Ypresian flysch. It is composed of a polymictic ophiolitic melange at its base, and Cretaceous limestones in its upper part (Photiades 1986). The ophiolitic melange contains metresized lens-shaped blocks, mainly including subduction-related ophiolitic rocks (serpentinized harzburgites, dunites, subordinate basalts and boninites: Dostal et al. 1991), cherts, carbonates, quartz greywackes, marbles, amphibolites and micaschists.

Structure and stratigraphy of the Middle Unit The Middle Unit of the central-northern Argolis Peninsula crops out mainly in the synclinal structures of Dhimaina and Lygourio-Palea Epidavros

TRIASSIC BASALTS OF PINDOS BASIN, GREECE

111

Fig. 1. Structural zones of the Albanide-Hellenide Alpine orogenic belt (modified after Robertson & Shallo 2000). The locations of the study areas expanded in Figure 2 are also indicated.

(Fig. 2a), and in the Vothiki-Radho graben (Fig. 2b). The tectonic emplacement of the Middle Unit over the Lower Unit is generally represented by a refolded thrust surface, probably related to a Late Jurassic-Early Cretaceous tectonic phase (Aubouin et al 1970). By contrast, the overthrusting of the Upper Unit onto the Middle Unit marks a subsequent, post-Eocene tectonic phase (Photiades 1986). The Middle Unit is a composite

tectonic unit consisting of several imbricated thrust sheets separated by shear zones. The basal and central parts of the Middle Unit consist predominantly of sheets of volcanic rocks, whereas wedges of Cretaceous-Eocene sedimentary successions are prevalent at the top of the unit. Contacts between the volcanic rocks and the Cretaceous-Eocene sedimentary sequences generally appear tectonic; no clear evidence of strati-

112

E. SACCANI ETAL.

Fig. 2. Simplified geological maps showing the main outcrops of the Middle Unit of the northern (a) and central (b) Argolis Peninsula.

TRIASSIC BASALTS OF PINDOS BASIN, GREECE graphic relationships can be observed in the field. None the less, previous workers (Baumgartner 1985; Photiades & Skourtis-Coroneou 1994) have suggested that the Cretaceous-Eocene sequence represents a diachronous, transgressive (i.e. 'mesoautochthonous') sedimentary package, deposited upon the volcanic sequences after their early emplacement onto the continental Lower Unit. In the lower and central parts of the Middle Unit, the intense fragmentation resulted in development of tectonic slivers of large blocks of various lithotypes, including mafic volcanic rocks, cherts and subordinate serpentinites, as well as marbles and limestones. These limestone blocks have been considered 'exotic' by Baumgartner (1985), and their occurrence further supports the hypothesis of a tectonic nature for the Middle Unit. The volcanic sequences include pillow lavas, massive lavas and pillow breccias, and are generally dismembered by several major thrust sheets that are, in turn, affected by second-order internal thrusts, shear zones and overturned folds. Many outcrops of pillow lavas and pillow breccias are spatially associated with ore deposits of Mn, FeCu and Ba, probably generated during sea-floor hydrothermal metamorphism (Photiades 1986; Robertson et al. 1987; Photiades & Economou 1991). Relatively continuous volcanic sequences, locally topped by radiolarian cherts, are preserved in places. Some of these basaltic sequences display a MORB magmatic affinity (Baumgartner 1985; Dostal et al. 1991) and are associated with Kimmeridgian-Tithonian radiolarian cherts (Baumgartner 1985). Consequently, the overall lower and central part of the Middle Unit has been regarded as a Jurassic ophiolitic unit (i.e. Migdhalitsa Unit: Baumgartner 1985). None the less, a number of basaltic sequences topped by Ladinian-Norian ribbon-radiolarian cherts have recently been found (Bortolotti et al. 2001, 2002), commonly in an overturned setting, as slices in the Middle Unit. Although the nature of these volcanic sequences is poorly understood, the occurrence of tectonic sheets composed of both Triassic volcanic chert and Jurassic ophiolitic sequences further supports the tectonic nature of the Middle Unit.

Sampling and methods Location of samples Analyses were carried out on the basaltic rocks from the Middle Unit of the northern (Fig. 2a) and central (Fig. 2b) Argolis areas. In particular, sampling was performed in those outcrops where

113

clear stratigraphic relationships between volcanic rocks and Triassic radiolarian cherts (Bortolotti et al. 2001, 2002) were observed. Various samples were collected from each outcrop to achieve maximum diversity between samples along the volcanic sections. The sampled stratigraphic sequences commonly appear partially dismembered by local shear zones. None the less, samples were also taken from tectonized sections, but only where small degrees of displacement were recognized, and where field relationships were clearly established to make sure that the whole sequence could be ascribed to a single thrust sheet. According to these criteria, seven sections were chosen (Figs 2 and 3) for sampling the mafic extrusive rocks overlain by Triassic radiolarian cherts. These sections are located: (1) to the south of Dhimaina, along the road to Lyghourio (sections GR195, GR56 and GR51); (2) to the west of Palea Epidavros, along the road to Lyghourio (sections GR47 and GR50); (3) in the southern part of Vothiki village (sections GR71 and GR181). The lithostratigraphy of the measured logs and the position of the collected samples inside each volcanic sequence are indicated in Figure 4, together with the position of the dated radiolarian cherts. The sampled volcanic facies include mafic pillowed (samples GR 47b, 51a, 51b, 56a, 195a, 71a, 71b, 71c, 71d, 181a, 181b, 181d) and massive flows (samples GR 56b, 56c, 181c), as well as pillow breccias and hyaloclastites (samples GR 47a, 50c, 50d, 50e, 56d). According to Bortolotti et al. (2001, 2002), sections GR71, GR181 and GR195 cannot be assigned to a precise age, but the radiolarian fauna are sufficiently preserved to allow their attribution to a Triassic sensu latu age. By contrast, the ages of sections GR47, GR50, GR51 and GR56 are well constrained by radiolarian dates, and range from Ladinian to Norian.

Analytical methods Samples were analysed for major and some trace elements (Ni, Co, Cr, y Rb, Sr, Ba, Th, Nb, Zr, Y, Zn) by X-ray fluorescence (XRF) using pressedpowder pellets (Table 1). The matrix correction methods proposed by Franzini et al. (1972) were applied. Accuracy is better than 2% for major oxides, and better than 5% for trace element determinations. The detection limits for trace elements range from 1 to 2 ppm. Both accuracy and detection limits were determined using results from international standards. Thirteen representative samples (Table 1) were chosen for additional trace-element analyses, including rare earth elements (REE), Sc, Nb, Hf, Ta,

114

E. SACCANI ETAL.

Fig. 3. Geological cross-sections of the northern and central Argolis Peninsula. No vertical exaggeration. (See Fig. 2 for legend and locations.)

Th and U, using inductively coupled plasma-mass spectrometry (ICP-MS). The precision and accuracy of the data were evaluated using results for international standard rocks, duplicate runs of several samples, and the blind standards included in the sample set. Accuracy varies from 1 to 8%, whereas detection limits (in ppm) are: 0.29 for Sc;

0.02 for Nb, Hf and Ta; <0.14 for REE; 0.011 for Th and U. All analyses were performed at the Department of Earth Science of Ferrara University using a Philips PW1400 automated X-ray spectrometer and a VG Elemental Plasma Quad PQ2 Plus spectrometer.

TRIASSIC BASALTS OF PINDOS BASIN, GREECE

115

Fig. 4. Measured logs of the sampled sections, showing the stratigraphic relationships between basalts and Triassic radiolarian cherts.

E. SACCANI ETAL.

116

Table 1. Bulk-rock major and trace element analyses of Triassic basaltic rocks from the Middle Unit of the centralnorthern Argolis Peninsula GR51

GR50

Section:

GR47

Sample: Age: Note:

GR47b Lad-Car Pillow

GR50c Nor Pill. Brec.

GR50d Nor Pill. Brec.

GR50e Nor Pill. Brec.

47.02 1.41 15.70 1.12 7.46 0.22 6.60 11.85 3.70 1.06 0.22 3.66 100.00 61.2 77 81 92 90 289 17 163 108 2 12 90 32 n.a. n.a.

50.51 1.60 14.79 1.16 7.72 0.19 8.36 8.53 4.43 0.67 0.22 1.82 100.00 65.9 81 61 71 82 293 7 225 143 2 11 99 34 n.a. n.a.

48.80 1.52 15.92 1.33 8.84 0.17 5.69 9.51 4.50 1.34 0.22 2.17 100.00 53.5 91 56 50 68 312 18 176 119 n.d. 13 99 35 8 14

48.46 1.56 15.59 1.31 8.74 0.16 5.60 9.57 4.61 1.39 0.21 2.80 100.00 53.3 91 66 48 91 319 20 138 161 2 11 98 35 14 n.a.

47.93 1.18 16.02 1.09 7.28 0.14 7.22 11.56 3.89 0.23 0.20 3.26 100.00 63.9 59 104 41 389 252 4 242 95 n.d. 8 84 26 4 3

21.5 10.6 7.08 17.2 2.61 12.3 3.89 1.40 4.36 0.81 5.24 1.05 3.16 0.44 2.87 0.42 2.66 5.44 0.94 0.44

25.9 9.69 7.24 18.2 2.71 13.1 4.09 1.52 4.69 0.87 5.71 1.12 3.47 0.49 3.16 0.46 2.86 2.44 1.01 0.34

25.4 9.38 7.86 18.5 2.73 13.1 4.21 1.51 4.67 0.89 5.78 1.16 3.57 0.51 3.31 0.49 2.90 2.34 1.01 0.40

23.1 6.87 5.20 13.0 1.96 9.5 2.94 1.12 3.36 0.62 4.07 0.81 2.44 0.34 2.17 0.33 2.16 2.22 0.65 0.20

XRF analyses Si02 TiO2 A12O3 Fe203 FeO MnO MgO CaO Na20 K2O P205 LOI Total Mg no. Zn Ni Co Cr V Rb Sr Ba Th Nb Zr Y La Ce ICP-MS analyse'S Sc Nb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Th U

XRF analyses SiO2 Ti02 A1203 Fe203 FeO MnO MgO CaO Na 2 O K20 P205 LOI Total

GR51b Nor Pillow

GR56a Car-Nor Pillow

GR56b Car-Nor MLF

GR56C Car-Nor MLF

GR56d Car-Nor Pill. Brec.

48.56 1.17 15.12 1.15 7.65 0.13 7.05 10.62 4.57 0.21 0.33 3.44 100.00 62.2 47 97 46 360 256 5 180 70 n.d. 9 82 23 8 13

43.53 1.68 14.01 1.40 9.32 0.17 6.57 12.94 4.43 0.06 0.42 5.47 100.00 55.7 64 85 38 240 281 4 98 25 n.d. 7 123 36 6 31

46.99 1.61 16.68 1.49 9.91 0.18 6.77 10.80 3.07 0.28 0.39 1.83 100.00 54.9 54 72 47 142 311 5 200 33 3 6 109 35 n.a. 18

48.67 1.68 14.78 1.41 9.38 0.14 6.90 10.15 4.07 0.12 0.28 2.44 100.00 56.7 73 65 42 139 305 5 81 21 2 6 116 35 n.a. 7

47.84 1.52 16.55 1.42 9.45 0.16 6.33 9.91 3.46 0.68 0.24 2.43 100.00 54.4 67 72 40 186 298 7 247 71 2 5 109 37 n.a. 5

23.8 5.87 6.74 17.9 2.84 14.0 4.39 1.49 4.79 0.90 5.75 1.15 3.48 0.48 3.15 0.46 3.56 1.28 0.71 0.50

29.4 4.23 4.93 13.8 2.28 11.8 4.07 1.43 4.55 0.89 5.88 1.20 3.69 0.53 3.40 0.51 3.17 1.04 0.46 0.14

30.2 3.94 5.04 14.0 2.32 11.8 4.03 1.44 4.63 0.89 5.97 1.22 3.69 0.53 3.37 0.52 3.07 0.81 0.43 0.12

GR181

GR71

Section: Sample: Age: Note:

GR51a Nor Pillow

GR56

GR 195

GR71a Triassic Pillow

GR71b Triassic Pillow

GR71c Triassic Pillow

GR71d Triassic Pillow

GR71e Triassic Pillow

GR181a Triassic Pillow

GR 181b Triassic Pillow

GR181c Triassic MLF

GR181d Triassic Pillow

GR 195a Triassic Pillow

46.84 2.82 14.44 1.73 11.52 0.47 6.87 10.26 3.24 0.04 0.29 1.47 100.00

45.70 2.79 14.30 1.78 11.90 0.47 6.82 10.96 3.07 0.04 0.31 1.85 100.00

48.81 2.36 14.43 1.54 10.26 0.46 7.24 8.53 4.09 0.08 0.25 1.94 100.00

46.26 2.80 14.56 1.75 11.66 0.48 7.12 11.19 2.83 0.04 0.30 1.01 100.00

45.76 2.89 14.79 1.80 11.98 0.46 6.66 10.37 3.26 0.05 0.31 1.67 100.00

44.46 1.38 15.50 1.19 7.91 0.18 6.61 13.12 3.53 0.66 0.21 5.26 100.00

48.52 1.50 16.30 1.27 8.48 0.16 7.32 11.17 3.51 0.10 0.19 1.47 100.00

46.37 1.35 15.70 1.17 7.79 0.19 6.64 11.68 3.91 0.64 0.21 4.35 100.00

41.49 1.30 14.19 1.11 7.43 0.17 6.21 17.01 3.25 0.24 0.26 7.34 100.00

50.55 2.43 13.91 1.63 10.85 0.24 6.55 10.10 2.57 0.10 0.30 0.78 100.00

TRIASSIC BASALTS OF PINDOS BASIN, GREECE GR71

Section: Sample: Age: Note: Mg no. Zn Ni Co Cr V Rb Sr Ba Th Nb Zr Y La Ce ICP-MS analyses Sc Nb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Th U

GR71b Triassic Pillow

GR71c Triassic Pillow

GR71d Triassic Pillow

GR71e Triassic Pillow

GR 181a Triassic Pillow

GR 181b Triassic Pillow

51.5 132 76 39 144 467 4 114 72 n.d. 5 195 49 n.a. n.a.

50.5 120 73 40 131 462 3 115 71 n.d. 5 193 49 n.a. 20

55.7 109 80 47 188 418 3 166 53 n.d. 4 169 45 58 23

52.1 117 75 48 137 457 5 125 19 4 6 199 52 8 n.a.

49.8 136 71 44 159 467 3 124 123 n.d. 5 198 52 6 n.a.

59.8 67 83 42 346 248 8 176 67 n.d. 3 107 33 3 n.a.

60.6 76 88 48 381 280 4 152 36 n.d. 4 115 34 n.a. 18

39.0 3.27 6.06 18.9 3.30 17.7 6.02 2.14 7.31 1.38 9.33 1.87 5.75 0.79 5.00 0.72 4.91 0.61 0.32 0.26

GR195

GR181

GR71a Triassic Pillow

35.1 3.74 9.18 24.4 3.78 19.8 6.57 2.22 7.63 1.49 9.86 2.00 6.25 0.90 5.96 0.88 5.77 0.82 0.35 0.28

117

41.2 3.66 7.89 22.1 3.64 19.7 6.70 2.30 8.19 1.55 10.5 2.14 6.69 0.92 5.92 0.89 5.51 0.57 0.34 0.13

GR 181c Triassic MLF

95.3 4.18 7.49 23.3 4.07 21.8 7.49 2.76 8.93 1.71 11.1 2.22 6.97 0.99 6.55 0.97 5.81 0.97 0.35 0.16

60.3 57 90 46 355 265 8 366 55 2 3 108 34 2 n.a. 41.9 1.65 3.45 10.6 1.88 10.1 3.55 1.34 4.16 0.80 5.33 1.08 3.27 0.45 2.91 0.41 2.72 0.32 0.15 0.53

GR 181d Triassic Pillow

59.8 50 85 45 352 245 4 76 42 n.d. 3 101 30 4 23

GR 195a Triassic Pillow

51.8 111 63 45 103 437 4 102 45 3 6 149 48 n.a. 49 46.9 4.35 7.20 19.9 3.28 17.6 6.03 1.98 7.41 1.39 9.09 1.93 6.03 0.84 5.37 0.82 4.70 0.43 0.50 0.52

Sample locations are shown in Figures 2 and 4. Fe2O3/FeO = 0.15; Mg no. = 100Mg/(Mg + Fe2+), where Mg is MgO/40 and Fe is FeO/72. Pill. Brec., pillow breccia; MLF, massive lava flow; Lad, Ladinian; Nor, Norian; Car, Carnian; n.a., not analysed; n.d., not detected.

Petrography The volcanic rocks studied in this paper have undergone low-grade greenschist-facies oceanic metamorphism; however, the prefix 'meta' is hereafter omitted for simplicity. Primary minerals are variably altered in most of the studied samples, although primary igneous textures are well preserved. Mineral assemblages include chloritized clinopyroxene + albitized plagioclase + chlorite ± clay minerals ± epidote ± carbonate. In some samples, secondary carbonates also occur both in filled fractures and disseminated in the groundmass. A few samples display vesicles or amygdules filled by chlorite or zeolites. Pillowed flows and pillow breccias are predominantly aphyric with hyalophitic and hyalopilitic textures. Plagioclase microlites and granular clinopyroxenes are commonly recognized in chloritized or recrystallized vitric groundmass. A few pillow breccias have trace amounts (<5%) of phenocrysts, including plagioclase and clinopyroxene, as well as chlorite, probably replacing olivine.

Massive flows are mainly aphyric, intergranular to sub-ophitic, with euhedral plagioclase (up to 1 mm in size) and anhedral to interstitial clinopyroxene. In summary, from these observations it appears that the magmatic crystallization order for all samples is: olivine —» plagioclase —> clinopyroxene ± Fe-Ti-oxides; that is, the typical MORE crystallization sequence.

Geochemical results All samples are basaltic in composition (Table 1), with SiO2 ranging from 41.49 to 50.55 wt%, although the lowest values correlate closely with high loss on ignition. In the analysed basalts, CaO, K2O, Rb, Sr, Ba and, to a lesser extent, Na2O contents display considerable scatter, clearly related to secondary processes. In particular, CaO is highly variable, ranging from 8.69 to 18.36 wt%, and is generally higher than would be expected for fresh basaltic rocks of this type.

118

E. SACCANI ETAL.

Moreover, the higher values clearly correlate with the abundance of secondary carbonates. Although many samples are variably altered, many incompatible elements (e.g. Th, U, Ta, Nb, REE, Hf, Zr, Ti, Y) and transition metals (e.g. Ni, Cr, V) can be effectively used to describe the primary geochemical features of the studied volcanic rocks, as their concentrations are relatively immobile during metamorphic processes (Beccaluva et al. 1979; Pearce & Norry 1979). Pillow lavas and pillow breccias from the volcanic sections GR47, GR50 and GR51 are characterized by rather uniform composition (Table 1), and by slightly decreasing patterns, from Nb to Yb, of incompatible element abundance with respect to normal MORE (N-MORB) (Fig. 5a). As can be seen in Figure 5b, these rocks have REE abundance from 12 to 25 times that of chondrite composition (Sun & McDonough 1989), and display light REE (LREE) enrichments similar to those observed in transitional MORB (TMORB). This feature is exemplified by the LaN/ SmN and LaN/YbN ratios, which are 1.18-1.21 and 1.64-1.77, respectively. Pillowed and massive lavas from section GR56 display highly uniform composition for both major and trace elements (Table 1). These rocks are characterized by rather flat incompatible element patterns in comparison with N-MORB (Fig. 5c), and by a very slight LREE depletion with respect to middle REE (MREE; Fig. 5d), with LaN/SmN ratios ranging from 0.78 to 0.99. From Figure 5d it can also be observed that the overall REE patterns appear to be intermediate between typical N- and T-MORB types. High field strength elements (HFSE) range from one to two times NMORB abundance. Pillow lavas from section GR71 can be distinguished from the other studied samples by their high TiO2 content (2.36-2.89 wt%) and relatively low Mg number (49.8-55.7), as can be seen from Table 1 and Figure 6. They are characterized by HFSE abundance ranging from two to four times N-MORB composition (Fig. 5e) and variably depleted LREE patterns (Fig. 5f), with LaN/SmN ratios ranging from 0.65 to 0.90, and LaN/YbN ratios ranging from 0.87 to 1.10. Although their REE abundance is high (20-50 times chondrite composition), the overall REE patterns are similar to those displayed by N-MORBs. Pillowed and massive lava flows from sections GR181 and GR195 are characterized by HFSE patterns similar to those displayed by N-MORB, with HFSE abundance varying from one to three times N-MORB composition. These samples also show marked Th and Nb negative anomalies (Fig. 5g), although the concentrations of these elements are comparable with those of N-MORBs. How-

ever, the most striking geochemical feature exhibited by these basalts is a marked LREE depletion with respect to MREE (Fig. 5h), as testified by the LaN/SmN ratios ranging from 0.63 to 0.77, and by the LaN/YbN ratios ranging from 0.82 to 0.96. The pillow basalt from volcanic section GR195 displays high TiO2 (2.43 wt%) and V (437 ppm) contents, and other geochemical features (Table 1) similar to those observed in basalts from volcanic section GR71. Basaltic rocks from section GR181 are characterized by relatively low TiO2 (1.30-1.50 wt%) and V (245-280 ppm) contents, coupled with high Cr (346-381 ppm) abundance.

Tectonomagmatic interpretation and petrogenesis of basalts Tectonomagmatic interpretation The geochemistry of mafic extrusive rocks has been used by many researchers to infer the tectonic setting of eruption (Beccaluva et al. 1979; Pearce & Norry 1979; Shervais 1982). Analogously, we use the geochemical characteristics of the Triassic basalts from the Argolis Middle Unit to deduce the tectonic setting in which they formed. A variety of geochemical and petrographical indicators suggest that the analysed rocks derived from melts generated in a midocean ridge tectonic setting(s). The discrimination diagrams shown in Figure 6 (Shervais 1982) and Figure 7 (Pearce & Norry 1979) support this interpretation. On the basis of trace element composition (Fig. 5), Triassic basalts from the Argolis Middle Unit can essentially be subdivided into two compositional groups: (1) enriched (transitional) MORBs (T-MORB); (2) normal MORBs (N-MORB). The first group is found in volcanic sections GR47, GR50 and GR51 (Figs 2 and 3), and is characterized by Ta/Yb ratios >0.30 and Nb content >6.5 ppm, whereas the second group is found in sections GR56, GR71, GR181 and GR195 (Figs 2 and 3) and displays Ta/Yb ratios <0.22 and Nb content <6 ppm. This interpretation is supported by the discrimination diagram of Figure 8 (Meschede 1986), where basalts from sections GR47, GR50 and GR51 plot in the field for T-MORB compositions, and basalts from sections GR56, GR71, GR181 and GR195 plot in the N-MORB field. The T-MORBs have a more enriched composition, with higher values of Th, U, Ta, Nb and LREE with respect to the N-MORB samples. The latter, like present-day N-MORBs (Pearce 1983), are characterized by variable Th negative anoma-

TRIASSIC BASALTS OF PINDOS BASIN, GREECE

119

Fig. 5. N-MORB-normalized incompatible element abundance patterns and chondrite-normalized REE abundance patterns for the Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula. Normalization values from Sun & McDonough (1989).

lies and variably depleted LREE patterns (Fig. 4c-h). Petrogenesis Analyses of basalts from within individual volcanic sections display very limited chemical

variations. Although single successions are chemically broadly similar to each other, different chemistries between individual sequences are observed (Table 1), suggesting that there is no geochemical correlation between the volcanic sequences represented in different thrust sheets.

120

E. SACCANI ETAL.

Fig. 6. V v. Ti/1000 discrimination diagram for the Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula. Modified after Shervais (1982). Ti/V < 20 field for convergent plate margin basalts, 20 < Ti/V < 50 field for MORBs, Ti/V > 50 field for alkaline within-plate basalts.

Fig. 7. Zr/Y v. Zr discrimination diagram for the Triassic basaltic rocks from the Middle Unit of the central—northern Argolis Peninsula. Modified after Pearce & Norry (1979). V-A B, volcanic-arc basalts; MORB, mid-ocean ridge basalts; WPB, within-plate basalts.

Ratios of highly incompatible trace elements, such as Ce/Y, Zr/Nb and La/Yb, are generally little influenced by small extents of fractional crystallization and are expected to reflect either

Fig. 8. Zr/4-Y-2 X Nb discrimination diagram for the Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula. Modified after Meschede (1986). Al, within-plate alkali basalts; A2, within-plate alkali and tholeiitic basalts; B, within-plate tholeiitic and calc-alkaline basalts; Cl, transitional midocean ridge basalts (T-MORB); C2, normal mid-ocean ridge basalts (N-MORB).

the source characteristics or the degree of partial melting (Saunders et al. 1988). The restricted range of incompatible trace element abundance (Table 1, Fig. 5) indicates that the studied samples originated from similar sources; none the less, the Ce/Y, Zr/Nb and LaN/YbN ratio covariations plotted in Figure 9 reveal some differences between samples from single volcanic sections. Ce, Nb and La are more incompatible than Y, Zr and Yb, respectively. Consequently, the rocks representing the smallest degree of partial melting or derived from more enriched sources exhibit the lowest Ce/Y ratios and the highest Zr/Nb and LaN/ YbN ratios. In Figure 9 it can be observed that basalts from sections GR47, GR50 and GR51 display elemental ratios very similar to those observed in modern primitive T-MORB (Sun & McDonough 1989), and possibly represent melts derived from more enriched sources or, alternatively, from higher degrees of partial melting when compared with those of rocks from other volcanic sections. Analogous basalts from Dinaride-Hellenide ophiolitic melanges have been interpreted as originating from depleted mantle sources variably metasomatized by OIB-type components (Pe-Piper 1998, and references therein). A comparable case of interaction between depleted mantle and OIBtype material has been observed in present-day MORBs from the Easter microplate. In fact, available isotopic and incompatible element data

TRIASSIC BASALTS OF PINDOS BASIN, GREECE

Fig. 9. (La/Yb)N v. Ce/Y (a) and Zr/Nb (b) diagrams for Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula. Compositions of modern T-MORB and N-MORB from Sun & McDonough (1989). Variations of elemental ratios of TMORBs from the East Rift (EMER) and N-MORBs from the West Rift (EMWR) of the Easter microplate (Haase 2002) are reported for comparison.

(Haase 2002) indicate that T-MORBs erupted in the East Rift of the Easter microplate are generated from depleted mantle containing variable volumes of the Easter plume materials. A similar genesis can also be postulated for the T-MORBs studied in this paper. The close similarity between the (La/Yb)N, Ce/Y and Zr/Nb ratios of T-MORBs from the Easter microplate East Rift and Argolis Middle Unit further supports this conclusion (Fig. 9). The elemental ratios plotted in Figure 9 for basalts from sections GR56, GR181 and GR195 are similar to those of modern primitive N-MORB (Sun & McDonough 1989). However, one sample from section GR181 (GRlSlb) and all samples from section GR71 show high Zr/Nb ratios. These samples may represent fairly evolved rocks, as testified by their high REE abundance (Fig. 4f and h) and high FeOt contents (Table 1). Therefore, the high Zr/Nb ratios observed in these rocks may have been relatively influenced by Zr enrichment

121

during fractional crystallization. However, the elemental ratios are generally compatible with a genesis from primary magmas originating from depleted N-MORB type sub-oceanic mantle sources, with no or very little influence of enriched OIB-type material. This conclusion is also supported by the close similarities between the (La/Yb)N, Ce/Y and Zr/Nb ratios of N-MORBs studied in this paper and present-day N-MORBs from the West Rift of the Easter microplate (Fig. 9). In fact, both isotopic and incompatible element abundances in the Easter West Rift basalts (Haase 2002) indicate that, in spite of the proximity of the Easter plume to the West Rift spreading ridge, the erupted basalts bear no influences of any plume material. LaN/SmN ratios (Fig. 4c, e, f and h) indicate that the N-MORBs from the different volcanic sequences are variably depleted in LREE. Consequently, they probably represent melts derived from compositionally distinct mantle sources. An estimation of the composition of primary magmas and relative mantle sources can be obtained using hygromagmatophile element ratios. These elements are weakly fractionated during fractional crystallization and, therefore, in a diagram where ratio v. ratio of hygromagmatophile elements are plotted, the population of samples originating from chemically different mantle sources will plot along distinct correlation lines representing the elemental ratios in the source (Allegre & Minster 1978). The Th/Ta v. Th/Tb ratios for the analysed basalts are plotted in Figure 10, where three distinct groups of samples can be recognized. The first

Fig. 10. Th/Ta v. Th/Tb diagrams for Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula.

122

E. SACCANI ETAL.

group includes T-MORBs from sections GR47, GR50 and GR51; the second group is represented by N-MORBs from section GR56; the third group includes N-MORBs from sections GR71, GR181 and GR195. In particular, the comparison of Th/ Ta v. Th/Tb ratios for N-MORBs indicates that these basalts derived from slightly heterogeneous mantle sources. Because the data on basalts from each volcanic sequence are limited, it is not possible to evaluate their fractional crystallization in detail. Nevertheless, very limited extents of fractionation can usually be observed in each sequence. The main influence of plagioclase (and olivine) fractional crystallization in both N- and T-MORB sequences is evidenced in the Ti v. Zr diagram (Fig. 11); this conclusion conforms to the petrographic observations. The variation of Ni compared with that of incompatible elements (e.g. Zr, Y, Th) shows rather low contents of Ni (56-104ppm) in the whole set of samples (Table 1), indicating that these basalts had already fractionated a significant amount of olivine before their extrusion. The variation of Cr v. incompatible elements (Table 1) shows much more scatter than that observed for Ni. However, the general trend of Cr variation indicates that the different basaltic sequences (generated from compositionally distinct mantle sources) have fractionated variable amounts of Cr-bearing phases, such as Cr-spinel and, possibly, moderate amounts of clinopyroxene.

Fig. 11. Ti v. Zr diagram for Triassic basaltic rocks from the Middle Unit of the central-northern Argolis Peninsula. Fractional crystallization trends for mineral phases are shown, opx, orthopyroxene; plag, plagioclase; ol, olivine; cpx, clinopyroxene; mgt, magnetite.

Discussion Triassic magmatism in the Dinaride-Hellenide orogenic belt Triassic magmatic rocks are widespread throughout the Dinaride-Hellenide orogenic belts, from former Yugoslavia to Greece. They are mainly represented by: (1) subalkaline basalt, andesite, dacite and rhyolite series, possibly originating in a back-arc setting (Pe-Piper 1982, 1998); (2) high-K shoshonites associated with calc-alkaline andesites, originating in an extensional setting from a mantle source bearing subduction-derived components (Bebien et al 1978; Pe-Piper & Mavronichi 1990; Pe-Piper 1998); (3) alkali basalts, commonly referred to either oceanic island or rift settings (Pe-Piper 1983; Jones & Robertson 1991). Very rare Triassic MORBs have also been recognized in NW Croatia (Halamic et al. 1998), in the Hellenide Subpelagonian Zone (Avdella Melange: Jones & Robertson 1991; KerassiesMilia ophiolitic belt: Pe-Piper & Hatzipanagiotou 1993), in the island of Samos (Pe-Piper & Kotopouli 1991), and in central Evia island (Danelian & Robertson 2001). All these rocks are somewhat enriched in large ion lithophile elements (LILE) and display no depletion in LREE, typical of NMORBs. In general, their overall geochemical characteristics are similar to those of T-MORBs rather than N-MORBs, suggesting variable chemical contributions from an OIB-type mantle source (Pe-Piper 1998). A chemical contribution from both depleted mantle and OIB-type components is consistent with Nd and Pb isotopic values, which are intermediate between depleted asthenosphere and HIMU OIB (Pe-Piper 1998). Another peculiar geochemical feature of many of these basalts is a Th enrichment coupled with Nb negative anomaly, interpreted by Pe-Piper (1998) as a geochemical feature of the subcontinental lithospheric mantle inherited from a former Permian subduction. These basalts have been related by many workers to the initial stages of oceanic spreading (Jones & Robertson 1991; Robertson & Karamata 1994; PePiper 1998). The Triassic ages suggested for these MORBs are, in most cases, poorly constrained. For example, the radiolarian cherts overlying the basaltic sequence in the Kerassies area have not been dated, but have been attributed to the Triassic by comparison with radiolarites from Othris (Beck 1980; Pe-Piper & Hatzipanagiotou 1993). In Samos, Pe-Piper & Kotopouli (1991) and Pe-Piper (1998) have demonstrated that MORB-like rocks occur as tectonic intercalations, and are possibly not Triassic in age. In other cases, the stratigraphic relationships between basalts and Triassic

TRIASSIC BASALTS OF PINDOS BASIN, GREECE radiolarian cherts are either presumed (Danelian & Robertson 2001) or unclear (Jones & Robertson 1991; Pe-Piper & Kotopouli 1991). On the other hand, it should be taken into account that a number of available ages for Dinaride-Hellenide ophiolites and related rocks indicate that both MOR and SSZ ophiolites were generated during the Mid-Jurassic (Spray et al. 1984; Marcucci & Prela 1996; Bebien et al 2000).

Geodynamic implications of Triassic MORBs from the central-northern Argolis Middle Unit Unlike many cases previously reported, the Triassic basalt-chert sequences described in this paper represent coherent successions, and, although intrusive rocks are lacking, the chemistry of the basalts leaves no doubt as to their ophiolitic nature. On the basis of the major oxides and trace element compositions presented in this paper, two distinct types of Triassic MORBs can be recognized in the Middle Unit of central-northern Argolis: T-MORBs and N-MORBs. This finding has important implications for the evolution of the Hellenide sector of the Pindos oceanic basin. Triassic MORBs from the Middle Unit of the central-northern Argolis peninsula represent the oldest unequivocally dated oceanic crust in the Hellenide sector of the Pindos basin. Moreover, the occurrence of Triassic N-MORBs in the Argolis Middle Unit testifies that, at least in some sectors of the Pindos basin, the oceanization was already fully developed in the Mid?- Late Triassic. On the other hand, Triassic T-MORBs, already found as clasts and blocks in the DinarideHellenide melanges, have commonly been attributed to the incipient stages of formation of the Pindos ocean (Bebien et al. 1978; Robertson & Karamata 1994). However, T-MORBs have been documented in some fully developed oceanic basins, and their occurrence is commonly related to the interaction between N-MORB mantle sources and near-axis mantle plumes (Haase 2002, and references therein). In the light of the new data presented herein, a model for the evolution of the Pindos oceanic basin in the south Hellenide belt can be summarized as follows (Fig. 12). From the Early Triassic, extensional tectonics caused rifting of the continental lithosphere between the future Apulian and Pelagonian microplates. During this stage the magmatic activity was characterized by the development of shoshonitic and calc-alkaline series (Bebien et al. 1978; Pe-Piper 1998). Available data suggest that the

123

subduction-related geochemical characteristics of these series were inherited from the Hercynian subduction below Gondwana, which chemically modified the subcontinental lithospheric mantle through hydration processes and metasomatic enrichment in LILE, LREE and Th (Pe-Piper 1998, and references therein). Contemporaneously, the eruption of chemically enriched alkaline basalts associated with the rifting (Fig. 12a) indicates that an OIB-type mantle source was active since the Early?- Mid-Triassic (Jones & Robertson 1991; Pe-Piper 1998). The continuous extension between Apulia and Pelagonia microplates resulted in the generation of the early oceanic crust of the Pindos ocean, starting from the Mid-Triassic. The interaction between the uprising asthenosphere and the OIBtype mantle source resulted in the production of enriched MORBs, similar to modern T-MORBs (Fig. 12b). At this stage, enriched alkaline basalts may have been erupted either onto the passive continental realms or in oceanic islands (Pe-Piper 1998). In addition, N-MORBs with primitive asthenospheric geochemical characteristics were produced starting from the Mid?- Late Triassic (Fig. 12c). This implies that oceanic spreading had already reached a quasi-steady state, involving only suboceanic mantle sources and their partial melt derivatives. At the same time, off-axis eruption of both T-MORBs and enriched alkaline basalts was still occurring.

Modern oceanic analogue The data presented in this paper, along with those of the literature, suggest that three types of magmatic activity (i.e. enriched alkaline basalts, T-MORBs and N-MORBs) were contemporaneously present in the Pindos basin during the Mid-Late Triassic. This interpretation implies the spatial and temporal association of three types of possible mantle sources (i.e. OIB, intermediate OIB-depleted asthenospheric mantle and depleted asthenospheric mantle) very close to one another. Similar associations of N-MORBs and oceanic island-type alkaline basalts have also been described in the Turkish sector of the Neo-Tethys (Tankutef 0/. 1998). A modern analogue for this tectonomagmatic setting is represented by the embryonic ocean of the Red Sea. Available data indicate that the southern part of the Red Sea is characterized by ocean-floor spreading with the development of an axial rift valley, in which both T-MORBs and NMORBs are extruded (Eissen et al. 1989). By contrast, the northern Red Sea consists of thinned continental lithosphere and is characterized by the

124

E. SACCANI ETAL.

Fig. 12. Schematic evolution of the early oceanization phases of the Pindos basin during the Triassic. (See text for explanation.)

occurrence of rift-related alkaline and transitional basalts (Maury et al 1985). Similar 'along-strike' chemical variations of the erupted basalts can be postulated also for the Pindos basin during the Triassic phases of continental rifting, break-up and oceanic opening. Similar to the modern Red Sea, the opening of the Pindos rift-ocean system probably occurred diachronously from south to north. In fact, the predominance of within-plate basalts in the melanges along the Dinaride-Hellenide belt suggests that rifting processes prevailed in many areas during the Triassic. In addition, the occurrence of Triassic MORBs is rarely recorded: most of them are represented by enriched-type MORBs commonly referred to the very early stages of spreading (Bebien et al. 1978; Jones & Robertson 1991; Pe-Piper 1998), whereas true Triassic N-MORBs, originating from a depleted MOR-type mantle source, are found only in the southern Hellenides.

These observations argue in favour of a development during the Triassic of well-established spreading centres in the southern portion of the Pindos basin, whereas in the other parts full oceanization was attained only in the Jurassic.

Concluding remarks From the results discussed above, the following conclusions can be drawn. (1) The Middle Unit of the central-northern Argolis Peninsula is composed of multiple thrust sheets and does not represent a distinct ophiolitic unit, as previously reported in the literature (Baumgartner 1985). Nevertheless, coherent ophiolitic sequences including mafic volcanic rocks topped by radiolarian cherts are preserved in many tectonic slices. These sequences display both Jurassic (Baumgartner 1985; Dostal et al.

TRIASSIC BASALTS OF PINDOS BASIN, GREECE

125

BEBIEN, J., BLANCHET, R., CADET, J.-P., CHARVET, J., CHORDwicz, J., LAPIERRE, H. & RAMPNOUX, J.-P. 1978. Le volcanisme Triasique des Dinarides en Yougoslavie: sa place dans 1'evolution geotectonique Peri-Mediterraneenne. Tectonophysics, 47, 159-176. BEBIEN, J., DIMO-LAHTTTE, A., VERGELY, P., INSERGUEIX-FILIPPI, D. & DUPEYRAT, L. 2000. Albanian ophiolites. I—magmatic and metamorphic processes associated with the initiation of a subduction. Ofioliti, 25, 47-53. BECCALUVA, L., OHNENSTETTER, D. & OHNENSTETTER, M. 1979. Geochemical discrimination between ocean-floor and island-arc tholeiites—application to some ophiolites. Canadian Journal of Earth Sciences, 16, 1874. BECK, C.M. 1980. Essai d'interpretation structurale et paleogeographique des 'roches vertes du Pinde d'Etoilie' (Grece continental meridionale). Annales de la Societe Geologique du Nord, 99, 355-365. BORTOLOTTI, V., CHIARI, M., MARCUCCI, M., PHOTIADES, A. & PRINCIPI, G. 2001. Triassic radiolarian assemblages from the cherts associated with pillow lavas in the Argolis Peninsula. Ofioliti, 26, 75. BORTOLOTTI, V., CARRAS, N., CHIARI, M., FAZZUOLI, M., MARCUCCI, M., PHOTIADES, A. & PRINCIPI, G. 2002. New geological observations and biostratigraphic data on the Argolis Peninsula: palaeogeographic and geodynamic implications. Ofioliti, 27, 43. CAPEDRI, S., GRANDI, R., PHOTIADES, A. & TOSCANI, L. 1996. 'Boninitic' clasts from the Mesozoic olistostromes and turbidites of Angelokastron (Argolis, Greece). GeologicalJournal, 31, 301-322. CLIFT, P.O. & ROBERTSON, A.H.F. 1989. Evidence of a late Mesozoic ocean basin and subduction-accretion in the southern Greek Neo-Tethys. Geology, 17, 559-563. CLIFT, P.O. & ROBERTSON, A.H.F. 1990. Deep-water This research project was funded by MURST-COFIN basins within the Mesozoic carbonate platform of (project leader L. Beccaluva). Argolis, Greece. Journal of the Geological Society, London, 147, 825-836. DANELIAN, T. & ROBERTSON, A.H.F. 2001. Neotethyan References evolution of eastern Greece (Pangodas Melange, Evia Island) inferred from radiolarian biostratigraALLEGRE, C.J. & MINSTER, J.F. 1978. Quantitative phy and the geochemistry of associated extrusive models of trace element behavior in magmatic rocks. Geological Magazine, 138, 345-363. processes. Earth and Planetary Science Letters, 38, DEGNAN, P.J. & ROBERTSON, A.H.F. 1998. Mesozoic1-25. early Tertiary passive margin evolution of the AUBOUIN, J., BONNEAU, M. & CELET, P. ET AL. 1970. Pindos ocean (NW Peloponnese, Greece). SedimenContribution a la geologic des Hellenides: le tary Geology, 111, 33-70. Gavrovo, le Pinde et la zone ophiolitique subpelaDOSTAL, J., TOSCANI, L., PHOTIADES, A. & CAPEDRI, S. gonienne. Annales de la Societe Geologique du 1991. Geochemistry and petrogenesis of Tethyan Nord, 90, 277-306. ophiolites from northern Argolis (Peloponnesus, BAUMGARTNER, P.O. 1980. Late Jurassic Hagiastridae Greece). European Journal of Mineralogy, 3, 105and Patulibrachiidae (Radiolaria) from Argolis Pe121. ninsula (Peloponnesus Greece). Micropaleontology, DOUTSOS, T., PE-PIPER, G., BORONKAY, K. & KOUKOU23, 274-322. VELAS, I. 1993. Kinematics of the central HelleBAUMGARTNER, P.O. 1985. Jurassic sedimentary evolunides. Tectonics, 12(4), 936-953. tion and nappe emplacement in the Argolis PeninsuElSSEN, J.-P., JUTEAU, T., JORON, J.-L., DUPRE, B., la (Peloponnesus, Greece). Memoires de la Societe HUMLER, E. & AL'MUKHAMEDOV, A. 1989. PetrolHelvetique des Sciences Naturelles, 99, 1-111. ogy and geochemistry of basalts from the Red Sea BAUMGARTNER, P.O. 1987. Age and genesis of Tethyan axial rift at 18° North. Journal of Petrology, 30, Jurassic Radiolarites. Eclogae Geologicae Helve791-839. to, 80, 831-879.

1991) and Triassic (Bortolotti et al 2001, 2002) ages, as inferred from radiolarian assemblages. (2) Petrographical and geochemical analyses of Mid-Late Triassic basalts show that these volcanic rocks were generated in a mid-ocean ridge setting. None the less, two chemically distinct groups of lavas can be recognized: T-MORBs and N-MORBs. (3) T-MORBs were generated from partial melting of a primitive mantle source variably enriched by an OIB-type component. By contrast, NMORBs originated from partial melting of suboceanic depleted mantle sources only. Highly incompatible elements suggest that the various NMORB sequences were generated from chemically heterogeneous sources. (4) The Triassic MORBs in the Middle Unit of central-northern Argolis represent the oldest remnants of the Pindos oceanic crust unambiguously dated in the Dinaride-Hellenide belt. (5) The data presented in this paper suggest that the oceanic spreading of the Pindos basin had locally reached a quasi-steady state during the Mid?- Late Triassic. (6) The contemporaneous occurrence during the Mid?- Late Triassic of both T- and N-MORBs, as well as alkaline basalts, can be explained by both 'along-strike' and off-axis chemical variations in melt sources in the mantle beneath the spreading axes within the Pindos basin. (7) Similar to the present-day Red Sea, the development of the Pindos rift-ocean system probably occurred diachronously from south to north.

126

E. SACCANI ETAL.

FERRIERE, J. 1982. Paleogeographie et tectonique superposes dans les Hellenides internes: les massifs de I'Othrys et du Pelion (Grece continentale). Societe de la Geologic du Nord, Lille, Publication, 8, 1-941. FRANZINI, M., LEONI, L. & SAITTA, M. 1972. A simple method to valutate the matrix effect in X-ray fluorescence analysis. X-Ray Spectrometry, 1, 151-154. HAASE, K.M. 2002. Geochemical constraits on magma sources and mixing processes in Easter Microplate MORE (SE Pacific): a case study of plume-ridge interaction. Chemical Geology, 182, 335-355. HALAMIC, J., SLOVENEC, D. & KOLAR-JURKOVSEK, T. 1998. Triassic pelagic limestones in pillow lavas in the Oresje quarry near Gornja Bistra, Medvednica Mt. (Northwest Croatia). Geologica Croatica, 51, 33-45. JONES, G. & ROBERTSON, A.H.F. 1991. Tectono-stratigraphic evolution of the Mesozoic Pindos ophiolite and related units, northwestern Greece. Journal of the Geological Society, London, 148, 267-288. MARCUCCI, M. & PRELA, M. 1996. The Lumi i zi (Puke) section of the Kalur Cherts: radiolarian assemblages and comparison with other sections in northern Albania. Ofioliti, 21, 71-76. MAURY, R.C., BOUGAULT, H., COUTELLE, A., GUENNOC, P., JORON, J.-L. & PAUTOT, G. 1985. Presence de ferrobasalte tholeiitique dans la fosse Jean Charcot (26°15'N, 35°22'E): signification dans le contexte geodynamique de la Mer Rouge. Comptes Rendus de I 'Academic des Sciences, 300, 811-816. MESCHEDE, M. 1986. A method of discriminating between different types of mid-ocean ridge basalts and continental tholeiites with the Nb-Zr-Y diagram. Chemical Geology, 56, 207-218. PEARCE, J.A. 1983. Role of the sub-continenetal lithosphere in magma genesis at active continental margin. In: HAWKESWORTH, C.J. & NORRY, M.J. (eds) Continental Basalts and Mantle Xenoliths. Shiva, Nantwich, 230-249. PEARCE, J.A. & NORRY, M.J. 1979. Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology, 69, 33-47. PE-PIPER, G. 1982. Geochemistry, tectonic setting and metamorphism of the mid-Triassic volcanic rocks of Greece. Tectonophysics, 85, 253-272. PE-PIPER, G. 1983. The Triassic volcanic rocks of Tyros, Zarouhla, Kalamae, and Epidavros, Peloponnese, Greece. Schweizerische Mineralogische und Petrographische Mitteilungen, 63, 249-266. PE-PIPER, G. 1998. The nature of Triassic extensionrelated magmatism in Greece: evidence from Nd and Pb isotope geochemistry. Geological Magazine, 135(3), 331-348. PE-PIPER, G. & HATZIPANAGIOTOU, K. 1993. Ophiolitic rocks of the Kerassies-Milia belt, continental Greece. Ofioliti, 18, 157-169. PE-PIPER, G. & KOTOPOULI, C.N. 1991. Geochemical characteristics of the Triassic igneous rocks of the Island of Samos, Greece. Neues Jahrbruch fur Mineralogie, Abhandlungen, 162, 135-150.

PE-PIPER, G. & MAVRONICHI, M. 1990. Petrology, geochemistry and regional significance of the Triassic volcanic rocks of the Western Parnassos isopic zone of Greece. Ofioliti, 15(2), 269-285. PHOTIADES, A. 1986. Contribution a I'etude geologique et metallogenique des unites ophiolitiques de I'Argolide septentrionale (Grece). PhD thesis, Universite de Besancon. PHOTIADES, A. 1989. The diversity of the Jurassic volcanism in the inner parts of the Hellenides: the northern Argolis ophiolitic unit (Peloponnesus, Greece). Bulletin of the Geological Society of Greece, 23, 515-530. PHOTIADES, A. & ECONOMOU, G. 1991. Alteration hydrothermale sous-marine des basalts et des dolerites (facies zeolitique) de 1'Unite moyenne 'volcanique' de 1'Argolide Septentrionale (Peloponnese, Grece. Bulletin of the Geological Society of Greece, 25,301-319. PHOTIADES, A. & SKOURTSIS-CORONEOU, V. 1994. Stratigraphic and paleogeographic evolution of the Northern Argolis (Greece) during the CretaceousPaleogene. Bulletin of the Geological Society of Greece,^, 135-146. ROBERTSON, A.H.F. 1994. Role of the tectonic facies concept in erogenic analysis and its application to Tethys in the Eastern Mediterranean region. EarthScience Reviews, 37, 139-213. ROBERTSON, A.H.F. & KARAMATA, S. 1994. The role of subduction accretion processes in the tectonic evolution of the Mesozoic Thetys in Serbia. Tectonophysics, 234(1-2), 73-94. ROBERTSON, A.H.F. & SHALLO, M. 2000. MesozoicTertiary evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-254. ROBERTSON, A.H.F., VARNAVAS, S. & PANAGOS, A. 1987. Ocean ridge origin and tectonic setting of Mesozoic sulphide and oxide deposits of the Argolis Peninsula of the Peloponnesus, Greece. Sedimentary Geology, 53, 1-32. ROBERTSON, A.H.F., CLIFT, P.D., DEGNAN, P.J. & JONES, G. 1991. Palaeogeographic and palaeotectonic evolution of the Eastern Mediterranean Neotethys. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 289-343. ROBERTSON, A.H.F., DIXON, J.E. & BROWN, S. ET AL. 1996. Alternative tectonic models for the Late Paleozoic-Early Tertiary development of Tethys in the Earstern Mediterranean region. In: MORRIS, A. & TARLING, D.H. (eds) Paleomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Pubblications, 105, 239-263. SAUNDERS, A.D., NORRY, M.J. & TARNEY, J. 1988. Origin of MORB and chemically-depleted mantle reservoirs: trace element constraints. Journal of Petrology, Special Issue, Oceanic and Continental Lithosphere: Similarities and Differences, 414-445. SHERVAIS, J.W. 1982. Ti-V plots and the petrogenesis of modern ophiolitic lavas. Earth and Planetary Science Letters, 59, 101-118. SPRAY, J.G., BEBIEN, J., REX, D.C. & RODDICK, J.C. 1984. Age constraints on the igneous and meta-

TRIASSIC BASALTS OF PINDOS BASIN, GREECE morphic evolution of the Hellenic-Dinaric ophiolites. In: DIXON, I.E. & ROBERTSON, A.H.F. (eds) The Geological Evolution of Eastern Mediterranean. Geological Society, London, Special Publications, 17, 619-627. SUN, S.-S. & McDoNOUGH, W.F. 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: SAUN-

127

DERS, A.D. & NORRY, MJ. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 313-345. TANKUT, A., DILEK, Y. & ONEN, P. 1998. Petrology and geochemistry of the Neo-Tethyan volcanism as revealed in the Ankara melange, Turkey. Journal of Volcanology and Geothermal Research, 85, 265-284.

This page intentionally left blank

Structural and microstructural analysis of a palaeo-transform fault zone in the Neyriz ophiolite, Iran KHALIL SARKARINEJAD Department of Earth Sciences, College of Sciences, University ofShiraz, Shiraz, Iran (e-mail: Sarkarinejad@geology. susc. ac. ir, sarkarinejad@yahoo. com) Abstract: The Neyriz ophiolite in SW Iran includes a NNW-SSE-trending, steeply dipping oceanic palaeo-transform fault zone that consists mainly of deformed and sheared gabbros and peridotites. Mylonitic rocks within the fossil palaeo-transform fault display C-, C'- and S-band shear cleavages containing green hornblende- and feldspar-rich bands. Deformed hornblendemantled porphyroclasts have symmetrical tails or cr-type porphyroclast systems, which show clockwise stair-stepping rotation and oblique foliation with asymmetrical complex stripes and wings. These microstructures indicate dextral shearing on the palaeo-transform fault. The caxes of hornblende in high-grade S-C mylonitic rocks exhibit strong lattice-preferred orientation (LPO) patterns with M- and G-type origin. These LPO patterns are asymmetrical with respect to the shear zone foliation and also indicate dextral shearing. The NNW-SSE trend of the palaeo-transform fault is perpendicular to the general NE strike of sheeted dykes and is parallel to the average harzburgite foliation. These observations suggest a noncoaxial shear orientation of mantle flow, which progressively rotated toward parallelism with the fossil fracture zone. This dextral transform fault is inferred to have connected ENE-trending oceanic spreading centre segments in a Neo-Tethyan ocean basin.

Oceanic transform faults and fracture zones play an important role in formation and evolution of oceanic lithosphere. However, direct observations of rocks and structures developed within these zones in modern oceanic crust are highly limited. Some ophiolites include well-preserved fossil transform faults and fracture zones, giving us access to examine in four dimensions the rock types and structures associated with the evolution of ancient oceanic lithosphere and to document the nature of related geological processes. The east-west-striking Arakapas fault zone in southern Troodos is interpreted, for example, as a fossil transform fault (Moores & Vine 1971). Simonian & Gass (1978) and Gass et al. (1994) observed that approximately north-south-oriented sheeted dykes in the Troodos ophiolite curve into an easterly attitude as they approach the Arakapas transform fault to the south, and they interpreted this phenomenon either as a primary feature caused by intrusion of the dykes along curved stress trajectories or as a later feature caused by shear against the active portion of the transform fault. The Coastal Complex associated with the Bay of Islands ophiolite in Newfoundland has also been interpreted as a palaeo-oceanic transform fault (Karson 1984). This fossil transform fault zone consists of deformed and metamorphosed gabbros, ultramafic rocks, and metasedimentary

rocks intruded by plutonic rocks and overlain by less deformed upper-crustal rocks. Another fossil oceanic fracture zone occurs in the Bogota Peninsula and NE ophiolite district of New Caledonia. This is a large-scale shear zone in which a dextral sense of movement was deduced from the following observations: (1) rotation of foliation on each side of the highly deformed zone; (2) a consistent dextral shear sense measured in rocks from the shear zone; (3) preferred direction of injection of various dykes. The Wadi Tayan massif in the Oman ophiolite contains a shear zone in the upper-mantle unit that has been interpreted as a transform fault (Boudier & Coleman 1981; Nicolas et al. 1988). The hightemperature character of deformation in this zone, the moderate slope of the cold wall, and clockwise rotation of the sheeted dyke complex all indicate that the transform motion was accompanied by significant extension (Nicolas 1989). Studies of fossil oceanic transform faults and fracture zones exposed in ophiolites are generally limited to mesoscopic structures based on field observations, as the examples above illustrate. Documentation of structures at microscopic scales from oceanic transform faults is extremely rare in the literature. Our studies of deformed and metamorphosed oceanic rocks in a fossil transform fault zone in the Neyriz ophiolite in Iran have

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 129-145. 0305-8719/037$ 15 © The Geological Society of London 2003.

130

K. SARKARINEJAD

shown that microscopic structures reveal significant information on the nature of the processes associated with transform shearing in oceanic lithosphere. The objectives of this paper are to: (1) present evidence for hornblende grain boundary sliding and dynamic recrystallization under high-grade metamorphic conditions during the formation of S-C shear-band cleavages in this fossil oceanic transform fault in the Neyriz ophiolite, and (2) present microscopic evidence of dextral shear using hornblende c-axes. The results of this structural and microstructural analysis of deformed rocks from the fossil transform fault can be used to infer the sea-floor spreading tectonics of the Neyriz ophiolite.

Alignment of these blocks as stretched, lensoidal bodies has produced a strong foliation in the sedimentary melange. Shear sense indicators from these lensoidal blocks reveal a dextral sense of shearing. The contact between the metamorphic melange and the ophiolite is marked by a low-angle thrust fault forming duplexes. The metamorphic melange includes blocks of amphibolite, kyanite schist, granulite, blueschist and eclogite. The basal metamorphic melange crops out in the Seh-Qalatoun area (Fig. 1), where it is separated from an orbitolina-bearing Cretaceous limestone by the right-lateral Mashkan fault, which is parallel to the Zagros Thrust Zone.

Regional geology The Neyriz ophiolite occurs west of the Zagros Thrust Zone, which separates the Sanandaj-Sirjan crystalline complex of the Zagros mountain range from the Zagros fold and thrust belt. The ophiolite is exposed in several large, isolated enclaves surrounded by Quaternary deposits. Approximately 2.5-5 km ENE of the Neyriz ophiolite, a melange formation crops out along the Zagros Thrust Zone. Excellent exposures of this melange are preserved in roadcuts along the Hassan Abad and Naghare-Khaneh passes (Fig. 1). The melange is thrust over the Pichkun Formation (Ricou 1968a,6), a continental slope assemblage composed of radiolarian chert, highly fossiliferous limestone-turbidite, middle Jurassic oolitic and micro-brecciated limestone, and middle Cretaceous limestone. The Neyriz ophiolite is thrust over both the Pichkun Formation and the melange, and this tectonic assemblage rests tectonically on the Cenomonian-Turonian shallow marine, shelf-platform carbonate-facies rocks of the Sarvak Formation along the northeastern margin of the AfroArabian continent (Ricou 1968a, 1968b; Stoneley 1981;Babaieef a/. 2001). The melange formation is divided into two units: a coloured sedimentary melange on top and a high-grade metamorphic melange at the bottom. The coloured melange (upper sedimentary melange unit in Fig. 1) consists of lenses, blocks and/ or ribbons of radiolarian chert, limestone, sandstone, pillow lava, tuff, serpentinite, shale and marl in a matrix made of greywacke and mudstone. These lenses and blocks range in size from a few millimetres to several hundred metres. This coloured melange unit displays a well-defined CS fabric represented by shear bands and mineral stretching foliation or lineation (Berthe et al. 1979). C-S shear-band fabrics in Hassan Abad Pass strike S71°E and those in Naghare-Khaneh strike N52°W with dips around 10-27° to the NE.

Geology of the Neyriz ophiolite The Neyriz ophiolite occupies an area c. 43 km long and 14 km wide, and has been dismembered along the Zagros Thrust Zone. It is composed, from bottom to top, of a tectonized mantle section consisting of harzburgite, Iherzolite and dunite, a transition zone including interlayered mafic-ultramafic rocks, and a crustal sequence made of isotropic gabbros, sheeted dykes, pillow lavas and radiolarian chert (Sarkarinejad 1985). Deformed and serpentinized harzburgite is the dominant ultramafic rock in the ophiolite (Fig. 2). It displays a distinct lattice-preferred orientation (LPO) of bladed orthopyroxene and olivine that gives the rock its foliated appearance. Some moderately foliated harzburgites have layers with discontinuous compositional bands of orthopyroxene-rich and orthopyroxene-poor sections. These layers are 12-20 cm thick and their modal compositions vary from O170 Opx30 to O185 Opxi5 (Sarkarinejad 1985). They are planar to slightly curved and traceable for up to 6 m in outcrop. The mean orientation of the harzburgite foliation is N50°W/54°NE (Fig. 3), and it appears to have developed as a result of plastic deformation of orthopyroxene and olivine in the mantle (Sarkarinejad 1994). Lenses and veins of deformed orthopyroxenite are also widespread in the harzburgite. Their width varies from 2 to 30 cm, and the contact between the harzburgite and orthopyroxenite veins is easily identified by a change in colour and grain size. Orthopyroxenite veins are generally coarse grained with orthopyroxene grains up to 7 cm long showing multiple kink bands. Highly serpentinized Iherzolite is also common in the mantle sequence, especially in the northeastern part of the ophiolite. The boundary between the Iherzolite and harzburgite is concordant although the precise contact is difficult to distin-

STRUCTURE OF THE N E Y R I Z OPHIOLITE, IRAN

131

Fig. 1. Geological map of the Neyriz area showing the distribution of the ophiolite, melange units and major tectonic features.

132

K. SARKARINEJAD

Fig. 2. Geological map of the Neyriz ophiolite with its oceanic palaeo-transform fault zone. The area is located 12 km NW of Neyriz. The inset map of Iran shows major tectonic elements and the location of the Neyriz ophiolite.

guish in the field because of similarities in colour and foliation. Ultramafic cumulate units composed of dunite, chromite pods, wehrlite, websterite and clinopyroxenite make up a transition zone between the residual mantle units below and the basal layered gabbros above. Clinopyroxenite is the most abundant rock type in the transition zone, which varies from 4 to 220 m in thickness in the northeastern part of the ophiolite and from 25 to 300 m in the southern part. The contact between the harzburgite tectonite and the clinopyroxenite is marked by the

gradual disappearance of the harzburgite foliation and the appearance of thick blocky clinopyroxenite. The clinopyroxenite is medium grained and has small but variable modal proportions of orthopyroxene and plagioclase. With an increase in modal per cent of these minerals, the clinopyroxenite grades upward into layered gabbro. The basal cumulate sequence is composed predominantly of anorthosite, pyroxene gabbro, gabbro norite, gabbroic troctolite and olivine gabbro. Anorthosite and anorthositic gabbros range in thickness from 2-7 m in the southern part of the

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN

133

Fig. 3. Equal-area, lower-hemisphere projection of the measured sheeted dyke orientations. •, poles of planar structures; 1, 2 and 3 dashed great circles represent the computed average Vj, ¥2 and ¥3 (eigenvector) direction. Sheeted dyke margin orientation (Sheet): 77 measurements, contours 1, 2, 4, 6 and 8%. Rotated in a clockwise sense about a horizontal axis of sheeted dyke margin orientation (RSD 3): 77 measurements, contours 1, 2, 4, 6 and 8%. Gabbro layering orientation (gb-1): 23 measurements, contours 1, 2, 4 and 6%. Gabbro layering orientation (gb-4): 20 measurements, contours 2, 4 and 6%. Harzburgite foliation (hzl): 98 measurements, contours 1, 2, 4 and 6%.

ophiolite to hundreds of metres in the main gabbro body. The basal gabbros exhibit both modal layering and a planar alignment of clinopyroxene, olivine and plagioclase crystals forming a distinc-

tive lineation in the rock. This generally NE-SWtrending mineral lineation plunges gently to the NE in the northern part of the ophiolite and about 25° to the NW in the southern part (Fig. 3).

134

K. SARKARINEJAD

The sheeted dyke complex occupies an area of c. 1 km2 and emanates from plagiogranite and high-level pegmatitic gabbros in the upper part of the plutonic complex. Elsewhere in the ophiolite sheeted dykes are in fault contact with the underlying serpentinized harzburgite and basal layered gabbros. In the central part of the ophiolite the sheeted dykes pass upward into pillow lavas, which are overlain by red radiolarian chert. In this sequence, sheeted dykes have well-developed onesided chilled margins. In general they strike N46° and dip steeply to the NE (Fig. 3). Sheeted dykes and pillow lavas are weakly altered, with chlorite replacing the mafic minerals.

Intra-ophiolitic shear zone In the southeastern part of the ophiolite volcanic rocks are juxtaposed against serpentinized peridotites along a NW-trending shear zone that has a distinctive surface expression marked by a topographic relief. The steeply north-dipping shear zone is exposed over a distance of 6.5 km before disappearing under the Quaternary deposits. It cuts the harzburgite and high-level gabbros (Fig. 2), which contain flattened and stretched hornblende and plagioclase grains that display a strong S-C fabric (Lister & Snoke 1984). The stretching lineation, defined by hornblende- and feldspar-rich bands, is commonly subhorizontal with plunges varying from 12° to 25°NW. Traces of the Cplanes have curved trajectories. The S-plane is rarely exposed on the mesoscopic scale, but there are traces of an oblique foliation around hornblende porphyroclasts. In this case, the stretched and elongated ribbons of feldspar define a fabric oblique to the C-planes. Many hornblende porphyroclasts show a stair-stepping geometry in outcrop that reveals a dextral sense of shear. In the southwestern margin of the shear zone, harzburgite is strongly sheared and mylonitized. The original foliation of the harzburgite, defined by flattened and stretched orthopyroxene, spinel, and fine-grained olivine and orthopyroxene wrapped around orthopyroxene porphyroclasts, is oblique or parallel to the C-plane shear-band cleavage. The harzburgite shear-band cleavage generally lies at a low angle to the harzburgite foliation of the SZB. The mylonitic foliation generally strikes N85°E to N30°W.

Petrography and chemistry of mylonitic rocks in the shear zone Medium- to fine-grained mylonite in the shear zone is essentially composed of green hornblende, andesine, actinolite, albite, pseudotachylyte, epi-

dote and accessory sphene. Hornblende porphyroclasts are elongated and range in size from 0.15 to 2.2 mm. They have been rotated parallel to their c-axes defining a lineation. They typically show bending and kinking with slip along their cleavage planes. Locally, the hornblende is partly to completely replaced by actinolite. The porphyroclasts are commonly enclosed by fine to ultra-fine, euhedral to anhedral hornblende neoblasts ranging in size from 1 to 50 um. Pseudotachylyte veins in the mylonite are dark brown, glassy and relatively homogeneous. They are small and curved to irregular in shape, and cut the ultra-fine-grained hornblende matrix. They are commonly parallel to the foliation plane. Inclusions in the pseudotachylyte consist of very finegrained hornblende, sphene and opaque minerals. A few porphyroclasts of plagioclase are also present. They have a subhedral to anhedral shape and range in size from 12 to 250 uni. They are enclosed in a matrix of fine to ultra-fine albite neoblasts, locally accompanied by small grains of sericite. Electron microprobe analysis of amphibole minerals in the mylonite are given in Table 1; analysed amphiboles are classified according to Leake (1978). Amphibole porphyroclasts are composed mainly of magnesio-hornblende [(Cai 96 Fe 2 + o.64Mg3.6Si7.14Al I V 0.8 4 Al V I o.37022)(OH) 2 ],

whereas matrix amphiboles are actinolite [(Cai 97 Fe2+o.9Mg3.69Si7.38AlIvo.6AlVIo.32022XOH)2] or actinolitic hornblende [(Ca2Fe2+o.94Mg3 76 817.73 Al^o 27 AlVIo.22O22) (OH)2]. Porphyroclastic magnesiohornblende typically has (Na + K) <0.5, Ti <0.5, and Mg/(Mg + Fe2+) ratios of 0.79-0.85. The Mg number is generally related to the Al content of calcic amphiboles (Grapes & Graham 1978). Another factor that plays a role in the composition of calcic hornblende is the grade of metamorphism, which affects the substitution of A1IV and A1VI. With a higher grade of metamorphism, A1IV increases whereas A1VI decreases in transition from actinolite to hornblende. Fleet & Barneth (1978) placed the boundary between low- and highpressure hornblende at 5 kbar, which corresponds to A1IV/A1VI >2.0. The A1IV/A1VI ratio of magnesio-hornblende in the mylonite is 2.27, suggesting that these hornblende grains formed at relatively high pressures. The main mineral assemblages in the mylonite are hornblende and plagioclase (andesine), suggesting amphibolite-facies conditions (temperatures around 450-650 °C and pressures about 57 kbar) during their recrystallization (Bowes, 1989). The presence of secondary epidote and actinolite in the rocks indicates lower-grade hydrothermal circulation after deformation and mylonitization.

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN

135

Table 1. Microprobe analyses and structural formulae ofamphiboles of the mylonite ML2

ML4

ML7

ML10

MK1

51.941 5.806 0.162 8.101 17.535 13.176 0.741 97.462

51.984 5.481 8.359 17.317 12.746 0.681 96.568

54.389 2.925 8.606 17.730 13.048 96.698

50.883 6.096 0.151 7.053 17.440 12.895 0.482 95.000

7.2566 0.7434 0.2125 0.0176 0.6975 0.2488 3.6513 1.9721 0.2007 15.000

7.3411 0.6589 0.2533 _ 0.5921 0.3950 3.6449 1.9284 0.1861 15.000

7.7377 0.2623 0.2281 _ 0.0344 0.9895 3.7595 1.9887 15.000

7.2814 0.7186 0.3095 0.162 0.5103 0.3336 3.7197 1.9769 0.1337 15.000

ML3

wt% 50.567 51.407 SiO2 7.755 7.139 A1203 Ti02 0.184 Fe203 FeO 8.237 8.286 MnO 0.762 17.241 MgO 17.221 CaO 12.998 13.225 6.709 Na20 0.762 K2O 97.691 Total 98.802 Number of ions on the basis of 23 oxygens 7.0409 Si 7.0934 A1IV 0.9591 0.9666 0.3134 A1VI 0.2543 _ Ti 0.0193 3+ 0.7986 Fe 0.8561 0.1605 Fe2+ 0.1000 Mn 0.0890 Mg 3.5780 3.5417 1.9389 Ca 1.9556 0.1914 Na 0.2038 K 15.000 Total 15.000

Structure of the shear zone Mesoscopic observations of the shear zone rocks indicate the presence of a well-developed S-C fabric. The C-type foliation is well displayed in mylonitic, banded gabbros. Individual bands are up to 50 cm in length and 0.5-3 mm thick (Fig. 4). Some of the thinnest bands feather out into 'horse-tail' patterns. S-type shear bands are about 12 mm wide and are oblique to the C-band foliation (White 1979; Platt & Vissers 1980). The angle between the two foliation types ranges from 35° to 60°. This type of shear-band cleavage develops mainly in strongly foliated mylonites (Passchier & Trouw 1996). It is difficult to determine the sense of shear because the transition from ultramylonite to S-C fabric takes place typically over a width of 1 mm. To study the fabric types of the mylonitic rocks, samples were oriented in the field relative to the observed foliation and lineation. In the laboratory, oriented thin-sections were cut parallel to the stretching lineation in the JfZ-plane to determine the shear sense. Thin sections were also cut normal to the stretching lineation in the 7Z-plane. In XZ thin sections, deformed hornblende porphyroclasts appear elongated with stair-stepping along cleavage planes and stretched in the direction of

the foliation. The grain aspect ratios along the cand a-axis directions range from 1.3:1 to 4:1. Typical long- and short-axis dimensions of the grains are 0.5-1.5 mm and 0.2-0.4 mm, respectively. Long axes are generally oriented oblique to the foliation. Porphyroclasts are stretched out into long narrow stripes or lines, which connect the hornblende grains. The thickness of stripes varies from 2 to 10 urn. Stretched and recrystallized stripes or line fabrics make up the tails and wings of mantled porphyroclasts commonly by forming symmetrical tails or cra-type and Ob-type porphyroclast systems (Figs 4a and c). The <7a-type porphyroclasts are surrounded by a matrix composed of fine to ultrafine grains of hornblende and plagioclase, whereas hornblende porphyroclasts of ob-type tend to occur in groups bounded by shear bands. The fine-grained tails and wing stripes that extend on both sides of hornblende porphyroclasts are also bounded by recrystallized ultra-fine-grained feldspar. The mantled porphyroclasts show anticlockwise stair-stepping rotation and oblique foliation trends from bottom right to top left. (5-type porphyroclast systems may be classified as <5a-type and db-type. In (5a-type systems (Fig. 4d), the hornblende porphyroclasts show extensional, asymmetrical stripes and wings, and have complex

Fig. 4. (a) Cross- and plane-polarized light photomicrographs of a mylonite composed of hornblende (Am) and plagioclase (PI) (XZ section). White lines indicate C-, C'- and Stype shear-band cleavages and (5-type (lower left) mantled porphyroclast system of hornblende. The dark veins (lower left) are pseudotachylyte. C-type and S-type shear-band cleavage geometry indicates dextral shear sense, (b) Photomicrograph of slip and microfaulting along the hornblende cleavages forming a bookshelf (bs) and leading to slip and rotation of hornblende porphyroclast in dextral sense of shear. <5-type (lower right) mantled porphyroclast system of hornblende showing stair stepping. White lines indicate C-, C'- and S-type shear-band cleavages, (c) Photomicrograph of cra-type porphyroclast system (XZ section). The complex cra-type porphyroclast system reflects inhomogeneous simple shear deformation showing bending, kinking, rotation and slip along the hornblende cleavages, (d) Photomicrograph of a-type mantled porphyroclast. The tails of mantled porphyroclast are composed of ultra-fine- to fine-grained dynamic recrystallized of hornblende grains. The dark veins are pseudotachylyte.

138

K. SARKARINEJAD

structures, involving folding, kinking and stair stepping of the wings. In <5b-type hornblende porphyroclast systems, asymmetrical wings occur on both sides with strong deflection of finegrained foliation feldspar wrapping around the hornblende wings.

Shear-band cleavages Three types of shear-band cleavages are distinguished in mylonitic rocks of the shear zone: Ctype, C'-type and S-type. The C-and C'-type shear bands correspond to the C- and C'- bands of Berthe et al. (1979) and the S-types correspond to Type-I S-C fabrics of Lister & Snoke (1984). Ctype shear bands are parallel to shear zone boundaries and at an angle of 45° to the S-type, whereas C'-type shear bands are at an angle of 100° to S-surfaces (Fig. 4a,b and c). C-type and C'-type shear bands are at an angle of 35°. It is difficult to determine precisely the age relationships between the S- and C-surfaces. Microscopic observations suggest that the two surfaces developed nearly synchronously. However, there is evidence that the S-planes bent and crosscut hornblende grains aligned along the C-surfaces, suggesting that the S-surfaces might have developed later than the C-planes.

Hornblende c- and a-axis preferred orientations The orientation of optical axes was measured on a representative number of hornblende grains (3000 grains) in thin sections parallel to the JfZ-plane (Fig. 5a) and FZ-plane (Fig. 5b) with a universal stage. All data are presented as equal-area, lowerhemisphere projections in which the plane of projection contains the stretching lineation (L) and the pole to the foliation (S). Optical fabric elements were analysed statistically in terms of eigenvalue ratios, which express the relationship of the geometry of LPO patterns as a function of the value of K on a modified Flinn diagram. K is defined as ln(S3/S2)/m(S2/Si), with confidence intervals of C = In^/Si), based on standard deviations, modified after Woodcock (1977). The measured c-axis K values range from 0.406 ± 0.011 to 1.385 ±0.27 near the southern part of the shear zone boundary (amc5, amc6, mc7) and from 0.162 ± 004 to 0.850 ± 0.08 near the centre of the shear zone (amcl, amc2, amc3 and amc4). This means that the c-axis texture has weakly developed clusters near the boundary of the shear zone, and a moderately to weakly developed girdle in the centre of the shear zone. The K values for a-axis fabrics ranges from 1.48 ± 0.031 to

4.38 ±0.146. This observation means the a-axis texture has a moderately developed girdle near the shear zone boundary and a moderately developed cluster near the centre of the shear zone. As shown in Figure 5a and b, hornblende exhibits a strong LPO with M- and G-type origin (Shelley 1993). The LPO may form under noncoaxial conditions near the shear zone boundary and at the centre of the shear zone. The oaxis pole figures indicate a moderately developed girdle characterized by a strong maximum at a high angle to the mylonitic foliation. The maximum point density pole to the foliation shows a trend from 50°NE to 49°SW. This maximum lies at a small angle to hornblende caxes and lies in (010) (Fig. 5a). This plane contains two prominent point maxima with a single, moderately developed girdle. The trend and plunge of sub-maxima concentrations is 9°NE to 29°NE (amcl, amc3, amc6). The two submaximal concentrations of (110) and (001) poles define a partial girdle with a cluster at 70-90° to the stretching lineation.

Sense of shear Microscopic examination of oriented thin sections (JfZ-plane) shows that there are two shear-band cleavages: C-type shear-band cleavages, which are nearly parallel to the SZB, and S-type shear-band cleavages, which are oblique to the C-type shear bands. The cleavages reflect the external asymmetry, which is due to the interplay of passive flattening and rotation of grains in a non-coaxial flow (Means 1981). S-type shear-band cleavages are at an angle of 32°-40° to C-type shear bands. The external asymmetry of the shear bands, which are inclined to the SZBs, is used as a shear-sense indicator. The geometry of S- and C-type shearband cleavages indicates a dextral shear sense. The da-type and at,-type mantled porphyroclast systems, as well as the (5-type that is part of the S— C fabric, were also used as shear-sense indicators. They also confirm a dextral shear sense (Fig. 5). Measurement of the angle between the shear zone plane and the LPO patterns of hornblende caxes in samples taken from near the SZB indicates that this angle varies from 35° at the shear zone margin to 70° or more in the centre. The angular relation between the shear zone foliation and the LPO of hornblende c-axes measured in thin section shows marked internal asymmetry (Fig. 5a and b), which indicates a dextral sense of shear. This asymmetry suggests an increase in strain magnitude during noncoaxial or simple shear deformation. The relationship between macroscopic shear zone foliations and hornblende LPO patterns was used to infer the shear strain magni-

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN

139

Fig. 5. (a, b) Hornblende c-axis and a-axis pole figures. Equal-area, lower-hemisphere projection showing three principal directions. The minimum eigenvector (Vj) points in the direction normal to the best-fitting (in the leastsquares sense) girdle through the data. Asymmetrical c-axis and a-axis LPO patterns relative to the reference foliation plane indicate dextral shear sense.

140

Fig. 5. Continued

K. SARKARINEJAD

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN

141

Fig. 6. Tracing of the XZ section of the sample from the Neyriz transform zone showing mylonitic foliation. This dextral mylonitic foliation was overprinted by a younger dextral shear zone foliation. The LPO of hornblende c-axes is also given (equal-area, lower-hemisphere projections).

tude. Using the shear strain angle (t/0 and the formula y = tan ip, it is possible to show that the shear strain magnitude near the SZB is 0.61, and that this value increases to 1.96 and then to 3.65 in the centre of the shear zone.

Discussion on the microstructures Hornblende and feldspar, generally classified as among the least plastic of the rock-forming minerals, were deformed by grain boundary sliding, intracrystalline glide and dynamic recrystallization in a noncoaxial strain path in different parts of the shear zone in the Neyriz ophiolite. The grain boundary sliding and dynamic recrystallization occurred under amphibolite-facies conditions. Dynamic recrystallization may have produced fine to ultrafine grains of hornblende and plagioclase that were deformed by grain boundary sliding. Dynamic recrystallization generates fine to ultrafine aggregates of hornblende and plagioclase neoblasts in a strongly foliated matrix wrapping around hornblende porphyroclasts. Both porphyroclasts and the matrix produced a-type and cr-type porphyroclast systems, indicating a change in strain that reflects multiple episodes of deformation of different strain magnitude. <5-type systems and complex objects mainly occur in high-strain mylonites, whereas cr-type objects mainly occur at lower strain (Passchier & Trouw 1996). Experimental evidence also indicates that the shape of

separator (imaginary surfaces, separating different flow patterns during inhomogeneous flow) depends on several factors such as the initial shape and orientation of the porphyroclasts, the change of these factors with time, the rheology of the matrix, the vorticity and the finite strain (Passchier & Sokoutis 1993; Passchier 1994). Changes of strain magnitude and possibly a change in the direction and magnitude of strain during multiple episodes of deformation produced small-scale shear zones superimposed on a previously strained shear zone (Fig. 6). Minor shears may have developed late during shear zone activity after a strong mineral shape preferred orientation was already established. It probably represents an energetically favourable flow partitioning in strongly isotropic materials (Platt & Vissers 1980; Platt 1984; Passchier 1991). The presence of such minor shear zones within the main shear zone is considered to reflect two phases of movement and simple shear deformation under amphibolite-facies conditions.

Kinematic model for the origin ofmylonite In the shear zone within the Neyriz ophiolite hornblende and plagioclase have been deformed by plastic grain boundary sliding and dynamic recrystallization of the hornblende and feldspar matrix and by sliding along cleavage planes in hornblende porphyroclasts. This plastic deforma-

142

K. SARKARINEJAD

tion took place under high-temperature amphibolite-facies conditions (450-650 °C). Several slip systems have been documented in the hornblende grains, mainly [001], but also (010) [100], which, according to Skrotzki (1992), suggest high temperature or low strain rate. Sub-grains are elongated parallel to the oaxis and sub-grain boundaries of amphibole consist of a simple array of [001] [100] (Reynard et al 1989). Measurements of hornblende orientation (LPO) show that the (100) poles are concentrated close to Z (the tectonic fabric X-, Y- and Z-axes) and two submaxima of the (001) poles are concentrated close to Y. Both LPO patterns seem to have formed at high temperatures in a progressive non-coaxial deformation. Such conditions of deformation are absent in other parts of the Neyriz ophiolite. Mesoscopic observations of the shear zone show a pervasive oblique foliation with a mean orientation of N77°W. This oblique foliation in the axis of the shear zone is characterized by compositional layering and a subhorizontal hornblende stretching lineation (XZ strain axes) marked by elongate porphyroclasts and matrix, and these features collectively indicate a dextral sense of shear. Study of minor superimposed shear zones within the major shear zone also indicates dextral sense of shearing. Microscopic examination of the mylonite shows a mantled porphyroclast system. The morphology of the mantled porphyroclasts clearly shows Cand C'-type shear-band cleavages, which correspond to the C- and C' bands of Berthe et al. (1979). Under the microscope, cr-type, Ob-type and (5-type systems have been recognized (Fig. 4). The stair-stepping of their wings is used to confirm the dextral shear sense. The preferred orientation of hornblende [001] poles indicates a moderately developed girdle with a maximum concentration close to Z, which is inclined to the reference foliation plane. The foliation plane here is interpreted as the Si—£2 plane of the finite strain ellipsoid. Deformation was under noncoaxial conditions because the [001] maximum is inclined to the foliation. Therefore, the LPO of hornblende oaxes clearly indicates a dextral shear sense.

(Abbate et al. 1980) and in the Bogota peninsula of New Caledonia (Prinzhofer & Nicolas 1980). It is also similar to modern oceanic transform faults (Van Andel et al. 1971; Cannat et al. 1991; Gudmundsson et al. 1993; Jancin et al. 1995). The Neyriz shear zone is hence interpreted as a fragment of a fossil oceanic transform fault zone. This palaeo-transform fault shows a dextral shear sense, suggesting that it acted as a NNW-SSEtrending (in present co-ordinate system), rightlateral strike-slip fault system connecting two ridge segments (Fig. 7). Radiolarite, pillow lavas and sheeted dykes of the ophiolite are located c. 2 km north of this palaeo-transform fault. The sequence is relatively undisturbed and undeformed, showing no sign of ductile deformation. Chilled margins on the sheeted dykes have a mean strike of N46°W. The dip of individual dykes is moderate to subvertical, indicating that they might have undergone postintrusion tilting. Several procedures have been suggested for untilting sheeted dykes (Bonhommet et al. 1988; Allerton 1989). The most commonly used method involves plotting the rotation axis on a stereonet (Fig. 3). According to Borradaile (2001), a single rotation axis inferred from structural analysis is perhaps a better approximation to the true geological rotation history. Despite using this simplified approach, Borradaile (2001) emphasized that correction for tectonic rotation requires recognition of the actual axes and the sequence of tectonic activities. The mean orientation of the restored sheeted dykes is N66°E, dipping 89° (Fig. 3). This mean orientation is nearly perpendicular to the direction of the palaeo-transform fault. This transform fault is approximately parallel to the mean orientation of the harzburgite tectonite foliation, which is N50°W/54°NE (Sarkarinejad 1994). The measured mean of hornblende oaxes is N40°W to N57°W, indicating that the trends of the harzburgite foliation and the hornblende oaxis LPO are parallel (Fig. 8). The direction of mantle flow, which reflects a noncoaxial simple shear rotation parallel to the dextral transform fault, is inferred to have been N50°W/54°NE. The oceanic palaeo-transform fault is inferred to have linked ENE-oriented spreading centre segments (Fig. 7) in a NeoTethyan ocean basin.

Neyriz shear zone as a fossil oceanic transform fault

Conclusions

Formation of the mylonitic fabric and LPO under amphibolite-facies conditions in the Neyriz shear zone is similar to the formation of other on-land fossil oceanic transform faults, such as in Troodos (Moores & Vine 1971; Murton 1990; Gass et al. 1994; Borradaile 2001), in the northern Apennines

The hornblende c-axes of high-grade S-C mylonite from the Neyriz palaeo-transform fault exhibit strong LPO with M- and G-type (mechanical and growth) origins. The hornblende porphyroclasts are commonly mantled and have either symmetrical (aa-type and db-type) or asymmetrical tails

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN

143

Fig. 7. Kinematic model of the Neyriz ridge-transform intersection modelled according to all microstructural and structural data for the shear zone: harzburgite foliation orientation, sheeted dyke margin orientations and mapping of the ophiolite sequence.

(da-type and (5b-type). The mantled porphyroclast systems show clockwise stair-stepping rotation and slip along the cleavages and grain boundary sliding. The hornblende porphyroclasts are connected by narrow bands or stripes of stretched and elongated hornblende neoblasts. The ultrafine grains of hornblende (1-10 urn size), which form stripes and lines, have no fabric. A decrease in grain size enhances grain-size-dependent deformation such as diffusion creep and grain boundary sliding (White et al 1980). The size of the neoblasts, which formed by dynamic recrystallization, is a function of differential stress (Passchier & Trouw 1996). Grains in the fine-grained matrix can slide past each other to form stripes and lines. The NNW-SSE-trending dextral transform fault with a nearly vertical dip originated at high temperatures (amphibolite-facies conditions) and at high differential stress. Numerous minor dextral

shear zones were later superimposed on the original shear zone, indicating several stages of reactivation. Movement along the palaeo-transform fault progressively rotated the asthenospheric mantle flow toward parallelism with the transform fault (Fig. 8). The asthenospheric mantle flow shear strain may have been smaller than that within the transform fault.

I would like to thank Y. Dilek, B. John, P. T. Robinson and D. Shelley for their valuable suggestions and constructive comments on earlier versions of this paper. Thanks are also due to A. Nadimi for his effective cooperation on this project. Financial support of the Research Consul in Shiraz University is gratefully acknowledged.

144

K. SARKARINEJAD

Fig. 8. Schematic diagram of the interpretative dextral transform fault in the Neyriz ophiolite; dashed lines depict shear-band cleavage; continuous lines are asthenospheric mantle flow. Volcanic rocks were extruded along the transform fault.

References ABBATE, E., BORTOLOTTI, V. & PRINCIPI, G. 1980. Apennine ophiolites: a peculiar oceanic crust. Ofioliti, 1, 59-96. ALLERTON, S. 1989. Fault block rotations in ophiolites: results of paleomagnetic studies in the Troodos complex, Cyprus. In: KISSEL, C. & LAJ, C. (eds) Paleomagmetic Rotation and Continental Deformation. NATO Conference Volume. Kluwer, Dordrecht, 393-410. BABAIE, H.A., GHAZI, A.M., BABAIE, A., LA TOUR, T.E. & HASSANIPAK, A. A. 2001. Geochemistry of volcanic rocks of the Zagros Crush zone, Neyriz, Iran. Journal of Asian Earth Sciences, 19, 61-75. BERTHE, D., CHOUKROUNE, P. & JEGOUZO, P. 1979. Orthogneiss, mylonite and non-coaxial deformation of granites: the example of the south Armorican shear zone. Journal of Structural Geology, 1, 31-42. BONHOMMET, N.,

ROPERCH, P. & CALZA, F.

1988.

Paleomagnetic arguments for block rotations along the Arakapas fault (Cyprus). Geology, 16, 422-425. BORRADAILE, G.J. 2001. Paleomagnetic vectors and tilted dikes. Tectonophysics, 333, 417-426. BOUDIER, F. & COLEMAN, R.G. 1981. Cross section through the peridotites in Samail ophiolite, southeastern Oman. Journal of Geophysical Research, 86, 2573-2592. BOWES, D.R. (ed.) 1989. The Encyclopedia of Igneous and Metamorphic Petrology. Van Nostrand Reinhold, New York. CONNAT, M., MEVEL, C. & STAKES, D. 1991. Stretching of the deep crust at slow-spreading Southwest Indian Ridge. Techonophysicts, 190, 73-94.

FLEET, M.E. & BARNETH, R.L. 1978. A1IV/A1/VI partitioning in calciferous amphiboles from the Frood Mine, Ontario. Canadian Mineralogy, 16, 527-532. GASS, L, MCLEOD, C.J., MURTON, B.J., PANAYIOTOU, A., SIMONIAN, K.O. & XENOPHONTOS, C. 1994. The Geology of the Southern Troodos Transform Fault Zone, Cyprus. Geological Survey Department Memoir, 9, 218. GRAPES, R.H. & GRAHAM, C.M. 1978. Actinolitehornblende series in metabasites and so-called miscibility gap: a review. Lithos, 11, 85-97. GUDMUNDSSON, A., BRYNOJOLFESSON, S. & JONSSON,

M.T. 1993. Structural analysis of a transform faultrift zone junction in north Iceland. Tectonophysics, 220,205-221. JANCIN, M., YOUNG, K.D., VOIGHT, B. & ORKAN, N.I. 1995. Dikes, minor fault and mineral veins associated with a transform fault in north Iceland. Journal of Structural Geology, 17, 1627-1631. KARSON, J.A. 1984. Variations in structure and petrology in coastal complexNewfoundland: anatomy of an oceanic fracture zone. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 131-144. LEAKE, B.E. 1978. Nomenclature of amphibole. Mineralogical Magazine, 42, 533-563. LISTER, G.S. & SNOKE, A.W. 1984. S-C mylonites. Journal of Structural Geology, 6, 617-663. MEANS, W.D. 1981. The concept of steady-state foliation. Tectonophysics, 78, 179-199. MOORES, E.M. & VINE, FJ. 1971. The Troodos Massif, Cyprus and other ophiolites as oceanic crust: evaluations and implications. Philosophical Transactions of the Royal Society of London, Series A,

STRUCTURE OF THE NEYRIZ OPHIOLITE, IRAN 268, 443-466. MURTON, BJ. 1990. Was the southern Troodos transform fault a victim of microplate rotation? In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Oceanic Crustal Analogues: Proceedings of Symposium, Troodos 1987. Geological Survey Department, Nicosia, 53-64. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere: Petrology and Structural Geology. Kluwer, Dordrecht, 4. NICOLAS, A., CEULENEER, G., BOUDIER, F. & MISSERI, M. 1988. A structural mapping in the Oman ophiolites: mantle diapirism along an oceanic ridge. Tectonophysics, 151, 27-56. PASSCHIER, C.W. 1991. Geometric constraints on the developments of shear bands in rocks. Geologic en Mijnbouw,7Q,2Q3-2U. PASSCHIER, C.W. 1994. Mixing in flow perturbations: a model for development of mantled porphyroclasts in mylonites. Journal of Structural Geology, 16, 733-736. PASSCHIER, C.W. & SOKOUTIS, D. 1993. Experimental modeling of mantled porphyroclasts. Journal of Structural Geology, 15, 895-910. PASSCHIER, C.W. & TROUW, R.AJ. 1996. Microtectonics. Springer, Berlin. PLATT, J.P. 1984. Secondary cleavages in ductile shear zones. Journal of Structural Geology, 6, 439-442. PLATT, J.P. & VISSERS, R.I.M. 1980. Extensional structures in anisotropic rocks. Journal of Structural Geology, 2, 397-410. PRINZHOFER, A. & NICOLAS, A. 1980. The Bogota Peninsula, New Caledonia: a possible oceanic transform fault. Journal of Geology, 88, 389-397. REYNARD, B., GILLET, P. & WILLIAME, C. 1989. Deformation mechanisms in naturally deformed glaucophanes: a TEM and HREM study. European Journal of Mineralogy, 1, 611-624. Ricou, L.E. 1968a. Sur la mise en place au Cretace Superieur d'importantes nappes a radiolarites et ophiolite dans les Monts Zagros (Iran). Comptes Rendus de I'Academic des Sciences, Serie D, 267, 2272-2275. Ricou, L.E. 1968b. Une coupe a travers les series a radiolarite des Monts Pichakun (Zagros Iran).

145

Bulletin de la Societe Geologique de France, 10(Fr. 7 Series), 478-485. SARKARINEJAD, K. 1985. The geology and tectonic settings of the ophiolites and associated rocks in the Neyriz area. PhD thesis, University of Wales, Cardiff. SARKARINEJAD, K. 1994. The geology and tectonic settings of the Neyriz ophiolite, southeastern Iran. In: ISHIWATERI, A., MALPAS, J. & ISHIZUKA, H. (eds) Proceedings of 29th International Geological Congress, Part D. Kyoto, Japan, VSP International Science Publishers, 221-234. SHELLEY, D. 1989. Plagioclase and quartz preferred orientation in lowgrade schist: the role of primary growth and plastic deformation. Journal of Structural Geology, 11, 1029-1038. SHELLY, D. 1993. Igneous and metamorphic rocks under the microscope. Chapman & Hall, London. SIMONIAN, K.O. & GASS, I.G. 1978. Arakapas fault belt, Cyprus: a fossil transform fault. Geological Society of American Bulletin, 89, 1220-1230. SKROTZKI, W. 1992. Defect structure and deformation mechanism in naturally deformed hornblende. Physical State of Solids, 131, 605-628. STONELEY, R. 1981. The geology of the Kuh-e Dalneshin area of southern Iran, and it bearing on the evolution of southern Tethys. Journal of the Geological Society, London, 138, 509-526. VAN ANDEL, T.H., VON HARZEN, R.P. & PHILLIPS, J.D. 1971. The Vema Fracture Zone and the tectonics of transverse shear zone in oceanic crustal plates. Marine Geophysical Researches, 1, 261-283. WHITE, S.H. 1975. Tectonic deformation and recrystallization of oligoclase. Contributions to Mineralogy and Petrology, 50, 287-304. WHITE, S.H. 1979. Large strain deformation: Report on a techtonic studies group discussion meeting held at Imperial College, London, introduction. Journal of Structural Geology, 4, 333-339. WHITE, S.H., BURROWS, S.E., CARRARAS, N.D. & HUMPHREYS, FJ. 1980. On mylonite in ductile shear zones. Journal of Structural Geology, 2,175-187. WOODCOCK, N.H. 1977. Specification of fabric shape using eigenvalue. Geological Society of American Bulletin, 88, 1231-1236.

This page intentionally left blank

Stratigraphic and sedimentological constraints on the age and tectonic evolution of the Neotethyan ophiolites along the Yarlung Tsangpo suture zone, Tibet JONATHAN C. AITCHISON 1 , AILEEN M. DAVIS 1 , ALEXANDRA V. ABRAJEVITCH 1 , JASON R. ALI 1 , BADENGZHU 2 , JIANBING LIU 1 ' 3 , HUI LUO 1 ' 4 , ISABELLA R. C. MCDERMID 1 & SERGEY V. ZIABREV 1 1

Tibet Research Group, Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong SAR, PR. China (e-mail: [email protected]) 2 Geological Team No. 2, Tibet Geological Survey, Lhasa, Tibet, P.R. China ^Present address: Institute of Geology and Mineral Resources, Bureau of Geology and Mineral Resources of Xinjiang, 16 Youhao Beilu Road, Urumqi, Xinjiang, PR. China ^Present address: Nanjing Institute of Geology and Palaeontology, Laboratory of Palaeobiology and Stratigraphy, Academia Sinica, Nanjing 210008, PR. China Abstract: Ophiolitic rocks distributed along the Yarlung Tsangpo suture zone in southern Tibet are the few remaining fragmentary remnants of many thousands of kilometres of the ocean space that formerly existed between India and Eurasia. Portions of mid-Jurassic and midCretaceous intra-oceanic island arcs can be recognized amongst those rocks that have been studied in detail. Complete suprasubduction zone ophiolite successions are preserved in the Dazhuqu terrane, which crops out both east and west of Xigaze. Radiolarians in inter-pillow cherts and immediately overlying sedimentary rocks indicate a Barremian ophiolite generation event. Palaeomagnetic data show that this ophiolite formed at equatorial latitudes south of the Lhasa terrane before its south-directed emplacement onto the northern margin of India. Highly refractory ultramafic rocks in the Luobusa ophiolite appear to be of Mid-Jurassic age and are potentially related to intra-oceanic island arc remnants in the nearby Zedong terrane. Ophiolitic massifs along the suture in western Tibet are thrust southwards onto northern India and record Late Jurassic ocean-floor development. Miocene north-directed back-thrusting associated with India-Asia collision has further complicated interpretation of regional geology. The Ophiolitic rocks of the Yarlung Tsangpo suture zone provide evidence for the former existence of multiple oceanic island arc segments within Neotethys and suggest that consumption of the oceanic space between India and Asia was more complicated than has been predicted by existing models.

The vast Neotethyan Ocean that once separated Eurasia and India finally closed during the Cenozoic along the Yarlung-Tsangpo suture zone (YTSZ) in Tibet. The suture traverses southern Tibet meridionally, at c. 29°N, and is marked by a discontinuous belt of Ophiolitic bodies (Fig. 1). It extends westwards beyond Tibet into northern India and experiences a major change in strike near Namche Barwa (7782 m 29°40'N, 095°10'E) in eastern Tibet, where it heads southward towards Burma. Although most of what once lay within Neotethys was subducted, and otherwise destroyed during continental collision, a few remnants of Tethyan affinity are preserved along, and marginal to, the YTSZ. The suture is characterized by the presence of a series of outcrops of rocks of

Ophiolitic affinity, the nature and distribution of which are reviewed in this paper, The Ophiolitic rocks along the suture (Table 1) have locally been mapped in detail by geologists of the Bureau of Geology and Mineral Resources of the Xizang Autonomous Region during the search for Cr resources in this region (Badengzhu 1979, 1981; Zhang & Fu 1982; Wei & Peng 1984). The first detailed studies of the ophiolites were carried out as part of investigations during the 1980s into the nature of the India-Asia collision zone. In particular, this work concentrated on the Xigaze region, where a major zone of outcrops, up to 25 km wide, includes several Ophiolitic massifs. These rocks are near continuously exposed for over 175 km along strike from

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 147-164. 0305-8719/037$ 15 © The Geological Society of London 2003.

148

J. C. A I T C H I S O N ^ r ^ f L .

Fig. 1. Geological sketch map of the Yarlung Tsangpo suture zone across southern Tibet together with settlements and places mentioned in the text. YTSZ, Yarlung Tsangpo suture zone; BNS, Bangong-Nujiang suture; MCT, Main Central thrust of the Himalaya; MBT, Main Boundary thrust of the Himalaya. The approximate position of the Yarlung Tsangpo is shown as a continuous line and the suture is shown as a subparallel grey zone.

near Renbung, 250 km SW of Lhasa, to Ngamring Tso. Numerous descriptions of the petrography, geochemistry, structure, metamorphism, palaeomagnetism and tectonic evolution of these rocks emanated from investigations by Chinese, French, British and German researchers (Nicolas et al. 1981; Shackleton 1981; Tapponnier et al. 1981; Burg 1983; Allegre et al. 1984; Girardeau et al. 1984b, 1984c, 1985a, 1985b, 1985c; Wang et al. 1987; Girardeau & Mercier 1988; Einsele et al. 1994; Diirr 1996). Although technically disrupted and heavily attenuated, sections locally display a complete ophiolite sequence from fresh Cr diopside-rich harzburgites to marine sedimentary cover developed on top of pillowed mafic volcanic rocks (Nicolas et al. 1981; Girardeau et al. 1984b, 1985a, 1985c). Interpretations of these rocks were much influenced by models available at that time for understanding ophiolite generation. The ophiolite was generally regarded to have been generated at a mid-ocean ridge and was considered to represent the basement to forearc turbidites of the Xigaze Group. More recent investigations include studies of the Luobusa ophiolite east of Zedong (Huang et al. 1981; Bai et al. 1993, 2000; Zhou & Robinson 1994; Zhou 1995; Zhou et al. 1996, 2002; Griselin et al. 1999; Hu 1999; Hebert et al. 2000, 2001) and re-examinations of the geochemistry and mineralogy of ophiolitic occurrences along the suture zone (Wang et al. 1999; Hebert et al. 2000, 2001). The suprasubduction zone affinity of this ophiolite has been widely recognized in these contemporary studies. The first information regarding the age of the ophiolite became available in reports of SinoFrench expeditions to this area in the early 1980s (Marcoux et al. 1982). The cherts as well as finegrained clastic deposits that depositionally overlie

the ophiolite were accorded a late Albian to possibly early Cenomanian age based on radiolarian faunas. Other radiolarian fossils described later from the same deposits were interpreted as upper Albian-lower Cenomanian (Li & Wu 1985) or lower Cenomanian (Wu 1986) based on correlation with the Archaeospongoprunum techamaensis Zone of California. Over the past six summers, members of the Tibet Research Group of the Department of Earth Sciences at the University of Hong Kong have made detailed field investigations of stratigraphic relationships amongst rocks exposed along the Yarlung Tsangpo suture zone (Aitchison et al. 2002a). Research began with the aim of testing existing models and the hope of being able to develop a new and improved understanding of the tectonic evolution of this region. In particular, when compared with areas such as the modern western Pacific, the existing model involving consumption of the Tethys at a single long-lived north-dipping subduction zone developed along the southern margin of the Lhasa block appeared likely to be oversimplified. Detailed regional investigations have resulted in the better discrimination of several discrete terranes together with other important units such as syncollisional conglomerates along the suture (Fig. 2). Micropalaeontological studies have provided high-precision age constraints for ophiolitic and subduction complex rocks along the suture. Together these results can be used to place constraints on the timing of interactions between the terranes present and the development of the collision zone between India and Asia. This work has resulted in the recognition of the possibility that remnants of more than one subduction system that may have existed within Neotemys are preserved within the suture zone (Aitchison et al. 2000).

Table 1. Summary of key characteristics of various ophiolitic rocks along the Yarlung Tsangpo suture zone Area

Lithologies

Longitude

Lithological-tectonic unit

Age constraints

Affinity

Primary contacts

Jungbwa

81°E

Jungbwa ophiolite

MOR

83°E

Dangxiong ultramafic

Late Jurassic 152 ± 33 Ma (Ar/Ar, Miller et al. 2003) Undetermined

Xigaze

Ultramafic (harz and cpx-poor 1hz) Mostly ultramafic, minor basalt Complete ophiolite

85°E-90°E

Dazhuqu terrane

Mid-Cretaceous, Barremian SSZ (radiolarians, Ziabrev 2001)

Zedong

Mostly harzburgite

92°E

Luobusa ophiolite

Mid- Jurassic 177 ± 31 Ma SSZ (Sm-Nd, Zhou et al. 2002)

Nappe thrust southwards over northern edge of Indian terrane Thrust southwards over northern edge of Indian terrane Thrust southwards over Bainang terrane accretionary complex and Indian terrane; later (Miocene) north-directed backthrusting over Gangrinboche facies conglomerates Originally southwards onto Indian terrane; later (Miocene) north-directed backthrusting over Gangrinboche facies conglomerates onto Lhasa terrane; close (faulted) spatial relationship with late MidJurassic Zedong terrane intra-oceanic arc rocks

Dangxiong

Unknown

Fig. 2. (a) Geological map of the Xigaze-Renbung area (modified after Wang et al. 1987) indicating the disposition of major tectonic entities in this region. GCT, Great Counter thrust. The distribution of Yamdrok melange is not shown but it mostly crops out within the northern Indian terrane within the 20 km south of the suture zone, (b) Geological map of the Zedong-Luobusa area indicating the disposition of the main tectonic entities in this region (modified after Badengzhu 1979, 1981). RZT, Renbu-Zedong thrust.

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES

Regional tectonic framework Several terranes and tectonically significant units that developed before, and during, the IndiaEurasia collision are recognized within and bounding the suture (Aitchison et al 2000, 2002a; Fig.

151

3). As some knowledge of these rocks is useful in the context of evolution of the suture zone a brief description is provided below. To the north of the suture lies the Lhasa terrane, a microcontinental block that had detached from the northern periphery of Gondwana and docked

Fig. 3. Summary time-space plot for terranes and tectonically significant units recognized within, and bounding, the Yarlung Tsangpo suture that developed before, and during, the India-Eurasia collision.

152

J. C. AITCHISON ETAL.

with Asia by the Late Jurassic (Allegre et al 1984; Yin & Harrison 2000). It is reported to contain a Middle Proterozoic to Lower Cambrian metamorphic basement overlain by Palaeozoic to middle Cretaceous shallow-marine and terrestrial deposits. The southern portion of the Lhasa terrane, immediately north of the suture, is marked by Upper Jurassic-Lower Cretaceous metasedimentary and metavolcanic rocks (Sangri Group), which are intruded by the regionally extensive calc-alkaline Gangdese (Trans Himalayan) batholith. These igneous rocks record extensive magmatism that resulted from northward subduction of Neotethyan oceanic lithosphere beneath the Lhasa terrane. Andesites within the Sangri Group are intercalated with fossiliferous Upper Jurassic Lower Cretaceous clastic and carbonate deposits (Badengzhu 1979; Bureau of Geology and Mineral Resources of Xizang Autonomous Region Geological Team No. 2 1979; Badengzhu 1981; Pearce & Deng 1988) and represent the oldest subduction-related volcanic rocks in the southern Lhasa terrane. Radiometric ages reported from Gangdese plutons range from 153 ± 6 Ma (Murphy et al. 1997) to 30.4 ±0.4 Ma (Harrison et al 2000). Extrusive units within remnants of the volcanic carapace of the batholith occur in the Lingzizong Formation and range from 60 to 48 Ma (Maluski et al 1982; Xu et al 1985). These volcanic rocks overlie the Upper Cretaceous Takena Formation with an angular unconformity indicating a latest Cretaceous deformational episode in the Lhasa terrane. Cretaceous units in the southern Lhasa terrane are also reported to contain andesitic volcanic rocks (Yin et al 1988) and volcanogenic detritus (Coulon et al 1986). Radiometric and biostratigraphic data thus both appear to indicate that subduction-related magmatism, associated with the consumption of Neotethyan oceanic lithosphere along the southern margin of the Lhasa terrane, commenced in the Late Jurassic and lasted until mid-Oligocene. The southern margin of the Lhasa terrane is marked by a regionally extensive sequence of conglomerates; the 'Gangrinboche' facies (Aitchison et al 2002b). These conglomerate units, known by a variety of local geographical names (Luobusa, Dazhuqu, Qiuwu and Kailas formations) crop out along the strike of the suture for at least 1500km (Aitchison et al 2002b). They record the latest Oligocene-earliest Miocene deposition of coarse clastic sediments. Each unit lies unconformably upon a basement of Lhasa terrane rocks that initially was the sole source of sediment. Up-section detritus from sources south of the suture then becomes increasingly dominant. Development of these conglomerates was a direct result of the India-Asia collision and our under-

standing of these siliciclastic rocks has significant implications for regional tectonic models (Aitchison & Davis 2001; Aitchison et al 2002b, 2003). The southern extent of outcrop of the Gangrinboche facies is typically marked by a north-directed thrust fault known variously as the Great Counter thrust or Renbu Zedong thrust system (Gansser 1964; Yin et al 1994). Numerous strands of this fault exist. It carries rocks of both suture zone and Indian terrane affinity in its hanging wall and places them over a footwall of southern Lhasa terrane rocks. In some areas, northward displacement along this Miocene fault system has totally obscured or removed remnants of the YTSZ, which is marked only by a zone containing schistose serpentinite-matrix melange. To the south of the northernmost strand of this fault system, a succession of several kilometres thickness composed of volcaniclastic turbidites (Xigaze terrane) abuts the Lhasa terrane and Gangrinboche conglomerates. These rocks extend westwards along the northern flank of the suture zone from around 90°E. The Xigaze terrane is thrust northwards over upper Oligocene-lower Miocene Gangrinboche facies conglomerates (Aitchison et al 2002b). The original geometry of the India-Asia collision suggests that the Xigaze terrane should have been thrust southwards over the suture zone ophiolites and northern Indian terrane. However, at its southern boundary the terrane at present lies in the footwall of a younger (late Micoene) north-directed thrust with ophiolitic rocks of the Dazhuqu terrane or Paleogene Liuqu Conglomerate (Davis et al 2002) in the hanging wall. Sparse fossils indicate that the turbidites have an upper Albian to Coniacian stratigraphic range (Bassoullet et al 1984; Wiedmann & Durr 1995; Wan et al 1998). As the oldest fossils reported are not from the base of the section, it is possible that sedimentation may have commenced before the late Albian. The top of the Xigaze turbidite sequence is truncated by erosion. These rocks are interpreted as a forearc succession that developed in association with north-directed subduction of Neotethyan oceanic lithosphere beneath the Lhasa terrane (Shackleton 1981; Burg & Chen 1984; Girardeau et al 1984b; Einsele et al 1994; Durr 1996; Wang et al 1999). Development of the terrane is likely to have been intimately related to evolution of the magmatic arc along the southern edge of the Lhasa terrane (Einsele et al 1994; Durr 1996). Although the Xigaze terrane is conventionally regarded as being floored by the Dazhuqu ophiolite (e.g. Burg & Chen 1984; Girardeau et al 1984a; Einsele et al 1994; Durr 1996), these two units are ubiquitously in tectonic contact and are best regarded as separate terranes (Aitchison et al 2000).

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES A tectonic sliver of Middle Jurassic intraoceanic island arc rocks (Zedong terrane) crops out at the northern edge of the suture between Zedong and Luobusa. This terrane is entirely bounded by north-directed thrusts related to the Miocene Renbu-Zedong thrust system (Yin et al. 1994). The base of the section is faulted and begins with a thin (several metres) succession of arc tholeiitic lavas overlain by a c. 15m thick sequence of red ribbon-bedded chert then 1000+ m of volcaniclastic breccias of shoshonitic affinity. The succession is cut by numerous shoshonitic dykes with minor intrusions of diorite and leucogranite (McDermid et al. 200Ib). Both radiometric and biostratigraphic data indicate the onset of magmatism in the late Mid-Jurassic. Radiometric ages of between 152 and 161 Ma (McDermid 2002; McDermid et al. 2002) are in accord with Bajocian-early Callovian radiolarian faunas recovered from the underlying chert. The terrane is interpreted as remnants of an intraoceanic magmatic arc (Aitchison et al. 2000; McDermid et al. 200la; McDermid 2002) potentially similar to other terranes known from elsewhere along the suture in NW India and Pakistan. In most areas, the main trace of the suture is marked by ophiolitic rocks such as those associated with the Dazhuqu terrane. These rocks are discussed in more detail elsewhere in the text. Accretionary prism rocks are common on the southern side of the suture zone, and include those of the Bainang terrane (Aitchison et al. 2000). This terrane lies between ophiolitic rocks of the Dazhuqu terrane to its north and the Indian terrane to the south. Early contacts between these terranes are south-directed thrust faults (Burg 1983; Burg & Chen 1984; Ratschbacher et al. 1994). It contains units previously referred to as infraophiolitic thrust sheets of radiolarites (Burg & Chen 1984) or Upper Jurassic-Lower Cretaceous red radiolarites (Girardeau et al. 1984a). Good exposures exist near Donglha, Xialu and Bainang. Radiolarians reported from siliceous rocks near Xialu have ages ranging from the Mid-Jurassic to Cretaceous (Aptian) (Wu 1993; Matsuoka et al. 200la, 200Ib). Detailed geological mapping and investigations of radiolarian biostratigraphy elucidate the structure, stratigraphy and evolution of the terrane (Ziabrev et al. 2000, 2003a; Ziabrev 2001). It contains numerous north-facing and chiefly south-verging tectonic slices that experienced structural imbrication during the Cretaceous as the Bainang terrane accretionary prism developed. These slices incorporate various lithologies: chert, siliceous and tuffaceous mudstones, limestone, siliceous and calcareous shales. The most complete and well-exposed piece of the terrane near Bainang has been mapped and sampled for

153

radiolarian biostratigraphic study (Ziabrev 2001). Five lithotectonic units can be discriminated based on their characteristic lithologies and structural styles. Radiolarians allow reconstruction of a relict stratigraphy that records a long history of sedimentation in different portions of Tethys since the Late Triassic (Ziabrev 2001). This stratigraphy records the northward travel of an oceanic plate and its mid-Cretaceous approach towards a subduction zone in a remote intra-oceanic setting (Ziabrev et al. 2003a). Passive margin rocks of the Indian terrane or Tethyan (Tibetan) Himalaya lie south of the suture. They include thick Permian to Lower Cenozoic continental rise deposits (Liu 1992), which merge southward into a continuous Ordovician to Eocene shelf sedimentary succession of marine carbonates, sandstone, siltstone and shale (Bureau of Geology and Mineral Resources of Xizang Autonomous Region 1993; Jadoul et al. 1998). The development of the passive continental margin facing the Tethyan domain was punctuated by a series of rifting episodes related to Gondwana disintegration and associated with intra-plate volcanism (Gaetani & Garzanti 1991). In the western Himalaya the Zanskar shelf merges northward with Mesozoic slope-rise deep-sea deposits of the Lamayuru complex and its distal equivalent, the Karamba complex (Danelian & Robertson 1997; Robertson 1998). Fragments of carbonate-topped Permian seamounts in the Ladakh Himalaya (Robertson & Sharp 1998) and south Tibet (Aitchison et al. 2000), which locally fringe the northern Indian passive margin, may have developed on the oldest Tethyan oceanic crust. Passive margin was terminated by the Cenozoic IndiaAsia collision. Disruption of northern Indian margin rocks into widespread regional melange zones accompanied Paleogene collision events (Liu 2001; Liu et al. 2002). However, continuing deposition of marine carbonate sediments in the Indian terrane until at least the Eocene-Oligocene boundary (Wang et al. 2002) suggests that continent-continent collision and final closure of Tethys was a later event. Extensive zones of disrupted rocks occur as the Yamdrok melange to the south of, and parallel to, the Yarlung Tsangpo ophiolites. They extend several hundreds of kilometres with a width of several to tens of kilometres (Liu & Einsele 1996). Field investigations near Gyangze indicate that these mud-matrix melanges are sandwiched within zones of Mesozoic rocks of India passive margin affinity (Liu 2001). Contacts are irregular and transgress bedding, indicating a likely tectonic and/or diapiric origin for the melange (Pini 1999). The ages of cherty blocks in the melange range from Late Jurassic to Late Cretaceous.

154

J. C. AITCHISON E7.4L.

Siliceous siltstones and mudstones within the melange matrix have yielded lower Aptian radiolarians. A new uppermost Paleocene radiolarian fauna from the matrix of the melange near Gyantze places important maximum age constraints on the timing of formation of this unit (Liu & Aitchison 2002). Several conglomerate units are exposed along the YTSZ and each one possibly documents discrete tectonic events. Rapidly deposited coarse clastic rocks of the Liuqu Conglomerate record a Paleogene phase of sedimentary basin development along the suture (Davis et al 1999, 2001, 2002). Sediment is dominated by conglomerates, and deposition occurred in both terrestrial and sub-aqueous settings. Facies changes are commonly abrupt, with rapid changes in clast types, grain size and depositional styles. Proximal deposits are locally offset relative to their original source terranes. These coarse clastic sedimentary rocks are interpreted as having been deposited in oblique-slip basins that developed as an intraoceanic island arc collided with the leading (northern) edge of the Indian margin (Davis et al. 1999, 2001, 2002). Sediments were derived from both areas and the conglomerates record aspects of the history of collision between these terranes. The absence of clasts derived from terranes to the north of the Yarlung-Tsangpo suture suggests that basins in which the coarse clastic units accumulated developed before the main collision between India and Asia. The original disposition of terranes within the suture zone has been greatly disrupted and former relations between terranes are not well constrained. Reconstruction of the tectonic evolution of the area is therefore difficult. Most early models (Allegre et al. 1984; Searle et al. 1987) suggested that a single Andean-type convergent plate margin along the northern side of Tethys was responsible for destruction of the Neotethys, although the possibility of additional subduction zones was considered by some workers (Proust et al. 1984). The co-occurrence and north-south distribution of the Zedong (magmatic arc), Dazhuqu (suprasubduction zone ophiolite) and Bainang (subduction complex) terranes led to their interpretation as evidence of the former existence of a south-facing intra-oceanic subduction system that lay within the Tethys (Aitchison et al. 2000) and the possible existence of more than one convergent margin. Analogy with the modern western Pacific and SE Asia suggests that reality may have been even more complex. As more details and constraints on the evolution of terranes within the YTSZ become available, the complexity and sophistication of models for this zone is likely to further increase.

Ophiolitic rocks along the Yarlung Tsangpo suture zone Dazhuqu terrane: Xigaze Ophiolitic rocks in the Xigaze region are the best known occurrences along the suture because of their proximity to one of the few highways through Tibet (Wang et al. 1987). Outcrop is near continuous from Ngamring-Tso eastwards to Renbung. The ophiolite in this area was the subject of detailed investigations by Chinese and French geologists particularly in the early 1980s. Early investigations recorded details of the petrology and geochemistry of lithologies present, the metamorphic and structural history of the ophiolitic massifs (Nicolas et al. 1981; Shackleton 1981; Tapponnier et al. 1981; Burg 1983; Allegre et al. 1984; Girardeau et al. 1984b, 1984c, 1985a, 1985b, 1985c; Wang et al. 1987; Girardeau & Mercier 1988; Einsele et al. 1994; Diirr 1996). Numerous massifs of mostly ultramafic rocks are accompanied by all other elements of an ophiolite suite and together these rocks are assigned to the Dazhuqu terrane (Aitchison et al. 2000). Ultramafic massifs dominate the zone of outcrop with major occurrences at Dazhuqu (Fig. 4e), Qunrang, Xigaze, Tiding, Liuqu and Ngamring. Other elements of an ophiolite suite are also present, although lower-crustal sections are not well preserved and have locally been eliminated by tectonic attenuation of the section. Sheeted dykes are well developed both SE of Qunrang (Fig. 4b) and SW of Sagui. An important observation of Girardeau et al. (1985c) was that diabase dykes occur at all levels throughout the entire ophiolite section, indicating the possibility of more than one phase in the magmatic evolution of this terrane. Most massifs preserve fragments of a basaltic section with a thin sequence of overlying sedimentary strata. Marine siliceous and fine-clastic deposits that cover the ophiolite mostly crop out along the northern margin of the Dazhuqu terrane. These deposits are referred to as the Chongdu Formation (R. L. Cao 1981, cited by Bureau of Geology and Mineral Resources of Xizang Autonomous Region 1993). Sections of pillow basalt overlain by a few metres of chert then volcaniclastic sediments are best exposed in the Qunrang district. Other sections are known from south of Tiding at Donglha (two sections; Fig. 4c), Sagui, Beimarang, Bainang, Dazhuqhu and Polio. In most cases, the pseudo-stratigraphies of ophiolitic sections are north-facing although south- and westfacing sections occur at Donglha and Bainang, respectively. Early investigations of the Chongdu Formation presented preliminary age data

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES

155

Fig. 4. (a) Ophiolitic rocks exposed at Dangxiong near the headwaters of the Yarlung Tsangpo, western Tibet. Photograph taken from 30°20'N, 082°55'E looking eastwards. Ultramafic rocks crop out in the low hills on the far side of the river flats, (b) Steeply east-dipping dykes exposed in upper-crustal levels of the Dazhuqu ophiolite SE of Qunrang. Photograph taken from 29°09'N, 089°01'E looking eastwards, (c) An overturned sequence of pillow basalts overlain by a thin (<20 m) sequence of uppermost Barremian to upper Aptian ribbon-bedded radiolarian cherts exposed at Donglha (29°08'N, 088°25'E) along a canal located SSE of Tiding, (d) Rocks of the Dazhuqu ophiolite observed looking westwards from a garnet amphibolite knocker locality (29°09'N, 089°20'E) in serpentinite-matrix melange at the base (south) of the ophiolite. Serpentinite-matrix melange lies in the foreground; rocks on the ridge on the left of the photo are dominated by cherts of the Bainang terrane. The centre right of the photo is dominated by serpentinized harzburgites and the mountain over the river valley in the distance is mostly a north-facing section of diabase and gabbro. (e) Serpentinized harzburgites of the Dazhuqu terrane (viewed looking west), which form the high (5000+ m) table-topped mountain, have been thrust southwards over the northern margin of the Indian terrane near Renbung. Thick beds within the Dazhuqu conglomerate crop out on the hill slopes on the sides of the Yarlung Tsangpo in the foreground. (f) View looking southwards towards the Purang valley from between Mapham Tso (Lake Manasarovar) and Lhanak Tso (Lake Raksas) south of Gangrinboche (Mt Kailas, 6714 m). The Jungbwa ophiolite crops out on the hills to the SW of Lake Raksas. (g) View from the Luobusa Chromite mine looking westwards up the Yarlung Tsangpo. Outcrops of the Lhasa terrane and Gangdese batholith lie in the distance beyond the ophiolite. (h) Northward-dipping tectonic slices of basalt, pillow breccia and basaltic hyaloclastite overlain by red ribbonbedded chert crop out on the southeastern side of ultramafic massifs near Dangxiong (30°09'N, 083°14'E).

156

J. C. AITCHISON ETAL.

(Marcoux et al. 1982) and suggested that the ophiolite had formed in the Cenomanian. On the basis that the Xigaze ophiolite sections are thinner than those reported from other classic ophiolites, such as those of Cyprus or Oman, the Sino-French group considered the ophiolite to be somewhat different (Nicolas et al. 1981). They suggested that it had formed at a slow-spreading ocean ridge located proximal to the margin of the southern Lhasa block (Pozzi et al. 1984). The presence of diabase dykes throughout the section suggests primary structural attenuation of the ophiolite in response to tectonic extension at the ridge system where it was generated. Further section removal probably occurred during tectonic emplacement onto India or later India-Asia collision. More recent workers have re-examined the geochemistry of some of these rocks (Hebert et al. 2000, 2001) and undertaken detailed investigations of the radiolarian biostratigraphy (Zyabrev et al. 1999; Ziabrev 2001; Ziabrev et al. 2003b) and palaeomagnetism (Abrajevitch et al. 2001, 2002, 2003) of sediments immediately overlying the ophiolite. The northern margin of the ophiolite lies in the hanging wall of a steep north-directed Miocene 'back'-thrust. A tectonic contact with a footwall of fore-arc basin turbidites of the Xigaze terrane is ubiquitous (Aitchison et al. 2003). Although the ophiolite is typically northfacing with its uppermost levels exposed near the contact, no depositional continuity with the Xigaze terrane can be demonstrated. Earlier suggestions of original depositional continuity between Dazhuqu and Xigaze terranes can be dismissed. The Dazhuqu terrane consists of numerous discrete and seemingly unrelated massifs. Individual ophiolitic massifs preserve discrete structural levels in different proportions. Palaeomagnetic data indicate that massifs had origins at different locations and have experienced varying degrees of tectonic rotation (Abrajevitch et al. 2001, 2002, 2003). The purportedly overlying Xigaze terrane is, on the other hand, structurally coherent and contains bedding packages, which can be traced along strike for several tens of kilometres. The southern margin of the ophiolite is typically dominated by ultramafic rocks with development of a wide (up to several hundreds of metres) zone of serpentinite-matrix melange. Locally this melange contains rare knockers, which have experienced amphibolite-facies metamorphism (Fig. 4d). In the Bainang district, an S-shaped sigmoidal bend in the YTSZ preserves earlier contacts with the Bainang terrane that predate later northdirected back-thrusting. Ophiolitic rocks and ser-

pentinite-matrix melange lie in the hanging wall of a south-directed thrust that is locally truncated by strike-slip faults (Girardeau et al. 1985b; Ratschbacher et al. 1994). Together with rocks of the Bainang terrane the ophiolite was emplaced southwards onto the northern edge of India before the India-Asia collision. Sedimentary rocks immediately overlying the ophiolite have been the subject of recent investigations by a research team from the University of Hong Kong, which show that they differ even between nearby localities (Ziabrev et al. 2003b). Overall, the sedimentary sections exhibit a coarsening-upward trend from chert and siliceous mudstones intercalated with, and immediately above, pillow basalts. Up-section sedimentary rocks become predominantly volcaniclastic and include basaltic agglomerates, felsic tuffs, and volcanolithic turbidites indicating an island-arc depositional environment. Detailed investigations of well-preserved, abundant radiolarians (Fig. 5) provide high-precision biostratigraphic age constraints (Fig. 6) on the timing of the eruption of ophiolitic basalts in this region (Zyabrev et al. 1999; Ziabrev 2001; Ziabrev et al. 2003b). These results collectively show that ophiolitic rocks in the Xigaze district were generated in an intraoceanic suprasubduction zone setting within a relatively short (c. 6 Ma) interval between the late Barremian (c. 123 Ma) and mid-Aptian (c. 117 Ma). Accumulation of sediments upon the newly generated ophiolite initially occurred in a series of discrete rift-controlled sub-basins associated with various spreading centres. An increasing flux of arc-derived volcaniclastic sediment up-section indicates activity in a nearby volcanic arc. The Dazhuqu terrane appears to have developed in an intra-oceanic setting within Tethys where it was isolated from any continental influence. Application of high-precision radiolarian biostratigraphy in constraining the ages of basalts and overlying sediments in the Dazhuqu terrane has made it possible to reinvestigate the palaeomagnetism of these rocks so as to better constrain where they might have formed. Results show that the ophiolite formed at equatorial to very low northern latitudes (Abrajevitch et al. 2001, 2002, 2003). Some fragments of the ophiolite have clearly experienced a counter-clockwise rotation whereas sedimentary successions within the Xigaze terrane, previously reported to conformably overlie the ophiolite, are coherent over many kilometres of their strike length. All sections from which conformable contacts with the Xigaze terrane had been reported have been re-examined and north-directed thrusting is ubiquitous between these two terranes.

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES

157

Fig. 5. Some representative mid-Cretaceous (lowermost upper Aptian UA8) radiolarians from siliceous mudstones intercalated with volcaniclastic sediments, which overlie pillow basalts of the Dazhuqu terrane SE of Qunrang (29°08'N, 089°02'E) (scale bars represent 100 |Lirn). (a) Acaeniotyle umbilicata(Rusi)', (b) A. diaphorogona Foreman; (c) Crucella hispana O'Dogherty; (d) Hiscocapsa grutterinki (Tan); (e) Pseudodictyomitra hornatissima (Squinabol); (f) Pseudodictyomitra pentacolaensis Pessagno; (g) Stichomitra communis Squinabol; (h) Thanarla brouweri (Tan); (i) Turbocapsula costata (Wu); (j) Xitus clava (Parona).

Dazhuqu terrane: Ngamring Tso-Saga-

Zongba —Jungbwa Ophiolitic rocks crop out more or less continuously in a westerly direction from Ngamring Tso to Saga. The general distribution of these rocks has been mapped (Zhang & Fu 1982; Burg 1983) but they have not been the subject of any detailed investigation. Outcrop occurs mostly as ophiolitic melange although a large ultramafic massif is present a few kilometres south of the river crossing at Sangsang (29°24'N, 086°43'E). From Ngamring west, outcrop lies north of the Yarlung Tsangpo but to the south of the main road leading to Saga (29°20'N, 085°25'E). Just east of the turnoff for the northern road to Tsochen and western Tibet the main road crosses to the south of the ophiolite. Outcrop is mostly obscured by Quaternary alluvium in a swampy area. West of the turnoff, ophiolitic rocks, mostly diabases and basalts, are exposed on the uppermost southern slopes of a large peak (6000+ m) and outcrop continues further westwards on high peaks. From Saga

westwards to beyond Zongba (29°45'N, 083°55'E), limited occurrences of ophiolitic rocks occur well to the north of the main (only) road and these rocks have not been studied. West of Zongba, a settlement in the large graben leading northwards from Mustang in Nepal, a significant occurrence of ophiolitic rocks is encountered near Dangxiong (30°09'N, 083°14'E). The ophiolite here is dominated by ultramafic massifs (Fig. 4a), which include Cr-bearing dunite bodies. This mineralization has been explored in detail by geologists of the Bureau of Geology and Mineral Resources of the Xizang Autonomous Region. The ophiolitic rocks have been emplaced southwards onto rocks of Indian affinity and lie several kilometres south of the suture zone. Numerous, south-vergent, thin (decametre-scale) tectonic slices of northward-dipping basalt, pillow breccia and basaltic hyaloclastite overlain by red ribbon-bedded chert crop out on the SE side of the ultramafic massifs (Fig. 4h). Although we have examined these rocks for radiolarian faunas, no identifiable material has been recovered so far from acid residues.

158

J.C.AITCHISON ET AL.

Fig. 6. Radiolarian-based ages of supra-ophiolitic sedimentary sections associated with the Dazhuqu terrane ophiolite. Unitary associations (UA) after Baumgartner et al. (1995) and O'Dogherty (1994). Chronological scale after Gradstein et al. (1994).

Further westwards, ophiolitic rocks are again encountered when Mapham Tso (Lake Manasarovar) is reached south of Gangrinboche (Mt Kailas, 6714m, 31°04'N, 081°19'E). The ophiolitic rocks here are referred to as the Jungbwa (Yungbwa) ophiolite (Gansser 1964; Searle et al. 1987; Murphy et al. 2002; Miller et al. 2003; Murphy & Yin 2003). Outcrop occurs mostly to the south and west of Lhanak Tso (Lake Raksas, Fig. 4f), where it is dominated by a large (>3500 km2) ultramafic massif. Harzburgite and cpx-poor Iherzolite are the dominant lithologies and isotopic data for these rocks (Miller et al. 2003) suggest that peridotite melting occurred during the Early Jurassic. Ultramafic rocks in this region are locally cut by pegmatitic gabbronorite and rare normal mid-ocean ridge basalt (N-MORB) type tholeiitic basalt dykes for which Miller et al. (2003) reported a Late Jurassic 40Ar/39Ar age of 152 ±33 Ma. Additional outcrops of ultramafic rocks occur further NW in the Kiogar region, where no reports of the upper levels of any ophiolitic succession exist and the most detailed descriptions of these rocks are from reconnaissance studies only (Gansser 1964). All the ophiolitic rocks lie in the hanging wall of southdirected thrusts, which transported them over rocks of Indian affinity (Gansser 1964). The Jungbwa ophiolite has been thrust 30-40 km south of the YTSZ (sensu stricto) and lies atop passive margin sediments of the Indian terrane (Murphy & Yin, 2003). The Kiogar ophiolite has

been thrust a similar distance southwards over the Indian terrane and development of extensive mudmatrix melange (Gansser 1964) may have accompanied its emplacement. The northern side of the YTSZ (sensu stricto) in the Kailas region is marked by the north-directed South Kailas thrust along which zones of sheared serpentinite and ophiolitic melange occur (Yin et al. 1999).

Dazhuqu terrane: Renbung-Quxu Outcrops of ophiolitic rocks, which occur as blocks in a serpentinite-matrix melange, can be traced eastwards intermittently from Renbung to west of Lhasa airport. The largest zone of outcrop is located on the SW side of the Yarlung Tsangpo to the south of a 6126 m peak on which Miocene Gangrinboche facies conglomerates lie unconformably upon Lhasa terrane rocks (Aitchison et al. 2002b). South of Quxu, serpentinite-matrix melange with large blocks of ultramafic rocks crops out together with diabase dykes and basalts where the YTSZ passes through a narrow col (Jiangdanyako, 29°18'N, 090°43'E) between Lhasa terrane granites and the Indian terrane. The ophiolitic rocks lie within shear zones associated with the north-directed Renbu-Zedong thrust system.

Ophiolitic rocks: Zedong-Luobusa A further major occurrence of ophiolitic rocks along the YTSZ crops out to the SE of Lhasa near

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES Zedong and Luobusa. The largest chromite deposit (Fig. 4g) in China is currently being worked at Luobusa and rocks there have been the subject of numerous investigations (Badengzhu 1979, 1981; Huang et al. 1981; Bai et al. 1993, 2000; Zhou & Robinson 1994; Zhou 1995; Zhou et al, 1996, 2002; Griselin et al 1999; Hu 1999; Hebert et al. 2000, 2001). The Luobusa ophiolite is dominated by a harzburgitic mantle section that has experienced Late Miocene India-Asia collision-related north-directed thrusting over a dunite transition zone and ophiolitic melange (Zhou et al. 1996). Lenses and pods of dunite are abundant within the ultramafic section and these rocks contain the chromite mineralization. Restricted occurrences of gabbros are also known but no diabase dykes have been reported from Luobusa. Together, the entire package has been thrust northwards over Lower Miocene Luobusa conglomerates (Gangrinboche facies) on the southern margin of the Lhasa terrane (Aitchison et al. 2002b) and is itself overthrust by Triassic Indian terrane rocks. The original structural relationship between the ophiolitic rocks and the Indian terrane is not preserved. The geochemistry and petrology of the Luobusa ophiolite are distinctive. Spinels from the Luobusa massifs have Cr numbers as high as 0.69-0.94 and olivines are more forsteritic than elsewhere along the YTSZ (Wang et al. 1999; Hebert et al. 2000, 2001). Such values are similar to those for rocks produced from boninitic melts and are considerably more refractory than for rocks from elsewhere along the suture. One of the most unusual features of the Luobusa ophiolite is the reported occurrence of small amounts of microdiamonds, graphite, SiC and other rare minerals found in association with chromitites (Bai et al. 1993, 2000; Hu 1999). Ophiolitic rocks at Luobusa and Zedong are intimately associated with island arc tholeiitic rocks of the Zedong terrane. At Luobusa, this terrane is limited to a small zone of overturned variolitic pillow basalts, which are stratigraphically overlain by red mudstones. Outcrop is much more extensive near Zedong, where the terrane consists of a sequence in which island arc tholeiitic pillow basalts are overlain by a thin succession of red radiolarian cherts, which are themselves overlain by an up to 1 km thick pile of shoshonitic autoclastic breccias (McDermid 2002). The exact nature of the original relationship between these two terranes remains indeterminate, as all contacts are faulted. Nevertheless, limited age data from the Luobusa ophiolite (Sm-Nd age of 177 ± 31 Ma on gabbro dykes given by Zhou et al. 2002) are similar to U-Pb, Ar-Ar and fossilbased ages for rocks of the Zedong terrane (McDermid 2002; McDermid et al 2002). The

159

relationship of ophiolitic rocks in the ZedongLuobusa district to other ophiolitic rocks along the suture remains ambiguous. Although ophiolitic rocks in the Zedong to Luobusa area lie in an identical structural position along the YTSZ, both the compositions (Wang et al 1999; Hebert et al 2000, 2001) and ages of these rocks (Ziabrev 2001; McDermid 2002; McDermid et al 2002; Ziabrev et al 2003b) differ from those of other ophiolitic rocks of the YTSZ. Further work is clearly required to resolve this enigma. We note reports of similar ages for the oldest ophiolitic rocks of the Spontang ophiolite in Ladakh, NW India (Pedersen et al 2001), and suggest that it is entirely possible that remnants of more than one intra-oceanic island arc are preserved between India and Asia. Emplacement

Kinematic indicators along the southern margin of the Dazhuqu terrane ophiolites near Bainang, SE of Xigaze, indicate its southward emplacement onto the feather-edge of the northern Indian continent (Girardeau et al 1985c; Ratschbacher et al. 1994). Ophiolites further west along the suture have similarly been emplaced southwards onto the northern Indian passive margin. Precise constraints on the timing of this emplacement event have proved difficult to obtain with seemingly contradictory datasets. Traditionally the ages of amphibolite-facies metamorphic rocks within melange zones at the base of ophiolitic successions have been interpreted as indicative of emplacement events. In this case, Late Cretaceous ages for amphibolites from near Bainang (Wang et al. 1987) would seemingly indicate obduction at that time. A tectonic event of this nature, however, seems inconsistent with any events recorded in regional sedimentary successions, which should reflect such an event. An alternative interpretation of the Late Cretaceous metamorphic ages known from various localities along the suture within Tibet (Zhou cited by Aitchison et al 2000) and further afield in NW India and Pakistan (Searle et al 1999) is that they represent an ophiolitespecific thermal event such as the subduction of a spreading centre (Shervais 2001). The timing of emplacement of ophiolitic rocks onto the northern margin of India is perhaps better constrained from the sedimentary record within, and south of, the suture. Distal Indian passive margin sediments in the northern Tethyan Himalayan zone, north of Qomolangma (Mt. Everest) record a significant tectonic disturbance and the development of a major disconformity around the CretaceousPaleogene boundary (Wan et al 2002) before resumption of normal carbonate accumulation in a

160

J. C. AITCHISON £T,4L.

quiescent environment until the time of IndiaAsia collision. Along the suture itself, ophiolitic rocks as well as strata of the Indian and Bainang terranes are unconformably overlain by Paleogene Liuqu conglomerates. These coarse clastic sediments are interpreted to have accumulated during the collision of the ophiolitic terrane and the Indian continental margin, which lay to its south (Davis et al. 2002). Other Lower Miocene conglomerates along the suture contain clasts from all nearby terranes, including those north of the suture, and constrain the timing of collision between India and Asia (Aitchison et al. 2002b). Regional development of mud-matrix melange was extensive along the northern Indian margin in the Paleogene (Gansser 1964; Shackleton 1981; Liu & Einsele 1996; Liu 2001). These melanges possibly developed in response to ophiolite obduction and are analogous to similar rocks in contemporary arc-continent collision zones such as Timor (Barber et al. 1986) and Taiwan (Chang et al. 2000).

Discussion As a result of detailed investigations along the YTSZ over the past few years, it is now possible to provide a better understanding of ophiolites in the region. Improved precision for Upper Mesozoic radiolarian biostratigraphy (O'Dogherty 1994; Baumgartner et al. 1995) allows more precise age constraints to be placed on any interpretations of the development of this region. Radiolarian ages from sediments immediately overlying the ophiolite constrain the timing and duration of the ophiolite generation event. Biostratigraphic investigations of numerous sections indicate that the Dazhuqu terrane ophiolite formed in the Barremian (mid-Cretaceous) with accumulation of volcaniclastic sediments continuing into the Aptian (Zyabrev et al. 1999; Ziabrev 2001; Ziabrev et al. 2003b). Results of palaeomagnetic investigations indicate generation of the Dazhuqu terrane ophiolite at equatorial to low northern latitudes at least 1000-1500 km south of Asia's margin (Abrajevitch et al. 2001, 2003; Abrajevitch jevitch 2002). Some fragments of the ophiolite have experienced a counter-clockwise rotation and individual ophiolitic massifs appear to have different travel histories. Although many workers had previously inferred depositional continuity between ophiolitic rocks of the Dazhuqu terrane and volcaniclastic turbidites of the Xigaze terrane, detailed investigations of all known localities where this relationship has been inferred unequivocally reveal that this contact is tectonic. Sedimentary rocks overlying the ophiolite are volcaniclastic, as are those in the Xigaze terrane,

but the sources of the two suites of rocks were different. Furthermore, the structural evolution of both terranes was remarkably different and there is no evidence that these two entities shared a common history before India-Asia collision. Present interpretation of the Dazhuqu terrane ophiolite as having an origin within a suprasubduction zone setting is supported by detailed mineralogical and geochemical studies in the Xigaze area (Wang et al. 1987; Hebert et al. 2000, 2001). The close spatial association of these rocks with the subduction complex assemblage of the Bainang terrane suggests the existence of a southfacing intra-oceanic island arc at near equatorial latitude within Neotethys during the mid-Cretaceous (Aitchison et al. 2000). As more data become available from along the YTSZ, interpretation of this zone becomes increasingly complicated. Ages for ophiolitic assemblages in essentially the same structural positions at Xigaze, Jungwa and Zedong are distinctly different. It now seems that the YTSZ contains remnants of at least one, and possibly two, intra-oceanic island arc systems. Given the complexity of analogous modern settings such as the western Pacific, it is clear that early interpretations were oversimplified. Multiple arcs may have coexisted in a wide Tethys with no single simple convergent margin. It is also probable that plate boundaries were not necessarily parallel to one another. The suture extends from one side of Tibet to the other and beyond. Despite a few detailed studies of isolated occurrences, the Yarlung Tsangpo ophiolites have received remarkably little attention relative to their significance in interpretation of the evolution of Tethys and the India-Asia collision. Continuing Tibet research at the University of Hong Kong is supported by grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project Nos. HKU7102/98P, 7299/99P and 7069/0IP). We thank colleagues in the Geological Society of Tibet for their assistance in arranging the logistics related to this research. The authors gratefully acknowledge the constructive reviews of Y. Dilek and P. Robinson, which helped to improve the manuscript.

References ABRAJEVITCH, A. 2002. Paleomagnetism of the Dazhuqu Terrane, Yarlung Zangbo Suture Zone, Southern Tibet. MPhil thesis, University of Hong Kong. ABRAJEVITCH, A., AITCHISON, J.C. & ALI, J.R. 2001. Paleomagnetism of the Dazhuqu Terrane, Yarlung Zangbo Suture Zone, Southern Tibet. EOS Transactions, American Geophysical Union Fall Meeting Supplement, GP11A-0188. ABRAJEVITCH, A., AITCHISON, J.C. & ALI, J.R. 2002. Paleomagnetism of the Dazhuqu Terrane, Yarlung Zangbo Suture Zone, Southern Tibet. In: Tectonic

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES Processes in the Evolution of China, Croucher Advanced Studies Institute Abstracts. University of Hong Kong, Hong Kong. ABRAJEVITCH, A., AITCHISON, J.C., ALI, J.R., BADENGZHU,, DAVIS, A.M., Liu, J.B. & ZIABREV, S.V. in press. Neotethys and the India-Eurasia collision— new insights following a palaeomagnetic investigation of the Dazhuqu Terrane (southern Tibet). Geophysical Journal International. AITCHISON, J.C. & DAVIS, A.M. 2001. Orogenic conglomerates indicate timing of collision in Tibet. Journal of Asian Earth Sciences, 19(3 A), 1-2. AITCHISON, J.C., BADENGZHU, & DAVIS, A.M. ET AL. 2000. Remnants of a Cretaceous intra-oceanic subduction system within the Yarlung-Zangbo suture (southern Tibet). Earth and Planetary Science Letters, 183, 231-244. AITCHISON, J.C., ABRAJEVITCH, A. & ALI, J.R. ET AL. 2002a. New insights into the evolution of the Yarlung Tsangpo suture zone, Xizang (Tibet), China. Episodes, 25(3), 90-94. AITCHISON, J.C., DAVIS, A.M., BADENGZHU, & Luo, H. 2002b. New constraints on the India-Asia collision: The Lower Miocene Gangrinboche conglomerates, Yarlung Tsangpo suture zone, SE Tibet. Journal of Asian Earth Sciences, 21(3), 253-265. AITCHISON, J.C., DAVIS, A.M., BADENGZHU, & Luo, H. 2003. The Gangdese thrust: a phantom structure that did not raise Tibet. Terra Nova, 15, 155-162. ALLEGRE, C.J., COURTILLOT, V. & TAPPONNIER, P. ET AL. 1984. Structure and evolution of the Himalaya-Tibet erogenic belt. Nature, 307(5946), 17-22. BADENGZHU (Compiler) 1979. Xizang Autonomous Region Zhanang—Sangri Regional Geology Reconnaissance Map 1:50,000. Xizang Geological Survey Geological Team No. 2, Group 2, Lhasa. BADENGZHU (Compiler) 1981. Xizang Autonomous Region Sangri—Jiacha Regional Geology Reconnaissance Map 1:50,000. Xizang Geological Survey Geological Team No. 2, Lhasa. BAI, W., ZHOU, M.F. & ROBINSON, P.T. 1993. Possibly diamond-bearing mantle peridotites and podiform chromitites in the Luobusa and Donqiao ophiolites, Tibet. Canadian Journal of Earth Sciences, 30(8), 1650-1659. BAI, W., ROBINSON, P.T. & FANG, Q. ET AL. 2000. The PGE and base-metal alloys in the podiform chromitites of the Luobusa ophiolite, southern Tibet. Canadian Mineralogist, 38(3), 585-598. BARBER, A.J., TJOKROSAPOETRO, S. & CHARLTON, T.R. 1986. Mud volcanoes, shale diapirs, wrench faults, and melanges in accretionary complexes, eastern Indonesia. AAPG Bulletin, 70(11), 1729-1741. BASSOULLET, J.P., COLCHEN, M., MASCLE, G., MATTE, P. & WANG, N.W. 1984. Les ensembles sedimentaires de la zone du Tsangpo (Lhaze, Lhasa, Linzhu). In: MERCIER, J.L. & Li, G.C. (eds) Mission Franco-Chinoise au Tibet 1980. Etude geologique et geophysique de la croute terrestre et du manteau superieur du Tibet et de I'Himalaya. CNRS, Paris, 133-153. BAUMGARTNER, P.O., O'DOGHERTY, L., GORICAN, S., URQUHART, E., PILLEVUIT, A. & DE WEVER, P.

161

(eds) 1995. Middle Jurassic to Lower Cretaceous Radiolaria of Tethys: Occurrence, Systematics, Biochronology. Memoires de Geologic, Universite de Lausanne, 23. Bureau of Geology and Mineral Resources of Xizang Autonomous Region 1993. Regional Geology of Xizang (Tibet) Autonomous Region. Peoples Republic of China, Ministry of Geology and Mineral Resources Geological Memoirs Series, Geological Publishing House, Beijing, 1, 31. Bureau of Geology and Mineral Resources of Xizang Autonomous Region Geological Team No. 2. 1979. Regional Geological Report for Zhanang-Sangri County, Xizang (Tibet) Autonomous Region, 10, 188. BURG, J.P. 1983. Tectogenese comparee de deux segments de chaine de collision:—Le Sud du Tibet (Suture du Tsangpo)—La Chaine Hercynienne en Europe (sutures du Massif Central). Doctorat d'Etat thesis, Universite des Sciences et Techniques du Languedoc, Montpellier. BURG, J.P. & CHEN, G.M. 1984. Tectonics and structural zonation of southern Tibet, China. Nature, 311(5983), 219-223. CHANG, C.P., ANGELIER, J. & HUANG, C.Y. 2000. Origin and evolution of a melange: the active plate boundary and suture zone of the Longitudinal Valley, Taiwan. Tectonophysics, 325, 43-62. COULON, C, MALUSKI, H., BOLLINGER, C. & WANG, S. 1986. Mesozoic and Cenozoic volcanic rocks from central and southern Tibet: 39Ar-40Ar dating, petrological characteristics and geodynamic significance. Earth and Planetary Science Letters, 79, 281-302. DANELIAN, T. & ROBERTSON, A.H.F. 1997. Radiolarian evidence for the stratigraphy and palace-oceanography of the deep-water passive margin of the Indian Plate (Karamba Formation, Indus suture zone, Ladakh Himalaya). Marine Micropaleontology, 30(1-3), 171-195. DAVIS, A.M., AITCHISON, J.C., BADENGZHU,, Luo, H., MALPAS, J. & ZYABREV, S. 1999. Eocene obliqueslip basin development, Tibet: terrane tracks on the roof of the world. In: EVENCHICK, C.A., WOODSWORTH, GJ. & JONGENS, R. (eds) Terrane Paths 99. Circum Pacific Terrane Conference, Okanagan Valley, B.C., Canada. Circum Pacific Terrane Conference Abstracts and Program, 28. DAVIS, A.M., AITCHISON, J.C., BADENGZHU, & Liu, H. 2001. Late Cretaceous-Paleocene island arc collision related conglomerates, Yarlung-Tsangpo suture zone, Tibet. Journal of Asian Earth Sciences, 19(3 A), 13. DAVIS, A.M., AITCHISON, J.C., BADENGZHU,, Luo, H. & ZYABREV, S. 2002. Paleogene island arc collisionrelated conglomerates, Yarlung-Tsangpo suture zone, Tibet. Sedimentary Geology, 150(3/4), 247273. DURR, S.B. 1996. Provenance of Xigaze fore-arc basin clastic rocks (Cretaceous, South Tibet). Geological Society of America Bulletin, 108(6), 669-684. EINSELE, G., Liu, B. & DURR, S. ET AL. 1994. The Xigaze forearc basin; evolution and facies architecture (Cretaceous, Tibet). Sedimentary Geology, 90,

162

J. C. AITCHISON £T,41.

1-32. GAETANI, M. & GARZANTI, E. 1991. Multicyclic history of the northern India continental margin (Northwestern Himalaya). AAPG Bulletin, 75(9), 14271446. GANSSER, A. 1964. The Geology of the Himalayas. Wiley-Interscience, New York. GIRARDEAU, J. & MERCIER, J.C.C. 1988. Petrology and texture of the ultramafic rocks of the Xigaze Ophiolite (Tibet); constraints for mantle structure beneath slow-spreading ridges. Tectonophysics, 147(1-2), 33-58. GIRARDEAU, J., MARCOUX, J. & ALLEGRE, CJ. ET AL. 1984a. Tectonic environment and geodynamic significance of the Neo-Cimmerian Donqiao Ophiolite, Bangong-Nujiang suture zone, Tibet. Nature, 307(5946), 27-31. GIRARDEAU, J., MARCOUX, J. & ZAO, Y. 1984b. Lithologic and tectonic environment of the Xigaze Ophiolite (Yarlung Zangbo suture zone, southern Tibet, China), and kinematics of its emplacement. Eclogae Geologicae Helvetiae, 77(1), 153-170. GIRARDEAU, J., MERCIER, J.C. & XIBIN, W. 1985a. Petrology of the mafic rocks of the Xigaze ophiolite, Tibet; implications for the genesis of the oceanic lithosphere. Contributions to Mineralogy and Petrology, 90(4), 309-321. GIRARDEAU, J., MERCIER, J.C.C. & ZAO, Y.G. 1985b. Origin of the Xigaze Ophiolite, Yarlung Zangbo suture zone, southern Tibet. Tectonophysics, 119, 407-433. GIRARDEAU, J., MERCIER, J.C.C. & ZAO, Y.G. 1985c. Structure of the Xigaze Ophiolite, Yarlung Zangbo suture zone, southern Tibet, China; genetic implications. Tectonics, 4(3), 267-788. GIRARDEAU, J., NICOLAS, A. & MARCOUX, J. ET AL. 1984c. Les ophiolites de Xigaze et la suture du Yarlung Zangbo (Tibet). In: MERCIER, J.L. & Li, G.C. (eds) Mission Franco-Chinoise au Tibet 1980. Etude geologique et geophysique de la croute terrestre et du manteau superieur du Tibet et de I'Himalaya. CNRS, Paris, 189-197. GRADSTEIN, P.M., AGTERBERG, P.P., OGG, J.G., HARDENBOL, J., VAN VEEN, P., THIERRY, J. & HUANG, Z.

1994. A Mesozoic time scale. Journal of Geophysical Research, 99B, 24051-24074. GRISELIN, M., DA VIES, G.R. & PEARSON, D.G. 1999. REE and Sr-Pb-Nd isotope geochemistry of Tibetan peridotites: implications for melting processes. Ofioliti,24(lA), 102-103. HARRISON, T.M., YIN, A., GROVE, M., LOVERA, O.M., RYERSON, FJ. & ZHOU, X. 2000. The Zedong Window: a record of superposed Tertiary convergence in southeastern Tibet. Journal of Geophysical Research, 105(B8), 19211-19230. HEBERT, R., BEAUDOIN, G., VARFALVY, V., HUOT, P., WANG, C.S. & Liu, Z.F. 2000. Yarlung Zangbo ophiolites, southern Tibet revisited. Earth Science Frontiers, 1, 124-126. HEBERT, R., WANG, C., VARFALVY, V., HUOT, P., Liu, Z., BEAUDOIN, G. & DOSTAL, J. 2001. Yarlung Zangbo Suture Zone ophiolites and their suprasubduction zone setting. Journal of Asian Earth Sciences, 19(3A), 27-28.

Hu, X.F. 1999. Origin of diamonds in chromitites of the Luobusa Ophiolite, southern Tibet, China. MSc thesis, Dalhousie University, Halifax, N.S. HUANG, Y.H., ZHON, X.z. & WANG, X.B. 1981. Pumpellyite from the contact metamorphic zone of ultramafic bodies of Luobusha Ophiolite suite, South Tibet. Bulletin de Mineralogie, 104(4), 441-444. JADOUL, P., BERRA, P. & GARZANTI, E. 1998. The Tethys Himalayan passive margin from Late Triassic to Early Cretaceous (South Tibet). Journal of Asian Earth Sciences, 16, 173-194. Li, H.S. & Wu, H.R. 1985. Radiolaria from the Cretaceous Congdu Formation in southern Xizang (Tibet). Acta Micropalaeontologica Sinica, 2(1), 61-76. Liu, G. 1992. Permian to Eocene sediments and Indian passive margin evolution in the Tibetan Himalayas. Tubinger Geowissenschaftliche Arbeiten, 13, 1268. Liu, G.H. & EINSELE, G. 1996. Various types of olistostromes in a closing ocean basin, Tethyan Himalaya (Cretaceous, Tibet). Sedimentary Geology, 104, 203-226. Liu, J.B. 2001. Yamdrok Melange, Gyantze district, Xizang (Tibet), China. MPhil thesis, University of Hong Kong. Liu, J.B. & AITCHISON, J.C. 2002. Upper Paleocene radiolarians from the Yamdrok melange, south Xizang (Tibet), China. Micropaleontology, 48(Supplement 1), 145-154. Liu, J.B., AITCHISON, J.C., BADENGZHU,, DAVIS, A.M. & ZIABREV, S.V. 2002. Origin and tectonic significance of the Yamdrok Melange, South Xizang (Tibet), China. In: Tectonic Processes in the Evolution of China, Croucher Advanced Studies Institute Abstracts. University of Hong Kong, Hong Kong. MALUSKI, H., PROUST, P. & XIAO, X.C. 1982. 39Ar/40Ar dating of the trans-Himalayan calc-alkaline magmatism of southern Tibet. Nature, 298(5870), 152154. MARCOUX, J., DE WEVER, P. & NICOLAS, A. ET AL. 1982. Preliminary report of depositional sediments on top of volcanic member; Xigaze Ophiolite (Yarlung Zangbo suture zone); South Xingang (Tibet). Ofioliti, 2(3), 395-396. MATSUOKA, A., KOBAYASHI, K., NAGAHASHI, T., YANG, Q., WANG, Y.J., TAKEI, M. & ZENG, Q.G. 2001a. Early Middle Jurassic (Aalenian) radiolarian fauna from the Xialu chert in the Yarlung Zangbo Suture Zone, southern Tibet. In: METCALFE, I., SMITH, J.M.B., MORWOOD, M. & DAVIDSON, I. (eds) Faunal and Floral Migrations and Evolution in SE Asia-Australasia. Balkema, Lisse, 105-110. MATSUOKA, A., YANG, Q., KOBAYASHI, K., TAKEI, M., NAGAHASHI, T., ZENG, Q.G. & WANG, YJ. 2001b. Jurassic-Cretaceous radiolarian biostratigraphy and sedimentary environments of the Ceno-Tethys: records from the Xialu Chert in the Yarlung-Zangbo Suture Zone, southern Tibet. Journal of Asian Earth Sciences, 20(3), 277-287. MCDERMID, I., AITCHISON, J.C., BADENGZHU, & DAVIS, A.M. 200la. The Zedong Terrane: an intra-oceanic magmatic arc assemblage, Tibet. Journal of Asian Earth Sciences, 19(3A), 44.

YARLUNG TSANGPO SUTURE ZONE OPHIOLITES MCDERMID, I., AITCHISON, J.C., DAVIS, A.M., HARRISON, T.M. & GROVE, M. 200Ib. The Zedong Terrane: a Jurassic intra-oceanic magmatic arc within the Yarlung-Zangbo Suture Zone of Southeastern Tibet. Journal of Asian Earth Sciences, 19(3A), 44-45. MCDERMID, I.R.C. 2002. Zedong terrane, south Tibet. PhD thesis, University of Hong Kong. MCDERMID, I.R.C., AITCHISON, J.C., DAVIS, A.M., HARRISON, T.M. & GROVE, M. 2002. The Zedong terrane: a Late Jurassic intra-oceanic magmatic arc within the Yarlung-Zangbo suture zone, southeastern Tibet. Chemical Geology, 187(3-4), 267-277. MILLER, C, THONI, M., FRANK, W., SCHUSTER, R., MELCHER, F., MEISEL, T. & ZANETTI, A. 2003. Geochemistry and tectonomagmatic affinity of the Yungbwa ophiolite, SW Tibet. Lithos, 66(3-4), 155-172. MURPHY, M.A. & YIN, A. 2003. Structural evolution and sequence of thrusting in the Tethyan foldthrust belt and Indus-Yalu suture zone, southwest Tibet. Geological Society of America Bulletin, 115(1), 21-34. MURPHY, M.A., YIN, A. & HARRISON, T.M. ET AL. 1997. Did the Indo-Asian collision alone create the Tibetan plateau? Geology, 25, 719-722. MURPHY, M.A., YIN, A. & KAPP, P. ET AL. 2002. Structural evolution of the Gurla Mandhata detachment system, southwest Tibet: implications for the eastward extent of the Karakoram fault system. Geological Society of America Bulletin, 114(4), 428-447. NICOLAS, A., GIRARDEAU, J. & MARCOUX, J. ET AL. 1981. The Xigaze ophiolite (Tibet); a peculiar oceanic lithosphere. Nature, 294(5840), 414-417. O'DOGHERTY, L. 1994. Biochronology and Paleontology of Mid-Cretaceous Radiolarians from Northern Apennines (Italy) and Betic Cordillera (Spain). Memoires de Geologic, Universite de Lausanne, 21, 1-415. PEARCE, J.A. & DENG, W.M. 1988. The ophiolites of the Tibet Geotraverses, Lhasa to Golmud (1985) and Lhasa to Kammandu (1986). Philosophical Transactions of the Royal Society of London, Series A, 327(1594), 215-238. PEDERSEN, R.B., SEARLE, M.P. & CORFIELD, R.I. 2001. U-Pb zircon ages from the Spontang Ophiolite, Ladakh Himalaya. Journal of the Geological Society, London, 158, 513-520. PINI, G.A. 1999. Tectonosomes and Olistostromes in the Argille Scagliose of the Northern Apennines, Italy. Geological Society of America, Special Papers, 225, 1-70. Pozzi, J.P., WESTPHAL, M., GIRARDEAU, J., BESSE, J., YAO, X.Z., XIAN, Y.C. & Li, S.X. 1984. Paleomagnetism of the Xigaze ophiolite and flysch (Yarlung Zangbo suture zone, southern Tibet); latitude and direction of spreading. Earth and Planetary Science Letters, 70(2), 383-394. PROUST, F., BURG, J.P., MATTE, P., TAPPONNIER, P., Li, T.D., Li, G.C. & CHEN, G.M. 1984. Succession des phases de plissement sur une transversale du Tibet meridional, implication geodynamiques. In: MERCIER, J.L. & Li, G.C. (eds) Mission Franco-Chinoise

163

au Tibet 1980. Etude geologique et geophysique de la croute terrestre et du manteau superieur du Tibet et de I'Himalaya. CNRS, Paris, 385-392. RATSCHBACHER, L., FRISCH, W. & Liu, G. 1994. Distributed deformation in southern and western Tibet during and after the India-Asia collision. Journal of Geophysical Research, 99(B10), 1991719945. ROBERTSON, A.H.F. 1998. Rift-related sedimentation and volcanism of the North Indian margin inferred from a Permian-Triassic exotic block at Lamayuru, Indus suture zone (Ladakh Himalaya) and regional comparisons. Journal of Asian Earth Sciences, 16(2-3), 159-172. ROBERTSON, A.H.F. & SHARP, I. 1998. Mesozoic deepwater slope rise sedimentation and volcanism along the North Indian passive margin; evidence from the Karamba Complex, Indus suture zone (western Ladakh Himalaya). Journal of Asian Earth Sciences, 16(2-3), 195-215. SEARLE, M.P., WINDLEY, B.F. & COWARD, M.P. ET AL. 1987. The closing of Tethys and the tectonics of the Himalaya. Geological Society of America Bulletin, 98(6), 678-701. SEARLE, M.P., KHAN, M.A., FRASER, J.E., GOUGH, S.J. & JAN, M.Q. 1999. The tectonic evolution of the Kohistan-Karakoram collision belt along the Karakoram Highway transect, north Pakistan. Tectonics, 18, 929-949. SHACKLETON, R.M. 1981. Structure of southern Tibet; report on a traverse from Lhasa to Khatmandu organised by Academia Sinica. Journal of Structural Geology, 3(1), 97-105. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, 2000GC000080. TAPPONNIER, P., MERCIER, J.L. & PROUST, F. ET AL. 1981. The Tibetan side of the India-Eurasia collision. Nature, 294(5840), 405-410. WAN, X., WANG, L., WANG, C. & JANSA, L. 1998. Discovery and significance of Cretaceous fossils from the Xigaze forearc basin, Tibet. Journal of Asian Earth Sciences, 16(2-3), 217-223. WAN, X.Q., JANSA, L.F. & SARTI, M. 2002. Cretaceous and Paleogene boundary strata in southern Tibet and their implication for the India-Eurasia collision. Lethaia, 35, 131-146. WANG, C.S. & Liu, Z.F. ET AL. 1999. Xigaze Forearc Basin and Yarlung Zangbo Suture Zone, Tibet. Geological Publishing House, Beijing. WANG, C.S., Li, X.H., Hu, X.M. & JANSA, L.F. 2002. Latest marine horizon north of Qomolangma (Mt Everest): implications for closure of Tethys seaway and collision tectonics. Terra Nova, 14(2), 114-120. WANG, X.B., BAG, P.S. & XIAO, X.C. 1987. Ophiolites of the Yarlung Zangbo (Tsangbo) River, Xizang (Tibet). Publishing House of Surveying and Mapping, Beijing. WEI, B.J. & PENG, Y.H. (Compilers) 1984. Xizang Autonomous Region Xigaze-Quxu Geological Traverse Map 1:200,000. Xizang Geological Survey Geological Team No. 2, Lhasa. WIEDMANN, J. & DURR, S.B. 1995. First ammonites

164

J. C. AITCHISON ETAL.

from the Mid- to Upper Cretaceous Xigaze Group, South Tibet, and their significance. Newsletters on Stratigraphy, 32(1), 17-26. Wu, H.R. 1986. Some new genera and species of Cenomanian Radiolaria from southern Xizang (Tibet). Acta Micropalaeontologica Sinica, 3(4), 347— 360. Wu, H.R. 1993. Upper Jurassic and Lower Cretaceous radiolarians of Xialu chert, Yarlung Zangbo ophiolite belt, southern Tibet. In: BLUEFORD, J.R. & MURCHEY, B.L. (eds) Radiolaria of Giant and Subgiant Fields of Asia, Nazarov Memorial Volume. Micropaleontology Special Paper, 6, 115-136. XU, R.H.,

SCHAERER, U. &

ALLEGRE, CJ.

1985.

Magmatism and metamorphism in the Lhasa Block (Tibet); a geochronological study. Journal of Geology, 93(1), 41-57. YIN, A. & HARRISON, T.M. 2000. Geologic evolution of the Himalayan-Tibetan orogen. Annual Review of Earth and Planetary Sciences, 28, 211-280. YIN, A., HARRISON, T.M., RYERSON, F.J., CHEN, W.J., KIDD, W.S.F. & COPELAND, P. 1994. Tertiary structural evolution of the Gangdese thrust system in southeastern Tibet. Journal of Geophysical Research, 99, 18175-18201. YIN, A., HARRISON, T.M. & MURPHY, M.A. ET AL. 1999. Tertiary deformation history of southeastern and southwestern Tibet during the Indo-Asian collision. Geological Society of America Bulletin, 111(11), 1644-1664. YIN, J., Xu, J., Liu, C. & Li, H. 1988. The Tibetan Plateau; regional stratigraphic context and previous work. Philosophical Transactions of the Royal Society of London, Series A, 327(1594), 5-52. ZHANG, S.M. & Fu, X.L. (Compilers) 1982. Xizang Autonomous Region Xigaze—Saga Geological Traverse Map 1:200,000. Xizang Geological Survey Geological Team No. 2, Lhasa. ZHOU, M.F. 1995. Petrogenesis of the podiform chromitites in the Luobusa ophiolite, southern Tibet. PhD thesis, Dalhousie University, Halifax, N.S. ZHOU, M.F. & ROBINSON, P.T. 1994. Petrogenesis of podiform chromitites in the Luobusa Ophiolite, Tibet;

a melt-rock reaction model. Geological Association of Canada, Program with Abstracts, 19 125. ZHOU, M.F., ROBINSON, P.T., MALPAS, J. & Li, Z. 1996. Podiform chromitites in the Luobusa Ophiolite (southern Tibet); implications for melt-rock interaction and chromite segregation in the upper mantle. Journal of Petrology, 37(1), 3-21. ZHOU, S., Mo, X.X., MAHONEY, J.J., ZHANG, S.Q., Guo, TJ. & ZHAO, Z.D. 2002. Geochronology and Nd and Pb isotope characteristics of gabbro dikes in the Luobusha ophiolite, Tibet. Chinese Science Bulletin, 41(2}, 143-146. ZIABREV, S. 2001. Tectonic evolution of Dazhuqu and Bainang terranes, Yarlung Zangbo suture, Tibet as constrained by radiolarian biostratigraphy. PhD thesis, University of Hong Kong. ZIABREV, S., AITCHISON, J., BADENGZHU,, DAVIS, A.M., Luo, H. & Liu, J. 2000. Tethyan relics in the Yarlung-Tsangpo suture, Tibet: structural setting, radiolarian ages and their tectonic significance. In: CARTER, E.S., WHALEN, P., NOBLE, P.J. & CRAWFORD, A.EJ. (eds) Ninth Meeting, The International Association of Radiolarian Paleontologists INTERRAD 2000, Blairsden, CA. Program with Abstracts. 72. ZIABREV, S.V., AITCHISON, J.C., ABRAJEVITCH, A., BADENGZHU, DAVIS, A.M. & Luo, H. 2003a. Bainang Terrane, Yarlung-Tsangpo suture, southern Tibet (Xizang, China): a trace of intra-Tethyan subduction on the Roof of the World. Journal of the Geological Society, London. ZIABREV, S.V., AITCHISON, J.C., ABRAJEVITCH, A., BADENGZHU, DAVIS, A.M. & Luo, H. 2003b. Precise radiolarian age constraints on the timing of ophiolite generation and sedimentation in the Dazhuqu terrane, Yarlung-Tsangpo suture zone, Tibet. Journal of the Geological Society, London, 160, 591-599. ZIABREV, S.V., AITCHISON, J.C., BADENGZHU, ., DAVIS, A.M., Luo, H. & MALPAS, J. 1999. Radiolarian biostratigraphy of supra-ophiolite sequences in the Xigaze area, Yarlung-Tsangpo suture, Southern Tibet (preliminary report). Radiolaria, 17, 13-19.

Yarlimg Zangbo ophiolites (Southern Tibet) revisited: geodynamic implications from the mineral record REJEAN HEBERT 1 , FRANCOIS HUOT 1 , CHENGSHAN WANG 2 & ZHIFEI LIU 3 1

Departement de Geologic et de Genie Geologique, Universite Laval, Sainte-Foy, Quebec, Canada, G1K IP4 (e-mail: [email protected]) 2 Institute of Sedimentary Geology, Chengdu University of Technology, Chengdu, Sichuan, 610059, PR. China 3 Department of Marine Geology and Geophysics, Tongji University, 1239 Siping Road, Shanghai 200012, P.R. China Abstract: We present mineral chemistry data and petrological evidence from the Yarlung Zangbo suture zone ophiolites (Southern Tibet) suggesting that they represent a collage of heterogeneous massifs. Mantle sections in these ophiolites consist of harzburgite and Iherzolite cut by several generations of gabbroic to diabasic intrusions, all affected by high-temperature deformation. Pyroxenitic bands are parallel to the mantle foliation. Crustal plutonic sections, consisting of dunite, wehrlite and gabbro, are thin or absent and have been observed only in the Dazhuqu massif. Plagioclase is an additional phase associated with crustal peridotites. The mineral chemistry of silicate minerals and spinel in the mantle and crustal rocks varies widely and is believed to reflect complex melt percolation and reaction. The massifs record polybaric exhumation steps from at least 50 km depth to the plagioclase stability field. Pyroxene has reequilibrated compositions from 1200°C down to medium-grade metamorphic conditions. The mantle peridotites are interpreted as the residues of 10-40% partial melting of a fertile Iherzolitic source. High Cr number, low TiC>2 content and relatively high Fe3+ number of spinels suggest that the ophiolitic massifs were generated in a suprasubduction zone (arc or back-arc) environment.

The Yarlung Zangbo suture zone (YZSZ) represents one of the major tectonic features of the Tibetan Plateau and records the collisional event between the Indian and Eurasian plates (Molnar & Tapponnier 1975; Gansser 1974; Fig. 1). Vestiges of the Tethyan oceanic domain, such as ophiolitic massifs, are partly preserved along the YZSZ. A better understanding of their origin and evolution from intra-oceanic to collisional settings is critical for development of plausible geodynamic models for the Mesozoic and Tertiary tectonics of the Tibetan Plateau. The common characteristics of the YZSZ ophiolitic massifs (Fig. 2), as defined by the collective results of the studies following the 1980 Sino-French Cooperative Investigation of the Himalayas (Wu & Deng 1980; Nicolas et al. 1981; Mercier & Li 1984; Girardeau et al. 1985a, 1985b), include: (1) a thin (2-4 km) crustal section; (2) the absence or low volume of crustal gabbroic components; (3) the occurrence of multiple gabbroic and diabasic dykes and sills that cut both the mantle and the crustal sections (Fig. 3).

The Xigaze ophiolites were interpreted to have formed at a slow-spreading mid-ocean ridge in a small basin located near the Eurasian palaeocontinent (Nicolas et al. 1981; Girardeau et al. 1984, 1985a, 1985c; Pozzi et al. 1984; Xiao 1980). The ophiolite belt formed some 120 ± 10 Ma ago as inferred from a U-Pb age (Gopel et al. 1984) from the Xigaze massif. Marcoux et al. (1982) published an Albian-Cenomanian age for radiolarian cherts overlying the Xigaze massif, whereas Ziabrev et al. (2000) reported an age of 121 Ma (Barremian) on Bainang radiolarian cherts. The ophiolite is believed to have been initially obducted southward during the Late Cretaceous-Eocene periods and backthrust in OligoMiocene times (Gansser 1974; Nicolas et al. 1981; Allegre et al. 1984). Subduction of the NeoTethys caused the development of the Gangdese continental arc on the southern margin of the Lhasa block at least from 95 to 40 Ma (Allegre et al. 1984; see also Wang et al. 2000 for a review) and possibly as early as 153 ± 6 Ma (Murphy et al. 1997). Part of the Gangdese arc was built on

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 165-190. 0305-8719/03/$15 © The Geological Society of London 2003.

166

R. HEBERT ETAL.

Fig. 1. Location map of the Yarlung Zangbo ophiolite massifs, (a) Western part; (b) eastern part. Modified after Gansser (1991) and Wang et al (1999). QT, Qiahgtang Terrane; LT, Lhasa Terrane; BNSZ, Baugong-Nujrang Suture Zone; YZSZ, Varlung Zangbo Suture Zone; MCT, Main Central Thrust; MET, Main Boundary Thrust.

an older metamorphosed ophiolitic crust (Proust et al. 1984) possibly accreted by the Late Jurassic to Early Cretaceous (Wang et al. 2000). As many as four compressional phases together with younger strike-slip and east-west extensional episodes have been recorded (Tapponier et al. 1981), all of which result from the closing of the Tethyan and Neo-Tethyan domains, and subsequent collision between Eurasia and India. These deformational episodes led to the partial dismemberment, faulting and shortening of the ophiolitic sequences (Girardeau et al. 1984). We conducted fieldwork in the eastern segment of the YZSZ during 1998 and 1999 to reassess its geology and to shed new light on the preserved fragments of Neo-Tethyan ocean floor (Hebert et al. 1999, 2000, 2001b, 2001c; Huot et al. 2002). Progress in ophiolite research and developments from modern oceanic crust studies guided us in the fieldwork. The investigated segment extends from Jiding to Luobusa, which comprises a strip of 250 km length. This work significantly extends the area of previous investigations on the Xigaze ophiolite, which herein we call the YZSZ ophiolites. The aim of this paper is to present the mineral chemistry of peridotites and some mafic plutonic rocks of the Yarlung Zangbo ophiolites and to discuss their petrogenetic and geodynamic significance. We investigated six lithologically and chemically distinct massifs along the YZSZ; these are, from west to east: Jiding, Beimarang, Qunrang,

Dazhuqu, Jinlu and Luobusa (Fig. 1). These massifs are named according to the nearest village and/or river valley to provide a geographical reference for each studied section.

Overview of the crustal unit of the YZSZ ophiolites The ophiolitic belt is clearly heterogeneous along strike and reveals a complex geological history. No single stratigraphic column is representative of the belt as a whole. The massifs are overprinted by late-stage deformational structures such as thrust, backthrust and strike-slip faults, folds and shear zones that make primary features difficult to decipher. Here we present important features of two highly distinct massifs (Fig. 2). The uppermost portion of the crustal unit, which crops out on the northern part of the ophiolites, is represented by a volcanic sequence and associated radiolarites. According to Tapponier et al (1981) and Marcoux et al. (1982), the ophiolitic volcano-sedimentary unit is stratigraphically overlain by the Xigaze Group, a sequence consisting of turbidites deposited in a forearc basin south of the Gangdese arc. On the other hand, Aitchison et al. (2000) strongly favoured a fault contact between these units. The volcanic products in most massifs include massive and pillowed basalts together with their autobrecciated facies. Volcaniclastic deposits are only locally

Fig. 2. (continued

overleaf).

Fig. 2. (continued overleaf}.

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES

169

Fig. 2. Geological maps of the studied ophiolite massifs showing sample locations, (a) Segment from Tiding to Beimarang massifs; (b) Segment from Xialu to Dazhuqu massifs. Modified after Wang et al. (1984); (c) Jinglu (Zedang) massif. Modified after Aitchson et al (2000).

present in the ophiolitic sequence but, as reported by Aitchison et al. (2000), are dominant in the Zedong terrane. This Early Cretaceous terrane is composed of island arc volcanic and volcaniclastic rocks having compositions ranging from basaltic andesite to rare dacite. The Zedong terrane is thought to be technically juxtaposed (Aitchison et al. 2000) with the clinopyroxene-phyric basalts belonging to the Zedang ophiolitic volcanic sec-

tion (Jinlu massif in this paper). In the YZSZ ophiolitic belt, diabasic sills and rare dykes locally intrude the volcanic pile and generally increase downward, passing into a sill complex. Among the studied massifs, Qunrang has relatively thick volcanic and diabasic sections (1-2 km). The plutonic section is rather thin and even absent in the YZSZ ophiolitic belt (e.g. Qunrang). According to Nicolas et al. (1981), the thin plutonic

170

R. H E B E R T ^ r ^ L .

Fig. 3. Idealized cross-sections of the Beimarang and Dazhuqu massifs showing similarities and differences along the YZSZ ophiolite belt. Modified after Girardeau et al. (1985a, 1985b).

section reflects a low magma supply and is not a consequence of tectonic thinning. In the study area, the Dazhuqu massif is the only one with a plutonic section (<300 m thick). The plutonic rocks include plagioclase dunite, wehrlite, plagioclase wehrlite, olivine websterite, olivine gabbronorite, troctolite and gabbro, all of which show complex patterns of intergrowth complicated by isoclinal plastic deformation. Several generations of locally plagioclase- or pyroxene-phyric dykes intrude the gabbroic part of the Tiding massif. In some localities, such as Jiding, hydrothermal circulation was locally intense; Cu mineralization and epidosites are observed within, or close to, shear zones. The boundary between crustal and mantle sections is either tectonic (eg. Qunrang, Jinlu) or marked by a Moho transition zone (eg. Jiding, Dazhuqu). In the Jinlu massif, the volcanic unit is technically juxtaposed against the mantle peridotites, without any intervening crustal plutonic rocks.

Ultramafic unit of the YZSZ ophiolite: similarities and differences between the studied massifs

grained harzburgites with ormopyroxenite banding. High-temperature plastic foliation (generally oriented NW-SE) and associated lineation display intense folding in the peridotites. Numerous gabbroic and diabasic intrusions are observed in sheared contacts within the peridotites. The hightemperature planar and linear structures in the mafic rocks do not parallel the plastic foliation in the mantle, suggesting rotation of the blocks during shearing. A Moho transition zone of 350 m thickness, consisting of screens of mantle peridotites and various types of crosscutting gabbro and diabase, is believed to be a syntectonically intrusive sequence. Gabbroic intrusions locally postdate serpentinization of peridotites, an observation contrasting with that of Girardeau et al. (1985a), who stated that gabbro injections preceded the serpentinization. Ultramafic xenoliths are locally enclosed within intrusive gabbro bodies. The outer margin of these xenoliths reacted with the mafic intrusion to form pegmatitic hornblende gabbro at the contact. A few centimetres from the gabbroxenolith contact, the amphibole crystals disappear. This observation suggests that the xenoliths were already serpentinized, supplying the required I^O for amphibole crystallization.

Jiding massif

Beimarang massif

The mantle section of the Jiding massif is made of partly to totally serpentinized, coarse- to fine-

The Beimarang massif is mainly composed of harzburgite and clinopyroxene-bearing harzbur-

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES

171

gite. Coarse-grained Iherzolite, showing local spinel or orthopyroxene enrichment, is more abundant than in other massifs. Chromite pods are concordant with the foliation. Local dunite bodies, both concordant and discordant, are not systematically enriched in chromite. The mantle rocks are cut by porphyritic gabbro veins and diabase intrusive rocks. The apparent upper part of the mantle is technically complex and is injected by variously deformed diabasic and gabbroic intrusive rocks. The sequence is spatially associated with a well-exposed tectonic ophiolitic melange with a serpentinite matrix. The predominance of ultramafic, gabbroic and diabase blocks suggests that the melange material was mostly derived from the dismemberment of an ophiolitic mantle unit (Huot et al. 2003). Based on isotopic values Agrinier et al. (1988) concluded that serpentinization of these rocks may have occurred in both oceanic and continental settings.

Jinlu massif

Qunrang massif

Luobusa massif

The mantle section is composed of coarse-grained Iherzolite, harzburgite, dunite, and cross-cutting vari-textured gabbro bodies. High-temperature foliation and lineation in the peridotites are highly variable, very different from the regular orientation usually observed in mantle peridotites (Nicolas 1989). These structures could well reflect late tectonic activities.

The Luobusa massif mainly consists of coarsegrained protogranular to porphyroclastic harzburgite and lesser dunite. This massif has a clear predominance of harzburgite and dunite compared with other peridotite bodies along the YZSZ. The Luobusa massif is in fault contact with a sedimentary melange unit to the south, and an ophiolitic melange to the north (Zhou et al 1996). The massif is chromite rich and hosts the largest active chromite mine in China. Chromitite occurrences are associated with dunitic aureoles (Zhou et al. 1996). The chromitites display granular, massive, nodular and banded structures. High-temperature isoclinal folds and lobate structures are observed in the layered facies. Silicates and non-silicates containing diamond inclusions are locally present (Hu et al. 1998). Numerous platinum group element (PGE) alloys were documented in the chromitites by Bai et al. (2000).

Dazhuqu massif The Dazhuqu massif is mainly composed of mantle harzburgite (having <5% modal clinopyroxene), dunite and Iherzolite (having >5% modal clinopyroxene) with concordant orthopyroxenite banding. Because of a strong mantle foliation, these lithologies show a regular alternation, which probably reflects plastic isoclinal folding of the mantle sequence. The mantle foliation in protogranular to mylonitic Iherzolite, harzburgite and dunite is mainly oriented NE-SW, parallel to the foliation measured in the gabbroic intrusive rocks. The average orientation of the foliation is oblique to that of the petrological Moho, which trends in an east-west direction. The foliation in the crust, although variable, is mainly oriented NW-SE, and tends to parallel the mantle foliation close to the crust-mantle boundary. Even though the transition from upper mantle to lower crust is not observed because of the injection of numerous diabasic sills, we believe that it is less than 350 m thick.

The Jinlu massif comprises a mantle sequence of roughly 3 km thickness, which contains very fresh coarse-grained, protogranular to porphyroclastic and mylonitic Iherzolite and harzburgite. The foliation is mainly oriented NW-SE with a mineral lineation plunging to the SE. Serpentinization is restricted to shear zones. The massif is underlain by an ophiolitic melange containing blocks of plutonic and volcanic rocks, and of conglomerates including clasts of chert, limestone and radiolarite. These blocks are set in a serpentinite and/or a black to red shaly matrix. A spatially undelimited, feldspar-bearing ultramafic zone is tentatively interpreted as a slice of uppermost mantle impregnated by basaltic melts that is technically sandwiched between the ophiolitic melange to the south and the upper-mantle rocks to the north. Dykes and veins include successive injections of orthopyroxenite, sub-pegmatitic hornblende gabbro and fine-grained dark lavas.

Mineral chemistry Apart from a few mineral references given by Mercier & Li (1984), Girardeau et al. (1985a), Girardeau & Mercier (1988) and Zhou et al. (1996), the mineral chemistry of the Yarlung Zangbo ophiolites is still poorly known. In this section, we present new microprobe mineral analyses of 76 mafic and ultramafic samples from the ophiolites. Over 500 analyses of mostly primary phases were performed on an SX-100 Cameca microprobe at Universite Laval, Quebec. Operating conditions included a counting time of 10 s at

Table 1. Summarized petrography of the Yarlung Zangbo Suture Zone ophiolite samples Estimated primary modes (%) Serp. (%)

01

Spl

Opx

Cpx

PI

Amp

Harzburgite Cpx-harzburgite

90

85

3

9

3

—

—

Cumulates Cumulates Cumulates Cumulates Cumulates

Cpx-troctolite Dunite Dunite Troctolite Wehrlite

80 99 99 85 85

65 90 97 73 83

2 98 3 2 2

2 0 -

18

15

5 15

20

98-D-17E

Cumulates

Pl-wehrlite

80

65

3

-

20

12

-

98-D-21

Cumulates

Ol-Pl-websterite

35

45

<1

10

35

10

-

98-D-24A2

Mantle

Cpx-harzburgite

80

85

2

12

1

-

98-D-26A

Mantle

Cpx-harzburgite

75

80

1

16

3

-

V-99-DAZ-3A

Cumulates

Ol-gabbronorite

10

20

<1

10

25

45

-

V-99-DAZ-4

Mantle

Diabase/gabbro

-

-

-

-

42

55

3

V-99-DAZ-5A

Cumulates

Dunite

97

95

3

2

_

_

_

-

V-99-DAZ-5C

Cumulates

Pl-wehrlite

90

77

3

-

8

12

-

-

V-99-DAZ-6

Cumulates

Pl-wehrlite

90

75

3

-

14

8

-

-

Zedong 98-Z-10

Mantle

Cpx-harzburgite

15

73

3

20

<5

98-Z-ll 98-Z-12

Mantle Mantle

Harzburgite Cpx-harzburgite

70 10

71 65

2 2

30

3

98-Z-14-1

Mantle

Harzburgite

15

76

3

20

Sample no.

Unit

Rock type

Dazhuqu 98-D-1A 98-D-10

Mantle Cumulates

98-D-13 98-D-13E 98-D-13F 98-D-15A 98-D-16A

-

Op

-

-

1

-

General comments

Brown allotriomorphic Spl; small irregular Cpx and altered Opx Wehrlitic trend Serpentinized Serpentinized Serpentinized Rounded altered Ol in Cpx; Cpx partly replaced by Amp; dark Spl Rounded altered Ol in Cpx and altered PI; black sub-idiomorphic Spl; PI replaced by Prh+Chl+Ep Interstitial PI replaced by Prh+Act; small Spl Allotriomorphic brown Spl; weakly foliated; rounded Ol in Opx Allotriomorphic brown Spl; mostly aligned Cpx Rounded Ol in Cpx (and vice versa); interstitial Ol; Prh+Amp Diabasic texture; Act-Hbl-Prh-Qtz (greenschist); altered Cpx Recrystallized undeformed olivine; idioblastic dark Spl Impregnated dunite; interstitial Cpx+Pl; mantle Ol; dark Spl mostly in Ol Impregnated dunite; interstitial Cpx+Pl; mantle Ol; dark Spl mostly in Ol Plastic deformation and neoblasts; interstitial and idiomorphic brown Spl Brown sub-idiomorphic Spl; in exsolution and large Cpx Plastic deformation; neoblasts; interstitial greenish brown Spl and altered PI

98-Z-15

Mantle

Cpx-harzburgite

<2

72

2

25

1

98-Z-18-1

Mantle

Cpx-harzburgite

<10

82

2

15

<1

98-Z-19

Mantle

Dunite

<5

98

2

-

98-Z-20-2

Mantle

Dunite

<3

98

2

98-Z-22

Mantle

Harzburgite

<2

85

Luobusa 99-LUO-01A 99-LUO-01B

Mantle Mantle

Harzburgite Cpx-dunite

<5 <5

99-LUO-01C1

Mantle

Dunite

99-LUO-01D

Mantle

99-LUO-01E

-

-

-

-

-

<1

tr

-

-

1

<15

tr?

-

-

85 97

2 3

13

<1 <1

5

82

18

-

-

-

-

Cpx-harzburgite

<5

89

1

10

<1

-

-

Mantle

Dunite

<5

85

5

-

_

99-LUO-02A

Mantle

Chromitite

18

18

80

-

2

-

-

99-LUO-02B2 99-LUO-02C

Mantle Mantle

Cpx-dunite Cpx-harzburgite

15 5

96 77

4 2

20

<1 1

-

-

99-LUO-03B

Mantle

Dunite

30

96

2

2

-

-

-

99-LUO-5A

Mantle

Melagabbronorite

-

-

-

5

60

35

-

99-LUO-5B

Mantle

Melagabbronorite

-

-

-

15

22

63

-

99-LUO-5C Jiding 98-J-18

Mantle

Cpx-dunite

97

96

4

Cumulates

Dunite

98

95

3

2

98-J-18A 98-J-18B

Cumulates Cumulates

Harzburgite Lherzolite

97 65

90 70

1 2

<10 20

98-J-22B

Mantle

Harzburgite

85

85

3

17

Plastic deformation; neoblasts; interstitial greenish brown Spl; Cpx in exsolution Plastic deformation; neoblasts; undeformed brown Spl Plastic deformation; neoblasts; stretched Ol; brown equant Spl Plastic deformation; neoblasts; large undeformed brown Spl Plastic deformation; neoblasts; Act and Chi on Opx; rounded and interstitial brown Spl Large dark brown Spl; deformed Ol Calcite veins; plastically deformed Ol; Ol neoblasts; sub-idiomorphic dark brown Spl Spl banding; dark brown idiomorphic Spl; deformed Ol; Ol neoblasts Cataclastic zones; brown Spl; plastically deformed Ol Deformed Ol; Ol neoblasts; cataclastic zones; allotriomorphic brown Spl Dark Spl with inclusions of Ol and Cpx; large brown Spl Dark idiomorphic Spl; deformed Ol Brown interstitial Spl; deformed Ol; slightly recrystallized Highly stretched Ol; dark sub-idiomorphic Spl; altered interstitial Opx Fresh PI veinlets crosscut by cataclastic zones; altered PI in host rock; plastic deformation Altered PI; mostly interstitial Cpx and Opx Black sub-idiomorphic Spl

<1

tr 8

-

-

_

_

_

Dark sub-idiomorphic Spl; recrystallized Ol Mantle peridotite; brown Spl Brown subrounded Spl; Cpx in exsolution and granular Porphyroclastic texture

Table 1. (continued) Estimated primary modes (%) Sample no.

Unit

Rock type

Serp. (%)

01

Spl

_

70

_ 1

Opx 25 25

Cpx

PI

Amp

Op

75 4

_

_

98-J-29-1 98-J-35

Cumulates Melange

Websterite Cpx-harzburgite

_ 75

Beimarang V-99-BEI-2A

Mantle

Cpx-harzburgite

35

82

2

15

<1

-

V-99-BEI-2B

Mantle

Harzburgite

90

85

2

13

-

-

V-99-BEI-2C

Mantle

Gabbro

-

30

-

15

45

V-99-BEI-5B

Mantle

Dunite

99

98

2

-

-

-

V-99-BEI-5C

Mantle

Websterite

-

-

tr

30

70

-

V-99-BEI-7

Mantle

Pegmatitic gabbro

-

-

-

-

43

V-99-BEI-10A

Mantle

Harzburgite

98

78

2

20

-

_

V-99-BEI-10B

Mantle

Dunite

99

96

4

-

-

-

V-99-BEI-14A

Mantle

Cpx-harzburgite

60

80

1

16

3

-

V-99-BEI-14C

Mantle

Cpx-harzburgite

15

85

2

11

2

V-99-BEI-14E

Mantle

Cpx-harzburgite

15

77

<1

20

2

-

V-99-BEI-15A V-99-BEI-15B2

Mantle Mantle

Cpx-harzburgite Dunite

<5 <5

78 95

1 5

20 -

1 -

-

V-99-BEI-17

Mantle

Lherzolite

0

69

6

20

5

-

V-99-BEI-18 V-99-BEI-24A

Mantle Mantle

Orthopyroxenite Cpx-harzburgite

10

70

5

20

5

-

-

10

55

_

-

2

-

-

<1

-

General comments Recrystallized with neoblasts Dark brown Spl; Cpx in clusters; mantle peridotite Brown idioblastic Spl; granular and interstitial Opx; pyroxene partly replaced by Act Brown allotriomorphic Spl; oxidized serpentine; granular and interstitial Opx Fine-grained interstitial PI and Cpx; cumulate texture Oxidized serpentine; dark subidiomorphic Spl Opx partly replaced by Act; mainly undeformed large pyroxenes; few neoblasts Prehnite veinlets; interstitial primary Hbl; Act Porphyroclastic and interstitial altered Opx Equant recrystallized Ol; plastically deformed Spl along a foliation Oxidized serpentine; brown allotriomorphic Spl; Opx porphyroclasts; interstitial Cpx Reddish brown allotriomorphic Spl; Opx partly replaced by Tic Brown allotriomorphic Spl; Cpx associated with Opx; neoblasts Recrystallized Ol and Opx; brown Spl Large irregular brown Spl; recrystallized Ol Large irregular brown Spl; large Opx; interstitial small Cpx; neoblasts Brown Spl; Opx as porphyroclasts and in matrix; mantle foliation

V-99-BEI-51A

Mantle

Cpx-harzburgite

60

85

1

12

100

98

1

1

Qunrang (Xialu) 98-X-1A Mantle

Dunite

98-X-1B

Mantle

Lherzolite

50

57

3

35

98-X-5 98-X-8

Mantle Mantle

Gabbro Harzburgite

99

80

1

19

98-X-9

Mantle

Harzburgite

99

89

1

10

98-X-18 98-X-20

Mantle Mantle

Harzburgite Harzburgite

98 99

82 80

3 2

15 18

98-X-21 98-X-22A

Mantle Mantle

Harzburgite Cpx-harzburgite

99 97

84 82

1 2

15 15

98-X-22B V99-QUN-4

Mantle Mantle

Harzburgite Dunite

98

98

2

<1

V99-QUN-4A

Mantle

Harzburgite

98

78

2

20

V99-QUN-8A

Mantle

Cpx-harzburgite

96

82

1

13

V99-QUN-9

Mantle

Harzburgite

99

85

3

12

V99-QUN-10

Mantle

Harzburgite

98

84

1

15

2

5

1

4

tr

Allotriomorphic black Spl; porphyroclastic Opx; fine-grained Cpx; weak foliation Vermicular Spl associated with altered Opx; dark altered Spl Dark Spl; vermicular Spl associated with Opx; irregular Cpx and Opx Orthopyroxenite vein; Opx partly replaced by a colourless mica; dark brown Spl Dark brown allotriomorphic Spl; altered Opx; Opx partly replaced by a colourless mica Allotriomorphic brown Spl; altered Opx Prh veinlets; allotriomorphic dark Spl; vermicular Spl; altered neoblasts of Opx Brown Spl; altered Opx Brown allotriomorphic Spl; altered granoblastic and interstitial Opx Small sub-idiomorphic dark brown Spl; equant altered Ol Large allotriomorphic brown Spl; altered porphyroclastic and interstitial pyroxenes Large altered Opx; small fresh allotriomorphic Cpx; brown Spl Allotriomorphic brown Spl; altered porphyroclastic Opx Large allotriomorphic brown Spl; altered porphyroclastic Opx; fibrous Act

Serp. (%) is percentage of serpentine in ultramafic rocks. Ol, olivine; Spl, spinel; Opx, orthopyroxene; Cpx, clinopyroxene; PI, plagioclase; Amp, amphibole; Spn, sphene; Op, opaque minerals; Prh. prehnite; Act, actinolite; Chi, chlorite; Hbl, hornblende; Qtz, quartz; Tic, talc.

176

Fig. 4. (continued overleaf).

R. HEBERT ETAL.

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES

177

Fig. 4. Compositional variations of spinels from peridotites and mafic plutonic rocks, (a) Variation diagrams of Cr number v. Mg number. Fields for spinels in abyssal peridotites, normal-type mid-ocean ridge basalt (N-MORB) and boninites are taken from Dick & Bullen (1984). Others have been drawn from data for spinels in forearc peridotites (Ishii et al. 1992), island-arc tholeiites (IAT) and back-arc basin basalts (BABB) (Allan 1994), and ferrogabbro veins in peridotites and peridotite wall-rock of ferrogabbro veins. The curve with ticks represents the percentage of melting of the host peridotite (Hirose & Kawamoto 1995). (b) Variation diagrams of Fe2Os v. TiCVCr number. Fields outline spinel compositions in abyssal peridotites (Juteau et al. 1990), forearc peridotites (Ishii et al. 1992), island-arc tholeiites and back-arc basin basalts (Allan 1994), N-MORB (Allan et al. 1988), and alkaline and transitional alkaline basalts (Thy 1983).

15kV accelerating potential and 20 nA beam current. Pure oxide and natural phases were used as standards. Table 1 summarizes petrography of the analysed samples, and the representative analyses can be obtained from the Society Library or

the British Library Document Supply Centre, Boston Spa, Wemerby, West Yorkshire LS23 7BQ, UK as Supplementary Publication No. SUP 18194 (9pp) and can also be accessed online at http:// www.geolsoc.org.uk/SUP 181947.

178

R. HEBERT ETAL.

Spinel Spinel shows an unusual compositional range that spans almost the whole spectrum of terrestrial spinels, extending the limited range reported by Girardeau & Mercier (1988; Fig. 4a). The most aluminous spinels, with Cr number (Cr/(Cr -j- Al)) 0-13, are found in Iherzolites from the Qunrang massif, whereas the most chromiferous spinels, with Cr number around 0.94, are found in dunites from the Luobusa massif Two compositional trends are observed and they appear to depend on the origin of the host rocks. The 'Luobusa trend', defined by a slight decrease in Mg number and a large variation in Cr number, is similar to that of spinels hosted in mantle peridotites devoid of textural features indicating melt percolation, such as poikilitic clinopyroxene and plagioclase. The Luobusa trend parallels the array given by Hirose & Kawamoto (1995) that is defined by spinels (continuous-line arrow in Fig. 4) in variously depleted mantle rocks. The Cr number tends to increase from Iherzolite, through clinopyroxene-

bearing harzburgite, to harzburgite and to dunite. Some peridotite samples show vermicular spinel intergrowths with orthopyroxene (Fig. 5a) and/or spinel rimmed by chlorite, presumably a pseudomorph after plagioclase (Fig. 5b). Similar features were reported from the peridotites recovered on Ocean Drilling Program Legs 149 (Cornen et al. 1996) and 173 (Hebert et al 20010). Both phase relationships are suggestive of melting processes (Nicolas 1989). Based on calculations by Hirose & Kawamoto (1995), we infer that the host peridotites of these spinels may have experienced as much as 40% melt extraction. All of the spinels lying along the Luobusa trend have significantly low contents of Fe2O3 (<8 wt%) and TiO2 (<0.3wt%), and low TiO2/Cr number ratios (<0.4 wt%) (Fig. 4b). Figure 4b shows that such spinels formed under low/O2 (low Fe2C>3 content) and/or interacted with low-TiO2 magmas (low TiO2/Cr number ratios). Interaction between melt and residual mantle rock is shown by the presence of poikilitic clinopyroxene and pseudomorphs of plagioclase

Fig. 5. Microphotographs of different textures in peridotites and mafic plutonic rocks, (a) Vermicular Cr-spinel associated with orthopyroxene in a Iherzolite (98-X-01B). (b) Allotriomorphic spinel associated with plagioclase (replaced by chlorite) in a clinopyroxene-bearing harzburgite (99-LUO-02C). (c) Equant and non-poikilitic clinopyroxene in a clinopyroxene-bearing harzburgite (98-Z-10). (d) Interstitial poikilitic clinopyroxene and pseudomorphed plagioclase in a plagioclase wehrlite (98-D-17E). The rounded olivine inclusions now partly replaced by antigorite should be noted.

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES (discussed below). Spinels in the impregnated rocks have compositions defining the 'Dazhuqu trend', which is marked by a significant decrease in Mg number and a slight increase in Cr number. Such a trend is also observed in other ophiolites such as the Bay of Islands (Varfalvy et al. 1996, 1997; Bedard & Hebert 1998), and Thetford Mines ophiolite (Laurent & Kacira 1987) and in oceanic domains such as the Garrett, Terevaka and Pito transform faults (Constantin 1999) and the Mid-Atlantic Ridge (Takahashi et al. 1987; Komor et al. 1990; Juteau et al. 1990).

Olivine Most olivine compositions in mantle peridotites from all of the YZSZ massifs are in the range of

179

Fo90-92, with NiO contents between 0.30 and 0.52 wt% (Fig. 6). Olivines in some spinel-rich dunites (99-LUO-1C1 and 99-LUO-1E) from Luobusa, with forsteritic values as high as FOQS, are thought to have undergone subsolidus re-equilibration with spinels (see Lehmann 1983). Taken as a whole, all these analyses are similar to those of olivines from both forearc and abyssal peridotites. Relatively less forsteritic (c. Fogg) and lower NiO (0.14-0.27 wt%) olivines were found in plagioclase wehrlite (V99-DAZ-6) and olivine websterite (98-D-21) from the Dazhuqu massif. Such compositions are similar to those of olivines in oceanic mafic intrusions and associated impregnated peridotites from Pito and Terevaka. The high Ni content in olivine from sample 98-X-22B is not understood at present.

Fig. 6. Variations of NiO v. Mg number in olivines from peridotites and mafic plutonic rocks. Fields outline olivine compositions in forearc peridotites (Ishii et al. 1992) and various mantle peridotites, ultramafic and mafic plutonic rocks from the oceanic domain. The arrow is compatible with the fractional crystallization trend (Constantin 1999).

180

R. HEBERT ETAL.

Orthopyroxene Orthopyroxene, either as porphyroclasts or matrix mineral in the mantle peridotites, commonly shows evidence for high-temperature plastic deformation and clinopyroxene exsolution lamellae. The composition of the Orthopyroxene varies with that of olivine and falls in the range Mg number (Mg/(Mg + Fe2+)) 0.9-0.92, with A12O3 and Cr2O3 between 5.5 and 1.5wt%, and 1.1 and 0.02 wt%, respectively (Fig. 7a and b). When compared with orthopyroxenes from modern tectonic settings, the Tibetan samples are similar to those from abyssal and forearc peridotites. A^Oa contents vary widely at relatively fixed Mg numbers. Orthopyroxenes in

impregnated wehrlite, websterite and gabbronorite from Dazhuqu massif are more iron rich (En9i-78) with A^Os and C^Os contents below 1.6 wt% and 0.3 wt%, respectively (Fig. 7a and b).

Clinopyroxene Clinopyroxene has two distinct shapes in the ultramafic rocks. Some crystals are large, equant and devoid of mineral inclusions (Fig. 5c). This type occurs in deformed Iherzolite and clinopyroxene harzburgite of the mantle section. The other type, typically found in wehrlite, gabbronorite and, more rarely, in clinopyroxene-bearing harzburgite, is interstitial with respect to olivine

Fig. 7. Compositional variations of orthopyroxenes from peridotites and mafic plutonic rocks, (a) Variation diagrams of A^Os v. Mg number. Fields outline Orthopyroxene compositions in abyssal peridotites (Johnson et al. 1990) (dashed line field), forearc peridotites (Ishii et al. 1992) (solid line field), boninites (Van der Laan et al. 1992), and ultramafic and mafic plutonic rocks (grey fields) from Pito and Terevaka and Garrett (Constantin 1999). (b) Variation diagrams of Cr2Os v. Mg number. Fields are the same as in (a). The number on the contour line refers to the percentage of analyses outside the field. Curves in (a) and (b), compatible with the fractional crystallization trend, are taken from Constantin (1999).

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES and orthopyroxene. Such clinopyroxenes, commonly associated with pseudomorphed plagioclase, are poikilitic and include rounded and resorbed grains of olivine, orthopyroxene and spinel (Fig. 5d). The small abundance of these grains is compatible with a low degree of melt percolation. However, as percolation increased, the residual olivine and pyroxenes appear to have decreased in modal abundance such that the rock evolved from dunite or harzburgite to plagioclase wehrlite, troctolite, olivine websterite and olivine gabbronorite. The two types of clinopyroxene have distinct chemical compositions (Fig. 8a and b). Clinopyroxenes with an equant shape have compositions following a diopsidic-endiopsidic trend with Mg number ranging from 0.96 to 0.87. The TiO2 and A12C>3 contents are proportional to the abundance of clinopyroxene. For instance, TiO2 and A12O3 contents are extremely low in clinopyroxene from dunite but may reach values up to 0.3 wt% in

Fig. 7. (continued).

181

TiO2 and 4.9 wt% in A12O3 in Iherzolite. Interstitial clinopyroxene in impregnated and crustallike rocks have lower Mg number (0.92-0.75), higher TiO2 (up to 0.7 wt%), and lower Cr2O3 (<0.3 wt%). Those in gabbronorites from Dazhuqu and Luobusa are particularly low in Mg number and Cr2O3, and fit the general trend for progressive crystal fractionation reported by Constantin (1999).

Plagioclase Fresh plagioclase has been analysed only in gabbro. The plagioclase composition varies from very calcic (An97) to very sodic (An3). The most calcic grains are in gabbro (sample V99-BEI-2C), where they are accompanied by forsteritic olivine and Mg-rich orthopyroxene. The nearly pure albite grains provide evidence of hydrothermal metamorphism. Some ultramafic samples also contain small amounts of interstitial plagioclase but, be-

182

Fig. 8. (continued overleaf}.

R. HEBERT ETAL.

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES

183

Fig. 8. Compositional variations of clinopyroxenes from peridotites and mafic plutonic rocks, (a) Variation diagrams of TiC>2 v. Mg number; (b) A12O3 v. Mg number; (c) C^Os v. Mg number. Fields outline clinopyroxene compositions in abyssal peridotites (Johnson et al. 1990), forearc peridotites (Ishii et al. 1992), and ultramafic and mafic plutonic rocks from Garrett (Constantin 1999). The number on the contour line refers to the percentage of analyses outside the field. Curves in (b) and (c), compatible with the fractional crystallization trend, are taken from Constantin (1999).

cause it is totally replaced by chlorite, its composition remains unknown (see previous section and Fig. 5b). Amphibole

Amphiboles of magmatic and metamorphic origin were analysed in gabbros from Dazhuqu, Qunrang and Beimarang massifs. Magmatic amphiboles occur as medium- to coarse-grained crystals that are devoid of secondary inclusions. Metamorphic amphiboles partially or completely replace magmatic amphibole or clinopyroxene. They are typically green, poikilitic, interstitial crystals. Magmatic amphiboles are mainly tschermakitic hornblende to tschermakite, and are locally rimmed by hornblende, actinolitic hornblende and actinolite accordingly to nomenclature of Leake (1978). Magmatic amphiboles show the highest TiO2 contents (2.2-3.2 wt%), whereas metamorphic amphiboles have <2 wt% TiC>2. Amphiboles have <0.1 wt% Cl. They belong to the lowK trend defined by Cannat & Casey (1995).

Discussion Geothermobarometry Geothermometric information can be obtained by using the quadrilateral projection of Lindsley (1983) (Fig. 9). Clinopyroxene and orthopyroxene show important increases and decreases in Ca, respectively, with falling temperatures. Clinopyroxene temperatures range from almost 1200 °C to below 500 °C, whereas those for orthopyroxene range from almost 1300°C to below 500 °C. The highest temperature pyroxenes have magmatic compositions whereas the lower temperature Carich clinopyroxenes and Ca-poor orthopyroxene reflect subsolidus to metamorphic compositions as a result of re-equilibration during cooling and/or recrystallization. Major elements such as Ca cannot be used for magmatic modelling. On the other hand, these re-equilibrated compositions are helpful in tracing the metamorphic evolution of the massifs. Mg-Fe exchange clearly was effective down to medium-temperature conditions.

184

R. HEBERT ETAL.

Fig. 9. Compositional variations of clinopyroxenes and orthopyroxenes from ultramafic and mafic plutonic rocks plotted in the Di-En-Fs-Hd quadrilateral. Temperature curves (at 1 atm) are taken from Lindsley (1983) and the nomenclature is from Morimoto et al. (1989).

Metamorphic amphiboles in gabbros from Beimarang and Qunrang yield pressures of 3.25.5 kbar from various barometers (Probamph©). These pressures are obviously too high for crustal levels (around 2 kbar for modern oceanic crust) and record a tectonic overpressure that could be related to the initial detachment or tectonic emplacement of the ophiolitic thrust sheets. The former process seems more likely, as Huot et al. (2002) have shown that the tectonic emplacement of the Beimarang ophiolite massif on crustal units of the Indian continent occurred under low pressure (<2 kbar) in the prehnite stability field.

Petrogenesis Mantle reactions show variable degrees of completion as shown by the extreme compositional variation of spinel and variable modal abundance of clinopyroxene. In addition, melt percolation heterogeneously affected the mantle section, as suggested by sporadic occurrences of clinopyroxene or plagioclase (now pseudomorphed by chlorite), which locally forms haloes around spinel or occurs as interstitial grains. Similar plagioclase replacement by chlorite has been described by Cornen et al. (1996) and Hebert et al. (200la). Thus, the different massifs con-

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES tain diverse mantle sections, each of which evolved independently. The depth of origin of the diverse pieces of upper-mantle rocks preserved along the YZSZ is also important. We used spinel-olivine pairs to better constrain the barometric conditions. This barometric grid is based on the experiments by Sobolev & Batanova (1995) and is complemented by additional experimental or empirical data (Jaques & Green 1980; Arai 1994; Berstein et al 1998). Most of the YZSZ massifs record a polybaric history of exhumation (Fig. 10). For instance, Luobusa and Beimarang massifs show the deepest mantle provenance of > 15 kbar. In fact, these two massifs plot outside the range of available experimental data. However, the chromium-rich spinel peridotites from Luobusa appear to have been equilibrated within a pressure range from >15 kbar to 5 kbar. This pressure range suggests several stages of re-equilibration. However, most Tibetan peridotite massifs seem to have recorded equilibration pressures around 15 kbar (about 50 km) very similar to those recorded in the Troodos ophiolite, Cyprus (Sobolev & Batanova 1995). Such pressures also generally fit the field of equilibration of upper mantle defined by Arai (1994). This observation suggests that the resulting peridotite compositions are consistent with a general scheme of progressive depletion through exhumation. Impregnated peridotites contain more fayalitic olivine compositions. This shift towards more iron-rich compositions is attributed to partial or total equilibration of the peridotites by percolation of magmas through the mantle (Bideau & Hekinian 1995). These compositions cannot be used for palaeobarometric investigations. Palaeogeodynamic settings Mineral compositions can also help us define the geodynamic provenance of ophiolite sequences. Spinel is one of the most useful geodynamic indicators (Arai 1994). Spinels from YZSZ ophiolites have compositions that plot largely outside the field defined for abyssal peridotites (Dick & Bullen 1984) (Fig. 4). Spinel compositions show both more aluminous and chromiferous end-members as a result of several overlapping or successive processes including partial melting, spinelsilicate subsolidus re-equilibration, variable /C>2, melt-mantle interaction, etc. However, it appears that the extent to which these processes affect spinel compositions reflects the geodynamic environment of formation. For instance, suprasubduction zone settings are more complex environments than mid-ocean ridges because fluids are abundant and varied in composition. The compositions of Tibetan spinels suggest that they were crystallized

185

from magmas originating in back-arc, arc, and/or forearc settings. Thus, the results presented here suggest a strong arc component in the genesis of YZSZ ultramafic rocks. This component explains the Cr end-member (Luobusa trend) but does not provide an explanation for the Al end-member (Dazhuqu trend), although Arai et al, (2003) suggested that the Al end-member reflects an oceanic lithosphere imprint. We believe that percolation of tholeiitic magma is capable of significantly modifying the composition of spinels. These conclusions contrast with those of Nicolas et al. (1981) and Girardeau et al (1985a, 1985b, 1985c), who postulated a mid-ocean ridge setting for the Xigaze ophiolite. However, our conclusions on the original setting of the YZSZ ophiolites are in agreement with those of Zhou et al. (1996) and Aitchison et al. (2000). The TiO2 content of YZSZ spinels also supports a suprasubduction origin (Fig. 4). Most spinels are very poor in TiC>2 (<0.2 wt%). Only spinels in the Dazhuqu massif have TiC>2 contents similar to those of mid-ocean ridge environments as compiled by Arai (1992). The analytical data overlap the fields for spinels from island arc and boninitic magmas. It is worth noting that Dazhuqu massif, which has the most titaniferous spinels, is the only one with feldspar-bearing rocks in the lower part of the crust, suggesting a tholeiitic affinity and low-P fractionation environment. Arai et al. (2003) also used the calculated Fe3+ content of spinel to identify major geodynamic settings such as ocean floors and island arcs. They showed that island arc magmas evolve under a much more elevated yC>2, causing oxidation of iron. On an Al-Cr-Fe3+ ternary plot (not shown here), the YZSZ spinels show a distribution very similar to island arc and forearc spinels. As we noted above, the Fe3+ content of YZSZ increases significantly at Cr number >0.5. This pattern is similar to that of the Oman ophiolite, which has mid-ocean ridge as well as arc signatures (Pearce et al. 1981). Summary and geodynamic model

YZSZ upper-mantle peridotites are heterogeneous at all scales. They are generally coarse-grained, granular peridotites with local minor mylonitic zones. These textures are typical of low-stress, high-temperature environments. The mylonitic bands suggest heterogeneous deformation patterns (Nicolas et al. 1980). The mantle structures have been reworked by deformational events related to initial oceanic decoupling, transport and tectonic emplacement onto the Indian Plate (Girardeau et al. 1984). The most magnesian silicate minerals and chromiferous spinels are found in the Zedong and Luobusa massifs. The pyroxene compositions

186

R. HEBERT ETAL.

Fig. 10. Polybaric origin of Yarlung Zangbo ophiolitic massifs as deduced from the Cr number in spinels and coexisting Mg number in olivines. The olivine-spinel mantle array (OSMA; dotted field) and compositional range in abyssal peridotite (continuous field) are from Arai (1994). The compositional ranges in pre-oceanic (rectangle with diagonal lines) and subduction margin peridotites (grey rectangle) are taken from Bonatti & Michael (1989). The dashed line represents the mean compositional variations in the Troodos peridotites (Sobolev & Batanova 1995). The 5 and 10 kbar curves are after Sobolev & Batanova (1995), and the 15 kbar curve is after Jaques & Green (1980). The black continuous and large open arrows represent variations induced by fractional crystallization and metasomatism, respectively (Arai 1994), whereas the dotted arrow shows variations caused by Mg-Fe subsolidus exchange.

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES follow a trend along the diopside-endiopside line without significant Fe enrichment. Spinels have low TiO2 and relatively high Fe3+, and vary widely in Cr number. Spinel compositions suggest relatively high /C>2 mantle conditions, and either low melt circulation or TiO2-poor melt interaction. Estimates of melting of a fertile Iherzolitic source range from a few percent to more than 40%. The peridotites equilibrated at about 10-15kbar, suggesting that the massifs were exhumed from depths of more than 50 km. Mantle sections contain numerous gabbroic to diabasic intrusions, now largely boudinaged and hydrothermally rodingitized. Because crustal sections in the YZSZ ophiolites are thin or non-existent, we suggest that most of the magmas were not extracted from the mantle. The mineral chemistry suggests that all YZSZ ophiolitic massifs originated in a suprasubduction environment with superimposed back-arc and arc signatures. The Dazhuqu massif has a plagioclasebearing crustal sequence with intermediate Cr number and higher TiO2 contents, suggesting a more mid-ocean ridge basalt (MORB)-type signature compatible with a back-arc setting. However, it is clear from this study that the original geochemical signatures of these ophiolite massifs were overprinted by complex magmatic processes. For instance, the complete compositional range of Al-rich to Cr-spinels is not fully understood. However, a similar large spinel compositional spectrum

187

was documented by Varfalvy et al. (1996, 1997), and was interpreted as the result of melt-mantle interaction. Tholeiitic melts seem to be responsible for the Al-rich end member, whereas the Crrich end member probably reflects boninitic-type melts (Varfalvy 2000). Geochemical work in progress is focused on solving some of these problems. Our model for the formation and evolution of the YZSZ ophiolites is presented in Figure 11. At 120 Ma a subduction zone south of the active continental margin was forming the calc-alkaline Gangdese igneous belt. This subduction event led to the opening of a small Neo-Tethyan ocean basin in which the Jiding, Beimarang, Qunrang and Dazhuqu massifs were generated. They all have tholeiitic compositions but retain a suprasubduction zone signature. Opening of a forearc basin produced the Jinlu and Luobusa massifs, both of which show a boninitic or an arc component. This geodynamic setting is capable of explaining the compositional variability among the studied massifs. This model contrasts with that of Pozzi et al. (1984) in that the inferred subduction event began earlier.

Conclusions The YZSZ ophiolite massifs record multi-scale heterogeneities. Mantle sections have been structurally reworked by late-stage ductile and brittle

Fig. 11. Simplified geodynamic model for the evolution of the Yarlung Zangbo suture zone ophiolites.

188

R. HEBERT ETAL.

deformation events. Structures produced by mantle flow produced both foliation and pyroxenite layers. The mantle units were intruded by variously metamorphosed diabasic and gabbroic bodies. Similarly, sills and dykes are ubiquitous in the crustal section whereas plutonic rocks are rare or absent. Only Dazhuqu massif has an embryonic crustal plutonic section. Multiple phases of intrusion are observed in both the mantle and crustal sections. The mineral chemistry points to a complex magmatic and petrogenetic history for the mantle rocks. Mantle peridotites were exhumed from depths of more than 50km and experienced 10— 40% partial melting as well as melt percolation within the suprasubduction zone wedge. Fe3+ number in spinel and the presence of amphibole in the gabbros suggest that significant amounts of water were present, and that relatively high /C>2 conditions prevailed. The geochemistry of magmatic rocks from the Beimarang ophiolitic melange (Huot et al. 20032) suggests a back-arc environment of formation.

rocks as a potential guide to magma chemistry. Mineralogical Magazine, 56, 173-184. ARAI, S. 1994. Characterization of spinel peridotites by olivine-spinel compositional relationships: review and interpretation. Chemical Geology, 113, 191204. ARAI, S., KADOSHIMA, K. & MORISHITA, T. 2003. How and to what extent non-oceanic is mantle section of the northern Oman section? Earth and Planetary Science Letters, submitted. BAI, W.J., ROBINSON, P.T. & FANG, Q.S. ET AL. 2000. The PGE and base-metal alloys in the podiform chromitites of the Luobusa ophiolite, southern Tibet. Canadian Mineralogist, 38, 585-598. BEDARD, J.H. & HEBERT, R. 1998. Formation of chromitites by assimilation of crustal pyroxenites and gabbros into peridotitic intrusions: North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland, Canada. Journal of Geophysical Research, 103, 5165-5184. BERSTEIN, S., KELEMEN, P.B. & BROOKS, C.K. 1998. Depleted spinel harzburgite xenoliths in Tertiary dykes from East Greenland: restites from high degree melting. Earth and Planetary Science Letters, 154, 221-235. BIDEAU, D. & HEKINIAN, R. 1995. A dynamic model for generating small-scale heterogeneities in ocean floor We thank the organizers of the 16th Himalaya-Karakorbasalts. Journal of Geophysical Research, 100, am-Tibet Workshop held in March 2001 at Schloss 10141-10162. Seggau, Austria, for making this meeting a scientific success. R. Hebert benefited from an NSERC grant (No. BONATTI, E. & MICHAEL, PJ. 1989. Mantle peridotites from continental rifts to ocean basins to subduction 1253). J. Dostal and S. Arai provided constructive zones. Earth and Planetary Science Letters, 91, reviews that greatly improved the manuscript. We also 297-311. thank C. Dupuis and V Dubois-Cote for their help in CANNAT, M. & CASEY, J.F. 1995. An ultramafic lift at drafting some of the figures. This work was also the Mid-Atlantic Ridge: successive stages of magsupported by the Outstanding Youth Scientific Fund of matism in serpentinized peridotites from the 15°N China (49625203). region. In: VISSERS, R.L.M. & NICOLAS, A. (eds) Mantle and Lower Crust Exposed in Oceanic Ridges and in Ophiolites. Kluwer, Dordrecht, 5-34. References CONSTANTIN, M. 1999. Gabbroic intrusions and magAGRINIER, P., JAVOY, M. & GIRARDEAU, J. 1988. matic metasomatism in harzburgites from the GarHydrothermal activity in a peculiar oceanic ridge: rett transform fault: implications for the nature of oxygen and hydrogen isotope evidence in the the mantle-crust transition at fast spreading ridges. Xigaze ophiolite (Tibet, China). Chemical Geology, Contributions to Mineralogy and Petrology, 136, 71,313-335. 111-130. AITCHISON, J.C., BADENGZHOU, & DAVIS, A.M. ET AL. CORNEN, G., BESLIER, M.-O. & GIRARDEAU, J. 1996. 2000. Remnants of a Cretaceous intra-oceanic Petrology of the mafic rocks from the ocean/ subduction system within the Yarlung-Zangbo sucontinent transition in the Iberia Abyssal Plain. In: ture (southern Tibet). Earth and Planetary Science WHITMARSH, R.B., SAYER, D.S., KLAUS, A. & Letters, 183,231-344. MASSON, D.G. (eds) Proceedings of the Ocean ALLAN, J.F. 1994. Cr-spinel in depleted basalts from the Drilling Program, Scientific Results, 149. Ocean lau backarc basin: Petrogenetic history from Mg-Fe Drilling Program, College Station, TX, 449-469. crystal-liquid exchange. In: HAWKINS, J., PARSON, DICK, H.J.B. & BULLEN, T. 1984. Chromian spinel as a L., ALLAN, J., ET AL. (eds) Proceedings Ocean petrogenetic indicator in abyssal and alpine-type Drilling Program, Scientific Results, 135, Ocean peridotites and spatially associated lavas. ContribuDrilling Program, College, Station, Texas, 565-583. tions to Mineralogy and Petrology, 86, 54-76. ALLAN, J.F., SACK, R.O. & BATIZA, P. 1988. Cr-rich GANSSER, A. 1974. The ophiolitic melange, a world-wide spinels in petrogenetic indicators: MORB-type lavas problem on Tethyan examples. Eclogae Geologicae from the Lamont Seamount Chain, eastern Pacific. Helvetiae, 67, 479-507. American Mineralogist, 73, 741-753. GANSSER, A. 1991. Facts and theories on the Himalaya. ALLEGRE, C.J., COURTILLOT, V. & TAPPONNIER, P. ET Eclogae Geologicae Helvetiae, 84, 33—60. AL. 1984. Structure and evolution of the HimalayaGIRARDEAU, J. & MERCIER, J.-C.C. 1988. Petrology and Tibet orogenic belt. Nature, 307, 17-22. texture of the ultramafic rocks of the Xigaze ARAI, S. 1992. Chemistry of chromian spinel in volcanic ophiolite (Tibet): constraints for mantle structure

189

MINERAL CHEMISTRY OF TIBETAN OPHIOLITES beneath slow-spreading ridges. Tectonophysics, 147, 33-58. GIRARDEAU, J., MARCOUX, J. & ZAO, Y.G. 1984. Lithological and tectonic environment of the Xigaze ophiolite (Yarlung Zangbo suture zone, Southern Tibet, China): kinematics of its emplacement. Eclogae Geologicae Helvetiae, 77, 153-170. GIRARDEAU, J., MERCIER, J.C.C. & WANG, X.B. 1985a. Petrology of the mafic rocks of the Xigaze ophiolite, Tibet. Contributions to Mineralogy and Petrology, 90, 308-321. GIRARDEAU, J., MERCIER, J.C.C. & ZAO, Y.G. 1985b. Structure of the Xigaze ophiolite, Yarlung Zangbo suture zone, southern Tibet, China: genetic implications. Tectonics, 4, 267-288. GIRARDEAU, J., MERCIER, J.C.C. & ZAO, Y.G. 1985c. Origin of the Xigaze ophiolite, Yarlung Zangbo suture zone, southern Tibet. Tectonophysics, 119, 407-433. GOPEL, C., ALLEGRE, C.J. & Xu, R.H. 1984. Lead isotopic study of the Xigaze ophiolite (Tibet): the problem of the relationship between magmatites (gabbros, dolerites, lavas) and tectonites (harzburgites). Earth and Planetary Science Letters, 69, 301-310. HEBERT, R., WANG, C.S. & Liu, Z.F. 1999. Xigaze ophiolites, southern Tibet revisited. GAC-MAC Annual Meeting, Sudbury, Abstract Volume, 24, 52-53. HEBERT, R., BEAUDOIN, G., VARFALVY, V., HUOT, F., WANG, C.S. & Liu, Z.F. 2000. Yarlung Zangbo ophiolites, southern Tibet revisited. 15th HimalayaKarakoram-Tibet Workshop, Chengdu, China. Earth Science Frontiers, 7, 124-126. HEBERT, R., GUEDARRI, K., LAFLECHE, M.R., BESLIER, M.-O. & GARDIEN, V. 200la. Petrology and geochemistry of exhumed peridotites and gabbros at non-volcanic margins: ODP Leg 173 West Iberia ocean-continent transition zone. In: WILSON, R.C.L., WHITMARSH, R.B., TAYLOR, B. & FROITZHEIM, N. (eds) Non-Volcanic Rifting of Continental Margins: a Comparison of Evidence from Land and Sea. Geological Society, London, Special Publications, 187, 161-189. HEBERT, R., WANG, C.S., VARFALVY, V., HUOT, F., Liu, Z.F., BEAUDOIN, G. & DOSTAL, J. 200 Ib. Yarlung Zangbo suture zone ophiolites and their suprasubduction zone setting. 16th Himalaya-KarakoramTibet Workshop, Sclauss Seggau, Austria. Journal of Asian Earth Sciences, 19, 27-28. HEBERT, R., WANG, C.S., VARFALVY, V., HUOT, F., Liu, Z.F., BEAUDOIN, G. & DOSTAL, J. 2001c. Yarlung Zangbo suture zone ophiolites, Tibet, and their suprasubduction zone setting. GAC-MAC Annual Meeting, St. John s, Abstract Volume, 26, 60. HEKINION, R. ACKERMOND, D. & STOFFENS, P. 1995. Mafic and ultramafic intrusions into upper mantle peridotites from fast spreading centre of the Eastern Microplate (South East Pacific). In: VISSERS, R.L.M. & NICOLAS, A. (eds) Mantle and Lower Crust Exposed in Oceanic Ridges and in Ophiolites, Kluwer, Dordrecht, 71-120. HIROSE, K. & KAWAMOTO, T. 1995. Hydrous partial melting of Iherzolite at 1 GPa: the effect of H2O on

the genesis of basaltic magmas. Earth and Planetary Science Letters, 133, 463-473. Hu, X.F., ROBINSON, P.T., BAI, W.J., FANG, Q. & CAMERON, S.T. 1998. Mineralogy of diamondbearing chromitites, Luobusa ophiolite, southern Tibet. GAC-MAC-APGGQ-IAH-CGUAnnual Meeting, Quebec City, Abstract Volume, 23, 81-82. HUOT, F., HEBERT, R. & VARFALVY, V. ET AL. 2003. The Beimarang melange (southern Tibet) brings additional constraints in assessing the origin, metamorphic evolution and obduction processes of the Yarlung Zangbo ophiolite. Journal of Asian Earth Sciences, Special Issue, 21, 307-322. ISHII, T., ROBINSON, P.T., MAEKAWA, H. & FISKE, M. ET AL. 1992. Petrological studies from diapiric serpentinite seamounts in the Izu-OgazawaraMariana forearc. In: FRYER, P., PEARCE, J.A. & STOKKING, L.B. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 445-485. JAQUES, A.L. & GREEN, D.H. 1980. Anhydrous melting of peridotite at 0-15 kbar pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology, 73, 287-310. JOHNSON, K.T.M., DICK, H.J.B. & SHIMIZU, N. 1990. Melting in the oceanic upper mantel: An ion microprobe study of diopsides in abyssal peridotites. Journal of Geophysical Research, 95, 2661-2678. JUTEAU, T., EAGER, E. & CANNAT, M. 1990. Serpeutinized residual mantle peridotiter from the M.A.R. median valley, ODP Hole 670A (21°02'W, leg 109): Primary mineralogy and geothermometry. In: DETRICK,

R.,

HONNOREZ, J. &

JUTEAU, T.

(eds)

Proceedings Ocean Drilling Program, Scientific Results, 106/109, Ocean Drilling Program, College, Station, Texas, 27^45. KOMOR, S.C., GROVE, T.L. & HEBERT, R. ET AL. 1990. Abyssal peridotites from Hole 670A: residue of mantle melting. In: DETRICK, R., HONNOREZ, J. & JUTEAU, T. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 106/109. Ocean Drilling Program, College Station, TX, 85-101. LAURENT, R. & KACIRA, N. 1987. Chromite deposits in the Appalachian ophiolites. In: STOWE, C.W. (ed.) Evolution of Chromium Ore Fields. Van Nostrand Reinhold, New York, 169-193. LEAKE, B.E. 1978. Nomenclature of amphiboles. Canadian Mineralogist, 16, 501-520. LEHMANN, J. 1983. Diffusion between olivine and spinel: Application to geothermometry. Earth and Planetary Science Letters, 64, 123-138. LINDSLEY, D.H. 1983. Pyroxene thermometry. American Mineralogist, 68, 477-493. MARCOUX, J., DE WAEVER, P. & NICOLAS, A. ET AL. 1982. Preliminary report of depositional sediments on top of volcanic member: Xigaze ophiolite (Yarlung Zangbo suture zone); south Xigang (Tibet). Ofioliti, 7, 395-396. MERCIER, J.L. & Li, G.C. 1984. Mission franco-chinoise au Tibet 1980. CNRS, Paris. MOLNAR, P. & TAPPONNIER, P. 1975. Cenozoic tectonics of Asia: effects of a contiental collision. Science, 189, 419-426.

190

R. HEBERT ETAL.

MORIMOTO, N. & SUBCOMMITTEE ON PYROXENES 1989. Nomenclature of pyroxenes. Canadian Mineralogist, 27, 143-156. MURPHY, M.A., YIN, A., KAPP, P.A., HARRISON, T.M., RYERSON, F.J., CHEN, Z.L. & WANG, W.F. 1997. Miocene evolution of the Kailar thrust and Gurla Mandhata detachment fault, NW Tibet: Implications for slip history of the Karakoram fault. Geological Society of America, Abstract with Programs, 29, 143. NICOLAS, A. 1989. Structures of Ophiolites and Dynamic of Oceanic Lithosphere. Kluwer, Kluwer, Dordrecht. NICOLAS, A., BOUDIER, F. & BOUCHEZ, J.L. 1980. Interpretation of peridotite structures from ophiolitic and oceanic environments. American Journal of Science, 280, 192-210. NICOLAS, A., GIRARDEAU, J. & MARCOUX, J. ET AL. 1981. The Xigaze ophiolite (Tibet): a peculiar oceanic lithosphere. Nature, 294, 414-417. PEARCE, J.A., ALABASTER, T., SHELTON, A. & SEARLE, M.P. 1981. The Oman ophiolite as a Cretaceous arc-basin complex: evidence and implications. Philosophical Transactions of the Royal Society of London, Series A, 300, 299-317. Pozzi, J.P., WESTPHAL, M., GIRARDEAU, J., BESSE, J., ZHOU, Y.X., CHEN, X.Y. & Li, S.X. 1984. Paleomagnetism of the Xigaze ophiolite and flysch (Yarlung Zangbo suture zone), southern Tibet): latitude and direction of spreading. Earth and Planetary Science Letters, 70, 383-394. PROUST, F., BURG, J.P., MATTE, P. & TAPPONNIER, P. 1984. Succession des phases de plissement sur une transversale du Tibet meridional: implications geodynamiques. In: MERCIER, J.L. & Li, G.C. (eds) Mission franco-chinoise au Tibet 1980. CNRS, Paris, 385-392. SOBOLEV, A.V. & BATANOVA, V.G. 1995. Mantle Iherzolites of the Troodos ophiolite complex, Cyprus: clinopyroxene geochemistry. Petrology, 3,440-448. TAKAHASHI, E., UTO, K. & SCHILLING, J.G. 1987. Primary Magma Composition and Mg/Fe Ratios of their Mantle Residues along Mid Atlantic ridge 29°N to 73°N. Technical Report ISEI, Okayama University, A-9, 1-12. TAPPONIER, P., MERCIER, J.-L. & PROUST, F. ET AL. 1981. The Tibetan side of the India-Eurasia collision. Nature, 294, 405-410. THY, P. 1983. Spinel minerals in transitional and alkali basaltic glasser from Iceland. Contributions to Mineralogy and Petrology, 83, 141-149. VAN DER LAAN, S.R., ARCULUS, R.J., PEARCE, J.A. & MURTON, BJ. ET AL. 1992. Petrography, mineral chemistry, and phase relations of the basement boninite serie of Site 786, Izu-Bonin forearc. In: FRYER, P., PEARCE, J.A. & STOKKINGS, L.B. (eds)

Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 171-201. VARFALVY, V. 2000. Interactions magma-manteau dans les peridotites du massif de North Arm Mountain, complexe ophiolitique de Bay of Islands, Terre Neuve, Canada: implications sur la genese des magmas en contexte de subduction. Ph.D. thesis, Universite Laval, Sainte-Foy, Que. VARFALVY, V., HEBERT, R. & BEDARD, J.H. 1996. Interactions between melt and upper-mantle peridotites in the North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland, Canada: implications for the genesis of boninitic and related magmas. Chemical Geology, 129, 71-90. VARFALVY, V., HEBERT, R., BEDARD, J.H. & LAFLECHE, M.R. 1997. Petrology and geochemistry of pyroxenite dykes in upper mantle peridotites of the North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland: implications for the genesis of boninitic and related magmas. Canadian Mineralogist, 35, 543-570. WANG, C.S., Liu, Z.F. & HEBERT, R. 1999. Xigaze Forearc Basin and Yarlung Zangbo Suture Zone, Tibet. Geological Publishing House, Beijing. WANG, C.S., Liu, Z.F. & HEBERT, R. 2000. The Yarlung Zangbo paleo-ophiolite, southern Tibet: implications for dynamic evolution of the Yarlung Zangbo suture zone. Journal of Asian Earth Sciences, 18, 651-661. WANG, X.B., XIAO, X.C., CAO, Y.G. & ZHENG, H.X. 1984. Geological map of the ophiolite zone along the middle Yarlung Zangbo (Tsangpo River), Xizang (Tibet). Academy of Geological Sciences, Beijing. Wu, H.R. & DENG, W.M. 1980. Basic geological features of the Yarlung Zangbo ophiolite belt, Xizang, China. In: PANAYOTOU, A. (ed.) Proceedings International Ophiolite Symposium, Cyprus, 1979. Ministry of Agriculture and Natural Resources, Geological Department, Nicosia, 462-472. XIAO, X.C. 1980. Les ophiolites de Xigaze au Tibet meridional et quelques problemes tectoniques les concernant. In: MERCIER, J.L. & Li, G.C. (eds) Mission franco-chinoise au Tibet 1980. CNRS, Paris, 167-188. ZHOU, M.F., ROBINSON, P.T., MALPAS, J. & Li, Z.J. 1996. Podiform chromitites in the Luobusa ophiolite (Southern Tibet): implications for melt-rock interaction and chromite segregation in the upper mantle. Journal of Petrology, 37, 3-21. ZlABREV, S.V., AlTCHISON, J.C. & BADENGZHU, ET AL.

2000. Oceanic deposits in the Yarlung-Tsangpo suture zone: structural setting, radiolarian ages and their tectonic implications (abstract). Earth Science Frontiers, 7, 118.

Geochemical and geochronological constraints on the origin and emplacement of the Yarlung Zangbo ophiolites, Southern Tibet JOHN MALPAS 1 , MEI-FU ZHOU 1 , PAUL T. ROBINSON 1 ' 2 & PETER H. REYNOLDS 2 1 Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong, PR. China (e-mail: [email protected]) 2 Department of Earth Sciences, Dalhousie University, Halifax, NS, Canada B3H 3J5 Abstract: The Indus-Yarlung Zangbo suture zone in southern Tibet marks the Eocene collision of the Indian continent and the Lhasa Block of Eurasia. It is characterized, particularly in its central portion, by an east-west belt of ophiolitic and related oceanic volcanic and sedimentary rocks that form a number of structurally juxtaposed geological terranes. Although tectonically disrupted in many places, almost complete ophiolite sequences exist at Luobusa and Zedong in the east and near Xigaze in the west. In Luobusa, the ophiolite sequence is thrust over the Tertiary molasse deposits of the Luobusa Formation or onto plutonic rocks of the Gangdese batholith. A mantle sequence dominates the ophiolite massif and consists chiefly of harzburgite and clinopyroxene-bearing harzburgite with abundant podiform chromitites enveloped by dunite. The Luobusa ophiolite formed the basement to an intra-oceanic volcanic arc, the Zedong terrane, which developed between the Mid-Jurassic and Mid-Cretaceous. Farther to the west, complete ophiolite sequences exist at Dazhuqu and near Xigaze. These ophiolites have suprasubduction zone geochemical signatures but there is no apparent development of a volcanic arc. Sensitive high-resolution ion microprobe U-Pb zircon analyses yield an age of 126 Ma for the crystallization of a quartz diorite from the Dazhuqu massif. Amphibolites that occur as large blocks in melanges at the base of the ophiolites are considered to be remnants of dynamothermal metamorphic soles produced early in the ophiolite obduction process. Ar/Ar geochronology on amphibole and biotite separates from these rocks yields ages of 80 and 90 Ma, respectively, for this event, which is considered to have occurred as the Indian continental margin entered the intra-oceanic subduction zone. Continued northward subduction of the remaining portion of the Neo-Tethyan ocean floor beneath the southern margin of Eurasia produced the Gangdese continental arc on the southern margin of the Lhasa Block and led to the final closure of the ocean with the collision of India and Eurasia in the Eocene.

The tectonic history of the collision between India and Asia that produced the Tibetan Plateau has been extensively studied over the past few decades (e.g. Yin & Harrison 2000, and references therein). Such investigations have not only advanced our understanding of the genesis of the Himalayas and the development of the Tibetan Plateau, but have extended our knowledge of the processes associated with uplift and exhumation of collision orogens everywhere. In the Early and Mid-Mesozoic, the Neo-Tethyan ocean, as much as 7000 km wide in a northsouth direction (Metcalfe 1999), separated the southern margin of Eurasia from the northern, passive margin of India, which was still part of Gondwanaland. Although most workers agree that the Neo-Tethyan ocean gradually disappeared by northward subduction along the southern margin of Eurasia, a process that finally resulted in

India-Asia collision, the geometry and timing of the subduction and collision are poorly constrained and have been subject to various interpretations. Existing models suggest a Late Cretaceous to Early Cenozoic subduction system along the Asian margin that resulted in final closure of the ocean in the Eocene and the formation of a series of east-west sutures of which the Indus-Yarlung Zangbo Suture (IYS) is the most significant (Allegre et al. 1984; Xiao 1988; Wang et al. 2000). The IYS is characterized by a discontinuous belt of ophiolitic bodies stretching east-west across southern Tibet and beyond. The ophiolites are everywhere tectonically disrupted but almost complete sequences are found at Luobusa, Zedong and near Xigaze (Fig. 1). In this paper, for the first time, the formation of one of the ophiolite massifs is precisely dated as

From: DiLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 191-206. 0305-8719/037$ 15 © The Geological Society of London 2003.

192

J. MALPAS ETAL.

Fig. 1. Geological map of the Indus-Yarlung Zangbo suture zone, southern Tibet, showing the distribution of geological terranes, including Dazhuqu terrane ophiolites, between Luobusa and Xigaze (modified from Aitchison et al. 2002). (a) Luobusa to Zedong; (b) Renbung to Donghia. YZSZ, Yarlung Zangbo suture zone; BNS, BanggongNujing Suture.

Early Cretaceous (126 Ma) using the sensitive high-resolution ion microprobe (SHRIMP) U-Pb zircon method. Additional Ar/Ar analyses suggest that the ophiolites were displaced originally between 80 and 90 Ma (Late Cretaceous) and later exhumed in the Early Neogene. Together with new geochemical analyses, these and other available ages allow us to improve previous models of the tectonic evolution of the IYS ophiolites and the collision between India and Eurasia.

Geological setting Regional geology The Tibetan Plateau comprises a number of eastwest-striking terranes (e.g. Dewey et al. 1989; Aitchison et al. 2000), separated by well-defined suture zones (Table 1). The IYS is the southern-

most of these and separates the Lhasa Block to the north from the Indian continent to the south. It is characterized by an almost continuous belt of ophiolites and related rocks (Dazhuqu terrane of Aitchison et al. 2000), a series of island-arcrelated rocks (Zedong terrane), slices of red and grey ribbon cherts and siliciclastic rocks (Bainang terrane), and a variety of coarse clastic sedimentary rocks (Liuqu conglomerate) and melange. The Lhasa terrane was itself accreted to the Qiangtang terrane to the north along the Banggong-Nujing Suture (BNS), characterized by ophiolites emplaced at 175-180 Ma (Zhou et al. 1997).

Dazhuqu terrane ophiolites Ophiolitic rocks are well exposed in the central part of the IYS particularly at Luobusa and

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

193

Table 1. Major terranes of the Indus-Yarlung Zangbo Suture showing their rock assemblages, tectonic settings and ages of formation (see text for sources of age data) Terrane

Assemblage

Tectonic setting

Ages

Indian terrane Bainang terrane

Flysch sedimentary rocks Radiolarian chert sequences

Passive continental margin Accretionary complex

Permian to Cretaceous Triassic to Early Cretaceous by fossils (Zyabrev et al. 2002) 126 Ma (U/Pb zircon) at Dazhuqu, Barremian by fossils, 170 Ma by Sm/Nd whole rock at Luobusa; 70-90 Ma (Ar/Ar amphibole) 162 Ma (U/Pb zircon) to 127 Ma (Ar/Ar hornblende) Mid-Late Cretaceous

Dazhuqu terrane Ophiolitic rocks

Suprasubduction zone environments

Zedong terrane

Volcanic rocks

Intra-oceanic arc

Xigaze terrane

Siliciclastic sedimentary rocks

Lhasa terrane

Forearc deposits on Lhasa terrane Gangdese plutonic and volcanic Continental arc rocks

Zedong to the east and near Xigaze to the west. As a result of the tectonic disruption that occurred during their emplacement, the original relationships between the Ophiolitic rocks and those of adjacent terranes are conjectural. Clearly, there has been convergence and collision of large-scale crustal blocks, but the history of this convergence and the nature of what once lay between the continents are not clear. For example, despite arguments for subduction of thousands of kilometres of oceanic crust, only recently has a subduction-related accretionary complex been recognized south of the ophiolite belt (the Bainang terrane of Aitchison et al 2000).

150 Ma to 41 Ma (U/Pb and AT/ Ar whole rock)

Luobusa massif and ophiolitic rocks near Zedong The Luobusa massif comprises a thrust complex at least 1 km thick, emplaced northward onto Oligocene-Lower Miocene molasse deposits (Luobusa Formation) and the Gangdese batholith, and technically overlain by Triassic flysch deposits to the south (Fig. 2). The northern contact with the Luobusa Formation generally dips southwards at angles of 15-20° whereas the southern contact is much steeper. Immediately north of the Yarlung Zangbo River, a small klippe of Luobusa ultramafic rocks lies directly on Gangdese granites

Fig. 2. Geological map of the Luobusa ophiolite, SE Tibet (after Zhou et al. 1996).

194

J. MALPAS ETAL.

with an almost horizontal thrust contact. To the south of the river, the sandstones and conglomerates of the Luobusa Formation are locally deformed, particularly along the thrust contact with the overlying ophiolite. From the base upwards, the ophiolite complex comprises a melange with knockers of pillow lava, amphibolites and peridotites, a suite of transition zone lithologies and a mantle sequence. Thus, the overall ophiolite stratigraphy is inverted, and the lithostratigraphies within units suggest that they are also overturned. The boundaries between these units are marked by fault breccia zones several metres thick that dip shallowly southwards. Together, these features suggest that each unit is an individual, downward-facing thrust slice. The melange consists of blocks and knockers of wehrlite and pyroxenite, layered and varitextured gabbros, pillow lavas, amphibolite and midCretaceous marine sedimentary rocks of the Zedong Formation, in a serpentinized ultramafic matrix. The ultramafic cumulate lithologies and gabbros are considered to be derived from the overriding transition zone thrust sheet. Based on their geochemistry, as discussed below, the pillow lavas are probably from the top of the ophiolite. The amphibolites occur as large blocks directly beneath the ultramafic rocks and are similar to metamorphic rocks that occur in basal dynamothermal aureoles formed beneath ophiolites elsewhere during their initial displacement. There are no metamorphic rocks of this nature in any of the related terranes from which these rocks could have been derived. The blocks of marine sedimentary rocks are probably derived from the nearby Zedong island-arc terrane. The melange therefore is thought to represent the further dismemberment of an already complex juxtaposition of lithologies from the top to the base of an ophiolite complex, and a related arc sequence. The melange is structurally overlain by transition zone dunites and pyroxenites (Zhou et al. 1996). Chromitites are found in the transition zone rocks but are much thinner and sparser than those found in dunite pods in the mantle sequence above. Very little of the upper, more felsic part, of the transition zone is observed in place, but gabbroic and dioritic cumulate rocks are present as knockers in the underlying melange. The transition zone sequence presumably originally superseded the mantle sequence in the ophiolite stratigraphy, but is now found structurally beneath ultramafic tectonites. The mantle sequence makes up the bulk of the ophiolite and is composed chiefly of deformed harzburgite and clinopyroxene-bearing harzburgite. Abundant lenses and pods of dunite and chromitite (Zhou et al 1996) associated with a series of ultra-high-

pressure minerals (Bai et al. 1993, 2000) occur mainly toward the top of the sequence. The mantle rocks display a series of metamorphic fabrics ranging from those produced at high temperature and high pressure (mantle tectonites), to subsequent low-pressure-low-temperature cataclasites (Zhou et al. 1996). The geochemistry and mineralogy of these rocks suggest that the ophiolite was formed in a two-stage process: spreading at a midocean ridge followed by magmatism associated with intra-oceanic subduction (Zhou et al. 1995). The ophiolite is structurally overlain by flysch of Triassic age derived from the Indian Plate, and can therefore be considered as part of a largescale, northward-directed duplex thrust stack sandwiched between the underlying Eurasian basement and the overriding Indian margin sedimentary rocks. Approximately 50 km west of Luobusa, near the town of Zedong, a similar structural arrangement exposes ophiolitic rocks beneath northward-thrust Triassic flysch deposits. Here, however, red ribbon cherts of the Bainang terrane (Triassic to Lower Cretaceous) are found as a thrust slice between the flysch and the ophiolite. In places, the ophiolitic rocks are also structurally underlain by igneous and sedimentary rocks of the Zedong terrane that are located between the ophiolite and rocks of the Lhasa terrane (McDermid et al. 2002; Fig. 3). The ophiolitic rocks include serpentinite, gabbros, sheeted diabases and sparse pillow lavas apparently in a right-way-up, condensed stratigraphic section. The Zedong terrane contains a suite of igneous and volcaniclastic rocks with island-arc affinities (Aitchison et al. 2000). In this area, a steeply dipping, inverted succession of boninitic pillow lavas, andesitic and dacitic dykes, gabbro, diorite and quartz diorite stocks, is associated with interbedded radiolarian cherts, andesitic breccias and tuffs, overlain by marine siliciclastic sedimentary rocks (Zedong Formation).

Dazhuqu massif, Xigaze ophiolite and sections SWofSagui The Dazhuqu massif is exposed along the Yarlung Zangbo River c. 50km east of Xigaze (Fig. 1). There, the ophiolite consists of a 2 km thick succession of stacked thrust slices, which are little deformed internally. The rock units, from the base of the section upward, include peridotite, ultramafic and mafic cumulate rocks, diabase dykes, and pillowed and massive lava flows. The pillow lavas are capped by several metres of radiolarian chert, which indicate a subequatorial location of formation from palaeomagnetic studies (Abrajevitch et al. 2001). Thus, a virtually complete ophiolite

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

195

Fig. 3. Geological map of the area immediately west of Zedong (after McDermid et al. 2002).

section is present. Our recent mapping indicates that, to the north, the ophiolite is thrust northward over siliciclastic turbidites of the Late Cretaceous Xigaze terrane, although a depositional relationship was previously reported (Nicolas et al. 1981; Girardeau et al. 1984). To the south, it is in fault contact with the overlying Triassic flysch deposits of the Indian Block. The mantle sequence consists of Iherzolite and clinopyroxene-bearing harzburgite with minor pods of dunite. A sequence of cumulate rocks overlies the mantle rocks, comprising dunite, troctolite and olivine gabbro. These are succeeded by varitextured gabbros, diorites and plagiogranites, and, on the northern side of the complex, sheeted dykes. The dykes grade upward into pillow lavas with minor sills and sheet flows. Farther west, the Xigaze ophiolite is formed of a number of individual massifs running generally east-west in a belt some 25 km south of the town of Xigaze. These massifs expose sections of mantle and crustal rocks averaging 10 km wide (as much as 25 km in places) generally facing to the north. The belt runs eastward towards Bainang and then NE to Dazhuqu (Fig. 1). To the west, the ophiolite section thins to about 5 km, SW of Sagui, where excellent exposures exist in a north-south valley south of Donghia. Thus, the Xigaze ophiolite forms a belt almost 130 km long, but in which the stratigraphy has been disrupted by a series of structural events including thrusting and post-emplacement normal and strike-slip faulting. Igneous rocks of the ophiolite are overlain by chert, mudstone, felsic tuffs and fine-grained

volcaniclastic sediments. These are considered part of the ophiolitic terrane and conformably overlie pillowed and massive lavas that pass downward through diabase dykes, into serpentinized dunite and harzburgite. Clinopyroxene-bearing harzburgite is the dominant ultramafic rock. It lies about 2 km structurally below the mafic unit, and thus about 5 km beneath the sedimentary cover of the ophiolite sequence. The harzburgite grades downward into a more Iherzolitic peridotite. In places, the ultramafic rocks are intruded over a thickness of about 1 km by thick diabase sills, which become progressively less abundant down section. Locally, the ultramafic rocks are also cut by diabase dykes that are in places metamorphosed to epidote-amphibolite facies. Both the sills and dykes are compositionally similar to the pillow lavas at the top of the section. The southern margin of the ophiolite is marked by a serpentinite melange where it is faulted against the red ribbon cherts of the Bainang terrane. The melange contains blocks of harzburgite, Iherzolite, diabase and a large mass of garnet amphibolite akin to those metamorphic rocks found in the melange at the base of the Luobusa ophiolite, and which are thought to be remnants of a dynamothermal metamorphic sole produced during ophiolite displacement. The Bainang terrane comprises radiolarian cherts of Late Jurassic to Early Cretaceous age, arranged in a series of north-facing, south-verging imbricate thrust slices, reminiscent of a subduction zone accretionary wedge. Some of the cherts are intruded by mafic sills related to intra-plate oceanic alkaline magmatism (Aitchison et al. 2002).

196

J. MALPAS ETAL.

The nature of the northern margin of the Xigaze ophiolite is equivocal because it is nearly everywhere hidden beneath Quaternary sediments. Earlier workers claimed that the turbidites of the Xigaze terrane were deposited on ophiolitic basement (Girardeau et al. 1985). Although these siliciclastic rocks do bear some resemblance to volcaniclastic sediments of the supra-ophiolite succession, our recent detailed field investigations (Aitchison et al. 2000) show that, everywhere it is exposed, the contact between the ophiolite and the Xigaze turbidites is a north-directed, south-dipping thrust with the ophiolite on top of the siliciclastic rocks. Thus, although the Xigaze terrane sediments were deposited in an oceanic basin at the southern edge of Eurasia, they are considered unrelated to the volcaniclastic sediments on top of the ophiolite. Volcanic detritus, which appears in the Xigaze turbidites at Aptian-Albian time, is synchronous with a period of volcanism in the Gangdese continental arc to the north. Palaeomagnetic data from the pillow lavas and supra-ophiolitic radiolarites of the Xigaze ophiolite show that they formed at a latitude of about 10-20°N (Pozzi et al. 1984), close to the southern margin of Eurasia during Mid-Cretaceous time.

Geochemistry Samples of basaltic lavas and sheeted dykes from sections of the Dazhuqu ophiolite terrane at Zedong (MRZ), Dazhuqu (D) and Xigaze (Y) have been analysed for major and trace elements. Representative samples of the pillow lavas from blocks in the melange zone at Luobusa (B), and from andesitic flows from the Zedong terrane (MZ), collected just east of the town of Zedong, have also been analysed (Table 2).

XRF and ICP-MS analyses Major element abundances were obtained using X-ray fluorescence (XRF) spectrometry on fused glass pellets at the University of Hong Kong. Trace elements Sc, V, Cr, Ni, Cu and Zn were also determined by XRF on pressed powder pellets. Other trace elements, including rare earth elements (REE), were analysed by inductively coupled plasma-mass spectrometry (ICP-MS) using a VG Elemental PlasmaQuad 3 system at the University of Hong Kong. The protocol of Jenner et al. (1990) was used, with pure elemental standards for external calibration, and BHVO-1 as an internal reference material. Accuracies of the XRF analyses are estimated as ±2% for major elements present in concentrations >0.5 wt.%; and ±5% for trace elements. The ICP-MS analyses yield accuracies better than ±5%. Major

oxides are normalized to 100% on a volatile-free basis.

Major element oxides and trace elements The basaltic volcanic rocks from the various ophiolite localities show a number of similar geochemical features. They all exhibit relatively low contents of Ni, Cr, Ti, Zr, Nb and Hf, but high, although variable, alkalis, Ba, Sr, U, Th and V relative to mid-ocean ridge basalt (MORB) (Table 2). They plot in an AFM diagram with an arc-tholeiite trend (Fig. 4) and in the island-arc tholeiite field on the Zr v. Zr/Y diagram of Pearce & Norry (1979) (Fig. 5). Their chondrite-normalized REE show moderate light REE (LREE) depletion (Fig. 6), spanning the range typical of MORB (Saunders 1984). There are, however, some differences between the lavas from Luobusa and the other massifs. The pillow lava blocks found in the melange at the base of the Luobusa massif show a more gentle negative slope in the LREE than the lavas and dykes from the other ophiolites, suggesting derivation from a less depleted mantle source. In the trace element spider diagrams (Fig. 7), the ophiolitic basaltic rocks from Zedong, Dazhuqu and Xigaze show concentrations of large ion lithophile elements up to 10 times MORB, clear depletion of Nb, and high field strength element concentrations of 0.3-1 times MORB. These patterns indicate that these basalts are island-arc tholeiites generated in a suprasubduction environment. The Luobusa samples, in contrast, have a much reduced Nb anomaly and generally higher contents of high field strength elements, supporting their generation from a less depleted source. All of these basalts are very different from the analysed samples from the Zedong terrane (Table 2), which are typical intra-oceanic arc andesites, with distinct trace element signatures as shown in Figures 6 and 7.

Geochronology Previous studies of the ophiolites along the IYS suggest that subduction took place during the Mid-Cretaceous (Coulon et al. 1986) or Late Cretaceous (Allegre et al. 1984; Dewey et al. 1989; Yin et al. 1994), and that the Neo-Tethyan ocean was finally closed during Paleogene continental collision (Molnar & Tapponnier 1975; Patriat & Achache 1984; Rowley 1996). Biostratigraphic dates have been obtained from supraophiolitic radiolarites in Xigaze (Zyabrev et al. 2002) suggesting this ophiolite formed in the MidCretaceous (Barremian). Radiometric ages of 177 ±31 Ma from diabase dykes cutting the cumulate section and Rb-Sr ages of 173 ± 11 Ma

Table 2. Major oxides (wt.%) and trace elemental abundances (ppm) of rocks from the Indus-Yarlung Zangbo Suture Zone Zedong ophiolite

Dazhuqu ophiolite

Xigaze ophiolites

Luobusa ophiolite

Zedong ;indesites

Sample

MRZ4

MRZ5

MRZ9

MRZ10

Bl

B2

B3

B4

Y7

Y10

Yll

Y15

Y16

Y17

D2

D9

MZ3

MZ8

SiO2 TiO2 A1203 Fe203 MnO MgO CaO Na20 K2O P205 Rb Ba Th U Nb Ta Sr Zr Hf Y V Cr Ni Cu Zn Sc La Ce Pr Nd Sm Eu Gd Tb

52.5 0.39 16.8 10.5 0.21 6.9 6.1 5.18 0.29 0.03 1.7 13 0.04 0.02 0.18 0.09 123 13 0.48 11 211 21 22 132 111 42 0.44 1.41 0.27 1.77 0.86 0.4 1.36 0.25 1.91 0.43 1.30 0.19 1.38 0.2

51.1 0.78 16.9 9.6 0.17 8.5 6.3 4.26 0.57 0.06 3.7 14 0.01 0 0.19 0.11 153 31 1.22 18 211 286 86 92 66 35 0.61 2.65 0.63 4.45 1.87 0.58 2.76 0.49 3.23 0.71 2.07 0.30 2.05 0.28

50.0 0.48 17.0 9.6 0.16 8.9 8.0 4.20 0.06 0.06 0.31 5 0.02 0.02 0.32 0.05 81 23 0.74 11 217 164 58 95 66 38 0.72 2.44 0.49 2.96 1.07 0.5 1.54 0.28 1.97 0.44 1.29 0.19 1.38 0.18

46.1 0.42 14.6 9.9 0.25 11.0 7.2 3.68 0.49 0.04 5.3 30 0.01 0.01 0.12 0.02 151 8 0.57 10 212 158 54 97 49 40 0.55 1.86 0.37 2.25 0.83 0.32 1.25 0.22 1.64 0.37 1.05 0.16 1.10 0.17

46.9 2 15.2 12.8 0.21 5.5 14.2 2.56 0.67 0.2 40 57 0.41 0.18 5.39 1.88 132 58 2.20 32

47.0 1 16.4 10.1 0.19 5.0 16.2 3.27 0.27 0.1 4.6 16 0.14 0.06 1.7 1.18 73 48 1.86 30

45.8 1 16.6 8.2 0.15 6.4 21.8 0.11 0.00 0.1 0.17 16 0.10 0.05 5.8 2.98 109 51 1.54 18

49.7 2 14.5 11.5 0.18 6.8 11.0 3.85 0.56 0.2 11 35 0.19 0.09 2.1 1.06 57 56 2.23 35

6.7 17 2.74 13.4 4.09 1.59 5.61 0.90 6.06 1.25 3.68 0.51 3.34 0.49

3.3 9.8 1.72 9.31 3.29 1.07 4.86 0.84 5.42 1.16 3.68 0.51 3.28 0.47

1.9 6.0 1.05 5.57 2.16 0.91 2.95 0.51 3.42 0.75 2.24 0.31 1.94 0.32

4.4 13 2.36 12.6 4.17 1.56 6.02 0.99 6.58 1.38 4.15 0.57 3.69 0.53

50.6 1.04 13.6 10.7 0.18 8.4 7.1 3.23 1.07 0.11 7.0 14 0.06 0.02 0.83 0.18 164 72 1.63 24 263 117 44 53 62 36 2.62 8.68 1.59 8.56 3.11 1.04 3.96 0.69 4.64 0.94 2.84 0.39 2.75 0.38

50.6 0.72 13.9 9.1 0.17 10.8 9.2 2.66 0.88 0.07 5.4 7 0.03 0.01 0.58 0.18 195 34 1.03 16 226 346 114 102 57 34 1.23 4.13 0.78 4.68 1.74 0.62 2.42 0.43 2.94 0.65 1.88 0.27 1.91 0.26

50.2 0.94 13.0 10.2 0.17 10.0 8.7 3.04 0.22 0.11 1.1 4 0.04 0.03 0.72 0.14 169 59 1.73 22 270 350 104 10 44 35 2.19 7.06 1.28 7.46 2.61 0.88 3.47 0.6 3.98 0.86 2.6 0.37 2.64 0.37

48.2 1.70 14.7 9.8 0.17 9.3 9.2 5.42 0.75 0.19 8.9 45 0.13 0.35 2.02 0.23 220 137 3.82 43 247 31 17 71 1001 31 4.40 14.35 2.73 15.23 5.28 1.73 6.67 1.15 7.65 1.61 4.71 0.69 4.77 0.66

50.8 0.97 16.2 9.5 0.15 5.4 12.2 4.03 0.64 0.10 10.1 23 0.16 0.1 0.93 0.31 255 59 1.94 21 224 244 51 52 69 35 2.05 6.42 1.21 6.98 2.47 0.93 3.24 0.56 3.75 0.8 2.35 0.34 2.36 0.34

51.5 1.05 16.3 9.7 0.13 7.0 6.9 5.11 0.18 0.12 1.7 40 0.07 0.06 1.11 0.19 180 74 2.3 24 224 182 54 31 58 38 2.54 8.23 1.49 8.31 2.9 1.04 3.77 0.65 4.49 0.97 2.73 0.39 2.71 0.37

51.1 0.86 16.1 9.2 0.14 7.8 7.5 4.69 0.02 0.10 0.09 3 0.06 0.03 0.75 0.12 98 59 1.51 21 211 107 47 70 74 34 1.90 6.33 1.15 6.38 2.28 0.84 3.10 0.55 3.77 0.8 2.32 0.33 2.27 0.33

51.7 0.49 16.3 6.1 0.12 9.4 11.1 2.42 1.57 0.04 5.7 12 0.02 0.02 0.36 0.1 272 23 0.68 12 196 647 107 84 35 45.3 0.95 2.73 0.57 3.46 1.35 0.54 1.88 0.35 2.35 0.49 1.4 0.21 1.38 0.19

57.3 0.69 17.2 7.3 0.09 4.3 6.1 3.96 3.64 0.39 46 868 3.88 0.86 2.68 0.41 712 151 1.98 17 197 60 19 52 29 21 24.2 57.4 8.19 39.5 7.73 1.74 5.63 0.64 3.42 0.61 1.63 0.21 1.45 0.18

57.8 0.84 15.8 9 0.14 5.5 7.2 2.89 1.84 0.35 27 226 2.83 0.65 2.59 0.56 619 107 1.33 19 257 54 20 115 59 29 14.8 34.9 4.99 22.6 5.2 1.34 4.53 0.64 3.82 0.76 2.09 0.29 1.99 0.28

Dy Ho Er Tm Yb Lu

198

J. MALPAS ETAL.

Fig. 4. AFM diagram of the basaltic rocks from Dazhuqu terrane ophiolite massifs. FeO*, total iron as FeO.

Fig. 5. Plot of Zr v. Zr/Y for basaltic rocks from the Dazhuqu terrane ophiolites. Fields of basalts from the various tectonic settings are from Pearce & Norry (1979). Symbols as in Figure 4.

for the pillow basalts of the Luobusa ophiolite (Zhou et al 2002), indicate that this massif may be somewhat older than Xigaze. We here report a number of additional age dates from a variety of rocks along the suture zone.

Analytical methods Dating of zircon was carried out by SHRIMP techniques whereas Ar/Ar analyses were carried out on hornblende and biotite. For the SHRIMP analyses, zircons were separated using conventional heavy liquid and magnetic techniques, and cathodoluminescence images were obtained using

Fig. 6. (a, b) Chondrite-normalized REE distribution patterns of basaltic rocks from the Dazhuqu terrane ophiolites. Symbols as in Figure 4. (c) Andesites from the Zedong terrane.

a Philips XL30 scanning electron microscope to investigate their internal structures. The instrumental techniques for isotopic analysis of zircons using the SHRIMP II ion microprobe at Curtin University of Technology are similar to those of Compston et al. (1984). Pb/U ages are based on a value of 564 Ma determined by conventional U-Pb analysis of the standard zircon CZ3. Uncertainties of 207Pb/206Pb ages are independent of the standard analyses, but

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

199

and uncertainties in mean ages are quoted at the 95% confidence level (2o). For the Ar/Ar dating, hornblende and biotite were handpicked from heavy liquid separates. The samples were irradiated in the McMaster University nuclear reactor and argon isotopic analyses were undertaken using a VG 3600 mass spectrometer coupled to an internal tantalum resistance furnace of the double-vacuum type, at Dalhousie University, Canada. Hornblende MMhb-1, with an assumed age of 520 ± 2 Ma (Samson & Alexander 1987) was used as a standard for all analyses. Other experimental procedures follow those described by Muecke et al. (1988). Errors are the 2o analytical uncertainties.

Dating results

Fig. 7. Spider diagrams of MORE-normalized trace element abundances in basaltic rocks from the Dazhuqu terrane (a, b) and andesites from the Zedong terrane (c). Symbols as in Figure 6.

are sensitive to the common Pb correction in lowU zircons that have been calculated for zircons that are < 1000 Ma. Thus, the 206pb_238u age is normally preferred. Common Pb was corrected using the 204 method discussed by Compston et al. (1984). U, Th and Pb concentrations were calculated using the methods given by ClaoueLong et al. (1991). Individual analyses (Table 3) are presented as lo error boxes on concordia plots

A quartz diorite, D13, from the cumulate section of the Dazhuqu massif, contains zircons with a variety of textures and morphologies. Most grains show distinct zoning but their internal complexity does not affect the ages obtained from different grains, whether from cores and rims, or from high- and low-U regions. All crystals, even those of different shape, give the same age within the uncertainties. The mean 206pb/238U age is 126 ±2 Ma, where the uncertainty is the 2a error on the mean. The results are concordant, with the mean 207Pb/235U age being 126 ± 3 Ma. The 207 Pb/206Pb age, which is very poorly defined in such young zircons, is 122 ± 47 Ma. The data are consistent with a single age population of zircons from D13. The x2 value for the 206Pb/238U age is 3.34, and for the 207Pb/235U age it is 0.46. The former value suggests there may be a very small amount of geological scatter in the 206Pb/238U ages. This is a result of the lowest-U zircons having 206Pb/238U ages around 107 Ma (see Table 3). These ages occur within both the cores and rims of the zircons analysed. The younger ages may reflect the difficulty of making 204Pb corrections for low-Pb samples as the mean 204Pb is likely to be an overestimate as a Poisson statistic. This problem would result in slightly lower 206Pb/ 238 U ages, and very young or negative 207pb/206Pb ages, which is the case for these few analyses. The rare low-U analyses do not perturb the mean age obtained for the high-U analyses. Based on the low x2 value of the 207Pb/235U age, it is assumed that the zircons are from a single Gaussian distribution. Figure 8 gives concordia plots of the data and shows a single group of concordant zircons. The low-U zircons have been removed so that the data for the high-U, lower uncertainty grains can be more easily seen. There is no evidence of any resetting of the ages since 126 Ma, thus the

Table 3. SHRIMP zircon analytical results for zircons from quartz diorite D13, from Dazhuqu massif, southern Tibet (uncertainties are lo) Spot

1 2 3 5 6 7 8 10 11 12 13 14 15 16 17 18 19

U (ppm)

341 50 288 221 565 92 82 56 430 830 795 149 483 130 76 124 264

Th (ppm)

854 31 1109

282 1308

202 205 172 2005 2080 1551

585 830 235 179 186 683

Pb

11 2 11 6 17 3 3 2 18 26 23 6 13 6 3 3 8

% cone.

94 0 53 126 78 0 0 0 143 179 315 0 102 109 0 0 0

Ages (Ma) 206pb_238U

±

207pb/235U

±

207pb/206pb

±

208pb/232U

±

129 107 124 126 127 119 108 108 127 130 130 128 125 124 117 127 123

2 10 2 2 2 3 5 5 2 2 2 3 2 5 5 3 2

130 66 129 125 129 83 59 38 125 127 125 115 124 124 48 113 101

9 159 3 17 9 40 77 83 3 5 6 29 3 71 72 39 21

138 0 232 100 162 0 0 0 89 72 41 0 122 114 0 0 0

155 46 49 309 164 82 53 92 38 95 92 51 39

125 63 121 125 124 112 100 110 122 123 127 133 118 120 106 130 123

3 105 2 6 3 8 14 13 2 2 2 5 2 17 14 11 4

1012

73 68 54

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

201

were formed at a convergent plate margin during the closure process. The formation age of the ophiolites in Xigaze was previously reported as 120 ± 10 Ma using conventional 238U/206Pb methods (Gopel et al. 1984) and 109 ± 21 Ma by the Nd/Sm method (Prinzhofer 1987, cited by Nicolas 1989). The SHRIMP date of 126 Ma for the Dazhuqu massif reported here is considered the most reliable age of formation of the Xigaze ophiolite. It is in accord with the Barremian biostratigraphic age reported by Zyabrev et al. (2002). However, evidence from the Luobusa and Zedong massifs suggests that these ophiolites are older (c. 175 Ma). Fig. 8. SHRIMP U-Pb zircon concordia plot for sample D13, a quartz diorite from the Dazhuqu massif. Amphibolites in the melange zones occur as large rafts directly below the ultramafic sections of the ophiolites, are basaltic in character, and are complex zonation of the zircons is interpreted as similar to metamorphic soles beneath ophiolites deuteric interaction of zircon and fluids during elsewhere. Because there is no other known source cooling and crystallization rather than later meta- for these rocks and they appear, particularly near morphic processes. Xigaze, to be disrupted locally, we interpret them A hornblende andesite from the Zedong terrane, to be fragments of once coherent metamorphic MZ2, has an Ar/Ar plateau age of 127.9 ± soles. Ar/Ar ages from amphiboles and biotites of 0.33 Ma (Fig. 9). When 40Ar/36Ar is plotted v. 94.8 Ma and 79.3 Ma, respectively, from amphibo39 Ar/36Ar, the age is 126.5 ± 0.35 Ma. lite blocks at Luobusa, and 87.9 Ma from the Samples BO-17 and LW1-2 are amphibolites amphibolites at Xigaze suggest that initial displafrom the melange zone at the base of the Luobusa cement of the ophiolites (Fig. lOc) was approxiophiolite. An amphibole separate from BO-17 has mately consanguineous along this part of the IYS. an Ar/Ar age of 85.7 Ma, whereas biotite in LW1The collision of India with Eurasia began in the 2 has an age of 80.6 Ma (Fig. 9). An amphibole Cenozoic at c. 55 Ma as shown by palaeomagnetic separate, A13, from a large knocker of meta- and other data (Nicolas et al. 1981; Allegre et al. morphic rocks at the base of the Xigaze ophiolite 1984; Patriat & Achache 1984), and the Neoyields a plateau age of 87.9 ± 0.4 Ma. A plot of Tethyan ocean basin was completely closed and 40Ar/36Ar v ^Ar/36Ar yielded an age of the ophiolites obducted before the end of the 88.0 ± 0.3 Ma. We tentatively interpret the amphi- Eocene (before 40 Ma) (Tapponnier et al. 1981). bolites as remnants of metamorphic soles and This is confirmed by the Ar/Ar biotite date of suggest that the dates obtained from these rocks 41 Ma for the Gangdese batholith near Luobusa, mark the age of initial displacement of the which probably represents waning arc magmatism ophiolites beneath the Zedong arc (Fig. lOc). in response to the final stages of subduction along A biotite separate from sample L36, a granite the southern Eurasian margin, and which is in from the Gangdese batholith near Luobusa, yields accord with granodiorites dated at 41 Ma (Scharer an Ar/Ar age of 41 Ma (Fig. 9). This is not unlike et al. 1984). In addition, a time gap of 30-40 Ma the ages previously obtained for the Gangdese between ophiolite displacement and final emplacemagmatism (Yin et al. 1994) and is much younger ment exists, during which there would have been than the interpreted displacement age of the accretion of ophiolite and arc terranes before their ophiolite. juxtaposition with the Eurasian margin.

Discussion Ophiolite formation, the closure of the NeoTethyan ocean and the collision between Eurasia and India A number of age determinations can be used to constrain the history of closure of the NeoTethyan ocean (Table 1). The suprasubduction zone signature of the ophiolites indicates that they

Tholeiitic and boninitic magmatism in the ophiolite suites The basaltic rocks from the ophiolite suites all show suprasubduction signatures. They are essentially island-arc tholeiites. In addition, accessory chromites in associated peridotites, particularly in Xigaze, are generally high-Al varieties (Cr numbers 22-65) indicating a tholeiitic affinity (see Dick & Bullen 1984). However, investigation of

202

J. MALPAS ETAL.

Fig. 9. Ar/Ar spectra of amphibole and biotite from rocks of the Yarlung Zangbo suture zone. See text for explanation.

the mineralogy and chemistry of the mantle peridotites and their podiform chromitite deposits from the Luobusa massif indicates that in addition to the arc moleiitic magmatism there was subsequent production of more depleted magmas (Zhou

et al. 1996). These chromitites have high-Cr chromite (Cr numbers 82-85) and low TiO2 (0.10.2 wt.%), suggesting crystallization from boninitic melts. Minor interstitial orthopyroxene in the chromitites also has low A^Os (1.0-1.2 wt.%),

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

203

Fig. 10. Schematic diagram of the plate tectonic evolution of the Neo-Tethyan ocean from c. 170 Ma to c. 26 Ma. (a, b) Northward intra-oceanic subduction of oceanic lithosphere initiates the formation of the Zedong island arc at c. 160 Ma. Farther to the north, northward subduction beneath the Eurasian continental margin (Lhasa Block) produces the Gangdese continental arc commencing at c. 155 Ma. (c) The Dazhuqu terrane ophiolites are thrust southward onto the Indian continental margin as the Indian Block collides with the intra-oceanic arc complex, (d, e) Continued subduction of Neo-Tethyan lithosphere to the north eventually juxtaposes the Indian Block and the Lhasa Block, and results in the northward emplacement of the ophiolites and associated arc rocks by 'flake tectonics'. B, Bainang terrane; D, Dazhuqu terrane; Z, Zedong terrane; X, Xigaze terrane; G, Gangdese arc (after Aitchison et al. 2000).

204

J. MALPAS ETAL.

supporting the view that they formed from boninitic melts. Percolation of these boninitic melts through the upper mantle and their focus along downward propagating cracks allowed reaction between the melts and the host peridotites. This reaction removed pyroxenes from the host rock, forming dunite dykes and dunite envelopes around the podiform chromitites. Modification of the melts by this process is believed to have triggered the precipitation of chromite in melt pockets now represented by the podiform bodies (Zhou et al. 1996). The boninitic melts reflect hydrous mantle melting in a suprasubduction zone environment (Dick & Bullen 1984) and formed either by high degrees of partial melting or by remelting of a progressively depleted source above a subduction zone (Crawford et al. 1989, and references therein). This suggests that the Luobusa podiform chromitites probably formed at depth beneath an island arc. The pyroxenite dykes associated with the podiform bodies show a progressive gradation from diopsidite to orthopyroxenite. These types of pyroxenite dykes occur in many erogenic Iherzolite massifs, such as that at Ronda, Spain, and indicate high-pressure fractionation, likewise suggesting formation beneath an island arc. The extensive melt-rock interaction reflects both disequilibrium between the melts and their host peridotites and the thickened lithosphere beneath an island arc. Passage of these melts through the Luobusa mantle section is believed to be the cause of high-temperature recrystallization. In simple terms, the available evidence indicates the presence of two magmatic suites in Luobusa, an older arc tholeiitic suite represented by the pillow lavas and a boninitic suite that gave rise to the podiform chromitites. Similar associations are common in many ophiolites produced in suprasubduction zone environments. The island arc originally above the Luobusa ophiolite is probably represented by the Zedong terrane, which includes andesites, dacites and boninites (McDermid et al. 2002), indicating the development of a mature arc sequence. The dates available for the Luobusa massif and the Zedong terrane appear to support this model. The Luobusa oceanic crust originally formed at c. 175 Ma and provided the basement for the development of the Zedong volcanic arc between the Mid-Jurassic and Mid-Cretaceous, during which time the mantle was modified by boninitic magmatism.

Tectonic model It has commonly been assumed that the ophiolites in the IYS represent lithosphere from the Neo-

Tethyan ocean (Nicolas et al. 1981; Allegre et al. 1984; Xiao 1988), and that they are the same age everywhere. However, Hsu et al. (1995) suggested that the IYS ophiolites are remnants of a number of collapsed back-arc basins, a view supported by our evidence that the ophiolites are of different ages. The difference in geochemistry between the lavas of Luobusa and the other ophiolites (Figs 6 and 7) also supports this view. We believe that the mantle peridotites in Luobusa were originally chemically similar to those in the other ophiolites, such as Xigaze, which are now found in the same suture zone, i.e. they had the composition of depleted MORB mantle trapped in suprasubduction wedges. Remelting of depleted peridotites in the wedge beneath the island arc of the Zedong terrane produced the boninitic magmas that formed the mantle chromitite deposits at Luobusa and the effusive volcanic rocks near Zedong (Fig. 10). To the west, no island-arc edifices were constructed and the ophiolitic rocks of Dazhuqu and Xigaze represent lithosphere extension above the subduction zone, contemporaneous with late arc volcanism at Zedong. The northward subduction of Neo-Tethyan oceanic lithosphere between the Indian margin and the intra-oceanic subduction complexes led to the accretion of the Bainang terrane to the south of the Dazhuqu terrane (Fig. lOa and b). Following the disappearance of the Neo-Tethyan oceanic lithosphere to the south of the subduction complex, continuous northward movement of the Indian Plate beneath the island arc provided a mechanism for the uplift of the Dazhuqu terrane ophiolites onto the Indian continental margin (Fig. lOc). Blocks of amphibolites in the melange zones beneath the ophiolites, probably derived from a metamorphic sole, suggest that the initial displacement of the ophiolites occurred at c. 80-90 Ma (i.e. Late Cretaceous) (Fig. lOc). This significantly predates India-Eurasia collision at c. 55 Ma (Eocene), which resulted in the closure of the NeoTethyan ocean and the formation of the Gangdese batholith as a result of northward subduction beneath the Eurasian margin (Fig. lOd). This suggests that the ophiolites were first accreted onto the Indian continent as its margin collided with and was partially subducted northward beneath the suprasubduction zone lithosphere. This modified Indian margin, including the accreted ophiolite terrane, was later juxtaposed against the Eurasian continental margin as a result of northward subduction of the remaining Neo-Tethyan lithosphere (Fig. lOd). Because of the subduction polarity, emplacement of the ophiolite onto the Eurasian continent must have been facilitated by a 'flake tectonic' mechanism resulting in northwarddirected thrusting (Fig. lOe).

ORIGIN AND EMPLACEMENT OF YARLUNG ZANGBO OPHIOLITES

Conclusions Formation of the Dazhuqu terrane ophiolites appears to have occurred at different times in suprasubduction zone environments associated with the collapse of the Neo-Tethyan ocean basin. Intra-oceanic subduction was an important component of this process of ocean closure, which was clearly more complicated than previously thought and perhaps longer lived. Given that the Luobusa massif shows a suprasubduction geochemical signature at c. 175 Ma, and the collision of India and Eurasia occurred at c. 55 Ma, there is a span of at least 120 Ma during which subduction of NeoTethyan oceanic lithosphere was taking place. Intra-oceanic subduction appears to have produced a mature island arc built on oceanic crust in some places (Luobusa), but resulted in lithosphere extension and the formation of ophiolites elsewhere (Xigaze), perhaps at the same time. The newly formed island arcs were probably the regions in which podiform chromitites formed as a result of melt-rock interaction. The Indian continent eventually collided in the Late Cretaceous with the subduction zone, resulting in the accretion of the Bainang, Dazhuqu and Zedong terranes to the continental margin and the displacement of the ophiolitic rocks. This juxtaposition of the intra-oceanic subduction-related terranes with the Indian continental margin significantly predated the final closure of the NeoTethyan ocean to the north. The collision of India and Eurasia resulted in emplacement of ophiolitic rocks onto the Eurasian margin and the production of the Gangdese batholith in the Eocene. This study was supported by research grants from the Research Grant Council of the Hong Kong SAR, China (HKU7086/01P to J.M.) and NSERC of Canada to P.T.R. and P.H.R. Fieldwork in Tibet in the last few years was assisted by H. Wu and Badengzhu from the Tibetan Geological Survey.

References ABRAJEVITCH, A., AITCHISON, J.C. & ALI, J. 2001. Paleomagnetism of the Dazhuqu Terrane, YarlungZangbo Suture Zone, Southern Tibet. EOS Transactions, American Geophysical Union Fall Meeting Supplement, Abstract GP11A-0118. AITCHISON, J.C., BADENGZHU, & DAVIS, A.M. ET AL. 2000. Remnants of a Cretaceous intra-oceanic subduction system within the Yarlung-Zangbo suture (southern Tibet). Earth and Planetary Science Letters, 183,231-244. AITCHISON, J.C., ABRAJEVITCH, A. & ALI, J.R. ET AL. 2002. New insight into the evolution of the Yarlung-Tsangpo suture zone, Xizang (Tibet), China. Episodes, 25, 90-94. ALLEGRE, C.J., COURTILLOT, V., TAPPONNIER, P. ET AL.

205

1984. Structure and evolution of the HimalayaTibet erogenic belt. Nature, 307, 17-22. BAI, W.-J., ZHOU, M.-F. & ROBINSON, P.T. 1993. Possible diamond-bearing mantle peridotites and chromitites in the Luobusa and Donqiao ophiolites, Tibet. Canadian Journal of Earth Science, 30, 1650-1659. BAI, W.-J., ROBINSON, P.T., FANG, Q., Hu, X., ZHOU, M.-F. & MALPAS, J. 2000. Origin of PGE and base metal alloys in podiform chromitites of the Luobusa ophiolite, southern Tibet. Canadian Mineralogist, 38, 585-598. CLAOUE-LONG, J.C., ZHANG, Z.C., MA, G.G. & Du, S.H. 1991. The age of the Permian-Triassic boundary. Earth and Planetary Science Letters, 105, 182-190. COMPSTON, W., WILLIAMS, I.S. & MEYER, C. 1984. UPb geochronology of zircons from Lunar Breccia 73217 using a sensitive high mass-resolution ion microprobe. Journal of Geophysical Research, 89, 525-534. COULON, C., MALUSKI, H., BOLLINGER, C. & WANG, S. 1986. Mesozoic and Cenozoic volcanic rocks from central and southern Tibet: 39Ar/40Ar dating, petrological characteristics and geodynamical significance. Earth and Planetary Science Letters, 79, 281-302. CRAWFORD, A.J., FALLOON, T.J. & GREEN, T.H. 1989. Classification, petrogenesis and tectonic setting of boninites. In: CRAWFORD, A.J. (ed.) Boninites and Related Rocks. Unwin Hyman, London, 1-49. DEWEY, J.F., CANDE, S. & PITMAN, W.C. 1989. Tectonic evolution of the India-Eurasia collision zone. Eclogae Geologicae Helvetica, 82, 717-734. DICK, HJ.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. GIRARDEAU, J., MARCOUX, J. & ZHAO, Y. 1984. Lithologic and tectonic environment of the Xigaze ophiolite (Yarlung Zangbo suture zone, Southern Tibet, China), and kinematics of its emplacement. Eclogae Geologicae Helvetica, 77, 153-170. GIRARDEAU, J., MERCIER, J.C.C. & WANG, X.B. 1985. Petrology of the mafic rocks of the Xigaze ophiolite, Tibet. Implications for the genesis of the oceanic lithosphere. Contributions to Mineralogy and Petrology, 90, 309-321. GOPEL, C., ALLEGRE, CJ. & Xu, R. 1984. Lead isotope study of the Xigaze ophiolite (Tibet): the problem of the relationship between magmatites (gabbros, dolerites, lavas) and tectonites (harzburgites). Earth and Planetary Science Letters, 6, 301-310. Hsu, K.J., PAN, G., SENGOR, A.M.G. ET AL. 1995. Tectonic evolution of the Tibetan Plateau: a working hypothesis based on the Archipelago model of orogenesis. International Geological Review, 37, 473-508. TENNER, G.A., LONGERICH, H.P., JACKSON, S.E. & FRYER, BJ. 1990. ICP-MS—a powerful tool for high precision trace element analysis in earth sciences: evidence from analysis of selected USGS reference samples. Chemical Geology, 83, 133-148. MCDERMID, I.R.C., AITCHISON, J.C., DAVIS, A.M.,

206

J. MALPAS ETAL.

HARRISON, T.M. & GROVE, M. 2002. The Zedong terrane: a Late Jurassic intra-oceanic magmatic arc within the Yarlung-Tsangpo suture zone, southern Tibet. Chemical Geology, 187, 267-277. METCALFE, I. 1999. The ancient Tethyan oceans of Asia: How many? How old? How deep? How wide? UNEACAsia Papers, 1, 1-9. MOLNAR, P. & TAPPONNIER, P. 1975. Cenozoic tectonics of Asia; effects of a continental collision. Science, 189, 419-426. MUECKE, G.K., ELIAS, P. & REYNOLDS, P.H. 1988. Hercynian/Alleghanian overprinting of an Acadian terrane: Ar/Ar studies in the Meguma zone, Nova Scotia, Canada. Chemical Geology, 73, 153-167. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer Academic, Dordrecht. NICOLAS, A., GIRARDEAU, J. & MARCOUX, J. ET AL. 1981. The Xigaze ophiolite (Tibet): a peculiar oceanic lithosphere. Nature, 294, 414-417. PATRIAT, P. & ACHACHE, J. 1984. India-Eurasia collision chronology has implications for crustal shortening and driving mechanisms of plates. Nature, 311,615-621. PEARCE, J.A. & NORRY, MJ. 1979. Petrogenic implications of Ti, Zr, Y and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology, 69, 33-47. Pozzi, J.P., WEATPHAL, M., GIRARDEAU, J., BESSE, J., YAO, X.Z. & Li, S.X. 1984. Paleomagnetism of the Xigaze ophiolite and flysch (Yarlung Zangbo suture zone, southern Tibet): latitude and direction of spreading. Earth and Planetary Science Letters, 70, 383-394. ROWLEY, D.B. 1996. Age of initiation of collision between India and Asia: a review of stratigraphic data. Earth and Planetary Science Letters, 145, 1-13. SAMSON, S.D. & ALEXANDER, E.G. JR 1987. Calibration of the interlaboratory Ar/Ar dating standard, Mmhb-1. Chemical Geology, 66, 27-34. SAUNDERS, A.D. 1984. The rare earth element characteristics of igneous rocks from the ocean basins. /W.HENDERSON, P. (ed.) Rare Earth Element Geochemistry. Developments in Geochemistry, 2, 205-236. SCHARER, U., Xu, R.H. & ALLEGRE, CJ. 1984. U-Pb

geochronology of the Gangdese (Transhimalaya) plutonism in the Lhasa-Xigaze region, Tibet. Earth and Planetary Science Letters, 69, 311-320. TAPPONNIER, P., MERCIER, J.L., PROUST, F. ET AL. 1981. The Tibetan side of the India-Eurasia collision. Nature, 294, 393-399. WANG, C., Liu, Z. & HEBERT, R. 2000. The YarlungZangbo paleo-ophiolite, southern Tibet: implications for the dynamic evolution of the YarlungZangbo suture zone. Journal of Asian Earth Sciences, 18,651-661. XIAO, X.C. 1988. General Review of the Tectonic Evolution of the Crustal-Mantle of the QinghaiXizhang (Tibet) Plateau. Geological Publishing House, Beijing (in Chinese). YIN, A. & HARRISON, T.M. 2000. Geologic evolution of the Himalayan-Tibetan orogen. Annual Review of Earth and Planetary Sciences, 28, 211-280. YIN, A., HARRISON, T.M., RYERSON, F.J., CHEN, W., KIDD, W.S.F. & COPELAND, P. 1994. Tertiary structural evolution of the Gangdese thrust system, southeastern Tibet. Journal of Geophysical Research,^, 18175-18201. ZHOU, M.-F., ROBINSON, P.T., REYNOLDS, P.H., et al. 1995. Supra-subduction zone origin of the Luobusa ophiolite, Indus-Yarlung Zangbo Suture zone, Tibet. GAC/MAC Abstract, 20, A-114. ZHOU, M.-F., ROBINSON, P.T., MALPAS, J. & Li, Z. 1996. Podiform chromitites of the Luobusa ophiolite (Southern Tibet): implications for melt/rock interaction and chromite segregation in the upper mantle. Journal of Petrology, 37, 3-21. ZHOU, M.-F., MALPAS, J., ROBINSON, P.T. & REYNOLDS, P. 1997. Dynamothermal aureole of the Donqiao ophiolite, Northern Tibet. Canadian Journal of Earth Sciences, 34, 59-65. ZHOU, S., Mo, X.X., MAHONEY, J.J., ZHANG, S.Q., Guo, TJ. & ZHAO, Z.D. 2002. Geochronology and Nd and Pb isotope characteristics of gabbro dikes in the Luobusha ophiolite, Tibet. Chinese Science Bulletin, 47, 143-146. ZYABREV, S.V., AITCHISON, J.C., BADENGZHU,, DAVIS, A.M., Luo, H. & Liu, J.B. 2002. Radiolarian biostratigraphy of supra-ophiolites sequences in the Xigaze area, Yarlung-Tsangpo suture, Southern Tibet (preliminary report). Radiolaria, 17, 13—19.

Tectonic implications of boninite, arc tholeiite, and MORE magma types in the Josephine Ophiolite, California-Oregon GREGORY D. HARPER Department of Earth and Atmospheric Sciences, State University of New York at Albany, Albany, NY 12222, USA (e-mail: [email protected]) Abstract: The Josephine Ophiolite is a large complete ophiolite flanked by arc complexes, including rifted arc fades, and overlain by volcanopelagic and volcaniclastic sedimentary rocks. The extrusive sequence and sheeted dyke complex record a wide range in magma types and degree of fractionation. The upper part of the extrusive sequence, as well as late dykes in the ophiolite, have mid-ocean ridge basalt (MORB) affinities and include unusual highly fractionated Fe-Ti basalts. The sheeted dyke complex and lower pillow lavas are dominated by transitional island-arc tholeiite (IAT) to MORB, but about 10% consist of low-Ti, high-Mg basalts and andesites. Whole-rock chemistry and Cr-spinel compositions indicate that the lowTi rocks range from boninite (BON) to primitive arc basalt. The low-Ti samples have trace element characteristics indicating a greater subduction component than the IAT-MORB or MORB samples, as well as derivation from a wide range of sources ranging from depleted to enriched relative to an average N-MORB mantle source. Mixing of low-Ti and MORB magmas may have produced the IAT-MORB magma type that is most characteristic of the ophiolite. Podiform chromites and late magmatic features in the mantle peridotite, described by previous workers, appear to have been formed from the low-Ti magmas. Regional geological relationships and the presence of boninitic magmas suggest that arc rifting and initial sea-floor spreading to form the Josephine Ophiolite occurred in the forearc of a west-facing arc built on edge of the North American plate. Arc magmatism appears to have jumped westward, at which time the Josephine basin became situated in a back-arc setting, analogous to the inferred evolution of the modern Lau back-arc basin. Alternatively, the Josephine Ophiolite may have formed in a setting analogous to the north end of the Tonga Trench or the south end of the North Fiji basin, both sites of modern boninites, where a back-arc spreading centre has propagated across an arc into the forearc. Rift propagation during formation of the Josephine Ophiolite is consistent with the presence of highly fractionated Fe-Ti basalts.

Ophiolites are generally believed to represent ancient ocean crust and upper mantle, yet the tectonic setting of many ophiolites is equivocal, Whereas some appear to have formed at midocean ridges, the geochemistry, petrology, and sedimentary sequences of many ophiolites suggest they formed above a subduction zone and they have therefore been called 'suprasubduction zone' (SSZ) ophiolites (Pearce et al. 1984). Spreading centres in back-arc basins are a likely tectonic setting for many ophiolites, where geochemistry of magmas varies from mid-ocean ridge basalt (MORB), especially in mature back-arc basins, to transitional between MORB and island-arc tholeiite (IAT; e.g. Hawkins et al. 1990; Pearce et al. 1994; Hawkins 1995). Recently, basalts erupted on segments of the Chile Ridge that are closest to where the ridge is being subducted have some of the geochemical characteristics of magmas erupted in magmatic arcs (Klein & Karsten 1995). Thus a mid-ocean ridge origin of some

ophiolites having arc-like geochemical signatures is possible (Sturm et al. 2000), especially those that show no sedimentary or regional geological evidence for a nearby arc. A number of ophiolites contain boninites (e.g. Cameron et al. 1979; Cameron 1989; Coish 1989; Meffre et al. 1996; Ishikawa et al. 2002), an unusual magma type that appears to be derived by melting of refractory peridotite ('second stage melts'; e.g. van der Laan et al. 1989). Tertiary boninites are common in forearcs of many of the western Pacific intraoceanic arcs (e.g. Umino 1986; Bloomer & Hawkins 1987; Murton et al. 1992). A few modern occurrences of boninites are known, including forearcs where a back-arc spreading centre has propagated across an arc axis (Falloon & Crawford 1991; Sigurdsson et al. 1993) and directly behind the Tonga arc where a spreading centre is propagating into rifted arc crust (Kamenetsky et al. 1997). The unusual conditions needed to produce boninitic magmas provide an

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 207-230. 0305-8719/037$ 15 © The Geological Society of London 2003.

208

G. D. HARPER

important constraint on the tectonic origin of ophiolites that contain boninites. The Josephine Ophiolite is part of a belt of c. 165 Ma ophiolites that extends from southern California to Washington (Fig. 1). It is the largest and most complete of these ophiolites, and is best developed in northwestern California and southwestern Oregon (Fig. 2). A suprasubduction zone setting for generation of the Josephine Ophiolite is indicated by the presence of flanking arc complexes of similar age and overlying tuffaceous

hemipelagic rocks that grade upwards into flysch of largely volcanic arc provenance (Dick 1976, 1977a; Harper & Wright 1984; Harper et al 1985, 1994; Wyld & Wright 1988; Saleeby 1992). Previous studies have shown that dykes and lavas of the Ophiolite have geochemical affinities transitional between IAT and MORB (Harper 1984, 1988; Wyld & Wright 1988). More recently, Harper (2003) has shown that the upper pillow lavas and late dykes in the Ophiolite form a MORB-affinity suite that includes unusual highly fractionated Fe-Ti basalts. This paper focuses on a suite of low-Ti pillow lavas and dykes, some of which are boninitic. A synopsis of geochemical data for all sheeted dykes and pillow lavas is presented for comparison. These data, along with the regional geological setting, are used to reevaluate possible settings for Ophiolite formation, specifically whether arc rifting and initial sea-floor spreading occurred in a forearc. The geochemical data for the Josephine Ophiolite also provide a basis for comparison with other Middle Jurassic ophiolites in the western USA (Fig. 1), including the Ingalls Ophiolite Complex (Metzger et al. 2002) and the Coast Range Ophiolite (e.g. Shervais 1990; Giaramita et al. 1998). In particular, geochemical comparison of the Coast Range Ophiolite with the Josephine Ophiolite may help resolve whether they are related (e.g. Harper et al. 1985; Saleeby 1992) or, as some argue, the Coast Range Ophiolite is exotic with respect to North America (e.g. Dickinson et al. 1996; Godfrey & Dilek 2000; Pessagno et al. 2000).

Geological setting

Fig. 1. Middle Jurassic Ophiolites of the western USA that are similar in age to the Josephine Ophiolite. JODEO, Devils Elbow Remnant of the Josephine Ophiolite (Wyld & Wright 1988). Modified from Metzger et al. (2002).

The Klamath Mountains of northwestern California and southwestern Oregon consist of ophiolitic and island-arc terranes, separated by east-dipping thrust faults, that young to the west (Burchfiel & Davis 1981; Irwin 1994). The Josephine Ophiolite and its overlying sedimentary rocks (Galice Fm) are part of the westernmost of these thrust sheets, the Western Klamath Terrane. The main body of the Josephine Ophiolite (Fig. 2) and remnants exposed farther south (Devils Elbow remnant; DEO in Fig. 2; Wyld & Wright 1988) define an along-strike length of c. 250 km. The roof thrust (Orleans thrust; Fig. 2) is a major crustal boundary; the outcrop pattern indicates greater than 40 km of displacement and geophysical data suggest as much as 100 km (Jachens et al. 1986). The upper plate of the thrust is the ophiolitic Rattlesnake Creek Terrane (Fig. 2). As a result of the underthrusting, the Josephine Ophiolite and overlying Galice Formation were regionally metamorphosed under prehnite-pumpellyite (north) to lower greenschist (south) facies conditions (Harper

BONINITES IN THE JOSEPHINE OPHIOLITE

209

Fig. 2. Generalized geological map of the west-central Klamath Mountains (Snoke 1977; Harper 1984; Yule 1996; and compilation by Irwin 1994).

et al 1988, 1994). The basal thrust (Madstone Cabin thrust, Fig. 2) juxtaposes the Josephine Ophiolite over a mafic intrusive complex of similar age (Chetco Intrusive Complex; Dick 1976, 1977a; Harper et al. 1990, 1996; Yule 1996). Geochronological data show that displacement on both the roof and basal thrusts overlapped in time (Harper et al. 1994). The Josephine Ophiolite is conformably overlain by a hemipelagic sequence consisting of siliceous argillite, chert, and tuffaceous chert (Harper 1984; Pinto-Auso & Harper 1985); similar rocks and locally volcanic-rich greywacke are locally intercalated with pillow lavas of the Ophiolite. The hemipelagic sequence grades upwards into a thick flysch sequence derived largely from an active arc (Snoke 1977; Harper 1983, 1984). The ages of the Josephine Ophiolite (162-

164 Ma), hemipelagic sequence (c. 162-157 Ma), Galice flysch (c. 157-153 Ma), and thrust emplacement (c. 153-150 Ma) are tightly constrained by biostratigraphic (Pessagno & Blome 1990; Pessagno et al. 2000) and radiometric age data (Wyld & Wright 1988; Harper et al. 1994). North of the Josephine Ophiolite is the RogueChetco island-arc complex (Dick 1976, 1977a; Garcia 1982; Harper & Wright 1984; Harper et al. 1994; Yule 1996). The Rogue Formation consists of submarine volcanic breccias, tuffs, volcaniclastic rocks and less abundant flows, and, like the Josephine Ophiolite, is overlain by flysch of the Galice Formation. The Chetco Intrusive Complex lies structurally beneath the Rogue Formation and its c. 200 Ma ophiolitic basement, and represents the core of the island arc. The age of RogueChetco arc magmatism is well constrained at

210

G. D. HARPER

160-153 Ma (Hacker & Ernst 1993; Harper et al 1994; Hacker ef al. 1995; Yule 1996). Structurally above and east of the Josephine Ophiolite and Galice Formation is a 174-159 Ma volcano-plutonic arc complex. The Western Hayfork Terrane (Fig. 2) was affected by a c. 165 Ma erogenic event (Wright & Fahan 1988; Hacker et al. 1995), and was followed by emplacement of the 164-159 Ma Wooley Creek plutonic belt. The basement for this arc complex is the Late Triassic to Early Jurassic Rattlesnake Creek Terrane (RCT) that largely consists of a disrupted ophiolite (Wright & Wyld 1994; Hacker et al. 1995). Ophiolitic rocks virtually identical in lithology and age to the RCT also form the basement for the younger Rogue-Chetco arc (Fig. 2; Yule et al. 1992; Yule 1996). Josephine 'rift facies' (Fig. 2) are present in the RCT structurally above the Josephine Ophiolite (Preston Peak complex; Saleeby et al. 1982) and within the basement for the Rogue-Chetco arc (Fiddler Mountain complex; Yule 1996). These rift facies consist of mafic dyke complexes, talus breccias, minor pelagic rocks, and locally olistostromes built on older ophiolitic (RCT) basement. The Devils Elbow Remnant of the Josephine Ophiolite in the southern Klamath Mountains (Fig. 1) is also considered a rift facies, consisting of sheeted dykes, pillow lavas, and ophiolite breccia overlying older RCT-like ophiolitic basement (Wyld & Wright 1988). U/Pb zircon ages of the Devils Elbow and Preston Peak rift facies are 164 ± 1 Ma (Wyld & Wright 1988; Saleeby & Harper 1993), just older than the single highresolution zircon age of 162 ± 1 Ma for the main body of the Josephine Ophiolite (Harper et al. 1994).

Low-Ti dykes and lavas Occurrence The proportion of low-Ti dykes and lavas in the Josephine Ophiolite is difficult to estimate, but they are clearly a minor component, making up perhaps 10% or less of the sheeted dyke complex and pillow lavas. In general, the low-Ti lavas are restricted to the lower pillow lavas. In most of the area, the upper pillow lavas form a late MORBaffinity suite that includes fractionated Fe-Ti basalts (Harper 2003). At one locality in the southernmost part of the study area, however, this late MORB/Fe-Ti suite is absent and low-Ti lavas occur at the top of the extrusive sequence (samples Z91a and Z91b). In the type extrusive sequence of the Josephine Ophiolite, where the most detailed stratigraphic sampling has been carried out, only a single, 3 m thick low-Ti unit is

present, consisting of a broken pillow breccia (sample Y5). It is situated in the middle of a c. 400m thick section, beneath the MORB/Fe-Ti upper pillow lavas (Harper 2003). One low-Ti pillow lava (sample R20) occurs as a screen in the upper part of the sheeted dyke complex, and lowTi dykes in the sheeted dyke complex are cut by other dykes (i.e. they are not late dykes). The field relationships of the low-Ti pillow lavas and dykes indicate they formed 'on axis'. As discussed below, most of the low-Ti pillow lavas and dykes have primitive compositions. Harper (1988) used this as an argument for periodic freezing of axial magma chambers, which would have allowed for mantle-derived melts to rise to the surface. Oceanic faults and related large-scale tilting of the crustal sequence are postulated to record structural extension during periods when magma chambers were absent (Alexander & Harper 1992).

Petrography Pillow lavas and sheeted dykes of the Josephine Ophiolite show the effects of extensive sub-seafloor hydrothermal alteration as well as subsequent prehnite-pumpellyite to lower greenschist facies regional metamorphism (Harper 1984; Harper et al. 1988). Nevertheless, igneous textures are often well preserved, and relict clinopyroxene and Crspinel are common. All low-Ti samples contain Cr-spinel, which occurs as inclusions in mafic phenocrysts (most common), as microphenocrysts, and/or in the groundmass. The low-Ti volcanic rocks are pillow lavas and pillow breccias. They are distinctive in the field by the presence of light macrovariolites that grade outward into dark, originally glassy pillow rims. Some of the glassy rims are unusually thick (up to 5 cm) and many are vesicular (Table 1). With only two exceptions, neither macrovariolites nor more than 2% vesicles are present in the IAT-MORB or MORB-affinity pillow lavas. Microphenocrysts of olivine (pseudomorphs), containing translucent reddish brown octahedra of Cr-spinel, are present in amounts up to 6%. The olivine has been completely replaced by chlorite, chlorite + quartz or, in sample L10, entirely by quartz. Sample Y5 has rare pseudomorphs that may be altered orthopyroxene based on rectangular and six-sided shapes. Clinopyroxene occurs as microphenocrysts in some samples, and is abundant as a groundmass mineral. Glassy margins of pillows (now mostly chlorite) contain prismatic, skeletal, acicular, or feathery clinopyroxene set in a matrix of altered glass. Variolites consist of dendritic clinopyroxene or radiating clinopyroxene intergrown with plagioclase (albitized). Zierenberg et al. (1988) de-

Table 1. Major (%) and trace element (ppm) analyses for low-Ti dykes and lavas Extrusive sequence

Sheeted dyke complex Sample: % vesicles: % phen: Si02 Ti02 A1203 FeOT MnO MgO CaO Na 2 0 K2O P205 LOI Total Ba Rb Sr Y Zr Nb Ni Cr V Sc Th Hf Ta La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Mg no. Ti/Zr Ti/V Zr/Sm Zr/Y T

A20

A22

A90

E2

F30

RAB1

14

9

17

1

11

2

F88c 11 1

47.93 0.38 11.07 8.12 0.15 18.15 8.77 1.33 0.21 0.05 3.18 99.34 39 4 94 12 34 0.62 503 1224 169 29 0.32 0.99 0.05 2.84 6.52 1.12 4.61 1.29 0.49 1.72 0.30 1.94 0.42 1.22 0.19 1.27 0.20 80 67 13 26 2.8

51.35 0.44 15.14 7.60 0.14 9.20 10.09 2.71 0.10 0.04 1.66 98.48 130 0 115 12 35 0.50 110 444 192 36 0.20 0.98 0.04 1.25 2.97 0.51 2.74 0.91 0.36 1.33 0.25 1.70 0.37 1.04 0.16 1.06 0.17 68 75 14 38 3.0

47.96 0.38 13.27 7.10 0.13 15.16 10.95 0.96 0.10 0.05 3.22 99.28 19 3 114 12 24 0.63 348 848 161 30 0.25 0.85 0.04 2.41 5.68 1.02 4.53 1.40 0.54 1.68 0.33 2.04 0.44 1.26 0.20 1.37 0.20 79 95 14 17 2.0

52.40 0.47 14.62 7.79 0.13 9.57 10.52 2.48 0.16 0.06 2.27 100.47 30 2 169 10 31 0.60 157 519 225 40 0.28 0.83 0.04 1.82 4.28 0.72 3.77 1.15 0.44 1.47 0.26 1.71 0.37 1.10 0.17 1.12 0.18 69 91 13 27 3.0

56.75 0.55 14.29 6.51 0.13 6.63 3.25 6.31 1.64 0.03 3.82 99.92 106 27 158 12 36 0.86 122 374 133 34 0.26 0.99 0.06 2.42 5.02 0.87 4.43 1.45 0.41 1.87 0.33 2.12 0.43 1.19 0.18 1.12 0.16 65 92 24 25 3.0

49.90 0.36 11.60 7.91 0.18 15.50 8.07 1.76 0.19 0.08 4.54 100.09 37 5 71 12 24 0.79 283 959 191 33 0.23 0.77 0.06 1.82 4.06 0.63 2.95 1.03 0.36 1.39 0.28 1.76 0.38 1.22 0.18 1.31 0.19 78 90 11 23 2.0

53.07 0.45 13.95 7.60 0.17 11.06 8.30 3.02 0.42 0.06 n.d. 98.10 52 5 111 12 33 0.41 162 556 203 38 0.19 0.86 0.03 1.47 3.40 0.54 3.00 1.17 0.43 1.67 0.33 2.26 0.49 1.37 0.21 1.34 0.21 72 81 13 28 2.7

2+

G25 11 4

52.70 0.33 13.70 6.45 0.18 11.32 6.52 2.34 2.90 0.02 3.10 99.56 60 49 37 14 28 n.d. 84 444 190 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 76 71 10 2.0 2+

L10 20 2

L4 11 9

R20 10 4

Y29c 19 3

Y5 <2 6

Z85 ND c. 8

Z91a ND 1

Z91b ND n.d.

Z94 <2 0

47.36 0.38 13.50 6.90 0.21 7.75 10.79 3.43 0.70 0.06 7.76 98.85 46 19 26 11 27 0.52 190 685 248 32 0.25 0.71 0.04 2.16 4.92 0.80 4.13 1.22 0.58 1.47 0.25 1.65 0.35 1.03 0.16 1.06 0.16 67 84 9 22 2.6

52.33 0.40 14.30 6.87 0.13 9.20 7.03 4.79 0.18 0.05 3.99 99.27 21 2 96 9 35 0.69 190 524 152 27 0.32 0.97 0.05 2.33 5.56 0.85 4.30 1.26 0.33 1.44 0.24 1.57 0.32 0.91 0.14 0.87 0.14 70 69 16 28 3.7

47.87 0.31 11.55 9.33 0.29 16.09 6.70 2.73 0.15 0.07 4.40 99.49 34 2 14 11 34 0.60 338 926 212 34 0.52 0.99 0.05 3.46 7.62 1.30 5.33 1.48 0.50 1.64 0.27 1.69 0.38 1.09 0.16 1.19 0.19 75 55 9 23 3.1

62.46 0.44 13.29 5.21 0.10 6.85 4.30 2.55 1.05 0.06 3.46 99.76 77 22 105 12 33 0.82 149 501 128 34 0.30 0.90 0.05 2.63 6.21 1.00 5.07 1.44 0.42 1.79 0.30 1.96 0.39 1.07 0.16 0.98 0.15 70 80 21 23 2.8

50.92 0.23 13.28 6.08 0.21 11.77 8.14 2.94 1.61 0.06 4.68 99.92 109 17 64 8 28 0.95 352 834 158 30 0.41 0.78 0.08 2.23 4.49 0.71 3.23 0.81 0.23 1.07 0.19 1.30 0.29 0.88 0.14 0.97 0.16 78 49 9 35 3.4

47.34 0.52 12.45 9.63 0.19 14.97 8.04 2.45 0.53 0.10 3.44 99.66 225 47 42 12 35 1.03 337 1134 203 37 0.39 1.05 0.07 3.96 9.46 1.55 7.88 2.03 0.67 2.21 0.35 2.20 0.45 1.24 0.18 1.22 0.18 73 89 15 17 2.9

59.32 0.35 15.70 7.69 0.06 4.17 3.06 8.26 0.11 0.06 0.81 99.59 21 1 46 15 29 0.50 73 276 168 35 0.27 0.86 0.03 4.75 7.41 1.64 8.26 2.17 0.75 2.68 0.42 2.58 0.53 1.42 0.20 1.20 0.17 49 72 12 13 1.9

56.70 0.26 12.09 6.80 0.11 11.28 7.14 1.95 0.67 0.04 2.57 99.61 79 13 97 8 30 0.45 157 490 169 32 0.19 0.87 0.04 1.27 2.87 0.47 2.38 0.75 0.30 1.05 0.18 1.22 0.26 0.74 0.12 0.79 0.13 75 52 9 40 3.9

57.54 0.22 15.22 7.12 0.13 6.42 3.82 6.19 0.48 0.02 2.24 99.39 57 1 87 8 37 0.75 92 551 109 37 0.29 1.05 0.04 1.90 3.55 0.62 3.02 0.87 0.22 1.18 0.20 1.30 0.26 0.72 0.10 0.66 0.10 62 36 12 43 4.9

FeO , total Fe as FeO. Mg number is atomic Mg/(Mg + Fe ) where all Fe is calculated as Fe . n.d., not determined. % phen is percent phenocrysts. Mostly pseudomorphs after olivine; some clinopyroxene and trace Cr-spinel phenocrysts also present in some samples. % vesicles (now amygdules) measured in chilled, originally glassy rim of pillows. Amygdules avoided in sample material crushed for analysis.

212

G. D. HARPER

scribed texturally similar rocks ('olivine basalt') from the Turner Albright Mine (Fig. 2) that are associated with massive sulphide mineralization, and one of their analysed samples is a low-Ti basalt. Low-TiO2 pillow lavas from the southern part of the study area (samples Z85, Z91a, Z91b, Z94) have undergone a higher grade of regional metamorphism (lower greenschist facies) and are foliated. Nevertheless, flattened variolites are evident in outcrop, and Cr-spinel is present within flattened pseudomorphs or as microphenocrysts. Most low-Ti sheeted dykes can be recognized in outcrop by the presence of mafic phenocrysts (pseudomorphs), in contrast to other sheeted dykes that are either aphyric or contain sparse microphenocrysts. Non-porphyritic low-Ti dykes are also present, however, and were recognized by the presence of groundmass Cr-spinel (sample E2) or by their whole-rock geochemistry (sample RABl). The most abundant phenocryst phase contains octahedra of relict Cr-spinel and has been completely replaced by chlorite, epidote + chlorite, or amphibole. Outlines of many of the pseudomorphs are characteristic of olivine, but sample A20 has a few equant six-sided shapes that may have been orthopyroxene. Fresh clinopyroxene occurs as phenocrysts and is abundant in the groundmass, except for samples A90 and E2 in which pyroxene has been completely replaced by actinolitic hornblende of hydrothermal origin (Harper et al. 1988). Plagioclase is mostly replaced by finegrained clinozoisite and albite, but relict plagioclase (Anyg-go) is present in sample F30. Groundmass textures are equigranular to subophitic. Textures indicate a crystallization order of Crspinel followed by olivine, clinopyroxene, and finally plagioclase. The late crystallization of plagioclase is characteristic of matic arc magmas (e.g. Pearce et al. 1984), and the pyroxene-charged glassy pillow margins are texturally similar to boninites (e.g. Cameron et al. 1980). Boninites also contain orthopyroxene or clinoenstatite, but if they were present in the Josephine samples they have been completely altered. Minor orthopyroxenite and orthopyroxene-olivine cumulates are present in the cumulate sequence of the Josephine Ophiolite (Harper 1984) and may have formed from boninitic magmas. In general, however, the crystallization order in the cumulate sequence is more typical of island-arc tholeiites, with olivine followed by clinopyroxene, orthopyroxene, and finally plagioclase.

Geochemistry A large database of whole-rock major and trace element data for dykes and lavas from the Josephine Ophiolite is available from previous studies

(Harper 1984; Harper et al. 1988; Wyld & Wright 1988; Zierenberg et al. 1988; Alexander et al. 1993). Most recently, Harper (2003) presented more complete analyses, which include rare earth element (REE) data, for upper pillow lavas and late dykes in the Josephine Ophiolite as well as for representative samples of the sheeted dyke complex and lower pillow lavas. Table 1 gives major element, trace element and REE analyses for 17 low-Ti sheeted dykes and pillow lavas, and representative analyses of upper and lower pillow lavas, sheeted dykes, and plagiogranites are given in Table 2. Major and trace elements (Ba, Sr, Zr, Ni, Cr, V, Sc) for samples in Tables 1 and 2 were determined by standard X-ray fluorescence (XRF) analysis on glass discs fused with a flux and on pressed powder pellets, respectively (University of Utah and McGill University). All other trace elements and REE were determined by inductively coupled plasma mass spectrometry (ICP-MS); low-Ti samples (Table 1) were analysed in duplicate at Union College. Some of the same samples were also analysed by ICP-MS at Durham University, UK, and Washington State University, and the results are very similar, with values generally agreeing within 5% or better (Kosanke 2000). Cr-spinel was analysed for most of the low-Ti samples by electron microprobe at the State University of New York at Binghamton (Table 3).

Element mobility during metamorphism Harper et al. (1988) and Harper (1995) have shown that extensive mobility of many elements occurred during sub-sea-floor hydrothermal metamorphism. Most samples from the extrusive sequence have undergone loss of Ca and Sr and gain of K2O, Rb and Ba. Elevated SiO2 in some sheeted dykes and lower pillow lavas, associated with high Na2O (e.g. R9b), is probably the result of SiO2 addition during albitization. In contrast, Ti, Zr, Y, Th, Hf, Ta, and REE (except for Eu) appear to have been immobile, and V, Cr, Ni, and P have been immobile except in epidosites (Harper et al. 1988; Harper 1995). These results are consistent with a number of empirical and experimental studies of element mobility during sub-sea-floor hydrothermal metamorphism (e.g. Cann 1970; Humphris & Thompson 1978; Mottl 1983; Seyfried 1987). For the purpose of evaluating the igneous geochemistry of the low-Ti sheeted dykes and lavas, it is assumed that Ti, Zr, Y, V, Cr, Ni, Th and REE (except Eu in some samples) have been immobile. Several of the low-Ti samples show negative Eu anomalies in chondrite-normalized REE diagrams that are probably due to Eu

Table 2. Selected chemical analyses of pillow lavas, sheeted dykes, and plagiogranites

Sample: Rock type:

D74.5 pb

L31 pb

SiO2 TiO2 A1203 FeOT MnO MgO CaO Na20 K2O P205 LOI Total Ba Rb Sr Y Zr Nb Ni Cr V Sc Th Hf Ta La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

51.20 1.18 16.60 10.19 0.23 6.60 5.49 4.33 0.82 0.11 3.54 100.29 333 13 318 28 71 3.2 46 100 310 42 0.33 1.92 0.17 3.88 10.70 1.75 9.40 3.00 0.98 3.72 0.66 4.32 0.94 2.79 n.d. 2.53 0.37

50.28 1.67 15.64 7.35 0.22 4.46 11.88 4.80 0.52 0.17 n.d. 96.99 178 13 373 33 110 4.1 103 81 373 46 0.39 2.42 0.23 5.72 14.13 2.09 10.77 3.83 1.40 4.60 0.93 5.80 1.25 3.58 0.50 3.04 0.47

Sheeted dyke complex

Lower pillow lavas

Upper pillow lavas Qlp Pb

G24 pb

R9b pb

R19a db

Z83a db

R42a Pg

Z59b Pg

49.80 2.59 14.40 11.75 0.20 5.43 5.14 4.51 0.62 0.21 5.94 100.92 688 10 115 42 110 4.5 13 18 476 40 0.41 2.34 0.25 7.15 18.14 2.65 13.92 4.88 1.66 6.06 1.14 7.55 1.60 4.47 0.61 3.76 0.57

51.67 0.73 16.01 8.18 0.12 8.06 4.67 4.61 1.37 0.10 3.63 99.15 100 9 127 21 60 1.8 65 269 208 37 0.34 1.61 0.12 3.53 8.32 1.38 7.09 2.21 0.77 2.83 0.48 3.28 0.69 2.00 0.32 2.02 0.31

57.10 0.56 15.80 7.12 0.11 5.12 5.77 5.35 0.14 0.08 2.90 100.35 23 3 85 13 49 1.0 38 33 222 34 0.25 1.06 0.09 2.29 5.24 0.79 4.05 1.45 0.57 1.77 0.36 2.36 0.50 1.45 0.20 1.25 0.21

52.40 1.24 15.90 8.61 0.18 8.05 5.64 2.89 0.05 0.16 4.78 100.27 53 1 179 25 102 2.0 43 75 203 40 0.43 2.10 0.16 4.64 10.97 1.65 8.27 2.92 1.07 3.40 0.66 4.23 0.90 2.66 0.37 2.30 0.37

46.48 1.71 14.19 12.18 0.20 8.80 9.89 2.22 0.17 0.15 3.21 99.20 12 2 598 25 80 8.3 43 254 416 73 0.71 2.51 0.45 7.09 16.90 2.72 13.70 3.90 1.38 4.69 0.74 4.60 0.96 2.53 0.37 2.39 0.37

58.50 1.31 15.90 8.37 0.08 3.10 7.71 3.76 0.18 0.17 1.28 100.73 54 2 259 31 128 3.5 10 17 291 28 0.66 2.94 0.24 6.61 14.99 2.21 11.18 3.81 1.16 4.30 0.82 5.38 1.14 3.30 0.47 2.91 0.46

66.00 1.13 15.50 2.65 0.04 1.60 5.18 5.25 0.06 0.34 0.90 98.65 81 3 157 50 259 5.2 <10 <10 69 15 1.18 5.06 0.37 9.16 23.45 3.64 18.54 6.33 1.79 7.40 1.34 8.59 1.80 5.03 0.74 4.75 0.73

Rock type: pb, pillow basalt; db, diabase; pg, plagiogranite (has magmatic quartz), n.d., not determined.

Table 3. Selected analyses of Cr-spinel from low-Ti dykes and lavas Sample: Occurrence: TiO2 A12O3

FeO Fe203 Cr203

MgO Total Cations per 4 oxygens

Ti Al 2+

Fe Fe3+

Cr Mg Cr/(Cr+Al) Mg/Mg+Fe2+)

A20 ol

A22 ol

0.30 13.05 16.26 4.48 53.28 13.35 100.72

0.29 13.90 16.67 3.35 52.12 12.87 99.20

0.007 0.488 0.385 0.153 1.336 0.631 0.73 0.62

0.007 0.526 0.410 0.119 1.323 0.616 0.72 0.60

A22 mphen

A90 mphen

F88 ol

0.26 14.47 16.26 1.95 53.44 12.98 99.36

0.29 11.51 15.99 5.42 53.87 12.95 100.03

0.43 16.44 16.74 4.64 48.44 13.35 100.04

0.006 0.545 0.410 0.071 1.350 0.618 0.71 0.60

ol, in olivine phenocryst (pseudornorph); mphen, microphenocryst.

0.007 0.437 0.378 0.184 1.372 0.622 0.76 0.62

0.010 0.610 0.393 0.157 1.204 0.626 0.66 0.61

G25 ol

L10 ol

L4 ol

R20 ol

Y29c ol

0.29 11.64 15.93 4.46 54.05 13.11 99.48

0.29 12.66 14.93 4.37 53.86 14.00 100.11

0.32 13.71 18.00 3.40 51.53 12.02 98.98

0.42 16.53 16.12 5.46 47.60 13.57 99.70

0.37 13.22 16.76 5.03 50.74 12.93 99.05

0.007 0.443 0.384 0.155 1.380 0.631 0.76 0.62

0.007 0.474 0.353 0.149 1.354 0.663 0.74 0.65

0.008 0.523 0.448 0.122 1.319 0.580 0.72 0.56

0.010 0.614 0.372 0.182 1.185 0.636 0.66 0.63

0.009 0.503 0.401 0.173 1.293 0.621 0.72 0.61

Y5 ol

0.21 12.71 15.90 3.50 55.10 13.31 100.73

0.005 0.476 0.384 0.122 1.383 0.630 0.74 0.62

Z85 mphen

0.30 10.57 15.76 5.13 55.05 13.12 99.93

0.008 0.403 0.375 0.176 1.407 0.631 0.78 0.63

BONINITES IN THE JOSEPHINE OPHIOLITE

215

mobility associated with Ca loss. One sample (L10) has patchy prehnite replacing part of the groundmass and a positive Eu anomaly, features that suggest Ca gain. Harper (2003) noted that pillow cores of most upper pillow lavas, in addition to many samples from the lower sheeted dyke complex, show relatively good correlation between MgO, A12O3, FeO, and TiO2, which is interpreted to represent igneous trends. The low-Ti dykes and lavas show good correlation between Cr and MgO, Cr and SiOa, and Cr and A12O3 (Fig. 3). Because Cr is relatively immobile, especially in these samples where Cr is primarily contained in relict Cr-spinel, this correlation suggests limited mobility of MgO and SiO2 during sub-sea-floor hydrothermal metamorphism.

Magmatic affinities and fractionation Arc magmas differ from MORB in several ways (e.g. Pearce 1982; Shervais 1982; Pearce & Peate 1995): (1) enrichment of large-ion lithophile elements (LILE) such as Ba, Rb, K, and Th relative to high-field strength elements (HFSE) such as Ti, Zr, Y, and HREE; (2) generally lower HFSE; (3) higher H2O and oxygen fugacity. These differences are generally attributed to the addition of a hydrous fluid from the subducting slab to the overlying mantle wedge (Pearce & Peate 1995). A number of discriminant diagrams, utilizing elements that are generally immobile during metamorphism, take advantage of these differences (Figs 4-7). Basalts from back-arc basin spreading centres mostly have MORB affinities, but include compositions ranging from IAT to MORB (e.g. Hawkins et al 1990; Hawkins & Allan 1994; Hawkins 1995; Fretzdorff et al 2002), especially where the spreading axis is close to the arc (Pearce et al. 1994). This range is evident in Figure 7, which shows fields for basalts erupted at spreading centres in the Mariana Trough and Lau Basin. Basalts erupted during arc rifting can be highly variable, ranging from MORB to IAT (e.g. Hawkins & Allan 1994; Hawkins 1995). High MgO and low TiO2 volcanic rocks from modern suprasubduction zone settings include boninite, transitional boninite to arc tholeiite, and arc picrite. Boninites are essentially high-MgO andesites. Crawford et al. (1989) defined boninites as having >53% SiO2 at Mg numbers >0.6, where Mg number is Mg/(Mg + Fe2+) with all Fe calculated as Fe2+. The recent IUGS definition of boninite is SiO2 >52%, MgO >8%, and TiO2 <0.5% (Fig. 3d; Le Maitre 2002). Arc picrites were defined by Le Maitre (2002) as having <52% SiO2 and >12% MgO (Fig. 3d), and they generally have TiO2 >0.5% (Fig. 8).

Fig. 3. MgO variation diagrams for low-Ti dykes (^) and pillow lavas and breccias (t>) from the Josephine Ophiolite. Oxides were recalculated to 100% anhydrous, (a) MgO v. Cr. (b) MgO v. TiO2. (c) MgO v. A12O3. (d) MgO v. SiO2. SDC, basal sheeted dyke complex. UPL, upper pillow lavas. SDC and UPL trends are from Harper (2003). Fields for boninite and arc picrite are from Le Maitre (2002).

216

G. D. HARPER

Fig. 4. Ti v. V discriminant diagram (Shervais 1982). IAT, island-arc tholeiite; MORE, mid-ocean ridge basalt; WPB, within-plate basalt. Data for Hole 786 boninites from the Izu-Bonin forearc are from Murton et al. (1992).

Boninites are characterized by enrichment of LILE and light REE (LREE) coupled with very low HFSE such as Ti, Y, and heavy REE (HREE). This is reflected, for example, in a La/Sm v. TiO2 plot (Fig. 8), which shows the enrichment of boninites in LREE at very low TiO2 The very low abundance of the HFSE Y and Yb in boninites is evident in Y-Cr and Yb-Cr plots (Figs 5 and 6). Fields for boninites are generally not included in discriminant diagrams, and thus a field denned by boninites from the Izu-Bonin forearc (Ocean Drilling Program (ODP) Hole 786; Murton et al. 1992) was added to the diagrams. These rocks are fairly representative of boninites because they include both low-Ca and high-Ca varieties, the latter characterized by lower SiO2 and higher CaO and FeOT (Crawford et al. 1989). There is a complete gradation from boninites into low-Ti island-arc basalts (Beccaluva & Serri 1988). The Josephine Ophiolite shows an unusually wide range in magma types and degree of fractio-

nation. A summary of the magma types and fractionation for sheeted dykes and lavas, other than the low-Ti samples, is first presented for comparison. Sheeted dyke complex and lower pillow lavas (except low-Ti samples). Fractional crystallization trends can be inferred using MgO, as MgO will decrease with fractionation of mafic phases. Harper (2003) showed that a subset of samples from the basal sheeted dyke complex, considered to have undergone the least element mobility, show increasing FeOT and TiO2 enrichment trends with decreasing MgO typical of tholeiitic suites (SDC in Fig. 3b). The increase in Al2Os with decreasing MgO (Fig. 3c) indicates late crystallization of plagioclase, consistent with the crystallization order in the cumulate sequence (Harper 1984). Plagiogranites, which make up less than 1% of the ophiolite, define the negative sloping part of the sheeted dyke trend in Figure 3b and c.

BONINITES IN THE JOSEPHINE OPHIOLITE

217

Fig. 5. Y v. Cr discriminant diagram (Pearce 1982). (See Fig. 4 for key to symbols.) Arrows A, B, and C represent crystallization paths for magmas fractionating Cr-spinel + olivine + pyroxene from MORE, IAT, and boninite magmas, respectively. IAT, island-arc tholeiite; MORB, mid-ocean ridge basalt. Data for Hole 786 boninites from the Izu-Bonin forearc are from Murton et al. (1992).

The subparallel REE and MORB-normalized patterns for a suite of samples from the sheeted dyke complex and lower pillow lavas (Fig. 9a) are consistent with fractionation from a similar parental magma (Pearce 1982, 1983), with negative Eu and Ti anomalies in the most evolved sample (plagiogranite Z59b) reflecting plagioclase and Fe-Ti oxide fractionation, respectively. Nearly all samples of the sheeted dyke complex and lower pillow lavas show characteristics transitional between IAT and MORB, as recognized in previous studies (Harper 1984; Harper et aL 1985; Wyld & Wright 1988), although a few samples consistently plot in the IAT or MORB fields in discriminant diagrams. Most samples have Ti/V values of 20 or higher, similar to MORB (Fig. 4), but plot in the IAT field in Cr-Yand Th/Yb v. Ta/ Yb diagrams (Figs 5 and 7). They show enrichment in Th and have negative Ta and Nb anomalies in a MORB-normalized diagram (Fig. 9a), both characteristics of arc magmas (e.g. Pearce

1982). Ta/Yb ratios indicate a mantle source that is enriched relative to N-MORB mantle, but less enriched than average E-MORB. Three samples collected from the same section along the southwestern limb of the ophiolite, however, have distinctly higher Ta/Yb; sample Z83a (Table 2) is an E-MORB, and another sample from the sheeted dyke complex and a late dyke in the peridotite have calc-alkaline affinities (Fig. 7). Upper pillow lavas and late dykes. The upper pillow lavas and late dykes show Fe and Ti enrichment trends (Fig. 3b) characteristic of tholeiites, and in contrast to the sheeted dyke complex show decreasing A^Oa with decreasing MgO (UPL in Fig. 3c) suggesting early onset of plagioclase fractionation (Harper 2003). Many samples are unusual highly fractionated Fe-Ti basalts (e.g. sample Qlp, Table 2), defined as having greater than 12% FeOT and 2% TiO2 (e.g. Sinton et al. 1983).

218

G. D. HARPER

Fig. 6. Low-Ti dykes (^) and pillow lavas and breccias (<]) plotted in Yb v. Cr diagram of Pearce & Parkinson (1993). Field for Hole 786 boninites from Murton et al. (1992) and field for low-Ti arc picrites from Ramsay et al. (1984) and Eggins (1993). MORE, mid-ocean ridge basalt; IAB, island-arc basalt; BON, boninite volcanic series; FMM, fertile MORE mantle. Samples that plot well above melting curves have probably accumulated Cr-spinel. This diagram shows that some of the Josephine low-Ti samples are sufficiently depleted in HFSE to be classified as boninites, whereas others are primitive arc basalts. Extrapolation upward from the boninite samples, parallel to the crystallization trend, to the intersection with modelled melting curves shows that their parental magmas must have been derived from partial melting of a depleted mantle source (i.e. they are 'second stage melts').

Samples of upper pillow lavas and late dykes cutting serpentinite shear zones in the Josephine Ophiolite mostly have MORB magmatic affinities (Harper 2003). They have Ti/V ratios that are greater than 20 (Fig. 4) and higher Y values at a given Cr concentration compared with most samples of the sheeted dyke complex and lower pillow lavas (Fig. 5). Fe-Ti basalts plot below the MORB field in a Y-Cr discriminant diagram (Fig. 5), probably because they are more fractionated than samples used by Pearce (1982) to define the MORB field. The upper pillow lavas and late dykes are somewhat transitional to IAT, however, as indicated by modest Th enrichment and small negative Ta and Nb anomalies in a MORBnormalized diagram (Fig. 9b), a feature that is observed in many MORB-affinity back-arc basin basalts (e.g. Pearce et al. 1994). Ta/Yb ratios are similar to those for samples from the sheeted dyke complex and lower pillow lavas (Fig. 7), and indicate derivation from a mantle source that was enriched relative to N-MORB mantle. This en-

riched mantle source is also indicated by flat REE patterns (Fig. 9a and b), compared with LREEdepleted patterns typical of N-MORB, and evident in Figure 8 from La/Sm ratios that lie above the N-MORB field. Low-Ti pillow lavas and sheeted dykes. The lowTi pillow lavas and dykes show a positive correlation between Cr and MgO (Fig. 3a) and between Ni and MgO, consistent with Cr-spinel and olivine fractionation, respectively. A^Os increases with decreasing MgO (Fig. 3c), implying the lack of plagioclase fractionation, consistent with the absence of plagioclase phenocrysts. SiO2 also increases with decreasing MgO and Cr (Fig. 3d), from 49 to 65% (anhydrous). TiO2 and FeOT do not correlate with MgO, and this feature distinguishes this suite from the tholeiitic IAT-MORB and MORB suites (Fig. 3b). With the exception of sample Z91a (an andesite), all samples have Mg numbers greater than 0.6, indicating that they are relatively unfractionated (primitive), consistent

BONINITES IN THE JOSEPHINE OPHIOLITE

219

Fig. 7. Th/Yb v. Ta/Yb diagram from Pearce (1982). (See Fig. 4 for key to symbols.) Fields for the Lau Basin and Mariana Trough back-arc spreading centres are based on data of Wood et al. (1982), Hawkins & Melchior (1985), Jenner et al. (1987), Boespflug et al (1990), Frenzel et al. (1990), Hawkins et al. (1990), Vallier et al. (1991), Pearce et al. (1994), and Hawkins (1995); for some samples from these sources, Yb and Ta were estimated using Y/10 and Nb/17 (e.g. Wood et al. 1982), respectively.

with their high Cr and Ni contents. Samples with the highest Mg numbers are likely to have accumulated mafic phenocrysts, especially two samples having greater than 10% phenocrysts (Table 1). Accumulation of Cr-spinel, which occurs primarily as inclusions in olivine phenocrysts, is likely for samples having Cr contents sufficiently high (>900 ppm) that they plot above modelled partial melting curves (Fig. 6). The majority of the low-Ti samples have SiO2 ^53% (anhydrous; Fig. 3d), and at least some of those having <53% SiO2 may have lower SiO2 as a result of accumulated phenocrysts. Thus, the SiO2 contents and Mg numbers greater than 0.6 suggest most of the low-Ti samples are boninites as defined by Crawford et al. (1989), especially high-Ca boninites, which can have SiO2 as low as 50%. About half the low-Ti samples plot in the boninite field as defined by Le Maitre (2002) in the MgO v. SiO2 diagram (Fig. 3d). The linear trends in the MgO diagrams (Fig. 3) suggest that low-Ti samples having MgO <8% may have been derived from boninite magma by crystal fractionation (i.e. part of the boninite volcanic series; Crawford et al. 1989; Murton et al. 1992). About

half the samples have Ti/V ratios near 10, similar to that for most boninites (Shervais 1982), including those from Hole 786 of the Izu-Bonin forearc (Fig. 4). Many of the samples have sufficiently high La/Sm at low TiO2 that they fall in the fields defined by boninite suites (Fig. 8). In the Th/Yb v. Ta/Yb diagram, all the low-Ti samples plot well into the fields for arc basalts and overlap the field defined by Izu-Bonin forearc boninites (Fig. 7). One of the most distinguishing characteristics of boninites is their depletion of HFSE such as Y and HREE such as Yb. Figures 5 and 6 show that about half of the low-Ti samples have Y and Yb values sufficiently low that they fall within or along the edge of the field for the boninite volcanic series. The remainder of the low-Ti samples are probably best classified as primitive island-arc basalts; many fall in the field of arc picrites in the Cr v. Yb diagram (Fig. 6), but they are less enriched in LREE (Fig. 8). Cr-spinel in the low-Ti pillow lavas and dykes (Table 3) is mostly characterized by high Mg/(Mg + Fe), reflecting the generally primitive nature of these samples, and by high Cr/(Cr + Al) ratios (Cr number; Fig. 10). Many of the samples have

220

G. D. HARPER

Fig. 8. TiO2 v. La/Sm diagram (Meffre et al. 1996) showing fields for N-MORB, Mariana and Lau back-arc basins, Mariana and Tonga-Kermadec island arcs, low-Ti arc picrites from the Vanuatu and Solomon island arcs, and three suites of boninites. (See Fig. 4 for key to symbols.) It should be noted that some of the Josephine low-Ti dykes and lavas (triangles) plot entirely within the fields defined by the boninite suites. Fields are based on data from Ramsay et al (1984), Ewart & Hawkesworth (1987), Sinton & Fryer (1987), Woodhead (1988), Hawkins et al (1990), Falloon & Crawford (1991), Arculus et al (1992), Falloon et al (1992), Eggins (1993), and Ewart et al (1994).

Cr-spinel compositions that fall within the field for high-Ca boninites, corresponding to a Cr number greater than c. 0.7. Other samples have lower Cr numbers that are similar to those in island-arc tholeiites from Hole 839 erupted during forearc rifting in the Lau Basin (Allen 1994). This is consistent with the trace element data, suggesting that the low-Ti samples span a range from primitive arc basalt to boninite. Three of the most clearly boninitic samples (Y5, Z91b, Z94), recognized by their low Y and HREE, Ti/V ratios near 10, and SiC>2 (anhydrous) greater than 53%, are plotted in REE and MORBnormalized diagrams in Figure 9c. These samples have unusually low Ti/Zr (Table 1), and samples Z91b and Z94 have unusually high Zr/Sm (Table 1) for their La/Sm; these are characteristics of many Tertiary boninites in western Pacific forearcs (e.g. Murton et al 1992). They also show negative Ti anomalies and sample Y5 has a concaveupward REE pattern, features common in boninites (Beccaluva & Serri 1988). The negative Eu anomaly in samples Y5 and Z94, which also

occurs in several other low-Ti pillow samples not plotted, is probably related to Ca loss during subsea-floor hydrothermal metamorphism (Harper et al 1988). Negative Ce anomalies in these and other low-Ti samples (Fig. 9d) might also be the result of hydrothermal metamorphism, but such anomalies are common in modern arc magmas and generally attributed to contribution from pelagic sediment to the subduction component (e.g. Hole et al 1984; Ben Othman et al 1989). Other low-Ti samples are plotted in Figure 9d. Sample A22 is LREE depleted and has a MORBnormalized pattern similar to island-arc tholeiites, whereas LREE-enriched samples (R20, A90, Z91a) have MORB-normalized patterns similar to those of calc-alkaline magmas (e.g. Pearce 1982, 1983). Thus the low-Ti dykes and lavas appear to span a range from true boninites to low-Ti islandarc basalts. Similar chemical heterogeneity of boninites and associated tholeiites has been observed in other ophiolites such as the Troodos (Cameron 1985) and Koh (Meffre et al 1996) ophiolites.

BONINITES IN THE JOSEPHINE OPHIOLITE

221

Fig. 9. REE- and MORB-normalized diagrams for selected dykes and lavas from the Josephine Ophiolite. (a) Least fractionated to most fractionated samples (base to top) of the dominant IAT-MORB magma type in the ophiolite (Table 2). (b) Samples from the uppermost pillow lavas showing more MORB-like affinities (Table 2). (c) Low-Ti samples that have boninite affinities, (d) Other low-Ti samples. MORB and chondrite normalizing values are fromSun & McDonough (1989).

222

G. D. HARPER

Fig. 10. (a) Mineral chemistry of Cr-spinel in low-Ti dykes and lavas of the Josephine Ophiolite (Table 3). It should be noted that Cr-spinel compositions fall partially within the field for high-Ca boninites of the Troodos ophiolite (type high-Ca boninites as defined by Crawford et al. 1989). Data sources are Dick & Bullen (1984) for MORE and back-arc basins, Cameron (1985) for Troodos, Allen (1994) for OOP Holes 834, 836, and 839 in the Lau Basin, and Umino (1986) for Chichijima (Izu-Bonin forearc). (b) Cr-spinel compositions for rocks from the mantle peridotite of the Josephine Ophiolite in California. Harzburgite data are from Ellis (1977) and Harper (unpublished data), whereas data for chromitite (podiform chromite deposits) and dunite are from Ellis (1977). (c) Cr-spinel compositions for rocks from the mantle peridotite of the northern part of the Josephine Ophiolite in Oregon. Data are from Dick (1976) and Dick & Bullen (1984).

Low-Ti magmas and mantle peridotite Cr-spinel data for mantle peridotites, dunites, and podiform chromites from the study area are shown in Figure lOb. The dunites formed by reaction of magma passing through the peridotite, resulting in replacement of the peridotite by dissolution of pyroxene and precipitation of olivine and Crspinel; podiform chromites, which occur within dunite, probably formed as cumulates in pockets of magma (Dick 1976, 1977a; Kelemen & Dick 1995). The range in Cr number, defined as Cr/(Cr 4- Al), of Cr-spinels in dunites and podiform chromites is similar to those in the low-Ti pillow lavas and sheeted dykes (Fig. lOb and c), and it is likely that they formed from the same magmas. Kelemen & Dick (1995) described a number of late magmatic features in the northern part of the Josephine peridotite, ranging from early olivine pyroxenite to late gabbro, that may also be related to the low-Ti dykes and lavas; they formed from high Mg/(Mg + Fe) magmas that were LREE enriched and depleted in HREE, Ti, and Zr relative to MORE, features evident in Figure 9c and d for low-Ti dykes and lavas. The Josephine peridotite represents the residue left from partial melting (Dick 1976, 1977a). Crspinel compositions are a sensitive indicator of degree of partial melting in mantle peridotites (Dick & Bullen 1984), with Cr number increasing with increasing degrees of partial melting. Figure 10 shows that Cr-spinel in the harzburgites from

the southern (California) part of the Josephine peridotite has Cr numbers that fall in and above the field defined by ocean-floor peridotites (mantle residue left after extraction of MORB). IAT in general appear to form from higher degrees of melting than MORB (e.g. Pearce & Peate 1995), and thus the Cr numbers in Cr-spinel of the California harzburgites are consistent with their being the residue of magmas parental to the transitional lAT-MORB-affmity sheeted dykes and pillow lavas as well as the late MORB upper pillow lavas, which are somewhat transitional to IAT. Cr-spinel in the northern (Oregon) part of the Josephine peridotite (Dick 1976; Dick & Bullen 1984) has Cr numbers that overlap the range for the California samples but also extend to much lower values. The range of magmatic affinities of the sheeted dykes and pillow lavas in the northern part of the ophiolite, however, is the same as in the study area based on the data of Zierenberg et al. (1988) and Yule (1996). Thus much of the northern part of the peridotite does not appear to be a residue for magmas parental to the transitional lAT-MORB-affimty sheeted dykes and lavas. Dick (1976, 1977b) suggested that this part of the peridotite underwent two stages of partial melting: the first as dry melting beneath a midocean ridge, and the second as wet melting above a subduction zone. If the age of the Josephine Ophiolite increases northward, as implied by an apparent transition into a rifted arc facies (Yule

BONINITES IN THE JOSEPHINE OPHIOLITE 1996), the northern part of the peridotite may represent older mantle rocks beneath the Klamath Mountains that predate Josephine rifting and seafloor spreading. This hypothesis is consistent with the MORE affinities of mafic rocks of the ophiolitic part of the Rattlesnake Creek Terrane that formed the pre-Josephine basement of the western Klamath Mountains and occurs as basement for the Josephine rift facies (Wyld & Wright 1988; Wright & Wyld 1994; Yule et al 1994; Hacker et al 1995; Yule 1996).

Petrogenesis A full evaluation of petrogenesis of Josephine sheeted dykes and pillow lavas is beyond the scope of this paper, but a few first-order interpretations can be made from the discriminant diagrams. The Th/Yb v. Ta/Yb plot is useful because these ratios are largely independent of variations caused by the degree of partial melting and crystal fractionation, and thus these ratios essentially record those of the mantle source (Pearce 1982; Pearce & Peate 1995). Ta/Yb is a measure of the degree of mantle enrichment or depletion relative to N-MORB source mantle. Addition of a subduction component, probably a hydrous fluid from dewatering of the slab, results in addition of Th, but not Ta, to the mantle source and is thus represented by a vertical vector (Fig. 7). The upper pillow lavas and a few of the lower pillow lavas fall within the mantle array in Figure 7, indicating MORE affinities, and their Ta/Yb ratios indicate an enriched source relative to N-MORB source mantle. The sheeted dyke complex and lower pillow lava samples (excluding low-Ti samples) have similar Ta/Yb ratios (except for a few high Ta/Yb samples indicating a source similar to that for E-MORB), but have higher Th/ Yb indicating a significant subduction component (Fig. 7). The greater subduction component in the lAT-MORB-affinity sheeted dykes and lower pillow lavas compared with the late pillow lavas is also evident from their fractionation trends (SDC, Fig. 3b and c), with A12O3 enrichment (late crystallization of plagioclase) and less Fe and Ti enrichment (earlier onset of Fe-Ti oxide crystallization) reflecting higher H2O contents in the former (Harper 2003). As noted by Stolper & Newman (1994), water is one of the most sensitive indicators of a subduction component in back-arc basin basalts. In contrast, the low-Ti dykes and lavas show a much wider range of TaAYb (Fig. 7), suggesting mantle sources ranging from depleted to enriched with respect to average N-MORB source mantle. The low-Ti samples also plot further above the

223

mantle array than other Josephine samples, indicating a larger subduction component as indicated by higher Th/Yb at a given Ta/Yb. This larger subduction component is also suggested by the vesicular pillow rims of many of the low-Ti pillow lavas, suggesting sufficiently high water contents for saturation at the depth of eruption (1.3— 2 wt.% H2O assuming water depths of 2-4 km, respectively; e.g. Danyushevsky et al. 1993). The low Ti/V ratios of most of the low-Ti samples, as well as the lack of Ti and Fe enrichment with decreasing MgO, indicates that their magmas were more oxidized than other Josephine magmas, consistent with their higher inferred water content (Shervais 1982). Anomalously low Ti/Zr and high Zr/Sm ratios in some of the Josephine low-Ti pillow lavas, which is a characteristic of many of the Tertiary boninites of the western Pacific forearcs, may be an indicator of a slab melt component in these samples (Pearce et al. 1992; Pearce & Peate 1995). The Th/Yb v. Ta/Yb diagram (Fig. 7) clearly shows that few, if any, of the low-Ti magmas can be related by crystal fractionation to the transitional IAT-MORB sheeted dykes and lower pillow lavas, and none could be parental to the MORB-affinity upper pillow lavas. This conclusion is supported by the generally lower Ti/V (Fig. 4) and Zr/Y (Tables 1 and 2) ratios of the low-Ti samples, which should not change significantly with fractionation unless an Fe-Ti oxide or Zrbearing phase is fractionated (Shervais 1982; Pearce & Norry 1979). Furthermore, in the Y-Cr diagram most of the low-Ti samples lie to the left of crystallization vectors that would relate them to other Josephine samples by fractionation (Fig. 5). The Yb-Cr diagram (Fig. 6) can be used to make inferences regarding degrees of partial melting and depletion of mantle sources for the low-Ti magmas. Degrees of melting can be inferred by extrapolating along the mafic phase crystallization trend to intersect melting curves. This diagram shows that unrealistically high degrees of melting of a fertile MORB mantle (FMM) are required to produce boninites. Thus the Josephine boninitic samples, and boninites in general, are considered to have formed by partial melting of mantle that has already been depleted by removal of a melt or, in other words, represent second stage melts (e.g. Crawford et al. 1989; Pearce & Parkinson 1993). Melting of such refractory mantle to produce boninitic magmas requires unusual conditions, specifically hydrous melting at very high temperatures at shallow depths (less than c. 50 km; e.g. Crawford et al. 1989). The dominant transitional IAT-MORB magma type of the Josephine Ophiolite was probably derived by higher degrees of partial melting of the

224

G. D. HARPER

same enriched mantle source as the MORBaffinity upper pillow lavas. Decompression melting of upwelling mantle beneath a suprasubduction zone spreading centre can produce MORB-affinity magmas, and the addition of a hydrous fluid from the subducting slab will lower the solidus and thus increase the degree of melting (e.g. Pearce & Peate 1995). An alternative origin for the transitional IAT-MORB magma type is that it represents the product of mixing of MORBaffinity magma with the low-Ti magmas. In the type section of the extrusive sequence, the basal pillow lava unit of transitional IAT-MORB affinity is overlain by a unit of mixed IAT-MORB and MORB-affinity basalts capped by a primitive boninite pillow breccia (sample Y5; Harper 2003). The presence of MORB-affinity flows immediately below the boninite implies that both magma types were available at the ridge axis. The primitive nature of Y5, as well as other low-Ti basalts, suggests that axial magma chambers periodically froze, allowing primitive magmas to erupt at the surface (Harper 1988) and seawater to penetrate through the crust to cause serpentinization of uppermost mantle peridotites (Coulton et al. 1995). Based on the presence of fine-grained chilled diabase dykes, which are geochemically related to the upper pillow lavas, cutting these serpentinites, Harper (2003) suggested that the late MORB/Fe-Ti suite was transported laterally along the ridge axis after the axial magma chamber froze. If MORB magmas were also delivered laterally during the formation of the sheeted dykes and lower pillow lavas, mixing of these MORB magmas with low-Ti magmas delivered from beneath the spreading centre could have produced the transitional IAT-MORB magmas. If this scenario were correct, then only laterally delivered MORB magmas would be erupted after magma chambers froze. Alternatively, the late MORB/Fe-Ti magmas were derived from a younger propagating rift, although the absence of sediment separating the lower and upper pillow lavas in the type section of the extrusive sequence suggests no significant hiatus (Harper 2003).

Tectonic setting for eruption of Josephine boninites Meffre et al. (1996) noted that boninite-tholeiite associations, as observed in the Josephine Ophiolite, are common in ophiolites. The presence of boninites indicates unusually high temperatures in the shallow mantle coupled with the presence of H2O (van der Laan et al. 1989). Such conditions are achievable by subduction of young, hot oceanic lithosphere beneath hot mantle or subduction

of a spreading ridge (e.g. Pearce et al. 1992). The Tertiary boninites in the forearcs of island arcs in the western and southwestern Pacific probably formed during initiation of subduction along fracture zones, providing favourable conditions for melting of depleted mantle to produce boninites. This scenario has been proposed for generation of the Middle Jurassic ophiolites of California (Fig. 1; Stern & Bloomer 1992). This model, however, appears unlikely for the Josephine Ophiolite in that arc magmatism in the Klamath Mountains was well established before ophiolite generation (e.g. Harper & Wright 1984; Wright & Fahan 1988; Saleeby 1992; Hacker et al. 1995). Modern (high-Ca) boninites have been dredged from off-axis seamounts along a spreading centre near the northern termination of the Tonga Trench (Falloon et al. 1987; Falloon & Crawford 1991; Sobolev & Danyushevsky 1994) and where the North Fiji spreading centre intersects the Hunter Ridge protoarc (Sigurdsson et al. 1993). The northern Tonga and Hunter Ridge boninites have a similar tectonic setting: a back-arc spreading system has propagated across an arc axis into the forearc near the termination of a trench. Boninitic magmas also occur in off-axis seamounts along the southernmost part of the back-arc spreading system in the Lau Basin (Valu Fa ridge) where the spreading centre is propagating southward into rifted arc crust (Kamenetsky et al. 1997). The latter occurrence is consistent with models of boninite generation during the initiation of backarc basins (Crawford et al. 1981, 1989; Coish et al. 1982; Hickey-Vargas 1989; Falloon et al. 1992), and such an origin is consistent with that for the Josephine Ophiolite as discussed below. The arc complexes associated with the Josephine Ophiolite (Fig. 2) clearly imply that the ophiolite formed within an island-arc system, but whether it formed by spreading in a forearc, in a back-arc (Harper & Wright 1984), or along an intra-arc transform system (Harper et al. 1985; Saleeby 1992) is uncertain. A northward transition of the Josephine Ophiolite into older ophiolitic basement, locally cut and overlain by mafic complexes similar in age to the ophiolite (rift facies, Fig. 2; Yule 1996), suggests that the main body of the Josephine Ophiolite (Fig. 2) is situated close to a rifted arc margin. This implies that the ophiolite formed during the early stages of seafloor spreading, consistent with the slightly younger age of the Josephine Ophiolite (162 ± 1 Ma; Harper et al. 1994) compared with ages of two of the rift facies (164 ± 1 Ma; Wyld & Wright 1988; Saleeby & Harper 1993). The position of the Josephine Ophiolite relative to the 160-153 Ma Rogue-Chetco arc and the 174-159 Ma Hayfork-Wooley Creek arc is, to a

BONINITES IN THE JOSEPHINE OPHIOLITE first approximation, consistent with a back-arc basin origin for the Josephine Ophiolite. In this model, following an orogenic event at c. 165 Ma, the NW-facing Hayfork-Wooley Creek arc underwent rifting that evolved into sea-floor spreading to form the Josephine Ophiolite, resulting in the arc axis migrating westward as the Rogue-Chetco arc and leaving the Hayfork-Wooley Creek arc behind as a remnant arc (Harper & Wright 1984). The current high-resolution age data are compatible, however, with rifting and initial sea-floor spreading taking place in the forearc of the Hayfork-Wooley Creek arc, in that: (1) the 164159 Ma Wooley Creek plutonic belt overlaps the 164-162 Ma ages of the rift facies and Josephine Ophiolite; (2) the 160-153 Ma age range of the Rogue-Chetco arc is younger than the Josephine Ophiolite; (3) Hayfork-age volcanic or plutonic rocks are rare in the Rogue-Chetco arc, even though both the Hayfork and Rogue-Chetco arcs are underlain by the same c. 200 Ma ophiolitic basement of the Rattlesnake Creek Terrane (Yule et al 1992; Yule 1996). Thus it appears likely that rifting and initial sea-floor spreading took place in the forearc, followed by a trenchward jump in the arc axis at c. 160 Ma to form the Rogue-Chetco arc on rifted forearc crust. This model is similar to the evolution of the Lau back-arc basin, in which rifting and spreading apparently initiated in the forearc of what is now the Lau Ridge remnant arc (Parson & Hawkins 1994). Fe-Ti basalts are characteristic of modern propagating rifts (Sinton et al. 1983). The presence of late-stage Fe-Ti basalts in the Josephine Ophiolite thus suggests the likelihood of propagating rift tectonics (Harper 2003). The Josephine Ophiolite may have formed during a single propagating rift event related to the transition from arc rifting to sea-floor spreading, analogous to the Valu Fa Ridge in the southern Lau Basin (Pearce et al. 1994; Parson & Wright 1996). Alternatively, the late-stage Fe-Ti basalts may record a separate younger propagating event, analogous to the Central Lau Basin Spreading Centre, in which Fe-Ti basalts are erupting where this ridge is propagating into slightly older back-arc basin crust (Pearce et al. 1984). The propagation of a spreading centre into a rifted arc or young back-arc basin lithosphere provides conditions favourable for boninite generation in that depleted residual mantle in front of the propagator would probably be near its solidus temperature (Falloon & Crawford 1991; Meffoetal. 1996). Other models for the origin of the Josephine Ophiolite take into account the east-west orientation of the Josephine spreading centre, inferred from the strike of sheeted dykes and oceanic faults (Harper 1982; Alexander & Harper 1992), which

225

is oriented at a high angle to the regional trends of the arc complexes. Based on this geometry, Harper et al (1985) and Saleeby (1992) postulated that the Josephine and other Mid-Jurassic ophiolites of California formed in pull-apart basins along an intra-arc transform resulting from oblique subdue tion, similar to the modern Andaman Sea. The orientation of the spreading axis is also consistent with a tectonic setting analogous to the northern Tonga and Hunter Ridge situations, where a spreading centre has propagated at a high angle across the arc axis into the forearc near the termination of a trench. The Hunter Ridge example may be particularly relevant in that samples dredged from the rift that has propagated into the forearc show the unusually wide range of magma types and fractionation observed in the Josephine Ophiolite, including N-MORB (including Fe-Ti basalt), E-MORB, transitional IAT-MORB, IAT, and high-Ca boninite (Sigurdsson et al. 1993). I thank J. Hawkins, P. Robinson, and J. Shervais for constructive reviews. This work was supported by NSF grants EAR-9706798 and EAR-0003444 and by a SUNY-Albany FRAP grant.

References ALEXANDER, RJ. & HARPER, G.D. 1992. The Josephine Ophiolite; an ancient analogue for slow- to intermediate-spreading oceanic ridges. In: PARSON, L.M., MURTON, BJ. & BROWNING, P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 3-38. ALEXANDER, R., HARPER, G.D. & BOWMAN, J.R. 1993. Oceanic faulting and fault-controlled subseafloor hydrothermal alteration in the sheeted dyke complex of the Josephine Ophiolite. Journal of Geophysical Research, 98, 9731-9759. ALLEN, J.F. ET AL. 1994. Cr-spinel in depleted basalts from the Lau backarc basin: petrogenetic history from Mg-Fe crystal-liquid exchange. In: HAWKINS, J., PARSON, L. & ALLAN, J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135. Ocean Drilling Program, College Station, TX, 565-584. ARCULUS, R.J., PEARCE, J.A., MURTON, BJ. & VAN DER LAAN, S.R. ET AL. 1992. Igneous stratigraphy and major-element geochemistry of Holes 786A and 786B. In: FRYER, P., PEARCE, JA. & STOKKING, L.B. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 143-169. BECCALUVA, L. & SERRI, G. 1988. Boninitic and low-Ti subduction-related lavas from intraoceanic arcbackarc systems and low-Ti ophiolites; a reappraisal of their petrogenesis and original tectonic setting. Tectonophysics, 146, 291-315. BEN OTHMAN, D., WHITE, W.M. & PATCHETT, J. 1989. The geochemistry of marine sediments, island arc

226

G. D. HARPER

magma genesis, and crust-mantle recycling. Earth and Planetary Science Letters, 94, 1—21. BLOOMER, S.H. & HAWKINS, J.W. 1987. Petrology and geochemistry of boninite series volcanic rocks from the Mariana Trench. Contributions to Mineralogy and Petrology, 97, 361-377. BOESPFLUG, X., DOSSO, L., BOUGAULT, H. & JORON,

J.L. 1990. Trace element and isotopic (Sr, Nd) geochemistry of volcanic rocks from the Lau Basin. Geologisches Jahrbuch, D92, 503-516. BURCHFIEL, B.C. & DAVIS, G.A. 1981. Triassic and Jurassic tectonic evolution of the Klamath Mountains—Sierra Nevada geologic terrane. In: ERNST, W.G. (ed.) The Geotectonic Development of California. Prentice-Hall, Englewood Cliffs, NJ, 5070. CAMERON, W.E. 1985. Petrology and origin of primitive lavas from the Troodos ophiolite, Cyprus. Contributions to Mineralogy and Petrology, 89, 239-255. CAMERON, W.E. 1989. Contrasting boninite-tholeiite associations from New Caledonia. In: CRAWFORD, AJ. (ed.) Boninites. Unwin Hyman, London, 314338. CAMERON, W.E., NISBET, E.G. & DIETRICH, VJ. 1979. Boninites, komatiites and ophiolitic basalts. Nature, 280, 550-553. CAMERON, W.E., NISBET, E.G. & DIETRICH, VJ. 1980. Petrographic dissimilarities between ophiolitic and ocean-floor basalts. In: PANAYIOTIOU, A. (ed.) Ophiolites, Proceedings International Ophiolite Conference. Geological Survey of Cyprus, Nicosia, 182-192. CANN, J.R. 1970. Rb, Sr, Y, Zr and Nb in some ocean floor basaltic rocks. Earth and Planetary Science Letters, 10,7-11. COISH, R.A. 1989. Boninitic lavas in Appalachian Ophiolites; a review. In: CRAWFORD, AJ. (ed.) Boninites. Unwin Hyman, London, 264-287. COISH, R.A., HICKEY, R. & FREY, F.A. 1982. Rare earth element geochemistry of the Betts Cove ophiolite, Newfoundland; complexities in ophiolite formation. Geochimica et Cosmochimica Acta, 46, 2117-2134. COULTON, AJ., HARPER, G.D. & O'HANLEY, D.S. 1995. Oceanic versus emplacement-age serpentinization in the Josephine ophiolite: implications for the nature of the Moho at intermediate and slow spreading ridges. Journal of Geophysical Research, 100, 22245-22260. CRAWFORD, AJ., BECCALUVA, L. & SERRI, G. 1981. Tectono-magmatic evolution of the West Philippine-Mariana region and the origin of boninites. Earth and Planetary Science Letters, 54, 346-356. CRAWFORD, AJ., FALLOON, TJ. & GREEN, D.H. 1989. Classification, petrogenesis and tectonic setting of boninites. In: CRAWFORD, AJ. (ed.) Boninites. Unwin Hyman, London, 2-49. DANYUSHEVSKY, L.V., FALLOON, T.J., SOBOLEV, A.V., CRAWFORD, AJ., CARROLL, M. & PRICE, R.C. 1993. The H2O content of basalt glasses from Southwest Pacific back-arc basins. Earth and Planetary Science Letters, 111, 347-362. DICK, HJ.B. 1976. The origin and emplacement of the Josephine Peridotite of southwestern Oregon. PhD thesis, Yale University.

DICK, HJ.B. 1977a. Partial melting in the Josephine Peridotite; I, The effect on mineral composition and its consequence for geobarometry and geothermometry. American Journal of Science, 211, 801-832. DICK, HJ.B. 1977b. Evidence of partial melting in the Josephine peridotite. In: DICK, HJ.B. (ed.) Magma Genesis 1977; Proceedings of the American Geophysical Union Chapman Conference on Partial Melting in the Earth's Upper Mantle. Oregon Department of Geology and Mineral Industries, Portland, 1977, 59-62. DICK, HJ.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DICKINSON, W.R., HOPSON, C.A. & SALEEBY, J.B. 1996. Alternate origins of the Coast Range Ophiolite (California). GSA Today, 6, 1-10. EGGINS, S.M. 1993. Origin and differentiation of picritic arc magmas, Ambae (Aoba), Vanuatu. Contributions to Mineralogy and Petrology, 114, 79-100. ELLIS, R.K. 1977. Podiform chromite occurrences in the Josephine Peridotite, Klamath Mountains, northwestern California. MS thesis, University of California, Berkeley. EWART, A. & HAWKESWORTH, CJ. 1987. The Pleistocene-Recent Tonga-Kermadec Arc lavas; interpretation of new isotopic and rare earth data in terms of a depleted mantle source model. Journal of Petrology, 28, 495-530. EWART, A., BRYAN, W.B., CHAPPELL, B.W. & RUDNICK, R.L. ET AL. 1994. Regional geochemistry of the Lau-Tonga arc and back-arc systems. /W.HAWKINS, J., PARSON, L. & ALLAN, J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135. Ocean Drilling Program, College Station, TX, 385-425. FALLOON, TJ. & CRAWFORD, AJ. 1991. The petrogenesis of high-calcium boninite lavas dredged from the northern Tonga ridge. Earth and Planetary Science Letters, 102, 375-394. FALLOON, TJ., GREEN, D.H. & CRAWFORD, AJ. 1987. Dredged igneous rocks from the northern termination of the Tofua magmatic arc, Tonga and adjacent Lau Basin. Australian Journal of Earth Sciences, 34, 487-506. FALLOON, TJ., MALAHOFF, A., ZONENSHAIN, L.P. & BOGDANOV, Y. 1992. Petrology and geochemistry of back-arc basin basalts from the Lau Basin spreading ridges at 15°, 18° and 19°S. Mineralogy and Petrology, 47, 1-35. FRENZEL, G., MUEHE, R. & STOPPERS, P. 1990. Petrology of the volcanic rocks from the Lau Basin, Southwest Pacific. Geologisches Jahrbuch, D92, 395-479. FRETZDORFF, S., LIVERMORE, R.A., DEVEY, C.W., LEAT, P.T. & STOPPERS, P. 2002. Petrogenesis of the back-arc East Scotia ridge, South Atlantic Ocean. Journal of Petrology, 43, 1435-1467. GARCIA, M.O. 1982. Petrology of the Rogue River island-arc complex, southwest Oregon. American Journal of Science, 282, 783-807. GIARAMITA, M., MACPHERSON, GJ. & PHIPPS, S.P. 1998. Petrologically diverse basalts from a fossil

BONINITES IN THE JOSEPHINE OPHIOLITE oceanic forearc in California: the Llanada and Black Mountain remnants of the Coast Range ophiolite. Geological Society of America Bulletin, 110, 2-20. GODFREY, N.J. & DILEK, Y. 2000. Mesozoic assimilation of oceanic crust and island arc into the North American continental margin in California and Nevada: insights from geophysical data. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 365-382. HACKER, B.R. & ERNST, W.G. 1993. Jurassic orogeny in the Klamath Mountains; a geochronological analysi. In: DUNNE, G.C. & MCDOUGALL, K.A. (eds) Mesozoic Paleogeography of the Western United States, II. Pacific Section, Society of Economic Paleontologists and Mineralogists, 71, 37-59. HACKER, B.R., DONATO, M.M., BARNES, C.G., Me WILLIAMS, M.O. & ERNST, W.G. 1995. Timescales of orogeny: Jurassic construction of the Klamath Mountains. Tectonics, 14, 677-703. HARPER, G.D. 1982. Evidence for large-scale rotations at spreading centers from the Josephine ophiolite. Tectonophysics, 82, 25-44. HARPER, G.D. 1983. A depositional contact between the Galice Formation and a Late Jurassic ophiolite in northwestern California and southwestern Oregon. Oregon Geology, 45, 3-7. HARPER, G.D. 1984. The Josephine ophiolite, northwestern California. Geological Society of America Bulletin, 95, 1009-1026. HARPER, G.D. 1988. Episodic magma chambers and structural extension during generation of the Josephine ophiolite. Geology, 16, 831-834. HARPER, G.D. 1995. Pumpellyosite and prehnitite associated with epidosite in the Josephine ophiolite—Ca metasomatism during upwelling of hydrothermal fluids at a spreading axis. In: SCHIFFMAN, P. & DAY, H. (eds) Low Grade Metamorphism of Mafic Rocks. Geological Society of America, Special Papers, 296, 101-122. HARPER, G.D. 2003. Fe-Ti basalts and propagating rift tectonics in the Josephine ophiolite. Geological Society of America Bulletin, 115, 771-787. HARPER, G.D. & WRIGHT, J.E. 1984. Middle to Late Jurassic tectonic evolution of the Klamath Mountains, California-Oregon. Tectonics, 3, 759-772. HARPER, G.D., SALEEBY, J.B. & NORMAN, E.A.S. 1985. Geometry and tectonic setting of the sea-floor spreading for the Josephine ophiolite, and implications for Jurassic accretionary events along the California margin. In: HOWELL, D.G. (ed.) Tectonostratigraphic Terranes of the Circum-Pacific Region. Circum Pacific Council for Energy and Mineral Resources, Earth Science Series, 1, 239257. HARPER, G.D., BOWMAN, J.R. & KUHNS, R. 1988. Field, chemical, and isotopic aspects of submarine hydrothermal metamorphism of the Josephine ophiolite, Klamath Mountains, California-Oregon. Journal of Geophysical Research, 93, 4625-4657. HARPER, G.D., GRADY, K. & WAKABAYASHI, J. 1990. A

227

structural study of a metamorphic sole beneath the Josephine ophiolite, western Klamath terrane, California-Oregon. In: HOWELL, D.S. & MILLER, M.M. (eds) Paleozoic and Early Mesozoic Paleogeographic Relations in the Klamath Mountains, Sierra Nevada, and Related Terranes. Geological Society of America, Special Papers, 255, 379-396. HARPER, G.D., SALEEBY, J.B. & HEIZLER, M. 1994. Formation and emplacement of the Josephine ophiolite and the age of the Nevadan Orogeny in the Klamath Mountains, California-Oregon: U/Pb zircon and 40Ar/39Ar geochronology. Journal of Geophysical Research, 99, 4293-4321. HARPER, G.D., GRADY, K. & COULTON, AJ. 1996. Origin of the amphibolite sole of the Josephine Ophiolite; emplacement of a cold ophiolite over a hot arc. Tectonics, 15, 296-313. HAWKINS, J.W. 1995. The geology of the Lau Basin. /W.TAYLOR, B. (ed.) Back-arc Basins: Tectonics and Magmatism. Plenum, New York, 63-138. HAWKINS, J.W. & ALLAN, J.F. ET AL. 1994. Petrologic evolution of Lau Basin Sites 834 through 839. In: HAWKINS, H., PARSON, L. & ALLAN, J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135. Ocean Drilling Program, College Station, TX, 427-470. HAWKINS, J.W. & MELCHIOR, J.T. 1985. Petrology of the Mariana Trough and Lau Basin Basalts. Journal of Geophysical Research, 90, 11431-11468. HAWKINS, J.W., LONSDALE, P.F., MACDOUGALL, J.D. & VOLPE, A.M. 1990. Petrology of the axial ridge of the Mariana Trough backarc spreading center. Earth and Planetary Science Letters, 100, 226-250. HICKEY-VARGAS, R. 1989. Boninites and tholeiites from DSDP Site 458, Mariana forearc. In: CRAWFORD, AJ. (ed.) Boninites. Unwin Hyman, London, 339356. HOLE, M.J., SAUNDERS, A.D., MARRINER, G.F. & TARNEY, J. 1984. Subduction of pelagic sediments: implications for the origin of Ce-anomalous basalts from the Mariana Islands. Journal of the Geological Society, London, 141, 453-472. HUMPHRIS, S.E. & THOMPSON, G. 1978. Trace element mobility during hydrothermal alteration of oceanic basalts. Geochimica et Cosmochimica Acta, 42, 127-136. IRWIN, W.P. 1994. Geologic Map of the Klamath Mountains, California and Oregon. Geological Survey Miscellaneous Field Investigations, 1-2148. ISHIKAWA, T., NAGAISHI, K. & UMINO, S. 2002. Boninitic volcanism in the Oman Ophiolite; implications for thermal condition during transition from spreading ridge to arc. Geology, 30, 899-902. JACHENS, R.C., BARNES, C.G. & DONATO, M.M. 1986. Subsurface configuration of the Orleans Fault; implications for deformation in the western Klamath Mountains, California. Geological Society of America Bulletin, 97, 388-395. JENNER, G.A., CAWOOD, P.A., RAUTENSCHLEIN, M. & WHITE, W.M. 1987. Composition of back-arc basin volcanics, Valu Fa Ridge, Lau Basin: evidence for slab-derived component in their mantle source. Journal of Volcanology and Geothermal Research, 32, 209-222.

228

G. D. HARPER

KAMENETSKY, V.S., CRAWFORD, A.J., EGGINS, S. & MUHE, R. 1997. Phenocrysts and melt inclusion chemistry of near-axis seamounts, Valu Fa Ridge, Lau Basin: insight into mantle wedge melting and the addition of subduction components. Earth and Planetary Science Letters, 151, 205-223. KELEMEN, P.B. & DICK, H.J.B. 1995. Focused melt flow and localized deformation in the upper mantle; juxtaposition of replacive dunite and ductile shear zones in the Josephine Peridotite, SW Oregon. Journal of Geophysical Research, 100, 423-438. KLEIN, E.M. & KARSTEN, J.L. 1995. Ocean-ridge basalts with convergent-margin geochemical affinities from the Chile Ridge. Nature, 374, 52-57. KOSANKE, S. 2000. The geology, geochronology, structure and geochemistry of the Rogue Wilderness ophiolite, SW Oregon: implications for the magmatic and tectonic evolution of the Coast Range ophiolite. PhD thesis, State University of New York at Albany. LE MAITRE, R.W. 2002. Igneous Rocks: a Classification and Glossary of Terms: Recommendations of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks, 2nd. Cambridge University Press, Cambridge. MEFFRE, S., AITCHISON, J.C. & CRAWFORD, A.J. 1996. Geochemical evolution and tectonic significance of boninites and tholeiites from the Koh ophiolite, New Caledonia. Tectonics, 15, 67-83. METZGER, E.L., MILLER, R.B. & HARPER, G.D. 2002. Geochemistry and tectonic setting of the ophiolitic Ingalls Complex and adjacent mafic rocks, North Cascades, Washington: implications for correlations of Jurassic Cordilleran ophiolites. Journal of Geology, 110, 543-560. MOTTL, MJ. 1983. Metabasalts, axial hot springs, and the structure of hydrothermal systems at mid-ocean ridges. Geological Society of America Bulletin, 94, 161-180. MURTON, B.J., PEATE, D.W., ARCULUS, R.J., PEARCE, J.A. & VAN DER LAAN, S.R. 1992. Trace element geochemistry of volcanic rocks from Site 786; the Izu-Bonin forearc. In: DEARMONT, L.H., MAZULLO, E.K., STEWART, N.J. & WINKLER, W.R. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 211-235. PARSON, L.M. & HAWKINS, J.W. ET AL. 1994. Two-stage ridge-propagation and geological history of the Lau backarc basin. In: HAWKINS, J., PARSON, L. & ALLAN, J. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135. Ocean Drilling Program, College Station, TX, 9-22. PARSON, L.M. & WRIGHT, I.C. 1996. The Lau-HavreTaupo back-arc basin; a southward-propagating, multi-stage evolution from rifting to spreading. Tectonophysics, 263, 1-22. PEARCE, J.A. 1982. Trace element characteristics of lavas from destructive plate boundaries. In: THORPE, E.S. (ed.) Andesites. Wiley, New York, 525-548. PEARCE, J.A. 1983. Role of the sub-continental lithosphere in magma genesis at active continental margins. In: HAWKESWORTH, C.J. & NORRY, M.J. (eds) Continental Basalts and Mantle Xenoliths.

Shiva, Nanrwich, 230-249. PEARCE, J.A. & NORRY, M.J. 1979. Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology, 69, 33-47. PEARCE, J.A. & PARKINSON, I.J. 1993. Trace element models for mantle melting: application to volcanic arc petrogenesis. In: PRICHARD, H.M., ALABASTER, T., HARRIS, N.B.W. & NEARY, C.R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 373-403. PEARCE, J.A. & PEATE, D.W. 1995. Tectonic implications of the composition of volcanic arc magmas. Annual Review of Earth and Planetary Sciences, 23,251-286. PEARCE, J.A., LIPPARD, SJ. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology: Volcanic and Associated Sedimentary and Tectonic Processes in Modern and Ancient Marginal Basins. Geological Society, London, Special Publications, 16, 77-94. PEARCE, J.A., VAN DER LAAN, S.R., ARCULUS, R.J., MURTON, B.J., ISHII, T., PEATE, D. & PARKINSON, I.J. ET AL. 1992. Boninite and harzburgite from Leg 125 (Bonin-Mariana forearc); a case study of magma genesis during the initial stages of subduction. In: FRYER, P., PEARCE, J.A. & STOKKING, L.B. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 623-659. PEARCE, J.A., ERNEWEIN, M., BLOOMER, S.H., PARSON, L.M., MURTON, B.J. & JOHNSON, L.E. 1994. Geochemistry of Lau Basin volcanic rocks: influence of ridge segmentation and arc proximity. In: MELLIE, J.L. (ed.) Volcanism Associated with Extension at Consuming Plate Margins. Geological Society, London, Special Publications, 81, 53-75. PESSAGNO, E.A. JR & BLOME, C.D. 1990. Implications of new Jurassic stratigraphic, geochronometric, and paleolatitudinal data from the Western Klamath terrane (Smith River and Rogue Valley subterranes). Geology, 18, 665-668. PESSAGNO, E.A. JR, HULL, D.M. & HOPSON, C.A. 2000. Tectonostratigraphic significance of sedimentary strata occurring within and above the Coast Range ophiolite (California Coast Ranges) and the Josephine ophiolite (Klamath Mountains), northeastern California. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 383-394. PINTO-Auso, M. & HARPER, G.D. 1985. Sedimentation, metallogenesis, and tectonic origin of the basal Galice Formation overlying the Josephine ophiolite, northwestern California. Journal of Geology, 93, 713-725. RAMSAY, W.R.H., CRAWFORD, A.J. & FODEN, J.D. 1984. Field setting, mineralogy, chemistry, and genesis of arc picrites, New Georgia, Solomon Islands. Contributions to Mineralogy and Petrology, 88, 386402.

BONINITES IN THE JOSEPHINE OPHIOLITE SALEEBY, J.B. 1992. Petrotectonic and paleogeographic settings of U.S. Cordilleran ophiolites. In: BURCHFIEL, B.C.,

LlPMAN, P.W. & ZOBACK, M.L.

(eds)

The Cordilleran Orogen; Conterminous U.S., The Geology of North America, G-3. Geological Society of America, Boulder, CO, 653-682. SALEEBY, J.B. & HARPER, G.D. 1993. Tectonic relations between the Galice Formation and the Condrey Mountain Schist, Klamath Mountains, Northern California. In: DUNNE, G.C. & MCDOUGALL, K.A. (eds) Mesozoic Paleogeography of the Western United States, II. Pacific Section, Society of Economic Paleontologists and Mineralogists, 71, 61-80. SALEEBY, J.B., HARPER, G.D., SNOKE, A.W. & SHARP, W. 1982. Time relations and structural-stratigraphic patterns in ophiolite accretion, west—central Klamath Mountains, California. Journal of Geophysical Research, 87, 3831-3848. SEYFRIED, W.E. JR 1987. Experimental and theoretical constraints on hydrothermal alteration processes at mid-ocean ridges. Annual Review of Earth and Planetary Sciences, 15, 317-335. SHERVAIS, J.W. 1982. Ti-V plots and the petrogenesis of modern and ophiolitic lavas. Earth and Planetary Science Letters, 59, 101-118. SHERVAIS, J.W. 1990. Island arc and oceanic crust ophiolites: contrasts in the petrology, geochemistry and tectonic style of ophiolite assemblages in the California Coast Ranges. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, A. (eds) Ophiolites: Oceanic Crustal Analogues. Geological Survey Department, Nicosia, Cyprus, 507-520. SIGURDSSON, LA., KAMENETSKY, V.S., CRAWFORD, A.J., EGGINS, S.M. & ZLOBIN, S.K. 1993. Primitive island arc and oceanic lavas from the Hunter-Ridge Fracture Zone. Evidence from glass, olivine and spinel compositions. Mineralogy and Petrology, 47, 149-169. SINTON, J.M. & FRYER, P. 1987. Mariana Trough lavas from 18 degrees N; implications for the origin of back arc basin basalts. Journal of Geophysical Research, 92, 12782-12802. SINTON, J.M., WILSON, D.S., CHRISTIE, D.M., HEY, R.N. & DELANEY, J.R. 1983. Petrologic consequences of rift propagation on ocean spreading ridges. Earth and Planetary Science Letters, 62, 193-207. SNOKE, A.W. 1977. A thrust plate of ophiolitic rocks in the Preston Peak area, Klamath Mountains, California. Geological Society of America Bulletin, 88, 1641-1659. SOBOLEV, A.V. & DANYUSHEVSKY, L.V. 1994. Petrology and geochemistry of boninites from the north termination of the Tonga Trench; constraints on the generation conditions of primary high-Ca boninite magmas. Journal of Petrology, 35, 1183-1211. STERN, R.J. & BLOOMER, S.H. 1992. Subduction zone infancy; examples from the Eocene Izu-BoninMariana and Jurassic California arcs. Geological Society of America Bulletin, 104, 1621-1636. STOLPER, E. & NEWMAN, S. 1994. The role of water in the petrogenesis of Mariana Trough magmas. Earth and Planetary Science Letters, 121, 293-325. STURM, M.E., KLEIN, E.M., KARSTEN, J.L. & KARSON,

229

J.A. 2000. Evidence for subduction-related contamination of the mantle beneath the southern Chile Ridge; implications for ambiguous ophiolite compositions. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 13-20. SUN, S.-s. & McDoNOUGH, W.F. 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 313-345. UMINO, S. 1986. Magma mixing in boninite sequence of Chichijima, Bonin Islands. Journal of Volcanology and Geothermal Research, 29, 125—157. VALLIER, T.L., JENNER, G.A. & FREY, F.A. ET AL. 1991. Subalkaline andesite from Valu Fa Ridge, a backarc spreading center in the southern Lau Basin; petrogenesis, comparative chemistry, and tectonic implications. Chemical Geology, 91, 227-256. VAN DER LA AN, S.R., FLOWER, M.F.J. & KOSTER VAN

GROOS, A.F. 1989. Experimental evidence for the origin of boninites: near-liquidus phase relations to 7.5 kbar. In: CRAWFORD, A.J. (ed.) Boninites. Unwin Hyman, London, 113-148. WOOD, D.A., MARSH, N.G., TARNEY, J., JORON, J.L., FRYER, P. & TREUIL, M. ET AL. 1982. Geochemistry of igneous rocks recovered from a transect across the Mariana Trough, arc, fore-arc, and trench sites 453 through 461. In: HUSSONG, D.M. & UYEDA, S. (eds) Initial Reports of the Deep Sea Drilling Project, 60. US Government Printing Office, Washington, DC, 611-642. WOODHEAD, J.D. 1988. The origin of geochemical variations in Mariana lavas; a general model for petrogenesis in intra-oceanic island arcs? Journal of Petrology, 29, 805-830. WRIGHT, J. & FAHAN, M.R. 1988. An expanded view of Jurassic orogenesis in the western United States Cordillera; Middle Jurassic (pre-Nevadan) regional metamorphism and thrust faulting within an active arc environment, Klamath Mountains, California. Geological Society of America Bulletin, 100, 859876. WRIGHT, J.E. & WYLD, S.J. 1994. The Rattlesnake Creek terrane, Klamath Mountains, California: an early Mesozoic volcanic arc and its basement of tectonically disrupted oceanic crust. Geological Society of America Bulletin, 106, 1033-1056. WYLD, S.J. & WRIGHT, J.E. 1988. The Devils Elbow ophiolite remnant and overlying Galice Formation: new constraints on the Middle to Late Jurassic evolution of the Klamath Mountains, California. Geological Society of America Bulletin, 100, 29-44. YULE, J.D. 1996. Geologic and tectonic evolution of Jurassic marginal ocean basin lithosphere, Klamath Mountains, Oregon. PhD thesis, California Institute of Technology. YULE, J.D., SALEEBY, J.B., JONES, D.L. & SILK, M. 1992. Correlation of basement terranes across the Late Jurassic Josephine inter-arc basin, southwestern Oregon and northern California. Geological Society

230

G. D. HARPER

of America, Abstracts with Programs, 24, 93. YULE, J.D., SALEEBY, J.B. & BARNES, C.G. 1994. Geochemistry of Rattlesnake Creek Terrane fragments contained within the western Jurassic belt, Oregon Klamath Mountains. Geological Society of America, Abstracts with Programs, 26, 106-107.

ZIERENBERG, R.A., SHANKS, W.C. Ill, SEYFRIED, W.E. JR, KOSKI, R.A. & STRICKLER, M.D. 1988. Mineralization, alteration, and hydrothermal metamorphism of the ophiolite-hosted Turner-Albright sulfide deposit, southwestern Oregon. Journal of Geophysical Research, 93, 4657-4674.

Forearc extension and sea-floor spreading in the Thetford Mines Ophiolite Complex JEAN-MICHEL SCHROETTER 1 , PHILIPPE PAGE 1 , JEAN H. BEDARD 2 , ALAIN TREMBLAY 3 & VALERIE BECU 2 l lnstitut National de la Recherche Scientifique—Eau, Terre et Environnement, 880 Chemin Sainte-Foy, Quebec, PQ, Canada, G1S 2L2 2 Natural Resources Canada, Geological Survey, 880 Chemin Sainte-Foy, Quebec, PQ, Canada, G1S 2L2 (e-mail: [email protected]) ^Departement des Sciences de la Terre, Universite du Quebec a Montreal, CP 8888, Succursale centre ville, Montreal, PQ, Canada, H3C 3P8 Abstract: The Ordovician Thetford Mines Ophiolite Complex (TMOC) is an oceanic terrane accreted to the Laurentian margin during the Taconic Orogeny and is affected by syn-obduction (syn-emplacement) deformation and two post-obduction events (Silurian backthrusting and normal faulting, and Acadian folding and reverse faulting). The southern part of the TMOC was tilted to the vertical during post-obduction deformation and preserves a nearly complete cross-section through the crust. From base to top we distinguish cumulate Dunitic, Pyroxenitic and Gabbroic Zones, a hypabyssal unit (either sheeted dykes or a subvolcanic breccia facies), and an ophiolitic extrusive-sedimentary sequence, upon which were deposited sedimentary rocks constituting the base of a piggy-back basin. Our mapping has revealed the presence of subvertically dipping, north-south- to 20°-striking faults, spaced c. 1 km apart on average. The faults are manifested as sheared or mylonitic dunites and synmagmatic breccias, and correspond to breaks in lithology. The fault breccias are cut by undeformed websteritic to peridotitic intrusions, demonstrating the pre- to synmagmatic nature of the faulting. Assuming that rhythmic cumulate bedding was originally palaeo-horizontal, kinematic analysis indicates that these are normal faults separating a series of tilted blocks. In the upper part of the crust, the north-south-striking fault blocks contain north-south-striking dykes that locally constitute a sheeted complex. The faults correspond to marked lateral changes in the thickness and facies assemblages seen in supracrustal rocks, are locally marked by prominent subvolcanic breccias, and have upward decreasing throws suggesting that they are growth faults. The base of the volcano-sedimentary sequence is a major erosional surface in places, which can penetrate down to the Dunitic Zone. The evidence for coeval extension and magmatism, and the discovery of a locally well-developed sheeted dyke complex, suggest that the TMOC formed by sea-floor spreading. The dominance of a boninitic signature in cumulate and volcanic rocks suggests that spreading occurred in a subduction zone environment, possibly in a forearc setting.

Ophiolites are fragments of oceanic lithosphere transported onto a continental margin during orogenesis (Coleman 1971) and they may preserve evidence of the tectono-magmatic evolution of the fossil oceanic crust. It is now known that many ophiolites formed in suprasubduction zone environments (Robinson et al. 1983; Pearce et al. 1984;Robinson & Malpas 1990; Bloomer et al. 1995). Because boninitic magmatism seems to characterize forearc environments, ophiolites that contain abundant boninitic rocks are interpreted as having formed in forearcs (Lytwyn & Casey 1993; Bedard et al. 1998). In such settings, the overriding plate can consist of older oceanic crust, an older arc, or can be a tectonic collage of off-

scraped material and accreted micro-terranes (Dickinson & Seely 1979; Dickinson et al. 1988), and its stress field can be compressional or extensional (Uyeda & Kanamori 1979; Hamilton 1988, 1995). Typically, the forearc basement is hidden beneath thick accumulations of volcanogenie sediments derived from the active arc (Stern & Bloomer 1992), hindering reconstructions of the previous geological history. The ease of access and quality of exposure available in forearc ophiolites afford an opportunity to understand the processes associated with these complex tectonic environments. However, the Ordovician ophiolites of the Canadian Appalachians have been overprinted by Ordovician (Taconic), Silurian (Salinic)

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 231-251. 0305-8719/037$ 15 © The Geological Society of London 2003.

232

J.-M. S C H R O E T T E R £ 7 M L .

and Devonian (Acadian) events, which make it difficult to recognize pre-obduction structures. None the less, pre-obduction structures have been recognized in the Betts Cove (Tremblay et al. 1997; Bedard et al 2000) and Bay of Islands (Karson et al. 1983; Bedard 1991; Berclaz et al. 1998; Suhr & Cawood 2001) ophiolites of Newfoundland. At Betts Cove, forearc extension appears to have proceeded to the point where seafloor spreading was initiated (Bedard et al. 1998; compare Stern & Bloomer 1992). The Thetford Mines Ophiolite Complex (TMOC) of southern Quebec also has abundant boninitic lavas (Church 1977, 1987; Hebert & Laurent 1989; Laurent & Hebert 1989) and cumulates derived from boninites (Bedard et al. 2001), and appears to represent another fragment of forearc crust. This paper presents field evidence for an extensive and complex pre-obduction history for the TMOC.

Geological setting The northern Appalachians are the result of successive Palaeozoic orogenic pulses. The Midto Late Ordovician Taconic Orogeny is thought to be due to the accretion of oceanic terranes (i.e. the Dunnage Zone) against the continental margin of Laurentia (i.e. the Humber Zone) (see inset map in Fig. 1). The tectono-stratigraphic zones of the Canadian Appalachians were defined on the basis of the Cambro-Ordovician geology (Williams 1979). The Baie-Verte-Brompton Line (BBL) separates the Humber and Dunnage Zones, is characterized by numerous ophiolite complexes, and has been interpreted as a continental-oceanic suture zone (Williams & St-Julien 1982). The Late Silurian, Mid-Devonian Acadian Orogeny is attributed to the final closure and destruction of lapetus Ocean, when the Composite Avalonian Terrane collided with the Taconian Orogen developed upon the Laurentian margin (Osberg 1978; Williams & Hatcher 1983; Robinson et al. 1998; van Staal £tf a/. 1998). In southern Quebec, the Dunnage Zone contains three principal ophiolite complexes (Thetford Mines, Asbestos and Orford) which are shown on existing maps as slivers embedded in the SaintDaniel Melange, interpreted by Cousineau & StJulien (1992) as the vestiges of a subduction zone accretionary complex. Other important components of the Dunnage Zone in southern Quebec are the Ascot Complex (c. 460 Ma), interpreted as a volcanic arc by Tremblay (1992), and the Upper Ordovician Magog Group, interpreted as a forearc sedimentary basin that developed during continuing convergence between the Laurentian margin and adjacent oceanic terranes (Cousineau & St-

Julien 1992). The Humber Zone is characterized by a first generation of NW-verging thrusts of Mid-Ordovician age (469-460 Ma), and coeval regional metamorphism associated with ophiolite emplacement and crustal thickening (Pinet & Tremblay 1995; Tremblay & Castonguay 2002). Subsequently, Late Silurian to Early Devonian (c. 430-415 Ma; Castonguay et al. 2001) backthrusts and associated normal faults (e.g. the St-Joseph fault) exhumed Laurentian margin metamorphic rocks (Pinet et al. 1996a, 1996b; Castonguay et al. 2001; Tremblay & Castonguay 2002). The ophiolites of southern Quebec occur in the hanging wall of the St-Joseph normal fault, which separates panels with a distinct metamorphic history, and which forms a composite structure with the BBL (Tremblay & Castonguay 2002). The Thetford Mines (TMOC) and Asbestos (AOC) Ophiolite Complexes are composed of thick mantle and crustal sections, whereas only upper-crustal rocks are preserved in the Mt Orford Ophiolite Complex (MOC) (Oshin & Crocket 1986; Church 1987; Laurent & Hebert 1989; Hebert & Bedard 2000). Lavas and dykes in the MOC have backarc to forearc geochemical affinities (Harnois & Morency 1989; Laurent & Hebert 1989; Huot et al. 2002). Zircons from a trondhjemite intruded into the gabbroic crust contain inherited Proterozoic cores, and the least contaminated fractions give a U/Pb age of 504 ± 3 Ma (David & Marquis 1994). The TMOC and AOC, on the other hand, are dated at 479 ± 3 Ma and 478-480 +3/-2 Ma (Dunning et al. 1986; Whitehead et al. 2000), respectively, and are dominated by boninitic (forearc) lavas and dykes (Church 1977, 1987; Hebert & Laurent 1989; Laurent & Hebert 1989), and by cumulates derived from boninites (Hebert & Bedard 2000; Bedard et al. 2001).

The Thetford Mines Ophiolite Complex (TMOC) The TMOC can be divided into the Thetford Mines Massif (TMM) in the north and the Mount Adstock-Ham Massif (AHM) in the south (Laurent 1975). The TMM has a thick mantle section (c. 5 km thick) and a thin crust (Fig. 1; Laurent et al. 1979). Pre-emplacement structures are difficult to recognize in the TMM, because they have been reworked by three types of syn- and post-emplacement structures (Brassard & Tremblay 1999; Schroetter et al. 2000, 2002), as follows. (1) Syn-emplacement structures are associated with dynamo-thermal metamorphism, with an inverted metamorphic gradient ranging from greenschist grade at the base to upper amphibolite

THETFORD MINES OPHIOLITE, QUEBEC, CANADA

233

Fig. 1. Geological and structural map of the Thetford Mines Ophiolitic Complex, showing the location of the stratigraphic columns of Fig. 2, and the location of Fig. 3. The map is based on three summers of fieldwork (20002002) by our team, with additional data collected in previous field campaigns, complemented by data from previous maps (Cooke 1938; Hebert 1983; Pinet 1995; Brassard & Tremblay 1999).

grade at the contact with the mantle units (Feininger 1981). The most recent age for the metamorphic aureole is 477 ± 5 Ma (Ar-Ar, Whitehead et al. 1995); an age that is interpreted to reflect emplacement (Taconian Orogeny) onto the continental margin (Pinet & Tremblay 1995; Tremblay & Castonguay 2002). (2) Post-emplacement structures include: (a) WSW-ENE- to east-west-striking, SE-directed

backthrusts and associated folds, which are probably coeval with the Upper Silurian-Early Devonian backthrusts that characterize the Humber Zone; (b) subvertical NE-SW-striking, Mid-Devonian Acadian reverse faults and folds (Schroetter et al. 2000, 2002; Tremblay & Castonguay 2002). The original stratigraphy of the TMM has been tilted to the vertical by these events, and the Ophiolitic rocks now occupy the hinge of a major

234

J.-M. SCHROETTER £r,4L.

(multi-kilometre), overturned, SE-verging fold (Laurent 1975; St-Julien 1987). According to Cousineau & St-Julien (1992), the black shales and pebbly mudstones of the SaintDaniel Melange represent a subduction zone accretionary complex, and its contact with the upper part of the TMOC is a fault. Others have interpreted this contact as stratigraphic and depositional (Derosier 1971; Hebert 1983; Schroetter et al 2000, 2002), implying rather that the Saint-

Daniel Melange is a piggy-back basin, and that it represents the lowermost unit of the forearc basin sequence, the Magog Group (Schroetter et al. 2000, 2002) (Fig. 2, column C4). Our field observations indicate that the northern margin of the AHM is indeed tectonic. In the Mt Adstock area (AHM, Fig. 1), an Acadian reverse fault juxtaposes ophiolitic rocks (SE side) against sedimentary rocks of the Saint-Daniel Melange (NW side; Schroetter et al. 2002), whereas in the Mt

Fig. 2. Stratigraphic columns. Column Cl, Thetford Mines Massif (TMM), Caribou Lake. C2, TMM, Lac de 1'Est. C3 and C4 are from the northern and southern (Bisby Lake) parts of the Adstock-Ham Massif, respectively.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA Ham area, metamorphosed continental margin rocks are backthrust onto the ophiolite. However, the southeastern contact between the Saint-Daniel Melange and the AHM is depositional, and is marked by an erosional unconformity (Fig. 2). The AHM is composed of ultramafic and mafic cumulates, layered to massive gabbroic rocks, ultramafic to mafic dyke swarms that locally grade into a sheeted dyke complex, tholeiitic to boninitic lavas and felsic pyroclastic rocks, upon which were deposited detrital sediments (Coleraine Breccia and the Saint Daniel Melange). Late structures are of the same type as in the TMM (type 2 described above), but are less intensely developed. Our mapping shows that these post-emplacement structures are superimposed upon pre-existing, north-south- to NNE-SSW-trending, brittleductile to brittle faults, which we attribute to a pre-emplacement (pre-obduction) extensional phase (Figs 1 and 3).

Stratigraphy The stratigraphy of the TMOC is shown in the four columns of Figure 2. Columns Cl and C2 are from the northern and southern parts of the TMM, respectively, and columns C3 and C4 are from the northern and southern parts of the AHM, respectively. The main points to note are: (1) the existence of significant lateral variations in the thicknesses of cumulate, volcanic and sedimentary facies; (2) the presence of an erosional surface (s-eS in Fig. 2) within the ophiolitic crust that penetrates down to the Dunitic Zone in column Cl, and to the Gabbroic Zone in column C4 (Fig. 2); (3) sedimentary rocks of the Saint-Daniel Melange are in depositional contact with underlying ophiolitic rocks, with the basal member (Coleraine Breccia) infilling a palaeo-topography, with coarse debris flow deposits above thick lava sequences, and thin-bedded siltstones and mudstones above thin lava sequences. In the next section we describe typical plutonic, hypabyssal and supracrustal facies of the AHM, with only brief mention of equivalent facies from the TMM. The early structures (faults) that dissect the AHM crust are then described. Swarms of hydrothermal veins containing amphibolite and greenschist mineral assemblages are common throughout the section, but are not evenly distributed. Most rocks are affected by pervasive intraoceanic hydrothermal metamorphism, but textural pseudomorphism and the absence of younger penetrative deformation generally allow primary features to be identified; for this reason we omit the prefix 'meta-' in the following.

235

Mantle section The harzburgite tectonites of the TMM show nearubiquitous planar and linear fabrics defined by alignment and elongation of orthopyroxene and chromite grains. Occasional compositional layering, pyroxenite dykes, podiform dunites and chromitites are discordant to this high-temperature tectonite fabric (Laurent et al. 1979). Emplacement-related fabrics are developed near the contact with the dynamo-thermal sole. Mineral chemical and whole-rock geochemical data imply that the harzburgite is residual from extensive partial fusion (Hebert 1985; Hebert & Laurent 1989). Peraluminous two-mica granite intrusions have high 87Sr/86Sr initial ratios and igneous zircons with low 208pb/206Pb ratios, and have yielded 469 ± 4 to 470 ± 5 Ma crystallization ages (U/Pb on zircon, Whitehead et al. 2000), suggesting that these granites were derived by fusion of continental margin sediments during emplacement of the hot ophiolite (Clague et al. 1985; Whitehead et al. 2000). The transition from mantle tectonite to cumulate crust is rarely observed, and where exposed (north of Lac Caribou), it is a post-emplacement backthrust fault (Schroetter et al. 2000).

Plutonic crustal section We divide the plutonic crustal section into three zones. The Dunitic Zone is up to 500 m thick (Fig. 1 and column Cl in Fig. 2). Above a thin (50 m) basal dunite subzone, it is common to find rhythmic alternations of dunite and massive chromitite beds. Chromitites are typically 1-10 cm thick, with a maximum of 12 m observed in the Reed-Belanger deposit (see Figs 3 and 4d). Chromitite beds may show modal grading but do not define a consistent polarity. Localized development of schlieren textures and tight isoclinal folds suggest an early, high-temperature deformation event (Kacira 1971). The origin of the rhythmic chromitite-dunite bedding is uncertain. It could reflect the interplay of fractional crystallization, magmatic currents, wallrock assimilation and chamber replenishment. In any case, it appears to have originated as near-horizontal primary bedding, and we use these orientations to define the palaeo-horizontal in the Dunitic Zone. It is noteworthy that immediately below the main chromitite layer of the Reed-Belanger Mine, there are at least four pyroxenite layers (10-30 cm thick) composed of orthopyroxenitic clots in a dunitic matrix. These pyroxenite layers are size-graded, are oriented subparallel to the chromitites, and all give the same, normal, way up. They are inter-

236

J.-M. SCHROETTER£T,4L.

Fig. 3. Geological map of part of the Adstock-Ham Massif, with equal area stereograms (Schmidt) of structural features, and the locations of Figs 4-6. It should be noted that because of Siluro-Devonian tilting to the vertical, this map-view is essentially a cross-section through the oceanic crust.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA

237

Fig. 4. Photographs and sketches illustrating syn-oceanic deformation features from the lower crust in Sector 1. (a) Schematic cross-section summarizing the different tectonic events and illustrating the relation between the major normal fault, the chromitite bedding, the minor and major intrusions, and deformation, (b) Chromitite beds are tilted to the west and crosscut by websterite dykes (E2 event), (c) Websterite dykes are truncated by sub-north-south normal faults, forming a horst-and-graben pattern (E3 event), (d) Chromitite beds are tilted and crosscut by peridotite dykes (E2 event); high-temperature synmagmatic(?) deformation (El event) parallel to chromitite beds should be noted. preted to represent some type of depositional event, and show no evidence of tectonic inversion or refolding. The disappearance of chromitite beds higher in the section marks the transition from the rhythmic dunite-chromitite subzone to the mas-

sive dunite subzone, which is dominated by dunites having only 1 — 1.5% of disseminated chromite. The top of the Dunitic Zone is characterized by the presence of disseminated orthopyroxene (c. 5%) and minor chromite (c. 1%).

238

J.-M. SCHROETTER£T,4L.

Fine-grained websterite dykes (Fig. 4b and c) are layers exhibit modal grading, with a concentration typically oriented sub-east-west (in the AHM) of cumulus-textured clinopyroxene at the top. Thin with subhorizontal to shallow dips. olivine-rich 'layers', schlieren or veins may sepaThe Pyroxenitic Zone has a maximum thickness rate pyroxenite layers. Embedded within these of 600 m but is commonly beheaded by an inter- massive websterites, there are repetitions of rhythnal erosional surface (s-eS in Fig. 2, column Cl). mically layered harzburgite-websterite sequences A basal rhythmic subzone (5-15 m) is character- (10-15 m thick) that are very similar to those of ized by alternations of harzburgite, orthopyroxe- the basal rhythmic subzone. The nature of these nite and websterite layers (30-60 cm) (Fig. 5bl), embedded harzburgite-websterite sequences is which are used to define the palaeo-horizontal. In uncertain. They may be the result of magma the harzburgite, cumulus orthopyroxene grains chamber replenishments, or they could represent may be aligned parallel to layering. The pyrox- sills affected by the high-temperature tectonoenite layers (3-8 mm grain size) are commonly magmatic deformation. Sills of massive dunite boudinaged and transposed by an early tectono- (20-30 m thick), and of undeformed harzburgite magmatic event, which is more strongly developed Iherzolite (2-3 m thick) intrude the massive webin the Thetford Mines Massif. Above these ortho- sterites. A size-graded Iherzolite sill gives a pyroxene + olivine-dominated rocks there is a way-up direction towards the SSW. Fine-grained thick, massive pyroxenite subzone, which is domi- dykes oriented perpendicular to layering are lonated by thickly layered (typically 1 -3 m) web- cally abundant. sterites. Layering generally reflects variations in The Gabbroic Zone is up to 1200m thick the orthopyroxene/clinopyroxene ratio. Some (depending on the depth of penetration of the

Fig. 5. Sketches illustrating syn-oceanic deformation features from the lower crust in Sector 2. (a) Dunitic Zone, websterite dyke in dunite is cut by north-south normal faults oriented parallel to the main fault structure, (bl) Pyroxenitic Zone, alternation of cumulate websterite and harzburgite layers, truncated by north-south normal faults that define a horst-and-graben pattern. (b2) Near the main fault, websterite-harzburgite layers are more chaotic.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA denudational surface) and is organized on a gross scale, with interlayered norites and gabbronorites at the base, gabbros sensu stricto in the middle, and an upper complex composed of hornblende gabbro, hornblendite, dykes, cataclasites, trondhjemitic intrusions and breccia veins. The norites, gabbronorites and gabbros are fine to medium grained, and thin, laterally discontinuous melagabbro to pyroxenite (1-5 cm) layers are common. The medium-grained to pegmatitic hornblende gabbros are either layered or chaotically varitextured, and are associated with coarse-grained hornblendite pods and layers. Trondhjemite forms intrusions of various size (1 cm to 30 m) within the uppermost gabbros, and may form a breccia fill between angular gabbroic clasts; but trondhjemites1 also occur to a lesser extent as dykes within the Dunitic and Pyroxenitic Zones. Zircons extracted from TMOC trondhjemites gave U/Pb ages of 478 +3/-2 and 480 ± 2 Ma, which were interpreted as crystallization ages by Whitehead et al. (2000).

Hypabyssal fades Two types of hypabyssal facies rocks occur above the Gabbroic and Pyroxenitic Zones: dyke swarms and breccias. The dykes (30 cm to c. 1m thick) are mafic to ultramafic, locally orthopyroxenephyric, but more commonly show microgabbroic or aphanitic textures. Microgabbro dykes have thin (1-2 cm) aphanitic chilled margins. Dyke margins are locally brecciated by late hydrothermal vein networks. Dykes are most commonly oriented north-south, and cut the magmatic foliation in host plutonic rocks at c. 60-90°. In some sectors, the dykes constitute 40-100% of the outcrop over hundreds of metres, and are mapped as a sheeted dyke complex (Figs 1 and 3). The breccia facies was not previously recognized in the TMOC. The breccias reach a maximum of 150 m in thickness and separate plutonic and volcanic sequences. Where we have studied them in detail (north of Lac Bisby, Fig. 3; column C4 in Fig. 2, and Fig. 6), the breccia facies caps the gabbroic sequence and is overlain by boninitic lavas and volcaniclastic deposits. In this area, the brecciated hypabyssal facies is characterized by alternations of breccia (0.5-3 m thick) and amygdaloidal sills (Fig. 6c). The breccia is composed of angular clasts (90-40%, 1-10 cm) of aphanitic 'dolerite', microgabbro and gabbro. The angularity of clasts indicates only limited transport, and jigsaw-puzzle textures imply in situ brecciation. The matrix is typically igneous (microgabbro), but hydrothermal assemblages are also prominent locally. Intra-breccia sills (Fig. 6b and d) (1 cm2 m) have very irregular shapes, contain clasts of

239

adjoining breccia at top and bottom, are commonly amygdaloidal and locally orthopyroxene porphyritic. Similar breccia facies occur above ultramafic rocks of the Dunitic Zone near Lac Caribou (Fig. 2, column Cl), but in this instance, the clasts in the breccia are composed of peridotites and pyroxenites, and the overlying debris flows and pyroclastic deposits are rich in gabbroic and pyroxenitic fragments. Because platinum group element (PGE) mineralization occurs within the Pyroxenitic Zone (e.g. Star Chrome showing: Page et al. 2001), erosion that penetrates to the Dunitic Zone implies that palaeo-placer PGE deposits might represent a new type of exploration target in the area.

Volcano-sedimentary facies The volcanic and volcaniclastic rocks of the ophiolite exhibit marked lateral changes in thickness and lithology (Fig. 2). North of Lac Bisby (Fig. 1), the volcaniclastic rocks are made of blocky tuffs (2-20 m thick) containing rounded pillow-lava fragments (10 cm average, with a few larger blocks), in a sandy volcaniclastic matrix (Fig. 2, column C4). Vesicular pillow lavas of 11.5 m in size alternate with smaller pillows (0.5 m average), with intercalated massive flows, hyaloclastite breccias, and possible submarine talus breccias. South of Lac Caribou there are abundant pyroclastic flow breccias containing rounded clasts of dacite, gabbro and pyroxenite, with intercalated fine-grained dacitic tufs (l-2m thick) and argillites (Fig. 2, column Cl). At the Lac de FEst section, a 1 -2 m red argillite separates a lower volcanic unit composed of tholeiites and boninites from an upper unit dominated by boninites (Hebert 1983; Hebert & Bedard 2000). The volcanic and volcaniclastic rocks of the TMOC are everywhere capped by a thick-bedded polygenie breccia (the Coleraine Breccia) that contains ophiolitic and metasedimentary fragments in an epiclastic matrix, which appear to represent submarine debris flows (Hebert 1981). The Coleraine Breccia grades up and laterally into the SaintDaniel Melange (Schroetter et al. 2002).

Deformation A major problem in the structural analysis of ophiolites is the identification of marker surfaces allowing the geometry of deformation to be constrained. We used the orientation of chromitite bedding as palaeo-horizontal markers in the Dunitic Zone, of harzburgite-websterite rhythmic layering and mineral foliation as palaeo-horizontal markers in the Pyroxenitic Zone, and of pyroxenite-gabbro rhythmic layering and mineral folia-

240

J.-M. SCHROETTER ETAL.

Fig. 6. Photographs and sketches illustrating syn-oceanic deformation features from the upper crust, (a) Detailed map of Sector 3. Y volcanic rocks; FHb, 'doleritic' and gabbroic breccias; G, upper gabbro; SD, Saint-Daniel Melange. (b) Alternation of amygdaloidal mafic sills and gabbro to microgabbro breccias, all crosscut by a brittle normal fault. (c) Alternation of breccia composed of fine-grained mafic clasts (grey) and gabbro to microgabbro clasts (light grey). (d) Boudinaged boninitic sill emplaced in microgabbro breccia, (e) Interpretative sketch showing formation of the brecciated hypabyssal facies. Tl, oceanic crust temperature; T2, intrusion temperature.

tion as palaeo-horizontal markers in the Gabbroic Zone. Dyke contacts were considered to have been originally palaeo-vertical and oriented perpendicular to the spreading direction (Cann 1974), allowing the determination of the palaeo-stress field in the upper crust. Finally, bedding in fine-grained

sedimentary rocks occurring in the volcanosedimentary superstructure was assumed to originally have been horizontal. Although none of these assumptions are fail-safe, the fact that they all provide similar answers suggests that they are reasonable.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA In the AHM (Fig. 1), there are two main generations of faults. Younger reverse faults are associated with open folds and a regionally distributed, vertically dipping fracture cleavage trending c. NE-SW. An older set of subvertically dipping faults that trend north-south to NNESSW had not been recognized hitherto. The field relations we describe next lead us to believe that these older structures are related to intra-oceanic extension and sea-floor spreading. The origin of the older, high-temperature, tectono-magmatic deformation that affects much of the lower crust will not be discussed in this paper, and remains a major unresolved problem. Considered on the scale of the AHM, chromitite beds do not display a consistent orientation, with each major fault block (see below) having its own attitude. Near major north-south faults, chromitite s appear to be transposed into parallelism with the fault (Figs 3 and 4d). The magmatic foliation in the uppermost Dunitic Zone and Pyroxenitic Zone is much more regular, trending c, 125-130°, and dipping steeply (85-90°) to the SSW. Most dykes from the AHM are oriented c. north-south, dip steeply to the west (Fig. 3), and cut the foliation and layering in host plutonic rocks. In contrast to the great regularity shown by the Pyroxenitic Zone, the magmatic foliation in the Gabbroic Zone is much more variable, although always very steeply dipping. North of Lac Bisby, the foliation in the gabbros strikes between 40° and 170°, whereas at Lac de 1'Argile, it strikes c. 60° (Fig. 3). At Mt Adstock (Fig. 1), the layering in the gabbros is more regular, striking c. 110°. Dykes at Mt Adstock dip subvertically and strike 80°, as does the pyroxenite-gabbro contact. Cross-cutting relationships suggest that the pyroxenite-gabbro contact is, at least locally, an intrusive one, with pyroxenites cutting older gabbros. Our mapping shows that faults striking from north-south to 20° commonly terminate the along-strike extension of rocks of the Dunitic, Pyroxenitic and Gabbroic Zones, or separate plutonic from volcanic rocks (Figs 1 and 3). These north-south- to 20°-striking faults are locally reactivated by post-emplacement (Silurian and Acadian) deformational events. We have mapped in some detail an area that appears to be largely free of post-emplacement deformation (Fig. 3) to illustrate the distribution of lithologies, and the orientation of the various marker surfaces. It should be noted that, because this part of the AHM was tilted to the subvertical during postemplacement deformation (Silurian and Acadian), the map view (Fig. 3) is effectively a cross-section through the crust. Offsets between unit boundaries indicate an apparent sinistral motion in map view (Fig. 3). Apparent fault throws decrease from the

241

base (north) to the top (south), and are capped by deposition of synconvergence sediments of the Saint-Daniel Melange. The apparent throw calculated from the Dunitic-Pyroxenitic Zone boundary offset is c. 300 m on average, with a maximum of c. 1 km for the fault on the western flank of Nadeau Hill (Fig. 3). It should be noted also that the thickness of various lithological packages varies to either side of these faults. The map patterns indicate that the upper part of the crust in this part of the AHM was dissected by these north-south faults into a series of 100 m to 1 km blocks. Detailed observations from three sectors (labelled 1, 2 and 3 in Fig. 3) allow a kinematic analysis of this deformation. Sectors 1 and 2 are from the Dunitic and Pyroxenitic Zones, which are characterized by higher temperatures, whereas Sector 3 is from the cooler and more brittle upper crust.

Lower-crustal deformation We now describe the two sectors we have analysed in detail, and then present some generalizations. Sector 1 (Fig. 4, location shown in Fig. 3) is entirely contained within the Dunitic Zone. From west to east, massive dunite with disseminated chromite gives way (over 0.5 m) to a thick (hundreds of metres) breccia composed of angular, 1-10 cm clasts of dunite, locally with chromitite beds within them, in an orthopyroxenitic stockwork. As the main fault plane is approached, sheared serpentinites appear in the dunite, culminating in a 2-3 m wide serpentinite mylonite that marks the core of the fault. These breccias and mylonites are cross-cut by undeformed, tabular websterite and Iherzolite intrusions (30-50m wide), which are oriented north-south, parallel to the main fault. The overall geometry is illustrated by Figure 4a. The fault kinematics are clearly expressed in outcrops located c. 50 m east of the main fault plane, where chromitite beds strike north-south and dip 50° to the west (Fig. 4b and d). The chromitites are cut by shallowly dipping, east-west-striking websterite dykes. The websterite dykes lack chilled margins against their dunitic hosts, suggesting that these rocks were still hot at the time of dyke emplacement. A series of faults associated with serpentinite veins are parallel to dyke contacts and offset chromitite beds to the east (Fig. 4b). On the same outcrop, these same websterite dykes are chopped up into centimetrescale horst-and-graben structures by a series of conjugate, steeply dipping, north-south-striking normal faults (Fig. 4c). The history of deformation in Sector 1 can thus be divided into three increments (E1-E3, Fig. 4a). The early event (El) corresponds to the localized

242

J.-M. S C H R O E T T E R £ T ^ L .

development of a high-temperature layering-parallel fabric (chromitite schlieren and isoclinal folds (Kacira 1971)). Restoration to the horizontal of the chromitite beds (50° tilts, Fig. 4b and d) gives the websterite dykes and parallel faults of increment E2 a subvertical orientation, and so E2 faults are interpreted as having originally been steeply dipping, normal faults. The last increment of deformation (E3) defines a horst-and-graben system (Fig. 4c), marking a continuation of extension. Sector 2 (Fig. 5, location shown in Fig. 3) contains a fault that crosses the Dunitic-Pyroxenitic Zone contact. The Dunitic Zone here is very similar to that of Sector 1 as described above, whereas the Pyroxenitic Zone (Fig. 5bl) is composed of steeply dipping websterite and harzburgite layers that strike 125°. The main fault plane is highlighted by the presence of strongly sheared blue dunite. Immediately to the west of the main fault (in the Dunitic Zone), east-west-trending, shallowly south-dipping websterite dykes are offset sinistrally by steeply dipping, northsouth-striking faults oriented parallel to the main fault plane (Fig. 5a). In the Pyroxenite Zone to the east of the main fault plane, websterite layers embedded in harzburgite are stretched and boudinaged, and a mineral foliation is developed parallel to layering. Immediately adjacent to the main fault plane, layering becomes more chaotic, with the development of disharmonic folds (Fig. 5b2). The high-temperature fabrics are cut into horstand-graben structures by steeply dipping faults (Fig. 5bl). Restoration to the horizontal of the websteriteharzburgite layering converts the main fault of Sector 2 into a steeply dipping normal fault, with downthrow of the eastern compartment. Shearsense indicators and the orientation of faults affecting rocks of the Pyroxenitic and Dunitic Zones of Sector 2 have the same orientation as faults of episode E3 in Sector 1, and we infer a correlation. However, it is not clear whether the high-temperature fabric in Sector 2 corresponds to E2 or El in Sector 1. To summarize, both sectors register several extensional deformation events. As fault blocks are variably rotated by events E2 and E3, this explains why the early tectono-magmatic fabric (El) is not uniform in the AHM. The E3 event is particularly prominent, and affects most of the plutonic crust of the AHM. Restoration to palaeohorizontal of marker horizons implies that the principal E2 and E3 structures were originally steeply dipping normal faults that controlled the map-scale distribution of lithologies. Our mapping shows that massive, undeformed intrusions of peridotite and pyroxenite were commonly em-

placed in the immediate vicinity of the main faults (Fig. 3), in some cases being injected within the fault breccias, or along fault planes (Fig. 4d). This demonstrates the synmagmatic nature of the faults, and suggests that they guided the ascent of magmas.

Upper-crustal deformation Upper-crustal deformation (Gabbroic Zone, Hypabyssal and Extrusive sequences) is characterized by brittle fractures and brecciation (Fig. 6a and c), and by a structural control on the orientation of dykes (Fig. 3). Brittle to brittle-ductile faults in the upper crust are steeply dipping and oriented roughly north-south. They cut rocks of the Gabbroic Zone, the sheeted dyke complex, and the hypabyssal breccia facies (Fig. 6b). Most have small throws (tens of centimetres) and show an apparent sinistral sense of motion, similar to the faults in the lower crust. In the hypabyssal breccia facies, amygdaloidal intrusions injected into the microgabbroic breccia are stretched and offset by the faults (Fig. 6b and d), suggesting that faulting and magmatism were coeval. The faults must have played a role in brecciation, because some breccias have a cataclastic matrix, and our mapping shows that the brecciated hypabyssal facies is preferentially developed along the extension of the major north-south normal faults described in the previous section. However, the breccia matrix is generally igneous, and the jigsaw-puzzle morphology of the rocks (Fig. 6b) seems more compatible with some type of magmatic hydro-fracture mechanism, perhaps complemented by phreatomagmatic explosions caused by ascent of magma into rocks impregnated with seawater, and possibly by volume expansion caused by vesiculation of ascending, water-rich magmas (Fig. 6e). In any case, the faults would still have played a role in brecciation, as they would represent pathways for the ascent of magma and infiltration of seawater. Throughout the TMOC, dykes are typically oriented approximately north-south, subparallel to the synmagmatic normal faults described above. An exception is a dyke swarm located south of Nadeau Hill (Fig. 3), where the main dyke trend is c. 40°. It is difficult to know if this difference in the trend of the dykes from Nadeau Hill is an original extensional feature (suggesting progressive rotation of fault blocks; see Dilek et al. 1998), or if it reflects a syn- or post-emplacement reactivation. None the less, the general parallelism between dykes and normal faults throughout the AHM suggests a genetic link, whereby both features would have formed synchronously during east-west crustal extension (sea-floor spreading).

THETFORD MINES OPHIOLITE, QUEBEC, CANADA

Discussion Our fieldwork (Figs 1 and 3) suggests that several aspects of the geological history of the Thetford Mines Ophiolitic Complex (TMOC) need revision.

Sequence of crystallization and magmatic affinity The sequence of lithologies in the sections we have examined is: dunite (± orthopyroxenite), chromitite, dunite, harzburgite, orthopyroxenite or websterite, gabbro-norite, gabbro, hornblendeoxide gabbro, trondhjemite. Clinopyroxenite is rare, occurring only at the top of websterite layers. Wehrlite is also rare, occurring as late intrusions in AHM gabbros (Fig. 3). The orthopyroxenitic layers embedded in the basal dunite are thought to represent magmatic debris flows of some type, perhaps the result of roof disaggregation into underplating sills? Thus, our mapping and petrographic studies imply that the dominant sequence of crystallization of the TMOC is: chromite + olivine, orthopyroxene, (olivine-out), clinopyroxene, plagioclase, hornblende (pyroxene-out), Fe-Ti-oxides, quartz. This differs from previous interpretations (Laurent 1975; Laurent et al. 1979; Hebert 1983), in which it was proposed that clinopyroxene generally appeared before orthopyroxene, and that wehrlites and clinopyroxenites were common lower-crustal lithologies. The early appearance and abundance of orthopyroxene, typically low in Ti (Hebert & Bedard 2000), together with the high Cr/Al ratio of chromite (Hebert & Bedard 2000), and results of trace element inversion modelling (Bedard et al. 2001), all indicate a boninitic affinity for the common plutonic rocks of the TMOC. Because most dykes (Bedard et al. 2001) and lavas (Hebert & Bedard 2000) are also boninitic, this finding implies that the dominant crust-forming magmatic phase of the TMOC had a boninitic affinity. This supports the inference made by Church (1977) that the TMOC represents an unusual oceanic crust of Betts Cove type (Bedard et al 1998,2000).

The sheeted dyke complex and crustal polarity Previous maps of the TMOC did not show a sheeted dyke complex (Laurent et al. 1979; Hebert 1980; Beulac 1982; Hebert 1983), although concentrations of minor intrusions were recognized locally at the interface between lavas and gabbros. Part of the problem was that the map-scale lithological polarity (mantle-dunite-

243

pyroxenite-gabbro-lavas) was interpreted to signify that 'way up' was towards the SE, which implied that the minor intrusions were sills. These interpretations then led to a debate on whether or not the TMOC was really an ophiolite formed by sea-floor spreading (Church 1977; Laurent 1978). Our fieldwork shows that the orientation of early layering structures, and contacts between major lithological packages (e.g. Dunitic-Pyroxenitic and Pyroxenitic-Gabbroic Zones) indicate a way up towards the south to SSW in the AHM (Fig. 1). As most minor intrusions are oriented roughly north-south, a south to SSW way up means that they are dykes, and that concentrations of these dykes (>40% on our map) constitute a sheeted dyke complex, which in turn implies formation by sea-floor spreading. Most dykes are parallel to syn-oceanic normal faults, which implies that they were emplaced into planes of minimum compressional stress during east-west crustal extension. The localized absence of a sheeted dyke complex at the plutonic-lava interface can be explained as the result of intra-oceanic erosion-denudation, compounded by fragmentation by subvolcanic explosions responsible for formation of the brecciated hypabyssal facies.

Volcanic stratigraphy The lava sequences in the Lac de 1'Est and Mt Ham areas are composed of a lower mixed tholeiitic + boninitic volcanic unit, a thin red argillite, and an upper boninitic volcanic unit (Laurent & Hebert 1977; Beulac 1982; Hebert 1983; Oshin & Crocket 1986). However, our mapping (Figs 1 and 3) indicates that the thickness of the lavas is extremely variable, with the greatest accumulations being located within faultbounded basins (graben). Furthermore, the two extrusive units found in the Lac de LEst and Ham areas are not everywhere present. The tholeiites appear to be restricted to the basal sections of the main graben, and we speculate that they might represent local preservation of a tholeiitic protocrust. Preliminary data suggest that a few gabbros in the TMOC also have tholeiitic affinities (though most are boninitic), and may also represent components of a proto-crust. This may explain the more complex structural pattern recorded in the Gabbroic Zone, relative to the Pyroxenitic Zone. Alternatively, the change in fabric orientations may represent a shift from a pattern dominated by crust-mantle shear in the ultramafic rocks, to a more localized control of deformation pattern in the upper crust (gabbros).

244

J.-M. SCHROETTER ETAL.

A preliminary model for the structural and magmatic evolution of the TMOC The style and sequence of syn-oceanic deformation we have documented (Figs 4 and 5) allow us to propose a possible evolutionary scenario for the main, boninitic phase of the TMOC (Figs 7 and 8). The starting point is a layered plutonic crust composed of dunite at the base, pyroxenite in the middle, and gabbro at the top, into which is rooted

a sheeted dyke complex that fed overlying lavas (Fig. 8a). During a first, high-temperature, synmagmatic deformation event, many of the plutonic rocks were transposed and recrystallized along planes subparallel to original bedding planes. The cause of this first event is not certain. It may reflect shear between crust and mantle driven by a diapir (Nicolas 1992; Rabinowicz et al 1993), extensional shear as crust slides away from the ridge axis, or may represent the way deformation

Fig. 7. (a) Interpretative palaeogeographical reconstruction of the Thetford Mines Ophiolitic Complex before its emplacement. Colours are the same as in Fig. 1. (b) Idealized map cross-section of the Adstock-Ham Massif.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA

245

is partitioned in an extending ductile crust at the spreading centre (Tapponier & Francheteau 1978; Harper 1985). The overall regime remains tensional, however, and continuing magmatism is expressed as dykes of peridotite and pyroxenite (E2, Fig. 8b), some of which were probably emplaced as sills within preexisting cumulates. The extent to which rocks of the Dunitic and Pyroxenitic Zones belong to the early cumulate 'stratigraphy', and the extent to which they represent under- and intra-plating intrusions is not certain (seeThy 1990; Bedard 1991, 1993). In many cases, these intra-cumulate intrusions would have been transposed and recrystallized during continuing high-temperature deformation, and so would no longer be recognizable as late intrusions. The presence of pyroxenitic debris within graded layers near the base of the Dunitic Zone implies that some dunites are cumulates formed within intra-plating or under-plating sills, with the pyroxenitic debris being derived from disaggregation of a pre-existing roof. The high-temperature foliation is truncated by the major north-south faults we have documented. On outcrop and map scales, these faults dissect the crust into horst-and-graben structures. Kilometre-scale tilted blocks develop in the rigid upper plutonic crust and overlying extrusive rocks. An upward decrease in throw on these faults indicates that they are growth faults, propagating from bottom to top. Associated intrusions (Figs 3 and 6) imply that these faults were coeval with magmatism and played a role in magma ascent. The Dunitic Zone is only partly affected by these faults, and the lower part of the Dunitic Zone could represent a decollement surface accommodating movement of the tilted blocks (e.g. Harper 1985). The presence of a brecciated hypabyssal facies in areas where the sheeted dyke complex is absent suggests that erosion and/or tectonomagmatic destruction of upper-crustal facies was probably an important part of the geological history of the TMOC (Fig. 8c). Fig. 8. Schematic illustrations of a possible evolutionary scenario for the main, boninitic, crust-forming event of the Thetford Mines Ophiolitic Complex, (a) Event El. Development of a high-temperature foliation in cumulate rocks, (b) Event E2. Synkinematic pyroxenitic intrusions (grey arrows) are emplaced and transposed by continuing high-temperature tectonomagmatic deformation, as are harzburgite and websterite cumulates. Faults begin to form when magmatism wanes. Subvolcanic and talus(?) breccias form in the uppermost crust, (c) Event 3. Main stage of faulting, with only minor late-kinematic intrusions and associated lavas. Crust breaks up into horsts and grabens, and blocks are tilted along a basal decollement. The tops of tilted blocks are eroded, locally removing much of the upper crust.

The TMOC, a slow-spreading environment? The morphology of the axial rift zone of spreading centres is dependent on spreading rate, with prominent fault scarps and a deep axial valley characterizing slow-spreading environments (Macdonald 1982). Seismic data imply that magma chambers are ephemeral at slow-spreading ridges, with the depth of seawater penetration, and the depth of the brittle-ductile transition depending on the presence or absence of magma chambers (Harper 1985; Toomey et al. 1988; Dick et al 1992; Dilek & Eddy 1992; Dilek et al 1998). The deep graben of the TMOC, the deep crustal

246

J.-M. S C H R O E T T E R ^ r ^ L .

Fig. 9. (a) Schematic illustrations of a possible configuration for the genesis of boninites (after Deschamps & Lallemand 2003). (b) Model of boninite formation for southern Quebec Appalachian ophiolites.

THETFORD MINES OPHIOLITE, QUEBEC, CANADA level to which synmagmatic extensional faults penetrate (Figs 1 and 3), and the evidence for near-pervasive lower-crustal hydrothermal metamorphism are all features compatible with a slow spreading rate. An unresolved problem is the nature and location of the accommodation zone (Harper 1985) for the normal faults we have documented (Figs 1 and 3). We have been able to follow the faults down into the Dunitic Zone, but because of the poor exposure and common post-emplacement reactivation of Dunitic Zone rocks, we cannot be certain whether rotation of upper-crustal fault blocks occurred along a decollement zone, whether it penetrated down into the mantle, or whether the deformation was accommodated by a wide, ductile, lower crust. The tectonic exhumation of lower-crustal and mantle rocks along shallowly dipping extensional faults seems to characterize slow-spreading environments (e.g. Dick et al 1992; Dilek et al. 1998). In the TMOC, the deposition of lavas and sediments directly upon mantle or lower-crustal rocks (Fig. 1) could perhaps be explained in this manner.

Formation ofboninitic crust in the northern Appalachians The formation of boninite lavas requires a hot, depleted mantle affected by subduction zone metasomatism (Crawford et al. 1989). Several tectonic environments have been suggested to explain boninite genesis, including: subduction of a spreading ridge; plume-subduction zone interactions; subduction zone initiation; propagation of a backarc spreading centre into a subduction zone; slab rollback caused by a decrease in convergence rate (e.g. Crawford et al. 1989; Stern & Bloomer 1992; Bedard et al. 1998; Macpherson & Hall 2001; Kim & Jacobi 2002; Deschamps & Lallemand 2003). The Taconian ophiolites of the northern Appalachians are unusual in that boninites are extremely abundant, may constitute the dominant magmatic suite over extensive regions and are roughly coeval (490-480 Ma, e.g. Church 1977; Bedard et al. 2000; Bedard & Kim 2002; Kim & Jacobi 2002). Any genetic model must explain the development of extensive boninitic magmatism at this juncture. The identification of sub-north-south extensional oceanic structures in the TMOC suggests that the spreading centre had a north-south trend and that extension and sea-floor spreading took place along an east-west flow direction. Northsouth lineaments have also been reported from other oceanic terranes in southern Quebec and northern New England (Doolan et al. 1982;

247

Tremblay & Malo 1991; Tremblay 1992), suggesting that this may be the orientation of the extensional (spreading) axis at this time. Ordovician palaeogeographical reconstructions (Tremblay 1992) suggest a SE-dipping subduction zone, which would produce a diachronous intersection between spreading centre and subduction zone (Fig. 9), a configuration compatible with the genetic model proposed for boninites by Deschamps & Lallemand (2003). However, this assumes that the north-south trend documented for the TMOC spreading axis has not been affected by rotation of the obducted oceanic terrane, as has been documented in other ophiolites (e.g. Perrin et al. 1993). Another objection to this model is that it requires a very stable geometry to account for development of similar, roughly coeval, boninitic oceanic crust along >1500 km of strike length in the northern Appalachians. In contrast, models involving collision of the Taconic arc against an irregular continental margin and coeval formation of forearc spreading centres in re-entrants (Fig. 9: Cawood & Suhr 1992; Bedard & Kim 2002; Kim & Jacobi 2002) may be more compatible with the apparently synchronous development of anomalous boninitic crust of Betts Cove type (Church 1977) over such an extensive region.

Conclusions Our data highlight the links that exist between the structural and magmatic history of the Thetford Mines Ophiolitic Complex (TMOC; Fig. 7). We have documented the manner in which synmagmatic normal faults dissected the upper crust into tilted fault blocks and controlled deposition of lavas and sediments. These observations imply that the ophiolite was partially dismembered by extensional tectonics before its emplacement onto the continental margin (Fig. 7a). The dominance of a boninitic signature in cumulates and lavas suggests that the TMOC is a forearc-related ophiolite. The evidence for coeval extension and magmatism, and the discovery of a locally well-developed sheeted dyke complex suggest that the TMOC formed when sea-floor spreading was initiated in a forearc, with slow magmatic extension being fed by boninitic, rather than tholeiitic magmas. Emplacement onto the continental margin followed very quickly after the ophiolite formed, and a piggy-back forearc basin was developed. Thanks are due to R. Hebert, R. Laurent and the late P. St-Julien for introducing us to the area and sharing their knowledge during earlier field seasons; to B. Brassard and Ressources Allican Inc. for initiating the project; to N. Pinet, Y. Hebert and F. Huot for numerous

248

J.-M. SCHROETTER ETAL.

discussions; to P. Cousineau and the late lamented G. Kessler for volcanological and sedimentological insights; and to Y. Dilek and P. Robinson for insightful and lightning-fast reviews. This project has been supported by the Geological Survey of Canada, by Canadian National Science and Engineering Research Council grant ESS 233685-99, by a Diversification de 1'Exploration Minerale au Quebec grant provided by ValorisationRecherche Quebec (project 2201-133), and by Ressources Allican Inc. This is Geological Survey of Canada Contribution 2002206.

References BEDARD, J.H. 1991. Cumulate recycling and crustal evolution in the Bay of Islands ophiolite. Journal of Geology, 99, 225-249. BEDARD, J.H. 1993. The oceanic crust as a reactive filter: multiple syn-kinematic intrusion, hybridization and assimilation in an ophiolitic magma chamber. Geology, 21, 77-80. BEDARD, J.H. & KIM, J. 2002. Boninites in NE Appalachian Notre-Dame Subzone ophiolites: tectonic models for lapetus terranes. Geological Society of America, Abstracts with Programs, 34, A60. BEDARD, J.H., LAUZIERE, K., TREMBLAY, A. & SANGSTER, A. 1998. Evidence for forearc seafloor-spreading from the Betts Cove ophiolite, Newfoundland: oceanic crust of boninitic affinity. Tectonophysics, 284, 233-245. BEDARD, J.H., LAUZIERE, K., TREMBLAY, A., SANGSTER, A., DOUMA, S.L. & DEC, T. 2000. Betts Cove Ophiolite and its Cover Rocks, Newfoundland. Geological Survey of Canada Bulletin, 550. BEDARD, J.H., TREMBLAY, A., SCHROETTER, J.-M., BECU, V. & PAGE, P. 2001. Structural and magmatic evolution of the Thetford Mines Ophiolite: preliminary results from new mapping and analysis. Geological Association of Canada & Mineral Association of Canada Joint Annual Meeting. StJohn s, NF, Canada. Program with Abstracts, 26, 11. BERCLAZ, A., BEDARD, J.H. & HEBERT, R. 1998. Importance and significance of deformation processes in the genesis of oceanic crust in a slowspreading environment: evidences from the Bay of Islands Ophiolite, Newfoundland. Geological Association of Canada & Mineralogical Association of Canada Joint Annual Meeting, Quebec, QC, Canada. Program with Abstracts, 23, A-15. BEULAC, R.X. 1982. Etude petrologique du Complexe ophiolitique du lac Nicolet. MSc thesis, Universite Laval, Quebec. BLOOMER, S.H., TAYLOR, B., MACLEOD, C.J., STERN, R.J., FRYER, P., HAWKINS, J.W. & JOHNSON, L. 1995. Early arc volcanism and the ophiolite problem: a perspective from drilling in the western Pacific. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 1-30. BRASSARD, B. & TREMBLAY, A., 1999. Synthese geologi-

que et metallogenique de la M.R.C. de I'Amiante. Rapport final prepare pour la societe d'aide au developpement de la collectivite de 1'Amiante (SADC), 1/50000 scale map, northern and southern sheets of the Amiante MRC. CANN, J.R. 1974. A model for ocean crustal structure developed. Geophysical Journal of the Royal Astronomical Society, 39, 169-187. CASEY, J.F., KARSON, J.A., ELTON, D.L., ROSENCRANTZ, E. & TITUS, M. 1983. Reconstruction of the geometry of accretion during formation of the Bay of Islands Ophiolite Complex. Tectonics, 2, 509-528. CASTONGUAY, S., RUFFET, G., TREMBLAY, A. & FERAUD, G. 2001. Tectonometamorphic evolution of the southern Quebec Appalachians: 40Ar/39Ar evidence for Middle Ordovician crustal thickening and Silurian-Early Devonian exhumation of the internal Humber Zone. Geological Society of America Bulletin, 113, 144-160. CAWOOD, P.A. & SUHR, G. 1992. Generation and obduction of ophiolites: constraints from the Bay of Islands Complex, western Newfoundland. Tectonics, 11(4), 884-897. CHURCH, W.R. 1977. The ophiolites of southern Quebec: oceanic crust of Betts Cove type. Canadian Journal of Earth Sciences, 14, 1668-1673. CHURCH, W.R. 1978. The ophiolites of southern Quebec: oceanic crust of Betts Cove type: Reply. Canadian Journal of Earth Sciences, 15, 18821883. CHURCH, W.R. 1987. The geochemistry and petrogenesis of ophiolitic volcanic rocks from Lac de 1'Est, Thetford Mines Complex, Quebec, Canada: Discussion. Canadian Journal of Earth Sciences, 24, 1270-1273. CLAGUE, D., FRANKEL, C.S. & EABY, J.S. 1985. The age and origin of felsic intrusions on the Thetford Mines ophiolite, Quebec. Canadian Journal of Earth Sciences, 22, 1257-1261. COLEMAN, R.G. 1971. Plate tectonic emplacement of upper mantle peridotites along continental edges. Journal of Geophysical Research, 76, 1212-1222. COOKE, M.H.-C. 1938. Region de Thetford, de Disraeli et de la Moitie orientate de Warwick (Quebec). Commission Geologique du Canada, Memoire, 211, 176. COUSINEAU, P.A. & ST-JULIEN, P. 1992. The SaintDaniel Melange: evolution of an accretionary complex in the Dunnage terrane of the Quebec Appalachians. Tectonics, 11, 898-909. CRAWFORD, A.J., FALLOON, T.J. & GREEN, D.H. 1989. Classification, petrogenesis and tectonic setting of boninites. In: CRAWFORD, AJ. (ed.) Boninites and Related Rocks. Unwin Hyman, London, 1-49. DAVID, J. & MARQUIS, R. 1994. Geochronologie U-Pb dans les Appalaches du Quebec: application aux roches de la zone de Dunnage. Revue Geologique du Quebec, 1, 16-20. DEROSIER, C. 1971. Etude geologique des breches de la region de Thetford Mines. PhD thesis, Universite de Paris. DESCHAMPS, A. & LALLEMAND, S. 2003. Geodynamic setting of Izu-Bonin-Mariana boninites. In: LAR-

THETFORD MINES OPHIOLITE, QUEBEC, CANADA TER, R.D. & LEAT, P.T. (eds) Infra-Oceanic Subduction Systems: Tectonic and Magmatic Processes. Geological Society of London, Special Publications, 219, 163-185. DICK, H.J.B., ROBINSON, P.T. & MEYER, P.S. 1992. The plutonic foundation of a slow-spreading ridge. In: DUNCAN, R.A. ET AL. (eds) The Indian Ocean: A Synthesis of results from the Ocean Drilling Program. Geophysical Monograph, American Geophysical Union, 70, 1-92. DICKINSON, W.R. & SEELY, D.R. 1979. Structure and stratigraphy of forearc regions. AAPG Bulletin, 63, 2-31. DICKINSON, W.R., KLUTE, M.A., HAYES, M.J., JANECKE, S.U., LUNDIN, E.R., McKiTTRicK, M.A. & OLIVARES, M.D. 1988. Paleogeographic and paleotectonic setting of Laramide sedimentary basins in the central Rocky Mountain region. Geological Society of America Bulletin, 100, 1023-1039. DILEK, Y. & EDDY, C.A. 1992. The Troodos (Cyprus) and Kizildag (S. Turkey) ophiolites as structural models for slow-spreading ridge segments. Journal of Geology, 100, 305-322. DILEK, Y., MOORES, E.M. & FURNES, H. 1998. Structure of modern oceanic crust and ophiolites and implications for faulting and magmatism at oceanic spreading centers. In: BUCK, R., KARSON, J., DELANEY, P. & LAGABRIELLE, Y. (eds) Faulting and Magmatism at Mid-Ocean Ridges. American Geophysical Union, Geophysical Monograph, 106, 219-266. DOOLAN, L.D., GALE, M.H., GALE, P. & HOAR, R. 1982. Geology of the Quebec Re-entrant: possible constraints from early rifts and the Vermont-Quebec serpentine belt. In: ST-JULIEN, P. & BELAND, J. (eds) Major Structural Zones and Faults of the Northern Appalachians. Geological Association of Canada, Special Papers, 24, 87-115. DUNNING, G.R., KROGH, T.E. & PEDERSON, R.B. 1986. U-Pb zircon ages of Appalachian-Caledonian ophiolites. Terra Cognita, 6, L51. FEININGER, T. 1981. Amphibolite associated with the Thetford Mines Ophiolite Complex at Belmina Ridge, Quebec. Canadian Journal of Earth Sciences, 18, 1878-1892. HAMILTON, W.B. 1988. Plate tectonics and island arcs. Geological Society of America Bulletin, 100, 15031527. HAMILTON, W.B. 1995. Subduction systems and magmatism. In: SMELLIE, J.L. (ed.) Volcanism Associated with Extension at Consuming Plate Margins. Geological Society, London, Special Publications, 81, 3-28. HARNOIS, L. & MORENCY, M. 1989. Geochemistry of Mount Orford Ophiolite Complex, Northern Appalachians, Canada. Chemical Geology, 77, 133-147. HARPER, G.D. 1985. Tectonics of slow spreading midocean ridges and consequences of a variable depth to the brittle/ductile transition. Tectonics, 4, 395409. HEBERT, R. 1980. Etude petrologique des roches d'Asbestos et du Mont Ham (Ham Sud), Quebec. MSc thesis, Universite Laval, Quebec. HEBERT, R. 1981. Conglomerats polygeniques ophiolitiques: anciens eboulis de talus de fond oceanique.

249

Canadian Journal of Earth Sciences, 18, 619-623. HEBERT, R. 1985. Petrologie des roches ignees oceaniques et comparaison avec les complexes ophiolitiques du Quebec, de Chypre et de I'Apennin. PhD thesis, Universite de Bretagne occidentale, Brest, France. HEBERT, R. & BEDARD, J.H. 2000. Les ophiolites d'avant-arc et leur potentiel mineral: exemple des complexes ophiolitiques du sud du Quebec. Chronique de la Recherche Miniere, 539, 101-117. HEBERT, R. & LAURENT, R. 1989. Mineral chemistry of ultramafic and mafic plutonic rocks of the Appalachian ophiolites, Quebec, Canada. Chemical Geology, 77, 265-285. HEBERT, Y. 1983. Etude petrologique du Complexe Ophiolitique de Thetford Mines, Quebec. PhD thesis, Universite Laval, Quebec. HUOT, F., HEBERT, R. & TURCOTTE, B. 2002. A multistage magmatic history for the genesis of the Orford Ophiolite (Quebec, Canada): a study of the Mt Chagnon massif. Canadian Journal of Earth Sciences, 39, 1201-1217. KACIRA, N. 1971. Geology ofchromitite occurrences and ultramafic rocks of the Thetford Mines-Disraeli area, Quebec. PhD thesis, University of Western Ontario, London. KARSON, J.A., ELTHON, D. & DELONG, S.E. 1983. Ultramafic intrusions in the Lewis Hills Massif, Bay of Islands Ophiolite Complex, Newfoundland: implications for the igneous processes at oceanic fracture zones. Geological Society of American Bulletin, 94, 15-29. KIM, J. & JACOBI, R.D. 2002. Boninites: characteristics and tectonic constraints, northeastern Appalachians. Physics and Chemistry of the Earth, 27, 109-147. LAURENT, R. 1975. Occurrences and origin of the ophiolites of the Southern Quebec, Northern Appalachians. Canadian Journal of Earth Sciences, 12, 443-455. LAURENT, R. 1978. The ophiolites of southern Quebec: oceanic crust of Betts Cove type: Discussion. Canadian Journal of Earth Sciences, 15, 18801881. LAURENT, R. & HEBERT, Y. 1977. Features of submarine volcanism in ophiolites from the Quebec Appalachians. In: BARAGAR, W.R. (ed.) Volcanic Regimes in Canada. Geological Association of Canada, Special Papers, 16, 91-109. LAURENT, R. & HEBERT, R. 1989. The volcanic and intrusive rocks of the Quebec Appalachian ophiolites (Canada) and their island-arc setting. Chemical Geology, 77, 287-302. LAURENT, R., HEBERT, R. & HEBERT, Y. 1979. Tectonic setting and petrological features of the Quebec Appalachian ophiolites. In: MALPAS, J. & TALKINGTON, R.W. (eds) Ophiolites of the Canadian Appalachians and Soviet Urals. International Geological Correlation Project, 39, 53-77. LYTWYN, J.N. & CASEY, J.F. 1993. The geochemistry and petrogenesis of volcanics and sheeted dikes from the Hatay (Kizildag) ophiolite, southern Turkey: possible formation with the Troodos ophiolite, Cyprus, along fore-arc spreading centers. Tectonophysics, 223, 237-272.

250

J.-M. SCHROETTER £T,4L.

MACDONALD, K.C. 1982. Mid-ocean ridges: fine scale tectonic, volcanic and hydrothermal processes within the plate boundary zone. Annual Review of Earth and Planetary Sciences, 10, 155-190. MACPHERSON, C.G. & HALL, R. 2001. Tectonic setting of Eocene boninite magmatism in the Izu-BoninMariana forearc. Earth and Planetary Science Letters, 186, 215-230. NICOLAS, A. 1992. Kinematics in magmatic rocks with special reference to gabbros. Journal of Petrology, 33, 891-915. OSBERG, P.H. 1978. Synthesis of the geology of northeastern Appalachians. In: TOZER, E.T. & SCHENK, P.E. (eds) Caledonian-Appalachian Orogen of the North Atlantic Region. Geological Survey of Canada, Papers, 78-13, 137-147. OSHIN, I.O. & CROCKET, J.H. 1986. The geochemistry and petrogenesis of ophiolitic volcanic rocks from Lac de 1'Est, Thetford Mines Complex, Quebec, Canada. Canadian Journal of Earth Sciences, 23, 202-213. PAGE, P., BECU, V., TREMBLAY, A., BEDARD, J.H. & BRASSARD, B. 2001. Occurrences of platinumpalladium-enriched chromitites in the Thetford Mines ophiolite, southern Quebec. Geological Society of America, Northeastern Section, Abstracts with Programs, 33(1), 58-59. PEARCE, J.A., LIPPARD, S.J. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction ophiolites. In: KOKELAAR, B.P. & HoWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PERRIN, M., PREVOT, M. & BRUERE, F. 1993. Rotation of Oman ophiolite and initial location of the ridge in the hotspot reference frame. Tectonophysics, 229, 31-42. PINET, N. 1995. Cartographic du secteur Lac BisbyPetit Lac Saint-Francois, Propriete Coleraine. QERDEM B Dossier d'exploration miniere, Ressources Minieres Coleraine Inc., GM 53715, 33. PINET, N. & TREMBLAY, A. 1995. Tectonic evolution of the Quebec-Maine Appalachians: from oceanic spreading to obduction and collision in the north Appalachians. American Journal of Science, 295, 173-200. PINET, N., CASTONGUAY, S. & TREMBLAY, A. 1996a. Thrusting and back thrusting in the Taconian internal zone, southern Quebec Appalachians. Canadian Journal of Earth Sciences, 33, 1283-1293. PINET, N., TREMBLAY, A. & SOSSON, M. 1996b. Extension versus shortening models for hinterland-directed motions in the southern Quebec Appalachians. Tectonophysics, 267, 239-256. RABINOWICZ, M., Rouzo, S., SEMPERE, J.C. & ROSEMBERG, C. 1993. Three-dimensional mantle flow beneath mid-ocean ridges. Journal of Geophysical Research, 98, 7851-7869. ROBINSON, P.T. & MALPAS, J. 1990. The Troodos ophiolite of Cyprus: new perspectives on its origin and emplacement. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crust Analogues, Proceeedings of Symposium 'TROODOS 1987'. Geological Survey

Department, Nicosia, 13-26. ROBINSON, P.T., MELSON, W.G., O'HEARN, T. & SCHMINCKE, H.-U. 1983. Volcanic glass compositions of Troodos ophiolite, Cyprus. Geology, 11, 400-404. ROBINSON, P., TUCKER, R.D., BRADLEY, D., BERRY, H.N. & OSBERG, P.H. 1998. Paleozoic orogens in New England. Geologiska Foreningens i Stockholm Forhandlingar, 120, 119-148. ST-JULIEN, P. 1987. Geologic des Regions de SaintVictor et de Thetford-Mines (moitie est). Ministere de 1'Energie et des Ressources du Quebec, Quebec, MM86-01. SCHROETTER, J.-M., BECU, V., PAGE, P., TREMBLAY, A.

& BEDARD, J.H. 2000. Chromitites Ophiolitiques de la Region de Thetford MinesRapport des Activites de Terrain de la Saison Ete. Ministere de 1'Energie et des Ressources du Quebec, Quebec, GM 58649. SCHROETTER, J.-M., PAGE, P., TREMBLAY, A. & BEDARD, J.H. 2002. Structural evolution of the ThetfordMines Ophiolitic Complex, Quebec: from synoceanic rifting to obduction and post-obduction deformation. Geological Society of America, Northeastern Section, Abstracts with Programs, 34(1), A21. STERN, R.J. & BLOOMER, S.H. 1992. Subduction zone infancy: examples from the Eocene Izu-BoninMariana and Jurassic California arcs. Geological Society of America Bulletin, 104, 1621-1636. SUHR, G. & CAWOOD, P.A. 2001. Southeastern Lewis Hills (Bay of Islands Ophiolite): geology of a deeply eroded, inside-corner, ridge-transform intersection. Geological Society of America Bulletin, 113, 1025-1038. TAPPONIER, P. & FRANCHETEAU, J. 1978. Necking of the lithosphere and the mechanics of slowly accreting plate boundaries. Journal of Geophysical Research, 83, 3955-3970. THY, P. 1990. Cryptic variation of a cumulate sequence from the plutonic complex of the Troodos ophiolite. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTES, C. (eds) Ophiolites, Ocean Crustal Analogues, Proceedings of the Symposium 'TROODOS 1987'. Publisher, Nicosia, 165-172. TOOMEY, D.R., SOLOMON, S.C. & PURDY, G.M. 1988. Microearthquakes beneath the median valley of the Mid-Atlantic Ridge near 23°N: tomography and tectonics. Journal of Geophysical Research, 93, 9093-9112. TREMBLAY, A. 1992. Tectonic and accretionary history of Taconian oceanic rocks of the Quebec Appalachians. American Journal of Science, 292, 229252. TREMBLAY, A. & CASTONGUAY, S. 2002. The structural evolution of the Laurentian margin revisited (southern Quebec): implications for the Salinian Orogeny and Appalachian successor basins. Geology, 30(1), 7-82. TREMBLAY, A. & MALO, M. 1991. Significance of brittle and plastic fabrics within the Masawippi Lake fault zone, southern Canadian Appalachians. Journal of Structural Geology, 13(9), 1013-1023. TREMBLAY, A., BEDARD, J.H. & LAUZIERE, C. 1997. Taconian obduction and Silurian exhumation of the

THETFORD MINES OPHIOLITE, QUEBEC, CANADA Betts Cove ophiolite, Canadian Appalachians. Journal of Geology, 105, 701-716. UYEDA, S. & KANAMORI, H. 1979. Back-arc opening and the mode of subduction. Journal of Geophysical Research, 84, 1049-1061. VAN STAAL, C.R., DEWEY, J.F., MACNIOCAILL, C. & McKERROW, W.S. 1998. The tectonic evolution of the Northern Appalachians and British Caledonides: history of a complex, west and southwest Pacifictype segment of lapetus. In: BLUNDELL, D.J. & SCOTT, A.C. (eds) Lyell: the Past is the Key to the Present. Geological Society, London, Special Publications, 143, 199-242. WHITEHEAD, J., REYNOLDS, P.H. & SPRAY, J.G. 1995. The sub-ophiolitic metamorphic rocks of the Quebec Appalachians. Journal of Geodynamics, 19, 325-350. WHITEHEAD, J., DUNNING, G.R. & SPRAY, J.G. 2000.

251

U-Pb geochronology and origin of granitoid rocks in the Thetford Mines ophiolite, Canadian Appalachians. Geological Society of America Bulletin, 112, 915-928. WILLIAMS, H. 1979. Appalachian Orogen in Canada. Canadian Journal of Earth Sciences, 16, 792-807. WILLIAMS, H. & HATCHER, R.D. JR 1983. Appalachian suspect terranes. In: HATCHER, R.D. JR, WILLIAMS, H. & ZIETZ, I. (eds) Geophysics of Mountain Chains. Geological Society of America, Memoirs, 158, 33-53. WILLIAMS, H. & ST-JULIEN, P. 1982. The Baie VerteBrompton Line: Early Paleozoic continent-ocean interface in the Canadian Appalachians. In: STJULIEN, P. & BELAND, J. (eds) Major Structural Zones and Faults of the Northern Appalachians. Geological Association of Canada, Special Papers, 24, 177-207.

This page intentionally left blank

Cr-spinel compositions, metadunite petrology, and the petrotectonic history of Blue Ridge ophiolites, Southern Appalachian Orogen, USA LOREN A. RAYMOND 1 , SAMUEL E. SWANSON 2 , ANTHONY B. LOVE 1 & JAMES F. ALLAN 1 1 Department of Geology, Appalachian State University, Boone, NC 28608, USA (e-mail: raymondla@appstate. edu) 2 Department of Geology, University of Georgia, Athens, GA 30602, USA Abstract: Resolution of the petrotectonic history of Blue Ridge ophiolites of the Southern Appalachian Orogen has remained enigmatic because of metamorphism and tectonic fragmentation of ultramafic bodies. Understanding of this history is confounded by the presence of five partial metamorphic overprints and by similar Ti enrichments in spinels from Blue Ridge and modern mid-ocean ridge basalt ultramafic rocks that result from different processes. Chrome spinels from oceanic ultramafic lithosphere show increases in Ti caused by metasomatism induced by passing mafic melts, which create both dunite melt channels within harzburgite wall rocks and associated troctolite impregnation zones. In the Blue Ridge Belt, the oldest metadunite mineral association generally lacks high-Ti spinel, whereas the higher Ti spinels are relatively low in Al and Mg and occur in three amphibolite- to greenschist-facies retrograde metamorphic associations that occur in deformed, metasomatized ultramafic bodies with high aspect ratios. Some spinel compositions in the oldest mineral association are similar to those from arc-suprasubduction zone ultramafic lithosphere. Together, available data are consistent with the hypothesis that: (1) the Blue Ridge ophiolites are fragmented, metamorphosed, very slow-spreading ridge, Xigaze-type ophiolites, consisting of mafic rocks, minor plutonic rocks, and a sublithospheric ultramafic tectonite base; (2) the metadunites represent sublithospheric melt channels and zones of high melt flux, perhaps formed in a suprasubduction zone setting; (3) pre-Taconic subduction may have been west-directed rather than east-directed. The Taconic orogenesis deformed, fragmented, and metamorphosed the ophiolites; and later Taconic, Acadian, and Alleghenian metamorphism hydrated the bodies, while associated deformation exaggerated their elongation.

More than 200 metamorphosed ultramafic rock bodies occur as isolated, scattered pods, lenses, and blocks forming a crudely linear array in the Blue Ridge Belt of the Southern Appalachian Orogen (Fig. 1) (Pratt & Lewis 1905; Hunter 1941; Hess 1955; Larrabee 1966; Misra & Keller 1978). Such linear arrays of ultramafic rocks are now recognized as lithological markers of plate sutures within orogens. In the Blue Ridge Belt, discoveries of eclogite, retrograded eclogite, and block-in-matrix structures within the ultramafic array support the interpretation that this zone marks an Ordovician (Taconic) suture (Raymond et al. 1989; Willard & Adams 1994; Adams et al. 1995; Abbott & Raymond 1997; Ryan et al. 2001). The purposes of this paper are: (1) to review the occurrences, mineralogy, petrology, and metamorphism of the ultramafic rocks; (2) to present and discuss new data and interpretations bearing on the history of the ultramafic and

associated rocks; (3) to suggest tectonic implications for the Neoproterozoic to Ordovician (Taconic) orogenic history implied by these data. The origins of alpine ultramafic rocks of the Blue Ridge Belt have been difficult to decipher, because of the fragmented nature and polymetamorphic histories of the rocks. Commonly, alpine ultramafic rocks in suture zones are parts of ophiolites that represent fragments of oceanic or suprasubduction zone (SSZ) mafic-ultramafic crust and subcrustal mantle (e.g. Moores 1970, 1981; Nicolas & Poirier 1976, chapter 11; Juteau et al. 1977; Pearce et al. 1984; Edwards 1995). Some alpine ultramafic bodies also represent subcrustal mantle slabs thrust to the surface (Davies et al 1993; Green et al 1997). In the Blue Ridge Belt, a few studies suggest that at least some of the alpine ultramafic rock masses represent ophiolite fragments (McSween & Hatcher 1985; Tenthorey et al 1996; Warner 2001). Here, we use

From'. DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 253-278. 0305-8719/037$ 15 © The Geological Society of London 2003.

254

L. A. RAYMOND ETAL.

Fig. 1. Map of the Blue Ridge Belt showing major faults, thrust blocks (and equivalent terrane names), and the locations of select ultramafic rock bodies. B, Boone; BZ, Brevard Fault Zone; FB, Fries Block; FGLF, Fries-Gossan Lead Fault; GLB, Gossan Lead Block, Toe Terrane; GMW, Grandfather Mountain Window; HFF, Hayesville—Fries Fault; MHT, Mars Hill Terrane; NC, North Carolina; GA, Georgia; SC, South Carolina; TN, Tennessee. Numbered and starred meta-ultramafic bodies, mentioned in this paper or extensively studied are: 1, Edmonds; 2, Greer Hollow; 3, Day Book; 4, Woody; 5, Webster-Addie; 6, Corundum Hill; 7, Buck Creek; 8, Hoots. •, locations of additional meta-ultramafic bodies. (Modified from Larrabee 1966; Brown et al. 1985; Hopson et al. 1989; Adams et al. 1995; Raymond 1998.)

the term 'ophiolite' in a broad sense to refer to a crudely layered sequence of mafic volcanic rocks underlain by a complex of mafic to ultramafic plutonic rocks and a basal peridotite tectonite unit. Evidence that Blue Ridge ultramafic rocks are ophiolitic in character is limited, but the available data more closely match the characteristics of ophiolitic rocks than those of other mafic-ultramafic complexes. Support for the hypothesis that these rocks are ophiolitic includes: (1) the petrotectonic association of the rocks within the regional tectonic setting; (2) petrological data; (3) geochemical data (including isotopic data) from both meta-ultramafic rocks and associated metamafic rocks (Hatcher et al. 1984; McSween & Hatcher 1985; Misra & Conte 1991; Tenthorey et al. 1996). For example, the ultramafic rocks are commonly associated with metamafic rocks interpreted to be metabasalts and metagabbros. The metabasalts (amphibolites) are tholeiitic and typically plot in ocean-floor, within-plate, or arc basalt fields on geochemical discrimination diagrams (Hatcher et al. 1984; Misra & Conte 1991). Metatroctolites associated with the metadunites of the Buck Creek body have trace element contents consistent with crystallization from a mantlederived melt in a rift setting (Tenthorey et al.

1996) and some associated amphibolites may have been gabbros (McElhaney & McSween 1983). Finally, the occurrence of dunite with harzburgite is compatible with an ophiolitic origin. None of the available evidence supports alternative interpretations that the Blue Ridge rocks represent other types of mafic-ultramafic complexes, such as layered lopolithic complexes, appinite-type complexes, Alaska-type complexes, or alkalinetype complexes. Studies conducted over the past 40 years have characterized the Blue Ridge ultramafic rocks and revealed several aspects of their histories. All of the ultramafic bodies have been metamorphosed and the bodies and rocks lack any definitive features of igneous intrusions, such as intrusive contacts, dykes and apophyses, contact metamorphism, chilled margins, or igneous textures (Swanson 1981, 1999, 2001; Abbott & Raymond 1984; Raymond 1995, p. 667ff; Raymond & Abbott 1997; Raymond & Warner 2001). The ultramafic bodies form small (<5 m) to large (>1 km) masses (Fig. 2). Mineral assemblages and textures are clearly metamorphic. Up to five metamorphic recrystallization events have affected the character of the ultramafic rocks, with older events representing amphibolite-facies (and per-

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

Fig. 2. Maps and sections of selected meta-ultramafic rock bodies of the Blue Ridge Belt, (a) Day Book (map modified from Swanson 1981). (b) Grassy Creek, a moderately hydrated body (map from Brobst 1962). (c) Greer Hollow, a hydrated body, d, dunite; G, granitoid rocks; S, mica schist; UM, ultramafic rocks (mixed).

255

256

L. A. RAYMOND ETAL.

haps eclogite-facies) conditions, and younger events representing greenschist-facies conditions. Serpentinites are late-stage rocks produced in greenschist-facies conditions. Critical to unravelling the full history and tectonic significance of Blue Ridge meta-ultramafic rocks is the determination of their igneous or metamorphic protoliths. Clues to any igneous history may be hidden in mineral chemistry: Hence we examine some of that chemistry. Clarification of any igneous history that may exist also requires recognition and isolation of metamorphic features, plus resolution of a number of questions related to the later metamorphic and structural histories of the bodies. Towards that end, we provide data on and discussion of the nature and distribution of both anhydrous and almost entirely hydrated metamorphic rock bodies, the recognition of which is important to the analysis of the metamorphic history. The anhydrous bodies are composed mainly of metadunite, with local metaharzburgite. These rocks contain the highest grade mineral assemblages (olivine + Cr-spinel ± orthopyroxene) (Table 1). Commonly, the olivine, orthopyroxene, and Cr-spinel have re-equilibrated with hydrous metamorphic phases such as tremolite and chlorite (Swanson 2001). However, core compositions of some of the chrome spinels may retain a vestige of an igneous heritage. If so, as we suggest here, one possible source of these cores is Cr spinels from SSZ ophiolitic ultramafic rocks. The data however, are not unambiguous. The uncertainty means that we must assess whether the abundant metadunites are dehydrated serpentinite bodies derived from ophiolite fragments, are highgrade metamorphosed ophiolite fragments, or represent ultramafic bodies of other types. Beyond resolving the issue of the general nature of a possible protolith, we also must determine whether the anhydrous, olivine-rich bodies and hydrated bodies represent different specific protoliths (e.g. dunite v. clinopyroxenite) or the same protoliths with differences in degrees of hydration and metasomatism. The anhydrous body v. hydrated body data provide some information about

this question, but do not unequivocably resolve it. These data include the aspect ratios of the ultramafic bodies, which aid in addressing the related question: does the mode of occurrence, as subequant blocks v. highly elongated layers and lenses, reflect metamorphic or structural histories? The answers to the above questions are seemingly linked, because the petrological features control the rheological properties and strength of the rocks. The features and chemistry of some mid-ocean ridge rocks and their minerals bear some similarities to the features and chemistries of the Blue Ridge rocks and minerals. The chromium spinels from crust of the Hess Deep (a 2500 m deep fault basin at the west end of the Cocos-Nazca spreading centre, Pacific Ocean, Ocean Drilling Program (ODP) Leg 147) and the chrome spinels from Blue Ridge meta-ultramafic rocks, for example, both show enrichment in TiO2. We show that similarities in spinel chemistries reflect different petrological histories. Nevertheless, as petrogenetic indicators, the Blue Ridge spinel Al contents (and ranges of TiO2) are consistent with the hypothesis that the enclosing metadunites are SSZ rocks. Another similarity between spreading centre rocks and those of the Blue Ridge Belt, revealed by crustal gabbroic and mantle peridotite rocks recovered from six drill holes in the Hess Deep, is the presence of masses of dunite. The Hess Deep holes reveal a mid-ocean ridge-style ophiolite with dunite melt channels (Allan & Dick 1996). A comparison of features of Blue Ridge, Hess Deep, and other ophiolitic sublithospheric mantle rocks suggests the hypothesis that Blue Ridge metadunite protoliths represent sublithospheric melt channels and melt flux zones beneath a spreading centre. In this paper, we present data supportive of a melt channel-melt flux zone, SSZ, slow-spreading centre origin for Blue Ridge metadunites and discuss the Ti enrichments in Hess Deep and Blue Ridge spinels so as to assess the meaning of these data in terms of possible igneous and metamorphic histories.

Table 1. Metamorphic associations in Blue Ridge meta-ultramafic rocks Association A-l: Olivine ± Chromite ± Orthopyroxene ± Clinopyroxene ± Hornblende ± Magnetite Association A-2: Olivine ± Orthopyroxene ± Clinopyroxene ± Tremolite ± Magnesiocummingtonite ± Chlorite ± Chromite Association A-3: ± Chlorite ± Magnesiocummingtonite ± Anthophyllite ± Tremolite ± Talc ± Phlogopite ± Magnetite ± Magnesite ± Garnet Association A-4: ± Serpentine (Antigorite) ± Magnetite ± Chlorite ± Talc ± Brucite ± Tremolite Association A-5: ± Serpentine (Lizardite) ± Serpentine (Chrysotile) ± Magnetite ± Talc ± Chlorite ± Tremolite ± silica minerals (e.g. Opal) ± Magnesite ± Aragonite ± Garnierite Modified from Raymond (1995, table 31.3).

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

Methods of analysis Data for this continuing study of Blue Ridge meta-ultramafic rocks were collected both in the field and in the laboratory. In the field, mapping was conducted using traditional compass and pace methods of interpretive mapping and attitude measurement with a Brunton pocket transit. Cross-sections were constructed using both data collected in the field and data obtained from published mapping. Aspect ratios of ultramafic bodies are defined as, and were calculated from, the long axis of the meta-ultramafic bodies (usually measured from maps) divided by the maximum short axis measured perpendicular to the foliation plane of enclosing rocks. In the case of older literature, the foliation attitudes are often unreported and we used regional foliations as a guide to determine the direction in which to measure the short axis. In a few cases, the only measure possible was one of map width. In data taken from the literature, where only text descriptions were available, we used the dimensions provided. The samples used in this study were collected in the Hess Deep, a 2500 m deep fault basin at the west end of the Cocos-Nazca spreading centre in the eastern Pacific Ocean, and in the North Caroli-

257

na section of the Blue Ridge Belt of North America (Fig. 1). Hess Deep samples are from drill cores. Blue Ridge samples are grab samples selected as representative samples of the bodies being studied. Chemical analyses on minerals were conducted using electron microprobe techniques. Hess Deep samples were analysed by James Allan on a Cameca SX-50 microprobe using a focused beam, sample current of 30 nA, long counting times (for most elements), pure metal standards, and Smithsonian natural mineral standards (see Allan & Dick 1996). Blue Ridge samples were analysed by Sam Swanson, Anthony Love, Renee McCarter, and Loren Raymond using the University of Georgia, Department of Geology JEOL JXA 8600 Superprobe, with an accelerating voltage of 15 ky sample current of 15 nA, and a 2 urn beam diameter. Natural minerals were used as standards and several chromite substandards were used to check the analytical routine during each analytical run. Data were reduced online using a Bence & Albee matrix correction. Total iron was measured as FeO by the microprobe and was recalculated based on ideal stoichiometry to give the FeO and Fe2Os values. Typically, totals from analyses were in the range of 98 to >99 wt%. Low totals were probably related to assumptions involving iron

Table 2. Electron microprobe analyses of Cr-chlorites and Cr-spinels from Blue Ridge meta-ultramafic rocks Day Book DB 134 Chlorite

DB134 Cr-spinel

Frank

Henson Creek HC5 Chlorite

HC 5 Cr-spinel

0.030 32.250 0.050 32.460 SiO2 n.d. 0.050 0.050 n.d. TiO2 3.070 14.350 2.750 12.260 A1203 55.800 4.310 58.330 2.950 Cr203 0.42* 10.080 0.27* 7.770 Fe203 2.52* FeO 23.760 1.68* 21.540 NiO 0.050 0.170 0.220 0.070 n.d. 0.000 0.000 MnO n.d. 33.870 MgO 5.130 32.440 8.790 Total 97.650 83.430 97.460 86.740 Cations based on 28 oxygens (chlorite); 4 oxygens (chromite), cations normalized to 3 6.194 0.001 Si 6.387 0.001 Ti 0.001 0.001 Al 3.227 0.118 2.860 0.130 Cr 0.444 1.601 0.675 1.656 3+ 0.276 0.210 Fe 2+ _ _ 0.722 Fe 0.647 0.447 0.309 Fe2+totai 0.001 Ni 0.025 0.036 0.002 Mn 0.000 0.000 9.633 Mg 0.278 9.577 0.353

F84-5 Chlorite

F84-5 Cr-spinel

32.050 n.d. 13.720 2.250 0.32* 1.96* 0.140 n.d. 32.830 83.270

0.030 0.200 3.670 46.780 17.660 25.460 n.d. 0.000 4.250 98.050

6.322 3.191 0.351 _ 0.360 0.022 9.658

0.001 0.021 0.156 1.347 0.485 0.776 0.000 0.230

n.d., not detected. *Data obtained as FeOtotai and calculated based on stoichiometry (eight positive charges, three cations).

L. A. RAYMOND ETAL.

258

partitioning, but may be related to some unanalysed component. Mineral formulae calculated from the spinel analyses give satisfactory totals (Tables 2 and 7) and suggest the analyses are reasonable.

where a blackwall exists. No dykes have been observed extending from any of the bodies into the surrounding country rock and no zones of thermal (contact) metamorphism in surrounding rocks have been documented.

Field setting of Blue Ridge ultramafic rocks

Metamorphic character of Blue Ridge ultramafic rocks

Blue Ridge ultramafic rock bodies occur as isolated, concordant to discordant, scattered pods, lenses, slabs, and blocks in country rock of the Eastern and Central Blue Ridge belts (Table 3). These belts are composed predominantly of hornblende schist and gneiss ('amphibolite'), pelitic schist, quartzo-feldspathic schist and gneiss, or combinations of these rock types. As indicated in Figure 1, the bodies actually occur in two major thrust blocks (called terranes by some workers), the Gossan Lead Block and the Fries-Hayesville Block. Body lengths range from <5 m to >1 km. Figure 2 shows maps and cross-sections of three typical Blue Ridge meta-ultramafic bodies. In general, the long axes of the bodies parallel the regional foliation and in the case of the Greer Hollow body, the meta-ultramafic rocks can be shown to have experienced folding and foliation development along with the enclosing rocks. Thus, this body (and probably others) was emplaced before deformation and some metamorphism. The contacts between country rocks and meta-ultramafic rocks are typically either sharp contacts or thin gradational zones of schist that locally form a 'blackwall' metasomatic reaction zone around the meta-ultramafic body (Sanford 1978; Swanson 1981). Contacts are relatively sharp, even in cases

The basic metamorphic character of Blue Ridge meta-ultramafic rocks has been delineated through a series of studies conducted over the past 30 years (Astwood et al. 1972; Dribus et al. 1976; Swanson & Raymond 1976; Honeycutt & Heimlich 1980; Swanson 1980, 1981, 1999, 2001; Abbott & Raymond 1984; Raymond & Abbott 1985, 1997; Raymond 1995, chapter 31; Raymond et al. 1988; Raymond et al 1999, 2001; Swanson et al. 1999; Warner 2001). Five metamorphic associations consisting of small domain, equilibrium assemblages are recognized (Table 1). These assemblages are preserved, because many of the bodies have resisted pervasive hydration and accompanying recrystallization. Thus, older assemblages have been only partly replaced by successively younger ones. The oldest assemblages are the highest grade assemblages and successively younger ones are hydrated and are of lower grade. The five groups of mineral assemblages (associations) in the meta-ultramafic rocks represent a declining P—T path of metamorphism that developed over almost 200 Ma (Abbott & Raymond 1984; see Adams et al 1995). The oldest assemblages (Association A-l) consist of olivine, chromium spinel (chromite), and, in some cases, ormopyroxene (clinopyroxene is extremely rare,

Table 3. Characteristic features of typical Blue Ridge ultramafic bodies and rock types Rock type and occurrence Hydration Dunite bodies

Low

Harzburgite bodies and lenses in dunite

Low

Orthopyroxenite bands in Low dunite and harzburgite Partially hydrated dunite Moderate and harzburgite bodies Chlorite and talc schist bodies Serpentinite bodies and veins in dunite bodies

High High

Contacts

Textures

Mineral associations

Discordant with blackwall, sheared, discordant- sharp

Equigranular, equigranular-tabular, porphyroclastic

A-l, A-2

Equigranular, equigranular-tabular, porphyroclastic Equigranular

A-l, A-2

Concordant with blackwall Discordant with blackwall, sheared, discordant-sharp Sheared, discordant-sharp Concordant to discordant, sharp with local blackwall

Equigranular, lepidoblastic to nematoblastic, felted Sharp, generally concordant Lepidoblastic to nematoblastic Sharp to gradational Lepidoblastic, meshtextured

A-l (Opx ± Ol ± Chr) A-l, A-2, A-3

A-3 A-4, A-5

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

259

Fig. 3. Photomicrographs and a photograph of meta-ultramafic rocks, (a) Photomicrograph of metadunite with Association A-l from the Day Book body, showing an equilibrium metamorphic texture with 120° grain boundary triple junctions between olivine grains, (b) Photomicrograph of metadunite of Association A-2 from Corundum Hill body showing Cr-chlorite associated with Cr-spinel. (c) Photomicrograph of porphyroblastic metaharzburgite of Association A-2 with chlorite and tremolite from the Hoots Body, (d) Photomicrograph of chlorite-anthophyllite schist of Association A-3 from the Greer Hollow body, (e) Photomicrograph of serpentinite of Association A-4 containing magnetite from the Greer Hollow body, (f) Photograph of aragonite vein (light needles) cutting metadunite of the Day Book Dunite. Coin diameter is 22 mm. Long dimension of field in (a), (b), and (d) is 6.5 mm, in (c) is 8 mm, and in (e) is 1 mm. Minerals: A, anthophyllite; Ag, aragonite, C, chrome spinel; Ch, chlorite; M, magnetite; O, olivine; Op, orthopyroxene; S, serpentine minerals; Tr, tremolite.

but does occur locally) (Table 1; Fig. 3a). These assemblages typically represent anhydrous upper amphibolite-, granulite-, or possibly eclogitefacies conditions of metamorphism. Textures in the rocks of the oldest association are clearly metamorphic (e.g. Dribus et al. 1976; Abbott & Raymond 1984; Raymond 1995, p. 667ff). Mineral assemblages of Association A-2, which partly replace those of Association A-l, are slightly hydrated assemblages containing olivine + Cr-

spinel + Cr-chlorite ± orthopyroxene + tremolite ± magnesiocummingtonite (Fig. 3b and c). Association A-2 represents amphibolite-facies conditions, but probably of lower P and T than Association A-l. Association A-3 commonly contains chlorite in abundance, plus additional hydrated phases such as anthophyllite, magnesiocummingtonite, phlogopite, and talc (Fig. 3d). This association represents low amphibolite-facies P—T conditions (Raymond 1995, chapter 31;

260

L. A. RAYMOND ETAL.

Fig. 4. Petrogenetic grid showing fields of contact metamorphism (C), greenschist facies (G), amphibolite facies (A), and eclogite facies (E), plus P-T—t metamorphic path for the Gossan Lead Block (Ashe and Alligator Back metamorphic suites of the Toe Terrane) and Fries Block (Pumpkin Patch metamorphic suite of the Cullowhee Terrane, i.e. Mars Hill Terrane) metamorphism. Metamorphic path based on Butler (1973), Abbott & Raymond (1984), Tenthorey et al. (1996), Adams & Trupe (1997), Goldberg & Dallmeyer (1997), Waters et al. (2000), Abbott & Greenwood (2001), Miller et al. (2001), and Warner (2001). Associations A-l-A-5 of Table 1.

Tenthorey et al 1996; Raymond & Abbott 1997; Warner 2001). Associations A-4 and A-5 represent hydrated conditions of lower grade (Fig. 4). Association A4 is typified by serpentine that occurs as a meshwork surrounding olivine grains, but is also represented by serpentine with magnetite in veins (Fig. 3e). The youngest mineral assemblages in many meta-ultramafic masses (Association A-5) fill veins in the metadunite or metaharzburgite and consists of serpentine and magnetite or, less commonly, minerals such as talc, aragonite, or garnierite (Fig. 3f). Veins of Association A-5 cut the serpentine meshwork of Association A-4. The serpentine-bearing assemblages probably represent P-T conditions of T <500 °C, P <0.5 GPa, considering the P-T path recorded by these rocks (Fig. 4) (Adams & Trupe 1997; Swanson 2001). Texturally, structurally, and mineralogically, these younger associations are also metamorphic in character. The petrogenetic grid for the ultramafic rocks (Fig. 4) depicts the estimated P-T conditions of recrystallization for Associations A-l to A-5. The mineralogy of A-l and A-2 rocks and the wholerock chemistries of dunites (available in the published literature) reveal low Ca and Al contents (Table 4) (Kulp & Brobst 1954; Carpenter & Chen 1978; Kingsbury & Heimlich 1978; Raymond et al 2001; Warner 2001). The wide range

of stability of assemblages of Association A-l in bulk compositions with low calcium and aluminium content does not narrowly limit the P-T conditions under which this first recorded metamorphic event occurred, but geothermometry on rocks in the Buck Creek area (Fig. 1) suggests temperatures in excess of 775 °C (Absher & McSween 1985; Tenthorey et al 1996; Warner 2001). Pressure estimates for these rocks fall in the range of 1.0-1.2 GPa. Rocks associated with ultramafic rocks to the north, in the vicinity of the Blue Ridge Bakersville eclogite locality west of the Grandfather Mountain Window, yield similar peak P-T values of >700 °C at 1.3 GPa (Adams et al 1995; Adams & Trupe 1997; Page et al 2001). Pelitic and mafic rocks exposed farther north, in the vicinity of the Hoots and Greer Hollow meta-ultramafic bodies, north of Boone, also yield similar values of 600-800 °C and 0.71.6 GPa (Raymond et al 1988, 1998; McSween et al 1989; Abbott & Greenwood 2001). The P-T path for the Bakersville eclogites and associated rocks was examined by Adams et al. (1995) and Adams & Trupe (1997). The P-T path for these rocks is one of relatively rapid rise to peak conditions, followed by rapid and subsequent slower descent to lower T and lower P-T, respectively. Essentially the same path is suggested by mineral inclusion mermobarometry (Page et al. 2001). Meta-ultramafic rocks and eclogites occur

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

261

Table 4. Selected whole-rock chemistries of Blue Ridge ultramafic bodies

SiO2 TiO2 A1203 Fe2O3 FeO MnO MgO CaO Na20 K2O P205 H2O+ H2CT Other Total Association % Hydration Rock type

Henson Creek HC-4

Day Book DB

Hoots RM116

Buck Creek BC-7

Greer Hollow GH

41.50 0.05 0.34 1.33 5.90 0.13 46.26 0.06 n.r. 0.00 n.r. 0.87 0.11 0.36 99.91 A-l <1 Dunite

40.67 0.01 0.75 1.15 6.56 0.12 48.77 0.00 0.00 0.03 0.02 1.38 n.r. 0.83 98.91 A-l <1 Dunite

41.46 0.02 0.98 1.43* 7.28* 0.10 46.90 0.13 0.07 <0.04 0.05 1.0* n.r. n.r. 99.46 A-l, A-3 16 Harzburgite

42.10 0.13 0.74 1.57* 7.99f 0.16 46.30 1.21 0.19 0.01 n.r. n.r. n.r. n.r. 100.24 A-2, A-4 20 Dunite

27.20 0.95 19.60 35.30 n.r. 0.22 10.40 0.60 0.05 0.04 0.03 5.30 n.r. 0.10 99.79 A-3 100 Chlorite schist

Sources: HC, Carpenter & Chen (1978); DB, Kulp & Brobst (1954); RM, Raymond et al. (2001); BC, Warner (2001); GH, Raymond (1995). All values in wt%. n.r., not recorded. * Calculated from FeOtotai. t Calculated from Fe2O3. f LOI includes minor CO2.

as blocks, locally in general association with one another in a matrix of pelitic, quartzo-feldspathic, and amphibolitic 'country' rocks. This physical association in field settings combined with the progressively lower-grade mineral assemblages in both the meta-ultramafic rocks and eclogites suggests that the meta-ultramafic rocks have been associated with the eclogites through all or most of this metamorphic history. Therefore, the placement of mineral association labels in Figure 4 is based on a combination of data from the eclogites, meta-ultramafic rocks, and the surrounding country rocks. Of particular importance in considerations of the progressive metamorphism of Blue Ridge meta-ultramafic rocks are the presence and movement of Ca, Al, Cr, and H^O. Clearly, lower-grade assemblages are all hydrated (although in the absence of I^O, olivine is metastable to surface conditions). As metamorphic conditions changed from peak conditions to lower P—T conditions, metamorphism progressed, at least in some rocks, through increasingly more hydrated states. An important question, currently unanswered, is: was most Ca and Al introduced with H2O or were Caand Al-bearing rocks more susceptible to hydration? The chemistry of many chlorite-amphibole rocks clearly reveals higher Ca and Al, which Scotford & Williams (1983) attributed to metasomatic metamorphism. Ca is generally known to be

mobile, whereas Al and Ti are not (Rollinson 1993, p. 73); yet the data suggest that Al and Ti can be mobilized under these metamorphic conditions (e.g. Kingsbury & Heimlich 1978). Some experimental work also suggests that Al is mobile in deep metasomatic environments (Manning 1999), just the setting in which metamorphism of these ultramafic rocks occurred. Under conditions of decreasing P and falling T, the first hydrous phases that appear in the rocks are tremolite and Cr-chlorites. The Cr-chlorites form rims on, or platy grains adjacent to, Crspinels (Fig. 3b). Mg, Cr, Al, and perhaps Fe derived from Cr-spinel contributed to these early chlorites (Table 2). Si and Fe were either introduced or were derived concomitantly from olivine. Movement of Fe from olivine into spinel, as Cr and Al moved into chlorite, would raise the Mg number of the olivines (Mg number = wt% Mg X 100/(wt% Mg + wt% Fe2+)), while lowering spinel Mg numbers and raising the spinel Cr numbers (Cr number = wt% Cr X 100/(wt% Cr + wt% Al)). Considered in bulk, these reactions may be isochemical except for the introduction of water. The appearance of tremolite, however, indicates that either Ca was available, perhaps in residual clinopyroxene present locally as grains and in exsolution lamellae, or it was introduced with the H2O. In the latter case, the reaction was allochemical. The presence of tremolite in some

262

L. A. RAYMOND ETAL.

metadunites and metaharzburgites suggests that by the time metamorphism produced Association A2, small amounts of Ca were present in these largely anhydrous rocks.

Hydrous v. anhydrous rocks A regional survey of the meta-ultramafic rock bodies of the North Carolina Blue Ridge Belt reveals that the population, locally, is slightly bimodal relative to degree of hydration (Fig. 5) (Swanson 2001; Raymond & Love 2002). Bodies dominated by anhydrous metadunite, plus less common but similarly anhydrous metaharzburgite, meta-orthopyroxenite, and rare metawebsterite and metaclinopyroxenite, contrast with completely or almost completely hydrated bodies of chlorite schist with one or more additional minerals such as talc, tremolite, anthophyllite, magnesiocummingtonite, magnesite, garnet, and magnetite (Fig. 3). The metadunites and related rocks contain >75% anhydrous phases (olivine, orthopyroxene, clinopyroxene), whereas the largely hydrated bodies and zones of hydrated rock contain >75% hydrous phases (chlorite, tremolite, talc, anthophyllite, serpentine). Several bodies consist of mixed anhydrous and hydrated rocks, with anhydrous phases in the range 75-25%. A geographi-

Fig. 5. Graph showing distribution of anhydrous (A), partially hydrated (M), and hydrated (H) ultramafic bodies in the Gossan Lead Block (Toe Terrane) and Fries Block (Mars Hill Terrane) of the North Carolina Blue Ridge Belt. It should be noted that the distributions of both hydrated and mixed bodies, each considered individually, are slightly bimodal. A, anhydrous bodies; M, mixed bodies with both anhydrous and hydrated minerals; H, largely hydrated bodies. (See Table 6 for data.)

cal contrast also exists, with hydrated bodies dominating north of Boone, and metadunites being far more abundant SW of Boone and the Grandfather Mountain Window (see Fig. 1). Locally, serpentinites form largely hydrated zones or veins in the dominantly anhydrous bodies and rarely form isolated serpentinite masses. Of particular interest to this study are some of the metamorphic, geochemical, and physical attributes of contrasting rock types. In terms of metamorphic grade, the anhydrous bodies record the oldest and highest grade mineral assemblages (Association A-l); yet they are also the best recorders of subsequent metamorphic events (Swanson 1980, 2001; Abbott & Raymond 1984; Raymond 1995, p. 667ff.). The hydrated bodies tend to record only Associations A-3 to A-5, with local remnants of Association A-2. Geochemically, there are significant differences between anhydrous and hydrous bodies, except for the serpentinites, which appear to have chemistries that differ from those of anhydrous bodies only in H2O content (Carpenter & Chen 1978; Raymond & Love 2002). Anhydrous bodies are lower in silica, lime, and alumina than hydrous bodies or the hydrous zones within partially hydrated bodies (Tables 4 and 5, Fig. 6; Kingsbury & Heimlich 1978). Elevated values of CaO occur in hydrated bodies rich in tremolite (Scotford & Williams 1983). Physically, some aspects of the rheological properties of anhydrous and hydrous ultramafic rocks and their minerals have been quantified (Birch 1966, section 7; Lockner 1995; Escartin et al. 2001). For example, chlorite schists and serpentinites have substantially lower moduli of rigidity than do dunites, and the former have higher compressibilities. Notably, the presence of serpentine greatly weakens dunitic rocks. As a proxy for strength, we examined the aspect ratios of meta-ultramafic rock bodies in the North Carolina Blue Ridge Belt (Table 6, Fig. 6). As noted in the Methods section, we measured the long axis and the maximum short axis of the bodies in relation to the foliation, dividing the former by the latter. Our field observations led us to assume that the weaker chlorite- and talc-rich bodies would be more prone to record higher degrees of visible strain reflected in high aspect ratios. High strain should also be revealed by fabric elements such as flattening, elongation, and shape preferred orientations (SPOs) of minerals. SPOs are typical of hydrated bodies. In contrast, the higher strength of the anhydrous, olivine-rich bodies makes them less prone to overall flattening and elongation, and the constituent minerals have lattice preferred orientations (LPOs) (see Bluhm 1976; Vrona 1979; but see Astwood et al 1972).

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

263

Table 5. Chemical data from anhydrous to hydrated Blue Ridge meta-ultramafic rocks Body name

Hydrated/ anhydrous

Wt% A1203

Wt% CaO

Reference

Addie (Webster) Bakersville Balsam Gap Blue Rock Road Blue Rock Mine Brushy Creek Buck Creek

M A A H H H A

1.23 n.d. 1.86 n.d. n.d. 2.74 2.09

0.12 n.d. 0.00 n.d. n.d. 0.08 0.79

Burton Lake Cane Creek Corundum Hill Crabtree Creek Cranberry Dark Ridge Day Book Deposit No. 9 Democrat Edmonds Elijay Creek Ennice Frank Grassy Creek North Grassy Creek South Greer Hollow Henson Creek Holcomb Branch Hoots Juno Laurel Creek Little Peak Creek Middleton Mine Creek Moores Knob Moores Knob Chi. Deposit Newdale Newdale Anthophyllite Newfound Gap Nigger Mountain Normanville Norton Phoenix Mountain Rattlesnake Mine Rocky Ridge Sapphire Asbestos Sapphire Mine Senia Shatley Springs Todd Twin Oaks Warrensville Webster Woodrow Bare Woody

M A A H H A A A A M M H A A M H A A A H M H A? M M H A M A H H H H A H M M A H H H H M H M

1.92 0.48 0.99 n.d. 4.84 0.38 2.18 1.36 n.d. 3.60 n.d. 3.61 0.80 n.d. n.d. n.d. 0.39 1.48 n.d. n.d. 2.10 8.20 n.d. n.d. 1.12 1.59 1.53 n.d. n.d. 5.84 n.d. 2.20 6.54 n.d. 7.10 n.d. n.d. 0.37 5.70 5.83 7.54 n.d. 10.29 6.12 n.d.

0.18 0.26 Trace n.d. 5.34 0.00 0.00 0.08 n.d. 3.68 n.d. 3.21 0.29 n.d. n.d. n.d. 0.19 0.00 n.d. n.d. 1.29 7.92 n.d. n.d. 0.05 0.22 0.24 n.d. n.d. 4.08 n.d. 1.32 5.83 n.d. 5.32 n.d. n.d. 0.19 3.21 5.53 11.05 n.d. 0.16 3.35 n.d.

Hunter 1941 Hunter 1941 Hunter 1941 Brobst 1962; Swanson 2001 Brobst 1962; Swanson 2001 Scotford & Williams 1983 Pratt & Lewis 1905; Hunter 1941; Hunter inMcElhaney & McSween 1983; Warner 2001 Hunter 1941 Hunter 1941 Pratt & Lewis 1905; Hunter 1941 Brobst 1962; Swanson 2001 Scotford & Williams 1983 Hunter 1941 Hunter 1941 Hunter 1941 Hunter 1941 Scotford & Williams 1983 Hunter 1941 Scotford & Williams 1983 Carpenter & Chen 1978 Brobst 1962; Swanson 2001 Brobst 1962; Swanson 2001 Raymond 1995 Carpenter & Chen 1978 Hunter 1941 Raymond et al. 2001 Hunter 1941 Hunter 1941; Hatcher et al. 1984 Scotford & Williams 1983 Hunter 1941 Brobst 1962; Swanson 2001 Hunter 1941 Hunter 1941 Hunter 1941 Brobst 1962; Swanson 2001 Hunter 1941 Scotford & Williams 1983 Brobst 1962; Swanson 2001 Hunter 1941 Scotford & Williams 1983 Pratt & Lewis 1905 Scotford & Williams 1983 Pratt & Lewis 1905 Pratt & Lewis 1905 Carpenter & Chen 1978 Scotford & Williams 1983 Scotford & Williams 1983 Scotford & Williams 1983 Scotford & Williams 1983 Hunter 1941 Scotford & Williams 1983 Brobst 1962; Swanson 2001

A, anhydrous; M, mixed; H, hydrated; n.d., no data available.

264

L. A. RAYMOND ETAL.

Fig. 6. Graphs showing the relationships between aspect ratio (long/short dimension) and long dimension (a), aspect ratio and A12C>3 content (b), and aspect ratio and CaO content (c) of Blue Ridge meta-ultramafic bodies. A, M, and H as in Figure 5. The Webster and Addie bodies typically plot outside of the normal fields and are labelled independently in the diagrams. (See Table 6 for data.)

Our data confirm our assumption that there should be a correlation between hydration and increase in aspect ratio, in that they show that hydrated bodies have a generally higher average and range of aspect ratios than do the anhydrous bodies (Fig. 6). Anhydrous bodies have low aspect ratios (<6), whereas the hydrous bodies have variable low to high aspect ratios (up to 47). This is the expected result, because the phyllosilicates that dominate the hydrous bodies are much weaker (more compressible) than the olivines and orthopyroxenes of the anhydrous bodies (Birch 1966, section 7; Lockner 1995). We note also that the overall pattern of aspect ratios relative to hydration conceals local bimodal distributions in aspect ratio, such as that north of Boone in Watauga and Ashe counties. Bodies in this area tend to fall into two populations, a high aspect ratio population

composed of hydrated bodies and a low aspect ratio one composed of anhydrous bodies. The aspect ratio data suggest that deformation results in significantly greater shortening and extension (higher overall strain) in hydrous bodies than in anhydrous bodies. We suggest also that high strain may have contributed to greater fragmentation of hydrous bodies and separation of hydrous parts of mafic-ultramafic masses from the anhydrous parts. We currently have no quantifiable measures of total strain in any of the ultramafic bodies.

Dunites and their spinel compositions The occurrence of metadunite as the dominant rock type in anhydrous North Carolina Blue Ridge meta-ultramafic rock bodies has long been puz-

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

265

Table 6. Aspect ratios of North Carolina Blue Ridge ultramafic bodies Body name Addie (Webster) Bakersville Balsam Gap Blue Rock Road Blue Rock Mine Brushy Creek Buck Creek Burton Lake Cane Creek Corundum Hill Crabtree Creek Cranberry Dark Ridge Day Book Deposit No. 9 Democrat Edmonds Elijay Creek Ennice Frank Grassy Creek North Grassy Creek South Greer Hollow Henson Creek Holcomb Branch Hoots Juno Laurel Creek Little Peak Creek Middleton Mine Creek Moore s Knob Moores Knob Chi. Deposit Newdale Newdale Anthophyllite Newfound Gap Nigger Mountain Normanville Norton Phoenix Mountain Rattlesnake Mine Rocky Ridge Sapphire Asbestos Sapphire Mine Senia Shatley Springs Todd Twin Oaks Warrensville Webster Woodrow Bare Woody

Aspect ratio

12.70 5.00 2.40 3.89 0.49 31.25 4.88 3.25 6.00 2.67 3.48 6.65 1.64 2.11 1.00 n.d. 20.31 2.50 6.45 3.95 3.81 4.99 2.22 3.68 2.31 3.84 13.33 2.08 6.45 2.22 5.27 3.00 n.d. 2.39 1.87 2.97 33.30 4.32 1.00 42.97 2.67 16.25 2.00 9.00 5.06 42.54 13.88 37.06 10.31 23.23 47.27 3.66

zling. Previous studies have suggested that these bodies may be either fragments of ophiolites or dehydrated serpentinite masses (Carpenter & Phyfer 1969; Swanson & Whittkopp 1976; Carpenter & Chen 1978; McElhaney & McSween 1983;

Reference Quinn 1991 Hunter 1941 Honeycutt et al. 1981 Brobst 1962 Brobst 1962 Scotford & Williams 1983 Bergeref a/. 2001 Hunter 1941 Hunter 1941 Hunter 1941 Brobst 1962 Scotford & Williams 1983 Schiering et al. 1982 Modified from Swanson 1981 Hunter 1941 Hunter 1941 Scotford & Williams 1983 Hunter 1941 Scotford & Williams 1983 Carpenter & Chen 1978 Brobst 1962 Brobst 1962 Raymond 1995 Carpenter & Chen 1978 Hunter 1941 Raymond et al. 2001 Hunter 1941 Hunter 1941 Scotford & Williams 1983 Hunter 1941 Brobst 1962 Hunter 1941 Hunter 1941 Vrona 1977 Brobst 1962 Hunter 1941 Scotford & Williams 1983 Brobst 1962 Hunter 1941 Scotford & Williams 1983 Pratt & Lewis 1905 Scotford & Williams 1983 Pratt & Lewis 1905 Pratt & Lewis 1905 Carpenter & Chen 1978 Scotford & Williams 1983 Scotford & Williams 1983 Scotford & Williams 1983 Scotford & Williams 1983 Quinn 1991 Scotford & Williams 1983 Brobst 1962

Lipin 1984; McSween & Hatcher 1985; Warner 2001). Both are possibilities, as ophiolites commonly contain serpentinized sections. Yet, the chemistries and textures of the spinels are largely incompatible with a deserpentinization hypothesis,

L. A. RAYMOND ETAL.

266

as are preliminary oxygen isotope analyses of olivines (Swanson et al. 1999; Swanson & Raymond, in prep). The 618O values of olivines range from +5.7 to +6.8, strongly suggesting that the olivine is not a product of deserpentinization (Swanson et al. 1999). If the Cr-spinels were to reveal remnants of an earlier phase of serpentinization, they should have ferritchromite or magnetite cores (with very low Mg numbers) characteristic of the spinels formed during serpentinization (Springer 1974). Instead, zoned Blue Ridge spinels have more iron-rich rims (Lipin 1984). In general, the Cr-spinels of the Blue Ridge metadunites show a wide range of compositions (e.g. Lipin 1984; Raymond et al. 2002; Tables 2, 7 and 8). Mg numbers range from <10 to 65 and Cr numbers range between 40 and 100. In the metaharzburgites, the Mg numbers are generally at the low end of this range. In the Hoots ultramafic body (Fig. 1), for example, Cr-rich spinels in metadunites and metaharzburgites have low Mg numbers of 20-26 (Table 8) (Raymond et al. 1999, 2001), values indicating that they are well out of high-J (low-P) equilibrium with associated olivines of Fo89-92- This implies significant metamorphic Mg loss (e.g. Ozawa 1983;

Allan & Dick 1996). Alumina contents of Cr-rich spinels are also very low, a condition consistent with metamorphic loss. Some Blue Ridge Cr spinels, however, show elevated TiO2 (in the 12% range), a condition that seems opposite to the loss of TiC>2 inferred to result from the oceanic metamorphic alteration of Cr-spinels (Allan & Dick 1996). Nearly all Cr-spinels from Blue Ridge metadunites show the effects of post-magmatic recrystallization. Lipin (1984) discussed the formation of chlorite at the expense of spinel in these rocks and illustrated the resulting zoning (loss of Al from spinel rims). A more extensive study of chrome spinels from hydrous and anhydrous meta-ultramafic rocks in the vicinity of the Spruce Pine area of the Blue Ridge (currently under way) shows extensive recrystallization of spinels related to the degree of hydration (Raymond et al. 2002). Thoroughly hydrated ultramafic bodies contain only magnetite or chromium magnetite. Anhydrous metadunites contain a range of spinel compositions from chromite with up to 15 wt% A12O3 to local ferritchromite and magnetite. Spinel grains in the metadunites are typically zoned and the highest alumina contents are in the cores.

Table 7. Chemistries of coexisting olivine and chrome-spinels Day Book

DB 153 Olivine 40.89 Si02 n.d. TiO2 A12O3 n.d. 0.01 Cr203 Fe2O3 n.d. FeO n.d. FeOt 8.01 NiO 0.26 0.09 MnO MgO 50.81 CaO 0.00 100.07 Total Cations based on */ oxygens (olivine); 0.994 Si Ti Al 0.000 Cr Fe3+ 2+ _ Fe 0.163 Fe2+t 0.005 Ni 0.002 Mn 1.841 Mg 0.000 Ca

DB 153 Cr-spinel

Henson Creek

HC 13 Olivine

HC 13 Cr-spinel

41.69 0.06 0.00 0.02 n.d. 0.02 n.d. 2.39 1.98 0.00 51.24 61.05 13.56 n.d. 5.95 n.d. 25.13 23.15 5.28 0.47 0.09 0.05 0.12 0.00 0.00 3.91 52.59 5.45 n.d. 0.00 n.d. 100.15 96.40 97.65 4 oxygens (chromite), cations normalized to 3 0.002 1.001 0.000 0.001 0.000 0.105 0.085 0.000 1.752 1.509 _ 0.380 0.163 _ 0.783 0.703 0.106 0.009 0.002 0.003 0.000 0.002 0.000 1.882 0.217 0.295 0.000 -

Newdale Anthophyllite

NA4 Olivine

NA4 Cr-spinel

40.50 n.d. n.d. 0.02 n.d. n.d. 8.84 0.39 0.09 50.76 0.00 100.60

0.01 0.36 1.92 44.76 21.60 27.09 0.21 0.00 3.12 n.d. 99.07

0.985 0.000 0.180 0.008 0.002 1.840 0.000

0.001 0.010 0.083 1.300 0.597 0.832 0.000 0.171 -

n.d., not determined. Fe2O3, FeO; calculated based on stoichiometry (eight positive charges, three cations). FeOt, total iron in olivine measured as FeO.

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES The loss of Al and the formation of ferritchromite and Cr-magnetite is related to the formation of chlorites during metasomatically enhanced, amphibolite-facies metamorphism. The transformation of more Cr-rich spinels to magnetite occurs during greenschist-facies serpentine formation. Most of the spinel compositions are related to metamorphic events and only the most Al-rich cores have any possibility of reflecting primary igneous compositions (Table 8). Considering Al mobility (demonstrated by Lipin 1984) and the values for Cr number and Mg number that seem to be related to metamorphic grade, it is unlikely that many cores represent igneous compositions. The behaviour of TiO2 in the spinels presents an interesting convergence between sublithosphere-lower-lithosphere processes and progressive retrograde amphibolite- to greenschist-facies metamorphism. Although in some cases metamorphic alteration of Cr-spinel results in TiO2 loss, the evidence from Blue Ridge rocks indicates the reverse (Raymond et al. 2002; work in progress). In general, Blue Ridge Cr-spinels, especially those in anhydrous rocks, have low values of TiO2 typical of ultramafic rocks associated with mid-ocean ridges (Table 8; see discussion below). The high TiO2 spinels that are present occur in hydrated rock bodies or in zones with significant amounts of hydration. Table 8 presents selected analyses of Cr-spinels from Hess Deep and Blue Ridge ultramafic rocks. Many of the Hess Deep dunite spinels (HDD samples) have TiO2 contents in excess of the 0.7 wt% that represents the typical upper limit of TiO2 contents in mantle dunites (Barnes & Roeder 2001; Kamenetsky et al 2001). In contrast, the Cr-spinels from harzburgites (HDH samples) have typical values. In the Blue Ridge Belt, it is Crspinels from metaharzburgites (e.g. RM 112-5c), rare spinel cores in metadunites, and spinel rims and grains in metasomatically altered ultramafic rocks that have higher TiO2 values (see Fig. 10, below; Raymond et al. 2002). A comparison of the characteristics of dunites of various origins and those of Blue Ridge metadunites and related rocks also yields some insight into potential origins of Blue Ridge rocks. In particular, studies of oceanic crust through the Ocean Drilling Program (ODP), studies of ophiolites, and rock reaction models provide an explanation for rocks that are in some ways analogous to those of the Blue Ridge (Dick 1977; Quick 1981; Lippard et al. 1986; Nicolas 1989, 1990; Kelemen 1990; Edwards 1995; Kelemen et al. 1995; Quick & Gregory 1995; Allan & Dick 1996; Cornen et al. 1996; Niida 1997; Varfalvy et al. 1997; Parkinson & Pearce 1998; Suhr et al. 1998; Constantin 1999; Braun & Kelemen 2001).

267

In ocean crust and ophiolites, many dunites appear to represent former melt channels and zones of high melt flux, where basaltic melt rose by porous flow through harzburgitic or Iherzolitic lithospheric or sublithospheric mantle. These observations and rock reaction models cited above, together suggest that as melt is produced from partial melting, it begins to rise and reaches a state of disequilibrium with the source rock minerals because of changes in asiO2 resulting, in part, from melt-wall-rock interaction. In addition, numerous studies have shown that the 0siO2 of basaltic melt increases with decreasing pressure (e.g. Kushiro 1964; O'Hara 1968; Basaltic Volcanism Study Project 1981; Kelemen et al. 1990). This change in activity results in profound meltwall-rock reaction during upward porous-media flow. Orthopyroxene incongruently melts during this reaction to form both olivine and an SiO2enriched, high-Mg melt that instantaneously reacts again with the wall rock as the melt rises and the temperature falls. Dunite melt channels and larger melt flux zones result. This theoretical model is supported by coring in uplifted and exposed portions of oceanic crust and mantle in both the Hess Deep and the MidAtlantic Ridge. The cores reveal melt veins and reaction zones within mantle rocks that show the effects of mafic melt migration (Fig. 7; Gillis et al 1993; Cannat et al 1995). At the Hess Deep, the cores show alternating patterns of gabbro, troctolite, dunite, and harzburgite, with gabbro representing frozen mafic melt (Fig. 8; also see Gillis et al. 1993; Dick & Natland 1996). The harzburgite represents depleted mantle peridotite wall rocks, whereas troctolite zones are produced by impregnation of olivine-rich rocks by plagioclase from the melt. The mafic magmas (represented by gabbros) that rose through the peridotite were olivine saturated but clinopyroxene undersaturated, resulting in dissolution of clinopyroxene with concomitant stabilization and precipitation of olivine (Kelemen et al. 1995; Allan & Dick 1996). Together with incongruent melting of orthopyroxene (described above), the processes convert peridotites to harzburgites and dunites along the zones through which magmas flow. Crystallization of olivine + spinel from the melts, combined with local magma mixing, may yield minor melts that produce orthopyroxenites that form dykes, which intrude the dunites and harzburgites (see Edwards 1995; Varfalvy et al. 1997; Raymond ef a/. 2001). Abundant and large (to 1 cm), often skeletal, chrome spinel is stable in the dunites. The abundance of the Cr-rich spinel in dunites and troctolites led Arai & Matsukage (1996) to propose that magma mixing, involving SiO2-rich

Table 8. Electron microprobe analyses of representative spinels from Hess Deep and Blue Ridge ultramafic rocks Sample Hess Deep HDD 108-1 HE HDD 108-11 ID HDD 108-1 11C HDD 101-105A HDD 101-105G HDH 10-12A HDH 123-126F HDH 10-12B HDH 7-1 OB HDH 34-37A Hoots RM 103A-1 RM 112-5R RM 112-5C RM 11-51 Frank F 84.5-1C Day Book DB 154B-6R DB 154C-4C DB 154A-8R DB 154D-5R DB 153-6R DB 154 A- 1C DB 154A-5C Mine Creek MC-6-3C Henson Creek HC 15-1C HC-2A-2C HC 15-2R HC-5-1C Newdale ND-20-5C ND-8-5C

MgO

CaO

Total

Fe no.

Cr no.

Mg no.

0.13 0.14 0.17 0.15 0.19 0.09 0.09 0.11 0.19 0.22

12.98 13.59 13.68 13.08 11.27 12.12 13.74 12.17 14.08 13.24

0.01 0.00 0.00 0.00 0.00 0.01 0.01 1.00 0.00 0.01

99.77 99.48 99.88 99.70 98.98 99.64 100.89 100.90 100.46 100.10

4.40 5.30 4.80 6.30 4.80 5.00 2.90 5.00 3.50 3.10

55.80 54.40 54.10 51.20 52.10 55.20 53.40 55.00 51.30 53.20

58.60 61.40 61.50 59.50 52.40 56.30 61.80 56.40 63.40 60.50

0.05 0.05 0.11 0.00

n.d. n.d. n.d. n.d.

4.73 5.94 5.78 5.47

0.00 0.02 0.00 0.02

95.31 97.48 97.08 97.95

n.d. n.d. n.d. n.d.

87.00 84.00 90.00 91.00

20.00 26.00 24.00 24.00

0.00

n.d.

n.d.

5.23

n.d.

82.06

n.d.

84.10

27.80

14.70 15.25 12.91 15.25 25.13 16.51 23.02

0.00 0.00 0.00 0.00 0.00 0.00 0.00

n.d. 0.05 0.08 0.00 0.09 n.d. 0.03

n.d. n.d. n.d. n.d. n.d. n.d. n.d.

11.74 11.79 11.60 11.79 4.22 11.00 5.80

n.d. n.d. n.d. n.d. n.d. n.d. n.d.

90.17 93.81 88.53 93.76 78.98 94.05 87.56

n.d. n.d. n.d. n.d. n.d. n.d. n.d.

76.40 77.70 84.91 77.00 91.20 79.10 n.d.

62.00 57.90 57.86 57.90 23.00 54.30 31.00

51.98

22.28

0.00

n.d.

n.d.

6.40

n.d.

86.07

n.d.

86.70

33.90

n.d. n.d. n.d. n.d.

55.06 56.53 52.58 61.19

19.74 22.13 21.57 20.60

0.00 0.00 0.00 0.00

0.03 0.05 0.10 0.00

n.d. n.d. n.d. n.d.

8.78 6.20 6.80 6.77

n.d. n.d. n.d. n.d.

93.63 88.23 85.10 90.87

n.d. n.d. n.d. n.d.

78.80 92.10 89.80 94.70

44.20 33.30 36.00 36.90

n.d. n.d.

58.02 56.85

21.19 22.24

0.00 0.00

n.d. n.d.

n.d. n.d.

6.67 6.36

n.d. n.d.

89.15 90.33

n.d. n.d.

92.30 88.70

36.00 33.80

FeO

MnO

MO

SiO2

TiO2

A1203

V203

Cr203

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02

1.10 0.94 0.90 0.47 0.44 0.07 0.06 0.05 0.05 0.03

22.65 23.37 23.81 25.18 24.61 23.32 25.47 23.51 26.61 25.31

0.26 0.24 0.25 0.21 0.21 0.25 0.23 0.25 0.22 0.24

42.69 41.59 41.83 39.43 39.96 42.82 43.54 42.82 41.74 42.82

19.64 19.26 18.88 20.74 21.87 20.60 17.46 20.60 17.27 17.90

0.18 0.17 0.17 0.33 0.31 0.27 0.21 0.31 0.20 0.22

0.13 0.18 0.19 0.11 0.12 0.09 0.08 0.08 0.10 0.09

0.02 0.00 0.00 0.00

0.19 1.11 1.95 1.17

5.15 6.91 4.08 3.93

n.d. n.d. n.d. n.d.

52.40 53.43 52.94 56.89

32.77 30.02 32.22 30.47

0.00 0.00 0.00 0.00

0.02

0.21

5.90

n.d.

46.47

24.23

0.00 0.00 0.00 0.00 0.03 0.00 0.02

0.23 0.19 0.20 0.19 0.11 0.11 0.07

6.86 10.72 6.79 10.72 2.99 9.98 5.00

n.d. n.d. n.d. n.d. n.d. n.d. n.d.

56.64 55.81 56.95 55.81 46.41 56.45 53.62

0.00

0.08

5.33

n.d.

0.00 0.01 0.00 0.02

0.09 0.05 0.04 0.01

9.93 3.26 4.01 2.28

0.00 0.00

0.02 0.01

3.25 4.87

ZnO

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES

269

Fig. 7. Photograph of core from the Hess Deep (ODP Leg 147) showing gabbroic melt (channels) (light) in ultramafic host (dark) (modified from Allan & Dick 1996).

Fig. 8. Schematic diagram showing relationship between peridotite mantle rock, dunite alteration (depletion) zones, and MORE gabbro in a sublithospheric melt channel (simplified from Dick & Natland 1996).

secondary melt generated during melt-wall-rock interaction, caused rapid spinel crystallization. This mechanism also may be one in which ophiolitic podiform chromitites are generated (Arai & Matsukage 1996; Arai 1997a, 1997b; compare Zhou et al. 1996). A striking feature of the dunite spinels produced by melt-wall-rock interaction is enrichment in TiOa. Values of TiOi in Hess Deep spinels are locally 10 times higher in dunite spinels (e.g. HDD 108-1 HE; HDD 108-111C) than in the surrounding harzburgite spinels (e.g. HDH 1012B; HDH 123-126F) (Table 8). Gabbroic dykelets in harzburgite from Leg 153 were sheathed successively by pyroxenite and dunite-wehrlite, with increasing enrichment of TiC>2 in Cr-rich spinel with decreasing distance to the dykelet (Cannat et al 1997; Casey 1997). Hence, it appears that the increase in TiO2 in the dunite spinels results from a transfer of TiO2 from the

270

L. A. RAYMOND ETAL.

impregnating mafic melt to the dunite chromium spinel (Allan & Dick 1996; Cannat et al. 1997; Niida 1997). The compositions of Cr-rich spinels from the Hess Deep and the Blue Ridge have both similarities and differences. Compositions of Hess Deep and Blue Ridge spinels are plotted on the Cr number-Mg number diagram in Figure 9. Hess Deep spinels plot near the middle of the ophiolite field: a field significantly overlapped by the field for mid-ocean ridge basalt (MORB) and oceanfloor peridotite spinels (Barnes & Roeder 2001). Blue Ridge Cr spinels plot in an array that begins in the ophiolite field and trends towards the upper right corner. The spinel cores and grains with the lowest Cr numbers in Blue Ridge rocks may be relict igneous spinels, inasmuch as they plot in the ophiolite field. In contrast, chrome spinels with low Mg number and high Cr number are typical of alpine ultramafic rocks. Although some alpine and ocean-floor peridotite spinels have similar Mg and Cr numbers

Fig. 9. Cr number-Mg number diagram showing typical Cr-spinel compositions from the Hess Deep and the Blue Ridge Belt. Fields in the diagram are based on those of Dick & Bullen (1984) and Leblanc (1987).

(Barnes & Roeder 2001), alumina and titania contents may differ. Figure 10 is a TiO2 v. A12O3 plot of spinel compositions from the Hess Deep and Blue Ridge Belt. The fields shown are those of Kamenetsky et al. (2001). MORB dunite and peridotite spinels contain up to 60 wt% A12O3 and generally <0.5 wt% TiO2, whereas SSZ dunite and peridotite spinels, although containing similar amounts of TiO2, contain only up to 40wt% A12O3. The TiO2 contents of chromium spinels from the dunites and harzburgites of the Hess Deep form an elongate array that plots within the MORB field and is paralleled by the field defined by the compositions of spinels from the Blue Ridge meta-ultramafic rocks. The Hess Deep spinels from ultramafic rocks, with between 18 and 30 wt% alumina, clearly fall within the MORB range of A12O3 contents, but some spinels from dunites have higher than normal TiO2 contents. In contrast, although locally containing

Fig. 10. TiO2-Al2O3 diagram showing the compositions of selected Hess Deep (open symbols) and Blue Ridge (filled symbols) Cr-spinels. Circles represent spinels from dunites and metadunites, whereas triangles represent spinels from harzburgites and metaharzburgites. Fields from Kamenetsky et al. (2001). LIP, large igneous provinces; OIB, ocean island basalts; MORB, mid-ocean ridge basalts; MORB peridotite, subbasaltic, ocean crust peridotites; ARC, volcanic arc rocks.

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES elevated TiO2 values, Blue Ridge spinels from metadunites and metaharzburgites typically have alumina contents in the 0-15wt% range, more typical of arc and SSZ rocks. Hess Deep spinels range from homogeneous to zoned grains, as do those from Blue Ridge rocks (Fig. 11). In some zoned grains, only rims are elevated (enriched) in TiO2. In other grains, TiO2 is elevated throughout the grain. Figure 10 reveals that both groups of spinels show local enrichment in TiO2. Initially, the similarities in enrichment suggested that similar processes operated in the Hess Deep mantle and the source region of the Blue Ridge rocks (Raymond & Allan 2001), but additional data reveal instead that this is a case of convergent chemical evolution (Raymond et al. 2002). As noted, the elevated TiO2 values in Hess Deep dunite spinels were interpreted by Allan & Dick (1996) to be the result of enrichment in titania facilitated by repeated passage of MORB melts through the dunite channels. This enrichment occurs in dunite spinels. In Blue Ridge rocks, however, the elevated TiO2 values occur in both dunite and harzburgite spinels and are related to metamorphic processes (Raymond et al. 2002). Alumina contents of Blue Ridge dunite spinels fall within the SSZ and arc fields of Figure 10 and a few fall outside any of the fields. This may suggest that the Blue Ridge ultramafic rocks represent an SSZ or arc setting. However, there is no evidence of a Taconic arc in the Blue Ridge Belt. There are, however, metamafic rocks with MORB-like chemistries associated with the metaultramafic rocks (Misra & Conte 1991) and these could represent SSZ rocks. An SSZ petrotectonic setting is supported by depleted values of Ti and other elements in these metamafic rocks (Gair & Slack 1984; Misra & Conte 1991). The fact that

Fig. 11. Backscatter image of chromium spinel from Frank ultramafic body showing zoning. Core has c. 6.0% A12O3 and 0.1% TiO2, whereas rim has c. 1.0% A12O3 and 0.5% TiO2

271

some alumina values for Blue Ridge spinels fall outside the SSZ field (or any field) is probably a testament to the mobility of aluminium during metamorphism and renders the SSZ assignment moot. Thus, the petrotectonic setting is not unambiguously resolved by these data.

Tectonic implications The petrotectonic implications of both the chromium spinel compositions and the petrology of the Blue Ridge meta-ultramafic rocks are twofold. First, spinel compositions suggest a possible petrogenetic setting for formation of Blue Ridge ultramafic bodies, an SSZ setting. Four possible petrogenetic models for formation of the Ashe Metamorphic Suite containing meta-ultramafic rocks were compared by Abbott & Raymond (1984). These included formation on a passive ensialic margin (Rankin 1970; Rankin et al. 1973), formation on a passive to transitional margin with a Laurentian sediment source (see Wehr & Glover 1985), formation on an active ensimatic margin with an eastern arc sediment source (Raymond & Swanson 1981), and formation in a back-arc basin (see Stevens et al. 1974). Although the spinel data are consistent with an arc or SSZ, back-arc basin model, in such models the emplacement of ultramafic rocks and the formation and emplacement of associated eclogites are difficult to explain. No NeoproterozoicEarly Palaeozoic, pre-Taconic arc existed within or to the west of the Blue Ridge terranes containing the ultramafic rocks. Thus arc and eastdirected SSZ models fail to explain the general geology of the region. Arcs proposed to the east would require west-directed subduction to produce an SSZ setting in the Blue Ridge Belt, in contrast to the east-directed subduction suggested in most models of Southern Appalachian Palaeozoic tectonics (e.g. Odom & Fullagar 1973; Hatcher 1978; Abbott & Raymond 1984; Hatcher 1987). Westdirected, Taconic subduction in the vicinity of the Blue Ridge Belt required to produce an SSZ setting for the meta-ultramafic rock protoliths would require a major revision of tectonic models. A second implication of the data, specifically the mineralogical, petrographic, and chemical data from the meta-ultramafic rocks and the associated mafic rocks (which have MORB affinities; Misra & Conte 1991; Tenthorey et al 1996), is that these rocks do not represent a classical ophiolite formed in a typical moderate- to fast-spreading, spreading centre setting. If the rocks represent ophiolites, as has been proposed (e.g. McSween & Hatcher 1985; Tenthorey et al. 1996), they appear to represent only the upper and lower parts of a typical ophiolite, because little metagabbro is

272

L. A. RAYMOND ETAL.

present in the Blue Ridge terranes containing the meta-ultramafic rocks. Only one plutonic complex is known, a troctolite-rich complex (Tenthorey et al. 1996). Although some metagabbros are present locally throughout the terrane, they are not abundant (Hatcher 1980; Hatcher et al 1984; Vance & Raymond 1994). This aspect of Blue Ridge rocks has been overlooked, perhaps because the existing rocks have been highly deformed and fragmented during deformation of the terrane. Deformation converted the meta-ultramafic rocks into the lenses and pods described above and locally intermixed them with metamafic rocks. A fundamental question remaining is: what protoliths are represented by the abundant metadunites and metaharzburgites and their phyllosilicate retrograde descendants? One possibility is that the Blue Ridge mafic-ultramafic assemblage represents one of the major types of mafic-ultramafic complex such as a layered lopolithic complex, Alaska-type body, alkaline complex, or an appinite-type complex. Neither the chemistry nor the minerals and rock associations of the Blue Ridge rocks are consistent with those aspects of non-ophiolitic, mafic-ultramafic complexes (see Jackson & Thayer 1972; Raymond 2002, table 9.1). The characteristics of Blue Ridge rocks are simply wrong for most of the alternative types of complexes, and the requisite and abundant gabbroic constituents of lopolithic bodies do not exist in the Blue Ridge Belt. Alternatively, the metadunites and metaharzburgites may represent: (1) basal ophiolite cumulates, (2) mantle dunite lenses, or (3) depleted mantle tectonite in sublithospheric(?) mid-ocean ridge melt flux zones of traditional ophiolites; or (4) the upper parts of a melt channel and melt flux zone-bearing ophiolite formed in a slow-spreading environment. In all cases, the abundant hornblende schists and gneisses would primarily represent basaltic oceanic crust. The relatively rare troctolites (Tenthorey et al. 1996; Berger et al. 2001) must represent either melt channel margins or altered parts of the plutonic layered complex of an ophiolite. Some apparent metagabbros exist within the amphibolite-rich areas, but a major plutonic complex required as a part of a traditional ophiolite is missing, was never formed, or has not been recognized. The Xigaze ophiolite of Tibet (Nicolas et al. 1981; Girardeau et al. 1985; Girardeau & Mercier 1988) may be an analogue for the Blue Ridge ophiolitic rocks. The Xigaze ophiolite consists, from top to bottom, of mafic rocks in the form of pillow lavas, dykes, and sills; an extremely sparse plutonic section consisting of sparse gabbros intruded by mafic dykes; serpentinized dunites and harzburgites; and a basal section consisting of Cr-

diopside bearing harzburgites overlying a serpentine matrix tectonic melange. Similar ophiolites occur in the Alps and Apennine regions (Lagabrielle & Lemoine 1997). The Xigaze ophiolite formed in a slow-spreading ridge where heat production was low and extensive seawater circulation altered the rocks (Girardeau et al. 1985; see Cannat 1993). The alteration, abundance of MORB-like mafic rocks with a dearth of metagabbro, and dunite-harzburgite base with the dunite formed as melt channel-melt flux zones, match the characteristics of the mafic-ultramafic parts of the Ashe and Pumpkin Patch metamorphic suites of the Eastern and Central Blue Ridge belts (Abbott & Raymond 1984; Gair & Slack 1984; Misra & Conte 1991; Berger et al. 2001; Raymond et al. 2001; Swanson 2001). High aspect ratios of the Blue Ridge rock bodies simply reflect later metamorphism and structural dismemberment with high strain that would have acted on the ophiolitic rocks during the Taconic and later orogenies. The Blue Ridge meta-ultramafic rocks and associated eclogites appear to represent a Taconic suture (Adams et al. 1995), but the kinematic history of such a suture zone remains to be resolved. Any model must explain coeval deep crustal (c. 30 km) metamorphism of MORB and associated meta-ultramafic rocks and mantle-depth (>50 km) metamorphism of similar rocks to produce eclogite and metadunite. These rocks must then be mixed in a suture zone. Displacement of some of the rocks for an undetermined distance along strike by an Acadian strike-slip fault, during a period of re-metamorphism may have been important in the final assembly of the orogen (Adams et al. 1995; Stewart & Miller 2001). If the meta-ultramafic rocks represent SSZ sublithospheric mantle, however, the geographical relationship between the only proposed arc rocks (to the east) and the Blue Ridge eclogites and meta-ultramafic rocks must be explained. Such an explanation might involve formation of an arc to the east as a result of west-directed (rather than east-directed) subduction. The small ocean basin hypothesized by many workers (e.g. Odom & Fullagar 1973; Hatcher 1978, 1987; Abbott & Raymond 1984) between a microcontinent or arc to the east and Laurentia to the west would represent a slow-spreading, back-arc rift basin that perhaps separated a fragment of Laurentia from the continental mass (see Thomas 1977).

Discussion and conclusions The Blue Ridge meta-ultramafic rocks are polymetamorphosed metadunites, metaharzburgites, and minor metawebsterites and meta-orthopyroxenites. No structural or textural features indicative

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES of an igneous protolith have been recognized in any of the more than 200 scattered lenticular, sheet-like, and podiform bodies in the Blue Ridge Belt. laconic metamorphism recrystallized these rocks and produced minor Ca-Mg amphibole and Cr-chlorite to accompany the existing olivine, orthopyroxene, and Cr-spinel. Acadian metamorphism was hydrous and metasomatic, producing rocks such as ferritchromite-bearing, tremolite-magnesiocummingtonite-talc and chlorite schists from their Taconic predecessors. Serpentine with accessory magnetite and late veins containing a diverse assemblage of minor minerals developed as further retrograde metamorphism again recrystallized the rocks during the Alleghenian Orogeny. Combining the data and interpretations here, we hypothesize that Blue Ridge metadunites represent sublithospheric mantle melt channel or melt flux zones, like those in the Bay of Islands, Oman, and the Trinity ophiolites (Nicolas 1989; Edwards 1995; Quick & Gregory 1995), in Xigaze-type ophiolitic complexes formed at a slow-spreading ridge, perhaps in an SSZ setting. The ophiolites were dismembered before or during emplacement into enclosing mafic, pelitic, and quartzo-feldspathic Blue Ridge rocks and, with eclogites, were assembled as an accretionary complex during the Taconic Orogeny. The mineralogy of the oldest mineral assemblages in metadunite and associated metaharzburgite is consistent with this interpretation. Alumina contents, Mg numbers, and Cr numbers for spinel cores in this oldest assemblage may also reflect an early igneous history, but the chemistry of the spinels and the bulk-rock compositions were probably altered during the multiple periods of metamorphism that changed these rocks during the Palaeozoic era. Elevated TiO2 values in Blue Ridge spinels resulted not from MORB meltinduced enrichment as magmas coursed through the melt channels, as was the case in the Hess Deep mantle, but in response to fluid-enhanced retrograde metamorphism following anhydrous peak metamorphic conditions. The major and trace element chemistries of mafic rocks associated with the meta-ultramafic rocks and regionally extensive alteration are consistent with these interpretations (Gair & Slack 1984; Misra & Conte 1991; Tenthorey et al. 1996). As demonstrated, the Blue Ridge rocks are highly metamorphosed. Rock distribution patterns within Blue Ridge terranes reveal block-in-matrix structures at various scales reminiscent of the structure of melanges (Raymond et al. 1989; Adams et al. 1995). These block-in-matrix patterns are overprinted by metamorphic minerals, textures, and structures, complicating their interpretation. In addition, thrust faults and an Acadian

273

dextral strike-slip fault cut the terrane (Rankin et al. 1973; Abbott & Raymond 1984; Adams et al. 1995; Stewart et al. 1997; Waters et al. 2000). Meta-ultramafic rocks form blocks, as do eclogites and hornblende schists and gneisses, within the overall terrane, leading to the interpretation that the ultramafic belt marks an Ordovician (Taconic) suture zone containing ophiolite fragments, perhaps from an SSZ setting. The distribution of anhydrous metadunite and metaharzburgite blocks and hydrous rocks, principally chlorite schists, is a result of both metamorphism and structural separation and high strain deformation of the weaker chlorite schists during melange formation and faulting in the suture zone. Specifically, Acadian deformation created greater stretching (strain) in the weaker hydrous rocks, leading to their greater aspect ratios. Lower-grade metamorphism with accompanying hydration in northern North Carolina and southern Virginia created a greater percentage of hydrated metaultramafic bodies in this region (Scotford & Williams 1983). The Alleghenian collision of Africa and North America led only to retrograde metamorphism that produced local serpentinization and alteration of the meta-ultramafic rocks. A new tectonic model for the southern Appalachian Orogen for the Cambrian to Mississippian periods is required to explain the geographical relationships implied by an SSZ setting for ophiolitic metadunites formed at a slow-spreading ridge, coeval metamorphism of ocean crust to both amphibolite-facies and eclogite-facies metamorphic grades, and emplacement and mixing of ultramafic rocks, mafic rocks (including eclogites, and pelitic and quartzo-feldspathic rocks) along a Taconic suture zone. An incomplete model allowing an SSZ setting for Blue Ridge rocks has been presented by Hibbard (2000) and an alternative model involving large-scale horizontal displacement has been offered by Stewart & Miller (2001). If the setting was SSZ, the Taconic subduction would have been west-directed, as suggested by Hibbard (2000), rather than east-directed, as previously proposed models have indicated. We thank Appalachian State University, the University of Georgia, and the National Science Foundation (Award Number PRM-8112182 to L.A.R.) for support. We thank F. Webb, M. McKinney, R. McCarter, and N. Johnson of Appalachian State University and C. Fleisher of the University of Georgia for technical assistance. We also thank C. Miller and two anonymous reviewers for helpful comments on an earlier version of the manuscript.

References ABBOTT, R.N. JR & GREENWOOD, J.P. 2001. Retrograde metamorphism of eclogite in the southern Appala-

274

L. A. RAYMOND ETAL.

chian Mountains, U.S.A.—a case involving seamount subduction? Journal of Metamorphic Geology, 19, 433-443. ABBOTT, R.N. JR & RAYMOND, L.A. 1984. The Ashe Metamorphic Suite, northwest North Carolina: metamorphism and observations on geologic history. American Journal of Science, 284, 350-375. ABBOTT, R.N. JR & RAYMOND, L.A. 1997. Petrology of pelitic and mafic rocks in the Ashe and Alligator Back Metamorphic Suites, northwest of the Grandfather Mountain Window. In: STEWART, K.G., ADAMS, M.G. & TRUPE, C.H. (eds) Paleozoic Structure, Metamorphism, and Tectonics of the Blue Ridge of Western North Carolina. Carolina Geological Society, 1997 Field Trip and Annual Meeting, 87-101. ABSHER, B.S. & McSwEEN, H.Y. JR 1985. Granulites at Winding Stair Gap, North Carolina: the thermal axis of Paleozoic metamorphism in the southern Appalachians. Geological Society of America Bulletin, 96, 588-599. ADAMS, M.G. & TRUPE, C.H. 1997. Conditions and timing of metamorphism in the Blue Ridge Thrust Complex, northwestern North Carolina and eastern Tennessee. In: STEWART, K.G., ADAMS, M.G. & TRUPE, C.H. (eds) Paleozoic Structure, Metamorphism, and Tectonics of the Blue Ridge of Western North Carolina. Carolina Geological Society, 1997 Field Trip Guidebook, 33-47. ADAMS, M.G., STEWART, K.G., TRUPE, C.H. & WILLARD, R.A. 1995. Tectonic significance of highpressure metamorphic rocks and dextral strike-slip faulting in the southern Appalachians. In: HIBBARD, J., VAN STAAL, C.R., CAWOOD, P. & COLMANSADD, S. (eds) Current Perspectives in the Appalachian-Caledonian Orogen. Geological Association of Canada, Special Papers, 41, 21-42. ALLAN, J.F. & DICK, H.J.B. 1996. Cr-rich spinel as a tracer for melt migration and melt-wall rock interaction in the mantle: Hess Deep, leg 147. In: GILLIS, M.C., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 157-172. ARAI, S. 1997a. Control of wall-rock composition on the formation of podiform chromitites as a result of magma/peridotite interaction. Resource Geology, 47, 77-187. ARAI, S. 1997b. Origin of podiform chromitites. Journal of Asian Earth Sciences, 15, 303-310. ARAI, S. & MATSUKAGE, K. 1996. Petrology of the gabbro—troctolite-peridotite complex from Hess Deep, equatorial Pacific: implications for mantlemelt interaction within the oceanic lithosphere. In: MEVEL, C., GILLIS, K.M., ALLAN, J.F. & MEYER, P.S. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 135-155. ASTWOOD, P.M., CARPENTER, J.R. & SHARP, W.E. 1972. A petrofabric study of the Dark Ridge and Balsam Gap dunites, Jackson County, North Carolina. Southeastern Geology, 14, 183-194. BARNES, S.J. & ROEDER, P.L. 2001. The range of spinel compositions in terrestrial mafic and ultramafic

rocks. Journal of Petrology, 42, 2279-2302. Basaltic Volcanism Study Project 1981. Basaltic Volcanism on the Terrestrial Planets. Pergamon, New York. BERGER, S., COCHRANE, D., SIMONS, K., SAVOV, L, RYAN, J.G. & PETERSON, V.L. 2001. Insights from rare earth elements into the genesis of the Buck Creek Complex, Clay County, NC. Southeastern Geology, 40, 201-212. BIRCH, F. 1966. Compressibility; elastic constants. /W.CLARK, S.P. JR (ed.) Handbook of Physical Constants. Geological Society of America, Memoirs, 97, 97-173. BLUHM, C.T. 1976. The structure of the Frank, Alpinetype dunite-lherzolite complex, Avery County, North Carolina. MS thesis, Southern Illinois University, Carbondale. BRAUN, M.G. & KELEMEN, P.B. 2001. Mapping dunite melt conduits in the mantle: constraints on dunite formation and MORE flux. EOS Transactions, American Geophysical Union, 82(20), S430. BROBST, D.A. 1962. Geology of the Spruce Pine District, Avery, Mitchell, and Yancey Counties, North Carolina. US Geological Survey Bulletin, 1122, 26. BROWN, P.M. ET AL. 1985. Geologic Map of North Carolina. North Carolina Geologic Survey, Raleigh. BUTLER, J.R. 1973. Paleozoic deformation and metamorphism in part of the Blue Ridge thrust sheet, North Carolina. American Journal of Science, 273, 72-88. CANNAT, M. 1993. Emplacement of mantle rocks in the seafloor at mid-ocean ridges. Journal of Geophysical Research, 98, 4163-4172. CANNAT, M., KARSON, J.A., MILLER, D.J., et al. (eds) 1995. Proceedings of the Ocean Drilling Program, Initial Reports, Sites 190-194. Ocean Drilling Program, College Station, TX. CANNAT, M., CHATIN, F., WHITECHURCH, H. & CEULENEER, G. 1997. Gabbroic rocks trapped in the upper mantle at the Mid-Atlantic Ridge. In: KARSON, J.A., CANNAT, M., MILLER, D.J. & ELTHON, D. (eds) Proceeding of the Ocean Drilling Program, Scientific Results, 153. Ocean Drilling Program, College Station, TX, 243-264. CARPENTER, J.R. & CHEN, H.S. 1978. Petrology and bulk rock geochemistry of the Frank ultramafic body, Avery county, N.C. and associated other ultramafic rock bodies of the southern Appalachians. Southeastern Geology, 20, 21-25. CARPENTER, J.R. & PHYFER, D.W. 1969. Proposed origin of the 'alpine-type' ultramafics of the Appalachians. Geological Society of America, Abstracts with Programs, 7, 261-263. CASEY, J.F. 1997. Comparison of major- and traceelement geochemistry of abyssal peridotites and mafic plutonic rocks with basalts from the MARK region of the Mid-Atlantic Ridge. In: KARSON, J.A., CANNAT, M., MILLER, D.J. & ELTHON, D. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 153. Ocean Drilling Program, College Station, TX, 181-241. CONSTANTIN, M. 1999. Gabbroic intrusions and magmatic metasomatism in harzburgites from the Garrett transform fault: implications for the nature of

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES the mantle—crust transition at fast-spreading ridges. Contributions to Mineralogy and Petrology, 136, 111-130. CORNEN, G., BESLIER, M. & GIRARDEAU, J. 1996. Petrologic characteristics of the ultramafic rocks from the ocean/continent transition in the Iberia Abyssal Plain. In: WHITMARSH, R.B., SAWYER, D.S., KLAUS, A. & MASSON, D.G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 377-395. DAVIES, G.R., NIXON, P.H., PEARSON, D.G. & OBATA, M. 1993. Tectonic implications of graphitized diamonds from the Ronda peridotite massif, southern Spain. Geology, 21, 471-474. DICK, HJ.B. 1977. Evidence of partial melting in the Josephine Peridotite. In: DICK, HJ.B. (ed.) Magma Genesis, Proceedings of the American Geophysical Union Chapman Conference on Partial Melting in the Earth's Upper Mantle. Oregon Department of Geology and Mineral Industries Bulletin, 96, 59-62. DICK, HJ.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DICK, HJ.B. & NATLAND, J.H. 1996. Late-stage melt evolution and transport in the shallow mantle beneath the East Pacific Rise. In: MEVEL, C, GILLIS, K.M., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 103-134. DRIBUS, J.R., GODSON, W.L., HAHN, K.R., KAMMER, T.W., LAWTON, D.L., PESCH, H.G., SCHICK, J.T. & TATA, S.S. 1976. Comparative petrography of some alpine ultramafic plutons in western North Carolina. Compass, 53, 33-45. EDWARDS, SJ. 1995. Boninitic and tholeiitic dykes in the Lewis Hills mantle section of the Bay of Islands ophiolite: implications for magmatism adjacent to a fracture zone in a back-arc spreading environment. Canadian Journal of Earth Sciences, 32, 2128-2146. ESCARTIN, J., HIRTH, G. & EVANS, B. 2001. Strength of slightly serpentinized peridotites: implications for the tectonics of oceanic lithosphere. Geology, 29, 1023-1026. GAIR, I.E. & SLACK, J.F. 1984. Deformation, geochemistry, and origin of massive sulfide deposits, Gossan Lead District, Virginia. Economic Geology, 79, 1483-1520. GILLIS, K., MEVEL, C., ALLAN, J., et al. (eds) 1993. Proceedings of the Ocean Drilling Program, Initial Reports, 147, Sites 894-895. Ocean Drilling Program,College Station, TX. GIRARDEAU, J. & MERCIER, J.C.C. 1988. Petrology and texture of the ultramafic rocks of the Xigaze ophiolite (Tibet): constraints for mantle structure beneath slow-spreading ridges. Tectonophysics, 147, 33-58. GIRARDEAU, J., MERCIER, J.C.C. & XIBIN, W. 1985. Petrology of the mafic rocks of the Xigaze ophiolite, Tibet: implications for the genesis of the

275

oceanic lithosphere. Contributions to Mineralogy and Petrology, 90, 309-321. GOLDBERG, S.A. & DALLMEYER, R.D. 1997. Chronology of Paleozoic metamorphism and deformation in the Blue Ridge Thrust Complex, North Carolina and Tennessee. American Journal of Science, 297, 488-526. GREEN, H.W. II, DOBRZHINETSKAYA, L., RIGGS, E.M. & ZIN, Z.-M. 1997. Alpe Arami: a peridotite massif from the mantle transition zone. Tectonophysics, 279, 1-21. HATCHER, R.D. JR 1978. Tectonics of the western Piedmont and Blue Ridge, Southern Appalachians: review and speculation. American Journal of Science, 278, 276-304. HATCHER, R.D. JR 1980. Geologic Map and Mineral Resources Summary of the Prentiss Quadrangle, North Carolina. North Carolina Department of Natural Resources and Community Development, Geological Survey Section Map, GM 167-SW, MRS 167-SW. HATCHER, R.D. JR 1987. Tectonics of the southern and central Appalachian internides. Annual Review of Earth and Planetary Sciences, 15, 337-362. HATCHER, R.D. JR, HOOPER, R.J., PETTY, S.M. & WILLIS, J.D. 1984. Structure and chemical petrology of three southern Appalachian mafic-ultramafic complexes and their bearing upon the tectonics of emplacement and origin of Appalachian ultramafic bodies. American Journal of Science, 284, 484-506. HESS, H.H. 1955. Serpentines, orogeny, and epeirogeny. In: POLDERVAART, A. (ed.) Crust of the Earth. Geological Society of America, Special Papers, 62, 391-408. HIBBARD, J. 2000. Docking Carolina; mid-Paleozoic accretion in the Southern Appalachians. Geology, 28, 127-130. HONEYCUTT, P.M. & HEIMLICH, R.A. 1980. Petrology of the Balsam Gap Dunite, Jackson County, North Carolina. Southeastern Geology, 21, 251-260. HONEYCUTT, P.M., HEIMLICH, R.A. & PALMER, D.F. 1981. Magnetic survey of the Balsam Gap Dunite, Jackson County, North Carolina. Southeastern Geology, 22,223-231. HOPSON, J.L., HATCHER, R.D. JR & STIEVE, A.L. 1989. Geology of the eastern Blue Ridge, northeastern Georgia and the adjacent Carolinas. In: FRITZ, W.J., HATCHER, R.D. JR & HOPSON, J.L. (eds) Geology of the Eastern Blue Ridge, Northeastern Georgia and the Adjacent Carolinas. Georgia Geological Society Guidebooks, 9(3), 1-40. HUNTER, C.E. 1941. Forsterite olivine deposits of North Carolina and Georgia. North Carolina Division of Mineral Resources Bulletin, 41, 117. JACKSON, E.D. & THAYER, T.P. 1972. Some criteria for distinguishing between stratiform, concentric and alpine peridotite-gabbro complexes. 24th International Geological Congress, Section 2, 289-296. JUTEAU, T., NICOLAS, A., DUBESSY, J., FRUCHARD, J.C. & BOUCHEZ, J.L. 1977. Structural relationships in the Antalya Ophiolite Complex, Turkey: possible model for an oceanic ridge. Geological Society of America Bulletin, 88, 1740-1748.

276

L. A. RAYMOND ETAL.

KAMENETSKY, V.S., CRAWFORD, AJ. & MEFFRE, S. 2001. Factors controlling chemistry of magmatic spinel: an empirical study of associated olivine, Crspinel and melt inclusions from primitive rocks. Journal of Petrology., 42, 655-671. KELEMEN, P.B. 1990. Reaction between ultramafic rocks and fractionating basaltic magma I. Phase relations, the origin of calc-alkaline magma series, and the formation of discordant dunite. Journal of Petrology, 31, 51-98. KELEMEN, P.B., JOYCE, D.B., WEBSTER, J.D. & HOLLOWAY, J.R. 1990. Reaction between ultramafic rock and fractionating basaltic magma; II, Experimental investigation of reaction between olivine tholeiite and harzburgite at 1150-1050°C and 5 kb. Journal of Petrology, 31, 99-134. KELEMEN, P.B., SHIMIZU, N. & SALTERS, VJ.M. 1995. Extraction of mid-ocean-ridge basalt from the upwelling mantle by focused flow of melt in dunite channels. Nature, 375, 747-753. KINGSBURY, R.H. & HEIMLICH, R.A. 1978. Petrology of an ultramafic body near Micaville, Yancey County, North Carolina. Southeastern Geology, 20, 33-46. KULP, J.L. & BROBST, D.A. 1954. Notes on the dunite and the geochemistry of vermiculite at the Day Book Dunite deposit, Yancey County, North Carolina. Economic Geology, 49, 211-220. KUSHIRO, I. 1964. The system diopside-forsteriteenstatite at 20kilobars. Carnegie Institution of Washington Yearbook, 63, 101-108. LAGABRIELLE, Y. & LEMOINE, M. 1997. Alpine, Corsican, and Apennine ophiolites: the slow-spreading ridge model. Comptes Rendus de VAcademic des Sciences, Sciences de la Terre et des Planetes, 325, 909-920. LARRABEE, D.M. 1966. Map showing distribution of ultramafic and intrusive mafic rocks from northern New Jersey to eastern Alabama. US Geological Survey Map, 1-476. LEBLANC, M. 1987. Chromite in oceanic arc environments: New Caledonia. In: STOWE, C.W. (ed.) Evolution of Chromium Ore Fields. Van Nostrand Reinhold, New York, 265-297. LIPIN, B.R. 1984. Chromite from the Blue Ridge Province of North Carolina. American Journal of Science, 284, 507-529. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Geological Society, London, Memoirs, 11. LOCKNER, D.A. 1995. Rock failure. In: AHRENS, TJ. (ed.) Rock Physics and Phase Relations. American Geophysical Union, Washington, DC, 127-147. MANNING, C.E. 1999. Solubility of corundum + kyanite in H2O at 700°C, 1 and 1.5 GPa: evidence for high Al mobility in deep metasomatic environments. Geological Society of America, Abstracts with Programs, 31(7), A3 54. MCELHANEY, M.S. & McSwEEN, H.Y. JR 1983. Petrology of the Chunky Gal Mountain mafic-ultramafic complex, North Carolina. Geological Society of America Bulletin, 94, 855-874. McSwEEN, H.Y. JR & HATCHER, R.D. JR 1985. Ophiolites(?) of the southern Appalachian Blue Ridge. In: WOODWARD, N.E. (ed.) Field Trips in the Southern

Appalachians. University of Tennessee, Department of Geological Sciences, Studies in Geology, 9, 144-170. MCSWEEN, H.Y. JR, ABBOTT, R.N. JR & RAYMOND, L.A. 1989. Metamorphic conditions in the Ashe Metamorphic Suite, North Carolina Blue Ridge. Geology, 17, 1140-1143. MILLER, B.V., STEWART, K.G., MILLER, C.F. & THOMAS, C.W. 2001. U-Pb ages from the Bakersville, North Carolina eclogite: Taconian eclogite metamorphism followed by Acadian and Alleghenian cooling. Geological Society of America, Abstracts with Programs, 32(2), A62. MISRA, K.C. & CONTE, J.A. 1991. Amphibolites of the Ashe and Alligator Back Formations, North Carolina. Geological Society of America Bulletin, 103, 737-750. MISRA, K.C. & KELLER, F.B. 1978. Ultramafic bodies in the southern Appalachians: a review. American Journal of Science, 278, 389-418. MOORES, E.M. 1970. Ultramafics and orogeny, with models of the US Cordillera and the Tethys. Nature, 228, 837-842. MOORES, E.M. 1981. Ancient suture zones within continents. Science, 213, 41-46. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht. NICOLAS, A. 1990. Melt extraction from mantle peridotites: hydrofracturing and porous flow, with consequences for oceanic ridge activity. In: RYAN, M.P. (ed.) Magma Transport and Storage. Wiley, Chichester, 159-173. NICOLAS, A. & POIRIER, J.P. 1976. Crystalline Plasticity and Solid State Flow in Metamorphic Rocks. John Wiley, London. NICOLAS, A., GIRARDEAU, J. & MARCOUX, J. ET AL. 1981. The Xigaze ophiolite (Tibet): a peculiar oceanic lithosphere. Nature, 294, 414-417. NIIDA, K. 1997. Mineralogy of Mark peridotites: replacement through magma channeling examined from Hole 920D, Mark area. In: KARSON, J.A., CANNAT, M., MILLER, D.J. & ELTHON, D. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 153. Ocean Drilling Program, College Station, TX, 265-275. ODOM, A.L. & FULLAGAR, P.D. 1973. Geochronologic and tectonic relationships between the Inner Piedmont, Brevard Zone, and Blue Ridge Belts, North Carolina. American Journal of Science, 273, 133-149. O'HARA, MJ. 1968. Are ocean floor basalts primary magma? Nature, 220, 683-686. OZAWA, K. 1983. Evaluation of olivine-spinel geothermometry as an indicator of thermal history for peridotites. Contributions to Mineralogy and Petrology, 82, 52-65. PAGE, F.Z., ESSENE, E.J. & MUKASA, S.B. 2001. Thermobarometry of eclogites from the eastern Blue Ridge, North Carolina. Geological Society of America, Abstracts with Programs, 33(6), 251. PARKINSON, I.J. & PEARCE, J.A. 1998. Peridotites from the Izu-Bonin-Mariana forearc (OOP Leg 125): evidence for mantle melting and melt-mantle interaction in a supra-subduction zone setting. Journal

CHROME-SPINEL AND BLUE RIDGE OPHIOLITES of Petrology, 39, 1577-1618. PEARCE, J.A., LIPPARD, S.J. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PRATT, J.H. & LEWIS, J.V. 1905. Corundum and the Peridotites of North Carolina. North Carolina Geological Survey Bulletin, 1. QUICK, I.E. 1981. The origin and significance of large, tabular dunite bodies in the Trinity peridotite, northern California. Contributions to Mineralogy and Petrology, 78, 413-422. QUICK, I.E. & GREGORY, R.T. 1995. Significance of melt-wall rock reaction: comparative anatomy of three ophiolites. Journal of Geology, 103, 187-198. QUINN, MJ. 1991. Two lithotectonic boundaries in western North Carolina: geologic interpretation of a region surrounding Sylva, Jackson County. MS thesis, University of Tennessee, Knoxville. RANKIN, D.W. 1970. Stratigraphy and structure of Precambrian rocks in northwestern North Carolina. In: FISHER, G.W., PETTIJOHN, F.J., REED, J.C. JR & WEAVER, K.N. (eds) Studies of Appalachian Geology: Central and Southern. Interscience, New York, 227-246. RANKIN, D.W., ESPENSHADE, G.H. & SHAW, K.W. 1973. Stratigraphy and structure of the metamorphic belt in northwestern North Carolina and southwestern Virginia: a study from the Blue Ridge across the Brevard Fault Zone to the Sauratown Mountains Anticlinorium. American Journal of Science, 273, 1-40. RAYMOND, L.A. 1995. Petrology: the Study of Igneous, Sedimentary, and Metamorphic Rocks. Brown, Dubuque. RAYMOND, L.A. 1998. Geology of the Blue Ridge Belt of northwestern North Carolina. In: MILLS, H.H. (ed.) Deposits and Landforms on the Piedmont Slopes of Roan, Rich, and Snake Mountains, Northwestern North Carolina and Northeastern Tennessee. Southeastern Friends of the Pleistocene 1998 Field Trip (Guidebook), 1-8. RAYMOND, L.A. 2002. Petrology: the Study of Igneous, Sedimentary, and Metamorphic Rocks, 2nd. McGraw-Hill Higher Education, Boston, MA. RAYMOND, L.A. & ABBOTT, R.N. JR 1985. Metamorphism of Alpine ultramafic rocks, North Carolina Blue Ridge: I—observed metamorphic histories. Geological Society of America, Abstracts with Programs, 17, 695. RAYMOND, L.A. & ABBOTT, R.N. JR 1997. Petrology and tectonic significance of ultramafic rocks near the Grandfather Mountain Window in the Blue Ridge Belt, Toe Terrane, Western Piedmont Zone, North Carolina. In: STEWART, K.G., ADAMS, M.G. & TRUPE, C.H. (eds) Paleozoic Structure, Metamorphism, and Tectonics of the Blue Ridge of Western North Carolina. Carolina Geological Society, 1997 Field Trip and Annual Meeting, 67-85. RAYMOND, L.A. & ALLAN, J.F. 2001. Spinel compositions as clues to the origin and tectonic significance of metadunite bodies in the Blue Ridge Belt of the

277

Southern Appalachian Orogen. Geological Society of America, Abstracts with Programs, 33(6), A227. RAYMOND, L.A. & LOVE, A.B. 2002. Aspect ratios of Blue Ridge metaultramafic rocks and their significance. Geological Society of America, Abstracts with Programs, 34, A92. RAYMOND, L.A. & SWANSON, S.E. 1981. Deformation history of ultramafic rocks, Blue Ridge Belt, Southern Appalachian Mountains: correlations with Appalachian tectonic events. Geological Society of America, Abstracts with Programs, 13, 535. RAYMOND, L.A. & WARNER, R.D. 2001. Preface (to Selected Ultramafic and Related Rocks of the Southern Appalachian Orogen: Petrology and Tectonic Significance). Southeastern Geology, 40(3), i. RAYMOND, L.A., LEATHERMAN, L.E., VANCE, K., COOK, T. & ABBOTT, R.N. JR 1988. The Greer Hollow ultramafic body, eastern Blue Ridge Province, North Carolina: petrography, field relations, and implications for metamorphic history of the Blue Ridge Province. Geological Society of America, Abstracts with Programs, 20, 310. RAYMOND, L.A., YURKOVICH, S.P. & MCKINNEY, M. 1989. Block-in-matrix structures in the North Carolina Blue Ridge Belt and their significance for the tectonic history of the Southern Appalachian Orogen. In: HORTON, J.W. JR & RAST, N. (eds) Melanges and Olistostromes of the U.S. Appalachians. Geological Society of America, Special Papers, 228, 195-215. RAYMOND, L.A., MCCARTER, R. & LOVE, A. 1998. Petrography and structure of the Hoots Ultramafic Body, Rich Mountain, northwestern North Carolina. Geological Society of America, Abstracts with Programs, 30(4), 56. RAYMOND, L.A., LOVE, A. & MCCARTER, R. 1999. Petrogenesis and mineral chemistry of the Hoots Ultramafic Body, Rich Mountain, northwestern North Carolina. Geological Society of America, Abstracts with Programs, 31(3), 61. RAYMOND, L.A., LOVE, A. & MCCARTER, R. 2001. Petrology of the Hoots Ultramafic Body, Blue Ridge Belt, northwestern North Carolina. Southeastern Geology, 40, 149-162. RAYMOND, L.A., SWANSON, S.E. & LOVE, A.B. 2002. Physical, mineralogical, and chemical trends during retrograde metamorphism of metaultramafic rocks, Blue Ridge Belt, North Carolina, USA. Geological Society of America, Abstracts with Programs, 34(6), 431. ROLLINSON, H.R. 1993. Using Geochemical Data: Evaluation, Presentation, Interpretation. Longman, Harlow. RYAN, J.G., PETERSON, V.L., YURKOVICH, S.P., BURR, J.L. & KRUSE, S.E. 2001. Mafic and ultramafic rocks of olistostromal assemblages of the Blue Ridge in SW NC: insights into convergent margin processes of the Paleozoic. Geological Society of America, Abstracts with Programs, 33(6), 262. SANDFORD, R.F. 1978. Metamorphism, diffusion and metasomatism in an ultramafic blackwall zone, Blandford, Massachusetts. Geological Society of America Abstracts, 10(2), 84. SCHIERING, M.H., HEIMLICH, R.A. & PALMER, D.F.

278

L. A. RAYMOND ETAL.

1982. Radiometric and magnetic surveys of the Dark Ridge Dunite, Jackson County, North Carolina. Southeastern Geology, 23, 89-98. SCOTFORD, D.M. & WILLIAMS, J.R. 1983. Petrology and geochemistry of metamorphosed ultramafic bodies in a portion of the Blue Ridge of North Carolina and Virginia. American Mineralogist, 68, 78-94. SPRINGER, R.K. 1974. Contact metamorphosed ultramafic rocks in the western Sierra Nevada foothills, California. Journal of Petrology, 15, 160-195. STEVENS, R.K., STRONG, D.F. & KEANE, B.F. 1974. Do some eastern Appalachian ultramafic rocks represent mantle diapirs produced above a subduction zone? Geology, 2, 175-178. STEWART, K.G. & MILLER, B.V. 2001. The tectonic implications of 460 Ma eclogite along the Taconian suture in the eastern Blue Ridge of North Carolina. Geological Society of America, Abstracts with Programs, 33(2), A65. STEWART, K.G., ADAMS, M.G. & TRUPE, C.H. 1997. Paleozoic structural evolution of the Blue Ridge Thrust Complex, western North Carolina. In: STEWART, K.G., ADAMS, M.G. & TRUPE, C.H. (eds) Paleozoic Structure, Metamorphism, and Tectonics of the Blue Ridge of Western North Carolina. Carolina Geological Society, 1997 Field Trip and Annual Meeting, 21-31. SUHR, G., SECK, H.A., SHIMIZU, N., GUNTHER, D. & JENNER, G. 1998. Infiltration of refractory melts into the lowermost oceanic crust: evidence from dunite- and gabbro-hosted clinopyroxenes in the Bay of Islands Ophiolite. Contributions to Mineralogy and Petrology, 131, 136-154. SWANSON, S.E. 1980. Petrology of the Rich Mountain ultramafic bodies and associated rocks, Watauga County, North Carolina. Southeastern Geology, 21, 209-225. SWANSON, S.E. 1981. Mineralogy and petrology of the Day Book Dunite and associated rocks, western North Carolina. Southeastern Geology, 22, 53-77. SWANSON, S.E. 1999. Mineral compositions from metaultramafic rocks in the Spruce Pine Thrust Sheet, western North Carolina. Geological Society of America, Abstracts with Programs, 31(3), A70. SWANSON, S.E. 2001. Ultramafic rocks of the Spruce Pine area, Western North Carolina: a sensitive guide to fluid migration and metamorphism. Southeastern Geology, 40, 163-182. SWANSON, S.E. & RAYMOND, L.A. 1976. Alteration of the Day Book dunite, Spruce Pine District, western North Carolina. Geological Society of America, Abstracts with Programs, 8, 282. SWANSON, S.E. & RAYMOND, L.A. 2003. Petrogenesis of chromite in the metaultramafic rocks of the Spruce Pine area, western North Carolina, USA. In preparation.

SWANSON, S.E. & WHITTKOPP, R.W. 1976. Day Book Dunite: an alpine ultramafic in western North Carolina. Transactions of the American Geophysical Union, 57, 1025-1026. SWANSON, S.E., MALOOF, T.L., CROSE, D.E., RICIPUTI, L.R. & HOCHELLA, M.F. 1999. Oxygen isotope analyses of olivine from the Blue Ridge of western North Carolina. Geological Society of America, Abstracts with Programs, 31(7), A225. TENTHOREY, E.A., RYAN, J.G. & SNOW, E.A. 1996. Petrogenesis of sapphirine-bearing metatroctolites from the Buck Creek ultramafic body, southern Appalachians. Journal of Metamorphic Petrology, 14, 103-114. THOMAS, W.A. 1977. Evolution of Appalachian-Ouachita salients and recesses from reentrants and promontories in the continental margin. American Journal of Science, 277, 1233-1278. VANCE, R.K. & RAYMOND, L.A. 1994. Field and geochemical constraints on the origin of Greer Hollow ultramafic rocks, Eastern Blue Ridge Province, North Carolina. Geological Society of America, Abstracts with Programs, 26(4), 67. VARFALVY, V., HEBERT, R., BEDARD, J.H. & LAFLECHE, M.R. 1997. Petrology and geochemistry of pyroxenite dykes in upper mantle peridotites of the North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland: implications of the genesis of boninitic and related magmas. Canadian Mineralogist, 35, 543-570. VRONA, J.P. 1977. Structure of the Newdale ultramafite, Yancey County, North Carolina. MS thesis, Southern Illinois University, Carbondale. WARNER, R.D. 2001. Mineralogy and petrology of metaultramafic rocks at Buck Creek, North Carolina. Southeastern Geology, 40, 201-212. WATERS, C.L., HEWITT, L.K., STEWART, K.G. & MILLER, B.V. 2000. Tectonomermal evolution of the Ashe Metamorphic Suite south of the Grandfather Mountain Window, NC: implications for Paleozoic orogenic events in the eastern Blue Ridge. Geological Society of America, Abstracts with Programs, 32(2), A81. WEHR, F. & GLOVER, L. Ill 1985. Stratigraphy and tectonics of the Virginia-North Carolina Blue Ridge: evolution of a late Proterozoic-early Paleozoic hinge zone. Geological Society of America Bulletin, 96, 285-295. WILLARD, R.A. & ADAMS, M.G. 1994. Newly discovered eclogite in the southern Appalachian Orogen, northwestern North Carolina. Earth and Planetary Science Letters, 123, 61-70. ZHOU, M.-F., ROBINSON, P.T., MALPAS, J. & Li, Z. 1996. Podiform chromitites in the Luobusa Ophiolite (southern Tibet): implications for melt-rock interaction and chromite segregation in the upper mantle. Journal of Petrology, 37, 3-21.

Multi-stage evolution of the Tertiary Mineoka ophiolite, Japan: new geochemical and age constraints N. HIRANO 1 , Y. OGAWA 2 , K. SAITO 3 , T. YOSHIDA 4 , H. SATO 5 & H. TANIGUCHI 6 1

Department of Earth and Planetary Sciences, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro, Tokyo 152-8551, Japan (e-mail: [email protected]) 2 Institute of Geoscience, University ofTsukuba, Tsukuba 305-8571, Japan ^Department of Earth and Environmental Sciences, Faculty of Science, Yamagata University, Yamagata 990-8560, Japan ^Institute of Mineralogy, Petrology and Economic Geology, Graduate School of Science, Tohoku University, Sendai, 980-8578, Japan 5 Ocean Research Institute, University of Tokyo, Tokyo 164-8639, Japan 6 Komazawa University Senior High School, Tokyo 158-0098, Japan Abstract: The Mineoka ophiolite in the southern Boso Peninsula is situated in a unique tectonic setting in the collisional zone between the Izu and Honshu arcs in Japan. The ophiolitic rocks are composed mainly of tholeiitic pillow basalts and dolerites, alkali-basaltic sheet flows, and calc-alkaline dioritic to gabbroic rocks. The tholeiitic basalts show variable trace element compositions ranging from mid-ocean ridge basalt to island-arc basalt, whereas the alkalibasalts have a within-plate affinity. High-Fe and -Ti tholeiitic basalt and within-plate alkalibasalt have Ar/Ar ages of 49 ± 13 Ma and 19.62 ± 0.90 Ma, respectively. Three plutonic rocks have K—Ar ages of c. 25, 35 and 40 Ma. These ages are inconsistent with the known ages from the Pacific or Philippine Sea Plate. We infer that the Mineoka ophiolitic assemblage was part of another Tertiary oceanic plate, the 'Mineoka Plate', which underwent island-arc volcanism in the Miocene as a result of subduction initiation at a fracture zone or a transform fault system due to a change in the position of the Euler rotation pole of the Pacific Plate at c. 43 Ma. Eruption of within-plate type alkali basalts on the Mineoka Plate took place near the palaeoJapan continental arc just before the emplacement of the Mineoka ophiolite into the Japanese continental margin.

The Mineoka ophiolite and related rocks in central Japan include mainly Tertiary mafic and ultramafic rocks with pelagic and terrigenous materials and occur in the area sandwiched between two forearcs of the Japan and Izu arcs (Ogawa 1983; Ogawa et al. 1985; Ogawa & Taniguchi 1987, 1988; Fig. 1). The Mineoka ophiolite constitutes a significant component of a fault belt (henceforth called the 'Mineoka Belt'), which lies near a trench-trench-trench (TTT) triple junction, the Boso triple junction (Ogawa et al. 1985; Seno et al. 1989), in the NW Pacific Ocean. Various kinds of mafic and ultramafic rocks are distributed in a complicated fashion together with alkali-basalt, diorite, calcareous and siliceous pelagic sedimentary rocks, and tuffaceous rocks within the Mineoka Belt. All the igneous rocks in the Mineoka Belt occur as tectonic blocks in Tertiary terrigenous clastic sediments or in serpentinite bodies

(Fig. 2). Various hypotheses on the origin of these igneous rocks in the Mineoka Belt have been advanced. Ogawa & Taniguchi (1988) proposed the existence of a Tertiary oceanic plate, 'the Mineoka Plate', based on geochemical data and preliminary radiometric ages of basalts (Kaneoka et al. 1980), and gravity data for the southern Boso Peninsula (Tonouchi 1981). This hypothesis was supported by a recent geophysical study by Fujiwara et al. (1999), who reported the magnetic structures in the southern Boso Peninsula and demonstrated that the Mineoka Belt consists of fragments of an oceanic plate, Arai and coworkers (e.g. Uchida & Arai 1978; Arai & Takahashi 1988; Arai 1991) reported petrological and mineralogical characteristics of ultramafic rocks in the Mineoka Belt and coneluded that they are analogues of abyssal peridotites. Arai (1991) proposed that peridotites

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 279-298. 0305-8719/037$ 15 © The Geological Society of London 2003.

280

N. HIRANO ETAL.

beneath the Shikoku Basin intruded into the sea floor and were then accreted to the Honshu arc. Fujioka et al. (1995) inferred that serpentinite in seamounts of the Izu-Ogasawara-Mariana forearc area had been accreted to the Honshu arc, based on the similar mode of occurrence of serpentinized peridotites in both areas. However, Sato & Ogawa (2000) disputed the forearc serpentinite seamount origin for peridotites in the Mineoka Belt, based on petrological and mineralogical characteristics, and concluded that the Mineoka Plate hypothesis proposed by Ogawa & Taniguchi (1988) is a plausible explanation for the origin of the Mineoka ophiolite. However, the tectonic setting and evolution of the ophiolitic rocks are still unresolved. To discuss the origin of the ophiolitic rock assemblage in the Mineoka Belt and its implications for the tectonics of the Paleogene NW Pacific area, we need precise geochemical and geochronological data from the igneous rocks. Thus, in this paper, we describe and discuss the major and trace element geochemistry and ages of the igneous rocks in the Mineoka ophiolite. We attempt to use these data to unravel the evolution of these complicated ophiolitic assemblages. Ocean-floor deformation of the Mineoka ophiolite has been documented in a companion paper by Takahashi et al. (2003).

Geology of the Mineoka Belt Fig. 1. Bathymetric map (500 m contour interval) of the NW Pacific and eastern Philippine Sea Plate.

The Mineoka Belt, denned as a 5 km wide, eastwest-trending fault zone, occupies the southern

Fig. 2. Index map of the Mineoka Belt after Hirano & Okuzawa (2002).

MINEOKA OPHIOLITE, JAPAN

Fig. 3. Outcrop photographs: (a) tholeiite pillow basalt; (b) dolerite sheeted dyke complex; (c) alkali-basalt sheet flows; (d) sandstone inclusions in alkali-basalt lavas.

281

N. HIRANO ETAL.

282

part of the Boso Peninsula in central Japan, extending from Kamogawa Harbour in the east to Hota Beach in the west (Fig. 2). Sedimentary rocks in the Izu forearc have accreted to the Japan forearc along the Sagami Trough, and several accretionary prisms have formed on the southern edge of the Mineoka Belt between the early Miocene and the present (Ogawa et al. 1985; Saito 1992). The forearc basin sediments cover the earlier accretionary prisms unconformably. Because the middle Miocene Sakuma Group in the western Mineoka Belt is the oldest geological unit, and because it includes clasts of the Mineoka ophiolite (Saito 1992), we infer that these rocks underwent the same processes of accretion and emplacement. The pelagic sedimentary rocks, such as limestone and chert, form an almost continuous sequence ranging in age from late Paleocene to mid-Miocene, or from 55 to 17 Ma (Mohiuddin & Ogawa 1998). In map view, most of the igneous rocks are rounded to lozenge-shaped and range from metre to kilometre scale. They are surrounded by, or in fault contact with, the Mineoka Group (part of the Shimanto Supergroup, probably Paleogene in age: Nakajima et al. 1981; Ogawa et al. 1985; Watanabe & lijima 1989), Hota Group (latest Oligocene to mid-Miocene: Saito 1992; Suzuki et al. 1996), Sakuma Group (mid-Miocene: Saito 1992)

and serpentinized ultramafic bodies (Fig. 2). Basaltic rocks consist mostly of tholeiitic and rarely alkali-olivine pillow lavas (Tazaki & Inomata 1980; Ogawa & Taniguchi 1988). In the Mineoka Sengen area (Fig. 2), Cyprus-type umber deposits locally occur above the pillow basalts (Tazaki et al. 1980; lijima et al. 1990). Pelagic to hemipelagic and terrigenous sediments were deposited from Eocene to mid-Miocene time, and are now incorporated into the Mineoka Belt. In the Hegurinaka and Toge areas (Fig. 2), a sequence of basaltic clastic sandstone and conglomerate, micritic limestone, glauconitic shale and siliceous shale overlie alkali-basalt sheet flows (Arakawa Formation in Fig. 2) (Ogawa 1983; Takahashi 1994; Hirano & Okuzawa 2002). Basaltic clastic sediments mainly include fragments of alkali-basalts, with some fossils of benthic foraminifers, oysters and corals (Takahashi 1994; Mohiuddin & Ogawa 1996; Hirano & Okuzawa 2002). Mohiuddin & Ogawa (1996) reported mid-Eocene to early Oligocene and early to mid-Miocene foraminifers from these micritic limestones in the Hegurinaka area. Ogawa (1981) and Saito (1992) reported early Miocene radioralians from the siliceous shale. Tholeiitic pillow lavas are the dominant basaltic rock type in the ophiolite (Table 1). They are concentrated in the eastern part of the belt as large

Table 1. Geochemical compositions of bulk-rock samples Sample: Rock type:

BM591 tholeiite

BM09BT1 tholeiite

BM23BT2 tholeiite

BM25BT3 tholeiite

BM10SY tholeiite

BM07SG1 tholeiite

BM08SG2 tholeiite

BM21SG3 BM24SG4 alkali-basalt tholeiite

BM22SG6 tholeiite

BM29SG5 tholeiite

wt% SiO2 TiO2 A1203 Fe2O3 FeO MnO MgO CaO Na2O K20 P205 H2O+ H 2 O~ Total

47.88 3.23 11.19 11.10 6.19 0.23 5.19 7.95 3.08 0.20 0.26 2.73 0.77 100.00

47.92 1.64 14.09 3.21 7.44 0.18 7.97 9.44 3.87 0.26 0.15 3.26 0.56 100.00

47.88 1.63 14.56 7.00 3.51 0.17 7.78 11.07 3.22 0.11 0.14 1.56 1.36 100.00

49.48 1.59 14.93 5.99 3.95 0.15 6.55 9.85 4.27 0.29 0.12 1.55 1.28 100.00

49.83 2.26 14.38 5.69 4.23 0.18 6.64 10.53 3.38 0.38 0.21 1.18 1.12 100.00

50.75 2.10 13.32 5.68 5.04 0.19 6.50 9.63 3.81 0.54 0.21 1.51 0.72 100.00

48.65 1.71 13.61 6.46 4.63 0.22 6.30 11.46 3.34 0.59 0.16 2.67 0.20 100.00

48.61 2.31 11.73 7.73 5.26 0.24 6.49 8.42 4.44 0.54 0.24 2.42 1.57 100.00

48.03 2.08 13.67 6.53 5.11 0.21 6.10 8.93 3.08 1.25 0.24 2.19 2.60 100.00

48.05 1.97 14.29 5.95 2.84 0.39 8.49 10.29 3.27 0.10 0.18 1.87 2.33 100.00

47.89 1.58 14.39 4.64 3.50 0.14 6.31 11.74 3.67 0.89 0.12 4.51 0.62 100.00

52.7 18.1 41.3 24.7 19.4 7.3 21.2 0.4 0.9 101 2.1 348 84.8 228

215 16.6 50.4 336 15.9 5.4 110 1.0 1.7 188 tr 281 34.7 98.6

46.1 9.3 51.1 356 17.7 4.3 96.7 0.6 tr 145 tr 357 33.8 94.6

31.5 19.7 47.5 136 19.1 5.9 65.0 1.0 3.7 133 tr 271 45.2 132

43.0 15.7 52.1 101 17.3 6.7 98.2 tr 6.3 167 tr 261 42.2 130

46.6 11.2 53.8 290 16.9 3.0 113 tr 13.0 118 tr 305 37.4 91.3

71.9 10.0 33.1 15.4 21.2 1.2 6.2 2.8 0.4 200 tr 346 39.0 66.9

224 21.5 51.9 147 20.2 5.7 54.6 1.2 17.9 111 tr 374 45.1 150

17.5 23.4 57.4 225 17.0 6.1 88.0 0.9 tr 180 tr 453 40.6 122

22.2 14.3 49.8 274 16.7 3.0 65.2 1.7 6.4 294 tr 346 33.5 105

ppm Ba Ce Co Cr Ga Nb Ni Pb Rb Sr Th V Y Zr

tr, trace.

2126 7.0 55.5 386 17.8 4.9 160 0.2 0.4 175 tr 326 32.4 45.5

MINEOKA OPHIOLITE, JAPAN blocks (several hundred metres to 1000 m scale) in fault contact with serpentinite and terrigenous or tuffaceous clastic rocks. In the Mineoka Sengen area in particular (Fig. 2), large basalt blocks, some 100 m across, are abundant. The pillow lavas are close-packed and only rarely intruded by doleritic dykes (Fig. 3a and b). In the Bentenjima, Kamogawa Harbour, at the eastern tip of the Mineoka Belt, three blocks of pillow basalt and dolerite dykes are in fault contact (Takahashi et al 2003). Alkali-basalts are distributed in the central Mineoka Belt and occur as sheet flows (Fig. 3c) or vesicular pillow basalt. Locally, some sheet flows are made of picritic lavas (Fig. 3c). Hirano & Okuzawa (2002) analysed sandstone xenoliths included in the alkali-basalt sheet flows in the Toge area (Fig. 3d), and concluded that they were picked up during eruption because the basalt is chilled against the xenoliths. The sandstone is terrigenous, mainly composed of quartz, plagioclase, datolite (a low-temperature secondary replacement mineral, borosilicate, CaBSiO^OH)) and silicic volcanic fragments, and is similar in composition to the sandstone in the Mineoka Group (Hirano & Okuzawa 2002). Although the lavas occur with massive ultramafic rocks or Tertiary sedimentary rocks, the relative ages are unknown because they occur as tectonic blocks. Diorites and gabbros in the serpentinite are interpreted as originally intrusive plutons or dykes (Nakajima et al. 1981; Sato et al. 1999). BM06HS tholeiite

49.35 1.72 13.75 4.02 6.19 0.28 7.96 9.41 4.10 0.08 0.14 2.32 0.70 100.00

42.6 10.4 49.7 180 17.3 2.7 66.3 tr tr 130 tr 317 37.2 96.6

BM32NSGM picrite basalt

41.11 1.54 7.26 6.90 4.47 0.15 24.28 5.68 0.34 0.10 0.17 7.31 0.69 100.00

43.2 27.5 106 2849 11.0 11.4 1110 0.3 3.3 99.6 0.1 241 16.8 92.2

BM03HG1 alkali-basalt

BM04HG2 alkali-basalt

283

Petrography and geochemistry The basaltic and other intrusive rocks from the Mineoka ophiolite were analysed by X-ray fluorescence (XRF) for major and trace elements. After excluding the altered parts and some secondary veins, all samples were split and pulverized. H2<3~ and loss on ignition (LOI) were determined at 110 and 900 °C (4 h). Specimens for analysis were prepared by mixing 3.6 g of flux (anhydrous lithium tetraborate) and 1.8 g of powdered sample and fusing the mixture into glass beads following the method of Kimura & Yamada (1996). A Rigaku RIX 2000 X-ray fluorescence spectrometer was used at the Department of Geology, Faculty of Education, Fukushima University. Analytical methods and statistical data have been given by Kimura & Yamada (1996). The results are recalculated to total 100% to insert the H2O" and LOI. FeO was determined by potassium permanganate titration. All geochemical data are listed in Table 1. For sample localities and their distribution, please see Figures A1-A5 in the Appendix. Based on Barker diagrams, the igneous rocks of the Mineoka ophiolite are divided into sub-alkali and alkali varieties (Miyashiro 1978) (Fig. 4a). However, Na and K may be prone to modification by alteration. To determine the original magma compositions, we used the P2O5-Zr diagram (Fig. 4b) (Winchester & Floyd 1976). Many samples, shown by squares and one circle, are 'high-alkali tholeiite', which occurs as close-packed pillow

BM05HG3 BM26SK tholeiite tholeiite

BM14FG gabbro

BM28SNJ diorite

BM31YD diorite

BM30RS gabbro

BM12GSCH gr.schist

41.92 2.80 11.83 6.69 5.64 0.27 11.81 9.43 2.34 0.27 0.32 5.53 1.16 100.00

47.23 3.20 13.50 4.51 6.85 0.16 8.04 8.00 3.80 0.93 0.37 2.88 0.52 100.00

47.94 2.09 13.87 4.86 6.08 0.20 7.28 9.16 4.17 0.11 0.23 2.97 1.04 100.00

48.71 1.74 14.51 5.97 3.33 0.15 7.60 10.21 3.78 0.59 0.16 2.47 0.79 100.00

47.00 0.32 25.53 1.70 2.27 0.07 4.86 9.11 4.06 0.50 0.01 4.16 0.40 100.00

59.89 0.54 17.04 2.11 3.28 0.09 3.78 6.41 4.83 0.47 0.06 1.37 0.13 100.00

58.46 0.38 18.75 2.98 2.31 0.08 3.81 7.38 3.89 0.48 0.04 1.23 0.22 100.00

49.33 1.02 14.25 3.16 5.26 0.15 8.72 9.14 4.46 0.25 0.02 4.05 0.19 100.00

48.10 1.29 14.46 4.90 5.49 0.17 8.29 11.40 2.70 0.44 0.10 2.16 0.51 100.00

224 35.8 67.7 832 22.6 23.1 389 1.2 3.1 238 1.5 238 25.6 155

375 41.8 55.2 146 19.6 27.6 109 tr 8.9 394 1.1 205 26.4 186

46.8 18.0 54.6 297 17.9 6.4 137 tr tr 262 tr 302 45.3 136

73.2 16.5 52.2 340 17.3 3.8 103 2.0 6.8 132 tr 374 36.6 98.2

16.9 tr 26,6 86.7 18.0 0.05 71.2 tr 2.6 659 tr 228 9.9 75.7

91.3 7.00 26.4 97.3 15.9 1.0 83.5 tr 1.4 211 tr 275 18.0 77.2

80.2 1.00 23.2 87.0 17.2 0.9 70.2 0.3 1.1 276 tr 340 11.5 65.2

92.0 1.6 41.4 478 13.2 1.0 149 1.1 tr 353 tr 355 33.9 63.2

42.8 7.2 50.9 338 15.2 2.1 98.6 0.3 6.0 119 tr 262 29.2 66.8

284

N. HIRANO ETAL. lavas and aphyric basalts in the Mineoka Belt. Also present are alkali-basalts (BM03HG1, BM04HG2, BM21SG3), shown by two triangles and one circle, which contain olivine and clinopyroxene phenocrysts and occur as sheeted lava flows with some picrite basalts or vesicular pillow lavas mainly in the western Mineoka Belt (Hirano & Okuzawa 2002). However, a non-vesicular, aphyric alkali-basalt sample (BM21SG3), shown by a circle, is from Mineoka Sengen, where tholeiitic basalts dominate. Another sample, BM32NSGM, not plotted in Figure 4, shows the composition of groundmass separated from the picrite basalt. Many tholeiitic samples are characterized by high FeO* (FeO + 0.9Fe2O3) and high TiO2 (FeO*/MgO >1.4wt% and TiO2 >2.0 wt%). Tholeiitic basalts lie along the trend of abyssal tholeiite in Figure 4c. Diorites are composed of euhedral plagioclase, subhedral to anhedral hornblende and anhedral quartz. Other minor minerals include orthopyroxene in sample BM31-YD2 and secondary chlorite replacing amphibole in BM28-SNJ. Gabbros are mainly composed of euhedral hornblende and anhedral plagioclase. The sample BM30RS includes rare orthopyroxene in hornblende crystals. Sample BM14FG has very high A12O3 (27 wt%; Table 1), suggesting that it is a plagioclase-rich cumulate; however, it is difficult to identify the original magma type of the gabbros. All the basaltic rocks were plotted on the discrimination diagrams of Pearce & Norry (1979) and Pearce & Cann (1973) (Fig. 5a-c) and normal mid-ocean ridge basalt (N-MORB) normalized spidergrams of Pearce (1983) (Fig. 5d-g). The basaltic rocks fall into the fields of ocean-floor basalts (OFB or MORB), island arc basalts (IAB or low K tholeiite (LKT); samples BM25BT3 and BM21SG3) and ocean island basalts (OIB or within-plate basalts (WPB); samples BM03HG1, BM04HG2 and BM32NSGM) (Fig. 5a and b). Alkali-basalts from the central Mineoka Belt reported by Hirano & Okuzawa (2002), are shown by open triangles in some plots and a fine line in Fig. 5e, and are the same type of rocks as OIB of this study.

Fig. 4. (a) SiO2-total alkali diagram. Nomenclature of volcanic rocks is after Cox et al. (1979). Continuous line dividing alkalic and sub-alkalic magma series is from Miyashiro (1978). Picrite basalt (BM32NSGM) is not plotted, (b) P2O5-Zr diagram, (c) FeO*-FeO*/MgO diagram (Miyashiro 1973). Continuous line and dotted lines show the abyssal tholeiite and volcanic rocks in several arcs (Miyashiro 1973). TH, tholeiite; AL, alkalibasalts; CA, calc-alkaline series. Squares and circles, tholeiitic basalts; triangles, alkali-basalts; stars and crosses, gabbro and diorite, respectively; open triangles, data from Hirano & Okuzawa (2002).

Geochronology K-Ar dating Hornblende and plagioclase separates from diorites (sample BM28SNJ from the Kawaguchi region and BM31YD2 from the Yamada region) and gabbro (BM14FG from the Futago region) were analysed by the K-Ar method. The samples were crushed to 30-50 mesh and minerals were separated into mafic and felsic groups using an

MINEOKA OPHIOLITE, JAPAN

285

Fig. 5. Discrimination diagrams for incompatible trace elements. Symbols are as in Figure 4. In this figure, tholeiitic basalts can be divided into ocean floor basalts (OFB and MORE) and island arc related basalts (LKT and IAB). (a) Zr/Y-Zr plot (Pearce & Norry 1979; Pearce 1983). (b) Ternary plot of Zr-Ti/100-3Y (Pearce & Cann 1973). (c) Ternary plot of Zr-Ti/100-Sr/2 for samples plotted in OFB & LKT & CAB area of (b) (Pearce & Cann 1973). (d-g) N-MORB normalized spider diagrams for each of the type basalts (OFB, WPB and IAB) and gabbros, respectively. Fine lines in (e) are after Hirano & Okuzawa (2002).

isodynamic separator. Hornblende was separated from the other mafic minerals by heavy liquids. Potassium analyses were carried out by XRF using a Philips PW1404 system at the University of Tsukuba. Based on Nakano et al. (1997), we estimated 5% error in the potassium content. Argon isotopic analyses with 38Ar spike were carried out at Yamagata University. The analytical

methods have been described by Saito et al. (1991). The K-Ar ages are summarized in Table 2. Experimental errors are shown as la. Some of these K-Ar dates will need to be confirmed by other methods, such as Ar/Ar. From sample BM28SNJ, we obtained ages of 27.9 ± 1.8 Ma for hornblende and 24.1 ±1.4 Ma for plagioclase.

Table 2. K-Ar age results Sample

Rock

Mineral

BM28SNJ

Diorite

BM31YD BM14FG

Diorite Gabbro

Hornblende Plagioclase (An50) Hornblende Hornblende

K (wt %)

40

Ar/36Ar

Radiogenic 40Ar Air (10~7 STP content cm3 g-1) (%)

Age (Ma)

0.171 0.312

405.2 ±6.1 373.3 ±2.6

1.87 ±0.08 2.95 ± 0.09

73 79

27.9 ± 1.8 24.1 ± 1.4

0.099 0.121

567.7 ± 6.0 455.3 ±4.3

1.59 ±0.03 1.68 ±0.03

52 65

40.9 ±2.1 35.3 ± 1.9

N. HIRANO ETAL.

286

Kara et al. (1989) reported a fission-track age of 34.9 ± 1.5 Ma on zircon from diorite in the Yamada area, which is from the same site as sample BM31-YD2, from which we obtained a K-Ar age of 40.9 ± 2.1 Ma on hornblende. Closure temperatures of plagioclase in the K-Ar system and zircon in the fission-track system are lower than the closure temperature of hornblende. Therefore, in plutonic rocks hornblende ages are older than plagioclase and zircon ages. The hornblende in gabbro sample BM14FG has an age of 35.3 ± 1.9 Ma. Ar/Ar dating Samples of tholeiitic basalt (BM591) and alkalibasalt (BM04HG2) were also analysed by the 40 Ar/39Ar method. All samples were crushed to 50-100 mesh grains, were wrapped in aluminium foil and sealed in quartz vials (70mm in length, 10 mm in diameter) under vacuum, with flux monitors biotite (HD-B1), K2SO4 and CaF2. The samples were irradiated for 24 h in the Japan Material Testing Reactor (JMTR), Tohoku University. During the irradiation, the samples were shielded by Cd foil to reduce thermal neutroninduced 40Ar production from 40K (Saito 1994). The Ar extraction and Ar isotopic analyses were carried out at Yamagata University. The samples were incrementally heated to 1500°C in a Mo crucible by induction heating. Gases were extracted in eight or nine steps between 600 and 1500°C. The analytical methods have been de-

scribed by Saito et al. (1991). The results of the Ar/Ar dating are summarized in Table 3. All errors are shown as 2a. Sample BM591, a tholeiite from Kamogawa Harbour, has an age of 49 ± 13 Ma (Fig. 6a). Because of the very low K in tholeiitic basalt, the error is large at the 95% confidence level. The two lower temperature fractions and the highest temperature fraction may contain excess argon and are omitted from the age determination. Sample BM04HG2, an alkali-basalt from the Hegurinaka region, yielded a good age of 19.62 ± 0.90 Ma based on three lower temperature fractions, which contain more than 88% 39Ar in the gas contents (Fig. 6b). Discussion We discuss here the relation between the mode of occurrence, geochemistry and radiometric ages of the basaltic and other igneous rocks from the Mineoka ophiolite to verify the origin and tectonic implications of these rock assemblages. As described above, the ophiolitic rocks are composed mainly of tholeiitic pillow basalts and doleritic sheeted dykes, alkali-basaltic sheet flows, and dioritic to gabbroic rocks. Some tholeiitic basalts are Fe-Ti-basalt, and are considered to be fractionation products of oceanic tholeiite magmas (Melson & O'Hearn 1979). Such rocks are rare in island-arc tholeiites. Moreover, most of tholeiitic basalts have trace element compositions typical of ocean-floor basalt. Two alkali-basalts and one

Table 3. Ar/Ar age results Temperature (°C)

36

Ar/40Ar (X 10~4)

37

Ar/40Ar (X 10~4)

BM591, tholeiite (J= 0.00457 ± 0.00014) 0.2 ± 4.5 600 25.9 ± 1.8 7.5 ± 2.2 900 26.7 ± 1.6 1000 19.1 ± 1.6 26.3 ± 1.6 1100 19.3 ± 1.3 26.2 ± 1.6 16.4 ±4.9 30.4 ± 1.1 1200 14.5 ± 6 28.9 ±1.6 1300 22 ± 10 1400 27.7 ± 1.0 26.1 ±4.5 31.09 ±0.86 1500 63 ± 12 1550 17.7 ± 1.6 BM04HG2, alkali-basalt (]= 0.002890 ± 0.000068) 5.32 ± 0.79 600 17.6 ±2.9 4.07 ± 0.36 3.6 ±2.9 900 8.81 ±0.66 6.4 ± 2.0 1000 20.8 ± 3.5 1100 13.2 ±3.8 5.0 ±2.4 27.4 ± 2.9 1200 5.2 ± 1.2 1300 28.0 ±2.3 1400 7.10 ±0.67 30.3 ± 1.9 4.00 ± 0.92 1500 33.4 ±2.1

39

Ar/40Ar

39

Ar (%)

Age (Ma)

0.24 ±0.14 1.91 ±0.13 4.01 ±0.12 3.65 ±0.19 2.32 ±0.12 1.73 ±0.11 2.778 ± 0.096 1.364 ±0.022 2.789 ± 0.054

2.4 10.4 14.2 25.7 13.1 9.3 6.5 7.0 11.4

(65 ± 33) X 10 89 ±21 45.5 ± 9.7 50± 11 36± 11 68 ±22 53.4 ± 9.2 48 ± 14 136 ± 14

11. 24 ±0.25 24.82 ± 0.26 21.56 ±0.20 7.92 ± 0.24 4.69 ±0.16 2.99 ±0.16 1.93 ±0.12 1.57 ±0.11

18.7 48.2 21.4 2.7 1.8 2.4 2.9 1.9

(x io-2)

22.1 ± 1.5 18.70 ±0.59 19.53 ± 0.70 39.8 ± 6.7 21.1 ±7.5 29.9 ± 7.3 28 ± 10 4 ± 17

For BM591, (36Ar/37Ar)Ca = (8.57 ± 0.32) X 1(T4, (39Ar/37Ar)Ca = (2.23 ± 0.67) X 10~4, (40Ar/39Ar)K = (0.43 ± 0.90) X 1(T2. For BM04HG2, (36Ar/37Ar)Ca = (3.032 ± 0.053) X 10~4, (39Ar/37Ar)Ca = (8.64 ± 0.48) X 10"4; (40Ar/39Ar)K is neglected.

MINEOKA OPHIOLITE, JAPAN

287

Fig. 6. Ar/Ar age results. 39Ar/40Ar-36Ar/40Ar isochrons and age spectra of sample BM591 (a) and sample BM04HG2 (b), respectively.

picrite basalt (samples BM03HG1, BM04HG2 and BM32NSGM) have a within-plate affinity. In the central Mineoka Belt, Hirano & Okuzawa (2002) reported a sequence of sheeted lavas of alkalibasalt and picrite basalt, suggesting that these types are related. Another alkali-basalt (sample BM21SG3), one tholeiite (sample BM25BT3) and the diorites are from island arc type magmas. The original magma type of the gabbros is unknown because they are mineralogicaly heterogeneous. Ar/Ar and K-Ar dates yield ages of 49 ± 13 Ma for the tholeiite (Fe-Ti-basalts; BM591), 19.62 ± 0.90 Ma for alkali-basalts (BM04HG2), and c. 25, 35 and 40 Ma for the plutonic rocks (two diorites and gabbro). A preliminary Ar/Ar age by Kaneoka et al (1980) (40-50 Ma) approximately agrees with our age for sample BM591. The pelagic sedimentary rocks, such as limestone and chert of the Kamogawa Group, range in age from late Paleocene to mid-Miocene, i.e. roughly from 17 to 55 Ma (Mohiuddin & Ogawa 1998). Some stratigraphic gaps, possible hiatuses,

are included within these successions, but the group may correspond to pelagic deposits on an oceanic plate. The mid-Eocene to early Oligocene micritic limestone (Hegurinaka Limestone; Mohiuddin & Ogawa 1998) is associated with conformable sequences of alkali-basalt, basaltic sandstone, glauconitic shale, micritic limestone and siliceous shale in the Hegurinaka area. The Hegurinaka Limestone differs in age from the Early Miocene alkali-basalt, micritic limestone (Heguri Formation; Mohiuddin & Ogawa 1998) and siliceous shale (Arakawa Formation; Saito 1992; Mohiuddin & Ogawa 1998), suggesting that it consists of tectonic blocks in fault contact with the Early Miocene sequences.

Where did the Mineoka ophiolite form? Previous studies suggested three models for the tectonic evolution of the Paleogene Philippine Sea Plate and the Pacific Plate. In model 1, the Philippine Sea Plate separated from the Kula Plate

288

N. HIRANO ETAL.

and was juxtaposed with the Pacific Plate by a north-south-trending transform fault before initiation of subduction of the Pacific Plate. The TTT triple junction has remained near its present position (Matsuda 1978). Model 2 suggests that the Philippine Sea Plate and the TTT triple junction drifted eastward along the SE Japan margin (Seno & Maruyama 1984; Hall et aL 1995), whereas model 3 calls for southwestward drift of the Philippine Sea Plate and the TTT triple junction along the NE margin of Japan (Otsuki 1990). Koyama et aL (1992) presented palaeomagnetic evidence for a clockwise rotation of the IzuOgasawara forearc region and northeastward drift of the Philippine Sea Plate after the early Oligocene, which supports model 2. In model 2, the Paleogene island arcs, such as the Ogasawara and Mariana Arcs and Kyushu-Palau Ridge, were originally distributed along the northern margin of the Philippine Sea Plate. This requires a southward subduction of the adjacent plate in the north (Fig. 7). However, the inferred southward subduction of the Pacific Plate under these island arcs is problematic, because the Pacific Plate was moving NNW before 43 Ma (see Wessel & Kroenke 1998) (Fig. 7). To infill the area north of the Paleogene Philippine Sea Plate, there might have been another oceanic plate, the Mineoka Plate (Ogawa & Taniguchi 1988), as the counterpart of the North New Guinea Plate (Seno 1984). In the Philippine Sea Plate at present, there are some extinct back-arc basins and active and extinct island arcs (Fig. 8). Paleogene basalts are distributed in the Ogasawara and Mariana Arcs and West Philippine Basin (e.g. Seno & Maruyama 1984). The Ogasawara and Mariana Arcs, however, did not produce Fe-Ti-basalts but rather boninite and low-Ti basalt or calc-alkaline rocks (Kuroda et aL 1978; Reagan & Meijer 1984). If the 49 ± 13 Ma tholeiitic basalt in the Mineoka ophiolite were part of the Philippine Sea Plate, it might have originated in the 35-60 Ma West Philippine Basin (Hilde & Lee 1984). However, the Cretaceous Amami Plateau isolated the Mineoka ophiolite from the West Philippine Basin even before the opening of the Shikoku and Parece Vela Basins (12-25 and 17-30 Ma, respectively; Okino et aL 1994, 1999). In addition, the Mineoka ophiolite could not have come from the Pacific Plate, because it contains only Jurassic and Cretaceous ocean-floor basalts and seamounts (Takigami et aL 1989; Pringle & Duncan 1995; Hirano et aL 2002). Because the ages of the igneous rocks in the Mineoka Belt are inconsistent with all known ages from the Pacific Plate and the Philippine Sea Plate, we infer that the Mineoka ophiolitic assemblage was part of another oceanic plate, desig-

Fig. 7. Plate tectonic reconstruction of the Philippine Sea Plate before 43 Ma (Seno & Maruyama 1984; Koyama et al. 1992; Hall et al. 1995). Another oceanic plate that subducted from the north under the Philippine Sea Plate is required because the Pacific Plate had NNW absolute motion (Seno & Maruyama 1984).

nated as the 'Mineoka Plate' by Ogawa & Taniguchi (1988) and Sato & Ogawa (2000).

Eruption environment of the OIB in the Mineoka ophiolite Hirano & Okuzawa (2002) reported sandstone xenoliths in OIB-type alkali-basalt sheet flows in the Mineoka Belt. The sandstone is composed of terrigenous fragments such as plagioclase, quartz, acidic volcanic fragments and datolite, and is not similar to the basaltic sedimentary rocks and shale conformably deposited on the alkali-basalts (Hirano & Okuzawa 2002). The sandstone xenoliths and the Ar/Ar age of sample BM04HG2 provide important information about the geological setting of this volcanism, indicating that the Mineoka Belt alkali-basalts erupted near the palaeo-Japan arc, which was a continental arc before the Japan Sea opened around 15 Ma (Otofuji & Matsuda 1983). OIB-like alkali-basalts are reported from the Kinan Seamounts in the central Shikoku Basin (Ishii et aL 2000; Sato et aL 2003). Although the Kinan Seamounts appear to be a linear seamount chain along the palaeo-spreading centre of the Shikoku Basin, they are not of hotspot origin because they show no age progression (Ishii et aL 2000). These seamounts were probably formed after cessation of spreading in the Shikoku Basin. The setting of the Kinan Seamounts near the Nankai Trough may be an analogue for the alkalibasalt with sandstone xenoliths. However, the alkali-basalts in the Mineoka Belt are around

MINEOKA OPHIOLITE, JAPAN

Fig. 8. Formation ages of the ocean basin and arcs shown on a 1000 m contour bathymetric map of the Philippine Sea Plate and its surrounding area. Ages are after Matsuda et al. (1975), Ozima et al. (1977, 1980), Marsh et al. (1980), Okino et al. (1994, 1999), Hicky-Vargas (1998) and Ishii et al. (2000).

289

290

N. HIRANO ETAL.

Fig. 9. Tectonic reconstruction models for the origin of the Mineoka ophiolitic rocks, (a) Formation of OFB-type tholeiite before 43 Ma. The Philippine Sea Plate had moved from the SW in Tertiary time (Seno & Maruyama 1984; Koyama et al 1992; Hall et al 1995). Left and right figures show the cases of MORE and BABB type tholeiite in this study, respectively, (b) Island arc stage on the Mineoka Plate after Euler pole change of the Pacific Plate, (c) Alkali-basalt erupted on the terrigenous sedimentary rocks, Mineoka Group (Hirano & Okuzawa 2002), near the palaeo-Japan Arc. The sites of three triple junctions relative to the Japanese coastline are approximate.

MINEOKA OPHIOLITE, JAPAN 20 Ma old, distinctly older than the OIB-like basalts of the Kinan Seamounts. The alkali-basalt volcanism presumably occurred after the cessation of back-arc rifting in the Mineoka Arc. In that case, the Mineoka Plate might have been a backarc basin before the alkali-basalt eruption at 20 Ma.

291

din, N. Takahashi, K. Okuzawa and S. Oyamada for their help in the field. Thanks are due to T. Nakano, the late K. Fukunaga, and S. Furusawa, T. Ohmura and M. Ishii for their help with the Ar/Ar and K-Ar dating. We also thank the staff of the Institute for Material Research, Tohoku University, for irradiating samples for Ar/Ar dating in the JMTR (Japan Material Testing Reactor). J. Kimura, T. Asaki, R. Takashima and K. Aizawa are thanked for the XRF analyses.

Conclusions Sometime in the Eocene, probably about 49 ± 13 Ma, tholeiitic basalts were generated as ocean floor of the Mineoka Plate (Fig. 9a). However, the trace element data do not discriminate between a MORE or back-arc basin basalt (BABB) setting for the Mineoka tholeiitic basalts. The Pacific Plate had a NNW absolute motion before 43 Ma (Wessel & Kroenke 1998), and at that time the very young Mineoka Plate and old Pacific Plate may have been juxtaposed either by a transform fault or by a subduction zone. The NNW-moving Pacific Plate may have subducted under the Mineoka Plate before 43 Ma (Fig. 9a). The Mineoka Plate experienced island-arc volcanism at 25 and 40 Ma as a result of subduction initiated at a fracture zone or a transform fault system, due to a change in absolute motion of the Pacific Plate from NNW to WNW at 43 Ma (Wessel & Kroenke 1998) (Fig. 9b). The diorites and island-arc basalts represent the products of this intra-oceanic arc magmatism (called the 'Mineoka Arc'), which may have been small, with only intermittent volcanism. Additional age dating is needed for a more detailed discussion of the Mineoka Arc. The two samples identified as island arc basalt in this study are not dated, but may also be from the Mineoka Arc. Rift volcanism associated with back-arc basin opening might have occurred within the Mineoka Plate after establishment of this island arc. Eruption of the withinplate-type alkali-basalts probably took place at around 20 Ma on the Mineoka Group near the palaeo-Japan continental arc (Fig. 9c). These new interpretations are based on the recognition of c. 25 and 40 Ma island arc magmatism associated with previous ophiolitic rocks, and continent-proximal alkali-basalt formed at 20 Ma (Fig. 9). Our data and interpretations support the idea of the existence of a Mineoka Plate as previously proposed by Ogawa & Taniguchi (1988) and Sato & Ogawa (2000). We conclude that the ophiolitic rocks in the Mineoka area were derived from the Mineoka Plate, not from the Philippine Sea Plate or Pacific Plate. We appreciate the informative and helpful discussions with Y. Dilek and M. F. J. Flower on the geology of the Mineoka Belt. We extend our thanks to M. M. Mohiud-

References ARAI, S. 1991. The Circum-Izu Massif peridotite, central Japan, as back-arc mantle fragments of the IzuBonin arc system. In: PETERS, TJ., NICOLAS, A. & COLEMAN R.G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Kluwer, Dordrecht, 801-816. ARAI, S. & TAKAHASHI, N. 1988. Relic plagioclasebearing harzburgite from the Mineoka Belt, Boso Peninsula, central Japan. Journal of Mineralogy, Petrology and Economic Geology, the Japanese Association of Mineralogists, Petrologists and Economic Geologists, 83, 210-214 (in Japanese with English abstract). Cox, K.G., BELL, J.D. & PANKHURST, R.J. 1979. The Interpretation of Igneous Rocks. Allen and Unwin, London. FUJIOKA, K., TANAKA, T. & AOIKE, K. 1995. Serpentine seamount in Izu-Bonin and Mariana forearcs— observation by a submersible and its relation to onland serpentinite belt. Journal of Geography, Tokyo Geographical Society, 104, 473-494 (in Japanese with English abstract). FUJIWARA, T., KlNOSHITA, H. & MORIJIRI, R.

1999.

Magnetic structure of the southern Boso Peninsula, Honshu, Japan, and its implications for the formation of the Mineoka ophiolite belt. Earth, Planets, Space, 51, 413-424. HALL, R., ALI, J.R., ANDERSON, C.D. & BAKER, S.J. 1995. Origin and motion history of the Philippine Sea Plate. Tectonophysics, 251, 229-250. KARA, Y., KUSADA, T. & NIREI, H. 1989. Fission track age of diorite in Mineoka-belt. Bulletin of Environmental Geology, Research Institute of Environmental Geology, Chiba, Japan, 20, 65-68 (in Japanese with English abstract). HICKY-VARGAS, R. 1998. Origin of the Indian Oceantype isotopic signature in basalts from Philippine Sea plate spreading centers: an assessment of local versus large-scale processes. Journal of Geophysical Research, 103, 20963-20979. HIRANO, N. & OKUZAWA, K. 2002. Occurrence of the sandstone included in the alkali-basalt lava flow from the western Mineoka Belt, Boso Peninsula, Japan, and its tectonic significance. Journal of the Geological Society of Japan, 108, 691-700 (in Japanese with English abstract). HIRANO, N., OGAWA, Y. & SAITO, K. 2002. Long-lived Early Cretaceous seamount volcanism in the Mariana Trench, Western Pacific Ocean. Marine Geology, 189,371-379. HILDE, T.W.C. & LEE, C.S. 1984. Origin and evolution

292

N. HIRANO ETAL.

of the West Philippine Basin: a new interpretation. Tectonophysics, 102, 85-104. IIJIMA, A., WATANABE, Y., OGIHARA, S. & YAMAZAKI, K. 1990. Mineoka umber: a submarine hydrothermal deposit on an Eocene arc volcanic ridge in central Japan. In: PARNELL, I.E., LIANJUN, Y. & CHANGMING, C. (eds) Sediment-hosted Mineral Deposits. Special Publication of the International Association of Sedimentologists, 11, 73-88. ISHII, T., SATO, H. & MACHIDA, S. ET AL. 2000. Geological and petrological studies of the Kinan and Izu-Ogasawara-back arc-echelon Seamount Chains. Bulletin of the Geological Survey of Japan, 51, 615-630 (in Japanese with English abstract). KANEOKA, I., TAKIGAMI, Y., TONOUCHI, S., FURUTA, T., NAKAMURA, Y. & HIRANO, M. 1980. Pre-Neogene volcanism in the central Japan based on K-Ar and Ar-Ar analyses. Abstracts 1981 IAVCEI Symposium—Arc Volcanism, Tokyo and Hakone. Volcanological Society of Japan, Tokyo, 166. KIMURA, J. & YAMADA, Y. 1996. Evaluation of major and trace element XRF analyses using a flux to sample ratio of two to one glass beads. Journal of Mineralogy, Petrology and Economic Geology, the Japanese Association of Mineralogists, Petrologists and Economic Geologists, 91, 62-72. KOYAMA, M., CISOWSKI, S.M. & PEZARD, P. ET AL. 1992. Palaeomagnetic evidence for northward drift and clockwise rotation of the Izu-Bonin forearc since the early Oligocene. In: TAYLOR, B. & FUJIOKA, K. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 126. Ocean Drilling Program, College Station, TX, 353-370. KURODA, N., SHIRAKI, K. & URANO, H. 1978. Boninite as a possible calc-alkaline primary magma. Bulletin ofVolcanology, 41, 563-575. MARSH, N.G., SAUNDERS, A.D., TARNEY, J. & DICK, H.J.B. 1980. Geochemistry of basalt from the Shikoku and Daito Basins, Deep Sea Drilling Project Leg 58. In: KLEIN, G. DE¥. & KOBAYASHI, K. (eds) Initial Reports of Deep Sea Drilling Project, Leg 58. US Government Printing Office, Washington, DC, 805-842. MATSUDA, J., SAITO, K. & ZASU, S. 1975. K-Ar age and Sr isotope ratio of the rocks in the manganese nodules obtained from the Amami Plateau, western Philippine Sea. In: NAKAZAWA, K. (ed.) Symposium on the Geological Problems of the Philippine Sea. Geological Society of Japan, Tokyo, 99-101 (in Japanese). MATSUDA, T. 1978. Collision of the Izu-Bonin arc with central Honshu: Cenozoic tectonics of the fossa magna, Japan. Journal of Physics of the Earth, 26 (Supplement), S409-S421. MELSON, W.G. & O'HEARN, T. 1979. Basaltic glass erupted along the Mid-Atlantic Ridge between 037°N: relationship between composition and latitude. In: TALWANI, M., HARRISON, C.G. & HAYES, D.E. (eds) Deep Sea Drilling Results in the Atlantic Ocean: Ocean Crust. American Geophysical Union, Maurice Ewing Series, 2, 273-284. MIYASHIRO, A. 1973. The Troodos ophiolitic complex was probably formed in an island arc. Earth and Planetary Science Letters, 19, 218-224.

MIYASHIRO, A. 1978. Nature of alkalic volcanic rock series. Contributions to Mineralogy and Petrology, 66,91-104. MOHIUDDIN, M.M. & OGAWA, Y. 1996. Middle Eocene to early Oligocene planktonic foraminifers from the micritic limestone beds of the Heguri area, Mineoka belt, Boso Peninsula, Japan. Journal of Geological Society of Japan, 102, 611-617. MOHIUDDIN, M.M. & OGAWA, Y. 1998. Early Miocene pelagic sequences in the Mineoka Belt, Boso Peninsura, Japan. Journal of Geological Society of Japan, 104, 1-12. NAKAJIMA, T., MAKIMOTO, H., HIRAJIMA, J. & TOKUHASHI, S. 1981. Geology of Kamogawa District. Quadrangle Series, scale 1:50,000. Geological Survey of Japan, Tokyo (in Japanese with English abstract). NAKANO, T., JEON, S.R. & SUENO, S. 1997. X-ray fluorescence analysis of rock samples using Philips PW 1404(1): simultaneous determination of major and trace elements using glass beads of GSJ igneous rock reference samples. Annual Report, Institute of Geoscience, University of Tsukuba, 23, 63-68. OGAWA, Y. 1981. Tertiary tectonics of the Miura and Boso Peninsulas: ophiolite and the Izu forearc sediments trapped to the Honshu arc. Chikyu (The Earth Monthly), Tokyo, 3, 411-420 (in Japanese). OGAWA, Y. 1983. Mineoka ophiolite belt in the Izu forearc area—Neogene accretion of oceanic and island arc assemblages on the northeastern corner of the Philippine Sea Plate. In: HASHIMOTO, M. & UYEDA, S. (eds) Accretion Tectonics in the CircumPacific Regions. Terra, Tokyo, 245-260. OGAWA, Y. & TANIGUCHI, H. 1987. Ophiolitic melange in the forearc areas and the development of the Mineoka belt. Science Reports, Department of Geology, Kyushu University, 15, 1-23 (in Japanese with English abstract). OGAWA, Y. & TANIGUCHI, H. 1988. Geology and tectonics of the Miura-Boso Peninsulas and the adjacent area. Modern Geology, 12, 147-168. OGAWA, Y., HORIUCHI, K., TANIGUCHI, H. & NAKA, J. 1985. Collision of the Izu arc with Honshu and the effects of oblique subduction in the Miura-Boso Peninsulas. Tectonophysics, 119, 349-379. OKINO, K., SHIMAKAWA, Y. & NAGAOKA, S. 1994. Evolution of the Shikoku Basin. Journal of Geomagnetism and Geoelectricity, 46, 463-479. OKINO, K., OHARA, Y., KASUGA, S. & KATO, Y. 1999. The Philippine Sea: new survey results reveal the structure and the history of the marginal basin. Geophysical Research Letters, 26, 2287-2290. OTOFUJI, Y. & MATSUDA, T. 1983. Paleomagnetic evidence for the clockwise rotation of southeast Japan. Earth and Planetary Science Letters, 62, 349-359. OTSUKI, K. 1990. Westward migration of the Izu-Bonin Trench, northward motion of the Philippine Sea Plate, and their relationships to the Cenozoic tectonics of Japanese island arcs. Tectonophysics, 180,351-367. PEARCE, J.A. 1983. Role of the sub-continental lithosphere in magma genesis at active continental

MINEOKA OPHIOLITE, JAPAN margins. In: HAWKESWORTH, CJ. & NORRY, MJ. (eds) Continental Basalts and Mantle Xenoliths. Shiva, Nantwich, 230-249. PEARCE, J.A. & CANN, J.R. 1973. Tectonic setting of basic volcanic rocks determined using trace element analyses. Earth and Planetary Science Letters, 19, 290-300. PEARCE, J.A. & NORRY, MJ. 1979. Petrogenetic implications of Ti, Zr, Y and Nb variations in volcanic rocks. Contributions to Mineralogy and Petrology, 69, 33-47. PRINGLE, M.S. & DUNCAN, R.A. 1995. Radiometric ages of basement lavas recovered at Lo-En, Wodejebato, MIT, and Takuyo-Daisan guyots, ODP Leg 144, northwestern Pacific Ocean. In: DAGER, E.L., FIRTH, J.V. & SINTON, J.M. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 144. Ocean Drilling Program, College Station, TX, 547-557. REAGAN, M.K. & MEIJER, A. 1984. Geology and geochemistry of early arc-volcanic rocks from Guam. Geological Society of America Bulletin, 95,701 —713. SAITO, K. 1994. Excess Ar in some metamorphic and plutonic rocks reduction of thermal neutron-induced 40 Ar by Cd shielding. Science Report, Research Institute ofTohoku University, A40, 185-189. SAITO, K., OTOMO, I. & TAKAI, T. 1991. K-Ar dating of the Tanzawa tonalitic body and some restrictions on the collision tectonics in the south Fossa Magna, central Japan. Journal of Geomagnetism and Geoelectricity, 43, 921-935. SAITO, S. 1992. Stratigraphy of Cenozoic strata in the southern terminus area of Boso Peninsula, Central Japan. Contributions of Institute of Geology and Palaeontology, Tohoku University, 93, 1-37 (in Japanese with English abstract). SATO, H. & OGAWA, Y. 2000. Sulfide minerals as an indicator for petrogenesis and serpentinization of peridotites: an example from the Hayama-Mineoka Belt, central Japan. In: DILEK, Y., MOORES, G.M., ELTHON, D. ET AL. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 419-429. SATO, H., TANIGUCHI, H., TAKAHASHI, N., MOHIUDDIN, M.M., HIRANO, N. & OGAWA, Y. 1999. Origin of the Mineoka ophiolite. Journal of Geography, Tokyo Geographical Society, 108, 203-215 (Japanese with English abstract). SATO, H., MACHIDA, S., KANAYAMA, S., TANIGUCHI, H. & ISHII, T. 2002. Geochemical and isotopic characteristics of the Kinan Seamount Chain in the Shikoku Basin. Geochemical Journal, 36, 519-526. SENO, T. 1984. Was there a North New Guinea Plate? Geological Survey of Japan Report, 263, 29-42. SENO, T. & MARUYAMA, S. 1984. Paleogeographic reconstruction and origin of the Philippine Sea.

293

Tectonophysics, 102, 53-84. SENO, T., OGAWA, Y., TOKUYAMA, H., NISHIYAMA, E. & TAIRA, A. 1989. Tectonic evolution of the triple junction off central Honshu for the past 1 m.y. Tectonophysics, 160, 91-116. SUZUKI, Y., AKIBA, F. & KAMIYA, M. 1996. Latest Oligocene siliceous microfossils from Hota Group in southern Boso Peninsula, eastern Honshu, Japan. Journal of Geological Society of Japan, 102, 1068-1071 (in Japanese). TAKAHASHI, N. 1994. 'Alkali basalt-clastic rock sequence' in the west end of the Mineoka tectonic belt, Boso Peninsula, Japan. Journal of the Natural History Museum and Institute, Chiba, 3, 1 — 18 (in Japanese). TAKAHASHI, A., OGAWA, Y., OHTA, Y. & HIRANO, N. 2003. The nature of faulting and deformation in the Mineoka ophiolite, NW Pacific Rim. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 299-314. TAKIGAMI, Y., KANEOKA, I., Ism, T. & NAKAMURA, Y. 1989. 40Ar-39Ar ages of igneous rocks recovered from Daiichi-Kashima and Erimo Seamounts during the KAIKO project. Palaeogeography, Palaeoclimatology, Palaeoecology, 71, 71-81. TAZAKI, K. & INOMATA, M. 1980. Picrite basalts and tholeiitic basalts from Mineoka tectonic belt, Central Japan. Journal of Geological Society of Japan, 86, 653-671 (in Japanese with English abstract). TAZAKI, K., INOMATA, M. & TAZAKI, K. 1980. Umbers in pillow lava from the Mineoka tectonic belt, Boso Peninsula. Journal of Geological Society of Japan, 86, 413-416. TONOUCHI, S. 1981. Paleomagnetic and geotectonic investigation of ophiolite suites and related rocks occurring in the south central Honshu, Japan. PhD thesis, University of Tokyo. UCHIDA, T. & ARAI, S. 1978. Petrology of ultramafic rocks from the Boso Peninsula and the Miura Peninsula. Journal of the Geological Society of Japan, 84, 561-570. WATANABE, Y. & IIJIMA, A. 1989. Evolution of the Tertiary Setogawa-Kobotoke-Mineoka forearc basin in central Japan with emphasis on the lower Miocene terrigenous turbidite fills. Journal of Faculty of Science, University of Tokyo, Section II, 22, 53-88. WESSEL, P. & KROENKE, L.W. 1998. The geometric relationship between hot spots and seamounts: implications for Pacific hot spots. Earth and Planetary Science Letters, 158, 1-18. WINCHESTER, J.A. & FLOYD, P.A. 1976. Geochemical magma type discrimination: application to altered and metamorphosed basic igneous rocks. Earth and Planetary Science Letters, 28, 459-469.

294

N. HIRANO ETAL.

Appendix Figure Al is an index map of the sample localities, which are shown in detail in Figures A2-A5.

Fig. Al. Index map of the sample localities shown in Figures A2-A5, and the legend for lithology for these figures. (This geological map is the same as Figure 2.)

Fig. A2. Sample locality map around the Kamogawa area. BM09BT1, dolerite; BM23BT2, aphyric basalt (pillow lava); BM25BT3, cpx-basalt (pillow lava); BM591, pl-cpxbasalt (basaltic dyke in dolerite); BM10SY, aphyric basalt; BM28SNJ, hb-quartz diorite; BM14FG, hb-gabbro; BM30RS, hb-opx-gabbro.

296

N. HIRANO ETAL.

Fig. A3. Sample locality map around the Mineoka Sengen area along the logging road of the eastern Mineoka Hill. BM07SG1, aphyric basalt (pillow lava); BM08SG2, aphyric basalt (pillow lava); BM21SG3, aphyric basalt (pillow lava); BM22SG6, aphyric basalt (pillow lava); BM24SG4, dolerite; BM29SG5, aphyric basalt (pillow lava); BM32NSGM, picrite basalt (pillow lava); BM06HS, aphyric basalt (pillow lava).

MINEOKA OPHIOLITE, JAPAN

297

Fig. A4. Sample locality map around the Hegurinaka area in the central Mineoka Hill. BM04HG2, ol-cpx-basalt (sheeted lava flow); BM03HG1, cpx-ol-basalt (sheeted lava flow); BM05SG3, dolerite; BM31YD, hb-quartz diorite.

298

N. HIRANO ETAL.

Fig. A5. Sample locality map around the Sakuma area in the western tip of the Mineoka Hill. BM26SK, aphyric basalt (pillow lava).

The nature of faulting and deformation in the Mineoka ophiolite, NW Pacific Rim AKIKO TAKAHASHI 1 , YUJIRO OGAWA 2 , YASUFUMI OHTA 3 & NAOTO HIRANO 4 ' 5 1 Doctoral Program in Earth Evolution Sciences, Graduate School of Life and Environmental Sciences, University ofTsukuba, Tsukuba 305-8571, Japan 2 Institute of Geoscience, University ofTsukuba, Tsukuba 305-8571, Japan (e-mail: [email protected]) 3 'Doctoral Program in Geoscience, University ofTsukuba, Tsukuba 305-8571, Japan 4 Ocean Research Institute, University of Tokyo, Tokyo 164-8639, Japan 5 Present address: Department of Earth and Planetary Sciences, Tokyo Institute of Technology, Ookayama, Meguro, Tokyo 152-8551, Japan Abstract: A belt of disrupted ophiolitie rocks occurs on the Boso Peninsula (Japan), currently located north of the oblique subduction boundary between the Philippine Sea and North American Plates, under which the Pacific Plate has been subducting westwards. This ophiolitie belt (Mineoka Belt) is composed of mafic-ultramanc rocks together with Tertiary chert and limestone and island-arc volcaniclastic rocks. Our detailed structural studies in and around the basaltic rock bodies within the ophiolite reveal three phases of deformation. The first phase is further divided into three stages, all related to oblique normal faulting associated with extensional tectonics at or near a spreading axis. Fluid pressures appear to have fluctuated in association with faulting and veining during this phase. The second phase of deformation is characterized by thrust-related shear zones with a significant strike-slip component and is probably related to the final emplacement of the ophiolite by oblique subduction-obduction processes. The third and final phase of deformation affected not only the ophiolite but also later terrigenous and island-arc pyroclastic rocks. This deformation involved large-scale transpressional dextral slip on forearc sliver faults, which are still active today.

The Boso Peninsula is located north of the oblique subduction boundary between the Philippine Sea and North American Plates, where the Pacific Plate is subducting westwards (Ogawa et al. 1989) (Fig. 1). In the southern part of the peninsula, an assemblage of mafic and ultramafic rocks and accompanying Tertiary chert and limestone forms the Mineoka ophiolite. This belt marks the boundary between forearc basin deposits to the north and the Miocene-Pliocene accretionary prisms to the south (Ogawa & Taniguchi 1988; Saito 1992) (Fig. 1). Although the lithologies in the Mineoka Belt have been studied earlier (Kanehira 1976; Ogawa & Taniguchi 1987, 1988), structural relationships and their tectonic implications for the ophiolitie rocks are poorly constrained because all internal contacts are faulted. Geochemical data indicate that the basaltic rocks in the Mineoka Belt include mid-ocean ridge basalt (MORB), back-arc basin basalt (BABB), within-plate basalt (WPB) and island-arc tholeiite (IAT) (Ogawa &

Taniguchi 1987, 1988; Hirano et al 2003). Whole-rock Ar-Ar ages of the volcanic rocks (Hirano et al. 2003) indicate that most tholeiitic basalts in the ophiolite are Eocene in age, whereas some alkaline varieties are of Miocene age. In the Mineoka Belt, all contacts between the ophiolite and the adjacent Miocene sedimentary rocks are faulted, although the latter were originally deposited over the already deformed and veined ophiolitie rocks. Ophiolitie rocks are composed dominantly of serpentinized harzburgite with minor dunite (Sato & Ogawa 2000), and of mafic lavas of various ages and compositions (Ogawa & Taniguchi 1987, 1988; Hirano et al 2003). In some parts of the belt, the basaltic rocks are in contact with serpentinite along cataclastic shear zones, but in other parts they are in fault contact with relatively uncompacted volcaniclastic rocks and/or continent-derived sandstone and conglomerate of mid-Miocene age (Yoshida 1974; M. M. Mohiuddin, pers. comm.) (Fig. 2). Some of the

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 299-314. 0305-8719/037$ 15 © The Geological Society of London 2003.

300

A. TAKAHASHI ETAL.

Fig. 1. Index map of the Mineoka Belt and its surroundings (adapted from Ogawa & Taniguchi 1988). The study area is in the easternmost part of the Mineoka Belt.

sandstones contain serpentinite fragments (Ogawa 1983). According to Kaneoka et al. (1980), whole-rock Ar-Ar ages of basaltic rocks are 3040 Ma and 40-50 Ma. These ages are generally supported by new whole-rock Ar-Ar dating by Hirano et al. (2003), but some alkaline basalts were found to be younger, around 20 m.y. old (Hirano & Okuzawa 2002). Ophiolitic rocks in the Mineoka Belt are accompanied by a number of other lithologies. For example, hornblende schist is exposed in the Kamogawa Harbour area and is estimated to have formed at temperatures of 500-550 °C and pressures of 500 MPa with high oxygen fugacity (Ogo & Hiroi 1991). Another rock type is weakly deformed and altered andesitic pumice exposed on Kojima Island in Kamogawa Harbour and sporadically throughout the belt (Fig. 2). These pyroclastic deposits, 12 m.y. old as inferred from similar rocks exposed on the Miura Peninsula (Kanie & Asami 1995), originally were deposited unconformably on the deformed and altered ophiolitic rocks and were subsequently faulted against other lithologies (Figs 1 and 2). In this paper, we document the different stages of faulting and veining in the basaltic rocks, and the kinematics of faulting. The data were collected from two well-exposed, large bodies of basalts in

the southern part of Kamogawa Harbour, Boso Peninsula (Figs 1 and 2). These volcanic bodies contain MORB or BABB-type pillow lavas at Shinyashiki (Fig. 3) and on Benten Island (Fig. 4). The pillow lavas and dolerite dykes on Benten Island probably also contain some lAT-type basalts. Our study of the limited but well-exposed basaltic bodies in the Mineoka ophiolite offers a good opportunity for understanding the nature of shallow crustal deformation in an ophiolite and the related hydrothermal alteration during oceanic crust generation.

Fault analysis Shinyashiki outcrops Basaltic pillow lavas are widely distributed at Shinyashiki, forming an outcrop 300 m across (Fig. 3). The pillow lavas, intruded by several dolerite dykes, are interlayered with massive flows several metres thick in the middle of the stratigraphic sequence. Pillows are commonly several tens of centimetres long, rarely up to 1 m. The sequence dips moderately to steeply to the SE. We have divided the Shinyashiki pillow lava body into four small areas (Fig. 3), which we informally name 'Bench', Teninsulet', 'South Face' and

FAULTING AND DEFORMATION OF M I N E O K A OPHIOLITE

Fig. 2. General lithological map and cross-section of the study area around Kamogawa Harbour, Boso Peninsula, central Japan. Map modified from Yoshida (1974).

301

302

A. TAKAHASHI ETAL.

Fig. 3. Location map of the outcrops at Shinyashiki, south of Kamogawa Harbour. The strike and dip of pillow lavas are marked.

'South Yooka Beach'. Several stages of faulting, mineral precipitation and veining were identified in these areas. Precipitation of calcite and quartz predates the first-stage faulting, and is common chiefly at 'Bench' and 'South Face'. Spherical quartz aggregates, 3-5 mm across, occur in the centres of pillows, whereas calcite occurs mainly in inter-pillow spaces. The inter-pillow calcite has been deformed by shearing (Fig. 4). The intra-pillow quartz and inter-pillow calcite occur particularly around 'Bench', suggesting local hydrothermal activity before the first stage of faulting. The first stage of faulting is characterized by a foliated cataclasite, exposed at 'Peninsulet', 'South Yooka Beach' and 'South Face'. The second stage of faulting is associated with conjugate sets of zeolite veins parallel to the faults and is developed at 'Bench' and 'Peninsulet'. Faults of the third stage, represented by crushed calcite veins along dextral-oblique normal faults, are well exposed at 'Bench'. (See Table 1 for summary of the deformational events.)

Fig. 4. Outcrop photograph showing deformation of inter-pillow calcite (Cal) and intra-pillow quartz (Qtz) formed during the first stage of precipitation; 'Bench' at Shinyashiki.

First-stage faults. Faults of the first stage are characterized by discrete zones of foliated cataclasite, c. 5 cm wide, and rare calcite veins in 'Peninsulet' and 'South Yooka Beach' (Fig. 5). The fault planes generally strike east-west and dip steeply. Microscale crushed basalt fragments are elongated as 'fish' on the shear surfaces and show a preferred orientation under the microscope. Strong shear fabrics are common, repre-

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

303

Table 1. Phases and stages of faulting, veining and other events in basaltic rock bodies of the Mineoka Belt Stage First phase First Second Third

Strike and dip

Mode of deformation and veining

Microscopic observation

East-west, high angle NE-SW and NW-SE, high angle East-west, high angle

Wide shear zones of foliated cataclasite Conjugate faults with various zeolites and calcite veins

Strong shear zones of Riedel and P-shears

Dragged pillow with brecciated fault zone of crushed calcite vein

Breccias composed of fragments of various size

NW-SE, moderate dip

Dextral oblique thrust with strong shear zone without mineral veins

Microscopic and mesoscopic Riedel shears

Calcite filling in pores of euhedral zeolites

Second phase Third phase East-west, NW- Large-scale dextral strike-slip SE, high angle without mineral veins

Shear zone with serpentinite-bearing fault

Fig. 5. (a) Plan view of 'Peninsulet' to the east of 'Bench' at Shinyashiki, showing the first-stage shear zones, which are cut by conjugate sets of second-stage faults seen in (b) and (c). (d) Photograph of area shown in (c).

sented by Riedel shears, P-shears and, particularly, duplex-like structures (Fig. 6). In general, within foliated cataclasite, calcite veins occur along the Riedel and P-shears. Second-stage faults. The first-stage faults are cut by the second-stage faults with attitudes of N30°E

or N50°W and steep dips. These NE and NW fault systems, which contain zeolite veins, crosscut each other, suggesting that they were conjugate and contemporaneous. Fault-related vein minerals are dominantly calcite and zeolites, such as anaIcime, stilbite and natrolite, and the veining is very complex. Calcite and analcime veins cut the

304

A. TAKAHASHI ETAL.

Fig. 6. (a) Polished surface of a hand sample showing first-stage faults; (b) photomicrograph of the sample in (a). Photograph is 5 mm wide

shear zone of the 'X-shaped vein' at 'Bench' (Fig. 7), and they are not crushed. In contrast, a thick analcime vein intrudes the sheared part, separating the shear zone with a straight boundary. These relationships suggest the following sequence of events: (1) shearing; (2) stilbite veining; (3) shearing; (4) analcime veining (Fig. 8). Analcime veins occur together with calcite veins and both vein types have been deformed by extensive shearing.

Another example of faulting and veining is observed along a second-stage fault at 'Peninsulef. In thin section, an analcime vein is observed to crosscut the shear zone without any offset. The vein itself is offset by the shear zone and is pulverized. The analcime has wavy extinction, indicating that shearing continued during its growth (Fig. 9). Euhedral natrolite was precipitated perpendicular to the vein wall, followed by

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

305

Fig. 7. (a) Mutual cross-cutting relationship of the second-stage faults in thin section; (b) close-up from 'Peninsulet'. Shinyashiki. (See text for details.)

the precipitation of anhedral calcite in the centre of the vein (Fig. 9). A fault with a shear zone of several centimetres width is exposed within the pillow lava sequence at 'South Face'. This fault curves from northsouth to NW-SE and displays a normal sense of shear based on asymmetric folding of a stilbite vein (Fig. 10), which has been rotated and cataclastically deformed. These relationships suggest

the following sequence of events associated with the second stage of faulting at 'Peninsulet' and 'South Face': (1) zeolite (analcime, stilbite and natrolite) vein formation; (2) crushing; (3) calcite vein formation. Third-stage faulting. An east-west-striking fault at 'Bench' was formed in the last stage of faulting because it crosscuts and dislocates many other

306

A. TAKAHASHI ETAL.

Fig. 8. Photomicrograph (plane-polarized light) of a sample from a fault zone and associated veins, showing several stages of shearing and veining: (a) shearing with stilbite vein (Stb); (b) shearing with analcime vein (Anl) and calcite vein (Cal). The dislocation of the vein by later shearing should be noted.

Fig. 9. Photomicrograph showing analcime (Anl) vein crosscutting the sheared, pulverized fault zone rock.

faults formed in the previous stages. It dips steeply southward and is up to 1 m wide. Displacement of pillow lavas shows that this fault has both normal and dextral components of movement (Fig. 11). A cataclastic zone, several centimetres wide and filled with microbreccia, occurs along the fault plane.

Microscopic observations suggest that brecciation of the host rock continued after precipitation of calcite between the breccia fragments. Some breccias are composed of aggregates of rock and mineral fragments of various sizes (Fig. 12). This style of deformation is characteristic of cataclastic

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

Fig. 10. Sample of a second-stage fault zone (strike N20°W, dip 30°E) at 'South Face', Shinyashiki. Deformation of the zeolite (analcime and stilbite) vein with a transposed structure should be noted. The vergence of associated folding suggests normal faulting.

307

flow without recrystallization and suggests brittle faulting. Veins along these faults may post-date the third stage of faulting. Euhedral crystals of analcime, oriented perpendicular to the vein walls, form thick veins, which are crosscut by thin analcime veins parallel to the vein axis. The veining at 'Bench' records the following progression of events: (1) precipitation of analcime; (2) deposition of thin veins of analcime; (3) calcite precipitation, because calcite fills cavities in the euhedral analcime (Fig. 13). The majority of the faults are subvertical with NE strikes (Fig. 14a). In the case of second-stage faults the inferred o\ (maximum principal stress axis) is vertical (Fig. 14b), suggesting that the conjugate fault systems at Shinyashiki were associated mainly with extensional normal faulting.

Fig. 11. (a) Interpretive sketch, and (b) and (c) outcrop photographs of the eastern part of 'Bench' and Shinyashiki (view to the east). Deformation of the pillows indicates an oblique normal fault with a dextral strike-slip component, (c) Close-up of calcite brecciation along this fault.

308

A. TAKAHASHI ETAL.

Fig. 12. Photomicrograph of a third-stage fault, showing various scales of brecciation.

Zeolite veins were simultaneously developed along settings, are present on Benten Island (Fig. 15): the faults of the second stage. In contrast, the (1) dolerite dykes with minor basaltic dykes of faulting of the third stage deformed pre-existing MORE or BABB composition crop out on the calcite veins of the earlier stages. This indicates main island; (2) black pillow lavas of IAT compothat the second-stage faulting was penecontem- sition occur in the NE section; (3) red-grey pillow poraneous with veining, but that third-stage fault- lavas of N-MORB composition are present in the ing occurred after veining. western section (Hirano et al. 2003). Each segment is characterized by a specific style of faulting and veining. The dolerite has sheared Benten Island outcrops laumontite veins, whereas the IAT lavas are Three blocks of basalt, each with different chem- characterized by inter-pillow silica and celadonite istry representing potentially different tectonic with worm-like fossils. The N-MORB pillow lavas

Fig. 13. Photomicrograph of a third-stage fault, showing first analcime (Anil), second analcime (Anl2) and calcite (Cal) veins.

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

Fig. 14. Stereographic projections of second-stage faults at Shinyashiki. (a) Kamb contour diagram of all faults, (b) Three major directions of normal faults.

have calcite and zeolite veins with a radial habit perpendicular to the pillow bedding. The dolerite dykes strike N60° to 80°W and dip steeply to the south. There are two types of dolerite dykes. The first is coarse grained with rare chilled margins. These are intruded by a later group of fine-grained dykes, which occur as single dykes with chilled margins on both sides or as multiple dykes with chilled margins only on the SW side. Some basaltic dykes are intruded parallel or oblique to the dolerite dykes at various locations (Fig. 15). The dolerite dykes are strongly

309

sheared and display several stages of veining. These rocks are altered to a slightly higher degree than those in the Shinyashiki body and the other basalts at Benten Island, as shown by the abundance of laumontite veins. Cataclastic textures are common along two fault systems, one parallel to the dyke orientation (N60°W) and the other perpendicular to it. This is similar to the secondstage conjugate faults at Shinyashiki, except that at Shinyashiki laumontite veins are rarely associated with this generation of faulting. A Kamb contour diagram (Fig. 16) representing poles of the fault planes in the dolerite dykes shows that steeply dipping faults are dominant. Thus, the fault system in the Benten Island dolerite and the conjugate faults at Shinyashiki are both steeply dipping and have the same inferred direction of o\. Faults in the N-MORB pillows in the NW part of Benten Island strike N50°E and dip steeply. These faults are characterized by euhedral anaIcime on both sides of the faults, followed by laumontite as the second-stage vein fill, then natrolite veining, and finally anhedral calcite (Fig. 17). Stilbite veins with calcite are parallel to the faults. These five minerals, including four zeolite types, all formed along the faults and within the pillows, suggesting changes in fluid compositions with time. A large-scale boundary fault, called the 'Aranami Fault', occurs between the main dolerite dyke block and the black pillow lavas. This fault strikes N60°W and dips 50°SW (Fig. 18). A remarkable shear zone occurs along this fault, characterized by a lenticular (phacoidal) fabric and Riedel shears on both large and small scales (Figs 19 and 20). The faulting produced a layered mixture of red and green basalt or dolerite. RI -shear, P-shear and Y-shear, and particularly duplex structures, indicate that progressive shearing and dislocation occurred as a result of dextral-reverse faulting. Along each shear zone, small lenticular 'fish' structures also indicate oblique thrusting. In thin section each 'fish' is composed of calcite lenses, crushed together with basalt and dolerite fragments. In the Aranami fault zone, early calcite was deformed by the primary RI and secondary shears, showing apparent dextral movement (N60°W). The small, crushed calcite fragments are concentrated along the primary RI shear (Fig. 20).

Discussion In the Shinyashiki body, three main stages of faulting and associated deformation features are recognized: (1) foliated cataclastic shear zones with rare calcite veins; (2) a set of conjugate shear zones with various zeolite minerals and calcite

310

A. TAKAHASHI ETAL.

Fig. 15. Lithological map of Benten Island, Kamogawa Harbour, showing three basaltic bodies in fault contact with each other. The dolerite dykes and I AT pillow lavas occur along the 'Aranami Fault'. The strike and dip of the pillow lavas and dolerite dykes are shown. White bars in the dolerite indicate basaltic dykes.

Fig. 16. Stereographic projection of faults in dolerite dyke swarms, Benten Island.

along the faults; and deformation first-stage shear which produced

(3) variable scales of brecciation of pillows along the faults. The zones indicate strain hardening, the widespread cataclastic defor-

mation without significant veining. This interpretation suggests that they formed at greater depth than the other zones and at low pore-fluid pressures because of strong frictional sliding. Microscale Riedel shears and duplex structures with a dextral sense of movement are common in the shear zones in 'Peninsulet', whereas those at 'Yooka Beach' show sinistral movement. These shear zones formed under semi-brittle or brittle ductile transitional conditions. The second-stage faulting occurred under high pore-fluid pressures, because secondary minerals are common in faultparallel veins. In some cases, mineral precipitation and faulting or shearing occurred repeatedly. In the Mineoka ophiolite, the first phase of deformation consists of three stages and is represented by repeated faulting in various directions. Many of the first-stage faults are associated with shear zones and are characterized either by cataclastic deformation or by vein formation. Such a sequence of faulting may be attributed to oscillation between states of differential stress, one of frictional sliding with semi-brittle conditions under low pore-fluid pressures and the other under considerably higher pore-fluid pressures. Multiple stages of zeolite veining subparallel to the faults within the shear zones strongly suggest that fluids

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

311

Fig. 17. Photomicrograph of N-MORB pillow lava, Benten Island, showing several stages of veining: analcime (Anl) on both sides of the fault, laumontite (Lmt), natrolite (Ntr) and calcite (Cal).

Fig. 18. Outcrop sketch and an areal photograph of the 'Aranami Fault', Benten Island, showing Riedel shears and Pshears.

312

A. TAKAHASHI ETAL.

Fig. 19. Schematic diagram of deformation of the 'Aranami Fault', Benten Island.

of differing compositions percolated along the shear zones during and after their formation. The faults in the Mineoka basalts appear to be normal faults. However, the present attitude of the normal faults may not represent their original orientation. By rotating the basalt bedding to horizontal we can restore the faults to their presumed original attitude, because the vein formation and related faulting is believed to have occurred at an early stage, just after formation of the oceanic crust formation. This interpretation is supported by the following observations. (1) Most of the pillow lavas are layered and locally intercalated with massive sheet flows. They strike N40° to 60°E and dip moderately to steeply. All of the lobes are more steeply inclined to the south and SE than the sheet flows, indicating that the lavas flowed in this direction. The ropy surface of one sheet lava indicates its flow in the same direction. Steeply dipping, NW-striking dolerite and basalt dykes are cut by faults with the same orientation. (2) Several stages of veining and faulting are observed and the dyking, veining and faulting occurred nearly simultaneously, indicating that they were controlled by the same tectonic processes. Clearly, the later faults of the second and third phases, with thrust or strike-slip compo-

nents, are not associated with mineral veining. In addition, the pyroclastic rocks and terrigenous sediments seldom show significant veining and shearing. These features indicate that the veining and faulting was coeval with the volcanic activity and dyke intrusion, long before emplacement of the ophiolite on land. Because most of the pillow lavas and dolerite dykes at Shinyashiki and on Benten Island have MORB and BABB affinities, the deformation and veining probably occurred at a spreading axis, supporting the assumption that the deformation took place when the lava flows were still nearly horizontal. After rotating the deformed basalts to horizontal, the normal fault systems in Shinyashiki become strike-slip faults with an inferred o\ trending NW-SE. The dolerite dykes on Benten Island are presumed to be in their original orientation and thus no rotation is necessary for these rocks. The second phase of deformation, represented by the Aranami Fault on Benten Island, becomes an oblique thrust that separates the IAT pillow basalt from MORB or BABB dolerite dykes. The thrust component may reflect regional convergent tectonics, related to the emplacement of the ophiolite. Rocks of the Kojima Formation, which is composed of andesitic pumice fall

FAULTING AND DEFORMATION OF MINEOKA OPHIOLITE

313

Fig. 20. Riedel shears on various scales from (a) outcrop size to (b) thin-section size along the 'Aranami Fault', Benten Island.

deposits, are neither altered nor deformed. Glassy pumice is preserved and there is no strong shear fabric (Ogawa & Taniguchi 1988). These rocks, which are widely distributed in and around the Mineoka Belt, are thought to be mid-Miocene in age (Kanie & Asami 1995). Therefore, the faulting and veining of the basaltic rocks must have occurred before the mid-Miocene. The presence of middle Miocene serpentine-bearing sandstones also supports this inferred emplacement age of the ophiolite (Ogawa 1983). The Boso triple junction may have been situated off Boso since the Miocene (Seno & Maruyama 1984; Ogawa et al 1989). The Mineoka plate, composed mainly of ophiolitic lithologies now exposed in the Mineoka Belt (Ogawa & Taniguchi

1988; Sato et al. 1999; Sato & Ogawa 2000; Hirano et al. 2003), may have been close to the triple junction; it was then emplaced onto the Honshu island arc. Following its emplacement, the ophiolitic bodies and the overlying Kojima Formation were cut by local shear zones (Fig. 2). These faults represent the third phase of deformation, beginning after the ophiolite emplacement and possibly continuing to the present. At least some of the faults that border the Mineoka Belt are thus thought to be still active. We are grateful for discussions with Y. Dilek, G. Moore, D. Curewitz, T. Ishii, H. Sato, H. Taniguchi, A. L. Abdeldayem and N. Takahashi. Comments and revisions by three reviewers, Y. Dilek, J. Wakabayashi and F. Huot,

314

A. TAKAHASHI ETAL.

are gratefully acknowledged. Part of this study was supported by a Grant-in-Aid from the Japan Society for Promotion of Science (A-10304037).

References HIRANO, N. & OKUZAWA, K. 2002. Occurrence of the sandstone included in the alkali-basalt lava flow from the western Mineoka Belt, Boso Peninsula, Japan, and its tectonic significance. Journal of Geological Society of Japan, 108, 691-700. HIRANO, N., OGAWA, Y., SAITO, K., YOSHIDA, T., SATO, H. & TANIGUCHI, H. 2003. Multi-stage evolution of the Tertiary Mineoka ophiolite at Boso TTT triple junction in the NW Pacific as revealed by new geochemical and age constraints. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, 218, 279-298. KANEHIRA, K. 1976. Modes of occurrence of serpentinite and basalt in the Mineoka Belt, southern Boso Peninsula. In: KIZAKI, Y. ET AL. (eds) Studies on the eboundary Area between East and Westt Japan. Memoirs of Geological Society of Japan, 13,43-50. KANEOKA, I., TAKIGAMI, Y., TONOUCHI, S., FURUTA, T., NAKAMURA, Y. & HIRANO, M. 1980. Pre-Neogene volcanism in central Japan based on K-Ar and ArAr analyses. In: IAVCEI Symposium—Arc Volcanism, Tokyo and Hakone, Abstracts. Publisher, Town, 166. KANIE, Y. & ASAMI, S. 1995. Biostratigraphy of the Miocene Hayama Group in the Miura Peninsula. Report of Culture and Natural Treasures of Yokosuka City, 29, 13-17. OGAWA, Y. 1983. Mineoka ophiolite belt in the Izu forearc area—Neogene accretion of oceanic and island arc assemblages in the northeastern corner of the Philippine Sea plate. In: HASHIMOTO, M. & UYEDA, S. (eds) Accretion Tectonics in the CircumPacific Region. Terra, Tokyo, 245-260. OGAWA, Y. & TANIGUCHI, H. 1987. Ophiolitic melange in forearc area and formation of the Mineoka Belt. Science Reports, Kyushu University, Geology, 15,

1-23. OGAWA, Y. & TANIGUCHI, H. 1988. Geology and tectonics of the Miura-Boso Peninsulas and the adjacent area. Modern Geology, 12, 147-168. OGAWA, Y., NAKA, J. & TANIGUCHI, H. 1985. Ophiolitebased forearcs: a particular type of plate boundary. In: NASU, N., KOBAYASHI, K., KUSHIRO, I. & KAGAMI, H. (eds) Formation of Active Ocean Margins. Terra, Tokyo, 719-746. OGAWA, Y., SENO, T., TOKUYAMA, H., AKIYOSHI, H., FUJIOKA, K. & TANIGUCHI, H. 1989. Structure and development of the Sagami Trough and off-Boso triple junction. Tectonophysics, 160, 135-150. OGO, M. & HIROI, Y. 1991. Origin of various mineral assemblages of the Mineoka metamorphic rocks from Kamogawa, Boso Peninsula, Central Japan— with special reference to the effect of oxygen. Journal of Japanese Association of Mineralogists, Petrologists and Economic Geologists, 86, 226240. SAITO, S. 1992. Stratigraphy of Cenozoic strata in the southern terminus area of Boso Peninsula, Central Japan. Contributions of Institute of Geology and Paleontology, Tohoku University, 93, 1-37. SATO, H. & OGAWA, Y. 2000. Sulfide minerals in peridotites as tectonic indicators for genesis of ophiolitic rocks: example from peridotite in the Hayama-Mineoka Belt, central Japan. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Oceanic Drilling Program. Geological Society of America, Special Papers, 349, 522pp. SATO, H., TANIGUCHI, H., TAKAHASHI, N., MOHIUDDIN, M.M., HIRANO, N. & OGAWA, Y. 1999. Origin of the Mineoka Ophiolite. Journal of Geography, Tokyo Geographical Society, 108, 203-215. SENO, T. & MARUYAMA, S. 1984. Paleogeographic reconstruction and origin of Philippine Sea. Tectonophysics, 102, 53-84. YOSHIDA, Y. 1974. Geology of the eastern end of the Mineoka Mountains. Chishitsu News, 233, 30-36.

Oxygen isotope and chemical studies on the origin of large plagiogranite bodies in northern Oman, and their relationship to the overlying massive sulphide deposits DEBRA S. STAKES 1 & HUGH P. TAYLOR, JR 2 1

Monterey Bay Aquarium Research Institute, 7700 Sandholdt Road, Moss Landing, CA 95039-9644, USA (e-mail: [email protected]) 2 Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA Abstract: The extraordinarily well-preserved and well-exposed Semail ophiolite of northern Oman hosts several large plagiogranite intrusions in proximity to economic copper sulphide deposits of the Lasail mining district. A progression of isotopic, chemical and mineralogical transformations observed within the plagiogranites and high-level gabbros (HLG), and a comparison of these effects with those in the lowermost dykes of the immediately overlying sheeted dyke complex (SDC) tracks the evolution of hydrothermal fluids and the alteration of overlying dykes and pillow lavas during discharge of these fluids on the sea floor. The largest hydrothermal alteration aureoles, and the greatest extent of metamorphic veins and metasomatic replacement features, are found adjacent to the largest high-level plagiogranite bodies, beneath and adjacent to the major ore bodies in northern Oman. The ubiquitous presence of metamorphic actinolitic hornblende, sodic plagioclase, epidote and titanite in metabasalts within the high-temperature alteration zones points to the most likely mineralogical and structural controls on the development and evolution of the hydrothermal fluids. Depleted Cu contents of the adjacent crustal rocks and Cu enrichments above the plagiogranite intrusions demonstrate the redistribution of heavy metals adjacent to the complexes. Field relationships implicate the formation of both the epidosites and plagiogranites in the genesis of the ore deposits. An important process inferred from the field and geochemical data is the assimilation of previously hydrothermally altered basaltic and gabbroic country rocks by stoping into the magma chambers developed near the SDC-gabbro horizon in the ophiolite. We suggest that this process of combined assimilation-fractional crystallization, together with replenishment and recharge by injection and quenching of basaltic magma 'pillows' into these plagiogranite magma chambers (i.e. RAFC), plays a major role in the development of these composite intrusions.

The Semail ophiolite in northern Oman is the largest, best-preserved and best-exposed ophiolite complex in the world. Understanding the fourdimensional aspects of construction and hydrothermal alteration of oceanic crust has benefited from almost two decades of interdisciplinary field examinations throughout the length and breadth of this ophiolite by international consortia of scientists. This effort has documented a complex magmatic history for the ophiolite based on its formation at a fast-spreading mid-ocean, back-arc or suprasubduction zone ridge (Coleman & Hopson 1981; Pallister & Hopson 1981; Alabaster et al. 1982; Lippard et al. 1986; Nicolas et al. 1994). This magmatic history is broadly interpreted as a series of overlapping spreading centres, microplates, propagating rifts and younger ('off-axis') intrusions (Alabaster et al. 1982; Ernewein et al.

1988; Juteau et al. 1988; Nicolas et al 1988; Vetter & Stakes 1990; Boudier et al. 2000), accompanied by a pervasive set of overlapping sea-floor hydrothermal systems throughout the oceanic crust during its extended and complex formation (Gregory & Taylor 1981; Alabaster & Pearce 1985; Nehlig & Juteau 1988; Stakes & Taylor 1992; Nehlig et al. 1994; Juteau et al. 2000). Particularly noteworthy are the numerous, widely varying plagiogranite intrusions (e.g. Coleman & Peterman 1975) emplaced in the northern part of the Semail ophiolite. The purpose of this paper is to use field, mineralogical and stable isotope characteristics of the upper portion of the ocean crust exposed in the northern part of the Semail ophiolite (Fig. 1) to demonstrate that: (1) plagiogranites are the result of extreme open-system magmatic processes

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 315-351. 0305-8719/037$ 15 © The Geological Society of London 2003.

316

D. S. STAKES & H. P. TAYLOR, JR

Fig, 1. Regional geological map of the northern half of the Semail ophiolite in Oman showing areas of study described in the present work: Bayda, Suhaylah, Lasail and Assayab (indicated in pink, and shown in greater detail in Figs 2-4). This figure is modified after Stakes & Taylor (1992), based on geology from Lippard (1980) and Smewing (1981). The rectangles (yellow) labelled Ragmi (Rajmi), Hilti and Shafan are the areas described in detail by Stakes & Taylor (1992). Wadi Jizi hosts the major economic ore deposits of the Lasail Mining District. The ore deposits are within a fault bounded valley called the 'Alley'. Plagiogranites studied for this work include intrusive bodies near the Lasail mine and Assayab prospect, the Aarja-Bayda mines and Suhaylah village. Other data cited are from Wadi Jizi, Wadi Shafan and Musafiyah, located south of Shafan in Wadi Far. MS, mantle sequence; CG, layered gabbro; G, isotropic (high-level) gabbro; D, sheeted dyke complex; V, volcanic rocks; BS, basal allochthonous complex of the Hawasina thrust sheets.

ORIGIN OF PLAGIOGRANITES IN OMAN that include assimilation of partially melted, hydrothermally altered wall rocks and dykes, together with recharge by contemporaneous mafic dykes and limited fractional crystallization and extraction of evolved magmas (e.g. assimilationfractional crystallization (AFC): Taylor 1980; DePaolo 1981; and recharge-assimilation-fractional crystallization (RAFC): Bohrson & Spera 2001; Spera & Bohrson 2001); (2) there is a spatial and causal linkage between the large plagiogranite intrusions that act as shallow point sources of heat and metals and the formation of overlying economic massive sulphide deposits. The stratigraphic zone defined by the upperlevel gabbros, diorites, and smaller plagiogranites in the Oman ophiolite lies just beneath the lowermost part of the sheeted dyke complex (SDC); we propose that the larger plagiogranite bodies and their immediate country rocks in this zone are the principal source regions for the Cu-rich high-temperature hydrothermal fluids responsible for the deposition of the overlying large massive sulphide deposits. Although earlier studies have noted a 'probable chance' relationship between the large plagiogranite bodies and the massive sulphide deposits (Aldiss 1978), or even suggested that the thermal driver for Oman's largest massive sulphide deposit (the Lasail body) was a plagiogranite intrusion (Alabaster & Pearce 1985), our model suggests that the RAFC processes inherent in plagiogranite formation play a critical role in providing the energy and the fluids involved in massive sulphide deposition. This conclusion suggests that there may be some fundamental differences between the genesis of the large ore deposits in ophiolites and the smaller ore deposits characteristic of active mid-ocean ridges. In fact, the smallest ore deposit in Oman (Bayda, at 600 000-106 tons) is 10 times larger than the largest well-mapped mid-ocean ridge deposit (Endeavour High Rise deposit, at 30 000 tons). This has broad implications regarding any correlations of ophiolite ore-forming processes with sulphide deposition at modern ocean ridges.

Regional setting In the Semail ophiolite, there is semi-continuous exposure of oceanic crustal rocks along strike for c. 500 km. The upper portion of the crustal section displays excellent preservation and exposures in the northern half of the ophiolite, and the largest massive sulphide deposits are also in the northern half of the ophiolite in the Lasail mining district (Fig. 1): the Lasail, Bayda and Aarja deposits are all hosted within a NE-trending fault-bounded structure referred to as the 'Alley' by Smewing et al. (1977). The ore deposits are confined to the

317

boundary between the lowermost lavas (referred to as Geotimes, VI or Ml by Alabaster et al. (1982), Ernewein et al. (1988) and Beurrier et al. (1989), respectively) and the overlying lavas (referred to as Lasail, V2 or M2). Primary field relationships have been described by Smewing (1981), Alabaster et al. (1982) and Lippard et al. (1986). The magmatic events in Oman were not contemporaneous: several younger magmatic and hydrothermal events overprint older events (Alabaster et al. 1982;Beurrier et al. 1989; Stakes & Taylor 1992). In addition, these multiple events are of varying importance in different parts of the ophiolite. In particular, the northern part of the ophiolite shown in Figure 1 has been shown to have a very complex magmatic and hydrothermal history (see Stakes & Taylor 1992). Because each lithological or structural boundary may in fact be diachronous and complex (e.g. dykes from multiple magmatic sources), a variety of field, petrological and bulk chemical characteristics must be used to distinguish different magmatic and hydrothermal events. Much of the lower plutonic section of the ophiolite is rhythmically phase-layered, as a result of either gravitational settling from a magma chamber (Pallister & Hopson 1981; Browning 1984; Smewing et al. 1984), or alignment from viscous deformation of a mostly crystalline mush coupled with the underlying mantle movement (Nicolas et al. 1988, 1994, 2000), or both. Late wehrlitic sills, dykes and layers within the lower plutonic section have been variously interpreted as: (1) primary melts recharging a mid-ocean ridge magma chamber (Pallister & Hopson 1981; Smewing 1981); (2) younger intrusions from a more hydrated mantle source region (Pearce et al. 1981; Juteau et al. 1988; Boudier et al. 2000); (3) hybrid magmas derived from syntectic assimilation reactions (Bedard et al. 2000). Diachronous portions of the plutonic complex can be identified by juxtaposed layered and isotropic gabbros at random depths, and by interfingered lithologies with mid-ocean ridge basalt (MORB) and wehrlitic compositions (e.g. Juteau et al. 1988; Reuber 1988; Boudier et al. 2000). Small plagiogranite bodies can be present within these deeper isotropic gabbros (e.g. as at Assayab; see Fig. 7, below). In this paper, these deeper gabbroic intrusions are referred to as 'deep isotropic gabbros' (DIG). Where preserved or inferred from pseudomorphs, the DIG primary mineralogy includes sulphides, orthopyroxene, calcic plagioclase and diopsidic augite, similar to the gabbronorites described by Juteau et al. (1988) and Boudier et al. (2000). The high-level gabbro (HLG) of the upper 100-500 m of the gabbroic sequence is structurally distinct, and this upper plutonic section is

318

D. S. STAKES & H. P. TAYLOR, JR

also extraordinarily texturally and mineralogically heterogeneous; it includes a variety of variably deformed and recrystallized dykes and sills from microgabbro to hornblende pegmatite and plagiogranite. In areas that are relatively unperturbed by late intrusions or major shear zones (e.g. Wadi Hilti; see Stakes & Taylor 1992), the upper gabbro is a faintly foliated clinopyroxene gabbro containing small, coarser-grained felsic patches. More commonly, however, the HLG unit varies greatly in size and shape, and includes truncated, rotated and boudinaged blocks of layered gabbro and isotropic gabbro that are crosscut by more silicic segregation pods and veins. Slightly larger felsic segregations are observed as small (0.5-5.Om) dykes that intrude upward into the base of the SDC and crosscut the gabbro-SDC boundary. When this fabric is fully developed, the outcrop can be characterized as a magmatic breccia (i.e. an agmatite; terminology after Pitcher 1993) containing sharp-bounded mafic blocks within a dioritic to aplitic matrix and reaction zones containing amphibole and Fe-Ti oxides. Such xenoliths can be relatively unaltered sodic plagioclase-calcic pyroxene hornfels, or they can be altered to albite-amphibole or epidosite assemblages (epidote-quartz-titanite ± albite). Quartz veinlets and porphyroblasts and quartz-epidote symplectite are conspicuous replacement textures. The largest plagiogranite-isotropic gabbro intrusions described here range from 0.5 to 8 km in diameter and are found in the Alley at Suhaylah and Lasail, west and SW, respectively, of the largest massive sulphide deposit in Oman (the Lasail mine, see Figs 1, 2, 4 and 6k). Suhaylah is the statigraphically deeper intrusion with a lower contact well within the cumulate gabbros. The Suhaylah body encloses partially recrystallized xenoliths with relict layered fabric as well as basaltic enclaves or 'pillows' (see Fig. 6b). The Lasail plagiogranite body occupies a stratigraphically higher position and extends much higher in the ophiolite stratigraphy, intruding the base of the lower volcanic sequence. The smaller plagiogranite-HLG body near the Aarja and Bayda mines (Fig. 3) is less than half the size of the Suhaylah or Lasail bodies and intrudes the sheeted dyke complex on all exposed margins.

Analytical techniques Samples were examined petrographically and results are provided in Table 1 (Lasail and Assayab), Table 2 (Bayda and Aarja) and Table 3 (Suhaylah; see Fig. 4). These tables summarize petrography, oxygen isotope compositions and mineral chemistry. A total of 140 samples were analysed for bulk 18 O/16O, supplemented by analyses of mineral

separates from 35 of these samples. The data are reported relative to SMOW (Standard Mean Ocean Water). Isotopic analyses of bulk rocks were carried out at either Caltech or the University of South Carolina using standard techniques and using Rose Quartz as an internal standard. Additional 18O/16O analyses of quartz and plagioclase separates were carried out at the University of Michigan. Mineral chemistry was determined on 26 samples using the electron microprobe facilities at the University of California, Davis. These data are included in Tables 1-3. Bulk chemical analyses of 109 samples were carried out at the Geological Survey of Canada. Major and trace elements analyses were performed by X-ray fluorescence (XRF), inductively coupled plasma mass spectrometry (ICP-MS) and ICPemission spectrometry (ICP-ES) using standard procedures. FeO, total water, carbon dioxide and loss on ignition (LOI) were obtained by standard chemical methods. Additional trace element data are available on request, and relevant portions are summarized in Tables 4 and 5. Samples for geochemistry include regional samples as well as those studied for mineralogy and stable isotopes. The regional samples from Musafiyah, Wadi Shafan and Wadi Ragmi been described by Stakes & Taylor (1992). Additional samples were collected from Wadi Fizh, SE of Wadi Ragmi. The Nd contents of samples with low total rare earth elements (REE) and high Sr content analysed by ICP-ES are abnormally high and in error as a result of interference from Sr. Either critical samples were reanalysed by ICP-MS or the Nd concentration is noted as analytically problematic.

General geology and petrology of the gabbro-SDC contact and the plagiogranites The structural relationships observed in the ophiolite are diagrammatically illustrated in Figure 5. The SDC is a continuous series of mostly contemporaneous, hydrothermally altered dykes that formed during the main ophiolite construction (Geotimes/Vl/Ml). Younger crosscutting dykes illustrated schematically in Figure 5 are associated with the waning phases of crustal construction or with a secondary (off-axis?) magmatic event (either Lasail/V2/M2 or Alley/V3/M3). Three types of gabbro-dyke transitions are illustrated in Figure 5: from left to right, these include (a) disrupted masses of variegated gabbro; (b) small plagiogranite segregation dykes; (c) large (10m10 km in scale) plagiogranite intrusions (these are commonly internally zoned to gabbro, diorite, and wehrlite). Types (a) and (b) have been described

Fig. 2. Simplified geological map of the Lasail-Assayab area (modified after Lippard (1980); see Fig. 1 for location) showing sample localities listed in Table 1. The Lasail mine is located c. 2 km NE of the plagiogranite (see Fig. 6k).

Fig. 3. Simplified geological map of the Aarja-Bayda area showing sample localities listed in Table 2 (modified after Lippard (1980); see Fig. 1). The copper mines at Aarja and Bayda are shown by the small black squares. The plagiogranite body is bisected by Wadi Bani Umar Gharbi, where good exposures of multiple generations of dykes can be shown to predate, be contemporaneous with and postdate the intrusion. The upper contact of the plagiogranite grades into a stockworks complex that includes pervasively altered pillow lavas and sheeted dykes with abundant epidote and chlorite. Gabbroic rocks along the margins of the plagiogranite are also extensively altered.

ORIGIN OF PLAGIOGRANITES IN OMAN

321

Table 1. Oxygen isotopic and petrogmphic data for samples collected from map localities 267-316 in the Lasail region (Assayab and Huwayl); see Figure 2 Map

Sample

<318O value (%o)

Rock type and petrography

Geological relations and mineral compositions

locality

WR 267

83-15

BAS, q, z, sm, ap, eel

15.6 11.7 11.2 11.6 8.0

83-19

AND, q, z, eel, ap, hm

83-20 83-21 83-23 83-24 83-26 83-31 83-171 83- 1 67

PLOT, q, f, ep, sulph, hm, chl, ap, sph BAS, q, ep, sulph, chl, mt, f AND, f, amph, chl, sulph, q, mt, ep BAS, f, cpx BAS, cpx, ol BAS, f, cpx [sulph, q] GB, f, cpx, hb, mt, amph XENO, q, amph, f, mt, hb, sph, d?

83-168 83-169 83-70 83-71 83-7 Ib 83-72*

IG, f, amph, mt, ep, ab, chl, q HLG, f, q, amph, hb, sph, mt PLOT, q, amph, sulph, sph, ap, mt, ep PLOT, q, f, amph, mt, sph XENOJ, q, amph, mt, sph, ep (83-71) PLOT, ep, hb, amph, sph, q, mt, ap, f

83-73 83-166 83-36 83-37 85-444*

[Y q, ep, sulph] XENO, q,f, ep, amph, mt, sph, di PLOT, q, ab, ep, act, mt BAS (83-36), f, q, amph, sm, ol?, sulph XENO, q,f, act, cpx, hb, sph, di

83-32* 83-33*

[V, q, ep, chl, amph, sulph, sph] [V, amph, sulph, chl]

83-34 83-35

AND, f, op, pr, act, sm [83-32] BAS, ep, hm, chl, ab, sph, cc, q

283

85-445 A 85-445B 83-39 83-41* 83-42 83-43 83-163*

XENOJ, q, act, chl, sph PLOT, q, f, op, act, chl, di PLOT, q, f, hb, sulph, sph PLOT, chl, mt, ep, f, q, sulph, sph XENO, chl, mt, q, ab?, sph, ep (83-41) BAS, ab, act, chl, sulph?, sm, hm BAS, ep, sph, chl, mt, q

6.4 7.5 9.2 6.5 10.9 6.7 6.9

284 285

83-44 83-78

AND EP-DK, ep, q, chl, ab, z

8.4 10.6

286

83-142 83-143 83-144 83-141 83-146 83-147

(DK, ab, act, chl, sulph?, hm) [ep] [V, ep, sulph] (DK, ab, q, ep, chl, sph) (DK, ol, hb, ep, sph, [pr]) BR, q, chl (DK, chl, sm, q, ep, sulph)

83-148

HLG, f, hb, act, cpx, op, ep

268 269 270 271

274 275 276

111

FLAG

13.4

5.7

7.5 6.2 5.5 9.2

PX AMPH QTZ OTHER

6.8

5.7

9.4 8.4

8.6

9.3

6.5 6.4 7.1

6.7

Massive flow mapped as [Alley' rhyolite; altered Massive flow in Wadi Lasail; altered Felsite stock in Wadi Lasail; sm replaces chl Adjacent to felsite stock Chlorite veins; pyrite crystals; vesicular Uppermost unaltered massive flows Cpx-phyric pillow lava; overlies massive flows Lasail pillow lavas; q and sulph in vesicles Cpx replaced by green hb; core of PLGT Dark XENO in PLGT margin; q megacrysts + relict Ca-plag Contact between PLGT and IG Q megacrysts; granophyre Sulph-rich XENO-rich part of intrusion Coarse-grained mafic xenolith Ep-rich (Ps]6-28); XENO-poor; act hb; An52~

36; An4 278 279

280

281 282

287 288 289 290

83-149* HLG, f, q, amph, ep, ap, ab, hb

293 294 295 296 300

301 302

303

304 305

[9.2]

8.3 6.3 4.2 4.9

8.0 6.2

9.9

8.3 10.1

(11.0)

(7.5) (6.8) 6.5 (2.0)

7.8

83-150 85-53 1 85-532 85-533 83-82 83-83 85-489 85-490*

(GB-DK, f, hb, cpx, opx, amph, chl) PLOT, f, amph, q, mt, pr [di?] GB, f, amph, hb (DK, f, amph, ep)[cc] BAS, ep, q, chl, act [V,jp,q] LGB, f, ol.cpx, amph LGB, f, ol, cpx, act, srp, chl, pr

(6.4)

85-492 85-494 85-495 85-496 85-497 85-475*

LBG, f, ol, cpx, amph LGB, f, cpx (DK, f, amph, sph, [pr]) DIG (85-495) LGB, f, hb (85-495) PEG, f, cpx, hb, act, chl, tc, mt

5.4 4.6 (3.6) 6.2 5.1 5.0

85-502 85-503 85-478 85-504 85-505 85-510 85-512

(DK, f, hb, ep, sph)[ep, sulph] LGB, f, cpx GB[q, sulph, ep] GB, hb, f, ep[ep, sulph] LGB DIG XENO, f, hb, pr, amph, chl

(6.8) 5.4 6.0 8.1 6.8 8.4

6.3 4.7

6.9

7.0 6.0

Top of main PLGT body Dark XENO in margin of PLGT 7.6 Central part of intrusion; sharp contact Sharp contact with PLGT (83-36) 7.5 Deformed; in PLGT 12 ft X5 ft; An75-8o; An6-65 [8.2] [3.6 chl] Near margin of Lasail PLGT; Ps32 -36 Chloritized basalt fragments(?) in sheared vein (83-32) Lasail andesite sheet; skeletal f + op Dark red pillows with green margins; vesicular Included in margin of PLGT; agmatite 7.5 Matrix for agmatite 7.3 Fine-grained; few mafic clots XENO-rich margin; chl-mt clots Chl-mt clots Sheared contact with PLGT, vesicular Altered zones in massive flow; Ps2o-27; Fechl; all f to q 9.9 vesicle Andesite massive flow Greyish green sheet with ep clots; ol pseudomorphs From fault with mineralization From mineralized fault SDC cut by mineralized fault Late mafic dyke Mineralized fault at SDC -HLG Adjacent to mineralized fault at SDC-HLG boundary Within mineralized shear zone; secondary plag From outside of faulted region; Ani -4; Clapatite; Ps2o—2 ?; An 2 i— 42 Cuts HLG; late Diorite; less abundant xenoliths Rare peg zones; no xenoliths SDC; north-south trending Epidotized pillow near ep-jasper; vesicular [8.1] Cuts jasper- qtz 'umber' Lowermost layered gabbro in peridotite Offset by silicified faults; An84-g8; An8! ; Fo87 35 m upsection from 490 80 m upsection from 492 Late mafic dyke Clots of hornblende; intrudes? 85-497 Weakly phase layered Parallels cumulate layers; ol or opx to tc/mt or tc/act/chl; An79-84; di; pg hb, mg hb N30W; associated with hb-sulph veins Faintly banded Massive ep veins Good phase layering Intruded by PLGT; zebra outcrop Zebra outcrop, blocks in PLGT (continued overleaf)

D. S. STAKES & H. P. TAYLOR, JR

322

Table 1. (continued) Map Sample locality

<518O value (%o)

Rock type and petrography WR

306

85-524* DIG, f, px, act hb, zeo, ab, pr

FLAG

Geological relations and mineral compositions

PX AMPH QTZ OTHER

10.8 6.1 (milky)

3.0ab

Diorite; foliated, px-hb myrmekite, cut by Cu-sulphide veins

4.5 (clear) 307

85-474* DIG [f, mg hb, act hb, di, tr-act, sph, pr]

6.1

308

85-508 85-509 85-506 85-470*

LGB DIG DIG, hb, f LG, f, mg hb act hb, act cpx, di, pr, sph (85-471) 85-471 (GB-DK, f, op, amph, ep, cpx, chl, pr) 85-468* XENO,f, hb, cpx, act; act, pr, chl 85-515 (DK, f, hb, amph, mt) [pr] 85-457 PLOT, f, pr, q, act, chl, ap, ep, sph

9.2 6.8

309 310

311 312

313

314 315

316

317 318

8.6

(6.6) 6.5 5.6 13.4

85-516 85-517 85-518 85-519a 85-459 85-536 85-537 85-461

(DK) (DK, amph, f) (DK, act, f, mt, pr, ep)[cc] PLOT, f, ep, hb, q GB DIG (85-537) (DK, mt, ab, amph, sph) (DK, f, amph)[pr]

85-460 85-538

PLGT (85-461) q, f, ep, act, op, splti IG, cpx, f, chl, pr, amph, ol? (85-539)

10.6 11.7

85-539 85-540 85-462 85-463 85-464

(DK, f, act, cpx, chl, ep) (DK) (DK, f, act, sph, op) (DK, f, act, op, chl) (DK, f, cpx, chl, pr, ep, op)

(6.5) (8.5) 8.1 (6.6) (8.8)

15.5

(10.6) (6.8) (5.7) 11.3 6.6 9.5 (7.0) (7.7)

13.4

6.8

6.2

7.4

85-465* HLG (85-464) ab.ep, sph, pr, q, 320 321 322 323 324 325 326 327 328

85-466 85-439 85-441 85-525 85-535 85-443 85-528 85-529 85-485 85-487 85-481 85-482

(DK, f, act, pr, op) (DK)[ep, q] (DK, ab, amph, ep, chl, sulph, q) (DK, sm, f, mt, cpx) (DK, ab, chl, amph) [ep, pr] (DK)[q, ep, sulph] PLGT, f, amph, q, cc (DK, ep, f, q, chl) P, ol, cpx, f, hb LG, f, ol, cpx, hb, pr GB, f, pr, cpx, ol (GB-DK, ol, cpx, opx, amph, sm)

85-483* LGB, cpx, di, gt, ep 329 330

18

85-521 85-522 85-599

(GB-DK, ol, cpx, amph, f) (GB-DK, ol, anth, chl, px, srp, op) (GB-DK, hb, chl, gt?)

85-598

P, ol, anth, op, chl, serp, f

(6.2) (6.8) (7.9) (9.8) (8.8) (9.1) 6.7 (7.3) 4.6 6.2 6.1 (7.8) 3.8

(9.7) (6.8) 4.0

Early hb veins, later q-ep; last q veins; FeCu sulphide; px to tr; strongly zoned plag, An22-g2 10 m from zebra outcrop and 509 Discontinuous layers; 10m from 508 Amphibole-rich with felsic patches Ango-yoi f deformed with neoblasts Anso-so; sodic rims Ani -8 Orthogonal to cumulate layer; zoned In HLG; act hb; mg hb; An30-7o SDC; coarsed-grained Thin sheet with 2 inch ep spherules; intrudes dykes SDC intrudes gabbro-diorite SDC; adjacent to 518 and 519; fine-grained SDC; adjacent to 517 & 519 Adjacent to 85-518 Contains PEG layers; adjacent to PLGT Hb aggregates; massive ep on fractures Fine-grained; ep'd margins Trends north—south; LMD; adj to ep-sulph veins At dyke-gabbro boundary; myrmekite Cut by late mafic dyke, faint relict layers; skeletal amphibole LMD LMD cuts PLGT, N20E SDC; trends north-south SDC; fine-grained margin of 10 ft wide dyke SDC; coarse-grained interior of 10 ft wide dyke Intrudes dykes; Ps24-32; all f to ab; cpx/ilm to di/sph to pr/sph Near PLGT and HLG Cuts HLG Outcrop cut by Q network; gabbro screens Intrudes GB R-19; SDC Late dyke(?) cuts meta pillows R-12 diorite; at contact with SDC R-13; dismembered dyke in diorite Feldspathic wehrlite Interlayered with wehrlite Gabbro lenses in wehrlite Gabbro pegmatite dyke in peridotite; f completely altered Sheared gabbro with faulted margins; di in veins Composite dyke cuts wehrlite (= R5) Composite dykes cut wehrlite (= R6) Gabbro dyke in mantle peridotite; rodingized (=R1) (=R2)

(3 O values: WR, whole rock; PLAG, plagioclase; PX, clinopyroxene; AMPH, amphibolite; QTZ, quartz; OTHER, mineral indicated by abbreviation. Rocks: GB, cumulate gabbro; DK, dyke; PEG, gabbroic pegmatite; PLGT, plagiogranite; BR, breccia; BAS, pillow basalt; V, vein; XENO, xenolith; HLG, high-level gabbro; LGB, conspicuously layered cumulate gabbro; SDC, sheeted dyke complex; AND, andesite; IG, isotropic gabbro. Prefixes: HB, hornblende; S, sheared, strongly foliated or deformed rock collected from a welldefined shear zone; EP, strongly epidotized, e.g. epidosite. Minerals: f, plagioclase (An, anorthite content); opx, orthopyroxene; cpx, clinopyroxene; ol, olivine; pr, prehnite; chl, chlorite; sm, smectite; sph, sphene; di, diopside; ct, cummingtonite; anth, anthophyllite; act, actinolite; hb, hornblende; act hb, actinolitic hornblende; pg hb, pargasitic hornblende; amph, green amphibole (fibrous or acicular); trem, tremolite; sulph, sulphides (pyrite, chalcopyrite, pyrrhotite); srp, serpentine; ep, epidote (Ps, pistacite content, Fe/(Fe + Al)); tc, talc; mt, magnetite; q, quartz; ab, albite; ap, apatite; jp, jasper; hm, hematite; gt, hydrogrossular garnet. Rock names and minerals enclosed in brackets, e.g. [V, q, ep, sulph], represent materials from crosscutting veins; if brackets contain a sample number, e.g. [81-74], it indicates that the rock at that locality is cut by the vein 81-74; if brackets contain a (518O value, e.g. [7.4], it indicates that the mineral or whole-rock 618O analysis is from vein material. Rock names and minerals enclosed in parentheses, e.g. (DK, f, amph, mt), represent a crosscutting intrusive body, usually a dyke; if the parentheses contain a sample number, e.g. (81-50), it indicates that the rock at that locality is cut by the dyke 81-50; if the parentheses contain a (518O value, it indicates a whole-rock 18O/ 16 O analysis of the dyke. Xenoliths and xenolith 18O/16O analyses are indicated by italics. * Samples for which electron microprobe compositional data were obtained.

ORIGIN OF PLAGIOGRANITES IN OMAN

323

Table 2. Oxygen isotopic and petrographic data for samples collected from map localities 200—232 in the AarjaBayda region (see Fig. 3) Map Sample locality

<518O value (%o)

Rock type and petrography WR

200

81-19 81-20

(DK, act, mt, sph, pr, ab, chl) HLG, f, cpx, opx, amph, hb (81-19)

(6.9) 4.3

201

81-21 81-22 81-23 81-24 83-63b 83-65*

(DK, chl, pr, ep, amph) [pr] GB, f, cpx, amph, hb, mt, chl [pr] (81-21) PEG, f, cpx, hb, chl, opx, amph [pr] PEG, hb, f, ep, amph, cpx, sulph [pr] [V, q, sulph] [V, q, ep, sulph]

(4.3)

83-66

BAS, chl, f, q, ep [83-63b, 83-65]

202 203

204 205

83-67 BAS, q, sulph, sm 83-100 [Y chl, q, sulph, cc, sm] 83-102* P, ol, cpx, opx, hb, bt, sm, tc, mt

[9.9] [11.0] 8.9

12.5 7.2

83-104* [V, ep, q, jp, z] 90-656* PLGT, q, ep, chl, sph, ilm

207

83-106* 83-107 83-108* 83-109* 83-110 83-113 83-114

XENO,f, amph, q, mt, sph XENO, q, ep, chl, amph, sh PLGT, sph, sulph, ep, f, amph, q, pr? PLGT, q, f, ep, mt, chl, sulph?, ap [V, ep, q, hm] PLGT, ep, q, mt, cc [cc] BAS, chl, q, ep, ab, mt, sulph? (83-218)

83-128 83-129

XENO,f, chl, amph [ep, q, sulph, sph, act, chl] [Y chl, ep, sulph, q, amph]

83-130 83-132 83-133

XENO,f, ep, sph, mt, hb, sulph?, ab? BAS, sulph, q [Y q, ep, sulph, sph, amph]

214 215

83-136 83-137 83-138 83-151 83-157

(EP-DK, ep, amph, q, f, sph, chl, pr?) (EP-DK, ep, q, sulph, chl, sph, hm) [Y ep, q, sulph, hm, chl] (DK, f, hb, chl, act, sph, pr, mt) (DK, f, ep, sulph, q, chl)

83-158 83-160 83-161

(PLGT-DK, ep, sulph?, hm, sm, f, q) [cc] (PLGT-DK, q, ab, amph, cc, shl, sph, mt) (PLGT-DK, chl, act, q, ep, sulph?, f)

(13.4)

216 217 218

83-162

(DK, chl, q, f, sulph, sm)

(18.5)

219 220 221 222 224

83-172 83-173* 83-176* 83-182 83-183*

(PLGT-DK, f, ep, chl?, mt) PLGT, ep, sph, chl, ab, , mt, q XENO, ep, q, sph, di BAS, f, q, sm, cc, hm, eel? HHLG, f, hb, cpx, q, ep, amph, chl [83184](83-185) [Y sulph, q, ep, amph, sm] (DK, ep, q, sulph, chl, sph) (DK, ep, q, sph, sulph, amph) (DK, amph, q, ep, chl) XENO, chl, ep, ab, sulph, [sulph]

208 209

210

211

212 213

83-184 83-185 83-186 83-187

4.3 6.5 6.9

[6.2]

[6.9]

5.1 [9.0] 9.2 8.4

(8.0) (9.8)

(6.0) (8.9)

(8.3) 10.9 7.6

13.2

4.3

3.8

(6.6) (5.8) (7.4) 7.4

83-189

(DK, f, amph, mt, sph) [pr]

(6.9)

83-190

XENO, q, ep,f, amph, hb, mt, sph, q

10.0

85-403

(DK, f, cpx, hb, amph, mt, sph)

(4.5)

227

85-405 85-409

(DK, f, hb, chl, amph, mt, sph) (DK, ab, ep, chl, sulph, pr) [ep]

(5.8) (7.2)

85-410

HLG, f, hb, amph, ep, q, chl, sph, mt (85409) XENO,f, q, mt, amph

85-411

6.4

[5.6]

225

229

FLAG PX AMPH QTZ OTHER

5.6 4.7

206

8.3 5.0

Geological relations and mineral compositions

7.1

3.6

Late mafic dyke with gabbro screen; Rahab Gabbro cut by mafic dyke; Rahab; px oikocrysts Dyke that parallels layers in gabbro; Rahab Gabbro with vertical layers PEG segregations or dyke; plastic deformation PEG in HLG; Rahab Thick vein between massive flows; hematized Vug fillings in chloritized basalt; north of Bayda mine North of Bayda mine; vesicular; plag microphenocrysts Heavily veined region adjacent to Bayda mine South of Aarja gossan; cuts lavas Wehrlite; ol > cpx > opx; near ramp into Bayda mine Ps3o-2s; stilbite; adjacent to late dyke Spindly epidote and q— ab myrmekite coronas on plag laths; Anj -2? ; Ps28 -34 From net-veined PLGT; qtz clots From net- veined PLGT; qtz rim Mafic-rich with XENOs West side of PLGT intrusion; ep-rich Cuts PLGT Ep-rich contact with pillow lavas Meta-pillow basalt with q + ep in quench margins Margin of PLGT; XENO was originally basalt XENO-rich margin of PLGT; sulphides are oxidized Mafic xenolith, very dark; originally gabbro Massive flow at PLGT-GB contact From margin of PLGT intrusion; contains altered XENO Epidosite sheet with pillow basalts Epidotized sheet; near felsite dyke From ep-rich horizon; north of PLGT Dykes (late?) with pillow screens Late dyke cuts fragmented pillows; cpx and f pseudomorphs Grey felsite sheets; strongly oxidized Felsite intruded into pillow basalt screen Grey felsite sheet intrudes HLG at base of PLGT Diabase sill cuts lower lavas; cpx pseudomorphs East margin of PLGT; felsite sheet Ep-rich region; f is altered; An2 — 6 ; Ps2s Ep-spherule from north margin of PLGT Altered pillow margin Primary An88-89; altered An 9 i- 56 ; secondary An20 Cuts pegmatoidal gabbro Deformed dyke near west margin of PLGT Siliceous deformed dyke, centre (coarse) Siliceous deformed dyke, margin (fine) Dark blocks in gabbro; epidotized margins; relict igneous Deformed dyke at PLGT/GB margin; protolith for 83- 187 Partially resorbed in PLGT, q megacrysts + poikiloblasts Wide early dyke adjacent to PLGT; f + cpx phenos Late mafic dyke near PLGT; no phenocrysts Late mafic dyke; epidotized; relict igneous texture GB pod, adjacent to late ep sheets; large skeletal mt In GB margin of PLaGT; relict igneous texture; q poikiloblasts (continued overleaf )

D. S. STAKES & H. P. TAYLOR, JR

324

Table 2. (continued) Map Sample locality

618O value (%o)

Rock type and petrography WR

85-412

231 232

233 234 235

PLOT, q, ab, ep, op

9.8

85-413 (DK, f, amph, sph, q) 85-413A (DK, q, act, f, chl) [q, ep] 85-417* (DK, act hb, sph, ab, ep, chl, q)

(7.2) (8.6) 8.6

85-418 (DK, amph, f, q, chl, op) 85-419 (DK, amph, q, chl, f) 85-434* HHLG, f, cpx, act, act hb, ep, q

9.3 (9.1) 2.9

85-435 90-619 90-620 90-616

(DK, f, q, act, sph) DK DK (DK, ep, ab, mt, chl)

12.7

90-623 90-638

AND BAS

Geological relations and mineral compositions

FLAG PX AMPH QTZ OTHER

12.6

7.7

PLGT; near contact, no xenoliths; q-ab myrmekite Late mafic dyke cuts marginal GB Chilled margin Coarse-grained dyke west of PLGT; cut GB; Psi5~29; sph 3>ep; chl-act hb replaces opx or hb; f to q Cuts PLGT; 3 ft from 419; cpx pseudomorphs Youngest dyke cuts PLGT East margin of PLGT; An^—j]; An26; f.g. granophyre N55W; dips to south; inside ore body Intrudes faulted HLG north of Bayda mine Diabase cut by fault 'Megadyke' north of Bayda mine; veined by ep and cut by fault Variolitic flow west of plagiogranite Blocky massive flow

Notes and abbreviations for 618O values, rock names, mineral names and prefixes are the same as given for Lasail region in Table 1, and: z, zeolite; cc, calcite; eel, celadonite. * Samples for which electron microprobe compositional data were obtained.

Table 3. Oxygen isotopic and petrographic data for samples collected from map localities 340-349 in the Suhaylah intrusion (see Fig. 4) Map Sample locality

(518O value (%o)

Rock type and petrography WR

340 341

81-07 81-11

XENO,f, q, hb, chl, sph, mt, ep (DK, ep, f, chl, amph, mt)

7.6 (6.8)

342 343

81-12 81-80

GB, f, amph, sulph, ep PLGT, f, q, sulph, ep, ab, amph, hb, sph,

6.5 10.2

344

81-81 81-82 81-83

XENO-DK, f, amph, mt, q, ap XENO-DK, f, amph, sph, mt, q, ap PLGT, f, amph, ab, q, mt, ap, hb

345

81-84*

PLGT, act, chl, ab, q, mt, , sph, act hb, ap 9.8

346

81-85

(DK, ab, q, amph, chl, sm)

347 348

81-86 81-88

LGB, f, hb, cpx, opx, ilm, amph (DK, f, amph, mt, q, sph)

81-89

GB, f, amph, mt, pr, sm, q?, hb [81-90]

8.9

349

81-90 81-91 81-92

(DK, f, hb, ilm, sph, opx, cpx, amph, ep) IG, f, hb, amph, cpx, z, tc, ep, di [pr, sm] HLG, f, amph, mt

(6.1) 5.4

FLAG

Geological relations and mineral compositions

PX AMPH QTZ OTHER

13.8

6.9

13.0

6.7

13.4

7.3

7.3

6.0 (10.7)

6.2 10.1

4.8 2.3

Large xeno with microgranophyric texture Late dyke swarm cuts PLGT; ep > chl » amph Strongly zoned FLAG Host for 81 -81 and 81 -82 mafic pillows (xenoliths); relict zoned plag Margin of dark oblong layer Centre of dark oblong layer Mafic-rich portion of intrusion; relict zoned plag + granophyre Mafic-poor portion of intrusion; relict zoned plag + granophyre; all f to ab; minor ilm adjacent to ap; possible pump intergrown with ab SDC; near PLGT margin; plag microphenocrysts Adjacent to PLGT; XENO protolith Late mafic dyke cuts PLGT; f + cpx phenocrysts Screens between dykes; abundant hb; secondary euhedral plag + cpx + mt Microgabbro; centre of wide dyke 10 m from 81-89; transition to HLG 10 m from 81-91; felsic

Notes and abbreviations for (318O values, rock names, mineral names and prefixes are the same as given for Lasail Region in Table 1, and: pump, golden acicular mineral possibly pumpellyite. * Samples for which electron microprobe compositional data were obtained.

most frequently as part of the 'normal' ophiolite stratigraphy. One of the best-developed segregation plagiogranites (type (b)) is exposed in a wadi near Musafiyah, as shown in Figure 6j, and described by Stakes & Taylor (1992). Over a distance of a

few tens of metres, such segregations can be traced stratigraphically upward to larger veins, which coalesce into a single plagiogranite dyke that intrudes the base of the sheeted dyke complex. A similar small plagiogranite intrudes the pegmatitic gabbro and overlying sheeted dykes in

ORIGIN OF PLAGIOGRANITES IN OMAN

325

Fig. 4. Simplified geological map of the Suhaylah area (modified after Lippard (1980)), showing the large plagiogranite body and the sample localities listed in Table 3. (For location see Fig. 1.) Letters in circles: A, exposures of faintly layered gabbro; B, transition from abundant hornblende pegmatites to plagiogranite; C, region with late mafic dykes dismembered in plagiogranite host; D, intrusive contact with basalts; E, isotropic gabbro cut by basaltic dykes with abundant hornblende pegmatite and segregation plagiogranites.

Wadi Lasail, west of the Lasail mining district. Here, small, mineralized (i.e. epidote-chloritesulphide; see Fig. 6e) shear zones can be traced from this intrusive contact to the top of the sheeted dyke complex. Representatives of the third type of body (type (c)) are the larger plagiogranites, typified by the intrusions in the Alley as well as the mediumsized intrusions at Wadi Shafan and Wadi Ragmi (see Stakes & Taylor 1992). These larger intru-

sions are commonly rooted beneath the SDCgabbro boundary, but can intrude rocks as shallow as the base of the pillow lavas (e.g. Lasail) or as deep as the layered gabbros (e.g. Suhaylah). Country rocks at the intrusive contacts are altered to massive epidote-quartz-titanite-sulphide assemblages (see Fig. 6g). Basaltic and gabbroic 'xenoliths' are extremely common in the larger intrusions and can make up 30-50% of the volume. These 'xenolims' have at least three

Table 4. Geochemistry of samples from northern Oman Sample no.

Lithology

Locale

Notes

Si02

OM81-4 OM81-13 OM81-30 OM81-63 OM81-75 OM81-80 OM81-83 OM81-84 OM81-156 OM81-157 OM83-36 OM83-173 OM85-400A OM85-402 OM85-403 OM85-404 OM85-405 OM85-409 OM85-410 OM85-411 OM85-412 OM85-413A OM85-413B OM85-414 OM85-415 OM85-416 OM85-417 OM85-418 OM85-419 OM85-420 OM85-421 OM85-424 OM85-428 OM85-429 OM85-431 OM85-432 OM85-433 OM85-434 OM85-435 OM85-437 OM85-439 OM85-441 OM85-443 OM85-444 OM85-445A OM85-445B OM85-457 OM85-459 OM85-460 OM85-461 OM85-463 OM85-466 OM85-468 OM85-469 OM85-470 OM85-471 OM85-474

PLOT PLOT PLOT PLOT PLOT PLOT PLOT PLOT PLOT PLOT PLOT PLOT BAS DK DK DK DK DK HLG XENO PLOT DK DK DK DK DK DK DK DK DK DK DK DK BAS HLG DK BAS HHLG DK BAS DK DK DK XENO XENO PLOT PLOT GB PLOT DK DK DK XENO GB LG GB-DK DIG

Jizi Kanut Musaf Ragmi Ragmi Suhaylah Suhaylah Suhaylah Shafan Shafan Lasail Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Bayda Lasail mine Lasail mine Lasail mine Lasail mine Bayda Bayda Bayda Bayda Bayda mine Bayda mine Lasail Lasail Lasail Lasail Lasail Lasail Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab

92T4 92T4 92T4 92T1 92T1 T3 T3 T3 92T3 w/ 81-156 Tl T2 F3/225 (1) F3/225 (1) T2(l) F3/225 (1) T2(l) T2(2) T2(2) T2 T2 T2 T2 F3/229 (3) F3/229 (3) F3/229 (3) T2 T2 T2 F3/231 (4) (5) (6) (1) nrF3/219 nr F3/212 nrF3/221 T2 T2(8) (8) Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl nrF2/311 Tl Tl Tl

71.05 62.55 77.26 73.59 65.38 72.25 63.60 68.83 64.49 64.45 71.42 75.01 56.60 55.70 48.30 55.40 45.80 43.60 49.00 57.50 71.10 49.20 53.10 52.40 52.70 65.90 55.30 50.30 53.20 54.50 55.40 53.70 53.30 51.70 50.70 49.80 56.80 47.20 50.50 52.00 51.90 52.80 53.40 55.70 60.90 65.70 56.70 45.40 65.70 53.20 51.70 50.70 52.10 49.30 50.60 45.50 48.30

TiO2 0.507 0.803 0.37 0,31 0.98 0.52 1.22 0.70 0.69 0.656 0.66 0.40 2.08 0.75 1.54 0.69 0.66 2.14 2.16 0.52 0.46 0.78 0.78 0.62 0.61 0.68 0.50 0.79 0.69 0.85 0.82 0.63 1.02 0.69 0.41 0.93 0.96 0.99 1.01 1.08 0.79 1.73 0.78 0.45 0.67 0.87 0.52 0.30 0.58 0.95 0.84 1.07 0.62 0.42 1.36 2.34 0.98

A12O3

Fe203t

13.12 16.11 12.81 13.78 15.31 13.66 14.84 14.58 16.46 16.75 13.36 12.12 13.90 14.60 15.20 14.40 15.60 17.80 14.40 15.20 12.10 17.00 15.00 15.20 15.50 12.90 15.20 15.50 15.20 15.00 14.00 14.80 13.70 15.20 14.30 12.80 14.40 14.40 16.50 17.00 15.50 13.70 15.20 15.20 13.90 13.60 20.10 20.40 15.10 16.30 15.70 16.00 16.80 15.60 20.70 15.40 15.70

6.157 6.768 2.08 1.98 6.69 4.58 7.81 5.28 5.16 3.501 4.82 4.84 12.40 14.50 12.50 14.90 14.70 16.30 14.30 9.50 6.70 13.50 12.50 9.50 11.60 7.80 9.00 10.70 9.14 9.60 10.50 14.10 12.60 12.60 8.30 8.10 10.90 15.70 12.10 10.90 8.90 13.10 9,50 7.10 9.30 7.40 3.30 6.20 6.70 9.20 9.00 9.90 7.50 6.30 7.80 16.30 10.30

MnO

0.061 0.124 0.02 0.03 0.08 0.05 0.12 0.06 0.05 0.034 0.04 0.02 0.14 0.12 0.18 0.13 0.24 0.11 0.17 0.14 0.02 0.11 0.10 0.14 0.10 0.05 0.30 0.12 0.12 0.07 0.17 0.18 0.19 0.14 0.16 0.14 0.16 0.16 0.26 0.32 0.12 0.18 0.20 0.13 0.14 0.10 0.01 0.07 0.02 0.12 0.15 0.13 0.12 0.12 0.06 0.16 0.14

MgO

CaO

Na2O

K2O

1.07 2.53 0.54 0.53 1.33 1.35 2.13 1.19 1.26 1.00 1.37 0.66 4.10 5.87 7.22 4.98 9.80 3.29 4.50 5.28 0.67 6.44 6.55 8.66 7.50 2.40 7.93 8.62 7.29 6.47 4.40 4.86 5.50 7.33 10.20 9.70 4.31 7.54 6.80 6.64 7.39 4.79 6.72 6.81 3.25 1.80 0.88 10.60 1.73 6.45 6.75 6.63 7.76 9.17 3.92 6.94 8.10

2.49 7.03 3.08 8.62 4.15 1.59 3.00 2.45 7.69 8.39 4.41 2.02 2.88 2.26 9.44 2.32 5.66 11.90 9.17 5.99 2.10 2.85 3.99 4.73 3.69 4.30 3.06 5.03 7.53 5.15 6.10 2.94 4.97 5.34 11.30 8.85 3.22 11.90 1.83 1.41 9.60 6.76 5.29 9.05 3.98 3.14 9.48 11.80 2.93 6.65 9.44 9.60 11.10 14.80 11.20 9.41 13.00

5.78 4.46 4.86 0.82 6.24 6.75 7.13 7.38 4.47 4.97 3.79 5.45 4.90 1.70 2.90 2.10 2.90 3.40 4.40 3.50 5.10 4.60 2.70 1.70 2,50 3.30 3.40 3.20 2.27 3.90 2.60 3.07 2.30 1.40 1.50 3.10 5.10 1.40 5.60 6.20 3.60 4.75 3.50 2.60 4.30 4.30 6.90 2.10 5.40 5.00 3.90 3.70 3.10 1.90 4.00 2.90 2.20

0.26 0.26 0.16 0.04 0.17 0.37 0.25 0.24 0.15 0.39 0.40 0.22 0.15 0.71 0.35 1.14 0.48 0.05 0.02 0.42 0.20 0.38 0.66 1.21 0.86 0.60 0.24 0.55 0.63 0.34 0.42 2.07 0.63 0.90 0.24 0.43 0.04 0.10 0.05 0.04 0.12 0.07 0.47 0.53 0.68 0.70 0.00 0.08 0.55 0.28 0.20 0.17 0.13 0.15 0.13 0.10 0.13

H2O

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 3.6 5.4 3.1 5.1 5.4 n.a. 2.8 3.0 1.4 5.9 5.2 6.9 5.2 2.1 5.9 4.9 4.6 4.4 6.2 5.4 6.1 .-• 5.7 3.7 4.9 3.3 1.9 5.0 5.2 3.3 n.a. 5.3 3.0 2.4 1.9 2.1 3.6 1.9 2.9 3.2 3.5 2.0 2.6 2.6 n.a. 1.9

Zr 139 50 207 51 221 211 175 240 167 173 113 87 140 36 94 47 28 99 130 52 83 43 50 38 38 57 19 45 43 55 51 27 48 28 12 47 59 7 75 81 35 45 35 45 82 98 500 14 180 44 42 66 52 21 49 26 24

Y

Rb

49.2 18,5 46 26.3 56.3 69.3 63.9 73.4 54.3 56.3 41.5 40.1 40 19 31 25 7 n.a. 16 24 43 20 18 14 18 24 12 20 18 27 20 12 13 6.2 8.9 17 22 8.9 28 25 16 32 15 14 33 33 82 5.8 44 16 18 23 17 12 11 11 13

1.9 2 0.7 0.5 1.4 1.4 1 1.2 0.9 2.1 2.5 2 6 10 0 4 7 n.a. 4 6 9 11 8 9 6 7 7 9 n.a. 6 7 n.a. 10 7 7 11 4 5 5 0 8 n.a. 7 9 9 8 6 9 7 8 8 9 5 0 4 9 10

Cu

Zn

Sr

S

12 9 3 8 0 0 0 0 58 2 0 2 0 0 0 100 99 0 0 0 3 2500 4 0 2 0 2 2 0 0 0 670 0 130 0 0 260 4 100 110 0 0 190 0 3 9 0 18 4 0 0 0 78 73 29 220 52

21 14 1 4 4 12 13 5 7 1 6 3 28 20 71 22 80 14 16 15 0 21 11 26 11 8 340 21 10 5 93 68 88 150 38 28 100 20 310 760 20 34 66 16 17 14 0 8 0 10 14 9 26 19 8 41 19

626 205 126 164 165 122 148 134 278 284 180 114 77 160 160 84 120 270 120 120 180 230 110 150 120 170 110 200 180 350 460 71 99 73 170 280 92 91 110 87 310 72 250 120 140 150 43 320 220 260 330 310 150 180 270 190 180

n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 117 222 n.a. 213 703 n.a. <50 466 n.a. 221 <50 80 55 <50 n.a. n.a. <50 <50 428 1170 221 n.a. 78 59 351 145 n.a. 404 112 n.a. 107 554 n.a. n.a. n.a. 171 52 139 139 197 125 n.a. 174 119 447

OM85-475 OM85-478 OM85-482 OM85-483 OM85-490 OM85-492 OM85-495 OM85-498 OM85-502 OM85-504 OM85-505 OM85-506 OM85-507 OM85-508 OM85-509 OM85-510 OM85-512 OM85-515 OM85-516 OM85-518A OM85-524 OM85-525 OM85-528 OM85-529 OM 85-531 OM85-536 OM85-537 OM85-538 OM85-539 OM85-540 OM90-616 OM90-619 OM90-620 OM90-622 OM90-623 OM90-625 OM90-626 OM90-631 OM90-632 OM90-633 OM90-635 OM90-637 OM90-638 OM90-641 OM90-642 OM90-643 OM90-644 OM90-645 OM90-646 OM90-649 OM90-651 OM90-652 OM90-653L OM90-653M OM90-656

PEG GB GB-DK LGB LGB LGB DK GB DK GB LGB DIG PLOT LGB DIG DIG XENO DK DK DK GB DK DK DK DIG DIG DK IG DK DK DK DK DK PLOT AND DK DK PLGT DK PERID GB GB BAS DK XENO GB DK MYL PERID DK PERID GB GB GB PLGT

Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Assayab Lasail Assayab Assayab Lasail Assayab Assayab Assayab Assayab Assayab Bayda Bayda Bayda Bayda Bayda W. Fizh W. Fizh W. Fizh W. Fizh W. Fizh W.Ragmi W. Ragmi Bayda W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi W. Ragmi Bayda

Tl Tl Tl Tl Tl Tl Tl nr F2/302 Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl Tl T2(9) T2 T2 nr F3/222 T2 (10) (10) T6(ll) (11) (11) (12) (13) T2 (14) (15) (12) (12) (13) (16) (15) (15) (16) (17) (17) nr F3/206

46.70 48.10 44.00 38.90 44.80 48.10 48.50 48.50 50.00 44.80 50.30 48.00 42.80 48.00 51.00 49.50 50.50 50.70 55.50 54.50 49.68 52.60 56.39 55.60 52.66 52.40 51.50 51.30 54.00 52.00 54.60 63.10 52.40 75.90 54.20 52.70 70.20 72.30 57.80 55.20 45.00 49.40 52.20 54.40 60.80 45.60 51.30 51.40 47.10 51.30 46.90 48.50 50.50 49.70 76.20

0.29 0.40 0.14 0.07 0.03 0.14 0.49 0.28 0.96 0.22 0.34 2.21 0.17 0.32 0.54 0.34 0.61 0.97 0.92 1.49 0.217 1.98 1.383 0.98 0.558 1.47 1.37 1.25 1.17 1.20 1.81 0.42 1.34 0.38 1.65 0.34 0.64 0.37 0.41 0.41 0.20 0.12 1.96 1.52 1.25 0.16 0.31 0.16 0.05 0.30 0.14 0.29 0.57 0.41 0.49

16.90 16.60 13.90 16.10 25.80 19.30 13.00 21.00 17.80 16.80 16.80 17.40 22.50 21.40 17.40 15.10 15.30 15.80 15.90 14.90 18.01 15.30 15.26 13.80 15.87 15.80 15.30 17.10 15.30 15.50 14.20 15.50 15.10 12.60 14.30 15.50 13.00 12.90 15.90 15.00 17.10 16.80 15.10 14.60 15.90 16.70 13.70 9.50 1.20 15.40 8.20 16.50 17.10 15.10 12.00

6.50 7.60 4.90 2.60 2.80 3.50 8.60 3.60 4.90 6.10 5.20 12.40 1.90 5.00 5.40 5.30 7.40 9.42 9.10 11.30 3.923 13.00 12.158 8.90 9.579 12.00 10.20 12.10 10.10 11.70 12.80 6.20 12.30 1.40 10.70 8.50 5.50 4.70 7.10 7.80 6.80 6.60 7.40 10.70 6.70 6.40 7.10 6.30 6.00 9.40 8.70 5.40 10.40 8.60 1.40

0.10 0.11 0.08 0.06 0.03 0.07 0.10 0.07 0.06 0.09 0.08 0.15 0.04 0.06 0.10 0.09 0.12 0.11 0.08 0.09 0.084 0.18 0.128 0.13 0.145 0.08 0.08 0.17 0.15 0.10 0.21 0.08 0.12 0.06 0.12 0.19 0.08 0.03 0.08 0.10 0.10 0.13 0.15 0.18 0.10 0.10 0.13 0.07 0.11 0.15 0.14 0.10 0.17 0.16 0.01

10.80 10.00 14.50 12.10 9.19 9.50 13.70 6.71 7.97 8.61 9.23 5.25 1.63 5.97 7.44 10.60 9.90 6.53 5.10 4.63 9.10 3.94 4.62 3.62 8.76 5.39 6.61 4.03 5.70 5.68 5.68 4.29 7.75 1.43 5.41 7.75 1.20 1.43 5.42 6.59 13.20 11.90 6.54 4.80 2.60 13.00 10.80 16.40 25.60 8.74 21.00 9.60 6.95 10.10 1.60

15.10 14.20 17.00 25.00 13.80 17.50 10.60 16.00 12.80 16.70 14.30 8.90 26.10 14.10 14.30 13.50 10.20 11.20 3.88 5.04 18.20 5.71 7.23 8.08 10.37 4.67 7.79 5.21 6.61 5.89 2.91 0.63 2.21 1.24 5.34 9.41 3.12 4.39 9.34 7.70 13.00 13.80 7.31 6.00 6.20 14.20 12.00 13.50 14.10 7.86 9.40 15.90 11.60 15.30 1.76

1.30 1.60 0.70 0.00 1.10 1.00 1.70 2.10 2.70 1.20 2.20 3.70 0.00 2.30 2.40 2.60 2.90 3.19 6.70 6.40 1.14 5.20 3.70 3.00 2.60 6.10 4.20 6.00 5.20 4.70 3.50 7.00 4.70 5.40 6.60 2.60 5.10 3.70 2.90 4.70 1.00 0.60 5.60 5.80 5.30 1.00 2.00 0.40 0.00 3.30 0.10 1.70 2.40 1.10 4.80

0.02 0.12 0.09 0.00 0.03 0.01 0.02 0.08 0.02 0.04 0.04 0.17 0.00 0.05 0.15 0.13 0.37 0.22 0.31 0.21 0.02 0.47 0.24 0.12 0.39 0.10 0.36 0.35 0.43 0.24 0.00 0.01 0.02 0.25 0.06 0.17 0.00 0.11 0.20 0.38 0.00 0.00 0.11 0.04 0.13 0.01 0.00 0.01 0.00 0.03 0.00 0.00 0.00 0.00 0.20

2.8 1.8 4.8 4.9 3.3 1.7 3.5 2.0 3.1 5.5 1.8 2.2 4.6 2.9 2.1 3.1 2.7 n.a. 3.0 2.1 n.a. n.a. n.a. 3.1 n.a. 3.0 3.2 3.0 2.5 3.7 5.0 3.0 5.0 1.5 2.4 3.6 1.6 1.1 1.6 2.8 4.1 1.6 3.6 3.2 1.2 3.3 3.4 2.8 5.4 4.7 5.8 2.2 1.3 0.7 1.6

11 19 0 0 0 1 33 10 45 5 19 46 7 10 24 14 33 65 91 96 18 140 63 37 33 93 72 98 74 68 120 19 77 79 110 12 92 47 22 14 4 0 120 120 210 4 6 4 4 5 4 0 5 4 80

6.7 8.9 3.8 2 0.6 3.3 12 6.9 19 5.3 6.2 10 4.6 5.9 12 7.2 12 19 21 28 6.4 40 21.1 20 14.4 30 26 32 26 25 41 11 27 36 42 11 33 18 16 16 4.9 3.2 46 38 49 4.5 10 2.6 1.5 8.7 4.6 5.1 11 10 40

0 7 9 4 9 9 7 7 6 6 6 10 0 8 4 6 7 n.a. 7 7 0.2 7 1 5 1.9 9 7 5 0 9 7 0 5 3 5 9 6 9 0 13 6 0 7 5 8 5 6 6 4 6 8 0 5 6 0

66 14 120 6 30 59 0 5 18 37 2 60 0 0 140 27 64 0 0 0 139 18 0 1 0 0 0 0 3 0 1 33 43 23 7 94 2 7 11 1 57 8 8 0 30 87 1 0 2 69 38 120 38 60 2

19 16 51 2 5 3 10 15 2 10 12 15 0 6 13 17 39 3 31 18 13 97 21 69 24 1 6 23 37 12 28 100 240 15 19 100 100 3 58 25 20 21 33 27 7 20 26 1 14 64 38 16 67 28 9

150 150 280 0 140 160 130 220 230 130 180 240 27 250 200 120 180 230 180 140 134 230 127 150 130 120 240 290 230 240 64 30 83 150 130 130 64 97 110 140 130 81 270 67 190 120 180 73 18 140 7 170 100 140 160

2152 309 n.a. n.a. 198 122 65 109 n.a. n.a. 110 320 n.a. 81 167 83 98 n.a. 55 117 n.a. 98 n.a. n.a. n.a. 88 100 144 63 125 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 96 n.a. n.a. n.a. n.a. 61 n.a. n.a.

Prefixes T and F refer to tables and figures in this paper; 92T and 92F refer to tables and figures of Stakes & Taylor (1992). (1) Outcrop of epidotized pillows (400, 401) cut by dykes (402, 403, 404). Dyke 405 is adjacent SDC. (2) Adjacent to mineralized fault; dyke (409) cuts gabbro (410). (3) Three adjacent dykes sampled along wall of wadi east of 413 (414, 415, 416). Dyke 416 is stained with altered pyrite. (4) Low-angle dyke at 100 m level of mine contemporaneous with sulphide-quartz ore. (5) Late siliceous dykes crosscut highly altered volcanic rocks and sulphide. (6) 'Ore body sheet' near footwall. (7) Mineralized sheet or flow may predate sulphides. (8) Massive flow or dykes inside ore body; 437 is from more Cu-rich portion of deposit. (9) Coarsegrained 'megadyke' north of Bayda mine; suggested by Koski et al. (1990) as related to Bayda ore formation. (10) Fizh road east of Ragmi. Sample 625 is syn-alteration sill described by Smewing (1990); sample 626 is late acid dyke from same outcrop. (11) Fizh road east of Ragmi; small plagiogranite (631) is cut by late mafic dyke (632) and late picrite (633). (12) Wadi Ragmi road at dyke-gabbro contact. Vertical foliation in gabbro is crosscut by late mafic dykes. Sample 643 is unfoliated gabbro at site, 635 is foliated gabbro and 644 is late mafic dyke. (13) In cumulate gabbros with interlayered gabbro and wehrlites. Sample 637 is a gabbro and 645 is a mylonite zone within the gabbro. (14) Sheeted dykes with consistent orientation of 100. Dyke is faulted and heavily veined with ep-q-sulphide. (15) Dykes with gabbro screens. Xenolith 642 has epidotized margins. Late hb-bearing dyke 649 is cut by picrite dyke 651. (16) Olivine websterite 646 cut by two-pyroxene gabbro pegmatite dykes. Adjacent gabbro 652 is highly foliated veined. (17) Within cumulate gabbro section outcrop of coarse-grained massive gabbro (653M) intruding layered gabbro (653L). n.a., not analysed.

D. S. STAKES & H. P. TAYLOR, JR

328

Table 5. Plagiogranites (with silica content given in parentheses) Sample no.

Locale

Description

OM81-30

Musafiyah

OM81-63 OM81-75 OM81-80 OM81-83 OM81-84 OM81-156 OM83-36 OM83-173 OM85-412 OM85-445B OM85-457 OM85-460 OM90-622 OM90-631 OM90-656 95GIA-11A* 95GIA-28A* 95GIA-28C* 95GIA-34B*

W. Ragmi W. Ragmi Suhaylah Suhaylah Suhaylah W. Shafan Lasail Aarja Aarja Lasail Assayab Assayab Aarja W. Fizh Aarja Aarja Aarja Aarja Aarja

Small segregation plagiogranite with transitional lower contact; intrudes base of dykes (76%) Deep plagiogranite cut by gabbro dyke (74%) Small body at base of sheeted dykes (65%) Large deep intrusion; see Table 3 (71%) Mafic enclave rich; see Table 3 (63%) Mafic enclave poor; see Table 3 (68%) Intermediate intrusion at dyke-gabbro boundary (64%) Large intrusion; see Table 1 (71%) Intermediate intrusion; see Table 2 (74%) Near upper contact with dykes; see Table 2 (71%) At contact with lavas; screen for mafic xenolith (65%) Small body intrudes upper dykes (57%) Small body intrudes upper dykes and HLG (66%) Screens cut by mineralized mafic dykes (76%) Screens cut by dykes (72%) Main intrusion; with abundant epidote (76%) Trondjemite from core of complex (73%) Mesotonalite; xenolithic in core of complex (57%) Isotropic gabbro; marginal to complex (52%) Altered gabbro; marginal to complex (55%)

*Data from Galley & Koski (1995) and Koski et al. 1995

different origins, as inferred from primary textural characteristics: (1) stoped blocks of the sheeted dyke complex; (2) stoped blocks of variegated gabbro derived either from the upper gabbro horizon or from coarse-grained, recrystallized dykes; (3) crosscutting or penecontemporaneous, dismembered or 'pillowed' mafic dykes emplaced within the much cooler plagiogranite magma (the enclaves shown in Fig. 6a-d). Xenoliths and enclaves found deeper in these intrusions are more likely to contain hornblende-plagioclase (± diopside ± ilmenite). Tabular xenoliths may originate either as stoped blocks or dismembered mafic dykes, but an unambiguous distinction can be made in the case of the net-veined basaltic 'pillows' in the plagiogranites. These dark ovoid enclaves commonly have mutually intrusive margins with their plagiogranite host, as is well shown in Figure 6c. Abrupt changes in mineralogy and texture within the larger intrusions are best interpreted as phantom relicts of stoped blocks almost completely recrystallized within the plagiogranite. The proportions of such mafic xenoliths and enclaves vary widely within the plagiogranite (e.g. Fig. 6a). The enclave-rich lower portions of the intrusions are typically mafic-rich, and the mafic-poor portions of the plagiogranite commonly make up the upper third of the body where xenoliths are less frequent (Fig. 6a and b). Gradational contacts between HLG layers and plagiogranites are com-

mon, including a transition from a laminated texture to a spotted 'porphyroblastic texture', with local deformed layers. This textural change has been interpreted as a metamorphic recrystallization, similar to the 'isotropization' of pre-existing layered gabbros by a younger intrusion described by Reuber (1988) in Wadi Ragmi. We suggest that these textures are representative of the latter stages of RAFC processes in which the extraction of anatectic melt becomes increasingly important (e.g. Bohrson & Spera 2001; Spera & Bohrson 2001).

Description of areas sampled Aarja-Bay da plagiogranite Excellent exposures of the Aarja-Bayda plagiogranite in Wadi Bani Umar Gharbi (Fig. 3) display multiple generations of dykes of 0.5-1 m width that crosscut more than 30% of the outcrop. These dykes predate, are contemporaneous with, and postdate the plagiogranite intrusion (Fig. 6a and b). Gabbroic rocks along the western margin of the plagiogranite are extensively altered to assemblages whose peak metamorphic grade is indicated by hornblende to pyroxene hornfels. Veins that contain epidote, quartz, Fe-actinolite and oxidized relicts of sulphides crosscut these gabbros. Malachite staining around the epidote-rich veins and epidotized xenoliths reveals the Cu-rich nature of

ORIGIN OF PLAGIOGRANITES IN OMAN

329

Fig. 5. Schematic diagram of the stratigraphic zone in the Semail ophiolite that is the main focus of this paper, showing the lateral variability in rock types and structure along the subhorizontal contact between the sheeted dyke complex (light blue) and the underlying cumulate gabbros (brown). From left to right, we illustrate (1) xenoliths and shattered remnants of isotropic gabbro (red) with quartz-rich segregations; (2) small plagiogranite dykes (white) that intrude the base of the sheeted dyke complex; (3) large zoned plagiogranite bodies (white) that intrude isotropic highlevel gabbro, sheeted dykes and, locally, even the overlying pillow lavas (grey and green). These large plagiogranites may be composed of up to 60% xenoliths, and are typically crosscut by late dykes. The upper contacts are hydrothermally altered to epidote- and chlorite-rich stockwork assemblages (light green).

many of the original sulphides. At the SW contact between the plagiogranite and the dykes and lavas, there is a massive alteration zone with interlayered chlorite- and epidote-rich replacement zones similar to epidosites described in the Solea graben in the Troodos ophiolite (Schiffman et al. 1988, 1990). These veins and replacement zones include chlorite-quartz, epidote-quartz, epidote-pyritechalcopyrite and epidote-specular hematite. This alteration zone can be traced from the plagiogranite directly into the overlying Aarja ore deposit, and it may also connect to the nearby Bayda ore deposit. East of the plagiogranite and north of the Bayda mine (Fig. 3) the SDC is delineated by weathered sulphides in dyke-parallel shears (Haymon et al. 1989) that are similar to those observed in Wadi Lasail (Fig. 6e). It is important to note that an extensive high-temperature hydromermal aureole extending for several hundred metres on every margin surrounds this relatively small, zoned plagiogranite intrusion.

The marginal gabbro in contact with the plagiogranite is strongly 18O-depleted, indicating alteration by high-temperature nydrothermal fluids (e.g. 85-434 at site 232 on the eastern side and 85-183 at site 224 on the western side). Primary plagioclase compositions vary in the range An7i-g9, and secondary plagioclase is An2o-26; the latter appears to be contemporaneous with abundant acicular actinolitic hornblende that replaces pyroxene. This assemblage (sodic plagioclase and actinolitic hornblende) is compatible with temperatures in excess of 400 °C, consistent with the 18O data (see below) and the fluid inclusion geothermometry on rocks from this area (Nehlig & Juteau 1988; Nehlig et al. 1994). Pale, interstitial (primary?) hornblende may be analogous to the hornblende observed in late wehrlitic intrusions (Beurrier et al. 1989). The most unusual characteristic of the marginal gabbros, however, is the presence of interstitial micrographic intergrowths of quartz and sodic plagioclase. These sodic granophyric

330

D. S. STAKES & H. P. TAYLOR, JR

intergrowths typically surround strongly zoned plagioclase crystals (An9i -56). The intrusive contact on the western edge of the plagiogranite exposes multiple generations of xenoliths, dismembered dykes, late crosscutting dykes and mafic enclaves. Thin late dykes are deformed and metamorphosed within the plagiogranite body, and are locally dismembered and boudinaged. Net-veined basaltic enclaves ('pillows') and xenoliths in the interior of the plagiogranite are prominent in exposures on the walls of the wadi. One outcrop on the NW contact of the plagiogranite and the gabbro is particularly interesting as it exposes a wide dyke that is dismembered and composed of numerous basaltic

'pillows' (Fig. 6a and b). These enclaves are ovoid and 0.2-0.5 m across. At the margins of some of the enclaves, the mafic intrusion is quenched against the felsic host, whereas elsewhere on the enclave margin, the plagiogranite mutually intrudes the enclave. These field characteristics demonstrate that the basaltic and plagiogranite magmas were contemporaneous (e.g. see Pitcher 1993). A similar outcrop, described by Aldiss (1978) for the Palekhori plagiogranite in the Troodos ophiolite, displays many of these characteristics. Although some of the basaltic enclaves in the plagiogranite were clearly formed from gabbroic or basaltic dykes emplaced into and then chilled

ORIGIN OF PLAGIOGRANITES IN OMAN

331

332

D. S. STAKES & H. P. TAYLOR, JR

Fig. 6. (a) Photograph (hammer in lower right for scale) of relationships along the western contact of the AarjaBayda plagiogranite body (at locality 220 in Fig. 3). On the left side of the photograph, one sees multiple generations of xenoliths, dismembered dykes, net-veined basaltic 'pillows' and late crosscutting dykes. The plagiogranite here varies in grain size and textures: it is mainly composed of coarse granophyric intergrowths of quartz and altered turbid feldspar, with abundant epidote and sphene, and usually some euhedral pyrite. Darker regions of the plagiogranite are richer in either actinolite or epidote. The still darker regions define the relict zone of a mafic dyke that is partially digested within the plagiogranite and is crosscut by thin late-stage dykes. Ovoid enclaves are found within the interiors of the earlier dyke; these are net-veined basaltic 'pillows' that formed when injections of mafic magma were chilled by the much cooler plagiogranite magma. Thin late dykes have a cherty appearance and are deformed and metamorphosed within the plagiogranite body (they have no relict igneous texture, and are composed of epidote, quartz, chlorite, prehnite and euhedral sulphides). Veins that are adjacent to the late dykes contain epidote (PS25-32), quartz, jasper and stilbite. (b) Explanatory diagram of outcrop photograph in (a): (1) homogeneous plagiogranite with quartz, albite, actinolite, epidote and magnetite as the primary minerals. To the left is the contact with a wide, partially digested basaltic dyke that was intruded into the plagiogranite melt before complete crystallization (2, 3, 4). These rocks are, in turn, crosscut by thin late dykes (5 and 6). (2) Mafic-rich plagiogranite— diorite host for small ovoid mafic enclaves. (3) Hybrid margin of the wide dyke. (4) Small, fine-grained, net-veined enclaves, showing mutually intrusive contacts with the adjacent plagiogranite. (5) A thin late basaltic dyke that crosscuts all lithologies. (6) Sheared composite dyke with altered margins (sample OM85-413). The late dykes are anastomosing and locally pinch out. (c) Net-veined enclaves ('basaltic pillows') at Suhaylah (near locality 345 in Fig. 4). This photograph is a close-up of one of the large ovoid mafic enclaves found within the core of the plagiogranite intrusion at Suhaylah (this example, with mutually interfingermg margins, is about 1 m in diameter). These enclaves result from quenching of mafic magma within the much cooler plagiogranite melt. Other adjacent xenoliths have patches of relict igneous layering and probably represent partially recrystallized blocks of cumulate gabbro. (d) Large deformed xenolith (4-5 m long X 1-1.5 m wide) within the Lasail plagiogranite body, near the NW contact of this body with the lower pillow lavas (Fig. 2). Sample OM85-444 is from this outcrop (location 279). The large deformed xenolith is surrounded by smaller xenoliths and a reaction zone rich in hornblende.

ORIGIN OF PLAGIOGRANITES IN OMAN

333

(e) Mineralized hydrothermal discharge zone in sheeted dykes exposed in Wadi Lasail near a contact with plagiogranite. Sample OM85-443 (locality 325, Fig. 2) is from the late-stage dyke that cuts pillow lavas adjacent to the plagiogranite contact. The discharge zone is confined to a fault with discontinuous lenses of gossan up to 2 tn wide. The concentration of sulphides and epidote along the margins of the chloritized dykes should be noted. The dykes strike N 25 E, placing this discharge zone directly on strike with the Lasail mine, (f) Sulphide-epidote fracture-filling from a stockwork on the margin of the Aarja-Bayda plagiogranite intrusion, near locality 214 in Figure 3. This alteration is characteristic of sheeted dykes and pillows that overlie the plagiogranite bodies. Scale is an Omani coin c. 2 cm in diameter, (g) A typical epidosite from the Oman ophiolite, from the massive epidosite zone that overlies the plagiogranite intrusion at Wadi Shafan (see Fig. 1). Epidote and quartz replace the dyke margins and fill voids within the dyke complex, (h) BSE (back-scattered electron) image of prehnite with titanite lamellae, replacing pyroxene-ilmenite intergowths in the epidotized gabbroic rocks assimilated by the plagiogranites (sample 85-465 from locality 318, Fig. 2). (i) Photomicrograph of plagiogranite with abundant xenoliths (sample 83-173 from locality 220, Fig. 3). This shows altered plagioclase phenocrysts surrounded by a micrographic (symplectite) intergrowth of albite and quartz, a common texture within the xenoliths and/or marginal rocks of the plagiogranites. The symplectites include quartz-albite, epidote-quartz, epidote-titanite and prehnite-titanite (X5; polarized light), (j) Photograph of a typical segregation-type plagiogranite in the Oman ophiolite, as exemplified by this occurrence at Musafiyah in Wadi Far, south of the region shown in Figure 1 and west of Rustaq. Here an irregular zone of plagiogranite intrudes the high-level gabbros a few metres from a much larger plagiogranite dyke emplaced into the base of the sheeted dyke complex, (k) Photograph of the Lasail copper mine, showing the gossan that overlies the massive sulphide ore body. The extensive greenschist alteration of the extrusive rocks that surround the ore body should be noted. On the horizon, the peaks of the Lasail plagiogranite (Fig. 2) can be seen c. 2 km downsection from the ore body.

334

D. S. STAKES & H. P. TAYLOR, JR

within a cooler silicic melt, there is also evidence that many are true xenoliths of pre-existing gabbro or diabase. These latter xenoliths have undergone partial digestion and extensive exchange with the plagiogranite magma. Typical mineralogy includes radiating aggregates of euhedral plagioclase and green, pleochroic, columnar to acicular grains of amphibole. Titanite is abundant and is observed mantling rare Fe—Ti oxides. Quartz forms large (porphyroblastic?) grains that also mantle and possibly replace the feldspar. We infer that these textures result from the systematic recrystallization and assimilation of altered country rocks, and, even more importantly, of the late mafic dykes that intruded into the upper levels of the oceanic crust at the same time as the plagiogranite magmas were being emplaced. Massive diabase sheets of 1-2 m width and late dykes of 0.5-1 m width crosscutting the eastern margin of the intrusion appear to be the of same generation as the diabase dykes or sheets intruding the adjacent volcanic rocks. There is no evidence of any silicic intrusions that might have been feeder dykes for the plagiogranite intrusions, and the silicic sheets found above the intrusion (and crosscutting the adjacent ore bodies) appear to be derived from the plagiogranite itself.

those that crosscut and concentrate the Cu-rich sulphides within the adjacent Lasail ore body. The core of the Lasail plagiogranite body is a two-pyroxene, hornblende-bearing isotropic gabbro that is similar in mineralogy to the late-stage hydrous gabbros described by Juteau et al, (1988) and Boudier et al. (2000). Plagioclase compositions range from An30 to An7Q. Amphibole varies from magnesio-hornblende to Fe-actinolite in composition. Apparent equilibrium pairs of actinolitic hornblende and sodic plagioclase (An^-39) suggest metamorphic alteration temperatures near 500 °C (Spear 1980; Maruyama et al. 1983; Holland & Blundy 1994). Southwest of the Lasail plagiogranite body is Assayab, a name taken from the fault zone-hosted ore prospect found in this area (Fig. 2). At this locality, a mineralized normal fault cuts the SDC and is obliquely crosscut by numerous late dykes that can be followed throughout the HLG (Fig. 7). Deep isotropic gabbros (DIG) beneath this mineralized zone can be traced down into the cumulate gabbro section where they host a small plagiogranite body surrounded by hornblende-bearing isotropic gabbro. The Assayab locality thus features an intrusive complex emplaced along a deep shear zone that was itself superimposed upon the regional ophiolite structure.

Lasail plagiogranite Although the plagiogranite intrusion in the Lasail area (Fig. 2) is at least four times larger than the Aarja-Bayda plagiogranite, these two bodies share many of the same mineralogical and structural features. These include an abundance of xenoliths and/or dismembered mafic dykes, pegmatoidal gabbro on some of the lower margins, and a spatial relationship to the sulphide ore deposits in the overlying pillow basalts. Hornblende-rich xenoliths locally form up to 50% of the volume of the Lasail intrusion. The mafic-rich upper contact of the intrusion (e.g. sample 83-72 at site 277) contains strongly zoned plagioclase (An52-36), zoned epidote (Psi6-2s)» and actinolitic hornblende grains that may be vestiges of the assimilated country rocks. These mineral phases are intermingled with granophyric intergrowths of albite and quartz, and crosscut by veins of quartz, epidote and pyrite. Blocks of relatively unaltered volcanic rocks at the upper intrusive contact confirm that this plagiogranite has intruded into the Geotimes/Vl/Ml Unit. Volcanic rocks (both Geotimes/Vl/Ml and Lasail/V2/M2) on the upper, NW margin of the plagiogranite are intruded by felsite dykes and sheets with a mineralogy that includes quartz, albite and euhedral pyrite. These felsic sheets originate from the top of the plagiogranite and have similar chemical compositions to

Suhaylah plagiogranite Behind the village of Suhaylah, just north of the Wadi Jizi road, is a lesser-studied plagiogranite body that is as large as the Lasail body, with dimensions of 10km by 8km (Fig. 4). The Suhaylah body was emplaced stratigraphically deeper than either Lasail or Aarja-Bayda, with a lower contact within the cumulate gabbros and an upper contact that varies from the dyke-gabbro boundary on the NE side to direct contact with basalts on the southern margin. On the easternmost margin, just north of the village, a 45 m high ridge of plagiogranite is crosscut by a series of 1 3 m thick dark layers of dismembered late mafic dykes that show variable degrees of recrystallization and exchange with the plagiogranitic host. These apparently are larger versions of the late mafic dykes described at Aarja-Bayda. On the southern margin the plagiogranite contact is a gradual transition from coarse diabase dykes that cut isotropic marginal gabbro to a 'normal' dykegabbro contact, with diabase dykes rooted in massive gabbro. The lower contact of this plagiogranite extends deeply into the cumulate gabbro sequence, where hornblende pegmatite dykes crosscut it. Layered gabbros are exposed within the lower boundaries of the plagiogranite and some xenoliths in the central core of the intrusion

ORIGIN OF PLAGIOGRANITES IN OMAN

335

Fig. 7. Schematic diagram of outcrop relationships in Assayab prospect area (See Figs 1 and 2 for location). The Assayab prospect is a fault-zone hosted sulphide deposit within the upper gabbros. Late-stage dykes (LMD) crosscut the gabbros and the sheeted dyke complex at an oblique angle. Evidence for high-temperature alteration is found associated with the deep shear zone with a Cu-rich mineralized zone (MZ) and a fault that detaches the sheeted dykes from the upper gabbros. CP, cumulate peridotite; CG, cumulate gabbro; SDC, sheeted dyke complex; HLG, high-level isotropic gabbro. Several types of late intrusions include the late-stage mafic dykes (LMD), the deep isotropic gabbro (DIG) and associated plagiogranite (PL). The transition from the phase layered gabbro to the DIG is observed as (1) local deformed layers with increasing abundance of hornblende (wavy lines); (2) radiating aggregates of hornblende and ilmenite porphyroblasts (X); to (3) more homogeneous isotropic gabbro (c). Adapted from Vetter & Stakes (1990).

have a relict layered texture. The interior of the plagiogranite is enclave-rich, making up more than 50% of the outcrop in some locations. Although some of the xenoliths are clearly partially recrystallized gabbro country rock, others are mafic net-veined ovoid 'pillows' formed by quenching of the mafic melt within the plagiogranitic liquid (Fig. 6c). The mafic enclaves at Suhaylah are unique in that they retain a hornfelsic texture of equigranular brown actinolitic hornblende and sodic feldspar with little lowtemperature hydrous overprint.

Oxygen isotope chemistry of plagiogranites, stockworks and the upper ophiolite Oxygen isotopic variations in rocks and their constituent minerals are a useful tool for tracking the flow and the amounts of aqueous fluids that interact with igneous rocks during hydrothermal circulation. The 18O/16O data from the rare plagiogranites found in oceanic dredge hauls and drill holes can be directly compared with the Oman plagiogranites (e.g. Muehlenbachs & Clayton

336

D. S. STAKES & H. P. TAYLOR, JR

1971; Muehlenbachs & Byerly 1982; Stakes 1991). Oxygen isotope data for the Lasail-Assayab region are given in Table 1 and plotted in Figure 8 according to their relative stratigraphic height in the ophiolite. Most of the (318O values for the upper part of the ophiolite are much higher than those for fresh MORE (<518O - 5.8%o; Taylor

1968) with (518O values up to +15.6%o. If this 18O enrichment is attributed to exchange with heated seawater having a d18O of 0%o, this requires that the final imprint of hydrothermal alteration (e.g. in the feldspars and alteration minerals) must have occurred at temperatures below 250 °C (O'Neil & Taylor 1967). Whole-rock d18O values lower than

Fig. 8. Plot of 618O v. relative structural height in the ophiolite for the mineral separates and whole-rock samples from the Lasail-Assayab traverses (Table 1). The data are arranged with depth only in a qualitative fashion, relative to the primary ophiolite lithostratigraphy. The layered gabbro and isotropic gabbro of the primary ophiolite magmatic series are shown in red. Isotropic gabbro that crosscuts the layered gabbro at deep (DIG) or high levels (HLG) is shown in green. Lavender is used for mafic dykes; red for gabbro dykes; light blue for plagiogranite; brown for peridotites. PD, late peridotitic wehrlite; PG, pegmatite; CHL, chlorite; C, dyke centre; M, dyke margin; V, vein mineral. Circles enclose samples associated with the late intrusions in Assayab.

ORIGIN OF PLAGIOGRANITES IN OMAN those of MORB are diagnostic of seawater interactions at temperatures above 250 °C. Such low <518O values are found in xenoliths and, most dramatically, in early dykes and gabbros that are in contact with the Lasail plagiogranite intrusion. In the Assayab region, relatively low <518O values are also found in dykes and gabbros near the heavily mineralized faults beneath the HLG, suggesting a deeper penetration and greater heating of seawater related to the zone of intense fracturing along this normal fault. For many of the crystalline rocks distal to the contacts and/or faults, the whole-rock <318O values are much higher, even though the mineralogy is greenschist or transitional greenschist-amphibolite in grade; such metamorphic assemblages require temperatures in excess of 300 °C (Liou et al 1974, 1983; Maruyama et al. 1983). In most cases, quartz separated from these samples is lower in <318O than that of the whole rock. This is inconsistent with the normal order of 18O enrichment in these minerals at equilibrium (quartz should be the most 18 O-rich mineral) and indicates that other minerals in the samples (particularly the plagioclase) have undergone a late-stage enrichment in 18O during this subsolidus alteration, either at lower temperatures or by a fluid enriched in 18O, or both. This phenomenon, namely, the existence of hydrothermally 18O-enriched ophiolitic rocks whose constituent minerals clearly must have formed at 300-400 °C or higher, can be observed throughout the Oman ophiolite, particularly near the level of the SDC-gabbro contact (Gregory & Taylor 1981; Stakes & Taylor 1992). We concur with Gregory & Taylor (1981) in explaining these observations by influx of ocean water into fractured gabbros much deeper in the ophiolite at temperatures of 400-800 °C, lowering these whole-rock (518O values from +5.8%o to the values of +2 to +5%o observed in many of the cumulate and isotropic gabbros at Wadi Ragmi, Wadi Shafan and Ibra (Stakes & Taylor 1992) and in the present study at Aarja-Bayda and Lasail-Assayab. By material balance, this produces 18Oenriched, evolved seawater that could then have <518O values anywhere in the range from zero (at very high water/rock ratios) to +7 or +8%o (in exchange equilibrium at low water/rock ratios with the more 18O-enriched dykes, gabbros and plagiogranites shown in Figs 8 and 9, many of which have d18O of +6 to -|-9%o). These plagiogranite magmas in Oman originally had (318O values as high as +7,5%o (see below), so it is also possible that true magmatic H2O with (518O of +7 or +8%o, under lithostatic pressure, could have been exsolved from these magmas at temperatures as high as 800 °C as they crystallized. The hydrogen in such H2O would also probably have

337

originally been derived from seawater (see the D/ H data of Stakes & O'Neil 1982), but because it would have been incorporated into the plagiogranite magmas by the dehydration reactions that accompany assimilation and stoping of hydrothermally altered roof rocks derived from the sheeted dyke complex and the gabbros, it would have attained a (318O similar to that of the magma from it was derived. Therefore, it is plausible that the high <518O H2O required to form the epidotequartz-titanite stockworks above the plagiogranites could represent a mixture of magmatic H2O and strongly 18O-shifted seawater. This cannot represent the whole story of 18O enrichment in the ophiolite, however, particularly of the striking hydrothermal change in the (518O values of the feldspars in the plagiogranites, from their primary (518O of +3.5 to +7.0%o, to their measured <518O values that are as high as +12 to + 16%o (see the discussion below and the data plotted in Fig. 11). At about 400 °C this would require H2O with a <518O of +9 to +13%o (e.g. Taylor 1997), and there is no evidence that H2O with that high a <318O value was ever present during the formation of the Oman ophiolite. Thus, superimposed upon the effects of high-18O fluids described above, there probably must be an additional, retrograde hydrothermal exchange event that proceeds at temperatures as low as 150200 °C. The albites in the plagiogranites are particularly susceptible to this retrograde hydrothermal exchange, as has been known for some time (Gregory & Taylor 1981). As an example, at 150°C the equilibrium albite-H2O 18O/16O fractionation is known to be about 15%o (Taylor 1997), which would be large enough to account for the observed data in Oman, even utilizing unshifted ocean water with a ($18O of about zero.

Plagiogranite magmas Figures 10 and 11 are compilations of mineral oxygen isotopic data from a variety of plagiogranite bodies in Oman. The highest temperature hydrothermal veins contain epidote and sulphide or oxide and are found near the margins of the plagiogranite bodies (Fig. 10). The relatively low whole-rock <518O values of some of the late mafic dykes within the plagiogranite body at AarjaBayda (Figs 9 and 10) are compatible with hightemperature alteration by a hydrothermal fluid near seawater in isotopic composition. The highly altered gabbro on the margins has the lowest (318O value of any samples from this part of the ophiolite (see Fig. 3). The intrusive contacts and hydrothermal aureoles that overlie the plagiogranite intrusions are locally strongly depleted in 18O, as would be expected in the case of intense

338

D. S. STAKES & H. P. TAYLOR, JR

Fig. 9. Plot of 618O of whole-rock and mineral samples in the vicinity of the plagiogranite body near the AarjaBayda ore deposits (Fig. 3), grouped according to location relative to the plagiogranite. Light blue is used for plagiogranite; green for high-level isotropic gabbro; red for isotropic gabbro; black for basalt; lavender for mafic dykes. Y vein mineral; PD, late wehrlite intrusion; PG, pegmatite; C, dyke centre; M, dyke margin.

circulation and hydrothermal exchange with seawater at high water/rock ratios. This is analogous to what is observed for practically all shallow subaerial intrusions on land (Taylor 1974, 1977), with the only difference being that much lower18 O meteoric waters are involved instead of ocean water. Table 5 summarizes the field and chemical characteristics of the Oman plagiogranites that have been analysed for oxygen isotopic composition and major and trace element chemistry. The silica contents of these plagiogranites vary from 63% to 76%. Plagioclase and quartz were separated from these plagiogranites to try to characterize the primary oxygen isotopic compositions of these ophiolitic silicic magmas. The expected equilibrium 18O fractionation between igneous quartz and plagioclase is A18O = 0.9-1.4%o (Taylor & Sheppard 1986; Clayton et al 1989). Deviations from this equilibrium fractionation reflect post-crystallization hydrothermal alteration, after the plagiogranite body was rigid enough to fracture and thereby admit the hydrothermal fluids

that were circulating in the surrounding country rocks. The <518O quartz values for the samples analysed in this study are confined to a relatively narrow range, from +5.8 to +8.2%o, with the great majority of samples lying between +6.5 and +7.5%o. If results from Wadi Saq are included (Gregory & Taylor 1981) this range is expanded to values as low as +4.8%o (Fig. 11). Only a couple of samples (from the Lasail intrusion) contain plagioclase with a <518O value that falls near the equilibrium line. For all the other samples, plagioclase is shifted to oxygen isotope values that are much higher than would be expected for equilibration with the coexisting quartz under igneous conditions. Most of the plagioclase separates are shifted upward in (518O in Figure 11 by 2-8%o, as a result of subsolidus interactions with hydrothermal fluids. As explained above, either these fluids are highly enriched in 18O compared with the original seawater and/or the alteration continued on down to temperatures below 250 °C (Stakes & O'Neil 1982; Taylor 1997).

ORIGIN OF PLAGIOGRANITES IN OMAN

339

Fig. 10. Composite plot of 618O for minerals and whole rocks from the three plagiogranite bodies described in this paper. Samples are plotted according to their specific lithological facies within and around the intrusion, i.e. the core, the xenoliths and host plagiogranite, the veins and the surrounding country rocks. A maximum fluid temperature is calculated based on the lowest-618O vein quartz (618O +4.1%o) and assuming that the fluid is seawater with a 618O of zero (after Clayton et al. 1989). Evolved or magmatic fluids with 618O >0 would raise the temperature estimate.

The initial <518O value of a MORB-type magma would be +5.8 (±0.2%o), and products from the closed-system fractional crystallization of a MORE magma could perhaps range from <518O of +5.6 to +6.7%o, an increase of no more than l%o (Taylor & Sheppard 1986). Thus a plagiogranite magma that formed solely from closed-system fractionation would be expected to have an initial (518O of about +6.4 (±0.3%o). The steeply dipping arrays of plagioclase-quartz <518O values in Figure 11 can be projected downward to the quartz plagioclase equilibrium band (at A18O quartzfeldspar = 0.9-1.4%o) to estimate the original isotopic compositions of the plagiogranite magmas before the high-18O/low-temperature overprint. If the original plagiogranite magma had an isotopic composition derived solely by fractional crystallization of a MORB-type gabbro, then the constituent plagioclase would be expected to have a <518O of about +6.1%o and the equilibrium quartz would be about +7.0 to +7.5%o. The actual range of quartz (518O values in Figure 11 is from +4.8 to +8.2%o. Based on these isotopic variations we can infer a <518O range for the plagiogranite magmas in Oman from as low as +4.0%o to as high as +7.5%o. The only mechanism that we

know of that could create such a broad range of magmatic (318O compositions, much lower than can be explained by closed-system fractional crystallization, is the incorporation and partial assimilation of altered dykes and isotropic gabbros with whole-rock compositions that vary from (518O of +2.0 to +6.0%o. These assimilated roof rocks would have had to become depleted in 18O by an earlier, high-temperature alteration involving heated seawater at high water/rock ratios, and thus, before stoping, such rocks undoubtedly would have contained abundant hydrous alteration minerals (amphibole, chlorite, epidote, etc.), making them readily susceptible to partial melting. Thus, assimilation and partial melting of such hydrothermally altered country rocks must play a fundamental role in the in situ generation of these ophiolitic plagiogranitic magmas.

Chemical characteristics A suite of samples was analysed for major and trace element chemistry at the Geological Survey of Canada (Table 4). An extensive discussion of the rock geochemistry is beyond the scope of this paper but will be provided separately. These

340

D. S. STAKES & H. P. TAYLOR, JR

Fig. 11. Plagioclase 618O v. quartz 618O for mineral separates from plagiogranites in northern Oman. The magmatie equilibrium 18O/16O fractionation between quartz and albite is expected to be A18O = 0.9~1.4%o. The narrow range of quartz 618O values should be noted. Two samples from Wadi Saq in the southern part of the Oman ophiolite (Gregory & Taylor 1981), shown for comparison, widen this range. New data from Alley plagiogranites in Wadi Jizi (see Stakes & Taylor, 1992) include OM81-2 with 618Oqtz +7.4%0; 618Opiag +16.0%0; OM81-4 with 618Oqtz +8.2%0; 6180plag +16.5%o.

samples were selected to identify the magmatie succession of mafic dykes based on the classifications for the volcanic stratigraphy. The three main eruptive and dyke series can be associated with broad changes in trace element chemistry. The earliest dykes and lavas (the Geotimes or VI/Ml series) have the highest concentrations of lithophiles. These lavas and dykes are characterized by TiC>2 contents greater than 1% and the highest total REE concentrations. The two subsequent eruptive units (the Lasail/V2/M2 series followed

by the Alley/V3/M3 magmatie series) contain lower light REE (LREE) and Ti, and higher H2O, etc., probably associated with a progressively depleted mantle source region. The TiOa contents for the Lasail lavas range from 0.4% to 1.2%, overlapping the values in the Geotimes series. The higher H^O contents of the Lasail and Alley lavas are reflected in crystalline rocks associated with the dykes that contain primary-appearing brown hornblende and trace amounts of orthopyroxene. Hydrous alteration tends to obscure the primary

ORIGIN OF PLAGIOGRANITES IN OMAN major element chemistry by increasing the relative Ca, Na and Ti contents. Figure 12 is a summary of trace element and REE variations for the dykes and gabbros. Most of the dykes have TiO2 contents between 0.4 and 2.0%, placing them into the Geotimes (VI/Ml) or

341

Lasail (V2/M2) magmatic series. The dykes with the highest TiO2 also contain higher total REE concentrations (see Fig. 12d), and these samples are also characterized by relatively flat REE patterns that vary from about 20 to 50 times chondritic values. Some dykes with high TiOa and

Fig. 12. (a) Zr/Y v. Mg number for all dykes and gabbros. (See Table 5 for data.) (b) Zr/Y v. TiO2 for all dykes and gabbros. (See Table 5 for data), (c) REE patterns for gabbros (TiO2 content shown in parentheses). High values of Nd represent analytical errors, resulting from high Sr interference; low Yb is result of alteration; 85-471 is a gabbro dyke from Assayab; 85-498, 85-536 and 85-538 are isotropic gabbros from Assayab; 85-459 is a hornblende gabbro from near the HLG in Assayab; 85-504 is a layered gabbro from Assayab; 85-506 and 85-509 are deep isotropic gabbros from Assayab; 90-635 is a foliated gabbro from Wadi Ragmi; 85-410, 85-434 are shallow isotropic gabbros from Bayda. Dashed line, high-level gabbro; continuous line, gabbro. (d) REE patterns for dykes (numbers in parentheses are TiO2 contents). OM 90-616 is a wide, coarse-grained dyke north of the Bayda mine and east of the Aarja-Bayda plagiogranite. OM85-403 is a large dyke from the west side of the plagiogranite. Sample OM90-638 is a basaltic andesite flow near 90-616. Samples OM 85-405, -413 and -419 are from the dyke complex that cuts the Bayda plagiogranite. Samples OM85-443, -525 and -463 are from Wadi Lasail west of the Lasail plagiogranite. Samples OM85-502 and 85-495 are from Assayab south of the Lasail plagiogranite. OM90-641 is from Wadi Fizh (west of Ragmi). OM85-516 is a dyke at Assayab near the SDC that is crosscut by OM85-539; OM85-539 and OM85-540 are late mafic dykes, the latter of which cuts a small plagiogranite body at the dike-gabbro contact.

342

D. S. STAKES & H. P. TAYLOR, JR

high FeO (Fe-Ti basalts?) contents were observed as swarms or crosscutting structures in the upper portions of the plutonic sequence in both the Lasail and Aarja-Bayda regions. These include some of the dykes that crosscut the highly mineralized pillows and massive flows adjacent to the plagiogranite. The dykes with lower TiO2 and REE contents tend to be slightly LREE depleted, and thus are more similar to MORE. All of these late dykes and the dykes that crosscut the plagiogranite body, as well as the dykes sampled from within the ore body, have TiOi contents below 1.00% and are probably associated with the 'Lasail' or V2/M2 event. Figure 13 depicts the chondrite-normalized REE patterns for 13 plagiogranites, along with two quartz diorites, and both a fresh and an altered marginal isotropic gabbro. The details of the sample locations and chemistry are provided

Fig. 12. (continued)

in Table 5. The REE patterns show broad variations from 5 to 50 times chondritic abundances, and they are generally flat or LREE depleted with small Eu and/or Ce anomalies (Fig. 13a). The REE patterns in the plagiogranites do not show the kinds of enrichments in the LREE that might be associated with extensive closed-system fractionation of MORB-type gabbro magmas. In fact, the inverse is observed: with one exception (81-80 from Suhaylah) the plagiogranite samples with higher concentrations of SiC>2 (i.e. >70wt.%; namely, 81-63, 81-30, 83-36, 83-173, 90-622 and 95GIA-11A) all show relatively smooth LREEdepleted patterns similar to the gabbros (but at somewhat higher concentrations, with La typically 5-20 times chondrite); they also tend to have lower overall enrichments in the REE than do the lower-SiO2 plagiogranites. The similarity in the shapes of the REE patterns of these smaller, high-

ORIGIN OF PLAGIOGRANITES IN OMAN

343

Fig. 13. (a) REE patterns for 13 plagiogranites having SiO2 >63.5 wt.%, and four other gabbros and quartz gabbros that have SiO2 of 52-57 wt.% (OM85-457 and 95GIA-28A, -28C and -34B). (See Tables 4 and 5 for localities and complete major and trace element chemistry.) It should be noted that six of the seven most SiO2-rich plagiogranites (>70 wt.%) all show LREE depletions (the only exception is OM81-80 from Suhaylah). Also, contrary to what might be expected from simple fractional crystallization, these higher-SiO2 plagiogranites also tend to have lower overall REE contents than the six plagiogranites with <70 wt.% SiO2. (b) REE patterns for plagiogranites and included mafic xenoliths. Plagiogranite OM85-412 and xenolith OM85-411 were collected from the same outcrop (locality 229), as also was xenolith OM85-445A and plagiogranite OM85-445B (locality 280). The striking similarities in the REE patterns between the plagiogranite bodies and their included mafic enclaves at these two outcrops should be noted. Less closely related are plagiogranite OM85-460 (from locality 315) and xenolith OM85-468 from the nearby locality 311, which lies 600 m away on the other side of the HLG body shown in Figure 2. Open symbols, xenoliths; filled symbols, plagiogranites.

silica bodies to that of the relatively unaltered isotropic gabbro (95GIA-28C), and the similarity of the latter to present-day samples of MORB, should be noted (e.g. Perfit et al 1998). With decreasing SiC>2 contents, the REE concentrations in the plagiogranites tend to increase, the LREE patterns flatten and the Eu anomalies become more distinct (Fig. 13a). Comparison with the REE patterns of altered isotropic gabbros (95GIA-34B and 85-410 near Aarja) and adjacent andesitic dykes (e.g. 85-405 and 90-616 adjacent

to the Bayda plagiogranite) suggests that in these lower-SiC>2 plagiogranites the REE patterns are more similar to the patterns of these country rocks and penecontemporaneous dykes. Thus, it is plausible that these REE patterns were inherited during the assimilation of these kinds of hydrothermally altered country rocks and/or during exchange with mafic enclaves intruded as basaltic magma 'pillows' into the plagiogranite melts. This is particularly well shown by the outcrops at localities 229 and 280, where the shapes and the concentrations

344

D. S. STAKES & H. P. TAYLOR, JR

of REE in the coexisting xenoliths and plagiogranite hosts are virtually identical (Fig. 13b).

Origin of plagiogranites A comprehensive survey of plagiogranite chemistry and mineralogy was carried by Aldiss (1978), comparing plagiogranites from the Troodos, Semail, Smartville and Point Sal ophiolites with samples from the Ballantrae complex, island arc complexes and rare dredge samples then available from the ocean crust. His data and observations highlight the many similarities of structure, lithology, mineralogy and tectonic setting to those that are detailed here for the plagiogranites of the Semail ophiolite. Although we disagree with the conclusions of Aldiss (1978) that plagiogranites are fundamentally the products of crystal fractionation processes alone, we agree with him that local saturation of the basaltic melt with water is a critical component in the formation of these oceanic plagiogranites. The paper by Aldiss (1978) was completed before the concepts of assimilation plus fractional crystallization (AFC) were widely accepted as being important in the formation of many SiOa-rich igneous differentiates (e.g. Taylor 1980; DePaolo 1981). Based on these concepts, Gregory & Taylor (1981) suggested that the large plagiogranite bodies in the Oman ophiolite are produced by partial melting or stoping of the hydrothermally altered country rocks into a differentiating magmatic body. Pallister & Knight (1981) demonstrated that the trace element chemistry of the Dasir plagiogranite was consistent with the partial assimilation model of Gregory & Taylor (1981), having inherited the REE pattern of the country rock. They concluded that only the smallest plagiogranite bodies could have formed solely by crystal fractionation. The above arguments probably prohibit a strictly crystal fractionation origin for all but the smallest of these bodies; therefore it is important to investigate the thermal requirements for extensive partial melting of crustal rocks to form the larger bodies. In this study, we have shown that most of the dykes adjacent to the plagiogranites and essentially all the dykes cutting or pillowed within the plagiogranitic intrusions have trace element compositions characteristic of the off-axis V2/M2 (Beurrier et al. 1989) magmas. This includes all of the mafic enclaves and xenoliths analysed for this study (except for one Suhaylah xenolith with gabbroic layering). The trace element compositions of dykes, xenoliths and netveined basaltic enclaves suggest that the majority of these mafic enclaves are in fact recrystallized late mafic dykes. The continual recharge of the evolving silicic magma chamber by mafic magmas

provides an environment conducive for development of an extreme end member in an RAFC process. The clear-cut field evidence for the existence of contemporaneous silicic and mafic magmas at a late stage in the magmatic history of the ophiolite indicates that such magmatic recharge would be readily available during the formation of the large plagiogranite intrusions, many of which are zoned to diorite and gabbro in their lower levels (Figs 5 and 14). Assimilation, partial melting, recrystallization and crystal fractionation are all viable and likely processes in such an environment, but it is clear that much of the heat required to dissolve or remelt large quantities of rock must be provided by the continued injection of off-axis basaltic magma. The volume of plagiogranite magma that can be produced is thus limited by the continued supply of these basaltic magmas, but we should note that most of these rocks at this stage will have cooled only to temperatures of 450-500 °C or higher, so that the additional heat requirements include mainly just the latent heat of fusion of the plagiogranite. It is important to realize that the latent heat of crystallization of basaltic magma is at least three times greater than that of hydrous granitic magmas (e.g. see data of Bohrson & Spera 2001; Spera & Bohrson 2001). Therefore, if we assume a specific heat of 1250 J kg"1 K"1 and a latent heat for basalt of 4.2 X 105 J kg"1, injection of 1 kg of basaltic magma at 1100 °C into a hydrous plagiogranite melt at 750 °C can produce (4.2 X 105J) + (350 X 1250J = 4.2 X 105J), for a total of 8.6 X 105 Jkg" 1 . It should be noted that 3.75 X 105 Jkg" 1 is sufficient to heat the country rocks from 450 to 750 °C, and only an additional 1.5 X 105 Jkg" 1 is required to produce a plagiogranite melt from such rocks, for a total of 5.25 X 105 Jkg" 1 . Thus 1 kg of basaltic magma can, in theory, produce 8.6/5.25 = 1.64kg of plagiogranite melt. These heat balance calculations underscore the critical role that the contemporaneous mafic dykes play in the development of the large plagiogranites and their associated ore bodies in Oman. Recently published quantitative models (Bohrson & Spera 2001; Spera & Bohrson 2001) provide support for these qualitative observations and simple heatbalance calculations (see Fig. 14). Bohrson & Spera assumed that most intrusions can be modelled as RAFC processes, and that thermal and geochemical considerations permit estimates of the extent of magma contamination by anatexis of country rocks. The Spera & Bohrson (2001) model provides a theoretical context for the observed variations in both oxygen isotopes and trace element data. Their model assumes that during the cooling of a magma body, heat is

ORIGIN OF PLAGIOGRANITES IN OMAN

345

Fig. 14. Modification of Figure 1 of Spera & Bohrson (2001) with field parameters based on the Aarja-Bayda plagiogranite body in northern Oman. Model assumptions are that plagiogranite magma forms the upper portion of the igneous body, with a liquidus of 1000 °C and a solidus of 750 °C. The country rocks are hydrated mafic oceanic crustal rocks with a liquidus of 1100 °C and a solidus of 950 °C. The intrusion is recharged by mafic lava (the late mafic dykes) with an initial temperature of 1200 °C, identical to the original temperature of the basalt primary melt. Field observations suggest that late mafic dykes can both chemically exchange with the magma to form xenoliths and quench within the magma to form mafic 'pillows'. The example provided in Figure 6a and b suggests that for a single dyke, the outer portion may recrystallize whereas the inner portion quenches.

transferred to the adjacent country rock. When the country rock (or xenoliths) is heated above the temperature of anatexis, the magma becomes increasingly contaminated by anatectic melt. Recharge will add additional heat to the magma, increasing the degree of anatexis and assimilation. In their model, incompatible elements (such as the REE) will become increasingly diluted by the anatectic melt such that the final evolved magma is lower in the incompatible elements than would be predicted by fractional crystallization alone. This is exactly what is observed for the large plagiogranites. Similarly, the continued assimilation of low-18O rocks would decrease the (518O of the contaminated magma even as the magma becomes increasingly fractionated. This again is what is observed for the plagiogranites.

Variations observed in trace elements, especially the rare earth analyses, are qualitatively consistent with predictions based on the Spera & Bohrson (2001) model, in that the large plagiogranite bodies typically display complementary chemistries to the lower diabasic dykes and variegated (high-level) gabbros that host them. For all the plagiogranites included in this study, the adjacent or closely associated high-level gabbros and dolerites are characterized by medium to low total REE concentrations with flat to slightly LREE-depleted compositions, as well as with positive Eu anomalies that can be attributed to plagioclase accumulation. In contrast, the plagiogranites themselves have moderate to high total REE, flat to dished patterns and negative Eu anomalies. Only the smaller plagiogranite segrega-

346

D. S. STAKES & H. P. TAYLOR, JR

tions are preferentially enriched in the LREE. Thus, whereas the smaller plagiogranite segregations could have formed by simple fractional crystallization the larger intrusions seem to require significant interaction with, and assimilation of, the hydrothermally altered and hydrated roof rocks of the oceanic crust, just as is required by the oxygen isotope evidence described above.

Relationship between plagiogranites and ore deposits The metamorphism and recrystallization of diabasic and/or gabbroic country rocks by the plagio-

granite-gabbro bodies drives local, high-level hydrothermal systems that transport Fe and Cu scavenged from the mafic minerals within the intrusion into the overlying crustal rocks. Coarsegrained dykes and gabbros from the adjacent crustal rocks are depleted in Fe, K and Cu, and enriched in H2O and Ca. These chemical changes are a result of the replacement of igneous pyroxene by chlorite and amphibole, ilmenite by titanite, and calcic plagioclase by albite, epidote and quartz. These epidote-quartz-titanite assemblages are strikingly similar to the epidosites described in the northern Troodos ophiolite by Schiffman & Smith (1988). Figure 15 is a schematic depiction of the structural relationship between such a

Fig. 15. Schematic depiction of the structural relationships between a plagiogranite body and its overlying ore deposit. Noteworthy features are the epidote- and chlorite-rich stockwork above the plagiogranite, abundance of late, partially recrystallized dykes and late dykes crosscutting the ore bodies (as the 'ore body sheets'). The plagiogranites, dykes and ore bodies appear to be associated with faults that parallel the spreading ridge axis.

ORIGIN OF PLAGIOGRANITES IN OMAN plagiogranite body and an overlying massive sulphide deposit, based on the field relationships observed in northern Oman. Surrounding the narrow discharge zone of chlorite, sulphide and epidote is a broad alteration zone dominated by epidote, titanite, albite and quartz. The replacement of calcic plagioclase by albite and clinopyroxene by amphibole would release excess Ca and Al into hydrothermal fluids to produce the observed epidote aureole. The consistent association of epidote and sulphides implicates this mineralogical transformation in the transport of metals to the circulating fluids that form the ore deposits. Such epidosites are depleted in base metals (Cu, Zn, etc.) compared with basalt, and thus their precursors are likely source rocks for the hightemperature Cu-rich fluids (Schiffman et al. 1990). This observation is supported by hydrothermal experimental data that demonstrate the importance of the presence of epidote-magnetite pyrite-anhydrite within the sub-sea-floor reaction zone to produce the appropriate fluid pH and redox conditions for Cu mobilization (see Seyfried et al. 1991; Seyfried & Ding 1993, for a review). In addition, the vein assemblages of epidotepyrite-magnetite-hematite are consistent with the experimental assemblages controlling the Fe/Cu ratio of the hydrothermal fluid. The hydrothermally metamorphosed isotropic gabbros and late dykes, whose mineral assemblages are composed of sodic plagioclase, amphibole, chlorite and epidote, provide a source of H/zO for the plagiogranite magmas, by their incorporation as xenoliths. This hydrous component is important during the evolution and differentiation of the silicic magma. Within the plagiogranite bodies, recrystallized xenoliths are composed of albite - quartztitanite, with lesser quantities of epidote-magnetite. Early phases recrystallized within the magma (e.g. sodic plagioclase, amphibole, diopside, magnetite) may subsequently be hydrothermally altered to epidosite mineralogies, presumably releasing more Cu and other metals into the hydrothermal system. Both the Bayda and Lasail ore deposits are emplaced into massive basalt flows and pillows crosscut by late felsic dykes or sheets. The mineralized units are non-magnetic and extensively chloritized. Within the Bayda mine shaft, the most Cu-rich mineralization lies directly adjacent to a fault cutting a unit of massive basalt interlayered with extensively mineralized pillow basalts. Silicic dykes and sheets around the Lasail and Bayda deposits confine the most Cu-rich portions of these deposits. This suggests that the final magmatic phase of plagiogranite formation (the injection of silicic dykes and sheets into the overlying pillow lavas) was contemporaneous with the loca-

347

lization of the major ore deposits and the concentration of the most Cu-rich ore. The locations of such felsite sheets, referred to as 'ore body sheets' by geologists from the Oman Mining Company, were utilized to identify potentially ore-rich zones during mining operations (P. Kinto, pers. commun.). Although the felsic sheets are themselves barren of mineralization, and their chemical compositions are substantially altered from their primary values, their low Ti and low total REE contents suggest an association with the late mafic dykes. In this connection, it should be noted that the margins of the late mafic dykes that crosscut the plagiogranite body at Aarja-Bayda are highly enriched in Cu (2500 ppm) compared with the dyke interiors (4 ppm), once again implicating these late intrusions in the mobilization and concentration of the Cu-sulphides. The ore deposits and plagiogranites in the Lasail mining district are located within a ridgeparallel graben that Boudier et al. (2000) have suggested formed at the tip of a propagating rift located to the north. Such an interpretation is consistent with the deep, high-temperature seawater penetration, strong 18O depletions, and alteration of gabbro dykes in Wadi Ragmi described by Stakes & Taylor (1992). These gabbro dykes penetrate the entire crustal section in the Wadi Ragmi area that lies to the north of the areas described in the present study (Fig. 1). Hightemperature seawater penetration to great crustal depths is evident in the striking 18O depletions ((518O as low as +1.8%o) observed in this area, and is consistent with a younger magmatic system disrupting slightly older oceanic crust by formation of a new spreading centre. The deep gabbro dykes have created a zone of extreme hydrothermal alteration within the gabbroic lower crust exposed in Wadi Ragmi. Rift propagation is a common phenomenon at fast-spreading mid-ocean ridges and is the mechanism by which oceanic plates accommodate changes in plate motion. In the modern crust, however, the lower crust is not generally accessible to geological observation and the magmatic and hydrothermal processes must be inferred from the composition and distribution of volcanic rocks and hydrothermal deposits. Based on extensive submersible mapping on the Galapagos propagating rift, Perfit et al. (1998) suggested an RAFC model to explain the extraordinarily fractionated and differentiated igneous rocks (from MORB to andesite). This model, presented as Figure 16, may provide a modern oceanic analogue for both the large plagiogranite bodies observed in Oman and their associated ore deposits (Ridley et al. 1994). The presence of Fe-Ti basalts and/or andesites in the upper volcanic section adjacent to Bayda (e.g. OM90-623) is

348

D. S. STAKES & H. P. TAYLOR, JR

Fig. 16. Proposed crustal model of the complex multiple magmatic systems for the modern Galapagos propagating rift from Perfit et al. (1998). The presence of a high-level silicic magma chamber is inferred, based upon the presence of andesites erupted slightly off-axis. The association of the high-level intrusion is also directly associated with ore mineralization defined by exposed stockworks (Ridley et al. 1994) and proposed subsurface plagiogranites. In this figure, the primitive MORE magma body that erupts within the neovolcanic zone is surrounded by a partially crystalline transition zone, which cools to form a gabbroic layer. The sheeted dyke complex and the volcanic rocks overlie the gabbros and are crosscut by late mafic dykes of the propagating rift. An evolved high-level magmatic centre coexists on the flanks of this central magmatic system. It is composed of Fe-Ti basalt to andesite that is contaminated by assimilation of up to 15% hydrothermally altered oceanic gabbros and dykes. Fractures above this high-level magmatic system allow the downflow of seawater and the shallow focused hydrothermal mineralization. Late andesite dykes intrude and volcanic rocks erupt proximal to the ore deposits.

consistent with this model. As pointed out by Boudier et al. (2000), the lower-crustal section of the Galapagos propagating rift exposed on the Hess Deep contains interlayed gabbro and gabbronorite, similar limologies to those observed in northern Oman.

Conclusions The chemical, isotopic and field relationships described above for the Semail ophiolite of northern Oman demonstrate the existence of multiple superimposed magmatic and hydrothermal events. Based on (518O values and mineral compositions, the most pervasive and highest temperature hydrothermal reactions are present in the isotropic gabbros, regardless of whether these occur in the vicinity of the SDC-gabbro boundary in the middle to upper part of the ophiolite lithostratigraphy, or as deep crosscutting bodies (DIG). The gabbros and diorites associated with the large plagiogranite bodies are particularly implicated in

these hydrothermal episodes. The diagnostic metamorphic assemblage is sodic plagioclase (An2o30) and actinolitic or magnesio-hornblende, suggesting temperatures of at least 400 °C. Both feldspar and amphibole are locally strongly depleted in 18O (to (518O values <+4%o) by the heated seawater. The abundance of metagabbroic assemblages adjacent to and included within the plagiogranites suggests that these silicic intrusions are the loci of the latest-stage, high-temperature seawater alteration events within the ophiolite. Detailed mineralogical, petrographic, geochemical and isotopic examination of the plagiogranites and their crosscutting dykes as well as their included mafic enclaves and xenoliths provide evidence of recharge, assimilation and fractional crystallization (RAFC) processes. Notable examples of the partial digestion and re-equibration of xenoliths include microgranophyric intergrowths of albite-quartz as coronas on tabular plagioclase and symplectic intergrowths of titanite and epidote. A ubiquitous component of all the plagiogra-

ORIGIN OF PLAGIOGRANITES IN OMAN nites are mafic enclaves with a variety of origins. These include mafic 'pillows' formed by the quenching of basaltic magma that has intruded into the cooler, more silicic plagiogranite magmas. This observation clearly points to the presence of contemporaneous mafic and plagiogranitic magmas. The continued addition of basaltic magmas to these latest-stage magma bodies in the ophiolite provides sufficient heat to the silicic magma to allow for extensive assimilation of previously altered, hydrous country rocks, particularly those from the overlying sheeted dyke complex (SDC). There is substantial field and laboratory evidence for the assimilation of hydrous mafic rocks as a fundamental process in the formation of these H2O-rich plagiogranite magmas, which probably remained partially liquid down to temperatures of c. 750 °C. This evidence includes the mafic pillows and recrystallized xenoliths, the variation in (518O of the magmatic quartz within the plagiogranites, and the inheritance of trace element and REE compositions from the assimilated materials. The pervasive replacement of clinopyroxene by epidote, of ilmenite by titanite, of calcic plagioclase by albite, and the dissolution of magnetite are consistent with the hydrothermal fluids observed to form sulphide deposits on the modern sea floor. Such metamorphic assemblages (epidosites) are ubiquitous near the SDC-gabbro boundary in the ophiolite, where they can commonly be linked to small fracture-hosted massive sulphide deposits within the sheeted dyke complex or the lower pillow lavas. The large plagiogranite bodies and their associated gabbroic and SDC country rocks, together with the contemporaneous late mafic dykes and net-veined enclaves that intrude them, provide a plausible shallow source of both heat and metals for the formation of the larger massive sulphide ore deposits. Our studies demonstrate that detailed future detailed work on ophiolites can continue to provide a wealth of information about oceanic crustal processes, both ancient and modern. J. Smewing, A. Shelton, R. Koski and R. Ressetar provided valuable field advice over the years. A. Gough produced many versions of the figures in this paper. K. Lima and J. Kela are acknowledged for assistance with the production of the isotopic and chemical tables. C. Gregoire, P. Belanger, I. Jonnason and the staff at GSC Analytical Chemistry Research Laboratories, Mineral Resources Division, Ottawa, are thanked for the XRF and ICP analyses. H. Jones from University of Michigan, L. Ember from University of South Carolina and J. Paduan assisted with isotopic analyses and sample preparation. Microprobe analyses were performed by D.S.S. at the University of South Carolina or by P. Schiffman at UC Davis. We thank A. Galley for

349

graciously providing us with his unpublished field and chemical data from Aarja. Fieldwork and analytical work were provided by NSF grants OCE80-19021, EAR-831306, EAR-88-16413, EAR-97-25811 and EAR-0106696. Additional analytical work and support for D.S.S. was provided by funds to MBARI from the Lucile and David Packard Foundation.

References ALABASTER, T. & PEARCE, J.A. 1985. The interrelationship between magmatic and ore-forming hydrothermal processes in the Oman Ophiolite. Economic Geology, 80, 1-16. ALABASTER, T., PEARCE, J.A. & MALPAS, J. 1982. The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex. Contributions to Mineralogy and Petrology, 81, 168-183. ALDISS, D.T. 1978. Granitic rocks of ophiolites. PhD dissertation, Open University, Milton Keynes. BEDARD, J.H., HEBERT, R., BERCLAZ, A. & VARFALVY, V. 2000. Syntexis and the genesis of lower oceanic crust. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 105-119. BEURRIER, M., OHNENSTETTER, M., CABANIS, B., LESCUYER, J.L., TEGYEY, M. & LE METOUR, J. 1989. Geochimie des filons doleritiques and des roches volcaniques ophiolitiques de la nappe de Semail: constrainte sur leur origine geotectonique au Cretace superieur. Bulletin de la Societe Geologique de France, 8, 205-219. BOHRSON, W.A. & SPERA, F.J. 2001. Energy-constrained open-system magmatic processes II: Application of energy-constrained assimilation fractional crystallization (EC-AFC) model to magmatic systems. Journal of Petrology, 42, 1019-1041. BOUDIER, F., GODARD, M. & ARMBRUSTER, C. 2000.

Significance of gabbronorite occurrence in the crustal section of the Semail ophiolite. Marine Geophysical Researches, 21, 307-326. BROWNING, P. 1984. Cryptic variation within the cumulate sequence of the Oman ophiolite: magma chamber depth and petrological implications. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 14, 71-82. CLAYTON, R.N., GOLDSMITH, J.R. & MAYEDA, T.K. 1989. Oxygen isotope fractionation in quartz, albite, anorthite and calcite. Geochimica et Cosmochimica Ada, 42, 613-621. COLEMAN, R.G. & HOPSON, C.A. (eds) 1981. The Samail Ophiolite, Oman. Journal of Geophysical Research, Special Issue, 86. COLEMAN, R.G. & PETERMAN, Z.E. 1975. Oceanic plagiogranite. Journal of Geophysical Research, 80, 1099-1108. DEPAOLO, D.J. 1981. A neodymium and strontium isotopic study of the Mesozoic calc-alkaline granitic batholiths of the Sierra Nevada and Peninsula Ranges, California. Journal of Geophysical Re-

350

D. S. STAKES & H. P. TAYLOR, JR

search, 86, 10470-10488. ERNEWEIN, M, PFLUMIO, C. & WHITECHURCH, H. 1988. The death of an accretion zone as evidenced by the magmatic history of the Sumail ophiolite (Oman). Tectonophysics, 151, 247-274. GALLEY, A.G. & KOSKI, R.A. 1995. Contrasting plagiogranite suites and their relationship to massive suphide formation. American Geophysical Union, Program with Abstracts, F450. GREGORY, R.T. & TAYLOR, H.P. JR 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: evidence for 618O buffering of the oceans by deep (>5 km) seawaterhydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research, 86, 2737-2755. HAYMON, R.M., KOSKI, R.A. & ABRAMS, M.J. 1989. Hydrothermal discharge zones beneath massive sulfide deposits mapped in the Oman ophiolite. Geology, 17,531-535. HOLLAND, T. & BLUNDY, J. 1994. Non-ideal interactions in calcic amphiboles and their bearing on amphibole-plagioclase thermometry. Contributions to Mineralogy and Petrology, 116, 443—447. JUTEAU, T., ERNEWEIN, M., REUBER, I., WHITECHURCH, H. & DAHL, R. 1988. Duality of magmatism in the plutonic sequence of the Sumail nappe, Oman. Tectonophysics, 151, 107-135. JUTEAU, T., MANAC'H, G., MOREAU, O., LECUYER, C. & RAMBOZ, C. 2000. The high-temperature reaction zone of the Oman ophiolite: new field data microthermometry of fluid inclusions, PIXE analyses and oxygen isotopic ratios. Marine Geophysical Researches, 21, 351-385. KOSKI, R.A., GALLEY, A.G., GIBSON, H.I. & GOULDING, H. 1995. Field studies in northern Oman 1995. Ministry of Petroleum and Minerals, Sultanate of Oman, 18 pp. KOSKI, R.A., HAYMON, R.M. & STAKES, D.S. 1990. Plagiogranites, hydrothermal discharge, and sulfide deposits in the Sohar Mining District, northern Oman, Excursion 3-Sypmosium on ophiolite genesis and evolution of oceanic lithosphere. Ministry of Petroleum and Minerals, Sultanate of Oman, 21 pp. Liou, J.G., KUNIYOSHI, S. & ITO, K. 1974. Experimental studies of the phase relations between greenschist and amphibolite in a basaltic system. American Journal of Science, 274, 613-632. Liou, J.G., KIM, H.S. & MARUYAMA, S. 1983. Prehniteepidote equilibria and their petrologic applications. Journal of Petrololgy, 24, 321-342. LIPPARD, S.J. (ed.) 1980. Open University Oman Project Map Sheet 2—The Wadi Jizzi Area (1:100000). Directorate of Overseas Surveys, UK. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Geological Society of London, Memoir, 11. MARUYAMA, S., SUZUKI, K. & Liou, J.G. 1983. Greenschist-amphibolite transition equilibria at low pressure. Journal of Petrology, 24, 583-604. MUEHLENBACHS, K. & BYERLY, G. 1982. O-enrichment of silicic magmas caused by crystal fractionation at the Galapagos Spreading Center. Contributions to Mineralogy and Petrology, 79, 76-79. MUEHLENBACHS, K. & CLAYTON, R.N. 1971. Oxygen

isotope ratios of submarine diorites, and their constituent minerals. Canadian Journal of Earth Sciences, 8, 1591-1595. NEHLIG, P. & JUTEAU, T. 1988. Flow porosities, permeabilities and preliminary data on fluid inclusions and fossil thermal gradients in the crustal sequence of the Sumail ophiolite, Oman. Tectonophysics, 151, 199-221. NEHLIG, P., JUTEAU, T., BENDEL, V. & COTTON, J. 1994. The root zones of oceanic hydrothermal systems: constraints from the Semail ophiolite (Oman). Journal of Geophysical Research, 99, 4703-4713. NICOLAS, A., REUBER, I. & BENN, K. 1988. A new magma chamber model based on structural studies in the Oman ophiolite. Tectonophysics, 151, 87-105. NICOLAS, A., BOUDIER, F. & ILDEFONSE, B. 1994. Evidence from the Oman ophiolite for active mantle upwelling beneath a fast-spreading ridge. Nature, 370,51-53. NICOLAS, A., BOUDIER, F., ILDEFONSE, B. & BALL, E. 2000. Accretion of the Oman ophiolite in a microplate system—discussion of a new structural map. Marine Geophysical Researches, 21, 147-179. O'NEIL, J.R. & TAYLOR, H.P. JR 1967. The oxygen isotope and cation exchange chemistry of feldspars. American Mineralogist, 52, 1414-1437. PALLISTER, J.S. & HOPSON, C.A. 1981. Samail ophiolite plutonic suite: field relations, phase variations, cryptic variation and layering, and a model of a spreading ridge magma chamber. Journal of Geophysical Research, 86, 2661-2672. PALLISTER, J.S. & KNIGHT, RJ. 1981. Rare-earth element geochemistry of the Samail ophiolite near Ibra, Oman. Journal of Geophysical Research, 86, 2661-2672. PEARCE, J.A., ALABASTER, T., SHELTON, A.W. & SEARLE, M.P. 1981. The Oman ophiolite as a Cretaceous arc-basin complex: evidence and implications. Philosophical Transactions of the Royal Society of London, Series A, 300, 299-317. PERFIT, M.R., RIDLEY, W.I. & JONASSON, I. 1998. Geologic, petrologic and geochemical relationships between magmatism and massive sulfide mineralization along the eastern Galapagos Spreading Center. Reviews in Economic Geology, 8, 75-100. PITCHER, W.S. 1993. The Nature and Origin of Granite. Blackie, Glasgow. REUBER, I. 1988. Complexity of the crustal sequence in the northern Oman ophiolite, (Fizh and southern Aswad blocks): the effects of early slicing? Tectonophysics, 151, 137-165. RIDLEY, W.I., PERFIT, M.R., JONASSON, I.R. & SMITH, M.F. 1994. Hydrothermal alteration in oceanic ridge volcanics: a detailed study at the Galapagos fossil hydrothermal field. Geochimica et Cosmochimica Acta, 58, 2477-2494. SCHIFFMAN, P. & SMITH, B.M. 1988. Petrology and oxygen-isotope geochemistry of a fossil seawater hydrothermal system within the Solea graben, northern Troodos Ophiolite, Cyprus. Journal of Geophysical Research, 93, 4612-4624. SCHIFFMAN, P., BETTISON, L.A. & SMITH, B.M. ET AL.

ORIGIN OF PLAGIOGRANITES IN OMAN 1990. Mineralogy and geochemistry of epidosites from the Solea Graben, Troodos ophiolite, Cyprus. In: MALPAS, J., MOORES, E.H., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites-Oceanic Crustal Analogues, Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 673-683. SEYFRIED, W.E. JR & DING, K. 1993. The effect of redox on the relative solubilities of copper and iron in Cl-bearing aqueous fluids at elevated temperatures and pressures: an experimental study with application to subseafloor hydrothermal systems. Geochimica et Cosmochimica Acta, 57, 1905-1917. SEYFRIED, W.E. JR, DING, K. & BERNDT, M.E. 1991. Phase equilibria constraints on the chemistry of hot spring fluids at mid-ocean ridges. Geochimica et Cosmochimica Acta, 55, 3559-3580. SMEWING, J.D. 1981. Mixing characteristics and compositional differences in mantle-derived melts beneath spreading axes: evidence from cyclically layered rocks in the ophiolite of North Oman. Journal of Geophysical Research, 86, 2645-2659. SMEWING, J.D. 1990 Field Guide to Excursion E4: Sections through the Northern Ophiolite: Wadi Rajmi and Wadi Fizh, Symposium on ophiolite genesis and evolution of oceanic lithosphere. Ministry of Petroleum and Minerals, Sultanate of Oman, Muscat, Jan. 1990. SMEWING, J.D., SIMONIAN, K.O., EL BOUSHI, L.M. & GASS, I.G. 1977. Mineralized fault zone parallel to the Oman ophiolite spreading axis. Geology, 5, 534-538. SMEWING, J.D., CHRISTENSEN, N.I., BARTHOLOMEW, I.D. & BROWNING, P. 1984. The structure of the oceanic upper mantle and lower crust as deduced from the northern section of the Oman ophiolite. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 14, 41-54. SPEAR, F.S. 1980. NaAl = CaAl exchange equilibrium between plagioclase and amphibole: an empirical model. Contributions to Mineralogy and Petrology, 72, 33-41. SPERA, FJ. & BOHRSON, W.A. 2001. Energy-constrained open-system magmatic processes I: General model and energy-constrained assimilation and fractional crystallization (EC-AFC) formulation. Journal of

351

Petrology, 42, 999-1018. STAKES, D.S. 1991. Oxygen and hydrogen isotope compositions of oceanic plutonic rocks: high temperature deformation and metamorphism of oceanic layer 3. In: TAYLOR, H.P. JR, O'NEIL, J.R. & KAPLAN, I. (eds) Stable Isotope Geochemistry: a Tribute to Samuel Epstein. Geochemical Society, Special Publication, 3, 77-90. STAKES, D.S. & O'NEIL, J.R. 1982. Mineralogy and stable isotope geochemistry of hydrothermally altered oceanic rocks. Earth and Planetary Science Letters, 57, 285-304. STAKES, D.S. & TAYLOR, H.P. JR 1992. The northern Samail ophiolite: an oxygen isotope, microprobe, and field study. Journal of Geophysical Research, 97, 7043-7080. TAYLOR, H.P. JR 1968. The oxygen isotope geochemistry of igneous rocks. Contributions to Mineralogy and Petrology, 19, 1-17. TAYLOR, H.P. JR 1974. The application of oxygen and hydrogen isotope studies to problems of hydrothermal alteration and ore deposition. Economic Geology, 69, 843-883. TAYLOR, H.P. JR 1977. Water-rock interactions and the origin of H2O in granitic batholiths. Journal of the Geological Society, London, 133, 509-558. TAYLOR, H.P. JR 1980. The effects of assimilation of country rocks by magmas on 18O/16O and 87Sr/86Sr systematics in igneous rocks. Earth and Planetary Science Letters, 47, 243-254. TAYLOR, H.P. JR 1997. Oxygen and hydrogen isotope relationships in hydrothermal mineral deposits. In: BARNES, H.L. (ed.) Geochemistry of Hydrothermal Ore Deposits, 3rd. Wiley, New York, 229-302. TAYLOR, H.P. JR & SHEPPARD, S.M.F. 1986. Igneous rocks: I. Processes of isotopic fractionation and isotope systematics. In: VALLEY, J.W., TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotopes in High-Temperature Geological Processes. Mineralogical Society of America, Reviews in Mineralogy, 16,227-271. VETTER, S. & STAKES, D.S. 1990. The northern Semail plutonic suite: field and trace element evidence for repeated magma injection in the construction of back-arc crust. In: MALPAS, J., MOORES, A., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites—Ocean crustal analogoues, Proceedings of the Symposium 'Troodos 1987'. Geological Survey, Nicosia, 397-412.

This page intentionally left blank

Ophiolites and global geochemical cycles: implications for the isotopic evolution of seawater R. T. GREGORY Stable Isotope Laboratory and Department of Geological Sciences, Southern Methodist University, PO Box 750395, Dallas, TX 75275-0395, USA (e-mail: [email protected]) Abstract: Isotopic profiles through ophiolite complexes provide the necessary link between the study of global geochemical cycles and plate tectonics. The hydrothermal circulation that occurs beneath the sea floor is the primary mechanism for exchange between the mantle of the Earth and the hydrosphere. The subduction of the hydrothermally altered crust and overlying sediments is the primary mechanism for crustal recycling. Oxygen and strontium isotopes of seawater track the competition between continental weathering and mid-ocean ridge hydrothermal exchange to control the composition of the oceans. Information derived from ophiolite studies on the elemental fluxes and the depth of seawater penetration into the oceanic crust provides constraints on the important rate constants associated with these competing processes. The same tectonic rates that account for the Sr isotope evolution of seawater indicate that the oxygen isotopic composition of the ocean is constrained to vary within narrow limits (per mil level). Isotopic analysis of dredge samples and ophiolite complexes demonstrates that seawater-ocean crust interactions result in a oxygen isotopic zonation of the oceanic crust with masses (concentration times volume) centred on the initial isotopic composition of the crust. This requires that the oxygen isotopic composition of the ocean resides at near steady-state conditions over Earth history. The inferences from ophiolite complexes contrast strongly with the results of measurements on carbonates from epicontinental seaways, particularly for the Palaeozoic. Ophiolites and greenstone belts track exchange processes between the ocean and the igneous crust whereas most carbonate measurements track the surface ocean on continental shelves. For oxygen isotopes, the mass of epicontinental seaways and the rates of meteoric water input suggest a resolution to the controversy that accounts for both data sets.

Isotopic studies of ophiolites (e.g. Muehlenbachs 1986, 1998; Alt & Teagle 2000) provide direct evidence of the interaction between seawater and oceanic crust and lithosphere. Mid-ocean ridges, as well as other centres of igneous activity, are heat engines that drive massive amounts of seawater throughout the oceanic crust. Black smoker hot spring vents are a direct manifestation of the geomermal energy available at mid-ocean ridges (e.g. Von Damm 1990). The chemical and isotopic composition of the seawater is dynamically maintained by the competition between the chemical weathering input to the oceans and the exchange between seawater and mantle-derived rocks at mid-ocean ridges (Muehlenbachs & Clayton 1976; Gregory & Taylor 1981; Holland 1984: Spencer & Hardie 1990; Gregory 1991; Hardie 1996, 1998; Holland et al 1996; Muehlenbachs 1998). Plate tectonic cycles provide a mechanism for chemical exchange between the mantle of the Earth and the lithosphere, hydrosphere and atmosphere (e.g. Gregor et al. 1988). On a planetary scale, the persistence of water in liquid, solid and

gas forms at Earth surface conditions over recorded geological time has profound implications for the distinctive evolution of the Earth when compared with other terrestrial planets. Ophiolites, no matter where they form, are easily accessible places to study the time-integrated effect of water-rock interaction between seawater and the oceanic crust. Ophiolite sequences consisting of pillow lavas, sheeted dyke complexes and gabbroic plutonic complexes form by extension of the lithosphere producing new crust by intrusion of magma into volcanic rift zones. This may occur in several sea-floor settings, the most important of which is the mid-ocean ridge setting, where greater than 60% of the crust of the Earth forms. Ophiolites allow for the examination of complete sections through oceanic crust and upper-mantle lithosphere to assess depth of seawater penetration and time-integrated fluxes from isotopic profiles through crustal sections, and to gain information about the starting compositions of material that enters subduction zones. The purpose of this paper is to review some of the key results of ophiolite studies as they relate

From: DiLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 353-368. 0305-8719/037$ 15 © The Geological Society of London 2003.

354

R. T. GREGORY

to the isotopic evolution of seawater, with particular emphasis on the isotopic profiles through the Samail ophiolite. The Samail ophiolite complex is one of the best examples of a well-preserved section of fossil oceanic lithosphere, and occurs as the structurally highest pre-Tertiary package of rocks in the 700 km long Oman Mountains. It provides a natural laboratory for studying the effects of seawater exchange with the crust including sources and sinks for important major elements and the depth of penetration of hydrothermal fluid into the crust. Comparisons of measured profiles through ophiolitic crustal sections with concentrations preserved in fresh basalt and gabbro yield data about elemental sources and sinks for seawater. The volumes of crust and water involved in these oceanic processes have rate constants that are much shorter than the age of the Earth so that the chemical composition of seawater resides near a limited range of quasi-steady-state compositions.

weathering. Even the longest residence time for a soluble element such as Na is short compared with the age of the Earth.

Boundary and initial conditions for hydrothermal activity Ophiolitic studies combined with modern marine data including seismic data, samples dredged from the sea floor, and more limited data from drill holes provide the constraints on the geometry of accretion of oceanic crust. In addition to the contrast in chemical composition between seawater and basalt, for the purposes of considering fluid-rock interaction on the sea floor, the important constraints are the contrast in temperature between the ocean and the magmas emplaced on the sea floor, and the geometric constraints imposed by the extension of the lithosphere.

Temperatures of interaction Ophiolites and global cycles: seawater composition The chemical and isotopic composition of seawater is the result of all of the interactions between fluid and rock where the resulting fluid finds its way back into the ocean reservoir that constitutes nearly all of the water present on the surface of the Earth. Seawater chemically exchanges with rocks made available for interaction by the rock cycle and therefore interacts with rocks directly derived from the mantle of the Earth, the continental crust, and with sediments, many of which are the products of previous precipitation from seawater. As such, seawater is a solution that is in chemical equilibrium with no particular rock. Each chemical species in the ocean is the result of the competition between processes that enrich its concentration in seawater and those that deplete its concentration in seawater. The residence time of an element is its abundance in the ocean divided by its total flux into or out of the ocean. Residence times vary by several orders of magnitude, from hundreds of millions of years to thousands of years, and depend on the solubility of the element and its abundance in rocks. The failure of calculations using the salt content of the oceans to date the Earth, at the end of the nineteenth century, indicates something about the time scale for fluid-rock interaction within the crust (Earth 1962). From plate tectonics and ophiolite complexes, we now know the importance of sea-floor processes as a sink for elements delivered to the oceans through chemical

Basalts are the most abundant igneous rocks in the Solar System. Evidence from meteorites and lunar samples indicates that production of basalt has occurred throughout the history of the Solar System on the terrestrial planets, dating back to times not preserved in the Earth record. With the exception of Earth, where crustal compositions appear to be bimodal (sima > sial), basaltic crusts dominate the other terrestrial planets so that from the beginning, seawater has interacted with basaltic crusts. Experimental studies constrain the solidus temperatures of basalts to within a range of 100 °C of 1200 °C (e.g. Kushiro 2001). Initial magma temperatures for mid-ocean ridge basalt (MORE) source regions are probably higher than 1400°C (Herzberg & O'Hara 1998), constraining, along with the latent heat content of the magma, the available heat for driving hydrothermal activity (e.g. Wolery & Sleep 1976). The existence of pillow lavas in greenstone terranes from some of the oldest Archaean successions on the Earth indicates eruption of basalt into liquid water. The presence of liquid water on the surface of the Earth and the ability of the mantle of the Earth to consistently deliver basalt to the surface of the Earth constrain the average contrast in temperature between magma and seawater reservoirs to be similar over Earth history.

Geometric constraints of accretion The average temperature of interaction between circulating seawater fluid and oceanic crust depends upon the geometry of accretion. Ophiolites such as the Samail ophiolite complex with later-

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES ally continuous sheeted dyke complexes require nearly 100% extension in order to form. This type of extensional setting (coupled with the initial primary permeability of the volcanic crust) provides the permeability necessary to allow for the circulation of seawater into the crust. The physical properties of water drive the circulation system towards the critical point of water (Norton 1984). The existence of permeability and a fracture system connected to the surface keep the fluidrock interaction confined to a hydrostatically pressured system. This is in contrast to the magma chambers and crystal mush systems that reside at lithostatic pressure. Studies of layered gabbro complexes such as the Skaergaard intrusion (e.g. Taylor & Forester 1979) clearly show that there is little direct infiltration of hydrothermal fluid into the magma chamber as long as it remains above the solidus. The heat content of the magma and the permeability contrast between the magma chamber and

355

rocks below the solidus promote vigorous hydrothermal activity above and to the sides of the magma chamber. These magmatic centres localize along the axes of mid-ocean ridges, the foci of magmatic activity of the ocean crust accounting for over 90% of igneous activity on the surface of the Earth. The permeability contrast between impermeable magma, fractured gabbro, and much more permeable dyke complex and pillow lava sets the conditions for hydrothermal exchange over a wide range of temperatures.

Profiles through the oceanic crust: the Samail ophiolite The Samail ophiolite was one of the first ophiolite complexes to have profiles of Sr, Nd and O isotopes (Fig. 1) measured through relatively complete sections of oceanic crust (Gregory & Taylor 1981; Lanphere et al 1981; McCulloch et

Fig. 1. Isotopic profiles through the Samail ophiolite modified after McCulloch et al. (1981) to include the data of Lanphere et al. (1981) from the Wadi Kadir traverse. Three different behaviours are illustrated: the ocean crust is a source and a sink for 18O, a sink for 87Sr and indifferent to 143Nd exchange. The symbols V, SD and Pg correspond to oceanic layer 2 consisting of pillow lavas, sheeted dyke complex and plagiogranite, respectively. Oceanic layer 3 rocks consist of high-level non-cumulus gabbro (HG), layered cumulate gabbro (G) and wehrlite-gabbro (WG) from the composite Ibra section of the Oman Mountains. For this composite section, D represents cumulus dunite at the base of the crust and H represents tectonite harzburgite beneath the palaeo-Moho. 618O = [(18O/16O)sampie/ (180/160)standard - 1] X 103 and eNd = [(143Nd/144Nd) sample/(143Nd/144Nd)standard - 1] X 104. The standards for O and Nd are Standard Mean Ocean Water and a standard chondritic earth reservoir (CHUR), respectively.

356

R. T. GREGORY

al. 1981). The shapes of the profiles illustrate behaviours distinct for each isotopic system that relates to the particular element's abundance in the crust and in seawater. These three elements illustrate three different behaviours with respect to sea-floor hydrothermal alteration.

Neodymium: trace element with a very short residence time Nd is an element with an extremely short residence time (300? years; e.g. Piepgras et al. 1979) in the ocean as a result of its insolubility in surface waters. It is present in such small concentrations because of the short residence time that the oceans are not well mixed with respect to Nd. The profile of Nd isotope ratios through the Samail ophiolite shows no effect of seawater interaction on the rocks (Fig. 1). The profile reflects the initial heterogeneity in Nd resulting from melting processes in the mantle. This is in contrast to Sr, for which the profile shows a gradient in isotopic ratios associated with sea-floor hydrothermal exchange.

Strontium isotopes: trace element with a million year residence time Sr is an element with a residence time of the order of a few million years so that the oceans are relatively well mixed with respect to Sr isotopes. Sr isotope ratios are enriched in the upper portions of the Samail oceanic crust, reflecting the exchange with seawater that had an isotopic ratio near 0.707 during the Cretaceous. Phases such as plagioclase in the layered gabbro have enriched Sr contents relative to pyroxene and olivine so that anorthositic layers deep in crustal layer 3 still record the mantle-derived signature of 0.70295. However, even at depths near the Moho, Srdepleted samples such as wehrlite layers exhibit a pronounced enrichment in 87Sr/86Sr ratios indicative of exchange with seawater Sr (Fig. 1). Seawater Sr is derived from two competing processes: input of radiogenic Sr from the weathering of Rb-enriched continental crustal rocks and input of mantle-derived Sr from Rb-depleted mantle. Because Cretaceous seawater had an isotopic ratio greater than that of the mantle, the altered oceanic crust is always enriched in 87Sr/ 86 Sr compared with its primary initial 87Sr/86Sr ratio. Exchange with mantle-derived Sr at midocean ridges drives the isotopic composition of Sr in seawater towards the mantle value and keeps the Sr isotopic composition of seawater from moving all the way to the average isotopic

composition of the river flux (c. 0.712-0.715; Goldstein & Jacobsen 1988; Palmer & Edmond 1989). Secular change curves for Sr isotopes show that the Phanerozoic Sr isotopic composition of seawater (0.706 < 87Sr/86Sr < 0.709) lies between the end-member compositions reflected by fluid in isotopic equilibrium with mantle Sr and fluids carrying Sr derived from continental weathering.

Oxygen isotopes: temperature-dependent fractionation and 100 Ma residence time The oxygen isotope profile is the most complex of all, reflecting exchange between a major element reservoir in seawater and the oceanic lithosphere (Fig. 1). Oxygen isotopes are light stable isotopes and in contrast to Sr isotopes exhibit a significant temperature-dependent fractionation between rock and water (>20%o, for plagioclase-water, O'Neil & Taylor 1967). For example, in the top parts of the section, low-temperature (e.g. 70 °C) albitization produces albite that is c. 20%o enriched relative to seawater. Andesine resulting from 400 °C exchange between dyke complex rocks and circulating fluid is only a few per mil enriched relative to the seawater-derived fluid. Anorthite surviving exchange in the deeper portions of ocean layer 3 will be depleted almost 3%o (using the exchange curve of O'Neil & Taylor 1967). As a result, the upper parts of the oceanic crust are enriched in 18O whereas the lower parts of the oceanic crust are generally depleted in 18O. There is a gradient in <518O values downward in the crust, decreasing from values >-|-10%o to values less than +4%o. After passing through a (518O minimum, still moving downward in the crust, the (518O curve then increases down-section back towards normal basaltic (318O values of c. +6%o. Three different elements with residence times spanning approximately six orders of magnitude produce very different exchange profiles. The Nd profile is virtually unchanged when compared with a profile for pristine basalt and gabbro. This is an element that would require more than 10000 masses of fluid to produce major isotopic change in the altered oceanic crust (McCulloch et al. 1981). Sr isotopes show an enrichment of the isotope ratio of the oceanic crust and a depletion of the oceans in 87Sr with respect to the input of Sr isotopes from continental weathering. This is a system that clearly cannot be buffered solely by exchange with the oceanic crust, i.e. the crust is always a sink for 87Sr. For oxygen isotopes, the amount of enrichment or depletion is a function of the average temperature of exchange, the permeability of the rocks, and the time-integrated fluid flux (water/rock

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES

357

ratio). With respect to seawater, the oceanic crust is both a major source and a sink for 18O. Because oceanic crustal production amounts to over 90% of the total crustal growth rate of the Earth, seawater must reach a steady-state value with respect to the mantle of the Earth.

Special significance of oxygen isotope profile for seawater When the profiles of oxygen isotopes are averaged, there are roughly complementary masses of enriched and depleted rocks that appear to balance each other so that the mean value of the altered oceanic crust is close to its primary 18O/16O ratio (Muehlenbachs & Clayton 1976; Gregory & Taylor 1981). In other words, the average bulk fractionation between seawater with a (518O value of c. 0 and mantle-derived rocks with (518O of c. 6%o is close to the difference between the two reservoirs, i.e. +6%o. If the balance is not exact, then some other reservoir is important in controlling the oxygen isotopic composition of the oceans. The observation that oceans near (518O c. 0 provide altered rocks with 18O enrichments and depletions requires that the $18O value of the ocean achieve some steady state with respect to the average (518O value of the mantle. Given the time constants for exchange (100 Ma time scale), the balance need not be exact in every section that only takes a fraction of a perturbation half-life to produce. As long as the actual measured bulk A value ((518O of the average crust minus the $18O value of seawater) between the altered crust and seawater is near +6%o for a long-term average, the oceans will migrate towards a (518O value of zero (Gregory 1991; Muehlenbachs 1998).

Comparison with oceanic layer 3 results Figure 2 compares the oxygen isotope results of coexisting mineral pairs from oceanic and Samail ophiolitic realms, and shows a great similarity in the style and spread along the non-equilibrium exchange arrays (e.g. Gregory et al 1989). These non-equilibrium arrays result from differing rates of isotopic exchange between seawater and plagioclase (easily exchanged) and pyroxene (more resistant to exchange). Non-equilibrium arrays are preserved only if the duration of the exchange event is short and if the exchange event is at a high enough temperature to allow pyroxene to remain stable in the presence of seawater-derived fluid. Both conditions are satisfied because the time scale for the cooling gabbroic crust is short enough, <106 years (Gregory et al. 1989), and

Fig. 2. Comparison between the oceanic layer 3 gabbro results (Stakes 1991; Lecuyer & Reynard 1996) and those from the Samail ophiolite complex (Gregory et al. 1989; Stakes & Taylor 1992; new data from the Dasir section of the ophiolite) for coexisting pyroxene and plagioclase. Both fast-spreading (Hess Deep) and slowspreading ridges (Mid-Atlantic Ridge, MAR; Indian Ocean dredge samples, Indian Ocean Hole 73 5B) are represented from the modern ocean basins. Irrespective of tectonic setting, the pyroxene-plagioclase pairs display non-equilibrium fractionations (differences in the c518O values between plagioclase and coexisting pyroxene) consistent with subsolidus exchange with seawater-derived hydrothermal fluid. Under equilibrium conditions, coexisting mineral pairs are forbidden to plot below the zero fractionation (A = 0). Similarly, more rapid exchange of plagioclase with strongly 18O-shifted hydrothermal fluids results in anomalously large differences between plagioclase and pyroxene (518O values. It should be noted that the Oman traverses from Wadis Rajmi and Hilti are c. 50 km apart in the northern Oman Mountains. The new Dasir data are from a section that lies c. 100 km SE of Wadi Hilti near the Samail Gap, which separates the Jabal Akhdar dome from the Saih Hatat dome. The Ibra data are from the southeastern Oman Mountains (45 km SE of Dasir) and represent results from several profiles within a 20 km block perpendicular to the strike of the sheeted dykes.

pyroxenes in the highest grade rocks are either petrographically unaltered or exhibit incipient alteration to brown amphibole. The results for several sections of the Samail ophiolite separated by few hundred kilometres normal (and several hundred parallel) to the spreading direction and oceanic crust are similar.

358

R. T. GREGORY

Whereas the original tectonic setting for the Oman ophiolite is controversial (e.g. Moores et al. 2000), both slow-spreading (Mid-Atlantic Ridge and Indian Ocean) and fast-spreading regimes (Hess Deep) are represented in the plagioclase pyroxene pairs from modern ocean basins. Ophiolite pairs and oceanic pairs exhibit similar nonequilibrium relationships in 6-6 plots of the <518O value of pyroxene plotted against that for plagioclase. The similarity in alteration styles also extends to the fine-scale development of millimetre-scale vein systems in gabbroic rocks of layer 3. Manning et al. (2000) showed that high-temperature veins in the Samail ophiolite were similar to those found in the Hess Deep, indicating that permeability development in layer 3 rocks was similar in ophiolite and sea floor. Manning et al. (2000) compared the Samail ophiolite to a fast-spreading mid-ocean ridge system. The data for layer 3 gabbros complement results summarized by Alt & Teagle (2000), who compared the profiles of the upper-crustal sections of ophiolites, specifically Troodos and Oman, with profiles derived from sea-floor studies. Similarly, Banerjee & Gillis (2001) have compared suprasubduction zone rocks from the Tonga Arc with ophiolites and mid-ocean ridge rocks and found similar alteration patterns, with the exception of more epidosite and trondjhemite in the forearc spreading environment. All of the studies show that hydrothermal alteration, the effects of hot rock interacting with hot fluids, is not a very good discriminator of tectonic setting. The question is whether the distribution of 6ISO values observed in altered oceanic crust, and more particularly ophiolites, is sufficiently sensitive to detect isotopic variability in the global ocean over geological time (e.g. Veizer et al. 1999; Wallmann 2001).

Comparison with continental layered gabbro complexes The flood basalt provinces, coastal dyke swarms and layered intrusions found in ancient rift zones share many similarities with ophiolites in terms of boundary and initial conditions for fluid-rock interaction. Continental layered gabbro complexes thus provide an answer to the question of whether ophiolites are sensitive to changes in seawater d18O values (Veizer et al. 1999; Wallmann 2001). The major difference in the settings is that meteoritic water is often the starting fluid instead of seawater. Over most of the Earth, meteoric fluids are on average about 4%o depleted with respect to modern seawater. However, in some continental regions, meteoric waters (particularly

at high latitudes and at high altitudes) exhibit depletions greater than 10%o with respect to modern seawater (e.g. Rozanski et al. 1993). The studies of Taylor and coworkers have defined the isotopic exchange characteristics of meteoric hydrothermal exchange in continental layered gabbro complexes (e.g. Criss & Taylor 1986; Criss 1999). The most complete of these studies is the work on the Skaergaard layered intrusion (Taylor & Forester 1979). Figure 3 shows a comparison of the pyroxene-plagioclase pairs from the Skaergaard intrusion with those from oceanic crust and ophiolite complexes. Mineral 6-6 diagrams can also display wholerock data by plotting their (518O values on the zero fractionation line. This facilitates comparison of mineral pairs with whole rocks from the entire crustal section. In this case, the whole-rock values are for hydrothermally altered basalts and diabase; the results are strikingly different for continental and oceanic environments. Through the examination of shifts in the spread in (518O values of isotopically exchanged rocks, it is possible to detect differences in the isotopic composition of the starting material at the few per mil level. This clearly shows that large (5-10%o) changes in the (318O value of seawater would be easily detectable in an ophiolite section.

Whole-rock 518O values of pillow lavas through time Figure 4 shows the whole-rock (518O values of pillow lavas through time. Pillow lavas typically exhibit zeolite- to greenschist-facies metamorphic assemblages. Bowers & Taylor (1985) were able to calculate the bulk fractionation factor for pillow basalt altered in seawater. This bulk-rock-seawater fractionation factor clusters near the value of +6%o; yet most greenstones exhibit (518O values that cluster around +9%o. For greenstones to be enriched in 18O relative to seawater, there must be an initial, low-temperature partial exchange followed by subsequent growth of greenschist mineral assemblages. From a petrological viewpoint, this is most probably accomplished by an initial hydration of the groundmass glass followed by conversion to sheet silicates such as chlorite accompanied by albitization of calcic plagioclase. A simple single-stage, hydrothermal exchange with seawater at greenschist facies should produce altered basalt or diabase with (318O values less than 6%o. Black smoker vent fluids typically exhibit d 18 O >0 (Shanks 2001) so that there must be a complementary 18O-depleted reservoir in the recharge volume for the black smoker vent fluid.

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES

359

Fig. 3. Comparison of Skaergaard intrusion mineral pair data with those from oceanic layer 3 rocks along with <518O values of altered basaltic whole rocks plotted along the zero fractionation line (•, East Greenland Tertiary lava; V, Oman pillow lavas). The Skaergaard intrusion was altered by fluids with an initial (518O value > — 11 using dD from altered rocks (Taylor & Forester 1979) that give calculated water dD c. -100, and assuming a meteoric water line with a D excess of +10%o. The fields for Oman pillow lava and East Tertiary lavas are strikingly different, as would be expected for pillow lava sections altered by —5 to — 10%o seawater. These results are inconsistent with the forward modelling shown by Wallmann (2001), where somehow the isotopic profile of the oceanic crust is rendered insensitive to changes in starting fluid isotopic composition by parameters in the model. Data from the Skaergaard intrusion are from Taylor & Forester (1979), and pillow lava data are from Gregory & Taylor (1981) and Stakes & Taylor (1992).

A plausible explanation for the 18O-enriched character of most greenstones is that there is a superposition of alteration regimes as the new crust spreads away from the ridge axis. As the crust spreads, the rocks encounter 18O-shifted fluids exiting from the newly solidified sides of the magma chamber where layer 3 gabbro has become depleted in 18O (e.g. Gregory & Taylor 1981). Two mechanisms are capable of producing greenschist-facies meta-basalt and diabase with (518O values greater than +6%o that involve superposition of exchange regimes in an evolving crust. Whatever the mechanism of the increase in (518O values of altered pillow lavas and diabase dykes, it is clear that if the (518O value of the starting seawater fluid was shifted downward by 5-6%o, the measured values in Figure 4 would be very different. The mean greenstone (518O value of c. +8-9%o would be shifted downward to values less than 6%o. The data from the Skaergaard instrusion and other continental volcanic provinces clearly show depleted volcanic rocks when fluids are 18Odepleted relative to seawater. Even though it is not possible to demonstrate complementary depleted and enriched 18O reservoirs in the ancient past for entire profiles through

the oceanic crust, the most common submarine rock type in the ancient rock record, greenstones (altered pillow lavas), exhibit (518O values consistent with seawater (318O values near zero over Earth history. The result is inconsistent with inferences from the carbonate record (e.g. Wallmann 2001) that require that the (518O value of the global ocean changed rapidly by as much as 6%o over certain intervals of Phanerozoic history.

The rate constant problem Attempts to explain the discrepancy between the greenstone-ophiolite isotopic record and the carbonate record group into three types of interpretations. The first is that the (518O value of seawater changes by as much as 5-10%o over geological time. This is a larger increase for a major constituent of water and rock than that observed for the trace element Sr. This interpretation is in clear conflict with the ophiolite record. Second, the surface temperature of the Earth as recorded by carbonates and cherts varies by several tens of degrees and, in general, surface temperatures were tens of degrees higher than today's if seawater maintains its $18O value near zero. This interpretation seems to run counter to global temperature

360

R. T. GREGORY

Fig. 4. c518O values of greenstones through geological time. Greenstones with (518O values >+6%o are common throughout Earth history and are inconsistent with exchange and alteration by fluids strongly depleted in 18O. T, Canyon Mountain (Permian) and Bay of Islands (Ordovician) ophiolite; A, Darb Zubaydah Ophiolite, Saudi Arabia, an accreted arc terrane; •, Purtuniq ophiolite; O, Fortescue pillow lavas basement for the Hamersley Iron Formation; •, Barberton Mountain Belt; figure modified from Gregory (1991; references therein). The Purtuniq ophiolite data are from Holmden & Muehlenbachs (1993). The frequency distribution in the inset shows the results for over 100 greenstone silicate analyses from the Pilbara block, mainly from the North Pole Dome, Western Australia. DSDP, Deep Sea Drilling Project.

change estimates from sedimentology and climate modelling, and from astrophysical inferences that suggest a faint young Sun (e.g. Sagen & Mullen 1972; Hoffman et al 1998). If anything, inferences about the snowball Earth suggest that mean global temperatures may have been lower than isotopic estimates, which have them double or triple back into the past (e.g. Knauth & Epstein 1976). Third, one of the records, either the ophiolite or the carbonate record, is irrelevant to the isotopic composition of the oceans. The major contributions of ophiolite studies to the seawater oxygen isotope controversy are twofold. First, ophiolite studies confirmed inferences from dredge sample work summarized by Muehlenbachs & Clayton (1976) that hypothesized the existence of complementary reservoirs of 18Odepleted and 18O-enriched rocks in the altered oceanic crust. Second, the demonstration of the depth of penetration of seawater in a more complete geological context allows the estimation of the rate constants of exchange. Table 1 shows

rate constant calculations for Sr and O isotopes using the same assumptions of oceanic crustal production rates and continental weathering rates translated into numbers that make sense technically following the analysis of Gregory (1991; references therein) on the profiles shown in Figure 1. A global spreading rate of 3 km2 a"1 yields a characteristic age of oceanic crust of about 100 Ma assuming an oceanic crustal area of 3 X 108 km2. A chemical weathering rate of 3 km3 a"1 yields a mean age of continental crust of about 2500 Ma for a constant crustal volume of 7.5 X 109 km3 (e.g. Armstrong 1991). Tectonic rates significantly different from these rates are not geologically sustainable, otherwise the rock record would change dramatically. The age distributions on the continents and in the ocean basins would be substantially different from the current distribution with young ocean basins (<150 Ma) and middle-aged continents (mean age of 1.8-2.6 Ga). Interpretations of measured iso-

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES topic ratios on carbonate rocks that require huge changes in the tectonic rates for the Earth suggest that there must be another explanation for these carbonate data sets. Table 1 shows that the same spreading rates and chemical weathering rates applied to the Sr and O isotopic systems yield dramatically different results consistent with observations on the isotopic composition of the oceans. The rates take into account the depth of penetration of seawater into the crust for each isotopic system, the integrated fluid and the elemental concentrations in fluid and rocks. The target steady-state values are calculated from reasonable choices for the isotopic composition of the reservoirs for the continental and oceanic crust (e.g. Goldstein & Jacobsen 1988; Palmer & Edmond 1989). For Sr, this is straightforward because there are minimal temperature effects. Weathering is more dominant (63:37) in controlling the Sr isotopic composition of seawater, which can change rapidly on geological time scales (<5 Ma). For oxygen isotopes, the target value for seawater is chosen by assuming an average value of rocks (<5i) dissolved minus a bulk-rock-fluid fractionation factor (Aiw) of +20 for continental weathering. This target value reflects all of the geological time that has passed so that the rock cycle has produced sediments with a memory of their previous interactions with the oceans, i.e. the bulk <518O value is no longer dominated by a normal igneous rock signature ((518O < +8%o; Taylor 1968). A frequency distribution diagram of sandstone whole-rock (518O values illustrates the shift upwards in isotopic composition from first cycle sediments such as the Archaean Lalla Rookh

361

Sandstone to the Phanerozoic Lachlan and Ouachita fold belt sandstones (Fig. 5). The net effect is that the target for the (518O value of the ocean driven by the weathering flux changes from a minimum value of -14 to — 4%o. The target value for oxygen isotopes for the ridge crest interaction is based upon the difference in the <518O values between the ocean and the mantle reservoir that partially melts to produce the oceanic crust. The ridge crest interaction is more dominant for O isotopes; the roles are reversed when compared with Sr isotopes and the time constant for the simple cycle is c. 100 Ma instead of c. 2 Ma for Sr isotopes. Other workers have concentrated on estimating the relative amounts of high- and low-temperature alteration and 18O exchange that will depend on the relative portions of oceanic layer 2 and 3 rocks (e.g. Holland 1984; Muehlenbachs 1986; Wallmann 2001). In analysis of Table 1, this variability is buried in the value of A between the crust and seawater. The fluctuations in the value A can be gauged by examining the variability in thickness of the oceanic layers deduced from seismic refraction surveys (e.g. Shor & Raitt 1969). These surveys sample much more of the oceanic crust than has been possible by drilling and dredging. The possible perturbations in the value of A induced by changes in the relative proportions of high- and low-temperature exchange must scale like the variability in the thickness of oceanic layers 2 and 3. Considering the mean age of the oceanic crust and the symmetry of spreading processes, these perturbations to the value of A are likely to be short compared with the half-life of a perturbation to the global ocean (518O value.

Table 1. Tectonic rates transformed into cycle rates, characteristic times, weighting factors and steady-state target isotopic ratios (after Gregory 1991) Rate (Ma^1)*

Time (Ma)t

/

Target§

3 km2 a"1 3 km2 a-1

0.008 0.160

125 6.3

0.73 0.37

0 0.703

3 km3 a"1 3 km3 a-1

0.003 0.278

350 3.6

0.27 0.63

-4 0.712

0.011 0.438

91 2.3

Tectonic rate Ridge crest Oxygen Strontium Weathering Oxygen Strontium Total Oxygen Strontium

-1 0.709

*The rate constant is the coefficient (k) for the damping factor (e kt ) for transient perturbations. tCharacteristic time is \lk. The process with the shortest time constant usually dominates. % §

fi = *//E*i.

Near steady state, the target (518O value for seawater after its exchange with the ith reservoir is (6[ — AI W ). Near steady state, the seawater (518O = 2/j((5j ~ A,- w ), where /,- = ki/'Zkj. Because of the rock cycle, particularly for the continents, the (3; can vary faster than the system can respond.

362

R. T. GREGORY

Fig. 5. A comparison between the (318O values of sandstones from an Archaean sandstone succession (new measurements from the Lalla Rookh Sandstone, Pilbara block, Western Australia) and Palaeozoic fold belts (the Lachlan fold belt, Australia, Gray et al. 1991; and the Ouachita Mountains, USA, Richards et al 2002). As a result of the rock cycle and the recycling of crustal quartz enriched in 18O owing to low-temperature interactions with surface fluids, the d 18 O values of Phanerozoic sandstones are higher than those from first cycle sands sourced from Archaean quartz-bearing igneous rocks. The effect of all of the recycling of crustal materials is to shift the target value of seawater resulting from continental weathering upward from <— 10%o to values closer to —4%o.

Figure 6 shows calculations for oxygen isotopes using a fixed 3 km3 a"1 chemical weathering rate (e.g. 26 tons km~2 a"1 actual rock dissolution, Berner & Berner 1996) with variable spreading rates from 1 to 5 km2 a"1. The area filled by the pattern represents the best estimate based upon spreading rates for the last 150 Ma. The inset shows the half-life of a perturbation for different choices of tectonic weathering rates and spreading rates. For nearly all combinations of spreading and weathering rates that produce the dichotomy in the ages between ocean basins and continents, the ridge interaction should dominate the oxygen isotopic composition of the oceans.

Global weathering rates Large changes in weathering rates (60-150 km3 a"1) are required to shift the <318O value of seawater by the magnitude (the absolute change and the time scale for the change) required by interpretations of the carbonate record (e.g. Walker & Lohmann 1989; Gregory 1991; Wallmann 2001). Chemical denudation rates measured from

today's rivers (see the review by Berner & Berner 1996), when extrapolated into global weathering rates, range from c. 50 km3 a"1 for dissolution of evaporites to 11 km3 a"1 for carbonate rocks and down to about 2-3 km3 a"1 for silicate rocks. Clearly, there are not enough evaporite rocks in the crust of the Earth to sustain such high dissolution rates. The mean value of the dissolved load is closer to 5 km3 a"1, which then is corrected down to about 3 km3 a"1 (26 tonskm"2 a"1, Berner & Berner 1996). These rates scaled to the volume of continental crust yield crustal residence times that range from 150 to 3750 Ma. Hercod et al. (1998) estimated a chemical denudation rate of 4 km3 a"1 (residence time 1.9 Ga) in a carbonate rock dominated watershed where serial measurements were made over a period of 1 year to take into account seasonal changes in solute export. This estimate is a factor of two lower than the average carbonate rate reported above. Clearly, geological rates that yield residence times significantly different from the mean age of the continental crust are not sustainable. Recycling of rocks already formed in the presence of seawater with a (518O value close to zero has little effect on the oxygen isotopic composition of the oceans, so that it is the processing of mantlederived silicate that has the biggest impact on the <518O value of seawater. The rates inferred for silicate rocks are in general less than 3 km3 a"1; this is particularly so if the continental crust has grown slightly with time (e.g. Albarede 1989). In Figure 6, the quasi-steady-state <518O value of seawater is less than the present-day value of the ocean because of the presence of ice sheets today that result in a c. l%o enrichment in the d 18 O value of seawater. During our global icehouse conditions (Fischer 1982) significant volumes of 18O-depleted ice are stored in the Antarctic and Greenland ice sheets (e.g. Gregory & Taylor 1981). Because most of the pre-Cretaceous carbonate 18O measurements are on continental shelf rocks, it is instructive to look at the water balance for the Earth.

Short-term water cycle effects on the 618O of seawater Waxing and waning of ice sheets produces an oxygen isotope ice volume effect that shows up in ice cores (Petit et al. 1999) and deep-sea cores (Emiliani 1978). Because the processes involving ice sheet advance or retreat involve time constants less than 105 years and are reversible, the net effect of the glacial advance and retreat is to add 'noise' to the long-term signal (Gregory & Taylor

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES

363

Fig. 6. (518O evolution of seawater for plausible choices of spreading rates and chemical weathering rates, assuming a starting ocean outgassed from the mantle during early Earth history. It should be noted that the +8%o starting composition could just as easily been — 8%o and the shape of the curve would be similar to the curve shown by Wallmann (2001) for the change in seawater (518O during the Phanerozoic. This curve and the curve of Wallmann (2001) are smoother than the actual curves inferred from carbonate data. The global cycle for oxygen isotopes in seawater cannot accommodate rapid changes (several per mil per 10 Ma) through fluid-rock interaction that occurs at any plausible geological rate (Walker & Lohmann 1989; Gregory 1991). The question is: why would the isotopic composition of the global ocean ever have moved to — 8%o in the first place? The system is damped for any reasonable combination of spreading and weathering rates (inset) so that the global ocean obtains $18O values near 0%o early in Earth history, probably by the beginning of the Archaean. The 18O-depleted rocks that must have resulted from high-temperature exchange in the crust are very difficult to find in the rock record. In particular, the lack of 18O-depleted greenstones and gabbros is troubling for global seawater changes inferred solely from carbonate, phosphate and chert (e.g. Muehlenbachs 1998). The inset shows perturbation half-life v. weathering and spreading rates cast as tectonic rates amenable to dimensional analysis. Any plausible combination of weathering and spreading rates must be consistent with the dichotomy of young ocean basins and middle-aged continental crust relative to the age of the Earth. Permissible combinations of tectonic rates drive the <518O value of seawater towards its current value.

1981). Long-term storage of 18O-depleted ice will shift the <518O value of the ocean towards more positive values. The long-term global cycle will work to reverse the increase in (518O, provided the ice caps can persist for significant portions of perturbation half lives, about the time scale of an era like the Cenozoic. Table 2 illustrates a mass balance calculation for the hydrosphere. Any perturbation to the ocean by partitioning water between the ocean and other reservoirs is simply its mole fraction times its <518O value. Ice volume changes have clearly imparted per mil level changes in the oxygen isotopic composition of the oceans; if only because the (318O value of ice is strongly 18O depleted with respect to seawater. The meteoric

water cycle provides a mechanism to produce masses of 18O-depleted water.

Epicontinental seaways In the geological past, particularly during times of inferred global greenhouse conditions, sea level was higher than it is today. Epicontinental seaways are much more extensive during times of more rapid sea-floor spreading. Land-locked epicontinental seas or very large shallow seaways could conceivably have significantly lower d 18 O values than the global ocean. The last row in Table 2 shows the impact of a large epicontinental sea (the surface area of all continents, 100m deep, that has a (518O value of —5%o close to the average of

364

R. T. GREGORY

Table 2. Water balance Reservoir Oceans Ice caps and glaciers Ground water Meteoric waters Total Maximum epicontinental sea*

Volume*

%

(5180

1400 43.4 15.3 0.2 1459 30

96 2.97 1.0 0.01 100 2.05

0 c. -35 -4 -4 -1.1 -5

X«180 0 -1.04 -0.04 -0.00 -1.08 -0.1

* Volumes from Berner & Berner (1996). tpor this purpose, a maximum epicontinental sea is a hypothetical 100m deep sea covering 3 X 108 km2; this is larger than any conceivable seaway. It should be noted that —0.1 is the change in seawater if the reservoir was instantaneously mixed back into seawater. X is the volume %, ~ the mole fraction of the reservoir.

modern meteoric water) on the (518O value of seawater. Such a hypothetical seaway would affect the <518O value of the global ocean by enriching it by only +0.1 %o. The results of these mass balance calculations indicate that it is possible that ophiolites and greenstones could be recording seawater isotopic signatures much different from those recorded by carbonates from epicontinental shelf regions from supercontinental masses. This suggests that there may be a wealth of palaeoenvironmental information present in the isotope and trace element signatures of pristine carbonate rocks that needs to be tested against plate reconstructions and global circulation models for the atmosphere and oceans. The carbonate record is at least a four-dimensional record that includes two spatial coordinates as well as (518O and time. Two documented examples exist for interior seaways in the Cretaceous, when most workers would agree that the global oceans were not less than — l%o. The first is the Late Cretaceous interior seaway of North America, where portions of the water mass may have been as low as -6%o (e.g. Tourtelot & Rye 1969; Wright 1987). The second example is the Late Aptian Eromanga Basin (Fig. 7), where DeLurio & Frakes (1999) found glendonite pseudomorphs after ikaite. Ikaite is a hydrated carbonate that only forms in water near freezing. Using the transformation of ikaite to glendonite (calcite), the measured oxygen isotopic composition glendonite shifts the oxygen isotope composition of the seaway downward to c. —3%o. As a result, calculated belemnite palaeotemperatures shifted downward by 10°C to become more consistent with the presence of dropstones in the Bulldog Shale and terrestrial palaeotemperatures near freezing ( Ferguson et al. 1999) from southeastern Australia determined from fluvial sediments with cryoturbation structures.

Fig. 7. A palaeoreconstruction of the Eromanga epicontinental sea during the Aptian-Albian for Australia showing the palaeolatitude and the position of the glendonite locality ( DeLurio & Frakes 1999) of the Australian interior epicontinental seaway. Terrestrial outcrops are represented by the shaded areas and the localities where low-<518O meteoric waters are inferred from analyses of syndepositional calcite concretions in fluvial sediments from the Otways and Strzlecki Ranges, southeastern Australia (two dots; Ferguson et al. 1999). Waters calculated using the transition temperature from ikaite to calcite yield seaway water compositions of c. -3%o, which put belemnite palaeotemperatures into better agreement with the presence of ice-rafted dropstones and cryoturbation structures in southeastern Australia.

Significance of ophiolite studies Ophiolites by their very nature, as pieces of ancient oceanic lithosphere that have escaped recycling that occurs at subduction zones, may

OPHIOLITES AND GLOBAL GEOCHEMICAL CYCLES form in anomalous tectonic settings when compared with the normal ridge systems that extend for over 40 000 km over the surface of the Earth. For obduction to occur, ophiolite complexes must form in regions where plate interactions must be changing on a regional scale (100-1000 km). However, the physical and chemical processes that occur when hot rock interacts with seawater (10 km scale) in a zone of magmatic intrusion and extension are similar whether they occur at midocean ridges, back-arc basins or in suprasubduction zone spreading centres. Alteration patterns seen in ophiolite complexes are similar to those found in dredge samples and drill cores from the sea floor. Geometric constraints from ophiolites have shaped interpretations of the structure of the sea floor. Importantly, ophiolite studies provide the three-dimensional information on the deeper levels of the oceanic crust and upper mantle difficult to obtain from sea-floor studies. Ophiolites with sheeted dyke complexes formed in ocean basins at water depths where seawater is likely to be the agent for fluid-rock interaction. All ophiolite and greenstone studies show similar patterns of fluid-rock interaction in terms of oxygen isotopes. If the global ocean ever was as depleted as interpretations of the carbonate record require (e.g. Veizer et al. 1999), there would be more 18O-depleted greenstones observed in the rock record. Figure 8 shows Bay of Islands ophiolite whole-rock data as a function of depth. The Cambro-Ordovician Bay of Islands ophiolite was hydrothermally altered during a time when the global oceans were presumably very strongly depleted in 18O (e.g. Veizer et al. 1999). The ophiolite simply does not confirm the existence of oceans with low (318O values, i.e. the rocks are not depleted as they should be (dashed curve on Fig. 8) if the oceans had the (518O value inferred by Veizer et al. (1999). There should be abundant oceanic layer 3 rocks from ophiolites that have 6nO values <0 (Fig. 8). So far, these rocks have not been found; indicating that there must be another explanation for the carbonate oxygen isotope record.

Conclusions Ophiolites provide a critical link between plate tectonics and global geochemical cycles because these rocks are important for understanding the most dominant crust-forming process for the Earth and for understanding its transformation from pristine mantle-derived rocks to the rocks that eventually are subducted at convergent margins. These processes enable the crust and hydrosphere to exchange elements with the mantle of the

365

Fig. 8. Schematic profile showing a d 18 O profile typical for near steady-state conditions, i.e. the crust is both depleted and enriched in 18O. The steady-state condition is typical of the Earth when the (518O value of seawater is close to zero. This is compared with a hypothetical profile (bold dashed line) calculated using mass balance and temperature constraints from the Samail ophiolite with the seawater fluid shifted downward to — 8%o. In this scenario, the crust becomes a major source for 18O so that the ocean would be in a transient state with a perturbation time scale of c. 100 Ma. Shown for comparison are Bay of Islands ophiolite data (dots with tie lines) cited by Gregory & Taylor (1981). This ophiolite is late Cambrian and dates from a time when the global oceans according to the carbonate record, if interpreted literally, were sufficiently depleted in 18O to produce the calculated dashed line profile, k, mainly from the North Pole Dome, Western Australia. DSDP, Deep Sea Drilling Project.

Earth. The behaviour of every element must be examined separately. Profiles through the Samail ophiolite show that different elements behave very differently during hydrothermal exchange at the spreading centre. Yet, through measurement and analysis, a common set of tectonic rates (rates of material exchange) can account for the apparently different controls on the isotopic abundances. An element such as Sr can exhibit concentrations indicative of river fluxes and continental weathering inputs as the dominant control on the Sr isotopic composition of the seawater. The same global material exchange rates explain the buffering of the (518O value of the ocean by ridge crest processes. Material balance calculations suggest that there is no reason to assume that epicontinental seaways faithfully record the isotopic composition, at least for oxygen isotopes, of the global ocean. Large changes in the isotopic composition of the interior seaways can be explained by processes related to cycling and storage of meteoric waters without affecting the oxygen isotopic composition of the global ocean. Pristine oxygen isotopic data from epicontinental seaways may record a wealth of palaeoenvironmental information.

366

R. T. GREGORY

S. Jacobsen provided the samples of the Bay of Islands ophiolite; these were analysed at Caltech along with those for M. McCulloch. Special thanks are due to H. P. Taylor, Jr, R. G. Coleman and C. A. Hopson for making the original work in Oman possible. D. S. Stakes and T. S. Bowers certainly extended it. The Australian Research Council and the SMU Stable Isotope Laboratory funded the work on the Pilbara craton. H. Al Azri, Ministry of Petroleum and Minerals and now the Ministry of Commerce, sponsors the work in the Sultanate of Oman. Comments by N. Banerjee, A. Basu, Y. Dilek, G. FriihGreen and K. Ferguson improved an earlier version of the manuscript.

References ALBAREDE, F. 1989. Sm/Nd constraints on the growth rate of continental crust. Tectonophysics, 161, 299-305. ALT, J.C. & TEAGLE, D.A.H. 2000. Hydrothermal alteration and fluid fluxes in ophiolites and oceanic crust. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies. Geological Society of America, Special Papers, 349, 273-282. ARMSTRONG, R.L. 1991. The persistent myth of crustal growth. Australian Journal of Earth Sciences, 38, 613-630. BANERJEE, N.R. & GILLIS, K.M. 2001. Hydrothermal alteration in a modern suprasubduction zone: the Tonga forearc crust. Journal of Geophysical Research, 106, 21737-21750. BARTH, T.F.W. 1962. Theoretical Petrology, 2nd. Wiley, New York. BERNER, K. & BERNER, R.A. 1996. Global Environment: Water, Air, and Geochemical Cycles. Prentice Hall, New Jersey, USA. BOWERS, T.S. & TAYLOR, H.P. JR 1985. An integrated chemical and stable-isotope model of the origin of mid-ocean ridge hot-spring systems. Journal of Geophysical Research, 90, 12583-12606. CRISS, R.E. 1999. Principles of Stable Isotope Distribution. Oxford University Press, New York. CRISS, R.E. & TAYLOR, H.P., JR. 1986. Meteoric-hydrothermal systems. In: VALLEY, J.W., TAYLOR, H.P., JR. & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geological Processes. Reviews in Mineralogy, 16, 373^24. DELURIO, J.L. & FRAKES, L.A. 1999. Glendonites as a paleoenvironmental tool: implications for early Cretaceous high latitude climates in Australia. Geochimica et Cosmochimica Acta, 63, 1039-1048. EMILIANI, C. 1978. The cause of the ice ages. Earth and Planetary Science Letters, 37, 349-352. FERGUSON, K.M., GREGORY, R.T. & CONSTANTINE, A. 1999. Lower Cretaceous (Aptian-Albian) secular changes in the oxygen and carbon isotope record from high paleolatitude, fluvial sediments, southeast Australia: comparisons to the marine record. In: BARRERA, E. & JOHNSON, C.C. (eds) Evolution of Cretaceous Ocean-Climate System. Geological Society of America, Special Papers, 332, 59-72.

FISCHER, A.G. 1982. Long term climatic oscillations recorded in stratigraphy. In: BERGER, W.H. & CROWELL, J.C. (eds) Climate and Earth History. National Academy Press, Washington, DC, 97-104. GOLDSTEIN, S.J. & JACOBSEN, S.B. 1988. Nd and Sr isotopic systematics of river water suspended material: implications for crustal evolution. Earth and Planetary Science Letters, 87, 249-265. GRAY, D.R., GREGORY, R.T. & DURNEY, D.W. 1991. Rock-buffered fluid-rock interactions in deformed quartz rich turbidite sequences, eastern Australia. Journal of Geophysical Research, 96, 19681 -19704. GREGOR, C.B., GARRELS, R.M., MACKENZIE, F.T. & MAYNARD, J.B. 1988. Chemical Cycles in the Evolution of the Earth. Wiley Interscience, New York. GREGORY, R.T. 1991. Oxygen isotope history of seawater revisited: 3.5 billion years of the greenstone record and its implication for the stability of seawater 18O. In: TAYLOR, H.P. JR, O'NEIL, J.R. & KAPLAN, I.R. (eds) Stable Isotope Geochemistry: a Tribute to Samuel Epstein. Geochemical Society, Special Publication, 3, 65-76. GREGORY, R.T. & TAYLOR, H.P. JR 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: evidence for 618Obuffering of the oceans by deep (>5 km) seawaterhydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research, 86, 2737-2755. GREGORY, R.T., CRISS, R.E. & TAYLOR, H.P. JR 1989. Oxygen isotope kinetics of mineral pairs in closed and open systems: applications to problems of hydrothermal alteration of igneous rocks and Precambrian iron formations. Chemical Geology, 75,1-42. HARDIE, L.H. 1996. Secular variation in seawater chemistry: an explanation for the coupled secular variation in the mineralogies of marine limestones and potash evaporites over the past 600 m.yO Geology, 24, 279-283. HARDIE, L.H. 1998. On the secular variations in the composition of Phanerozoic marine potash evaporites, Comment and Reply by H. D. Holland and H. Zimmerman. Geology, 26, 91-92. HERCOD, D.J., BRADY, P.V. & GREGORY, R.T. 1998. Catchment-scale coupling between pyrite oxidation and calcite weathering. Chemical Geology, 151, 259-276. HERZBERG, C. & O'HARA, MJ. 1998. Phase equilibrium constraints on the origin of basalts, picrites, and komatiites. Earth-Science Reviews, 44, 39-79. HOFFMAN, P.P., KAUFMAN, A.J., HALVERSON, G.P. & SCHRAG, D. 1998. A Neoproterozoic snowball Earth. Science, 281, 1342-1346. HOLLAND, H. 1984. The Chemical Evolution of the Atmosphere and Oceans. Princeton University Press, Princeton, NJ. HOLLAND, H., HORITA, J. & SEYFRIED, W.E. JR 1996. On the secular variations in the composition of Phanerozoic marine potash evaporites. Geology, 24, 993-996. HOLMDEN, C.F. & MUEHLENBACHS, K. 1993. The 18O/ 16 O ratio of 2-billion-year-old seawater inferred from ancient oceanic crust. Science, 259, 17331736.

OPHIOLITES AND GLOBAL G E O C H E M I C A L CYCLES KNAUTH, L.P. & EPSTEIN, S. 1976. Hydrogen and oxygen isotopic ratios in nodular and bedded cherts. Geochimica et Cosmochimica Acta, 40, 1095-1108. KUSHIRO, I. 2001. Partial melting experiments on peridotite and the origin of mid-ocean ridge basalt. Annual Review of Earth and Planetary Sciences, 29,71-107. LANPHERE, M.A., COLEMAN, R.G. & HOPSON, C.A. 1981. Sr isotopic tracer study of the Samail ophiolite, Oman. Journal of Geophysical Research, 86, 2709-2720. LECUYER, C. & REYNARD, B. 1996. High-temperature alteration of oceanic gabbros by seawater (Hess Deep, Ocean Drilling Program Leg 147): evidence from oxygen isotopes and elemental fluxes. Journal of Geophysical Research, 101, 15883-15897. MANNING, C.E., MACLEOD, C.J. & WESTON, P.E. 2000. Lower-crustal cracking front at fast spreading ridges: evidence from the East Pacific Rise and the Oman ophiolite. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies. Geological Society of America, Special Papers, 349,261-272. McCuLLOCH, M.T., GREGORY, R.T., WASSERBURG, GJ. & TAYLOR, H.P. JR 1981. Sm-Nd, Rb-Sr and 18O/ 16 O isotopic systematics in an oceanic crustal section: evidence from the Samail Ophiolite. Journal of Geophysical Research, 86, 2721-2736. MOORES, E.M., KELLOGG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency': a resolution to the 'ophiolite conundrum'. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies. Geological Society of America, Special Papers, 349, 3-12. MUEHLENBACHS, K. 1986. Alteration of the oceanic crust and the 18O history of seawater. In: TAYLOR, J.W., TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geological Processes. Mineralogical Society of America, Reviews in Mineralogy, 16, 425-444. MUEHLENBACHS, K. 1998. The oxygen isotopic composition of the oceans, sediments, and the seafloor. Chemical Geology, 145, 263-273. MUEHLENBACHS, K. & CLAYTON, R.N. 1976. Oxygen isotope composition of the oceanic crust and its bearing on seawater. Journal of Geophysical Research, 81, 4365-4369. NORTON, D.L. 1984. Theory of hydrothermal systems. Annual Review of Earth and Planetary Sciences, 12, 155-177. O'NEIL, J.R. & TAYLOR, H.P. JR 1967. The oxygen isotope and cation exchange chemistry of feldspar. American Mineralogist, 52, 1414-1437. PALMER, M.R. & EDMOND, J.M. 1989. The strontium isotope budget of the modern ocean. Earth and Planetary Science Letters, 92, 11 -26. PETIT, J.R., JOUZEL, J. & RAYNAUD, D. ET AL. 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature, 399, 429-436. PIEPGRAS, D.J., WASSERBURG, GJ. & DASCH, EJ. 1979.

367

The isotopic of composition of Nd in different ocean basins. Earth and Planetary Science Letters, 45, 223-236. RICHARDS, I.J., CONNELLY, J.B., GREGORY, R.T. & GRAY, D.R. 2002. The importance of diffusion, advection, and host-rock lithology on vein formation: a stable isotope study from the Paleozoic Ouachita orogenic belt, Arkansas and Oklahoma. Geological Society of America Bulletin, 114, 1343-1355. ROZANSKI, K., ARAGUAS-ARAGUAS, L. & GONFIANTINI, R. 1993. Isotopic patterns in modern global precipitation. In: SWART, P.K., LOHMANN, K.C., MCKENZIE, J.A. & SAVIN, S. (eds) Climate Change in Continental Isotopic Records. Geophysical Monograph, American Geophysical Union, 78, 1-36. SAGEN, C. & MULLEN, G. 1972. Earth and Mars: evolution of atmospheres and surface temperatures. Science, 175, 52-56. SHANKS, W.C. Ill 2001. Stable isotopes in seafloor hydrothermal systems: vent fluids, hydrothermal deposits, hydrothermal alteration, and microbial processes. In: VALLEY, J.W. & COLE, D.R. (eds) Stable Isotope Geochemistry. Reviews in Mineralogy and Geochemistry, 43, 469-525. SHOR, G., JR & RAITT, R. 1969. Explosion seismic refraction studies of the crust and upper mantle in the Paciific and Indian Oceans. In: PEMBROKE, J. (ed.) The Earth's Crust and Upper Mantle. American Geophysical Union Geophysical Monograph 13, 225-230. SPENCER, R. J. & HARDIE, L. A. 1990. Control of seawater composition by mixing of river waters and midocean ridge hydrothermal brines. In: SPECNER, R.J. & CHOU, I.-M. (eds) Fluid-Mineral Interactions: A Tribute to H.P. Eugster. Geochemical society Special Publicaion. 2, 409-419. STAKES, D.S. 1991. Oxygen and hydrogen isotope compositions of oceanic plutonic rocks: high temperature deformation and metamorphism of oceanic layer 3. In: TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotope Geochemistry: a Tribute to Samuel Epstein. Geochemical Society, Special Publications, 3, 77-90. STAKES, D.S. 1968. The oxygen isotope geochemistry of igneous rocks. Contributions Mineralogy and Petrology, 19, 1-71. TAYLOR, H.P. JR 1968. The oxygen isotope geochemistry of igneous rocks. Contributions Mineralogy and Petrology, 19, 1-71. TAYLOR, H.P. JR & FORESTER, R.W. 1979. An oxygen and hydrogen isotope study of the Skaergaard intrusion and its country rocks: a description of a 55 my. old fossil hydrothermal system. Journal of Petrology, 20, 355-419. TAYLOR, H.P. JR & FORESTER, R.W. 1979. An oxygen and hydrogen isotope study of the Skaergaard intrusion and its country rocks: a description of a 55 my. old fossil hydrothermal system. Journal of Petrology, 20, 355-419. TOURTELOT, H.A. & RYE, R.O. 1969. Distribution of oxygen and carbon isotopes in fossils of Late Cretaceous age, western interior region of North America. Geological Society of America Bulletin,

368

R. T. GREGORY

80, 1903-1922. VEIZER, J., ALA, D. & AZMY, K. ET AL. 1999. 87Sr/86Sr, (613C, and (618O evolution of Phanerozoic seawater. Chemical Geology, 161, 59-88. VON DAMM, K.L. 1990. Seafloor hydrothermal acitivity: black smoker chemistry and chimneys. Annual Review of Earth and Planetary Sciences, 18, 173-204. WALKER, J.C.G. & LOHMANN, K.G. 1989. Why the isotopic composition of seawater changes with time. Geophysical Research Letters, 16, 323-326.

WALLMANN, K. 2001. The geological water cycle and the evolution of marine (618O values. Geochimica et Cosmochimica Acta, 65, 2469-2485. WOLERY, TJ. & SLEEP, N.H. 1976. Hydrothermal circulation and geochemical flux at mid-ocean ridges. Journal of Geology, 84, 249-275. WRIGHT, E.K. 1987. Stratification and paleocirculation of the Late Cretaceous western interior seaway of North America. Geological Society of America Bulletin, 99, 480-490.

Hydrothermal circulation and metamorphism in crustal gabbroic rocks of the Bay of Islands ophiolite complex, Newfoundland, Canada: evidence from mineral and oxygen isotope geochemistry E. GIGUERE 1 , R. HEBERT 1 , G. BEAUDOIN 1 , J. H. BEDARD 2 & A. BERCLAZ 3 1

Departement de Geologic et de Genie Geologique, Universite Laval, Sainte-Foy, Quebec, Canada G1K 7P4 (e-mail: [email protected]) 2 Commission Geologique du Canada, Centre Geoscientifique de Quebec, 880 Chemin SainteFoy, bureau 840, Quebec, Quebec, Canada G1S 2L2 ^Ministere des Ressources naturelles du Quebec, 545 boul Cremazie E, b.1110, Montreal, Quebec, Canada H2M 2V 1 Abstract: The gabbroic crust of the Ordovician Bay of Islands ophiolite complex formed in an island-arc setting near the North American continental margin. Detailed structural studies on the North Arm Mountain massif provide us with a scheme of syn-oceanic deformation events recorded in the crust. During a first transtensional stage, which generated gabbroic rocks, sheeted dykes and lavas, the temperature of formation of amphiboles in the gabbroic unit fell with time in three steps from 880-745 °C, to 790-680 °C and to 550-500 °C. The Ti, Na and A1IV contents of amphiboles decreased, whereas the Si activity of the fluid increased with time. The first amphibole to form has typical mid-ocean ridge basalt <518OvsMow indicating equilibration with a magmatic fluid or evolved seawater at low fluid/rock ratio. Lower (518O values for some amphiboles (0-2.5%o) indicate the circulation of large volumes of seawater. The lowest (518O values are found in the inner part of the shear zones, which channelled deep infiltration of seawater into the gabbroic unit. During brittle deformation, infiltration of lowtemperature seawater produced prehnite, carbonate and quartz veins, and plagioclase with high (518O. This study documents that the hydration of ophiolitic crust in the Bay of Islands ophiolitic complex occurred mainly along pre-obduction oceanic structures in an intraoceanic setting.

Most oceanic and ophiolite gabbros experienced phibole crystallization in shear zones and the syn-oceanic hydrothermal alteration and meta- associated cracks. Amphibole composition is a morphism. According to spreading rates distinct function of bulk-rock composition, fluid phase and processes occur and display their own meta- temperature. Amphiboles evolve from pargasitic morphic characteristics. Evidence from both ocea- hornblende to hornblende sensu lato, and then to nic and ophiolitic rocks provides insight regarding actinolitic hornblende, suggesting lowering of the these particular problems. For modern oceanic temperatures from upper amphibolite facies to slow-spreading ridges, Mevel & Cannat (1991) greenschist-amphibolite transitional facies. An inhave proposed a model for high-temperature sea- crease in amphibole modal proportion suggests an water penetration distinct from the cracking front increase in water/rock ratio. As temperature falls, model (Lister 1974) in which thermal contraction ductile deformation becomes more localized. Late enhances cracks and allows seawater penetration, acidic rocks are emplaced during or after hydraThermal contraction occurs at temperature below tion associated with shearing. Ductile deformation 500 °C. According to the model of Mevel & ceases when the temperatures drop below 350Cannat (1991), early shearing starts at high tern- 400 °C. Under static conditions, hydrothermal peratures, between 800 and 900 °C, at low water/ metamorphism, from amphibolite facies through rock ratios, and produces granular textures with transitional facies between lower amphibolite and neoblast compositions similar to magmatic phases, greenschist, occurs when gabbro reacts with a With continued shearing, external fluid penetration fluid phase circulating through a crack network. In increases as temperature falls and enhances am- contrast, at fast-spreading ridges, fluid flow is From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 369^00. 0305-8719/037$ 15 © The Geological Society of London 2003.

370

E. GIGUERE ETAL.

uniformly distributed along microfracture networks (Mevel & Cannat 1991; Gillis 1995). At Hess Deep, hydrothermal fluids penetrated at temperature ranging from 450 to 650 °C, locally up to 800 °C. In Hess Deep alteration corresponds to the cracking front model proposed by Lister (1974), but at a higher temperature than predicted (Gillis 1995). This study deals with intraoceanic fluid circulation in natural channels predating obduction of an ophiolitic sequence. In this paper we characterize the massive, statically recrystallized, gabbroic units of North Arm Mountain ophiolitic crust and the compositions of the circulating fluids in deformed gabbroic units. Oxygen isotope geochemical data are available for several ophiolites and oceanic crustal sections. Oxygen isotope compositions mainly depend on the temperature and water/rock ratios (W/R) of water-rock interactions. Fluid temperatures constrain the (518O composition, such that at temperatures higher than 400 °C, rock and minerals are depleted in 18O relative to the initial mid-ocean ridge basalt (MORB) value (Gregory & Taylor 1981; McCulloch et al 1981). A low W/R ratio, however, induces less depletion in 18O relative to the initial MORB value (Vanko & Stakes 1991). Deformation induces lower apparent W/R ratios because a greater volume of minerals is recrystallized with the same volume of water. Downwelling cold seawater is believed to cause enrichment of rocks in 18O, whereas the upwelling high-temperature fluids deplete the rock in 18O (Schiffman & Smith 1988; Alt et al. 1989). These enrichments or depletions are linked to lower W/R ratios in downwelling zones, whereas higher W/R ratios can be hypothesized for upwelling zones. Oxygen isotope compositions from Ocean Drilling Program (ODP) samples and ophiolites show decreasing 18O with depth (Gregory & Taylor 1981; Alt et al 1986, 1989, 1995; Stakes & Taylor 1992). In ophiolites, the transition zone between sheeted dykes and gabbro is the upper limit between 18Oenriched and 18O-depleted rocks (Stakes & Taylor 1992). However, samples from Deep Sea Drilling Project (DSDP) Hole 504B show that the enriched to depleted transition occurs at the transition between lavas and sheeted dykes (Alt et al. 1986, 1989, 1995). In the underlying gabbroic unit, hydrothermal exchanges occur under open-system conditions characterized by high-temperature seawater infiltrating at depth along active shear zones and dyke margins (Harper et al. 1988; Stakes & Taylor 1992). Detailed mapping of the North Arm Mountain massif (NAM) of the Bay of Islands ophiolitic complex (BOIC), Newfoundland, Canada, documents several deformation events with related metamorphism and hydrothermalism. This paper

focuses on the complex crustal section, where deformational and hydrothermal events alternate in time with magmatic events. This complex structural, metamorphic and magmatic history has only recently been documented in the lowest part of NAM ophiolitic crust (Bedard 1991; Bedard & Constantin 1991; Bedard & Hebert 1996, 1998), which was previously described as a wellpreserved and undeformed system (Casey 1980; Casey et al. 1981). The metamorphic mineral assemblage and fabric petrology of representative shear zones corresponds to different deformation events, which can be placed within a comprehensive evolutionary tectonic context (Berclaz et al. 1994). This allows us to quantitatively constrain the temperature conditions during deformation and metamorphism of the ophiolitic crust. The oxygen isotope composition of rocks and minerals further documents the infiltration of fluids in the ophiolitic crust, the chronology of fluid circulation and the relative timing between deformation and fluid events.

Geological setting The BOIC is one of the best exposed and best studied ophiolites in the world. The complex is located on the west coast of Newfoundland and comprises a complete oceanic crustal section formed in a suprasubduction environment (Malpas 1976; Searle & Stevens 1984; Jenner et al. 1991; Berclaz et al. 1994; Bedard & Hebert 1996, 1998; Varfalvy et al. 1996, 1997; Berclaz et al. 1998). The BOIC formed in a slow-spreading environment (Berclaz et al. 1998), in a transtensional intra-arc spreading zone during collision of island-arc lithosphere with North American continental margin (Cawood & Suhr 1992). According to Berclaz et al. (1998), the restored palinspastic tectonic model for BOIC accretion is compatible with a continuous oblique subduction of the oceanic crust under an island arc near the North American margin; then boninitic magmas were underplated during continuous clockwise reorientation of the stress regime related to the oblique subduction. 206pb/238U crystallization ages of 500.6 ± 2.0 Ma and 503.7 ± 3.2 Ma were obtained on zircons from trondhjemite. An age of 485 ± 1 Ma was obtained on apatites from gabbro of the Lewis Hill massif (Kurth et al. 1998). This age is similar to the 2°7pb/206pb crystallization age of 484 ± 5 Ma obtained on zircons from gabbro of the Lewis Hill massif by Jenner et al. (1991). The dynamothermal metamorphic sole related to obduction is dated at 464 ± 5 Ma by 40 Ar/39Ar (Dallmeyer & Williams 1975). This age could represent intraoceanic detachment or a younger age reflecting ophiolitic slab transport

HYDROTHERMAL METAMORPHISM IN GABBROS onto the North American margin. The maximum age between crustal formation and its obduction is thus around 20 Ma. The BOIC comprises four ophiolitic massifs: Table Mountain, North Arm Mountain (NAM), Blow Me Down Mountain and Lewis Hills (Fig. 1). The NAM preserves a complete ophiolitic sequence subdivided into seven stratigraphic units: metamorphic sole (Unit 0); mantle tectonites (Unit 1); lower-crustal interlayered ultramafic and mafic cumulates and intrusions (Unit 2);

371

layered and foliated adcumulate crustal gabbros, porphyritic gabbros, hornblendites and plagiogranites (Unit 3); sheeted dykes (Unit 4); pillow and brecciated lavas (Unit 5); red chert and argillite (Unit 6) (Casey 1980; Casey et al. 1981; Bedard 1991; Berclaz et al. 1994). This study focuses on specific zones within the NAM sequence, where we studied recrystallization, deformation and fluid infiltration events that occurred within Unit 3, the crustal gabbroic section.

Fig. 1. Map of Bay of Islands ophiolitic complex modified from Casey et al. (1983, 1985).

372

E. GIGUERE ETAL.

Fig. 2. Structural map of the gabbroic unit of North Arm Mountain massif with the six structural domains from Berclaz (unpubl. data). Sampled shear zones are given by their names.

Lithological and structural units

The crustal gabbro section is about 5 km thick and varies in primary mineralogy and deformation.

The western section consists of olivine gabbros,

troctohtes and gabbros with minor trondhjemites, feldspathic wehrlites, anorthosites and pegmatites, The eastern section consists of plagioclase por-

HYDROTHERMAL METAMORPHISM IN GABBROS

373

Fig. 3. Map of sampled site TB09.16. <518O values (%o) are given near sample numbers for minerals (pi, plagioclase; am, amphibole) and for whole rock (wr). Plagioclase compositions in An are given near sample numbers.

phyritic microgabbros and Fe-Ti-oxide and hornblende-rich olivine gabbros (Berclaz et al. 1994). The crustal gabbroic unit is subdivided into eight lithological subunits (units 3a-3h) (Fig. 2). Each subunit is 200 m to 2 km thick and is composed of many intrusive sill-like igneous bodies (Bedard & Constantin 1991). The strike of all lithological subunits was transposed by folding and stretching. Several mylonitic shear zones cut the crustal section at various angles, from the Moho up to the sheeted dyke unit. Granulite-facies porphyroclastic gneisses irregularly alternate with hydrated mylonitic shear zones at amphibolite- to greenschistfacies assemblages. Most of the shear zones are narrow deformational corridors a few metres thick that are characterized by an extreme reduction of grain size (<0.05 mm). These mylonitic shear zones developed synchronously with magmatic activity. Magmatic bodies are found parallel to the

foliation of the porphyroclastic gabbro. Thus, the shear zones might have played a role in controlling the lateral and vertical migration of residual melts, and the ascent of primitive magmas. Berclaz et al. (1998) recognized six structural domains in units 2 and 3 based on intrusive relations and structural fabric orientations. We investigated syn-oceanic shear zones developed in structural Domains 3 and 4 (Figs 2 and 3). Domain 3 forms a structural block from the lower half of the gabbroic crustal unit 3 up to the sheeted dykes in the northeastern part of the map area (lithological subunits 3c and 3e in Fig. 2). In the lower half of Domain 3, porphyroclastic granulitic gneisses and mylonites are adjacent to proto- and ultramylonites characterized by a lower grade of alteration. Domain 3 is characterized by a north-NE planar fabric, which is steeply to moderately dipping and a linear fabric plunging

374

E. G I G U E R E £ T ^ L .

moderately to the east. Domain 3 is controlled by north-NE listric dip-slip shearing. Domain 4, found in the western part of the upper gabbroic unit (in lithological subunits 3a, 3b and 3c in Fig. 2), is characterized by ultramylonites of 20-50 cm thickness and protomylonites of >20 m thickness, with planar fabrics that dip steeply to the NW-SE and linear fabrics that plunge steeply to the east or shallowly to the NW. The later mylonites, developed in Domain 4, are at a high angle with respect to those developed in Domain 3. Domain 4 shear zones represent strike-slip faults slightly oblique to the palaeoridge (Berclaz et al. 1998). Analytical methods Electron-microprobe analyses of clinopyroxene, orthopyroxene, olivine, plagioclase, amphibole, chlorite, prehnite and epidote were carried out at Universite Laval with an ARL electron microprobe and at McGill University with a JEOL 8900 system. The core and rim of each grain were analysed to verify homogeneity, and neoblasts, reaction rims and veins were also analysed to study temporal changes in composition. Oxygen isotope analyses were made on whole rocks (n = 32) and pure mineral concentrates (plagioclase (n = 34), amphibole (n = 18) and clinopyroxene (n = 5)). Mineral separates from medium-grained rocks were handpicked under the binocular microscope. Plagioclase and amphibole from fine-grained rocks were concentrated using heavy liquids and handpicked under the binocular microscope. The sample purity was further verified using X-ray diffraction. Oxygen extraction was performed at Universite Laval using BrFs following the method of Clayton & Mayeda (1963). The reaction lasted at least 12 h at 500 °C for plagioclase and 600 °C for clinopyroxene, amphibole and whole rock. The oxygen is converted to CO2 by reaction with hot graphite. The CO2 was analysed by mass spectrometry at the G.G. Hatch isotope laboratory at Ottawa University. Based on replicate analyses of our internal standard quartz Kl, a precision of ±0.18%o is estimated. The results are given in 6 notation and reported relative to Vienna Standard Mean Ocean Water (V-SMOW) in per mil (%o). Petrography Static recrystallization Olivine gabbros and gabbros hosting the shear zones have undergone static recrystallization and hydrous replacement, and display no obvious fabric, with the exception of site MRL.24, where

the gabbro has a foliation perpendicular to a shear zone. In thin section, clinopyroxene can be pseudomorphed by colourless hornblende. Subsequently, a thin rim of brown hornblende formed around clinopyroxene, olivine, or colourless amphibole and later clinopyroxene is progressively pseudomorphed by green hornblende (Fig. 4b). Chlorite rims can form around mafic minerals. In olivine gabbro, the olivine is generally replaced by serpentine and chlorite or iddingsite and magnetite. Dynamically recrystallized gabbros Dynamically recrystallized gabbros are found in shear zones and protoliths are the same as gabbro hosting the shear zones. Shear zones show four main fabrics. Fabric type 1 ranges from extensively recrystallized gneiss to completely recrystallized mylonite. The mylonites display a fine-grained, equigranular, polygonal texture when completely recrystallized. Extensively recrystallized gneisses have pyroxene porphyroclasts surrounded by smaller polygonal plagioclase, pyroxene and hornblende neoblasts. In mylonite, magmatic plagioclase and clinopyroxene form porphyroclasts. Pre-kinematic argillization is locally suggested by the occurrence of plagioclase porphyroclasts, which are partly altered to clay, whereas plagioclase neoblasts are limpid and devoid of inclusions. The majority of the plagioclase, clinopyroxene and hornblende are recrystallized into neoblasts with a polygonal texture. Synkinematic brown or green hornblende is the major metamorphic mineral that pseudomorphed clinopyroxene. Fabric type 2 ranges from protomylonitic to porphyroclastic (Fig. 4c). In addition to recrystallization, the rock becomes very fine grained except for medium-grained clinopyroxene, plagioclase and amphibole porphyroclasts. Porphyroclasts are deformed and show 6-type kinematic indicators. Magmatic clinopyroxene is pseudomorphed by brown hornblende followed by green hornblende. No relics of olivine are present. The groundmass is composed of very fine-grained fibrous actinolite and fine-grained, recrystallized plagioclase. Fabric type 3 corresponds to mylonitic to ultramylonitic fabrics. Plagioclase and hornblende are the only major mineral phases found in the mylonitic to ultramylonitic rocks. Plagioclase neoblasts are very finely recrystallized. Very finegrained acicular green amphibole occurs in the matrix with plagioclase. No relics of olivine are present. Fabric type 4 corresponds to cataclastic to mylonitic fabric produced by brittle-ductile defor-

HYDROTHERMAL METAMORPHISM IN GABBROS

375

Fig. 4. (a) Site TA02.06 from structural Domain 4 showing a mylonite, 5-15 cm wide, cut by a dyke 1 m wide; (b) sample 94417 from site MRL.06 showing brittle deformation with fractures filled by green hornblende; (c) sample 92465 from site TB09.16 showing porphyroclastic texture.

mation. Samples are crushed to various degrees, producing porphyroclastic to mylonitic fabrics in the more recrystallized zones. This type is characterized by plagioclase and clinopyroxene porphyroclasts. Generally, clinopyroxene porphyroclasts are extensively replaced by brown or green hornblende. Early crystallized brown hornblende is surrounded by a corona of green hornblende. This fabric is also characterized by crystallization of plagioclase and green hornblende neoblasts. Most shear zones in Domain 3 have type 1 fabrics with polygonal textures. Clinopyroxene neoblasts are more abundant in shear zones from Domain 3. One shear zone from this Domain shows fabric type 2 with porphyroclastic to mylonitic texture. Fabric type 2 predominates in Domain 4 and is characterized by protomylonitic to porphyroclastic to ultramylonitic textures. Fabric type 3 is found in the upper zone of the gabbroic unit near the gabbro-sheeted dyke transition. Amphibolitization of clinopyroxene is most important in shear zones from Domain 4.

Mineral chemistry Clinopyroxene Clinopyroxene shows very restricted compositional variations from augite to diopside in Domain 3 (Wo44_5o En3o-^3 Fsi2-2o) and Domain 4 (Wo4i^8 En42-48 Fsog-14) (Fig. 5; Tables 1 and 2). Domain 4 clinopyroxenes are similar in composition to clinopyroxene found in Oman (Eri4o_54, Ti 0.002-0.017, Al 0.020-0.114; Stakes & Taylor 1992). Magmatic clinopyroxenes in Domain 3 occur in syn- to late-kinematic olivine gabbronorite and have core compositions of Wo43^4 En43 Fsi4 and rim compositions of Wo44_^5 £1142-43 Fsi2-i3- Clinopyroxene neoblasts in Domain 3 have composition of Wo48_5o En3o_36 Fsi6-2oThese clinopyroxenes have higher diopside contents than neoblasts from ODP Hole 735B (Wo45-t7 En36_48 Fs8-2o, Stakes et al. 1991). Clinopyroxene neoblasts and magmatic clinopyroxenes have the same Al content (Al 0.045-0.216 and 0.061-0.218, respectively). However, the Ti content is slightly higher in the rim of magmatic

376

E. GIGUERE ETAL.

Fig. 5. Pyroxene compositions for structural Domains (a) 3 and (b) 4, and (c) examples of compositional variations from two sites of Domain 4 (TA02.06-07, olivine gabbro; MRL.24, gabbro; r, rim; c, core; p, porphyroclast; n, neoblast) according to protolith (O, ferrogabbro; A, gabbro; n, olivine gabbro; o, gabbronorite; *, trondhjemite) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static) from 1 atm solvus of Lindsley (1983). Entire spectrum of data for North Arm Mountain Massif is also shown (grey field; Bedard, unpubl. data).

clinopyroxene compared with neoblasts and cores of magmatic clinopyroxene (Ti 0.017-0.028, 0.003-0.005 and 0.005, respectively). In Domain 4, magmatic clinopyroxenes in undeformed samples have core compositions of Wo45_^7 Eri4i^5 Fsg-14 and rim compositions of Wo42-4? En42^g Fs9_i4. In dynamically recrystallized samples, clinopyroxene porphyroclasts have core compositions of Wo4i_48 £012^7 Fs9_i2 and rim compositions of W042-48 Eri42-48 Fsg-ii- Clinopyroxene neoblasts have similar compositions to clinopyroxene porphyroclasts; however, Wo contents are slightly higher (Wo47_^9 £1142-45 £59-12). Magmatic clinopyroxenes have higher Ti and Al contents than neoblasts. Clinopyroxene porphyroclasts have compositions overlapping those of neoblasts and magmatic clinopyroxenes (Ti 0.002-0.038, 0.001-0.012 and 0.008-0.028; and Al 0.0260.191, 0.026-0.101 and 0.076-0.168, respectively). Compositions of clinopyroxene neoblasts in Oman (Eo^) (Stakes & Taylor 1992) and in Hess Deep (En4i_42) (Lecuyer & Reynard 1996) show the same variations as clinopyroxene neoblasts from Domain 4. Protolith compositional variations partly control the mineral compositions. For instance, the Mgnumber of ferrogabbro clinopyroxene (61-72) is at the lower end of Mg-number of clinopyroxenes in gabbro, olivine gabbro and olivine gabbronorite (64-88). In one specific gabbro, a more calcic clinopyroxene (Wo49 Eri42 Fsog) is cut by a trondhjemite

dyke. Porphyroblastic clinopyroxene in a trondhjemite dyke overlaps this composition (Wo49_50 En38^5 Fsos-12)Some clinopyroxenes display compositional variations across individual grains. Shear zones show an increase of clinopyroxene Mg-number and Wo content from undeformed to recrystallized samples (Fig. 5c). Clinopyroxene from undeformed samples shows a decrease of Mg-number and of Wo and En contents from core to rim (Fig. 5c).

Olivine Olivine in an olivine gabbro with static recrystallization texture has Fo7i_77 composition (Table 1) whereas olivine from an olivine gabbronorite intrusion is more iron rich (Foes). No within-grain zonation is observed.

Plagioclase A large variation of plagioclase An content is seen in Domain 3 (Fig. 6; Tables 1 & 3). Plagioclase neoblasts in samples with a lower abundance of amphiboles contain the more calcic plagioclase (Angg) whereas plagioclase neoblasts in samples with more amphibolitized clinopyroxene contain more sodic plagioclase (An^g), consistent with increasing replacement under amphibole-grade conditions. No variation of An content is found between partly argillized porphyroclasts and coexisting neoblasts.

Table 1. Synthesis of petrography, mineralogy and isotope composition for plagioclase, amphibole, pyroxene, olivine and whole rock Sample

Protolith

Texture

Magmatic minerals

Metamorphic minerals

Amphibole

pi

px

<5180

ol pi

am

cpx

wr

Structural Domain 3 TA17.STN09-

15 92494

ol gabbro

porphyroclastic (type 1)

pi 15, cpx 5, ol 1

92497

gabbro

nematogranoblastic (type 1)

pi 10, cpx 1

92498

ferrogabbro

nematogranoblastic (type 1)

pi 5, cpx 10

pi 35, cpx 22, br hbl 10, mgt 10, serp 2 cpx 5+br hbl 30^act 30 (Na+K)A: 0.240.33 A1IV: 0.91-1.23 Ti: 0.11-0.15 pi 20+cpx 23+br hbl 40 (Na+K)A: 0.660.73 A1IV: 1.65-1.74 Ti: 0.24-0.25 pi 30+cpx 14+br hbl 35 (Na+K)A: 0.530.70 A1IV: 1.54-1.80 Ti: 0.20-0.25 pi 30+cpx 20+br hbl 15

92495

gabbro

nematoblastic (type 1)

pi 35, cpx 5

92496

gabbro

nematogranoblastic (type 1)

pi 5

92499

gabbro

nematoporphyroclastic (type 1)

pi 24

pi 23+cpx 20+br hbl 27

92700

ferrogabbro

nematogranoblastic (type 1)

pi 5

pi 50+cpx 20+br hbl 20

92701

gabbro

nematogranoblastic (type 1)

pi 10

pi 35+br hbl 50

92702

ol gabbro-norite

granoblastic

pi 45, ol 20, cpx 10, opx 5

brhbl 15^act5

An54-59

4.9

7.4

An54-63

4.7

Wo49 En33 Fsis

An^o

An78^88

3.2

W048_50 En30-36

3.9

FS 16-20

Other shear zones 94703

gabbro

nematoblastic (type 1)

pi 30

94704

gabbro

nematoblastic (type 1)

pi 40

94705

gabbro

porphyroclastic (type 1)

pi 5, cpx 5

94706

gabbro

porphyroclastic (type 1)

pi 10

pi 35+gr hbl 33->pr 1+clay

1 pi 30+gr hbl 24^pr 5+clay 1 pi 25+cpx 25+br hbl 30-^gr hbl 10 pi 30+cpx 5+br hbl 15->act 40+pr

(Na+K)A: 0.280.35 A1IV: 1.16-1.26 Ti: 0.11-0.15 (Na+K)A: 0.570.58 A1IV: 1.57-1.58 Ti: 0.22-0.23 (Na+K)A: 0.390.46 A1IV: 1.33-1.49 Ti: 0.13-0.15 (Na+K)A: 0.710.72 A1IV: 1.59-1.60 Ti: 0.19-0.20

An53_65

An^o

5.7

Wo48 En32 Fs20

6.1

An57-6i

An61_62

W044^5

Fo63

5.4

5.6

4.3

En42-43 Fs28^33; Wo2,En65, Fs33

5.7 6.0 5.8 6.3

(continued overleaf)

Table 1. (continued} Sample

Protolith

Texture

Magmatic minerals

Amphibole

Metamorphic minerals

pi

px

<5 18 O

ol pi

94970 92758

gabbro gabbro

porphyroclastic (type 1)

pi 10 pi 10, cpx 27, opx 5

pi 40+act 35, ep 10 pi 40+cpx 22+opx 5+ br hbl 3^cl 1

Ans6~59

am

cpx

wr

6.1 6.4

W043^5 En42^3 FSi 3 _i5 Wo2_3 En69-71 Fs27-29

92759 94961

gabbro ol gabbro

granoblastic (type 1) granoblastic (type 1)

pi 5, cpx 5, mgt 5 pi 10

94925

gabbro

nematoporphyroclastic (type 2)

pi 25, cpx 2

94926 94928 94932

ol gabbro ol gabbro gabbro

granoblastic granoblastic (type 1) nematoporphyroclastic (type 2)

pi 50, cpx 35, ol 15

Structural Domain 4 TA02. 06-07 94400

trondhjemite

porphyroblastic

pi 60, sph 2

94689

ol gabbro

porphyroclastic (type 2)

pi 13, cpx 2

pi 15

pi 15+cpx 60^clay 10 pl40+cpx43H-o!5+brhbl 1—>serp 1 pi 25 + cpx 3+gr hbl 15^act

9.3 6.6 4.1

30 br hbl 1—>act 1—>serp+mgt 3 pi 50+cpx S^serp 5 pi 25+cpx 5+gr hbl 30->act 24^cl 1

act 5—>cl 2+ep 2->cpx 27-^cc 2 pi 35, br hbl 10^grhbl40

5.7 5.3 4.9

-1.0

6.6

(Na+K)A: 0.30- An73 77 0.55 A1IV: 0.39-1.34 Ti: 0-0.06

Wo42-47

(Na+K)A: 0.09- An73-77 0.24 A11V: 0.47-0.48 Ti: 0.05-0.08 (Na+K)A: 0.35- An32^3 0.57 A1IV: 1.16-1.40 Ti: 0.08-0.18

W042_47

94690 94691

diabase ol gabbro

granoblastic granoblastic

pi 24 pi 45, cpx 20, ol 20

hbl l->act 5^cl 5-^ep 65 brhb!2^act 13

94692

ol gabbro

porphyroclastic (type 2)

pi 15, cpx 5

pi 50, br hbl 10-+grhbl20

tdj vein

porphyroclastic

pi 75, br hbl 20

ep5

TA01.02 94603

gabbro

porphyroclastic (type 2)

pi 10, cpx 10

pi 38, hbl br 1-^grhbl 25^act 15->cl 1

94604 94605 94606 94607

gabbro gabbro gabbro-norite gabbro

porphyroclastic (type 2) granoblastic granoblastic granoblastic

pi 5, cpx 10 pi 30, cpx 30 pi 20, cpx 10, opx 5 pi 45, cpx 44

pi 55+gr hbl 29^cl 1 hbl 28^act lO^cl 2 hbl 35-^act 29-+cl 1 br hbl 2->gr hbl 2-^tr5-+cl (Na+K)A: 0.33- An8(M?3 2 0.59 A1IV: 1.71-1.74 Ti: 0.23-0.25

5.9

3.8

4.6

Ell44^t8 Fs9_n 4.8

En4i^6 Fsn-i4 W044^9

Fo71_72

5.0

8.0

5.5

6.6

5.2

5.9

7.9

5.8

En32 Fs9-i2

(Na+K)A: 0.27- An78_83 0.30 A1IV: 0.76-0.85 Ti: 0.05

5.0

7.6

W045^7 En44^6

Fs9

6.5 5.5

3.5

4.5

4.2 4.7

5.1

94608

gabbro

granoblastic

pi 60, cpx 22

am inc lO^br hbl 2-> act 5^cl 1

TB09.16 92458

gabbro

granoblastic

pi 33, cpx 5

92459 92460

gabbro gabbro

granoblastic granoblastic

pi 65 pi 40, cpx 7

brhb!2^pl 10+grhbl 20-^act 30 pi 15+act20 hbl 15-+act30^cl5+ep3 (Na+K)A: 0.00-

An74-8o

W046^7

An73-78

W046^7

3.8

1.7

3.4

0.57 IV

92463

gabbro

granoblastic

pi 55, cpx 15

hbl lO^act 15->cl 5

A1 : 0.15-1.83 Ti: 0.00-0.34 (Na+K)A: 0.00-

1.9

Ell43_44

Fsn 3.2

3.8

0.54 IV

A1 : 0.17-1.69 Ti: 0.00-0.29 92464 92465

br hbl vein gabbro

granoblastic mylonitic (type 2)

hbl br 100 pi 15

pi 35, hbl 15-»act 35^cl 1 (Na+K)A: 0.02-

En44^5

Fs9-10 An71_83

0.54 A1IV: 0.39- 1.37 92466

gabbro

porphyroclastic (type 2)

pi 20

pi 20+hbl 60

Ti: 0.02-0.11 (Na+K)A: 0.19-

Wo47

4.0

2.0 1.0

2.2

4.4

0.0

4.0

£044

Fs9 An75_76

0.37 0.2

A1IV: 0.71-1.39 Ti: 0.02-0.05 92468 92713 92715 92716

927 16b 92717 92718

tdj vein gabbro gabbro ol gabbro tdj vein gabbro ol gabbro

granoblastic granoblastic mylonitic (type 2) granoblastic granoblastic granoblastic granoblastic

pi 65, hbl br 30, sph 1 pi 40, cpx 30 pi 5, cpx 5 pi 60, cpx 15, ol 5 pi 60, cpx 30 pi 60, cpx 12 P 145,cpxl5,ol 15

grhblS br hbl 5-^act 20-^cl 5 pi 35+gr hbl 55 hbl lO^act 10

gr hbl 30 brhbl5^actl0^c!5 brhb!5->act 10->cl 5

(Na+K)A: 0.00-

An78

82

W045^7

Fo75_77

8.9

4.3

5.0

1.0

4.9

4.2

0.65 IV

A1 : 0.32-1.97 Ti: 0.00-0.26 92-719 MRL.06 94414 94415

ol gabbro

granoblastic

pi 60, cpx 10, ol 10

br hbl 5^act5-^cl5

ol gabbro ol gabbro

porphyroclastic (type 2) porphyroclastic (type 2)

pi 15, cpx 15 pi 15, cpx 15

pi 55+gr hbl 10 pi 35+gr hbl 32

£0(5^6 Fs9^o

4.4

(Na+K)A: 0.26-

An61_68

3.3

3.7 2.6

Wo46

0.41 A1IV: 0.61-1.14 Ti: 0.05-0.14 94416 94417

94418 94419

ol gabbro ol gabbro

tdj vein ol gabbro

porphyroclastic (type 2) granoblastic

granoblastic granoblastic

pi 25, cpx 25 pi 50, cpx 35

pi 60, cpx 7

pi 35+gr hbl 15 brhbl l->act 10-»cl 2

(Na+K)A: 0.00-

£1140^1 Fs ]3 _i4

An60^9

W044-.5

0.15 AF: 0.22- 1.32

En42-^4

Ti: 0.01-0.21

Fs,3-,4

pi 10+grhbl 10-*act 10—>cl

5.8

3.1

8.7 5.1

4.1

5.2

4.1 4.7

3 MRL.24 94653 94654

gabbro gabbro

porphyroclastic (type 2) granoblastic

pi 15, cpx 5 pi 50, cpx 17

94655

gabbro

cataclastic to porphyroclastic (type 3)

pi 15, cpx 15

pi 35, br hbl 35-^act 35 pi 10+cpx 3, br hbl 8—>act l->cl 1 pi 45+cpx 5, br hbl 10—>act (Na+K)A: 0.16-

10

An73_74

W04,_46

0.50

(continued overleaf)

Table 1. (continued) Sample

Protolith

Texture

Magmatic minerals

94656

gabbro

porphyroclastic (type 2)

pi 15, cpx5

94657

gabbro

granoblastic

pi 55, cpx 15

94658

tdj vein gabbro

granoblastic granoblastic

pi 75, hbl br 17 pi 45, cpx 15

Metamorphic minerals

Amphibole

A1IV: 0.75-1.64 Ti: 0.01-0.25 pi 30, br hbl 10-»gr hbl (Na+K)A: 0.000.33 30^tr 5 A1IV: 0.29-1.67 Ti: 0.00-0.27 pi 5, br hbl 15^act5^cl7 (Na+K)A: 0.230.64 A1IV: 0.99-1.69 Ti: 0.06-0.38 gr hbl 5, qtz 3 pi 5, hblbrl5-+grhb!7 (Na+K)A: 0.000.57 A1IV: 0.54-1.80 Ti: 0.00-0.31

CBR.05 94645

ol gabbro

granoblastic

pi 65, cpx 20

gr hbl S^act 5-»cl 5

94646 94647 94648 94649 94650 94651

ol gabbro diabase ol gabbro diabase leucogabbro ol gabbro

granoblastic microlitic porphyroclastic (type 2) granoblastic porphyroclastic (type 2) cataclastic to porphyroclastic (type 3)

pi pi pi pi pi pi

act 5—>serp 20 gr hbl 25->cl 2 + ep 20 pi 45, act 20^ep 10+ cc 3 act lO^ep 42^cl 3 act 10-^trem 10-^cl 5 pi 30, hbl 15-*act 5,cpy (Na+K)A: 0.000.41 A1IV: 0.25-1.83 Ti: 0.01-0.28

CBR.06 94562

gabbro

mylonitic (type 2)

pi 15, cpx 5

60, cpx 10 45, cpx 5 15, cpx 7 45 70, cpx 5 50

pi 45, act 35

(Na+K)A: 0.150.17 A1IV: 0.46-0.54 Ti: 0.06-0.11

(Na+K)A: 0.000.20 A1IV: 0.19-1.81 Ti: 0.00-0.039

pi

px

(5180

ol pi

am

6.0

4.1

cpx

wr

En44_47 FSg-12

An64-74

An56_73

W042^7

6.1

4.6

5.1 5.5

Ell42^8 FSiQ-28

An74-75

Wo45

8.3 5.2

5.1 5.1

En43^4 Fs12

An74_75

Wo45^7

6.8

6.2

8.8

6.8 7.7

En44^i5 Fs8_io

An74_84

7.6

3.9

5.3

An74_84

5.6

2.5

4.6

pi, plagioclase; am, amphibole; px, pyroxene; ol, olivine; wr, whole rock; cpx, clinopyroxene; opx, orthopyroxene; hbl, hornblende; act, actinolite; cl, chlorite; ep, epidote; pr, prehnite; mgt, magnetite; gr, green; br, brown.

HYDROTHERMAL METAMORPHISM IN GABBROS

381

Table 2. Clinopyroxene compositions Analysis: Sample:

1 92702

2 92498

3 94657

4 94657

5 94655

6 94655

7 94691

Si02 Ti02 A1203 Cr203

50.63 0.72 2.02 0.04 7.74 0.40 14.92 21.82 0.34 0.00 98.63 1.914 0.086 0.020 0.090 0.004 0.001 0.245 0.000 0.245 0.013 0.841 0.884 0.025 0.000 4.032 0.449 0.427 0.124

51.88 0.10 1.30 0.00 9.85 0.37 12.15 23.74 0.32 0.00 99.72 1.962 0.038 0.003 0.058 0.020 0.000 0.312 0.000 0.312 0.012 0.685 0.962 0.023 0.000 4.017 0.491 0.350 0.159

50.56 0.52 2.59 0.24 6.50 0.33 16.05 21.14 0.61 0.02 98.56 1.900 0.100 0.015 0.115 0.015 0.007 0.204 0.011 0.193 0.011 0.899 0.851 0.044 0.001 4.047 0.435 0.460 0.105

50.80 0.54 2.85 0.24 6.72 0.24 16.59 20.26 0.57 0.00 98.81 1.898 0.102 0.015 0.126 0.024 0.007 0.210 0.006 0.204 0.008 0.924 0.811 0.041 0.000 4.041 0.417 0.475 0.108

49.28 0.90 3.89 0.21 7.32 0.31 15.78 19.16 0.57 0.00 97.42 1.871 0.129 0.026 0.174 0.045 0.006 0.232 0.000 0.232 0.010 0.893 0.779 0.042 0.000 4.034 0.409 0.469 0.122

49.49 1.09 2.83 0.26 5.86 0.28 15.97 20.83 0.39 0.00 97.00 1.884 0.116 0.031 0.127 0.011 0.008 0.187 0.013 0.174 0.009 0.907 0.850 0.029 0.000 4.031 0.437 0.467 0.096

50.42 0.59 2.71 0.13 7.78 0.27 15.99 20.31 0.38 0.00 98.58 1.898 0.102 0.017 0.120 0.018 0.004 0.245 0.022 0.223 0.009 0.897 0.819 0.028 0.000 4.037 0.418 0.457 0.125

FeO MnO MgO CaO Na2O

K2O Total

Si A1IV

Ti Al A1VI

Cr Fe Fe3+ Fe2+

Mn Mg Ca Na K Total Wollastonite Enstatite Ferrosilite

8 94691

9 94692

10 94692

11 94692

49.09 53.58 54.10 53.55 0.07 0.59 0.07 0.05 3.82 0.71 0.61 0.61 n.a. 0.13 n.a. n.a. 8.76 6.33 6.93 6.00 0.33 0.24 0.33 0.22 15.53 15.00 15.33 15.43 24.74 22.78 24.31 24.95 0.29 0.32 0.36 0.29 0.01 0.00 0.00 0.05 100.73 101.52 102.11 101.15 1.834 1.964 1.966 1.965 0.166 0.034 0.036 0.035 0.017 0.002 0.002 0.001 0.168 0.026 0.030 0.026 0.002 0.000 0.000 0.000 n.a. 0.004 n.a. n.a. 0.274 0.212 0.192 0.184 0.107 0.006 0.009 0.012 0.167 0.206 0.183 0.173 0.008 0.010 0.010 0.007 0.841 0.836 0.838 0.844 0.912 0.955 0.963 0.981 0.023 0.026 0.021 0.020 0.000 0.000 0.000 0.002 4.075 4.033 4.027 4.032 0.482 0.451 0.476 0.488 0.421 0.413 0.418 0.420 0.135 0.092 0.106 0.096

Analyses from Domain 3: 1, rim of undeformed clinopyroxene in gabbronorite intrusion; 2, neoblast in mylonitic ferrogabbro. Analyses from Domain 4: 3 and 4, core and rim from same undeformed clinopyroxene in host gabbro; 5 and 6, porphyroclast core and rim in cataclastic gabbro; 7 and 8, rim and core from same undeformed clinopyroxene in host olivine gabbro; 9, 10 and 11, porphyroclast core and rim, and neoblast core in porphyroclastic olivine gabbro. n.a., not analysed. FeO, total iron calculated as FeO.

In Domain 4, the plagioclase composition range is An32_38 (Fig. 6b). Plagioclase compositions become more sodic from core to rim. Thus, porphyroclast core compositions are Any^^ and porphyroclast rim compositions are An7o_76- In individual samples, plagioclase becomes more sodic from porphyroclasts to neoblasts (Fig. 6c), but the entire field of neoblast compositions encloses primary and porphyroclast compositions (An32-43 and An6i_82, An68-so, An35_39 and Anyo_84, respectively). The more sodic plagioclase is found in samples cut by trondhjemite veins.

Amphibole Amphiboles in Domain 3 are aluminous and show little compositional variation. They are pargasitic hornblende, hastingsitic hornblende and magnesiohornblende according to the classification of Leake (1978) (Fig. 7, Tables 1 & 4). Generally, in TA17.STN09-15 shear zone, only one amphibole

type is present in each sample; however, one sample shows pargasitic hornblende replaced by hastingsitic hornblende. Hastingsitic hornblende coexists with An48 plagioclase whereas the magnesiohornblende occurs with more calcic plagioclase (An5^63)- Amphiboles with dynamic recrystallization textures have lower Mg-number. In Domain 4, olivine and clinopyroxene are replaced by pargasite, tschermakite, tschermakitic hornblende, hastingsite, hastingsitic hornblende and edenitic hornblende. These are rimmed by magnesio-hornblende, which in turn is replaced by activolitic hornblende, actinolite and tremolite. This amphibole succession is similar to that found in oceanic gabbros in slow-spreading setting such as the Mid-Atlantic Ridge, Mid-Cayman Rise and Southwest Indian Ridge (Mevel 1987; Hebert & Constantin 1991; Mevel & Cannat 1991; Stakes et al 1991; Gaggero & Cortesogno 1997). Some shear zones, however, do not show this amphibole zonation sequence. For example, in shear zone

382

E. GIGUERE ETAL.

Fig. 6. Plagioclase compositions for structural Domains (a) 3 and (b) 4, and (c) examples of compositional variations from two sites of Domain 4 (MRL.06, olivine gabbro; MRL.24, gabbro; r, rim; c, core; p, porphyroclast; n, neoblast) according to protolith (O, ferrogabbro; A, gabbro; n, olivine gabbro; o, gabbronorite) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static). Entire spectrum of data for North Arm Mountain massif is also shown (grey field: Bedard, unpubl. data).

TB09.16, magnesio-hornblende formed after actinolitic hornblende whereas in shear zone CBR.05, the magnesio-hornblende is replaced by tschermakite. Amphiboles in Domains 3 and 4 show a Titschermakitic substitution in a Ti-AlIV diagram (Fig. 8) with Ti contents ranging from 0.00 to 0.38 p.f.u. (per formula unit) and A1IV contents ranging from 0.15 to 1.97 p.f.u. Ti content in tschermakitic hornblende has a high compositional variability, ranging from 0.00 to 0.31 p.f.u., despite the high A1IV content with a small range from 1.53 to 1.97 p.f.u. Most of the amphiboles of Domains 3 and 4 plot along the pargasitic substitution trend in the AlIV-(Na + K)A diagram (Fig. 9). In Domain 4, hastingsitic hornblende, pargasitic hornblende and tschermakitic hornblende show an alkali variation at fixed A1IV composition and horizontal trajectories are seen between pargasitic and tschermakitic end-members (Fig. 9b). These amphiboles show inverse correlations of AlVI/Ti, Fe3+/Ti and Fe3+/(Na + K)A, and Mn/(Na + K)A in some shear zones. Mevel (1987) interpreted inverse correlation of AlVI/Ti for the same A1IV content in Mid-Atlantic Ridge amphibole as the result of a

reacting fluid phase Ti/Al ratio controlling amphibole composition at low water/rock ratios. Thus, trends between pargasitic and tschermakitic endmembers in the AlIV-(Na + K)A diagram and inverse correlations are probably produced by variations in Ti/Al, Ti/Fe and alkali/Fe ratios in the reacting fluid phase at low water/rock ratios.

Chlorite Chlorite from Domain 4 forms coronas around olivine, clinopyroxene and amphibole, or appears at the selvages of veins. Most chlorites fall in the fields of pychnochlorite and clinochlore according to the classification of Hey (1954). They have similar compositions in all shear zones (Table 5). Chlorites from coronas or veins within the same sample have uniform compositions. A more siliceous chlorite, diabantite, is found in a shear zone that also contains quartz veins, indicative of a higher silica activity.

Prehnite Prehnite occurs in veins and replaces plagioclase at vein selvages. It has low Fe contents with

Table 3. Plagioclase compositions Analysis: Sample:

1 92702

2 92701

3 94692

4 94692

5 94657

6 94657

7 94655

8 94655

9 92463

10 92463

11 92466

12 92465

SiO2 A1203 FeO MnO MgO CaO Na2O K2O Total Si Al Fe Mn Mg Ca Na K Total Anorthite Albite Orthoclase

54.05 29.48 0.22 n.a. 0.15 12.77 4.31 0.03 101.0 2.422 1.557 0.008 n.a. 0.010 0.613 0.375 0.002 4.987 0.620 0.379 0.002

51.18 29.97 0.54 0.00 0.00 11.31 4.59 0.15 97.74 2.374 1.639 0.021 0.000 0.000 0.562 0.413 0.009 5.017 0.572 0.420 0.009

58.60 25.79 0.06 n.a. 0.14 7.37 7.69 0.02 99.67 2.628 1.363 0.002 n.a. 0.009 0.354 0.669 0.001 5.026 0.346 0.653 0.001

58.23 26.52 0.09 n.a. 0.14 8.14 7.08 0.03 100.2 2.599 1.395 0.003 n.a. 0.009 0.389 0.613 0.002 5.011 0.388 0.610 0.002

54.15 31.19 0.23 n.a. 0.05 12.96 4.45 0.03 103.1 2.381 1.616 0.008 n.a. 0.003 0.611 0.379 0.002 5.001 0.616 0.383 0.002

54.35 30.66 0.17 n.a. 0.06 12.75 4.59 0.03 102.6 2.399 1.595 0.006 n.a. 0.004 0.603 0.393 0.002 5.001 0.604 0.394 0.002

48.97 32.03 0.60 n.a. 0.73 15.01 2.99 n.a. 100.3 2.237 1.724 0.023 n.a. 0.050 0.735 0.265 n.a. 5.033 0.735 0.265 n.a.

49.73 32.60 0.29 n.a. 0.09 15.56 2.98 n.a. 101.3 2.247 1.736 0.011 n.a. 0.006 0.753 0.261 n.a. 5.015 0.743 0.257 n.a.

49.47 33.10 0.33 n.a. 0.07 16.22 2.75 n.a. 101.9 2.225 1.754 0.012 n.a. 0.005 0.782 0.240 n.a. 5.018 0.765 0.235 n.a.

50.35 32.14 0.29 n.a. 0.07 13.70 2.86 n.a. 99.41 2.298 1.729 0.011 n.a. 0.005 0.670 0.253 n.a. 4.965 0.726 0.274 n.a.

49.06 31.82 0.06 0.00 0.00 15.35 2.86 0.01 99.17 2.260 1.728 0.002 0.000 0.000 0.758 0.255 0.001 5.004 0.747 0.252 0.001

50.79 28.02 1.13 0.00 2.69 14.35 3.04 0.01 100.0 2.330 1.515 0.043 0.000 0.184 0.705 0.270 0.001 5.048 0.722 0.277 0.001

Analyses from Domain 3: 1, core of undeformed plagioclase in gabbronorite intrusion; 2, neoblast in mylonitic ferrogabbro. Analyses from Domain 4: 3 and 4, porphyroclast core and rim in porphyroclastic olivine gabbro; 5 and 6, core and rim from same undeformed plagioclase in host gabbro; 7 and 8, porphyroclast rim and neoblast from same cataclastic gabbro; 9 and 10, core and rim from same undeformed plagioclase in host gabbro; 11, porphyroclast rim in mylonitic gabbro; 12, neoblast in mylonitic gabbro. n.a., not analysed. FeO, total iron calculated as FeO.

Fig. 7. Classification of amphiboles for structural Domains (a) 3 and (b) 4 according to protolith (O, ferrogabbro; o, gabbronorite; A, gabbro; n, olivine gabbro) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static) from the classification of Leake (1978). Entire spectrum of data for North Arm Mountain massif is also shown (grey field; Bedard, unpubl. data).

Table 4. Amphibole composition Analysis: Sample:

1 92702

2 92701

3 92701

4 94692

5 94692

6 94692

7 94657

8 94657

9 94655

10 94655

11 92463

12 92463

13 92466

14 92465

Si02 TiO2 A1203

43.17 1.83 10.77 12.15 0.21 13.49 11.59 2.78 0.39 n.a. n.a. 0.04 96.42 6.398 1.602 0.279 0.204 0.005 0.358 1.148 0.026 2.980 1.840 0.799 0.074 n.a. n.a. 2.000 17.71

45.07 1.33 10.49 12.57 0.25 13.09 11.50 2.00 0.17 0.00 0.12 0.06 96.65 6.604 1.396 0.415 0.146 0.007 0.468 1.072 0.031 2.860 1.806 0.569 0.032 0.000 0.029 1.971 17.41

45.58 1.21 13.65 9.97 0.23 10.49 14.92 1.47 0.12 0.01 0.09 0.03 97.76 6.602 1.398 0.932 0.132 0.003 0.000 1.208 0.028 2.265 2.315 0.413 0.022 0.002 0.022 1.976 17.32

45.53 1.69 9.29 12.16 0.39 14.36 12.55 2.10 0.14 n.a. n.a. n.a. 98.21 6.602 1.398 0.189 0.184 n.a. 0.325 1.150 0.048 3.104 1.950 0.590 0.026 n.a. n.a. 2.000 17.566

47.19 1.65 8.97 12.14 0.29 14.72 12.41 2.06 0.11 n.a. n.a. n.a. 99.54 6.716 1.284 0.221 0.177 n.a. 0.337 1.108 0.035 3.123 1.892 0.568 0.020 n.a. n.a. 2.000 17.481

48.02 0.84 9.46 10.64 0.25 16.06 12.32 1.76 0.16 n.a. n.a. n.a. 99.51 6.723 1.277 0.284 0.088 n.a. 0.613 0.632 0.030 3.352 1.848 0.478 0.029 n.a. n.a. 2.000 17.354

44.15 3.34 11.49 12.49 0.39 13.16 12.17 2.50 0.14 n.a. n.a. 0.18 100.01 6.325 1.675 0.265 0.360 0.020 0.213 1.283 0.047 2.811 1.868 0.694 0.026 n.a. n.a. 2.000 17.588

44.16 0.61 11.74 9.11 0.13 16.64 12.24 2.18 0.07 n.a. n.a. 0.06 96.94 6.306 1.694 0.282 0.066 0.007 0.913 0.175 0.016 3.542 1.873 0.604 0.013 n.a. n.a. 2.000 17.489

49.70 0.45 4.82 10.49 0.37 16.25 12.51 0.72 0.03 n.a. n.a. 0.14 95.48 7.248 0.752 0.077 0.049 0.016 0.442 0.837 0.046 3.533 1.955 0.204 0.006 n.a. n.a. 2.000 17.164

43.46 2.20 11.33 8.73 0.28 15.24 12.19 1.71 0.12 n.a. n.a. 0.18 95.44 6.358 1.642 0.311 0.242 0.021 0.498 0.570 0.035 3.323 1.911 0.485 0.022 n.a. n.a. 2.000 17.418

45.29 2.47 11.14 9.48 0.22 12.90 13.96 1.69 0.16 n.a. n.a. 0.14 97.45 6.585 1.415 0.493 0.270 0.016 0.000 1.153 0.027 2.796 2.175 0.476 0.030 n.a. n.a. 2.000 17.436

52.99

51.26 0.49 6.80 7.60 0.10 17.09 12.68 0.93 0.07 0.00 0.02 0.39 97.44 7.254 0.746 0.387 0.053 0.044 0.098 0.802 0.012 3.605 1.923 0.254 0.012 0.000 0.006 1.994 17.189

51.36 0.37 5.16 7.21 0.08 17.63 13.04 0.75 0.04 0.00 0.03 0.34 96.01 7.387 0.613 0.262 0.040 0.038 0.000 0.867 0.010 3.780 2.009 0.210 0.008 0.000 0.006 1.994 17.225

FeO MnO MgO CaO Na 2 O

K2O F Cl Cr203 Total

Si A1IV A1VI

Ti Cr Fe3+ Fe2+

Mn Mg Ca Na K F Cl OH* Total

0 4.12 8.96 0.54 17.92 12.61 0.12 0.00 n.a. n.a. 0.00 97.26 7.446 0.554 0.129 0.000 0.000 0.595 0.458 0.064 3.754 1.899 0.033 0.000 n.a. n.a. 2.000 16.931

Analyses from Domain 3: 1, brown hornblende rim of pyroxene pseudomorph from gabbronorite intrusion; 2 and 3, core of green hornblende neoblast and green hornblende vein from mylonitic ferrogabbbro. Analyses from Domain 4: 4, 5 and 6, core and rim of brown hornblende porphyroclast and rim of brown hornblende neoblast from porphyroclastic olivine gabbro; 7 and 8, brown hornblende rim of pyroxene pseudomorph and brown hornblende from vein of plagioclase and hornblende from host gabbro; 9 and 10, brown hornblende core of pyroxene pseudomorph and core of brown hornblende neoblast from cataclastic gabbro; 11 and 12, brown hornblende corona around clinopyroxene and fibrous actinolite from host gabbro; 13, core of green hornblende porphyroclast from mylonitic gabbro; 14, neoblast of green hornblende from mylonitic gabbro. Plagioclase-amphibole analysis pairs 1-1, 2-2, 4-6, 5-7, 8-10, 11-13 and 12-14 have been used for temperature calculations, n.a., not analysed. * Calculating to ideal stoechiometry.

386

E. G I G U E R E ^ r ^ I .

Fig. 8. Ti v. A1IV for amphiboles of structural Domains (a) 3 and (b) 4 according to protolith (O, ferrogabbro; A, gabbro; n, olivine gabbro; o, gabbronorite) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static). Entire spectrum of data for North Arm Mountain massif is also shown (grey field; Bedard, unpubl. data).

Fig. 9. A1IV v. (Na + K)A for amphiboles of structural Domains (a) 3 and (b) 4 according to protolith (O, ferrogabbro; A, gabbro; n, olivine gabbro; o, gabbronorite) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static). Entire spectrum of data for North Arm Mountain massif is also shown (grey field; Bedard, unpubl. data).

Fe3+/(Fe3++ A1IV) <0.04 and MnO contents <0.07. Prehnite veins in shear zones of Domain 3 are more iron rich than those in shear zones of Domain 4 (Table 6). These compositions are similar to prehnite from metagabbro of the MARK area (Gillis et al. 1993) and from gabbro of the Garrett Transform Fault (Bideau et al. 1991), and to the least Fe-rich prehnite of the Oman ophiolite (Stakes & Taylor 1992).

Epidote Epidote with or without calcite forms veins in two mylonitized samples. The epidote has low Fe content with a pistacite component between 8 and 10% (Table 6). These compositions are similar to Fe-poor epidotes of the Oman ophiolite (Stakes & Taylor 1992) and are similar to those found in gabbros of ODP Hole 735B (Vanko & Stakes

HYDROTHERMAL METAMORPHISM IN GABBROS

387

Table 5. Chlorite compositions Analysis: Sample: Position: Si02 TiO2 A12O3

FeO MnO MgO CaO Na2O

K20 Total

Si A1IV

Ti Al A1VI

Fe Mn Mg Ca Na K Total

9 92718

11 94645

2 94655

3 94656

4 94658

5 94658

6 94658

7 92460

8 92463

fc

vein 28.25 0.00 20.03 12.10 0.00 22.78 0.70 0.00 0.00 83.86 5.785 2.215 0.000 4.834 2.619 2.072 0.000 6.954 0.154 0.000 0.000 19.798

fc

cc 28.44 0.05 18.92 14.21 0.24 20.01 3.78 0.00 0.00 85.65 5.839 2.161 0.008 4.578 2.418 2.440 0.042 6.125 0.832 0.000 0.000 19.864

vein 27.63 0.02 20.66 15.92 0.30 20.99 0.34 0.16 0.00 86.02 5.639 2.361 0.003 4.970 2.609 2.717 0.052 6.386 0.074 0.063 0.000 19.905

fc

fc

28.11 0.00 20.08 15.85 0.18 21.50 0.38 0.00 0.00 86.10 5.721 2.279 0.000 4.816 2.537 2.698 0.031 6.523 0.083 0.000 0.000 19.871

cc 28.27 0.00 20.32 11.88 0.11 24.80 0.31 0.00 0.00 85.69 5.662 2.338 0.000 4.797 2.459 1.990 0.019 7.405 0.067 0.000 0.000 19.939

fc

28.49 0.00 21.00 17.37 0.32 22.29 0.40 0.09 0.00 89.96 5.592 2.408 0.000 4.858 2.450 2.851 0.053 6.522 0.084 0.034 0.000 19.996

33.21 0.00 14.15 14.61 0.39 19.67 2.42 0.16 0.00 84.61 6.835 1.165 0.000 3.432 2.267 2.514 0.068 6.035 0.534 0.064 0.000 19.481

29.03 0.00 20.49 10.59 0.07 24.57 0.16 0.10 0.21 85.23 5.797 2.203 0.000 4.821 2.617 1.768 0.013 7.313 0.035 0.039 0.055 19.840

36.05

0 15.19 15.58 0.34 21.77 0.93 0.06

0 89.91 6.924 1.076 3.439 2.363 2.503 0.055 6.233 0.191 0.022

0 19.368

Analyses from Domain 4: 1, fibrous chlorite in host olivine gabbro; 2, chlorite vein in cataclastic gabbro; 3, fibrous chlorite in porphyroclastic gabbro; 4 and 5, corona around magnesio-hornblende and vein in host gabbro; 7, corona around tremolite in host gabbro; 8, fibrous chlorite in host gabbro; 9, fibrous chlorite in host olivine gabbro. FeO, total iron calculated as FeO.

1991), of the MARK area (Gillis et al 1993) and of the Garrett Transform Fault (Bideau et al. 1991). However, these compositions are very different from epidote of the Mathematician Ridge (Stakes & Vanko 1986) and from the Fe-rich epidote of the Oman ophiolite (Stakes & Taylor 1992).

Geothermometry Lindsley (1983) provided a geothermometer for pyroxenes with Wo + En + Fs ^90% with isotherms for 1 atm, 5 kbar, lOkbar and 15 kbar with an accuracy of ±50 °C. We have used the 1 atm pressure curves for temperature calculations (Fig. 5), because this pressure is close to the 2 kbar lithostatic pressure prevailing in ophiolitic crust. For a typical NAM clinopyroxene composition, a 5 kbar pressure compared with a 1 atm pressure results in a temperature rise of 20 °C. Clinopyroxenes of Domain 3 give equilibration temperatures ranging from 550 °C to lower than 500 °C (Fig. 5a; Table 7). All clinopyroxenes are recrystallized such that the temperatures record the conditions of metamorphic re-equilibration. Primary orthopyroxene in intrusive olivine gabbro-

norite yields a temperature of 800 °C, which is slightly lower than the temperature (860-865 °C) obtained using the Holland & Blundy (1994) amphibole-plagioclase geothermometer in that sample (see below). Clinopyroxenes of Domain 4 yield temperatures ranging from 900 °C to lower than 500 °C for primary clinopyroxene in undeformed samples (Fig. 5b; Table 1). In samples with dynamic recrystallization textures, clinopyroxene porphyroclasts yield temperatures ranging from 900 to lower than 500 °C whereas clinopyroxene neoblasts yield temperatures ranging from 750 °C to lower than 500 °C (Fig. 5c). The plagioclase-amphibole geothermometers of Spear (1980) and Holland & Blundy (1994) were applied to plagioclase and amphibole in physical contact in samples devoid of quartz, except for site TA17.STN09-15 where plagioclase and amphibole are not in contact. This shear zone displays a high degree of recrystallization, and amphibole and plagioclase display small compositional variations. Equilibrium temperatures for these samples show small variation between 10 and 40 °C. The Spear (1980) geothermometer is based on the reaction

E. GIGUERE ETAL.

388 Table 6. Prehnite and epidote compositions Analysis: Sample: Position:

1 92495 vein

2 92495 c

3 92701 vein

4 92701 vein

5 94692 vein

6 94648 vein

7 94652 vein

8 94415 vein

9 94689 vein

Si02 TiO2 A1203 Fe203

42.01 0.00 23.05 1.41 0.01 0.03 24.90 0.02 0.01 91.41 3.018 0.000 0.000 1.952 1.952 0.076 0.001 0.003 1.917 0.002 0.001 6.969

45.77 0.00 25.06 0.11 0.00 0.00 23.00 0.90 0.02 94.85 3.118 0.000 0.000 2.012 2.012 0.005 0.000 0.000 1.678 0.119 0.002 6.934

41.49 0.12 22.98 0.39 0.05 0.09 26.40 0.03 0.00 91.55 2.988 0.012 0.006 1.951 1.939 0.021 0.003 0.010 2.037 0.004 0.000 7.021

43.01 0.05 23.69 0.48 0.02 0.21 26.34 0.00 0.00 93.80 3.012 0.000 0.002 1.955 1.955 0.025 0.001 0.022 1.977 0.000 0.000 6.995

42.84 n.a. 24.45 0.51 n.a. 0.13 28.53 n.a. n.a. 96.46 2.941 0.059 n.a. 1.978 1.919 0.026 n.a. 0.013 2.098 n.a. n.a. 7.057

42.15 n.a. 23.28 0.31 0.07 n.a. 28.26 n.a. n.a. 94.07 2.969 0.031 n.a. 1.933 1.902 0.016 0.004 n.a. 2.133 n.a. n.a. 7.056

42.30 n.a. 26.03 0.15 0.03 n.a. 26.38 n.a. n.a. 94.89 2.925 0.075 n.a. 2.121 2.046 0.008 0.002 n.a. 1.954 n.a. n.a. 7.010

43.06 n.a. 24.39 3.20 0.06 n.a. 24.80 n.a. n.a. 95.51 3.375 0.000 n.a. 2.253 2.253 0.189 0.004 n.a. 2.083 n.a. n.a. 7.904

37.70 0.02 28.30 4.87 0.12 0.50 24.52 0.01 0.00 96.04 2.981 0.000 0.001 2.637 2.637 0.290 0.008 0.059 2.077 0.002 0.000 8.055

MnO MgO CaO Na2O

K2O Total

Si A1IV

Ti Al A1VI Fe3 +

Mn Mg Ca Na K Total

Analyses from Domain 3: 1 and 2, prehnite vein and clast from mylonitic ferrogabbro; 3 and 4, prehnite vein from mylonitic ferrogabbro. Analyses from Domain 4: 5, prehnite vein from porphyroclastic olivine gabbro; 6, prehnite vein from host gabbro; 7, prehnite vein from mylonitic gabbro; 8, epidote vein from porphyroclastic olivine gabbro; 9, epidote vein from porphyroclastic olivine gabbro. n.a., not analysed. Fe2O3, total iron calculated as Fe2O3.

and >725 °C for these amphiboles. Temperatures for magnesio-hornblende and actinolitic hornblende range from 680 to 800 °C and from 510 to Equilibrium temperatures are determined from 650 °C, respectively. The lowest temperatures of a diagram of ln(XAn/XAbinplagioclase the Holland & Blundy (1994) geothermometer v. ln(Ca,M4/Na,M4) in amphibole calibrated for overlap with the highest temperatures of the Spear (1980) geothermometer. Because the Holland & temperatures between 300 and 725 °C. The Holland & Blundy (1994) geomermometer Blundy (1994) geothermometer is calibrated for yields temperatures for plagioclase-amphibole high-temperature amphiboles, it is more relevant assemblages with or without quartz and is valid for pargasite, hastingsite, tschermakite and magnefor temperatures between 500 and 900 °C. The sio-hornblende, whereas the Spear (1980) plagioclase composition must lie in the range XAn geothermometer is more suitable for actinolitic >0.1 and <0.9, amphiboles must have ^M4Na hornblende, actinolite and tremolite equilibrium >0.03, A1VI <1.8p.f.u., and Si in the range 6.0- temperatures. In Domain 3, the equilibrium temperatures 7.7p.f.u. (Holland & Blundy 1994). The reaction obtained for undeformed gabbro, for pyroxene as used is well as plagioclase-amphibole assemblages, are 800 °C and between 860 and 865 °C, respectively. The Holland & Blundy (1994) geothermometer In mylonitic samples, a greater variation is shown temperatures are higher than those determined between the pyroxene geothermometer and plagiousing the Spear (1980) geothermometer for all clase-amphibole geothermometer temperatures. amphibole types (Table 7). Pargasite, hastingsite Temperatures between 750 and 800 °C are oband tschermakite yield temperatures ranging from tained for hastingsitic hornblende, pargasitic horn745 to 880 °C with the Holland & Blundy (1994) blende and magnesio-hornblende, which are geothermometer, whereas the Spear (1980) within the temperature range for granulite facies geothermometer yields temperatures between 500 of regional metamorphism. However, pyroxenes

HYDROTHERMAL METAMORPHISM IN GABBROS

389

Table 7. Comparison of temperatures yielded by mineralogical and oxygene isotope geothermometers Amphibole

Sample Domain 3 TA17.STN09-15 92495 92496 92497 92498 92499 92700 92701 92702 Domain 4 TA2.STN06-07 94689 94691 94692 TA01.02 94603 94607 TB09.16 92460 92463 92465 92466

Lindsley (1983)

mylonite mylonite mylonite mylonite mylonite mylonite mylonite undeformed

brhbl brhbl brhbl

mylonite undeformed mylonite

grhbl brhbl

650-875 <500-850 <500-750

mylonite undeformed

grhbl brhbl

500-650

undeformed undeformed mylonite mylonite

927 16b 92718 MRL.06 94417 94415 94414 MRL.24 94658 94657

undeformed undeformed

94657b 94656 94655 CBR.05-06 94645 94651 94652

undeformed undeformed mylonite

Holland & Zheng (1993 a) Blundy(1994)

650 500-525 530-650

749-762 778-800 775-799

700 540 650->750 725

778 758-795 757-785 859-864

500-510

768-781

330

510-650

689-797

516

520 660

743-785 853-865

857

<500-550 brhbl brhbl brhbl brhbl

<500 800

gr hbl (po) gr hbl (po) gr hbl (ne)

500-550 500-550 500

1620

475 520 520-650

684-686 677-680 690-697

280 152 179

500-650

undeformed mylonite mylonite

grhbl grhbl

undeformed undeformed

brhbl brhbl actinolite

undeformed mylonite mylonite

<500

Spear (1980)

650-700 550

520-550 510

754-791

1032 650 500-900

grhbl brhbl

600-900

grhbl

500-650

grhbl

have re-equilibration temperatures between <500 °C and 550 °C. During dynamic recrystallization, pyroxene continues to re-equilibrate to temperatures lower than amphibole stability temperatures. In Domain 4, the first amphiboles to form are brown amphiboles (pargasite, hastingsitic hornblende, pargasitic hornblende and tschermakitic hornblende) that yield temperatures ranging from 855 to 880 °C in undeformed samples and from 745 to 800 °C in mylonitic samples. The second genera-

725 650 550

855-881 860-874

909

525 500-520

802 745-800

330 551 958

734-773 500-520

743-757

129 271

tion of amphiboles is green amphibole (magnesiohornblende) that yields temperatures ranging from 735 to 790 °C in undeformed samples and from 680 to 785 °C in mylonitic samples. Temperatures yielded by porphyroclast or neoblast pairs show the same range (680-765 °C and 690-740 °C, respectively). The third amphibole generation (actinolitic hornblende, actinolite and tremolite) crystallized at temperatures ranging from 520 to 550 °C in undeformed samples and from 500 to 525 °C in mylonitic samples. Thus, a gradual fall

390

E. GIGUERE ETAL.

in temperature, from granulite through amphibolite to greenschist facies, is observed with time.

Conditions of metamorphism We believe that metamorphism occurred under relatively low pressure for the shear zones studied, independent of the structural domain. The lithostatic pressure was <2 kbar if we suppose a crust of 5 km thickness and a water depth of 5 km. This low pressure is in agreement with the very low Na content of the clinopyroxenes (Jdo_2)In Domain 3, pyroxene and plagioclase-amphibole geothermometers and the A1IV content of amphiboles indicate high-temperature regimes (865-750 °C). Si activity was low whereas Na activity was high. In Domain 4, substitution relationships among amphiboles indicate an increase in fluid Si activity and a fall in temperature through time coupled with variable amphibole compositions. Brown amphiboles (pargasite, hastingsitic hornblende, pargasitic hornblende and tschermakitic hornblende) have temperatures between 880 and 745 °C, magnesio-hornblende, between 790 and 680 °C, and actinolitic hornblende, actinolite and tremolite, between 550 and 500 °C. Decreases in amphibole Na and A1VI contents are compatible with this temperature drop.

Oxygen isotope geochemistry Oxygen isotope results Structural Domain 3. Most of the whole-rock <318O values range from 3.9 to 6.6%o (Fig. 10; Table 1). These compositions are lower than or overlap with the average O-isotope composition of MORE and back-arc basin basalts (BABB) (5.66.3; Taylor et al 1994). Similar (318O values are found for shear zones and their host rock, ranging from 3.9 to 6.4%o and from 4.3 to 6.6%o, respectively. A late gabbro dyke has (518O values of 9.3%o. Plagioclase in this dyke is highly rodingitized, which accounts for the heavier whole-rock oxygen isotope composition. (518O values for plagioclase range from 3.2 to 7.4%o (Fig. 11; Table 1). All these plagioclases are secondary and come from mylonites. The lower (518O values come from plagioclase with high An contents (Fig. 11 a). The plagioclase with the highest (318O is partly rodingitized and prehnite veins cut the rock. High (518O plagioclase values are probably produced by low-temperature clays and prehnite with heavy O-isotope compositions. The other samples have slightly rodingitized plagioclase porphyroclasts with fresh plagioclase neoblasts. Samples with only fresh and recrystallized

Fig. 10. Histograms of oxygen isotope compositions for whole rock for structural Domains 3 and 4.

plagioclase have lower (318O. An amphibole with a <518O value of 5.4%o and a plagioclase with a <518O value of 6.1%o in the same sample have the typical isotope composition of MORB. Structural Domain 4. The same range of values is found for plagioclase, clinopyroxene and whole rock in host rocks and mylonites (pi 3.2-8.9%o and 4.0-8.0%o, cpx 4.1-4.5%o and 4.1%o, whole rock (wr) 3.1-6.896o and 2.2-6.5%o, respectively). A porphyroblastic clinopyroxene found in a trondhjemite vein has a (518O value of — 1.0%o. Analysed amphiboles from the host rock have (518O values similar to amphibole in mylonites (4.3%o and 0.1-5.596o). No correlation is found between An content and plagioclase 618O values (Fig. 11 a). Whole-rock, plagioclase, amphibole and clinopyroxene (518O values can be subdivided into three groups: (1) typical of MORB d18O; (2)

HYDROTHERMAL METAMORPHISM IN GABBROS

391

Fig. 11. (a) An v. d18O values for plagioclases, and (b) A1IV v. <518O values for amphiboles, for structural Domains 3 and 4 according to protolith (O, ferrogabbro; A, gabbro; n, olivine gabbro; o, gabbronorite) and process of recrystallization (open symbols indicate dynamic; filled symbols indicate static).

higher than MORE compositions; (3) depleted in 18 O compared with MORB (Fig. 10; Table 1). The first group comprises gabbro and olivine gabbro that have (518O values ranging from 5.5 to 6.2%o, typical of MORB. Plagioclase has (518O values ranging from 5.5 to 6.1%o, amphiboles have (318O values ranging from 5.1 to 5.8%o and clinopyroxene has a <518O of 5.1%o. MORB (518O values for amphibole are found in amphibole with the higher A1IV content (Fig. 1 Ib). The second group comprises partly rodingitized and some fresh plagioclase in gabbro, olivine gabbro and trondhjemite and crosscutting mylonite zones that have (518O values ranging from 6.5 to 8.9%o. The more 18O-enriched plagioclase is replaced partly by clays and prehnite. All these samples are cut by veins of prehnite, carbonate epidote, quartz or chlorite. Thus, their high values of 18O are related to the presence of these minerals. The third group comprises plagioclase with (318O values ranging from 3.2 to 5.2%o, whereas amphibole has <518O values ranging from 0.0 to 4.5%o. The lowest values of amphibole are found in the inner parts of shear zones (Fig. 12). Clinopyroxene has values slightly lower than MORB, with <518O ranging from 4.1 to 5.1%o, except for the porphyroblastic clinopyroxene with <518O value of -1.0%o. Whole-rock values of (318O for this group are between 2.2 and 4.8%o. A typical example of the low 18O group 3 is site TB09.16 (Figs. 3 and 12). Plagioclases have d 18 O values between 3.2 and 4.9%o in the host rock whereas plagioclases in the mylonite range from 4.0 to 4.4%o. Amphibole (318O values are

between 1.7 and 2.0%o in the host rock but range from 0.1 to 1.0%o in the mylonite samples. Granoblastic trondhjemite veins cutting the mylonite have <518O values for plagioclase (5.0%o) and amphibole (1.0%o) similar to those in the mylonite whereas another vein has 18O-enriched plagioclase with (318O values of 8.9%o for plagioclase and 4.3%o for amphibole.

Isotope equilibrium temperatures Isotope equilibrium temperatures were calculated using theoretical geothermometers of Zheng (1993a, 1993b) because a consistent set of experimental fractionation factors is not available for the minerals of interest. When the isotope equilibrium temperatures are lower than mineral equilibrium temperatures, it is likely that the O-isotope compositions of the minerals were re-equilibrated at lower temperature or are in isotopic disequilibrium. In Domain 3, one amphibole-plagioclase pair with magmatic compositions (am 6.1%o, pi 5.4%o) yields a high isotope temperature of 1620°C, above the solidus for this rock (Table 7). This isotope equilibrium temperature is higher than temperatures obtained by mineral equilibrium geothermometers that range from 760 to 785 °C for this sample and from 750 to 865 °C for all the samples taken from this shear zone. In Domain 4, isotope equilibrium temperatures yielded by plagioclase-clinopyroxene pairs for host rock and shear zones are similar albeit higher than mineral equilibrium temperatures (Table 7). A temperature of 860 °C is obtained from a

392

E. GIGUERE ETAL.

amphibole mineral compositions or amphibole plagioclase isotope disequilibrium. Amphiboleplagioclase disequilibrium is expected for temperatures ranging from 135 °C to 330 °C. Low 18O of amphiboles requires circulation of large amounts of water.

Oxygen isotope composition of the hydrothermal fluids

Fig. 12. (a) (518O values of plagioclase (+), amphibole (O) and whole rock (o) v. shear zone distance at site TB09.16; (b) fluid/rock ratios (F/R) for an initial composition of 0%o (A) and 2%o (•).

plagioclase-clinopyroxene pair that compares well with mineral equilibrium temperatures ranging from 855 to 865 °C in that sample. In another shear zone, a plagioclase-clinopyroxene isotope equilibrium temperature of 910 °C compares well with a mineral equilibrium temperature of 860875 °C. This undeformed rock is cut by a trondhjemite dyke for which a formation temperature of 800-1000 °C has been calculated (El Bilali 1995). Domain 4 mylonitic samples show isotope equilibrium temperatures on plagioclase-amphibole ranging from 135 to 550°C compared with mineral equilibrium temperature ranging from 500 to 800 °C. The lowest isotope equilibrium temperatures are found in the middle of the shear zones. Lower isotope equilibrium temperatures than mineral temperatures can represent either continued O-isotope exchanges at lower temperatures without modification of plagioclase and

(518O values of the hydrothermal fluids can be calculated from the isotope composition of coexisting plagioclase and amphibole or clinopyroxene in conjunction with the isotope equilibrium temperature. When these minerals are not in isotopic equilibrium, the mineral temperature of plagioclase-amphibole geothermometers is regarded as the isotopic closure temperature of the amphibole system and it is used to calculate the <518O value of the hydrothermal fluids. Three types of hydrothermal fluids can be identified from O-isotope systematics in shear zones of structural Domains 3 and 4. (1) An evolved hydrous magmatic fluid with an O-isotope composition between 6.9 and 7.6%o is detected in samples cut by trondhjemite and brown amphibole veins. Trondhjemite <518O values from ODP samples and ophiolites range from 7 to 8%o. These trondhjemites are produced by late melting and assimilation of altered rocks at high levels in magma chambers (Stakes et al. 1984) or by magmatic differentiation (Elthon et al. 1984; Kempton et al. 1991). This magmatic fluid composition is also found in a sample of Domain 3 that is not cut by trondhjemite veins, but this sample records circulation of aqueous fluid because magnesio-hornblende crystallizes at relatively high temperature at the expense of clinopyroxene. (2) High-temperature evolved seawater with (318O values ranging from 1.9 to 5.4%o. The <518O value of 2.0%o is within the error interval for the Ordovician seawater end-member at 0 ± 2%o (Muehlenbachs 1986, 1998a, 1998b) indicative of isotope exchange between seawater-derived fluid and rock. The high (518O indicates exchange with rock under low fluid/rock ratio. (3) Low-temperature seawater with compositions ranging from 1.1 to 2.2%o is also compatible with Ordovician seawater. In Domain 3, samples with polygonal recrystallization textures reveal isotopic re-equilibration reactions. Mineral geothermometric temperatures and plagioclase (318O values yield fluid (518O values of 1.2 and 4.2%o (Table 8). The higher value indicates that the hydrothermal fluid was evolved seawater. The lower value (1.2%o) is

HYDROTHERMAL METAMORPHISM IN GABBROS

393

Table 8. Calculated fluid/rock ratios and dl*O fluid compositions Fluid/rock ratio for 2 fluid compositions Amphibole Closed system Sample Domain 3 TA17.STN09-15 92495 92498 92701 Domain 4 TA2.STN06-07 94689 94691 94692 TA01.02 94603 94607 94608 TB09.16 92460 92463 92464 92465 92466 92468 927 16b 92718 MRL.06 94414 94417 94419 MRL.24 94655 94656 94657 94657b 94658 CBR.05-06 94651 94652

0%o

2%o

Plagioclase

Open system 0%o

2%o

Closed system

Al8/^» -cu^J

Open system

composition

0%o

HT

0%o

2%o

2%o

0.8 0.1

1.9 0.0

4.2

0.0 0.2

0.0 0.3

5.9

LT

1.2 0.0

0.0

0.3

0.5

7.6

1.1

7.3 0.1 0.2 0.7

1.1

0.6 0.9 1.3 0.2 0.9

1.0 1.6 3.4 0.2 1.6

0.2

0.5 0.6

0.8 1.1

3.6

0.4 0.3

0.7 0.5

2.8; 4.6 1.9; 5.0

0.2

0.3

0.3 0.0 0.2

0.6 0.1 0.3

5.6; 4.4 5.9 5.2

0.2

0.2

6.0 6.5 6.9

2.0 0.2

0.4

0.1 0.3

0.3

0.5

0.3

2.2 0.2 0.2 0.5

0.4 0.8

0.3 0.1

0.1

5.6 4.2

2.1

HT, high temperature; LT, low temperature.

calculated from a plagioclase with (318O of 7.4%o that indicates infiltration of low-temperature seawater. In Domain 4, fluid <518O values range from 1.9 to 7.6%o (Table 8). The heavier composition is found in samples cut by trondhjemite and brown amphibole veins, suggesting that magmatic fluid equilibrated with the host rock. Generally, the fluid has a heavier O-isotope composition in host rocks than in mylonites, reflecting lower fluid/rock ratio (Fig. 12). Disequilibrium amphibole and plagioclase pairs indicate that the two minerals were formed at distinct temperatures, that more

than one fluid has circulated in the shear zone and the two minerals did not form at the same time, or that the residence time of the fluid was too short to reach equilibrium with the rock. At site TB09.16 (Fig. 3), the fluid (318O value is calculated using mineral geothermometric temperatures. For a temperature of 700 °C, amphibole indicates a fluid with values 1.9 and 2.8%o. The <518O fluid is interpreted as Ordovician seawater, which has a value of 0 ± 2%o (Muehlenbachs 1986, 1998a, 1998b). Plagioclase at 4.4%o can be produced by reaction of the fluid (1.9-2.8%o) at lower temperatures (between 290 and 310°C);

394

E. G I G U E R E ^ r ^ L .

however, neoblast and porphyroclast plagioclases have a composition of An75_76. Another possibility is the circulation of a second fluid with a (518O between 4.6 and 5.0%o calculated from the plagioclase value at 700 °C. An initial circulating low d18O seawater signature is still identifiable. Because the reaction rate is faster in plagioclase than in amphibole (Cole & Ohmoto 1986), evolved seawater reacted with plagioclase and overprinted the earlier signature of the fluid at 1.9-2.8%o. Other sites record only one fluid, seawater or evolved seawater. Samples of site CBR.05-06 are in equilibrium with a fluid composition of 2%o. Fresh plagioclase with enriched 18O indicates circulation of seawater at low temperatures with (518O between 1.1 and 2.2%o at 200 °C.

Fluid/rock ratios Fluid/rock ratios (F/R) can be calculated from plagioclase and amphibole isotope compositions using the equations of Taylor (1977) for open and closed systems. The dplagioclase-damphibole, clinopyroxene (Fig. 13) diagram allows us to determine, for each structural domain, whether samples evolved under open or closed systems (Gregory & Criss 1986), which then dictates the relevant equation to compute fluid/rock ratio. In Domain 3, a plagioclase-amphibole pair has a MORB-like igneous composition. In Domain 4, a closed system is found in two shear zones (MRL.24 and TA02. 06-07). An example shows a plagioclase-amphibole pair (94656; 6.0 and 4.1%o, respectively) which re-equilibrated at lower temperature in a closed system from a plagioclase-clinopyroxene pair of the host rock (94657; 6.1 and 5.1%o, respectively) (Fig. 12). A sample of shear zone TA02.06-07 (94689; 5.9 and 3.8%o) shows the same trend with closed-system evolution from an initial composition similar to MORE (Fig. 13). An open system is found where fluid circulation re-equilibrated plagioclase O-isotope compositions to lower temperatures without reequilibration of amphibole O-isotope compositions. The rocks record the circulation of this fluid resulting in rodingitization of plagioclase and the precipitation of prehnite, chlorite and calcite in veins. However, amphibole found together with high <518O plagioclase probably records the circulation of a hydrous late magmatic fluid with an Oisotope composition similar to a trondhjemitic vein because the amphibole does not re-equilibrate at temperatures lower than 350 °C (Ito & Clayton 1983). An open system is also found in shear zones, where <518O amphibole values near 0%o are in equilibrium with high-temperature evolved seawater. The lower values are observed in the centre of the mylonitic shear zones (Fig. 12) where

Fig. 13. <518Oplagioclase v. (518Oclinopyroxene (filled symbol) or (518Oamphibole (open symbol) for samples of structural Domains 3 (squares) and 4 (circles). MORE (-AT) value is (5.5; 6.0). The arrow shows the trace of equilibrium values for plagioclase and amphibole with a fluid at 2%o. Trend 1 represents an open system at low temperature and trend 2 represents an open system at high temperature. A line joins sample pairs evolving in a closed system. Field values from ODP Hole 73 5B, Oman and Indian Ocean are shown for comparison (from Stakes et al. 1991). Isotherms are calculated for plagioclase of An60, clinopyroxene of Wo40 En50 FslO, hornblende and actinolite.

amphiboles coexist with plagioclase having (318O values between 4.0 and 4.4%o. These pairs are similar to amphibole-plagioclase pairs observed in Oman (Stakes & Taylor 1992). This is the clearest evidence that hot seawater was channelled through by the high-temperature shear zones. Temperatures of amphibole formation based on mineral compositions are used for fluid/rock ratio calculation derived from amphibole isotope composition (Table 8). Ito & Clayton (1983) used temperatures between 350 °C and 500 °C for fluid/ rock ratio calculation because these are regarded as the lowest temperatures of equilibration for amphibole. We used a temperature of 500 °C that yields similar fluid/rock ratio to mineral equilibrium temperatures. Fluid/rock ratios derived from plagioclase isotope compositions are computed at 500 °C, regarded as the lowest temperature of equilibration as suggested by Lecuyer & Reynard (1996). This temperature is used only for plagioclase without 18O enrichment or rodingitization

HYDROTHERMAL METAMORPHISM IN GABBROS resulting from low-temperature seawater circulation. Initial fluid isotope compositions of 0 and 2%o are used for samples that equilibrated with seawater and evolved seawater (Ito & Clayton 1983; Lecuyer & Reynard 1996). In Domain 3, high-temperature seawater circulation yields fluid/rock ratios of 0.8 and 1.9 for a fluid initial composition of 0 and 2%o, respectively. In Domain 4, high-temperature seawater circulation yields a fluid/rock ratio lower than 3.4. Fluid/rock ratios increase from the host rock to the centre of a shear zone (Fig. 12). The higher fluid/rock ratios are found where amphiboles have lower (518O. These two observations are consistent with the penetration of seawater into the shear zones.

Discussion and model for fluid circulation Structural Domain 3 Oxygen isotope compositions for structural Domain 3, except for plagioclase, which shows enrichment by low-temperature metasomatism, indicate circulation of high-temperature evolved seawater with a component of magmatic fluid. This fluid has a (318O value of 4.2%o. Active listric shear zones channelled seawater to deeper levels in the oceanic crust under low fluid/rock ratios less than 1.9. Recrystallization of magmatic clinopyroxene under low fluid/rock ratios produced neoblasts with high Wo. These low fluid/rock ratios and the high temperatures of isotope exchange enrich seawater in 18O by interaction with the rock (Fig. 14). Magnesio-hornblende and magnesio-hastingsite yield mineral temperatures ranging from 750 to 865 °C. In this domain, a fluid with low Si activity and high Na activity is inferred from amphibole composition.

Structural Domain 4 A wide range of amphibole compositions coupled with small variations in clinopyroxene compositions is found in this domain for A1IV, Ti, A1VI, Na and Mg-number. Na and A1VI decrease in amphibole with falling temperature (Laird & Albee 1981; Helz 1982; Deer et al 1992). The amphibole composition in mylonitic samples is dependent on the ambient isotherm, such that actinolite is found near the transition between the sheeted dykes and the gabbros, and magnesiohornblende deeper in the gabbroic unit (Fig. 14b). A succession in amphibole generation is also found, with crystallization of brown amphibole (pargasite, hastingsitic hornblende, pargasitic hornblende and tschermakitic hornblende) at temperatures ranging from 880 to 745 °C, magnesio-

395

hornblende from 790 to 680 °C, and actinolitic hornblende, actinolite and tremolite from 550 to 500 °C. The same temperature stages are found in oceanic gabbros from slow-spreading settings such as the Mid-Atlantic Ridge, the Mid-Cayman Rise and the Southwest Indian Ridge (Mevel 1987; Mevel & Cannat 1991; Stakes et al. 1991; Gaggero & Cortesogno 1997). Isotope compositions also show a wide range, indicating intense seawater circulation. Whole-rock (518O values range from 2.2 to 7.7%o, which are lower than $18O values obtained by Muehlenbachs (1998a) for Blow Me Down and Betts Cove gabbros (5.17.3%o) and the (518O values for the BOIC (4.510.9%o; Gregory 1980). The lower (518O values probably reflect our sampling strategy, which was focused on shear zones. These rocks experienced higher seawater-derived fluid percolation. Amphibole and plagioclase values can be separated into three groups: highly depleted in 18O (0.0-2.5%o and 3.2-3.8%o, respectively), slightly depleted in 18 O (3.3-5.8%o for amphibole) and enriched in 18 O (6.5-8.9%o for plagioclase). Amphibole values in samples depleted in 18O are between 0.0 and 2.5%o whereas plagioclase values are between 3.2 and 3.8%o. Low 18O amphibole oxygen isotope compositions are found in Oman (Stakes & Taylor 1992), in the Troodos complex (Friedrichsen & Rommel 1985; Vibetti et al 1989) and at the Mid-Cayman Rise (Ito & Clayton 1983). Stakes & Taylor (1992) have found depleted 18O composition in highly deformed and fractured crustal zones. In NAM, amphibole-plagioclase pairs yield relatively low isotope temperatures ranging from 150 to 475 °C, indicating disequilibrium. Interaction between rock and evolved seawater probably occurred at temperatures between 600 and 700 °C. These temperatures are in agreement with the presence of actinolitic hornblende. These amphibole (518O values can be produced by exchange with evolved seawater with <518O values ranging from 1.9 to 2.8%o. Numerous veins and ductile-brittle structures at the time of this event facilitated downward penetration of seawater. Thus, undeformed rocks cut by numerous veins have fluid/rock ratios ranging from 1.0 to 1.6 compared with non-fractured host rocks with fluid/rock ratios lower than 0.3. Amphibole in trondhjemite showing low (318O values similar to adjacent amphibole in mylonite reveals that circulation of evolved seawater occurred after trondhjemite injection. Amphibole O-isotope compositions in samples slightly depleted in 18O are between 3.3 and 5.8%o. These results can be explained by two hypotheses. One explanation involves circulation of high-temperature seawater under lower fluid/ rock ratios than in the highly depleted samples,

396

E. GIGUERE ETAL.

Fig. 14. Evolution of hydrothermal circulation, (a) In structural Domain 3, seawater infiltration occurred along listric shear zones under low fluid/rock ratio, (b) In structural Domain 4, active shear zones and dyke margins acted as natural conduits to allow seawater penetration at depth. Fluid/rock ratios increase and temperatures fall from the base to the top of unit 3. In several shear zones of Domain 4, magmatic water derived from trondhjemite veins controlled (518O rock composition.

promoting amphibole and plagioclase (518O increase (Schiffman & Smith 1988). This hypothesis is consistent with lower fluid/rock ratios (<1.1) for samples slightly depleted in 18O compared with fluid/rock ratios between 0.6 and 3.4 for samples more highly depleted in 18O. A second explanation involves fluid exchange between gabbro and trondhjemite. This magmatic fluid with

(518O values ranging from 6.9 to 7.6%o derived from trondhjemite veins would deplete plagioclase, amphibole or clinopyroxene slightly in 18O in surrounding rocks. For example, a whole rock (94657) with plagioclase (6.1%o) and clinopyroxene (5.1%o) is cut by a trondhjemite vein with amphibole (5.1%o) and plagioclase (8.3%o). Isotope and mineral temperatures for the host rock

HYDROTHERMAL METAMORPHISM IN GABBROS are 910 °C and 860-875 °C, respectively. These temperatures are similar to magmatic temperatures of NAM trondhjemite ranging between 800 and 1000°C (El Bilali 1995), which supports the hypothesis that magmatic fluids from the veins exchanged with the host rock plagioclase and clinopyroxene. Trondhjemite veins are found throughout the structural Domain 4 (Fig. 14b). In high (518O samples, plagioclase <518O values range from 6.5 to 8.9%o. In some shear zones, plagioclase isotope compositions show 18O enrichment toward the centre of the shear zone because seawater circulation is favoured by shearing. 18 O-enriched samples have prehnite, chlorite or carbonate veins, indicating circulation of lowtemperature seawater, and contain rodingitized and albitized plagioclase. At the time of lowtemperature seawater circulation, fluid/rock ratios are estimated to range from 1.0 to 2.2. Lydon & Lavigne (1990) obtained a mean isotope composition of 7.5 ± 0.2%o for plagioclase gabbro in the lower cumulate sequence near a mineralized zone in NAM in the SW of the area covered by this study (Yao 1995). They explained these compositions by seawater circulation in the mantle or by ascent of a distilled volatile fluid from melting of hydrous oceanic crust or sediments above a subduction zone. Late seawater circulation at low temperature is our preferred interpretation to explain the high <518O plagioclase values. Mylonite TA02.06-07 has an 18O abundance similar to that of a related cross-cutting dyke, which reveals evolved seawater circulation after mylonite formation and dyke injection (Fig. 4a). Thus, dyke margins might have acted as natural conduits to allow seawater penetration to depth (Fig. 14b; Harper et al 1988; Stakes & Taylor 1992). Textural relationships show that some metamorphic diopside porphyroblasts formed late at the expense of low-temperature assemblages, which developed during reaction with seawater. These clinopyroxenes with a (518O value of — 1.0%o are found in trondhjemite veins. The temperature rise caused by late intrusion induced dehydration reaction and clinopyroxene crystallization after the formation of the low-temperature assemblage. Evolution of shear zones in NAM supports the Mevel & Cannat (1991) model for shearingcontrolled fluid penetration. Thus, rocks in Domain 3 listric shear zones are characterized by polygonal textures with plagioclase, pyroxene and amphibole neoblasts. The shearing took place under low fluid/rock ratios (0.0-1.9) as indicated by (518O values and at high temperatures (750865 °C). Domain 4 mylonites show changes in amphibole composition from pargasite, tscherma-

397

kite and hastingsite to magnesiohornblende and then to actinolite, as temperatures fell from 880 to 500 °C and fluid/rock ratios increased from 0.4 to 3.4. (518O values are also lower in later amphibole, indicating higher fluid/rock ratios.

Conclusions In this study, structural and relative time control on the studied shear zones allows a detailed interpretation of fluid migration through time. The origin of the crustal section of NAM ophiolite is compatible with continuous oblique subduction of the oceanic crust, generating an island arc near the North American continental margin, and with continuous clockwise reorientation of the oblique subduction-related forces (Berclaz et al. 1998). A first dextral transtensional pull-apart episode generated gabbroic unit 3, sheeted dykes and lavas. Development of shear zones in structural Domains 3 and 4 is related to this episode, where Domain 3 mylonites represent dip-slip listric faulting and Domain 4 mylonites represents superimposed strike-slip faulting slightly oblique to the palaeoridge (Berclaz et al. 1998). Distinct textures are found in shear zones according to the structural domain. Polygonal textures are found in earlier mylonites and, with falling temperature, textures become porphyroclastic or mylonitic, then cataclastic. The first amphibole to form is enriched in Ti, Na and A1IV. Si content increases and Ti, Na and A1IV contents decrease with time as temperature falls. Early listric shear zones show only brown amphibole with little compositional variation. Later shear zones have amphibole with large compositional variations for A1IV, A1VI, Ti and Na contents and Mg-number. Later shear zones have several amphibole types from hastingsitic, pargasitic and tschermakitic hornblende in a first stage to magnesio-hornblende in a second stage and to actinolitic hornblende, actinolite and tremolite in a third stage. Temperatures of amphibole formation fell from 880 to 745 °C in the first stage, from 790 to 680 °C in the second stage, and from 550 to 500 °C in the final stage. Early amphiboles have near igneous oxygen isotope compositions including MORB or BABB compositions. However, in some early shear zones, evolved seawater circulation is recorded in some plagioclase ((518O 3.2%o). During this first listric faulting, active shear zones serve as conduits that allow fluid circulation deep into the gabbroic crust (Fig. 14). This hypothesis is supported by lighter oxygen isotope compositions for amphibole and by higher fluid/rock ratios in the inner part of the shear zones. Amphiboles with very low (518O

398

E. GIGUERE ETAL.

values (0-2.5%o) indicate that a significant volume of seawater has circulated through the crustal system. These depleted compositions are found in a shear zone where mylonitization and fracturing are important. Fracturing indicated by amphibole veins and by trondhjemite injections enhances seawater access to these zones. Hydrothermal events alternate in time with magmatic events such as basaltic and trondhjemite intrusions. Intrusions heat the system and promote convective circulation of hydrothermal fluids (Stakes & Taylor 1992). Gabbro cut by trondhjemite veins has oxygen isotope compositions near MORE composition and the gabbro O-isotope composition is probably influenced by these veins. During later brittle deformation, low-temperature fluid circulation in host rocks and shear zones is shown by prehnite, chlorite, carbonate and quartz veins. Plagioclase partly replaced by clay with heavy oxygen isotope composition also supports low-temperature hydrothermal fluid circulation. Clinopyroxene porphyroblasts with low (318O crystallized after clay alteration and after formation of prehnite, chlorite and quartz veins. Formation of metamorphic clinopyroxene requires higher-temperature events after the low-temperature alteration. That heat was probably supplied by a late intrusion. Thus, the low-temperature alteration is restricted to the intra-oceanic history. Evolution of mineralogical compositions, oxygen isotope compositions, fluid/rock ratios and mineralogical temperatures in NAM supports the Mevel & Cannat (1991) model for fluid penetration at slow-spreading ridges. Funding was provided through grants by the Natural Sciences and Research Council of Canada (NSERC) to R.H. and G.B., and by the Fonds pour la Formation de Chercheurs et FAide a la Recherche Equipe (FCAR) to R.H. Logistic facilities were provided by the Geological Survey of Canada. We thank D. Vanko, D. Stakes and C. Mevel for constructive and critical reviews of the manuscript.

References ALT, J.C., MUEHLENBACHS, K. & HONNOREZ, J. 1986. An oxygen isotopic profile through upper kilometer of the oceanic crust, DSDP Hole 504B. Earth and Planetary Science Letters, 80, 217-229. ALT, J.C., ANDERSON, T.F. & BONNELL, L. 1989. The geochemistry of sulfur in a 1.3 km section of hydrothermally altered oceanic crust, DSDP Hole 504B. Geochimica et Cosmochimica Acta, 53, 1011-1023. ALT, J.C., ZUEGLER, E. & ERZINGER, J.A. 1995. Mineralogy and stable isotopic compositions of the hydrothermally altered lower sheeted dyke complex, Hole 504B, Leg 140. In: BECKER, K., DICK, H.J.B. & STOCKING, L.B. (eds) Proceedings of the Ocean

Drilling Program, Scientific Results, 137/140. Ocean Drilling Program, College Station, TX, 155-166. BEDARD, J.H. 1991. Cumulate recycling and crustal evolution in the Bay of Islands ophiolite. Journal of Geology, 99, 225-249. BEDARD, J.H. & CONSTANTIN, M. 1991. Syn- and postkinematic intrusions of troctolite, gabbro, and peridotite into layered gabbroic cumulates in the Bay of Islands ophiolite: genesis of anorthosite by partial assimilation, and troctolite by hybridization. Current Research, part D, Geological Survey of Canada, Paper 91-1D, 79-88. BEDARD, J.H. & HEBERT, R. 1996. The lower crust of the Bay of Islands Ophiolite, Canada: petrology, mineralogy, and the importance of syntexis in magmatic differentiation in ophiolites and at ocean ridges. Journal of Geophysical Research, 101, 25105-25124. BEDARD, J.H. & HEBERT, R. 1998. Formation of chromitites by assimilation of crustal pyroxenites and gabbros into peridotitic intrusions: North Arm Mountain Massif, Bay of Islands ophiolite, Newfoundland, Canada. Journal of Geophysical Research, 103, 5165-5184. BERCLAZ, A., BEDARD, J.H., HEBERT, R., VARFALVY, V., MELANCON, B., YAO, K.R. & EL BILALI, L. 1994. Tectonic evolution of the plutonic section of the Bay of Islands Ophiolite. EOS Transactions, American Geophysical Union, 75, 650-651. BERCLAZ, A., BEDARD, J.H. & HEBERT, R. 1998. Importance and significance of deformation processes in the genesis of oceanic crust in a slowspreading environment: evidences from the Bay of Islands ophiolite, Newfoundland. Recueil des resumes—Association geologique du Canada; Association mineralogique du Canada; Association professionelle des geologues et geophysiciens du Quebec; Association Internationale des hydrogeologues; Union geophysique canadienne, Quebec, 23, 15. BIDEAU, D., HEBERT, R., HEKINIAN, R. & CANNAT, M. 1991. Metamorphism of deep-seated rocks from the Garrett ultrafast transform (East Pacific Rise near 13°25'S). Journal of Geophysical Research, 96, 10079-10099. CASEY, J.F. 1980. Geology of the southern half of North Arm Mountain Massif, Bay of Islands ophiolite complex, southwestern Newfoundland. PhD thesis, State University of New York at Albany. CASEY, J.F., DEWEY, J.F., Fox, P.J., KARSON, J.A. & ROSENCRANTZ, E. 1981. Hetcrogenous nature of the oceanic crust and upper mantle: a perspective from the Bay of Islands ophiolite complex. In: EMILIANI, C. (ed.) The Sea. Wiley, New York, 7, 305-338. CASEY, J.F., KARSON, J.A., ELTHON, D.L., ROSENCRANTZ, E. & TITUS, M. 1983. Reconstruction of the geometry of accretion during formation of the Bay of Islands ophiolite complex, North Arm Mountain, Newfoundland. Canadian Journal of Earth Sciences, 2, 509-528. CASEY, J.F., ELTHON, D.L., SIROKY, F.X., KARSON, J.A. & SULLIVAN, J. 1985. Geochemical and geological

HYDROTHERMAL METAMORPHISM IN GABBROS evidence bearing the origin of the Bay of Islands and Coastal Complex ophiolites of Western Newfoundland. Tectonophysics, 116, 1-40. CAWOOD, P.A. & SUHR, G. 1992. Generation and obduction of ophiolites; constraints for the Bay of Islands Complex, western Newfoundland. Tectonics, 11, 884-897. CLAYTON, R.N. & MAYEDA, T.K. 1963. The use of bromine pentafluoride in the extraction of oxygen from oxides and silicates for isotopic analysis. Geochimica et Cosmochimica Acta, 27, 43-52. COLE, D.R. & OHMOTO, H. 1986. Kinetics of isotopic exchange at elevated temperatures and pressure. In: VALLEY, J.W., TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geologic Processes. Mineralogical Society of America, Reviews in Mineralogy, 16, 41-90. DALLMEYER, R.D. & WILLIAMS, H. 1975. 40Ar/39Ar ages from the Bay of Islands metamorphic aureole, their bearing on the timing of Ordovician ophiolite obduction. Canadian Journal of Earth Sciences, 12, 1685-1690. DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. 1992. An Introduction to the Rock-Forming Minerals. Longman, Harlow, 696. EL BILALI, L. 1995. Origine des trondhjemites dans le massif North Arm Mountain, complexe ophiolitique de Bay of Islands, Terre-Neuve. MSc thesis, Universite Laval, Quebec. ELTHON, D., LAWRENCE, J.R., HANSON, R.E. & STERN, C. 1984. Modelling of oxygen-isotope data from the Sarmiento ophiolite complex, Chile. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Blackwell Scientific, Oxford, 185-197. FRIEDRICHSEN, H. & ROMMEL, U. 1985. Stable (O, H) and radiogenic (Sr) isotope investigations of the Troodos Ophiolitic complex. EOS Transactions, American Geophysical Union, 66, 1128-1129. GAGGERO, L. & CORTESOGNO, L. 1997. Metamorphic evolution of oceanic gabbros; recrystallization from subsolidus to hydrothermal conditions in the MARK area (OOP Leg 153). Lithos, 40, 105-131. GILLIS, K.M. 1995. Controls on hydrothermal alteration in a section of fast-spreading oceanic crust. Earth and Planetary Science Letters, 134, 473-489. GILLIS, K.M., THOMPSON, G., KELLEY, D.S. & 1993, A. 1993. view of the lower crustal component of hydrothermal systems at the Mid-Atlantic Ridge. Journal of Geophysical Research, 98, 19597-19619. GREGORY, R.T. 1980. Oxygen and hydrogen isotope study for Samail ophiolite, Oman: implications for origin and hydrothermal alteration of the oceanic crust. PhD thesis, California Institute of Technology, Pasadena. GREGORY, R.T. & CRISS, R.E. 1986. Isotopic exchange in open and closed systems. In: VALLEY, J.W., TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geologic Processes. Mineralogical Society of America, Reviews in Mineralogy, 16, 91-127. GREGORY, R.T. & TAYLOR, H.G. JR 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: evidence for 618O

399

buffering of the ocean by deep (>5 km) seawaterhydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research, 86, 2737-2755. HARPER, G.D., BOWMAN, J.R. & KUHNS, R. 1988. A field, chemical, and stable isotope study of subseafloor metamorphism of the Josephine ophiolite, California-Oregon. Journal of Geophysical Research, 93(B5), 4625-4656. HEBERT, R. & CONSTANTIN, M. 1991. Petrology of hydrothermal metamorphism of oceanic layer 3, implications for sulfide parageneses and redistribution. Economic Geology, 83, 472-485. HELZ, R.T. 1982. Phase relations and compositions of amphiboles produced in studies of the melting behavior or rocks. In: VEBLEN, D.R. & RIBBE, P.H. (eds) Amphiboles: Petrology and Experimental Phase Relations. Mineralogical Society of America, Reviews in Mineralogy, 9B, 279-346. HEY, M.H. 1954. A new review of the chlorites. Mineralogy Magazine, 30, 227-292. HOLLAND, T. & BLUNDY, J. 1994. Non-ideal interactions in calcic amphiboles and their bearing amphiboleplagioclase thermometry. Contributions to Mineralogy and Petrology, 116, 433-447. ITO, E. & CLAYTON, R.N. 1983. Submarine metamorphism of gabbros from the Mid-Cayman Rise: an oxygen isotopic study. Geochimica et Cosmochimica Acta, 47, 535-546. JENNER, G.A., DUNNING, G.R., MALPAS, J., BROWN, M. & BRACE, T. 1991. Bay of Islands and Little Port complexes, revisited: age, geochemical and isotopic evidence confirm suprasubduction-zone origin. Canadian Journal of Earth Sciences, 28, 1635-1652. KEMPTON, P.D., HAWKESWORTH, C.J. & FOWLER, M. ET AL. 1991. Geochemistry and isotopic composition of gabbro from layer 3 the Indian Ocean crust, Leg 118, Hole 735B. In: VON HERZEN, R.P. & ROBINSON, P.T. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 118. Ocean Drilling Program, College Station, TX, 127-144. KURTH, M., SASSEN, A., SUHR, G. & MEZGER, K. 1998. Precise ages and isotopic constraints for the Lewis Hills (Bay of Islands Ophiolite): Preservation of an arc-spreading ridge intersection. Geology, 26, 1127-1130. LAIRD, J. & ALBEE, A.L. 1981. High pressure metamorphism in mafic schist from northern Vermont. American Journal of Science, 86, 2777-2782. LEAKE, B.H. 1978. Nomenclature of amphiboles. Canadian Mineralogist, 16, 501-520. LECUYER, C. & REYNARD, B. 1996. High-temperature alteration of oceanic gabbros by seawater (Hess Deep, Ocean Drilling Program Leg 147): evidence from oxygen isotopes and elemental fluxes. Journal of Geophysical Research, 101, 15883-15897. LINDSLEY, D.H. 1983. Pyroxene thermometry. American Mineralogist, 68, 477-493. LISTER, C.R.B. 1974. On the penetration of water into hot rocks. Geophysical Journal of the Royal Astronomical Society, 39, 465-509. LYDON, J.W. & LAVIGNE, J. 1990. Magmatic sulphides and gabbro remelting in the Bay of Islands ophiolite, Newfoundland: evidence for seawater penetration. Symposium on Ophiolite Genesis and

400

E. GIGUERE ETAL.

Evolution of the Oceanic Lithosphere, Muscat, Sultanate of Oman, 7-18 January 1990. Ministry of Petroleum & Minerals, Muscat, 135. MALPAS, J.G. 1976. The petrology and petrogenesis of the Bay of Islands Ophiolite Suite, Western Newfoundland. PhD thesis, Memorial University of Newfoundland, St. John's. McCuLLOCH, T.M., GREGORY, R.T., WASSERBURG, GJ. & TAYLOR, H.P. 1981. Sm-Nd, Rb-Sr, and 18O/ 16 O isotopic systematics in an oceanic crustal section: evidence from the Samail ophiolite. Journal of Geophysical Research, 86, 2721-2735. MEVEL, C. 1987. Evolution of oceanic gabbros from DSDP Leg 82; influence of the fluid phase on metamorphic crystallizations. Earth and Planetary Science Letters. 83, 67-79. MEVEL, C. & CANNAT, M. 1991. Lithospheric stretching and hydrothermal processes in oceanic gabbros from slow-spreading ridges. In: PETERS, TJ., NICOLAS, A. & COLEMAN, R.G. (eds) Ophiolitic Genesis and Evolution of the Oceanic Lithosphere. Kluwer Academic, Dordrecht, 293-312. MUEHLENBACHS, K. 1986. Alteration of the oceanic crust and the 18O history of seawater. In: VALLEY, J.W., TAYLOR, H.P. JR & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geologic Processes. Mineralogical Society of America, Reviews in Mineralogy, 16, 425-444. MUEHLENBACHS, K. 1998a. The oxygen isotopic composition of the oceans, sediments and the seafloor. Chemical Geology, 145, 261—211. MUEHLENBACHS, K. 1998b. The oxygen isotopic composition of Paleozoic ophiolites and oceans. Recueil des resumes—Association geologique du Canada; Association mineralogique du Canada; Association professionelle des geologues et geophysiciens du Quebec; Association Internationale des hydrogeologues; Union geophysique canadienne, Quebec, 23, 130. SCHIFFMAN, P. & SMITH, B.M. 1988. Petrology and oxygen isotope geochemistry of a fossil seawater hydrothermal system within the Solea Graben, Northern Troodos Ophiolite, Cyprus. Journal of Geophysical Research, 93, 4612-4624. SEARLE, M.P. & STEVENS, R.K. 1984. Obduction processes in ancient, modern and future ophiolites. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Blackwell Scientific, Oxford, 303-319. SPEAR, F.S. 1980. NaSi = CaAl exchange equilibrum between plagioclase and amphibole: an empirical model. Contributions to Mineralogy and Petrology, 72,33-41. STAKES, D.S. & TAYLOR, H.P. JR 1992. The Northern Samail Ophiolite: an oxygen isotope, microprobe, and field study. Journal of Geophysical Research, 97, 7043-7080. STAKES, D.S. & VANKO, D.A. 1986. Multistage hydrothermal alteration of gabbroic rocks from the failed Mathematician Ridge. Earth and Planetary Science

Letters, 79, 75-92. STAKES, D.S., TAYLOR, H.P. JR & FISHER, R.L. 1984. Oxygen-isotope and geochemical characterization of hydrothermal alteration in ophiolite complexes and modern oceanic crust. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Blackwell Scientific, Oxford, 199-214. STAKES, D.S., MEVEL, C., CANNAT, M. & CHAPUT, T. ET AL. 1991. Metamorphic stratigraphy of Hole 735B. In: VON HERZEN, R.P. & ROBINSON, P.T. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 118. Ocean Drilling Program, College Station, TX, 153-180. TAYLOR, H.P. JR 1977. Water/rock interactions and the origin of H2O in granitic batholiths. Journal of the Geological Society, London, 133, 509-558. TAYLOR, R.N., NESBITT, R.W., VIDAL, P., HARMON, R.S., AUVRAY, B. & CROUDACE, I.W. 1994. Mineralogy, chemistry, and genesis of the boninite series volcanics, Chichijima, Bonin Islands, Japan. Journal of Petrology, 35, 577-617. VANKO, D.A. & STAKES, D.S. ET AL. 1991. Fluids in oceanic layer 3, evidence from veined rocks, hole 735B, Southwest Indian Ridge. In: VON HERZEN, R.P. & ROBINSON, P.T. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 118. Ocean Drilling Program, College Station, TX, 181-215. VARFALVY, V., HEBERT, R. & BEDARD, J.H. 1996. Interactions between melt and upper-mantle peridotites in the North Arm Mountain Massif, Bay of Islands Ophiolite, Newfoundland, Canada: implications for the genesis of boninitic and related magmas. Chemical Geology, 129, 71-90. VARFALVY, V., HEBERT, R., BEDARD, J.H. & LAFLECHE, M.R. 1997. Petrology and geochemistry of pyroxenite dykes in upper mantle peridotites or the North Arm Mountain Massif, Bay of Islands Ophiolite, Newfoundland: implications for the genesis of boninitic and related magmas. Canadian Mineralogist, 35, 543-570. VIBETTI, N.J., KERRICH, R. & FYFE, W.S. 1989. Oxygen and carbon isotope studies of hydrothermal alteration in the Troodos ophiolite complex, Cyprus. In: GIBSON, I.L., MALPAS, J., ROBINSON, P.T. & XENOPHONTOS, X. (eds) Cyprus Crustal Study Project: Initial Report, Hole CY-4. Geological Survey of Canada, Paper, 88-9, 221-228. YAO, K.R. 1995. Etude des parageneses sulfurees et silicatees dans les gabbros crustaux du massif de North Arm Mountain, ophiolite de Bay of Islands; Terre-Neuve (Canada). MSc thesis, Universite Laval, Quebec. ZHENG, Y.F. 1993a. Calculation of oxygen isotope fractionnation in hydroxyl-bearing silicates. Earth and Planetary Science Letters, 120, 247-263. ZHENG, Y.F. 1993b. Calculation of oxygen isotope fractionation in anhydrous silicate minerals. Geochimica et Cosmochimica Acta, 57, 1079-1091.

Ophiolites as faithful records of the oxygen isotope ratio of ancient seawater: the Sohmd-Stavfjord Ophiolite Complex as a Late Ordovician example KARLIS MUEHLENBACHS1, HARALD FURNES2, HEGE C. FONNELAND2 & BJARTE HELLEVANG 2 1 Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB T6G OE2, Canada (e-mail: [email protected]) ^Department of Earth Science, University of Bergen, Allegt. 41, 5007 Bergen, Norway Abstract: Fragments of the Ordovician sea floor preserved in the Solund-Stavfjord Ophiolite Complex in Western Norway serve as proxies for the d 18 O of Ordovician seawater. The pillow basalt sections at Oldra and Strand are both enriched in 18O, recording their alteration by seawater at low temperature on the sea floor. In contrast, the sheeted dykes and gabbros generally are depleted of 18O, reflecting the modal proportion of secondary, low-18O chlorite and epidote formed from seawater at high temperature. These isotopic contrasts simply reflect the high water to rock ratio of sea-floor alteration and the temperature dependence of the 18O partitioning between minerals and water. Superposition of high-<518O pillows on low-(518O dykes and gabbros is a necessary consequence of alteration at both low and high temperatures by a fluid near 0%o and is easily recognized in well-preserved ophiolites. Also, the d 18 O of seawater can be independently calculated from 18O fractionations among secondary minerals. Older, dismembered and highly metamorphosed segments of the oceanic crust may still retain the original seawater imprint because their subsequent obduction and metamorphism was relatively closed to external fluids. Suites of diamond-bearing nodules from kimberlites still have contrasting high- and low-d18O eclogites, proving that even subduction into the mantle is not sufficient to erase the seawater fingerprint. Inspection of the sea-floor, ophiolite and eclogite data reveals no secular trend in (518O, indicating that the (518O of seawater has not changed with geological age. Because the d18O of seawater itself is fixed by sea-floor-seawater exchange, the constancy of (518O of seawater implies that the scale and style of sea-floor-seawater interactions has not changed over time.

The composition and temperature of seawater can be thought of as reflecting the state of our planet. Changes in global climate or tectonic style should cause secular trends or cycles in seawater composition and temperature. A massive body of literature attempts to decipher changes in seawater over geological time from geochemical proxies preserved in contemporaneous fossils and sediments (e.g. Veizer et aL 2000). A key limitation to the unambiguous interpretation of the geochemical record is a lack of agreement, even after three decades of debate, on the oxygen isotope ratio of ancient oceans (see discussion by Muehlenbachs 1998). Samples of ancient seawater are no longer available for study so most researchers of secular trends have utilized marine fossils and sediments as geochemical proxies for seawater (Veizer et aL 1999; Jacobsen & Kaufman 1999, and many others). However, for Mesozoic and older oceans, these fossils and sediments may be poor proxies in that they invariably are from epicontinental seas

and thus may not be representative of the open global ocean even if the primary compositions of the fossils have not been diagenetically altered (Holmden & Muehlenbachs 1993a). In this paper we will demonstrate that fragments of ocean crust preserved in ophiolites are valid proxies for the isotopic composition of ancient ocean water using as a prime example the Ordovician ophiolites of Western Norway. As the sea floor forms and ages it comes into contact with seawater in whichever tectonic environment it is actually formed. The inevitable chemical and mineralogical reactions that follow cause a flux of elements to be exchanged between the crust and seawater that imprints a geochemical fingerprint of seawater on the crust. Mass balance demands for some elements that the alteration of the ocean crust dominates the chemical and isotopic composition of seawater itself. Oxygen isotope ratios are perhaps the clearest example of a geochemical balance between the altered sea

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 401-414. 0305-8719/037$ 15 © The Geological Society of London 2003.

402

K. MUEHLENBACHS ETAL.

floor and seawater (Muehlenbachs & Clayton 1976). We will review (518O studies on ophiolites to assess how well these rocks preserve their seafloor alteration history and evaluate their usefulness as proxies for the isotopic composition of ancient seawater. A new <518O profile through the Solund-Stavfjord Ophiolite Complex (SSOC) ophiolite will be presented and an estimate made of the contentious (318O value of Ordovician seawater.

Review of sea-floor-seawater 18O exchange and its preservation in ophiolites The alteration of the sea floor by seawater is pervasive and occurs over a wide temperature range with contrasting results on (518O. At low temperature (below 150°C) basalts are altered to high-(518O clays, zeolites and carbonates, stripping 18 O by mass balance from seawater (Muehlenbachs & Clayton 1972a). At higher temperatures (250-350 °C), <518O-depleted chlorites and epidotes form in the altered dykes and gabbros (Muehlenbachs & Clayton 1972b), adding 18O to seawater. Moreover, near 300 °C the equilibrium 18 O fractionation factor between water and altered basalt is nearly the same as the difference between seawater and unaltered mid-ocean ridge basalt (MORE), leading to massive scrambling of 18O between the crust and seawater but with no net change in (518O. The result is that alteration of the sea-floor 'buffers' the (518O of seawater to its present value of 0 ± 2%o (SMOW) (Muehlenbachs & Clayton 1976; Gregory & Taylor 1981; Muehlenbachs 1998). The most easily recognized impact of seawater alteration on the sea floor is this superposition of 18 O-enriched extrusive rocks on 18O-depleted intrusive rocks. Ophiolites can serve as proxies for (518O of seawater if they preserve the abovementioned contrasts in (518O. Also, the (518O of seawater can be independently estimated from the (318O of seawater-derived hydrothermal fluids as calculated from 18O fractionations between coexisting secondary minerals such as quartz and epidote. If, as Perry (1967), Veizer et al (1999) and many others have suggested, the (518O of Palaeozoic and older seawater was —3 to — 18%o, then rocks and minerals of old ophiolites should be proportionately lower in (518O compared with those of the modern sea floor. Even in a low-18O ocean, pillows in the ophiolites would still be 26%o higher in (318O than exceptionally low-18O cogenetic dykes and gabbros (Holmden & Muehlenbachs 1993b). Ophiolitic rocks are inherently altered and their (518O integrity needs to be examined. The state of

preservation in these rocks ranges from unaltered relict MORB glass on Macquarie Island (Cocker et al. 1982) to highly metamorphosed diamondbearing eclogite nodules in kimberlites (Garlick et al. 1971). To establish the utility of such altered rocks as seawater proxies one has to examine the processes that may alter the (518O of ophiolites. The alteration of (518O in an ophiolite can occur in at least four stages: (1) primary alteration by seawater at the time of formation at a spreading centre; (2) off-axis, low-grade hydrothermal and diagenetic alteration of the crust by seawater or pore water as it ages; (3) metamorphic alteration during subduction or obduction; (4) any one or more post-obduction processes depending on the vagaries of the geological history of the ophiolite. We briefly discuss below the effect these alteration stages may have on their (518O record. The first premise, that ophiolites can record the primary (518O imprint of seawater, is shown by the (518O profile at the well-preserved Samail, Oman ophiolite (Gregory & Taylor 1981; Stakes & Taylor 1992), that matches the (518O enrichments and depletions observed on the sea floor (Muehlenbachs & Clayton 1972a, 1972b; Alt et al. 1986). Gregory & Taylor were able to calculate the expected <518O of Cretaceous seawater from the mass balance of 18O enrichments and depletions observed in the altered rocks of the Samail profile. The Samail section is often cited as the 'type' (518O profile for the ocean crust. However, it should be noted that the (318O of hydrothermally altered oceanic crust is spatially heterogeneous on a 10-103 m scale. Gillis et al. (2001) contoured 'bull's eye' alteration patterns in (518O of 2.5-6.5%o in the East Pacific Rise crust exposed in the Hess Deep, and suggested marked localized zones of seawater recharge and discharge within the hydrothermal system. The changes of (518O in rocks of any one profile of the oceanic crust will show the complementary 18O enrichments and depletions, but need not add up to zero across a particular profile because of the aforementioned heterogeneity. The (318O of the oceanic crust can be further altered as it ages and moves off axis. On the flanks of the ridge there will be moderate- to lowtemperature alteration by off-axis hydrothermal activity (Elderfield et al. 1999). Further along, the pillow basalts beneath the abyssal sediments will continue to react with pore waters at ambient temperature (Lawrence & Gieskes 1981). The overall effect is to further increase the <518O in affected parts of the crust. This secondary alteration is by seawater, so such off-axis overprinting does not weaken in itself the utility of ophiolites as proxies for (518O of seawater, as demonstrated by Lecuyer & Fourcade (1991) in a study on the

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN Silurian Trinity ophiolite, California. An extreme example of 18O enrichment may be seen in the Xigaze ophiolite, Tibet, where Agrinier et al. (1988) could still deduce the (318O of Cretaceous seawater. Obduction clearly has not altered the <518O of the Macquarie or of the Samail ophiolite. The effects of subduction-related metamorphism and metasomatism on the (518O of less perfect exposures have been addressed by a series of papers on Tethyan and Alpine ophiolites by Cartwright & Barnicoat (1999), Miller & Cartwright (2000a, 2000b) and Miller et al (2001). They found that the (318O contrast between the 18O-enriched pillows and 18O-gabbros is preserved even across the gabbro-eclogite transition (Barnicoat & Cartwright 1997; Putlitz et al. 2000). Extraneous (518O effects by subduction processes were generally confined to cross-cutting veins. Isotopic resetting was observed on hand-specimen scale but not outcrop scale. These studies clearly show that even highly metamorphosed ophiolites preserve their (518O profiles and thus continue to serve as proxies for seawater because their obduction and metamorphism were relatively closed to external fluids. The robustness of the (518O profile during metamorphism and dismemberment is remarkable. Cartwright & Valley (1991) suggested that the low (518O signal imposed by seawater alteration on the lower parts of the oceanic crust was still evident in the Scourie dyke complex of the Lewisian complex. Diamond-bearing eclogites display the same low to high (518O values as the sea-floor (518O profile (Garlick et al. 1971), implying that their (518O values are not homogenized during subduction, storage in the mantle for 109 years, disruption and subsequent eruption in kimberlites. Many workers now cite the contrasting <518O values of eclogites as evidence that these nodules are relics of the oceanic crust (MacGregor & Manton 1986; Ongley et al. 1987; Neal et al. 1990; Jacob et al. 1994; Snyder et al. 1995; Beard et al. 1996; Earth et al. 2001). The diamondbearing eclogites may thus serve as proxies for the (518O of Archaean seawater. In this paper we present new results of a (518O study of the Ordovician Solund-Stavfjord Ophiolite Complex (SSOC) of Western Norway and deduce the (518O of Ordovician seawater. No secular trends in (518O are evident in ophiolites ranging in age from Miocene (Cocker et al. 1982) to Archaean (Hoffman et al. 1986; de Wit & Hart 1993). Heretofore the (518O values of all ophiolitic rocks are similar to their modern sea-floor equivalents but more studies are justified because the (518O of Palaeozoic seawater is controversial and its true value has direct bearing on several new

403

models on global climate change and history of the oceans.

The Solund-Stavfjord Ophiolite Complex (SSOC) The SSOC (Fig. la) of the lapetus oceanic lithosphere developed in a Caledonian marginal basin (Dunning & Pedersen 1988; Andersen et al. 1990; Fumes et al. 1990, 2000). In the southern and western parts of the ophiolite terrain the oceanic crust contains three structural domains that display different crustal architecture reflecting the mode and nature of magmatic and tectonic processes that operated during the multi-stage sea-floor spreading evolution of this basin (Dilek et al. 1997; Furnes et al. 1998). Domain I represents the youngest crust, with NNE structural grain defined by its dyke swarms and sheeted dyke complex. Domain II, which represents the oldest crust, contains a NW-trending sheeted dyke complex and high-level gabbros. Where Domains I and II approach each other on Tviberg (Fig. Ib), the dykes of Domain II curve from a NW direction to an ENE direction towards a serpentine-bearing, prominent shear zone, defined as Domain III, interpreted as part of a transform fault (Skjerlie & Furnes 1990: Dilek et al. 1997). Based on the overall construction of the volcanic successions, the tectonic style and geochemistry, Dilek et al. (1997) proposed that the NNE-trending Domain I represents the remnant of a spread rift system that propagated NNE-wards into the pre-existing oceanic crust, which was developed along the NWtrending doomed rift (Domain II) in the marginal basin (Fig. Ib). The samples for this study were collected from two sections within the SSOC, the Oldra section and the Strand section (Fig. Ic). Whereas the Oldra section belongs to Domain I, the position of the Strand section is ambiguous, i.e. it may belong either to Domain I or to Domain II. In the Strand area (Fig. Ic) the first generation of major folds resulted in a 90° tilting of the sequence, thus rotating the bedding of the volcanic rocks to a vertical position, whereas the orientation of the dykes is subhorizontal. Thus, even though the rocks of the Strand area are strongly deformed through the first and two subsequent phases of deformation (Furnes 1974), it has none the less been possible to establish a complete section of 500 m of volcanic rocks overlain by black schist and metasandstone, a 150m thick transition zone of volcanic rocks and dykes, and a 900 m thick sheeted dyke complex rooted in, or in places cut by gabbro (exposed thickness of 800 m).

404

K. MUEHLENBACHS ETAL.

Fig. 1. (a) Location map, showing the occurrence of the Solund-Stavfjord ophiolite within the Norwegian Caledonides. (b) Model of the domains in the Solund-Stavfjord ophiolite. (c) Geological map and profile of the islands of Oldra and Strand. Modified form Dilek et al. (1997).

In the Oldra area the sequence is isoclinally folded (in major folds during the first phase of deformation) and faulted, thus resulting in vertical attitude of the bedding of the volcanic rocks and the dyke orientation. Thus the exact stratigraphic position of the volcanic rocks and the depth of dyke and gabbro samples in the Oldra section are

uncertain, in contrast to the exact sample position in the Strand section. The proportion of pillow lavas, volcanic breccias (hyaloclastite and pillow breccias) and massive sheet flows of the SSOC varies greatly. Thus, in the Stand and Oldra profiles pillow lavas, volcanic breccias and massive sheet flows occur in

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN the proportions 35%, 30%, 35%, and 70%, 20%, 10%, respectively. Volcanologically, the volcanic sequence was constructed in a cyclic manner, in which the lowest part of a volcanic cycle is characterized by sheet flows and/or large pillows followed by lavas in which pillows become progressively smaller upward (Furnes et al. 2001). Where the volcanic sequence is thickest (c. 800m), the crust was constructed during at least seven volcanic cycles that resulted in stratigraphic units with thicknesses ranging from 40 to 225 m (Furnes et al. 2001). The least deformed section of Domain I is represented on the island of Oldra (Fig. Ic), and is composed of extrusive rocks, a transition zone, a sheeted dyke complex and high-level, vari-textural gabbros. For the detailed description given below of the various volcanic lithologies, dykes and gabbro, as well as their petrographic characteristics, the Oldra section is the most appropriate. The ophiolite pseudostratigraphic units are separated by steep-dipping faults or fracture zones (Fig. Ic). Extrusive rocks The extrusive sequence is composed of pillow lavas, pillow breccias and massive lava flows. Pillow lavas are predominant, whereas massive lava flows show a minor occurrence. Pillow lava flows range in thickness from 1 m to about 40 m and they are separated by jasper or chert horizons, faults or fractures, pillow breccias or massive flows. Epidote- and quartz-filled drainouts occur abundantly throughout the pillow lavas. The pillow lavas are non-amygdaloidal, and by applying the vesicle content-eruption depth relation of Moore & Schilling (1973), the Oldra samples would suggest that eruptions probably occurred at water depths greater than 2.5 km. Pillow breccias consist of both isolated-pillow breccias and broken-pillow breccias (Furnes 1972), with pillow fragments and isolated pillows of variable sizes set in a matrix of hyaloclastite. The pillow breccia layers vary from 1 to 16 m in thickness, and their occurrence varies from 15% in the middle profile to 30% in the southern profile. Foliation is more distinguished in the breccias than in the other extrusive members. Several concordant layers of massive greenstone occur in the volcanic pile, and their brecciated or pillowed contact relations to the adjacent pillow lavas or pillow breccias would indicate that they represent massive lava flows. Similar chill relations of massive flows have been reported by Baragar (1984). The massive flows occupy about 5% of the profile. Their thicknesses vary from 0.5 to 8m and they can be followed along-strike for 75-100 m.

405

Transition zone (TZ) The TZ is defined as the zone between the 100% sheeted dyke complex and the extrusive rocks of the ophiolite pseudostratigraphy. It contains both extrusive and intrusive rocks in a 200 m wide zone in the northwestern half of Oldra (Fig. Ic). The rocks in the TZ are foliated, and the foliation is subparallel to the dyke orientation. Epidote as veins, ranging in width from 1 to 5 mm, and as knolls occur throughout the TZ. The epidote veins generally define a net-vein-like appearance, or they are oriented parallel to the foliation. Sheeted dyke complex The NE-trending sheeted dyke complex contains subparallel, steeply dipping dykes with one- and two-sided chilled margins that display multiple intrusive relations. The dykes vary in width from a few centimetres to a couple of metres; the average width of unsplit dykes is 0.71 m (Ryttvad et al. 2000). The dyke complex can be separated into several fragments bounded by fractures or faults, which are subparallel to the dykes. There is no clear evidence for significant amount of rotation or back-tilting of dykes along these fractures or faults. Throughout the sheeted dyke complex, epidote veins of 1-5 mm width occur. These may be parallel, perpendicular or oriented at an angle of about 45° to dyke margins. Epidote lenses also occur in the central part of dykes and in breccia zones along dyke margins. Some dykes are offset along veins; at one location the epidote-filled fault is crosscut by a new dyke, clearly indicating oceanic origin of the fault. High-level gabbro The gabbros are subdivided into (1) vari-textured gabbro (Pedersen 1986; Skjerlie & Furnes 1996); (2) fine-grained gabbro. The fine-grained gabbro is intruded by 5-50 cm thick, irregular dykes, which are interpreted to be fossil dyke roots (Skjerlie & Furnes 1996). Veins of epidote, quartz and amphibole are common in the plutonic sequence, and oxides appear as rust-coloured patches (5 cm X 15 cm to 10 cm X 25 cm). A few epidosite zones (15-25 cm wide) occur within the gabbros. Petrography All the rocks of Oldra have been variably recrystallized under greenschist-facies conditions and commonly they exhibit the secondary mineral assemblage epidote + chlorite + quartz + actinolite + albite. Downward changes in mineral

406

K. MUEHLENBACHS ETAL.

assemblage and texture are present in the ophiolite pseudostratigraphy of Oldra. The extrusive rocks are slightly more altered than the rocks of the other units. In particular, the hyaloclastite breccias have developed a pronounced foliation. Most commonly their primary magmatic textures are completely obliterated, and chlorite is a dominant mineral. The mineral assemblage in the extrusive rocks consists of albite, chlorite, epidote, actinolite, quartz, hematite, magnetite, ± hornblende, ± sulphides and ± calcite. Some pillow rims show excellent quench textures, although most of the pillows are too deformed to preserve any primary magmatic textures. The rocks in the transition zone contain the lowest amount of chlorite, probably as a result of the smaller amount of chlorite-altered pillow rims. Textures in this unit are poorly preserved. Relict Fe-Ti oxide is rare and largely altered, and occurs only within the gabbros and sporadically in the extrusive rocks. Other relict minerals are rarely present, represented only by moderate occurrence of relict pyroxene in the gabbros. All of the primary calcic plagioclase seem to be replaced by the more sodic albite. The mineral assemblage of the dyke complex resembles that of the extrusive rocks, although the latter contain slightly more chlorite and quartz and smaller amounts of sulphides. The transition zone contains smaller amounts of albite, chlorite and magnetite compared with the extrusive rocks and the dyke complex. In the gabbros the mineral assemblage has changed, and the dominant minerals are albite and actinolite.

Isotopic data and discussion Whole-rock data Eighty-six samples have been analysed for <518O (Tables 1 and 2) from transects of the SSOC on the islands of Oldra and Strand (Fig. 1). The samples, selected from all of the representative ophiolite units of the SSOC (volcanic rocks, transition zone, the 100% dyke complex and gabbros), include rocks that show moderate to extensive hydrothermal alteration and their fresh equivalents. Quartz and epidote were separated and analysed from a set of epidosites spanning the ophiolite section (Table 3). Whole-rock powders were reacted with BrF5 at 625 °C to liberate oxygen (Clayton & Mayeda 1963). Oxygen was converted to CO2 and analysed for isotopic ratio on a Finnigan-MAT 252 mass spectrometer. The data are reported in the usual delta notation with respect to SMOW (Standard Mean Ocean Water) (Craig 1961) using a water/CO2 fractionation factor of 1.0407 to ensure compatibility of these

data with previous studies of mid-ocean ridge rocks. Analytical reproducibility was ±0.1 %o. The hydrogen isotope measurements on epidote were carried out by a method modified from Godfrey (1962) and Coleman et al (1982). The oxygen isotope compositions of whole-rock samples from the Oldra and Strand sections of the SSOC are given in Tables 1 and 2 along with their lithological description and depth beneath the sedimentary contact. The data range widely from 2.4 to 8.8%o, when compared with the unaltered MORE (5.7 ± 0.3%o), and show on a histogram (Fig. 2) a bimodal distribution with peaks at 4.5 and 7.5%o. Lithological variation can explain the isotopic distribution. As expected from previous studies of Deep Sea Drilling Project (DSDP), Ocean Drilling Program (ODP) and ophiolite samples (Muehlenbachs 1986), volcanic samples have higher (518O values because of alteration by cold seawater, whereas the dykes and gabbros have lower (518O values as a result of interaction with hydrothermal fluids. The peak at 4.5%o observed in SSOC dykes and gabbros is also found in the large data set of hydrothermally altered dykes of the East Pacific Rise (Gillis et al 2001) and is also the average (518O value for the hydrothermally altered crust compiled from the literature (Muehlenbachs 1998). The extrusive rocks of the SSOC are not quite as 18O enriched as are most sea-floor basalts (8-10%o, Muehlenbachs 1998). There could be a wide variety of explanations for this apparent discrepancy. Perhaps the pillows of the SSOC spent too short a time beneath the sea to be extensively weathered. The maximum time constraints between the construction and obduction of the SSOC, based on U/ Pb dating of zircons from quartz diorites (443 ± 3 Ma) from Domain II of the SSOC (Dunning & Pedersen 1988), and the occurrence of fossils (pentamerids) of Wenlock age in part of the obduction melange (Andersen et al. 1990), would be about 20 Ma. Perhaps the basalt weathering occurred at higher temperatures, or perhaps some 18O was removed during obduction and subsequent metamorphism, or we failed to sample the most altered pillows. That the (518O of SSOC pillows has not been pervasively reset is shown from the (518O values of three concentric samples from one deformed pillow and the adjacent material between pillows (Table 1, samples 4-7). These pillow samples retain an 18O gradient as seen in dredged but weathered MORE (Muehlenbachs & Clayton 1972a). The core of the pillow from Oldra has a (518O of unaltered basalts (5.6%o); towards the rim there is a slight increase (5.9-6.3%o) whereas the material between pillows has a (318O value of 8.0%o. Preservation of such 18 O gradients in meta-pillows is prima-facie evi-

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN

407

Table I . Oxygen isotopic composition of rocks from Oldra profile of the SSOC Sample no.

Rock type

1 3 4 5 6 7 8 9 10 13 14 15 16 17 19 22 24 25 27 28 29 30 32 33 34 35 36 37 38 39 40 41 43 44 45 46 47 48 49

Pillow & hyal. Pillow centre Pillow centre Pillow intermediate layer Pillow margin Interpillow Massive flow Massive flow Massive flow Massive flow Pillow & hyal. Pillow & hyal. Pillow & hyal. Pillow & hyal. pillow & hyal. Dyke Dyke Pillow & hyal. Dyke Dyke Dyke Dyke Dyke Breccia Breccia Dyke Dyke Dyke Dyke Breccia Breccia Dyke Gabbro Dyke Gabbro Gabbro Gabbro Gabbro Dyke

dence of the lack of geochemical resetting during obduction and metamorphism of Tethyan ophiolites (Miller & Cartwright 2000a). Figure 3 is a plot of the (518O of rock v. depth below the contact with sediment of both the Strand section of the SSOC and Hole 504B (Alt et al 1996) from the Costa Rica Rise. Coherence of (518O v. depth data is evident from both the 443 Ma ophiolite and the 5.9 Ma sea floor. With the exception of just one low-18O pillow at 200 m depth, the Strand and Hole 504B data overlap. The (518O data from the Oldra profile (Table 1) complement the comparison of isotopic profiles of the SSOC with Hole 504B if the comparison were

Depth below sedimentary contact (m)

618O (SMOW) (%o)

2 12 24 24 24 24 46 49 55 102 104 118 118 155 229 605 615 620 630 630 920 910 930 940 950 960 970 980 990 1000 1010 1020 1710 1720 1730 1740 1750 1760 1770

5.44 5.94 5.56 5.93 6.35 8.00 5.14 5.37 5.67 6.17 7.05 6.60 7.03 6.18 4.96 3.97 4.07 4.44 4.37 3.04 4.32 3.97 4.17 5.09 3.85 4.15 4.51 4.60 4.01 3.37 6.47 4.87 4.81 3.57 3.90 5.90 2.39 3.86 3.03

based on lithology but not metres measured in outcrop. The Oldra profile is more disrupted than the Strand profile and the sample depths below sediment contact are only crude estimates. The data from both Strand and Oldra demonstrate that obduction and Caledonian metamorphism have not homogenized or destroyed the <518O imprint of the Ordovician seawater on the SSOC.

Quartz and epidote mineral pair analysis Oxygen and hydrogen isotopic data for epidosites and separated quartz and zoisite mineral pairs from both sections of the SSOC are listed in Table

K. MUEHLENBACHS ETAL.

408

Table 2. Oxygen isotopic composition of rocks from the Strand profile of the SSOC Sample no.

Rock type

97-BH-O2 97-BH-03 97-BH-04 BH-06-97 BH-07-97 BH-08-97 97-BH-09 97-BH-10 97-BH-13 97-BH-14 97-BH15 97-BH-18 97-BH-20 97-BH-21 97-BH-33 97-BH-34 97-BH-35 97-BH-36 97-BH-38 97-BH-39 97-BH-42 97-BH-47 97-BH-54 98-BH-34 98-BH-30 98-BH-26 98-BH-22 98-BH-18 98-BH-15 98-BH-40 98-BH-44 98-BH-51

Pillow Massive flow Massive flow Massive flow Massive flow Massive flow Pillow HyaL/breccia Pillow HyaL/breccia Massive flow Pillow HyaL/breccia Massive flow HyaL/breccia HyaL/breccia HyaL/breccia hyal./breccia HyaL/breccia Massive flow Dyke Dyke Dyke Dyke Dyke Dyke Dyke Dyke Dyke Dyke Gabbro Dyke

Fig. 2. Histogram of 618O of extrusive and intrusive rocks from Solund-Stavfjord ophiolite.

Depth below sedimentary contact (m)

618O (SMOW) (%o)

1 34 41 73 108 124 149 159 203 221 235 302 308 315 398 409 420 431 454 465 498 560 666 694 784 874 963 1053 1120 1266 1288 1512

8.76 7.92 8.30 7.46 7.28 6.64 6.87 7.05 5.05 7.69 7.06 6.43 6.26 6.36 6.88 6.63 7.05 6.31 7.30 7.16 7.10 6.54 4.69 4.49 4.34 3.88 4.75 4.96 4.32 4.37 4.97 5.02

3. The (518O values of SSOC epidosites and separated minerals fall within the range previously reported for epidote from the sea floor as well as ophiolites and the Tonga forearc (Muehlenbachs & Clayton 1972b; Heaton & Sheppard 1977; Schiffrnan & Smith 1988; Harper et al 1988; Banerjee et al. 2000). No gradient in (518O with stratigraphic depth is obvious, with the exception of the one sample at the contact with high-18O sediments from Strand. Table 3 also lists the temperature of formation calculated from the 18O partitioning between coexisting quartz and epidote (corrected for measured mole fractionation of pistacite in the Oldra samples and assumed to be an analogous 0.25 at Strand) assuming they formed at equilibrium (Matthews & Schliestedt 1984). The calculated temperatures range from 260 to 470 °C and are plotted v. depth in Figure 4 for both sections. The temperature calculated from the anomalously high (518O mineral pair from within 1 m of the basalt-sediment contact at

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN

Fig. 3. 618O v. depth profile of rocks from the Strand section of Solund-Stavfjord ophiolite. For comparison, data from ODP Hole 504B (Alt et al 1996) are also given on the same depth scale.

409

ism. The uniform metamorphic mineral assemblage of the volcanic rocks is most probably a product of greenschist-facies conditions created during Caledonian deformation, which has overprinted the expected oceanic metamorphic facies (zeolite facies in extrusive rocks, greenschist facies in dyke complex, amphibolite facies in gabbros). However, the poor correlation between the temperature and final metamorphic facies, particularly in the extrusive rocks, implies that the (518O values of the bulk rocks or epidosite veins have not been reset and are representatives of oceanic hydrothermal flow. The (518O of the hydrothermal fluid altering the SSOC can be calculated from the quartz-epidote temperature and (518O of quartz (Clayton et al. 1972). The calculated <518O values of fluids (Table 3) are plotted against depth in Figure 5. The fluid at the top contact at Strand clearly shows equilibration with high-(518O sediments. Hydrothermal fluids calculated for the rest of the ophiolite range in (518O from -0.5 to +4.8%o, averaging +2.2 ± 1.6%o, a value within a per mille of the average (518O of fluids exiting modern black smokers (+1.2 ± 0.5%o, Jean-Baptiste et al 1997). A seawater origin of the hydrothermal fluids at the SSOC is implied by the (5D of -12 and -l%o of two epidotes from Oldra, from the extrusive sequence and transition zone, respectively (Table 3).

The (518O of Ordovician seawater Strand is relatively high, 370 °C. The quartzepidote temperatures deeper within the extrusive sequence are lower but increase with depth (from 260 to 310°C). The sample from the transition zone at Strand gives a slightly lower temperature (290 °C) whereas the epidosite from the TZ at Oldra yields an unexpectedly higher temperature (420 °C), compared with the extrusive rocks and the dyke complex (260-345 °C). The two epidosites from the gabbro complex cover a large temperature range from 290 to 470 °C. The temperature estimates on the pseudostratigraphically different rocks from the SSOC (Fig. 4) and their differing (318O values suggest a multistage history of hydrothermal circulation. The higher (518O of the pillows indicates that they were first altered by low-temperature seawater but then were cross-cut by high-temperature, low-18O epidote-quartz veins. Lack of polymetamorphic mineral associations through the Oldra ophiolite section does not support a retrograde evolution of the alteration temperature. The mineral-pair analyses give alteration temperatures of the extrusive rocks that are too low for greenschist metamorph-

The (518O profiles and mineral separates at Strand and Oldra sections of the Ordovician SSOC are remarkably similar to those observed in the modern oceanic crust, implying that Ordovician seafloor alteration processes were similar to those occurring today. The 18O enrichment at low temperature and simultaneous 18O depletion at high temperature observed in the profiles is possible only if seawater altering the sea floor in the Ordovician was near 0 ± 2%o. The oxygen and hydrogen compositions of quartz and epidote mineral pairs from the SSOC also imply that the (518O (and (3D) of Ordovician black smoker fluid was the same as today's value. The new quartzepidote data are especially significant because the calculated d18O of Ordovician black smoker fluid (2.2 ± 1.5%o) is independent of any putative model calculation based on mass balances or the proportion of sea-floor alteration on or off axis. These observations on the SSOC agree with previous 6nO studies on other Palaeozoic ophiolites, implying that Silurian, Ordovician and Cambrian seawater had a (518O near 0%o (Gregory & Taylor 1981; Lecuyer & Fourcade 1991; Lecuyer et al 1995; Giguere et al 1998). On a longer time

Table 3. Measured isotopic ratios of epidosites from the SSOC and calculated black smoker fluid temperatures and compositions Sample no.

Oldra section 2

O-5 18 O-l 21 23 O-7 0-6 O-4 O-3 0-2 Strand section BH-60 BH-08 BH-56 BH-07

Depth below sedimentary contact (m)

Epidosite (WR) 618O (SMOW)

4 74 207 229 600 610 625 930 970

4.76 3.83 4.11 4.03 3.98 3.38 3.91

1700 1720

0.95

1 450 460 500

WR, whole rock; Ps, pistacite.

Quartz 618O (SMOW)

Epidote 618O (SMOW)

10.03

2.9

10.17

3.78

7.25 9.41 9.26 6.78 3.56

2.85 2.87 3.94 0.52 -0.29

13.62 9.36 11.87 8.61

8.74 3.29 5.56 2.14

D/H 6D (SMOW)

Water 618O (no Ps)

Water 618O (16% Ps)

°C (no Ps)

°C(16%Ps)

190

240

-2.1

0.5

-13

220

270

-0.3

1.9

-1

320 220 270 230 360

380 260 320 270 420

290 230 220 220

350 280 270 260

1.1 -1.4

1.1 -3.4 -1.4

6.4 -0.4

1.6 -2.1

2.6 0.9 3.0 -1.2 -0.1

8.1 1.7 3.8 0.2

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN

Fig. 4. Temperatures calculated from quartz-epidote 18 O fractionations v. depth in the Solund-Stavfjord ophiolite.

scale, the (518O of Proterozoic, Archaean ophiolites and the diamond-bearing eclogites all indicate that the (518O of seawater has not changed since the Archaean (not counting short-term changes as a result of continental ice build-up during ice ages) because it is 'buffered' to today's value by 18O exchange between the seawater and the sea floor (Muehlenbachs & Clayton 1976; Gregory & Taylor 1981; Muehlenbachs 1998). The argument that (518O of seawater has not changed over time has profound consequences. Sea-floor spreading, alteration of the oceanic crust and subduction have been operating from Archaean times onwards. However, Perry (1967) first suggested, based on the analyses of cherts, that the Archaean ocean was depleted of 18O and had evolved with time to its present value. More recently, Veizer et al. (1999) and many others, have on the basis of thousands of carbonate analyses, given — 8%o for the (518O of seawater in the Cambrian. Using such a low seawater <518O value and their brachiopod isotope record, Veizer et al. (2000) calculated a

411

Fig. 5. 618O of hydrothermal fluid v. depth in SolundStavfjord ophiolite. 618O of water calculated from 618O of quartz and the quartz—epidote temperature. Also shown are 6D of two epidotes.

secular surface seawater temperature trend inconsistent with accepted palaeoclimate studies and concluded that global climate was decoupled from atmospheric CC>2 content. Wallmann (2001) assumed a — 8%o Cambrian ocean as a starting parameter for modelling the water budget of the oceans and concluded that the oceans must be changing in volume. Both suggestions are interesting but are not supported by the invariant (518O record of seawater as inferred from ophiolites. The suggestions of decoupling of atmospheric COa and global climate as well as shrinking oceans follow directly from two assumptions: (1) that the massive brachiopod (518O record has preserved its isotopic integrity; (2) that it has recorded open ocean water (518O at approximately present-day temperatures. Analyses of ophiolites imply that the (518O of seawater has remained near 0%o over all geological time. The secular trends frequently (but not always) observed in low <518O of fossils and sediments may reflect a combination of (1) diagenetic or later re-equilibration of the

412

K. MUEHLENBACHS ETAL.

carbonates, etc. with low-18O pore waters, etc.; (2) higher sea surface temperatures than current prejudice allows; (3) biased sampling. The innumerable studies of Palaeozoic and older fossils and sediments all have sampled by necessity continental shelves and seas. These may be water bodies whose d18O is only tenuously linked to (518O of the deep open ocean, which, however, is recorded by ophiolites.

Conclusions Pillows, dykes and gabbros of the 443 Ma Sohmd-Stavfjord Ophiolite Complex have preserved the d 18 O of the original Ordovician sea floor in both the Oldra and Strand sections. The (318O profile in this Ordovician ophiolite is comparable with the profile observed in the 5.9 Ma ODP Hole 504B as well as that inferred from modern seafloor samples recovered by dredging and by submersibles. Cross-cutting epidosite veins in the Ordovician ophiolite show that the temperature and (518O of Ordovician black smoker fluids were the same as today's values. The new SSOC data are compatible with previous (518O studies on other ophiolites. All of these studies on ophiolites require that the (318O of Palaeozoic seawater was near 0%o. The style of sea-floor alteration has not changed since the Cambrian, nor is it likely to have differed at earlier times. The ophiolite data show no secular trend in the (518O of seawater, in contrast to secular (518O trends observed in most brachiopods and sediments. The <518O record of most sediments may not be recording the true (518O of seawater but may reflect isotopic resetting, higher ocean temperatures or biased sampling of restricted basins. One need not invoke the decoupling of atmospheric CO2 from global climate nor shrinking oceans to understand the <518O history of the oceans. Financial support for this study has been provided through grants from the Canadian NSERC (K.M.), University of Bergen, Norway (H.F.) and the Norwegian Research Council (H.F.), and is gratefully acknowledged. We further thank O. Levner for help with the isotopic analyses.

References AGRINIER, P., JAVOY, M. & GIRARDEAU, J. 1988. Hydrothermal activity in a peculiar oceanic ridge: oxygen and hydrogen isotope evidence in the Xigaze Ophiolite (Tibet, China). Chemical Geology, 71, 313-335. ALT, J.C., MUEHLENBACHS, K. & HONNOREZ, J. 1986. An oxygen isotopic profile through the upper kilometer of the oceanic crust, DSDP Hole 504B. Earth and Planetary Science Letters, 80, 217-229.

ALT, J.C., LA VERNE, C. & VANKO, D.A. ET AL. 1996. Hydrothermal alteration of a section of upper oceanic crust in the eastern equatorial Pacific: a synthesis of results from site 504 (DSDP Legs 69, 70, and 83, and ODP Legs 111, 137, 140 and 148). In: ALT, J.C., KINOSHITA, H., STOKKING, L.B. & MICHAEL, J.P. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 148. Ocean Drilling Program, College Station, TX, 417-434. ANDERSEN, T.B., SKJERLIE, K.P. & FURNES, H. 1990. The Sunnfjord Melange: evidence for Silurian ophiolite accretion in the West Norwegian Caledonides. Journal of the Geological Society, London, 147, 59-68. BANERJEE, N.R., GILLIS, K.M. & MUEHLENBACHS, K. 2000. Discovery of epidosites in a modern oceanic setting, the Tonga forearc. Geology, 28(2), 151-154. BARAGAR, W.R.A. 1984. Pillow formation and layered flows in the Circum-Superior Belt of eastern Hudson Bay. Canadian Journal of Earth Sciences, 21, 781-792. BARNICOAT, A.C. & CARTWRIGHT, I. 1997. The gabbroeclogite transformation: an oxygen isotope and petrographic study of west Alpine ophiolites. Journal of Metamorphic Geology, 15, 93-104. BARTH, M.G., RUDNICK, R.L., HORN, I., MCDONOUGH, W.F., SPICUZZA, M.J., VALLEY, J.W. & HAGGERTY, S.E. 2001. Geochemistry of xenolithic eclogites from West Africa, Part I: a link between low MgO eclogites and Archean crust formation. Geochimica et Cosmochimica Acta, 65, 1499-1527. BEARD, B.L., FRARACCI, K.N., TAYLOR, L.A., SNYDER, G.A., CLAYTON, R.A., MAYEDA, T.K. & SOBOLEV, N.V. 1996. Petrography and geochemistry of eclogites from the Mir kimberlite, Yakutia, Russia. Contributions to Mineralogy and Petrology, 125, 293-310. CARTWRIGHT, I. & BARNICOAT, A.C. 1999. Stable isotope geochemistry of Alpine ophiolites: a window to ocean-floor hydrothermal alteration and constraints on fluid-rock interaction during highpressure metamorphism. International Journal of Earth Sciences, 88, 219-235. CARTWRIGHT, A. & VALLEY, J.W. 1991. Low 18O Scourie dike magmas from the Lewisian complex, northwestern Scotland. Geology, 19, 578-581. CLAYTON, R.N. & MAYEDA, T.K. 1963. The use of bromine penta-fluoride in the extraction of oxygen from oxide and silicates for isotope analysis. Geochimica et Cosmochimica Acta, 27, 43-52. CLAYTON, R.N., O'NEIL, J.R. & MAYEDA, T.K. 1972. Oxygen isotope exchange between quartz and water. Journal of Geophysical Research, 77(17), 3057-3066. COCKER, J.D., GRIFFIN, B.J. & MUEHLENBACHS, K. 1982. Oxygen and carbon isotope evidence for seawater-hydrothermal alteration of the Macquarie Island ophiolite. Earth and Planetary Science Letters, 61, 112-122. COLEMAN, M., SHEPPARD, T.M., DURHAM, J.J., ROUSE, I.E. & MOORE, G.R. 1982. Reduction of water with zinc for hydrogen isotope analysis. Analytical Chemistry, 54(6), 993-995.

OPHIOLITES AS FAITHFUL RECORDS OF OXYGEN CRAIG, H. 1961. Standard for reporting concentrations of deuterium and oxygen-18 in natural waters. Science, 133, 1833-1834. DE WIT, MJ. & HART, R.A. 1993. Earth's earliest continental lithosphere, hydrothermal flux and crustal recycling. Lithos, 30, 309-335. DILEK, Y., FURNES, H. & SKJERLIE, K.P. 1997. Propagating rift tectonics of a Caledonian marginal basin: multi-stage seafloor spreading history of the Solund-Stavfjord ophiolite in Western Norway. Tectonophysics, 280, 213-238. DUNNING, G.R. & PEDERSEN, R.B. 1988. U/Pb ages of ophiolites and arc-related plutons of the Norwegian Caledonides: implications for the development of lapetus. Contributions to Mineralogy and Petrology, 98, 13-23. ELDERFIELD, H., WHEAT, C.G., MOTTL, M.J., MONNIN, C. & SPIRO, B. 1999. Fluid and geochemical transport through oceanic crust; a transect across the eastern flank of the Juan de Fuca Ridge. Earth and Planetary Science Letters, 172, 151-165. FURNES, H. 1972. Meta-hyaloclastite breccias associated with Ordovician pillow lavas in the Solund area, west Norway. Norsk Geologisk Tidsskrift, 52, 385-407. FURNES, H. 1974. Structural and metamorphic history of the Lower Palaeozoic metavolcanics and associated sediments in the Solund area, Sogn. Norges Geologiske Unders0kelse, 302, 33-74. FURNES, H., SKJERLIE, K.P. & PEDERSEN, R.B. ET AL. 1990. The Solund-Stavfjord Ophiolite Complex and associated rocks, west Norwegian Caledonides: geology, geochemistry and tectonic environment. Geological Magazine, 127, 209-289. FURNES, H., DILEK, Y., SKJERLIE, K.P., RYTTVAD, H.L. & FONNELAND, H.C. 1998. Variations in basaltic geochemistry along a propagating rift of the Late Ordovician basin of the West Norwegian Caledonides. Terra Nova, 10, 21-26. FURNES, H., SKJERLIE, K.P. & DILEK, Y. 2000. Petrology, tectonics, and hydrothermal alteration of a fossil backarc oceanic crust: Solund-Stavfjord ophiolite complex of the western Norwegian Caledonides. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 443-460. FURNES, H., HELLEVANG, B. & DILEK, Y. 2001. Cyclic volcanic stratigraphy in a Late Ordovician marginal basin, west Norwegian Caledonides. Bulletin of Volcanology, 63, 164-178. GARLICK, G.D., MACGREGOR, I.D. & VOGEL, D.E. 1971. Oxygen isotope ratios in eclogites from kimberlites. Science, 172, 1025-1027. GIGUERE, E., BERCLAZ, A., HEBERT, R. & BEAUDOIN, G. 1998. Circulation hydrothermale et metamorphisme dans la sequence crustale du massif de North Arm Moutain, Bay of Islands, Terre-Neuve. AGC/GAC-AMC/MAC-APGGQ, Abstract Volume, 23, A-64-A-65. GlLLIS,

K.M.,

MUEHLENBACHS,

K.,

STEWART,

M,

GLEESON, T. & KARSON, J. 2001. Fluid flow patterns in fast spreading East Pacific Rise crust

413

exposed at Hess Deep. Journal of Geophysical Research, 106(B11), 26311-26329. GODFREY, J.D. 1962. The deuterium content of hydrous minerals from the East-Central Sierra Nevada and Yosemite National Park. Geochimica et CosmochimicaActa,26, 1215-1245. GREGORY, R.T. & TAYLOR, H.P. 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: evidence for 618O buffering of the oceans by deep (>5 km) seawater hydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research, 84(B4), 2737-2755. HARPER, G.D., BOWMAN, J.R. & KUHNS, R. 1988. A field, chemical, and stable isotope study of subseafloor metamorphism of the Josephine ophiolite, California-Oregon. Journal of Geophysical Research, 93(B5), 4625-4656. HEATON, T.H.E. & SHEPPARD, S.M.F. 1977. Hydrogen and oxygen isotope evidence for seawater-hydrothermal alteration and ore deposition, Troodos complex, Cyprus. In: JONES, M.I. (ed.) Volcanic Processes in Ore Deposits. Geological Society, London, Special Publications, 7, 42-57. HOFFMAN, S.E., WILSON, M. & STAKES, D.S. 1986. Inferred oxygen isotope profile of Archaean oceanic crust, Onverwacht Group, South Africa. Nature, 321, 55-58. HOLMDEN, C.E. & MUEHLENBACHS, K. 1993a. The 18Q/16Q ratio of 2-billion-year-old seawater inferred from ancient oceanic crust. Science, 259, 1733-1736. HOLMDEN, C. & MUEHLENBACHS, K. 1993b. The origin of oxygen isotope variations in Paleozoic carbonates. Geological Association of Canada Annual Meeting, Program with Abstracts, A-45. JACOB, D., JAGOUTZ, E., LOWRY, D., MATTEY, D. & KUDRJAVTSEVA, G. 1994. Diamondiferous eclogites from Siberia: remnants of Archean oceanic crust. Geochimica et Cosmochimica Ada, 58(23), 5191-5207. JACOBSEN, S.B. & KAUFMAN, A.J. 1999. The Sr, C, and O isotopic evolution of Neoproterozoic seawater. Chemical Geology, 161, 37-57. JEAN-BAPTISTE, P., CHARLOU, J.L. & STIEVENARD, M. 1997. Oxygen isotope study of mid-ocean ridge hydrothermal fluids: implications for the oxygen-18 budget of the oceans. Geochimica et Cosmochimica Acta, 61, 2669-2677. LAWRENCE, J.R. & GIESKES, J.M. 1981. Constraints on water transport and alteration in the oceanic crust from the isotopic composition of pore water. Journal of Geophysical Research, 86(B9), 7924-7934. LECUYER, C. & FOURCADE, S. 1991. Oxygen isotope evidence for multi-stage hydrothermal alteration at a fossil slow-spreading center: the Silurian Trinity ophiolite (California, USA). Chemical Geology (Isotope Geoscience Section), 87, 231-246. LECUYER, C., GRANDJEAN, P. & MARTINEAU, F. 1995. Seawater-sediment-basalt interactions: stable isotope (H,O) and elemental fluxes within the Ordovidian volcano-sedimentary sequence of Erquy (Brittany, France). Contributions to Mineralogy and Petrology, 120, 249-264.

414

K. MUEHLENBACHS ETAL.

MACGREGOR, I.D. & MANTON, W.I. 1986. Roberts Victor eclogites: ancient oceanic crust. Journal of Geophysical Research, 91, 14063-14079. MATTHEWS, A. & SCHLIESTEDT, M. 1984. Evolution of the blueschist and greenschist facies rocks of Sifnos, Cyclades, Greece. A stable isotope study of subduction related metamorphism. Contributions to Mineralogy and Petrology, 88, 150-163. MILLER, J.A. & CARTWRIGHT, I. 2000a. Distinguishing between seafloor alteration and fluid flow during subduction using stable isotope geochemistry: examples from Tethyan ophiolites in the Western Alps. Journal of Metamorphic Geology, 18, 467-482. MILLER, J.A. & CARTWRIGHT, I. 2000b. Large-scale fluid flow patterns within the Corsican Ophiolite. Journal of Geochemical Exploration, 69-70, 297-301. MILLER, J.A., CARTWRIGHT, I., BUICK, I.S. & BARNICOAT, A.C. 2001. An O-isotope profile through the HP-LT Corsican ophiolite, France and its implications for fluid flow during subduction. Chemical Geology, 178, 43-69. MOORE, J.G. & SCHILLING, J.-G. 1973. Vesicles, water and sulfur in Reykjanes Ridge basalts. Contributions to Mineralogy and Petrology, 41, 105-118. MUEHLENBACHS, K. 1986. Alteration of the oceanic crust and the 18O history of seawater. In: VALLEY, J.W., TAYLOR, H.P.J. & O'NEIL, J.R. (eds) Stable Isotopes in High Temperature Geological Processes. Mineralogical Society of America, Reviews in Mineralogy, 16, 425-444. MUEHLENBACHS, K. 1998. The oxygen isotopic composition of the oceans, sediments and the seafloor. Chemical Geology, 145, 263-273. MUEHLENBACHS, K. & CLAYTON, R.N. 1972a. Oxygen isotope studies of fresh and weathered submarine basalts. Canadian Journal of Earth Science, 9, 172-184. MUEHLENBACHS, K. & CLAYTON, R.N. 1972b. Oxygen isotope geochemistry of submarine greenstones. Canadian Journal of Earth Science, 9, 471-478. MUEHLENBACHS, K. & CLAYTON, R.N. 1976. Oxygen isotope composition of the oceanic crust and its bearing on seawater. Journal of Geophysical Research, 81(23), 4365-4369. NEAL, C.R., TAYLOR, L.A. & DAVIDSON, J.P. ET AL. 1990. Eclogites with oceanic crustal and mantle signatures from the Bellsbank kimberlite, South Africa, part 2: Sr, Nd, and O isotope geochemistry. Earth and Planetary Science Letters, 99, 362-379. ONGLEY, J.S., BASU, A.R. & KYSER, T.K. 1987. Oxygen isotopes in coexisting garnets, clinopyroxenes and phlogopites of Robert Victor eclogites: implications

for petrogenesis and mantle metasomatism. Earth and Planetary Science Letters, 83, 80-84. PEDERSEN, R.B. 1986. The nature and significance of magma chamber margins in ophiolites: examples from the Norwegian Caledonides. Earth and Planetary Science Letters, 11, 100-112. PERRY, E.G. JR 1967. The oxygen isotope chemistry of ancient cherts. Earth and Planetary Science Letters, 3, 62-66. PUTLITZ, B., MATTHEWS, A. & VALLEY, J.W. 2000. Oxygen and hydrogen isotope study of high pressure metagabbros and metabasalts (Cyclades, Greece): implications for the subduction of oceanic crust. Contributions to Mineralogy and Petrology, 138, 114-126. RYTTVAD, H.L., FURNES, H., SKJERLIE, K.P. & ROLFSEN, R. 2000. Geochemistry and petrogenesis of extrusive rocks, dykes and high-level plutonic rocks on the island of Oldra, Solund-Stavfjord Ophiolite Complex, western Norway. Norsk Geologisk Tidss&n/f, 80, 97-110. SCHIFFMAN, P. & SMITH, B.M. 1988. Petrology and oxygen isotope geochemistry of a fossil seawater hydrothermal system within the Solea Graben, Northern Troodos Ophiolite, Cyprus. Journal of Geophysical Research, 93(B5), 4612-4624. SKJERLIE, K.P. & FURNES, H. 1990. Evidence for a fossil transform fault in the Solund/Stavfjord Ophiolite Complex: West Norwegian Caledonides. Tectonics, 9, 1631-1648. SKJERLIE, K.P. & FURNES, H. 1996. The gabbro-dyke zone demonstrated on Tviberg, Solund-Stavfjord Ophiolite Complex. Geological Magazine, 133, 573-582. SNYDER, G.A., TAYLOR, L.A. & JERDE, E.A. ET AL. 1995. Archean mantle heterogeneity and the origin of diamondiferous eclogites, Siberia: evidence from stable isotopes and hydroxyl in garnet. American Mineralogist, 80, 799-809. STAKES, D.S. & TAYLOR, H.P. JR 1992. The northern Samail ophiolite: an oxygen isotope, microprobe, and field study. Journal of Geophysical Research, 97, 7043-7080. VEIZER, J., ALA, D. & AZMY, K. ET AL. 1999. 87Sr/86Sr, 613C and 618O evolution of Phanerozoic seawater. Chemical Geology, 161, 59-88. VEIZER, J., GODDERIS, Y. & FRANCOIS, L.M. 2000. Evidence for decoupling of atmospheric CO2 and global climate during the Phanerozoic eon. Nature, 408,698-701. WALLMANN, K. 2001. The geological water cycle and the evolution of marine 618O values. Geochimica et Cosmochimica Acta, 65(15), 2469-2485.

Bioalteration recorded in ophiolitic pillow lavas H. FURNES 1 & K. MUEHLENBACHS 2 1 Department of Earth Science, University of Bergen, Allegt. 41, 5007 Bergen, Norway (e-mail: Harald.Furnes@geo. uib. no) 2 Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB T6G 2E3, Canada Abstract: In this paper we summarize the present knowledge on bioalteration of basaltic glass from pillow lava rims of former oceanic crust, based on the study of four ophiolite complexes. These complexes range in age from Late Cretaceous to Mid-Proterozoic, and their metamorphic grades vary from non-metamorphosed to low greenschist- to low amphibolite-facies metamorphism. In the non-metamorphosed volcanic part of ophiolite complexes, in which undevitrified glass is still present in pillow lava rims and/or hyaloclastites, biogenerated textures are common. These textures (termed granular and tubular) are similar to those found in volcanic glass of recent to old (170 Ma) in situ ocean floor, and mimic microbes in terms of size and shape. In the metamorphic and completely recrystallized examples, the textural evidence of bioalteration is generally obliterated, although it may still be visible in littledeformed volcanic domains of low-grade greenschist-facies metamorphism. Where biogenerated textures are present, element mapping invariably reveals the presence of organic carbon, and sometimes nitrogen, sulphur and phosphorus. Carbon-isotope signatures ((513C) in the pillow lava rims from all four investigated ophiolites show lower values than those of the adjacent crystalline parts. This phenomenon may be attributed to bio-induced fractionation of carbon isotopes during preferential bioalteration of the pillow lava glass, and may further represent a feature that seems to survive metamorphism and deformation.

Basaltic glass, constituting a significant part of the upper oceanic crust, is an unstable component when exposed to water. Alteration of basaltic glass yields a pale yellow to dark brown material known as the multi-component palagonite, commonly divided into two types, gel-palagonite and fibropalagonite. Gel-palagonite is an isotropic, amorphous material, whereas fibro-palagonite is a crystalline material consisting of authigenic minerals such as zeolites and clays (e.g. Hay & lijima 1968a, 1968b; Zhou et al 1992). The palagonitization process takes place at low to high temperature (Jakobsson 1978; Zierenberg et al. 1995), and has traditionally been viewed as a purely chemical-physical process (e.g. Hay & lijima 1968a, 1968b; Zhou & Fyfe 1989; Thorseth et al 1991; Daux et al. 1994; Schiffman et al. 2000). However, within the temperature range allowing for microbial existence, glass is also subject to bioalteration (e.g. Thorseth et al. 1992, 1995, 2001; Furnes et al. 2001d). Bioalteration of basaltic glass from the upper in situ oceanic crust is a process that is now well documented. Independent documentation is provided by the following features: (1) alteration textures; (2) the presence of DNA within alteration

textures; (3) carbon and nitrogen within alteration zones; (4) carbon isotopes (generally very low 613C) in disseminated carbonate (Thorseth et al. 1995; Furnes et al. 1996, 1999, 2001b; Fisk et al. 1998; Torsvik et al. 1998); (5) identification of microbes inhabiting altered glassy margins of pillow lava (Thorseth et al. 2001). Based on the consideration of potassium and C-isotope signatures in altered glass of the deep Hole 504B at the Costa Rica Rift, Furnes et al. (1999) proposed that the present bioalteration takes place down to a depth of about 380 m into the volcanic basement. Staudigel et al. (2003) suggested that large portions of hydrothermal systems in the oceanic crust can be considered a giant bioreactor that mediates water-rock exchange and buffers the chemical composition of seawater. Bio-mediated alteration textures can be identified as such in alteration fronts into fresh glass, along cracks or from the outer pillow margins. These textures appear not to be preserved, or at least are not easily recognizable when the bio-altered glass is further converted to crystalline matter (e.g. birefringent palagonite) or metamorphic minerals (i.e. chlorite, amphibole, epidote, etc.). Attempts to quantify the significance of bioalteration relative to abiotic

From'. DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History, Geological Society, London, Special Publications, 218, 415^26. 0305-8719/037$ 15 © The Geological Society of London 2003.

416

H. FURNES & K. MUEHLENBACHS

alteration of basaltic glass throughout the volcanic basement have indicated that in the upper c. 300 m of the volcanic sequence bioalteration is significant to dominant, and the effects of bioalteration have been traced down to 550 m (Furnes & Staudigel 1999;Furnes et al. 200Id), thus giving evidence for a deep oceanic biosphere. Little is known about bioalteration of ophiolites, despite the fact that only in such sequences is it possible to trace this process back in time, far beyond that of the oldest (c. 170 Ma) in situ oceanic crust (Furnes et al. 2002b). So far, only limited information on this topic is available from two ophiolite sequences, i.e. the 90 Ma Troodos ophiolite complex (TOC) in Cyprus (Furnes et al. 200 Ic), and the 443 Ma Somnd-Stavfjord ophiolite complex (SSOC) in West Norway (Furnes et al. 2002a). In this paper we review the present knowledge of bioalteration in the two above-mentioned ophiolites, and also present some unpublished data from two other ophiolite complexes, i.e. the 160 Ma Mirdita ophiolite complex (MOC) in Albania, and the 1.95Ga Jormua ophiolite complex (JOC) in Finland. In all these ophiolite complexes traces of bioalteration are present, thus providing documentation of the bioalteration process in former oceanic crust over a substantial time period, i.e. Late Cretaeous to Mid-Proterozoic.

General geology of the ophiolite complexes The Troodos ophiolite complex (TOC) The Late Cretaceous (90 Ma) Troodos ophiolite of Cyprus contains all the components of a complete ophiolite (e.g.Malpas et al. 1990). More than half of the volcanic rocks are pillow lavas, the remainder being breccias associated with pillows, and sheet flows (Schmincke & Bednarz 1990). Fresh glass occurs throughout the volcanic sequence, in which various types of biogenerated features may exist. Most of the samples were collected from the well-exposed, >600m thick Akaki river section and the drill core CY-1 (for further details, see Furnes et al. 200Ic). A total of 58 samples were investigated microscopically, and of these 14 samples were analysed for C and O isotopes (see Furnes et al. 200Ic), seven were investigated by scanning electron microscopy and four were X-ray mapped.

The Mirdita ophiolite complex (MOC) All the samples, for the purpose of bioalteration studies of the Late Jurassic (c. 160 Ma) Mirdita ophiolite complex, were collected from pillow lava rinds from the western part of the volcanic

sequence, which exhibits a typical mid-ocean ridge basalt (MORE) geochemistry (Shallo 1995; Bebien et al. 1998). The volcanic sequence, resting on plutonic rocks of the ophiolite, or in part the mantle sequence, is about 700 m thick from bottom to the overlying cover of cherts and shales. The samples are equally distributed throughout the volcanic sequence from bottom to top. At present, 48 samples of pillow rims have been microscopically investigated. Most of the samples are completely recrystallized to low-grade authigenic minerals; only in a few samples are there remnants of fresh glass, which make biogenerated textures visible. In the few samples containing fresh glass, however, bioalteration textures are common. Thirty-four samples have been analysed for C and O isotopes, of which 18 represent original glass and 16 are crystalline samples, taken adjacent to the pillow rims. Data collection is still in progress, and the complete data set will be published elsewhere.

The Solund-Stavfjord ophiolite complex (SSOC) The samples were collected from the least deformed part of the Late Ordovician (443 ± 3 Ma) Solund-Stavfjord ophiolite complex (SSOC) in the western Norwegian Caledonides (see Furnes et al. 200la, and references therein). The SSOC displays a well-preserved volcanic sequence of basaltic pillow lavas, sheet flows and volcanic breccias, a sheeted dyke complex, and high-level gabbro. The investigated samples are from pillow lavas of the uppermost 290 m of the volcanic sequence (Furnes et al. 2002a). The volcanic rocks have suffered pervasive lower greenschist-facies metamorphism (chlorite, actinolite and epidote as the principal minerals), but the collected rocks are very little deformed. Seventeen samples were collected at various stratigraphic levels. For isotope analyses the samples were split into 18 glassy (chilled margin of pillows) and 15 crystalline samples (material adjacent to, and within the chilled margin), and two samples were X-ray mapped (Furnes et al. 2002a).

The Jormua ophiolite complex (JOC) The JOC contains all the principal components of an ophiolite, i.e. pillow lavas and volcanic breccias, a sheeted dyke complex, plutonic rocks, and mantle peridotites (Kontinen 1987). The age of the complex has been determined to be 1.95 Ga (Peltonen et al. 1996), and it may thus represent one of the oldest examples of crust formed by ocean-floor spreading in which all components of

BIOALTERATION IN OPHIOLITES an ophiolite are present. The thickness of the volcanic sequence varies, but appears to be at least 500 m at the thickest. Forty-six samples were collected. The original glassy samples represent rocks from pillow margins, fragments and matrix of hyaloclastite breccias, as well as material from the interior of pillows and dyke rocks. The volcanic rocks and dykes have been subjected to lower amphibolite-facies metamorphism, and the mineral assemblage is pale green amphibole, oligoclase-andesine plagioclase, epidote and chlorite (Kontinen 1987). The volcanic rocks also display a pronounced foliation. Hence, in none of the samples has it been possible to find preserved biotextures. For C isotopes five crystalline and 41 glassy samples (38 from the volcanic rocks and three from the chilled margin of dykes) were analysed. Data collection is still in progress, and the complete data set will be published elsewhere.

Evidence of bioalteration

417

Thus, colonizing microbes cause pitting of glass surfaces and finally the formation of granular and tubular structures. Along fractures within the originally glassy parts of the investigated metamorphic pillows from the SSOC there are textural features that strongly resemble biogenerated textures adjacent to fractures in fresh, non-metamorphic pillows (Fig. 3). These features appear as sphene-rich aggregates developed in a symmetric or asymmetric arrangement around fractures, with an irregular front against the original glass, now replaced by chlorite. The asymmetric development of alteration products around fractures, as well as the irregular contact with the fresh glass, are typical textural features related to biogenerated alteration in the quenched zone of pillows of in situ oceanic crust (Furnes et al. 1996). In the amphibolite-facies metamorphosed volcanic rocks of the JOC no textural traces of bioalteration can be traced.

Biogenerated textures

Organic remains

In samples still containing fresh glass, textures of anticipated biogenerated origin occur throughout the lava sequence of the TOC and MOC. These textures occur at the alteration front, defined as the boundary between fresh and altered glass. The most common type represents isolated and/or persistent zones consisting of coalesced spherical bodies about 1-3 urn in diameter, referred to by Furnes et al (200Id) as 'granular' texture (Fig. la and b). Associated with the granular textures of the altered glass of the TOC are abundant straight and curved thin (c. 2-5 urn) tubes that may attain lengths of up to 100 urn (Fig. 2a), referred to as 'tubular' texture (Furnes et aL 200Id). Another, less abundant type of tubular bodies may reach lengths up to 500 urn, diameters up to 20 urn, and show well-defined segmentation into plates 510 urn thick (Fig. 2b), and a rare type of tubular texture forms spirals (Fig. 2c). In the altered basaltic glass of the MOC are tubular features associated with zeolite occurrence (Fig. 2d). An explanation of the formation of secondary textures observed in basaltic glass, which resemble the granular and tubular textures shown in Figures 1 and 2, was first proposed by Thorseth et aL (1992). They argued that metabolic activity of colonizing microbes might cause localized changes in pH. These pH changes may be mostly to increase the acidity (down to pH 2) but they may also increase the pH to more alkaline values (up to 10) (Golubic 1973; Krumbein et aL 1991). Silicate glasses display a solubility minimum at near-neutral pH, and therefore increases or decreases in pH will result in enhanced dissolution.

Features that appear to represent organic remains have so far been found only within granular textures, at the alteration front in the glassy rims of pillow lavas from the TOC (Furnes et aL 200 Ic). These features occur as long (at least 6 urn) and c. 50-200 nm wide, branching filaments, occasionally twisted, which in many cases can be seen to be attached to the fresh glass at the alteration front (Fig. 4). The filaments are partly buried by spherical and/or elliptical bodies c. 50200 nm in diameter (Fig. 4). The presence of filaments coincides with high C content, as revealed by element mapping (see below).

Element mapping Chemical evidence for the interaction between microbes and basaltic glass includes the typical occurrence of C and N associated with the granular and tubular textures (Furnes et aL 200Id). For the purpose of demonstrating the effect of bioalteration of ophiolitic pillow lava rinds, we emphasize typical bio-elements such as C, N, P and S. At the time of writing, only two of the ophiolite complexes (TOC and SSOC) have been investigated by element mapping, and some of these results have been published (Furnes et aL 200Ic, 2002a). Figure 5a shows alteration of basaltic glass (from the TOC) around a fracture, of which the alteration front exhibits typical granular texture. Within the altered area, and in particular the alteration front, C and N are enriched (Fig. 5b and c), whereas P and S are depleted (not shown), relative to the fresh glass.

418

H. FURNES & K. MUEHLENBACHS

Fig. 1. Granular biogenerated textures: (a) from TOC; (b) from MOC. FG, fresh glass; GT, granular texture.

Along the alteration front, Ca is strongly depleted within the area of C enrichment (Fig. 5d), thus demonstrating the absence of calcite but suggesting the presence of organic C. In the top right in

Figure 5b and d, however, there is enrichment of C and Ca, respectively, probably showing the presence of a minor amount of calcite. Mg and Fe (not shown) are slightly depleted. Also, in the

BIOALTERATION IN OPHIOLITES

419

Fig. 2. Tubular textures of bio-origin: (a) abundant thin tubes; (b) a spiral structure; (c) large, segmented tubes. The tubular textures shown in (a)- (c) (from TOC) all penetrate fresh glass, (d) Tubular textures (from MOC), hosted in a SiO2-rich, A^Os-, Na2O- and K^O-poor zeolite (most probably stilbite). FG, fresh glass; T, tubes; ST, segmented tube; SP, spiral; Ze,zeolite.

metabasaltic glass of the SSOC pillow lavas, the presence of probable organic C can be demonstrated, in this case generally associated with high Fe and S (Fig. 6). In altered pillow lavas from the Costa Rica Rift, Alt & Mata (2000) documented secondary sulphides, for which they suggested a microbial origin. The association of C, Fe, S and P in the metamorphic material presented in this paper (Fig. 6) may indicate a similar origin. The presence of probable organic carbon further supports the supposition that bioalteration took place in the glassy rim of pillows. The preservation of the organic remains is probably due to their enclosure by crystals (McKinley et al. 2000), and/or because they became part of the crystalline structure (Collins et al. 1995; van Lith et al. 2001).

Carbon isotopes In previous studies the usefulness of carbon isotopes of extracted carbonate from bioaltered basaltic glass from in situ ocean-floor pillow lavas has been demonstrated (Thorseth et al. 1995;

Furnes et al. 1999, 2001b). In general, the (513C signatures of altered basaltic glass rims of pillows are significantly lower than those of the adjacent crystalline basalt (Furnes et al. 2002b). This phenomenon has been attributed to microbialprocessed glass, in which oxidation of organic matter (with already low (513C) leads to 12C enriched CO2 (Thorseth et al. 1995; Furnes et al. 200 Ib). In Figure 7 we present the (513C data from the original glassy rims and adjacent crystalline material of pillow lava from the four ophiolite complexes introduced above, ranging in age from 92 Ma to 1953 Ma. Although there is a predominance of analyses of glassy material, Figure 7 shows that this material defines a wider range in their (313C values, and in particular they are shifted towards lower (513C values, compared with the crystalline samples. This is particularly well demonstrated when all the ophiolitic samples are plotted together (Fig. 8a), and they show a similar pattern to those of in situ ocean-floor pillow lavas (Fig. 8b). The modes of the (513C values of the ophiolitic samples are —10.63 and —3.47 for the

420

H. FURNES & K. MUEHLENBACHS

Fig. 3. Backscattered electron scanning electron microscope image, showing inferred biogenerated texture (light grey sphene (Sph)), developed adjacent to fractures (Fr) in a dark chloritic (Chi) matrix (originally glass) in pillow lava from the SSOC. It is particularly the irregular boundary between sphene and chlorite that resembles the alteration front as seen in in situ, bioaltered pillow lava glass of the oceanic crust. Chlorite and sphene were identified by the use of optical microscopy and element maps of Ti, Ca, Mg and Si (not shown).

Fig. 4. Scanning electron microscope images showing organic-like remains (biofilm and filaments) (BF) within areas of typical granular textures with high carbon content. The organic remains are in places attached to the fresh glass (FG), and they are partly buried by bean- to spherical-shaped material. Both images are from the glassy rim of pillow lava samples of the TOC.

BIOALTERATION IN OPHIOLITES

421

Fig. 5. Scanning electron microscope image and X-ray maps that show the distribution and concentrations of C, N and Ca from a bioaltered area of a pillow lava rim from the TOC. Concentration of elements increases from dark blue to light blue. The boxed area in (a) shows the position of Figure 4a. FG, fresh glass; AG, altered glass.

glassy and crystalline samples, respectively. The difference of c. 7%o we regard as significant, and the large range of the data probably reflects different processes such as mixing of various carbon sources and oxidation of low-(513C organic carbon.

Perspectives of investigating bioalteration in ophiolites The oldest fossil micro-organisms have so far been found in sedimentary rocks. Some of the best preserved fossil micro-organisms come from the 3.3-3.5 Ga sedimentary rocks of the Barberton

greenstone belt (Westall et al 2001). The earliest indications of life on Earth come from carbonisotope signatures of organic carbon in Lower Archaean supracrustal rocks in Greenland, and it was suggested that fractionation of carbon by living organisms occurred prior to 3.8 Ga (Schidlowski 1988; Mojzsis et al. 1996). However, the birthplace of life may have been connected to volcanic environments, such as deep-sea hydrothermal vents (Russel & Hall 1997). Thus, if microbes in the earliest times also utilized mafic glass as a possible source of energy and nutrition, leaving bio-signatures behind, as demonstrated for the present-day volcanic glass (e.g. Thorseth et al. 1995; Furnes et al. 200 Ic), one might expect to

422

H. FURNES & K. MUEHLENBACHS

Fig. 6. Scanning electron microscope image and X-ray maps that show the distribution and concentrations of C, Fe, S, P and Ca from a bioaltered area of a pillow lava rim from the SSOC. Concentration of elements increases from dark blue to light blue to green to yellow to red to white.

BIOALTERATION IN OPHIOLITES

423

Fig. 7. Distribution of <513C in glassy and crystalline pillow lavas from the TOC, Cyprus (data from Furnes et al. 200Ic), MOC, Albania (unpublished data by H. Furnes), SSOC, West Norway (Furnes et al 2003a), and JOC, Finland (unpublished data). A more comprehensive description and discussion of the bioalteration history of the MOC and JOC will be completed when all available data are available, and published elsewhere.

find traces of such signals in the oldest volcanic rocks. The significance of basaltic glass as a substrate for microbes is still poorly known. Thorseth et al. (2001) showed geochemical evidence for iron oxidation of altered basaltic glass from the Knipovich Ridge, in which Gallionellalike stalks indicate the presence of iron-oxidizing bacteria, and ferrous iron as an energy source. The carbon signatures of the originally glassy volcanic rocks of the 1.95Ga JOC, showing a modal <513C value of —8.9, may also suggest that

isotopic fractionation took place as a result of bioalteration (Fig. 7). However, with only five analyses of the crystalline samples, compared with 41 analyses of the glassy samples of the JOC, no statistical distinction can yet be made between the data for the two types. The same statement also applies to the samples of the TOC. An important task is therefore to find out to what extent it is possible to back-trace the evidence of bioalteration of basaltic glass in the oldest oceanic crust, or any other volcanic rock. Candidates for such a

424

H. FURNES & K. MUEHLENBACHS

Fig. 8. (a) Distribution of all (513C data from the ophiolites as shown in Figure 7. (b) shows the distribution of all d 13 C data from pillow lavas from in situ ocean floor from the Atlantic and Pacific oceans (data from Furnes et al. 1999, 2001b).

study are probably abundant within the greenstone belts, of which one of the best examples may be the 3.5 Ga Jamestown ophiolite complex of the Barberton greenstone belt of South Africa (de Wit et al. 1987). Thus, if microbes have always played the same important role in the alteration of basaltic glass of the oceanic crust as that proposed for the present in situ oceanic crust (Furnes et al. 200Id), the scale of bioalteration in time and space is vast.

Conclusions The textures, organic remains, distribution of bioelements (C, N, P, S) and carbon-isotope signatures within altered pillow lava rims of ophiolitic

pillows all point to the effect of bioalteration. In the non- to low-metamorphosed TOC and MOC, biogenerated textures (granular and tubular) are similar to those found in in situ ocean-floor pillow lavas. Even in the low greenschist-facies metamorphic pillow lavas of the SSOC, textures reminiscent of bioalteration can be found. Also, as far as our limited information on the distribution of organic carbon goes, it appears that this element may survive at least low-grade greenschist-facies metamorphism, and thus give valuable bioalteration information. At the present state of knowledge of bio-signals in the alteration of ophiolitic pillow lavas, indicators of the original (513C signatures of extracted calcite seem to be present, despite age (Late Cretaceous to Mid-

BIOALTERATION IN OPHIOLITES Proterozoic) and metamorphic grade (up to lower amphibolite-facies metamorphism). The studies that have led to the conclusions of this work have been funded by grants (to H.F.) from the Research Council of Norway (grant 110833/40 and the Strategic University Program SUBMAR), from the University of Bergen (Meltzer H0yskolefond), and an Environmental and Earth Science and Technology Collaborative Linkage Grant (EST.CLG.97617) from NATO in support of the research in Albania. K.M. also acknowledges support from the NSERC of Canada. The reviewers (J. Alt and M. Fisk) and Y. Dilek made useful comments on an early version of the manuscript. J. Ellingsen kindly helped with the illustrations.

References ALT, J.C. & MATA, P. 2000. On the role of microbes in the alteration of submarine basaltic glass: a TEM study. Earth and Planetary Science Letters, 181, 301-313. BEBIEN, J., SHALLO, M., MANIA, K. & GEGA, D. 1998. The Shebenik Massif (Albania): a link between MOR- and SSZ-type ophiolites? Ofioliti, 23(1), 7-15. COLLINS, M.J., BISHOP, A.N. & FARRIMOND, P. 1995. Sorption by mineral surfaces: rebirth of the classical condensation pathway for kerogene formation? Geochimica et Cosmochimica Acta, 59, 2387-2391. DAUX, V., CROVISIER, J.L., HEMOND, C. & PETIT, J.C. 1994. Geochemical evolution of basaltic rocks subjected to weathering: fate of the major elements, rare earth elements, and thorium. Geochimica et Cosmochimica Acta, 58, 4941-4954. DE WIT, M.J., HART, R.A. & HART, R.J. 1987. The Jamestown Ophiolite Complex, Barberton mountain belt: a section through 3.5 Ga oceanic crust. Journal of African Earth Sciences, 6, 681-730. FISK, M.R., GIOVANNONI, S.J. & THORSETH, I.H. 1998. Alteration of oceanic volcanic glass, textural evidence of microbial activity. Science, 281, 978-979. FURNES, H. & STAUDIGEL, H. 1999. Biological mediation in ocean crust alteration: how deep is the deep biosphere? Earth and Planetary Science Letters, 166, 97-103. FURNES, H., THORSETH, I.H., TUMYR, O., TORSVIK, T. & FISK, M.R. 1996. Microbial activity in the alteration of glass from pillow lavas from Hole 896A. In: ALT, J.C., KINOSHITA, H., STOKKING, L.B. & MICHAEL, J.P. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 148. Ocean Drilling Program, College Station, TX, 191-206. FURNES, H., MUEHLENBACHS, K., TUMYR, O., TORSVIK, T. & THORSETH, I.H. 1999. Depth of active bioalteration in the ocean crust: Costa Rica Rift (Hole 504B). Terra Nova, 11, 228-233. FURNES, H., HELLEVANG, B. & DILEK, Y. 200la. Cyclic volcanic stratigraphy in a Late Ordovician marginal basin, west Norwegian Caledonides. Bulletin of Volcanology, 63, 164-178. FURNES, H., MUEHLENBACHS, K., TORSVIK, T., THORSETH, I.H. & TUMYR, O. 2001b. Microbial fractio-

425

nation of carbon isotopes in altered basaltic glass from the Atlantic Ocean, Lau Basin and Costa Rica Rift. Chemical Geology, 173, 313-330. FURNES, H., MUEHLENBACHS, K., TUMYR, O., TORSVIK, T. & XENOPHONTOS, C. 200Ic. Biogenic alteration of volcanic glass from the Troodos ophiolite, Cyprus. Journal of the Geological Society, London, 158, 75-84. FURNES, H., STAUDIGEL, H., THORSETH, I.H., TORSVIK, T., MUEHLENBACHS, K. & TUMYR, O. 200Id. Bioalteration of basaltic glass in the oceanic crust. Geochemistry, Geophysics, Geosystems, 2, Paper Number 2000GCOOO150. FURNES, H., MUEHLENBACHS, K., TORSVIK, T., TUMYR, O. & SHI, L. 2002a. Bio-signatures in metabasaltic glass of a Caledonian ophiolite, West Norway. Geological Magazine, 139, 601-608. FURNES, H., THORSETH, I.H., TORSVIK, T., MUEHLENBACHS, K., STAUDIGEL, H. & TUMYR, O. 2003b. Identifying bio-interaction with basaltic glass in oceanic crust and implications for estimating the depth of the oceanic biosphere: a review. In: SMELLIE, J.L. & CHAPMAN, M.G. (eds) Volcano-Ice Interaction on Earth and Mars. Geological Society, London, Special Publications, 202, 407-421. GOLUBIC, S. 1973. The relationship between blue-green algae and carbonate deposits. In: CARR, N.G. & WHITTON, B.A. (eds) The Biology of Blue-Green Algae. Blackwell, Oxford, 434-472. HAY, R.L. & IIJIMA, A. 1968a. Nature and origin of palagonite tuffs of the Honolulu Group on Oahu, Hawaii. Geological Society of America, Memoir, 116, 338-376. HAY, R.L. & IIJIMA, A. 1968b. Petrology of palagonite tuffs of the Koko Craters, Oahu, Hawaii. Contributions to Mineralogy and Petrology, 17, 141-154. JAKOBSSON, S.P. 1978. Environmental factors controlling the palagonitization of the Surtsey tephra, Iceland. Bulletin of the Geological Society of Denmark, 27, 91-105. KONTINEN, A. 1987. An early Proterozoic ophiolite— The Jormua mafic-ultramafic complex, northeastern Finland. Precambrian Research, 35, 313-341. KRUMBEIN, W.E., URZI, C.E. & GEHRMANN, C. 1991. Biocorrosion and biodeterioration of antique and medieval glass. Geomicrobiology Journal, 9, 139-160. MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) 1990. Ophiolites: Oceanic Crustal Analogues. Proceeding of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 783 pp. MCKINLEY, J.P., STEVENS, T.O. & WESTAL, F. 2000. Microfossils and paleoenvironment in deep subsurface basalt samples. Geomicrobiology Journal, 17, 43-54. MOJZSIS, S.J., ARRHENIUS, G., MCKEEGAN, K.D., HARRISON, T.M., NUTMAN, A.P. & FRIEND, C.R.L. 1996. Evidence for life on Earth before 3800 million years ago. Nature, 384, 55-59. PELTONEN, P., KONTINEN, A. & HUHMA, H. 1996. Petrology and geochemistry of metabasalts from the 1.95Ga Jormua ophiolite, Northeastern Finland. Journal of Petrology, 37, 1359-1383.

426

H. FURNES & K. MUEHLENBACHS

RUSSEL, MJ. & HALL, A.J. 1997. The emergence of life from iron monosulphide bubbles at a submarine hydrothermal redox and pH front. Journal of the Geological Society, London, 154, 377-402. SCHIDLOWSKI, M. 1988. A 3800 million-year isotopic record of life from carbon in sedimentary rocks. Nature, 333, 313-318. SCHIFFMAN, P., SPERO, H.J., SOUTHARD, RJ. & SWANSON, D.A. 2000. Controls on palagonitization versus pedogenic weathering of basaltic tephra: evidence from the consolidation and geochemistry of the Keanakako'i Ash Member, Kilauea Volcano. Geochemistry, Geophysics, Geosystems, 1, Paper Number 2000GC000068. SCHMINCKE, H.-U. & BEDNARZ, U. 1990. Pillow, sheet flow and breccia flow volcanoes and volcano-tectonic hydrothermal cycles in the Extrusive Series of the northeastern Troodos ophiolite (Cyprus). In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites Oceanic Crustal Analogues, Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 185-206. SHALLO, M. 1995. Volcanics and sheeted dykes of the Albanian SSZ ophiolite. Buletini i Shkencave Gjeologjike, 91, 99-118. STAUDIGEL, H., FURNES, H., KELLEY, K., PLANK, T., MUEHLENBACHS, K., TEBO, B. & YAYANOS, A. 2003. The oceanic crust as a bioreactor? In: EDITOR, A. (ed.) Deep Subsurface Biosphere at Mid-Ocean Ridges. American Geophysical Union, Monograph, in press. THORSETH, I.H., FURNES, H. & TUMYR, O. 1991. A textural and chemical study of Icelandic palagonite of varied composition and its bearing on the mechanism of the glass-palagonite transformation. Geochimica et Cosmochimica Acta, 55, 731-749. THORSETH, I.H., FURNES, H. & HELDAL, H. 1992. The importance of microbiological activity in the alteration of natural basaltic glass. Geochimica et Cosmo-

chimica Acta, 56, 845-850. THORSETH, I.H., TORSVIK, T., FURNES, H. & MUEHLENBACHS, K. 1995. Microbes play an important role in the alteration of oceanic crust. Chemical Geology, 126, 137-146. THORSETH, I.H., TORSVIK, T., TORSVIK, V., DAAE, F.L., PEDERSEN, R.B. & Keldysh-98 Scientific Party 2001. Diversity of life in ocean floor basalt. Earth and Planetary Science Letters, 194, 31-37. TORSVIK, T., FURNES, H., MUEHLENBACHS, K., THORSETH, LH. & TUMYR, O. 1998. Evidence for microbial activity at the glass-alteration interface in oceanic basalts. Earth and Planetary Science Letters, 162, 165-176. VAN LITH, Y., VASCONCEIOS, C., WARTHMANN, R. & MCKENZIE, J.A. 2001. Organic matter trapped in sedimentary dolomite: a potential tracer to recognize microbial dolomite. European Union of Geoscience XI, Journal of Conference Abstracts, 56. WESTALL, F., DE WIT, M.J., DANN, J., VAN DER GAAST, S., DE RONDE, C.E.J. & GERNEKE, D. 2001. Early Archean fossil bacteria and biofilms in hydrothermally-influenced sediments from the Barberton greenstone belt, South Africa. Precambrian Research, 106,93-116. ZHOU, Z. & FYFE, W.S. 1989. Palagonitization of basaltic glass from DSDP Site-335, Leg-37—textures, chemical composition, and mechanism of formation. American Mineralogist, 74(9-10), 1045-1053. ZHOU, Z., FYFE, W.S., TAZAKI, K. & VAN DER GAAST, S.J. 1992. The structural characteristics of palagonite from DSDP Site 335. Canadian Mineralogist, 30,75-81. ZlERENBERG, R.A., SCHIFFMAN, P., JONASSON, I.R., TOSDAL, R., PlCKTHORN, W. & McCLAIN, J. 1995.

Alteration of basalt hyaloclastite at the off-axis Sea Cliff hydrothermal field, Gorda Ridge. Chemical Geology, 126(2), 77-99.

What constitutes 'emplacement' of an ophiolite?: Mechanisms and relationship to subduction initiation and formation of metamorphic soles JOHN WAKABAYASHI 1 & YILDIRIM DILEK 2 1 1329 Sheridan Lane, Hayward, CA 94544, USA (e-mail: [email protected]) 2 Department of Geology, Miami University, Oxford, OH 45056, USA Abstract: Ophiolites have long been recognized as on-land fragments of fossil oceanic lithosphere, which becomes an ophiolite when incorporated into continental margins through a complex process known as 'emplacement'. A fundamental problem of ophiolite emplacement is how dense oceanic crust becomes emplaced over less dense material(s) of continental margins or subduction-accretion systems. Subduction of less dense material beneath a future ophiolite is necessary to overcome the adverse density contrast. The relationship of subduction to ophiolite emplacement is a critical link between ophiolites and their role in the development of orogenic belts. Although ophiolite emplacement mechanisms are clearly varied, most existing models and definitions of emplacement concern a specific type of ophiolite (i.e. Oman or Troodos) and do not apply to many of the world's ophiolites. We have denned four prototype ophiolites based on different emplacement mechanisms: (1) 'Tethyan' ophiolites, emplaced over passive continental margins or microcontinents as a result of collisional events; (2) 'Cordilleran' ophiolites progressively emplaced over subduction complexes through accretionary processes; (3) 'ridge-trench intersection' (RTI) ophiolites emplaced through complex processes resulting from the interaction between a spreading ridge and a subduction zone; (4) the unique Macquarie Island ophiolite, which has been subaerially exposed as a result of a change in plate boundary configuration along a mid-ocean ridge system. Protracted evolutionary history of some ocean basins, and variation along the strike of subduction zones may result in more complicated scenarios in ophiolite emplacement mechanisms. No single definition of emplacement is free of drawbacks; however, we can consider the inception of subduction, thrusting over a continental margin or subduction complex, and subaerial exposure as critical individual stages in ophiolite emplacement.

Ophiolites have been recognized as on-land fragments of oceanic crust since the advent of plate tectonics (e.g. Gass 1968; Dewey & Bird 1970; Moores 1970; Coleman 1971; Moores & Vine 1971). Incorporation of ophiolites into continental margins is a significant component of the tectonic evolution of orogenic belts and has been broadly defined as 'ophiolite emplacement' or 'ophiolite obduction' (e.g. Moores 1970; Dewey & Bird 1970, 1971; Coleman 1971). Scientific evaluation of ophiolite emplacement has played a key role in the formulation of plate tectonic theory, because ophiolites provide a critical link between the seafloor spreading evolution of oceanic plates and their demise at subduction zones and because the mechanisms of their incorporation into land constitute a first-order tectonic problem in plate tectonics. Ophiolite emplacement mechanisms were once a subject of vigorous debate, particularly with respect to the derivation of an ophiolite from the

lower (Coleman 1971) versus upper plate (Temple & Zimmerman 1969; Dewey & Bird 1970, 1971; Moores 1970) of a subduction system (Fig. 1). In the past two decades, however, controversy regarding the tectonic setting of ophiolite formation has greatly overshadowed any debate over emplacement mechanisms (e.g. Moores et al 2000 and references therein). The widespread acceptance of the suprasubduction zone (SSZ) ophiolite concept (e.g. Robinson et al. 1983; Pearce et al. 1984) has contributed to the swinging of the majority opinion on ophiolite emplacement toward the model of emplacement from the upper plate of a subduction system (e.g. Dewey 1976; Moores 1982; Searle & Stevens 1984) (Figs 1 and 2). Regardless of their original tectonic setting of igneous formation, ophiolites became incorporated into continental margins through complex interactions of lithospheric plates and hence the mechanisms of ophiolite emplacement should be expected to vary depending on the age, thickness and thermal state

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 427-447. 0305-8719/03/$15 © The Geological Society of London 2003.

428

J. WAKABAYASHI & Y. DILEK

Fig. 1. The preferred model of emplacing an ophiolite over a continental margin (same thrusting sense as subduction) contrasted with emplacement antithetic to the subduction polarity. Although these diagrams illustrate the case for Tethyan ophiolites, the same principles apply to Cordilleran ophiolites emplaced over subduction-accretion complexes (see Fig. 4).

of oceanic crust, the nature and geometry of plate boundaries involved, and the size and character (i.e. oceanic versus continental, microcontinent, island arc, seamount, etc.) of the interacting plates. Although ophiolite emplacement mechanisms have been debated for several decades, most of the arguments concern a specific type of ophiolite and do not apply to different types of ophiolites around the world. In this paper we examine the existing ideas and models on ophiolite emplacement mechanisms to better document the nature and order of the processes involved in the incorporation of fossil oceanic crust into continental margins as ophiolites. We define four prototypes of ophiolites based on their emplacement mechanisms, which deviate from each other as a result of different plate interactions in the past. We then

present a critical evaluation of the models on subduction initiation and metamorphic sole development, both of which constitute two major phases in ophiolite emplacement. Finally, we discuss the emplacement mechanisms of the four prototypes of ophiolites.

Ophiolite prototypes We follow in this paper the 1972 Penrose definition of an ophiolite (Penrose Conference Participants 1972) for simplicity, although we realize the obvious shortcomings of this restricted definition in ophiolite classification (Dilek 2003), because the discussion of various tectonic environments of ophiolite genesis is not directly relevant to emplacement mechanisms. Our discussion of ophiolites excludes thrust slices or blocks of pelagic sedi-

EMPLACEMENT OF AN OPHIOLITE

429

Fig. 2. Illustration that ophiolite emplacement and the environment of the emplacement do not constrain the tectonic setting of ophiolite genesis. Any type of ophiolite, whether it be nascent arc, interarc, backarc or mid-ocean ridge generated, can be emplaced over a continental margin or subduction-accretion complex.

mentary rocks, basalt and variably serpentinized ultramafic rock that are intercalated within accretionary wedges. Exposures of ophiolitic rocks in subduction-accretion systems are not treated as ophiolite complexes in this paper because: (1) accretionary wedge ophiolitic rocks most commonly comprise small blocks or thrust sheets of basalt with or without overlying chert or limestone; (2) serpentinite, although locally present as moderately large bodies or sheets (up to several kilometres of structural thickness and 30 km in along-strike length), seldom occurs in the same block or thrust sheet with basalt and chert; (3) gabbro or sheeted dykes are extremely rare in accretionary wedge sheets or blocks; (4) the largest dimensions of most thrust sheets of ophiolitic rocks in accretionary wedges are less than 10 km, whereas Penrose-type ophiolites can extend for hundreds of kilometres along-strike; (5) different scraps of oceanic rocks within the same accretionary wedge can vary greatly in age and origin. The lack of large ophiolite sheets, containing thick plutonic sections, in subduction complexes, is consistent with the conclusion of Cloos (1993) that all downgoing oceanic crust is subducted except for a few topographic highs from which basalt and pelagic sediments may be offscraped. Ultramafic rocks within accretionary

wedges may be the off-scraped remnants of peridotite-cored uplifts formed at ridge-transform intersections (Coleman 2000). We distinguish four prototypes of ophiolites based on their emplacement mechanisms and the nature of their underlying tectonic basements: (1) Tethyan; (2) Cordilleran; (3) ridge-trench intersection (RTI); (4) Macquarie Island-type. Moores (1982) recognized the differences between the Tethyan and Cordilleran types and provided several lines of criteria for their distinction that we follow herein. RTI ophiolites are special because both their igneous evolution and tectonic emplacement are strongly controlled by the spatial and temporal interactions between mid-ocean ridges and subduction zones (e.g. Forsythe & Nelson 1985; Lytwyn et al. 1997). The close association and interaction of ridges with trenches during the formation of RTI ophiolites are unrelated to subduction initiation. The Macquarie Island-type ophiolite presents a unique case (Varne et al. 1969, 2000) whereby relatively in situ and young oceanic crust has been exposed subaerially as a result of changing plate boundary configurations. Depending on the interpretation of the tectonic setting of the Macquarie Island ophiolite, an argument could be made that this ophiolite is subaerially exposed but not emplaced.

430

J. WAKABAYASHI & Y. DILEK

The terms Cordilleran and Tethyan traditionally carry geographical connotations, but we emphasize that we define 'Cordilleran' and 'Tethyan' ophiolites on the basis of their emplacement mechanisms, not their location. For example, the Brooks Range ophiolite, of Alaska in the North American Cordillera, has been emplaced over a continental margin (Wirth et al. 1993) reminiscent of Tethyan-type ophiolites, and hence we consider it a Tethyan ophiolite for the purposes of our discussion of emplacement. Some ophiolites in the Sierra Nevada of California, also part of the North American Cordillera, have been emplaced over continental margins or island arcs (e.g. Moores 1970; Moores & Day 1984) much as in Tethyan examples, and therefore we also would consider them as Tethyan ophiolites from the standpoint of their emplacement. On the other hand, the Cretaceous ophiolites of Neo-Tethys in the eastern Pontide belt of Turkey clearly have a protracted emplacement history typical of subduction-accretion systems in the Pacific Rim (Yilmaz et al. 1997), and we consider these ophiolites as Cordilleran in character regarding their emplacement histories. We briefly summarize the characteristic features of these four ophiolite prototypes below and discuss their emplacement mechanisms in a later section.

Tethyan-type ophiolites Tethyan ophiolites structurally overlie passive continental margins and their crystalline basement, microcontinental fragments, or island arcs. Tethyan ophiolites in the eastern Mediterranean region commonly display a Penrose-type complete pseudostratigraphy (defined as having upper-mantle rocks, cumulates, gabbros, sheeted dykes, volcanic rocks) and include some of the classic ophiolites of the world (e.g. Troodos ophiolite, Dilek et al. 1990b; Robinson & Malpas 1990; Oman ophiolite, Searle & Cox 1999; Dilek et al 1998; Bay of Islands ophiolite in Newfoundland, Casey et al. 1981). Ligurian-type ophiolites exposed in the western Alps and Apennines have Hess-type oceanic crust with MORB affinities (Dilek 2003, and references therein) and are also considered as Tethyan based on their emplacement mechanisms. Extrusive sections of most Tethyan ophiolites do not have volcaniclastic rocks that are typical of volcanic arcs (e.g. Dilek & Moores 1990, and references therein), but the upper-crustal rocks in many Tethyan ophiolites display the geochemical characteristics of subduction zone environments (e.g. Pearce 1975; Alabaster et al. 1982; Rautenschlein et al. 1985; Umino et al. 1990; Jenner et al. 1991). Metamorphic soles, thin (<500m thick)

sheets of high-grade metamorphic rocks, are present beneath most Tethyan ophiolites (e.g. Wil= Hams & Smyth 1973; Spray 1984; Jamieson 1986; Dilek et al. 1999). There is a significant break in metamorphic pressure between the ophiolite, which commonly exhibits negligible burial metamorphism, and the structurally underlying metamorphic sole (reviewed by Wakabayashi & Dilek 2000).

Cordilleran-type ophiolites Cordilleran ophiolites structurally overlie subduction-accretion complexes and range from rare complete ophiolite sections to those missing one or more of the major ophiolite lithologies (e.g. Irwin 1977; Saleeby 1992; Coleman 2000). Volcaniclastic and intermediate to silicic volcanic rocks that are generally associated with island arc development are widespread in the extrusive sections of Cordilleran ophiolites. Upper-crustal rock units in Cordilleran ophiolites display island arc tholeiite to calcalkaline chemical affinities indicating a subduction zone origin of their magmas (Shervais & Kimbrough 1985; Shervais 1990; Saleeby 1992). The existence of volcaniclastic rocks, including some subaerial depositions, indicates the construction of volcanic arc edifices during the evolution of these ophiolites. Crosscutting field and geochronological relations from the Jurassic ophiolites in the Sierra Nevada foothills in California show that the arc construction had occurred on and across a pre-existing, multiply deformed and heterogeneous oceanic basement (Dilek et al. 1990a, 1991). Metamorphic soles are present beneath many Cordilleran ophiolites, although in some cases the sole has been nearly completely dismembered (e.g. Platt 1975; Brown et al. 1982; Cannat & Boudier 1985; Wakabayashi & Dilek 2000). Blueschist-facies rocks are also present structurally beneath many Cordilleran ophiolites (e.g. Ernst 1971; Platt 1975; Brown et al. 1982; Ernst 1988). There is a significant break in metamorphic pressure between the ophiolite, which commonly exhibits negligible burial metamorphism, and the structurally underlying metamorphic sole or blueschist-facies rocks (e.g. Platt 1986).

Ridge-trench intersection (RTI) ophiolites Ridge-trench intersections are common in plate tectonics and may cause anomalous near-trench igneous activity (Marshak & Karig 1977), which may result in oceanic crust or ophiolite formation (Casey & Dewey 1984). Ridge-trench intersection (RTI) ophiolites may have a complete or nearly complete pseudostratigraphy. Accretionary wedge

EMPLACEMENT OF AN OPHIOLITE materials may be present both structurally above and below RTI ophiolites (e.g. Lytwyn et al. 1997). The ridge-trench intersections are associated with low-pressure, high-temperature metamorphism of the rocks structurally above (and inboard) of the ophiolite (e.g. Sisson & Pavlis 1993; Brown 1998). Examples of this type of ophiolites include the Resurrection Bay and Knight Island ophiolites in Alaska (Lytwyn et al. 1997) and the Taitao ophiolite in Chile (Forsythe & Nelson 1985; Nelson et al. 1993; Lagabrielle et al. 2000). The petrology and geochemistry of the ultramafic, gabbro and dyke sections of the Taitao ophiolite in Chile display a mid-ocean ridge basalt (MORB) affinity whereas the geochemistry of extrusive rocks, which are interpreted to have erupted after the spreading centre had intersected with the trench (LeMoigne et al. 1996), indicates a mixed MORB and island arc tholeiite affinity (LeMoigne et al. 1996). The degree of decompressional melting of MORB mantle, caused by ridge subduction, was apparently less rigorous than that typically occurring at mid-ocean ridges because of the capping of the melting column by the continental edge. This phenomenon, combined with crustal assimilation and fractional crystallization of enriched MORB melt, produced more silicic rocks in the extrusive sequence of the Taitao ophiolite (Kaeding et al. 1990; Lagabrielle et al. 1994). Sheeted dykes and lavas from the two Alaskan ophiolites have geochemistry similar to that of MORB with some arc influence that has been attributed to intersection of the spreading centre with a subduction zone and a slab-free window (Lytwyn et al. 1997). Although only two ridge-trench ophiolite localities have been described to date, this type of ophiolite may be more common in the rock record (van den Beukel & Wortel 1992).

431

The Macquarie Island ophiolite differs from Tethyan- and Cordilleran-type ophiolites in that it structurally overlies either in situ oceanic crust (Varne et al. 2000) or suboceanic mantle (Daczko et al. 2002), rather than a continental margin or a subduction-accretion complex. A metamorphic sole is not exposed. If such a sole were present, it would be beneath the present level of exposure. Igneous ages of the Macquarie Island ophiolite, determined from 40Ar/39Ar step heating ages of two different basalt outcrops, range between 9.7 and 11.5 Ma (Duncan & Varne 1988). These ages are consistent with the estimated age of the ophiolite from plate motion models, magnetic anomaly patterns and the ages of associated sedimentary rocks (Varne et al. 2000). The generation of oceanic crust in the Macquarie Island ophiolite apparently occurred at slow rates (5— 10 mm a"1 half-spreading rate) during the waning stages of sea-floor spreading activity at a midocean ridge (Varne et al. 2000). Goscombe & Everard (2001) suggested that, following generation of the 'Macquarie Island oceanic crust' and associated extensional faulting in the vicinity of the spreading centre, the ophiolite was subject to transtensional deformation, possibly during transition from the spreading environment to a transform environment. The transtension was followed by transpressional deformation along the transform fault plate boundary. In contrast, Daczko et al. (2002) interpreted all structures on Macquarie Island, including active ones, to be extensional or transtensional. Greenschist-grade lower pillow lavas and sheeted dykes have yielded 40Ar/39Ar step heating ages of 6.5-7.2 Ma, interpreted to indicate cooling at the end of greenschist-grade hydrothermal metamorphism (Duncan & Varne 1988). The transform plate boundary along which Macquarie Island is situated became transpressional at or after 5 Ma (Varne et al. 2000; Goscombe & Everard 2001).

Macquarie Island ophiolite The Macquarie Island ophiolite is exposed on the 37 km X 5 km Macquarie Island, which is situated about 950 km SSW of New Zealand and 1500 km SSE of Tasmania along the transform boundary between the Indo-Australian and Pacific Plates (Varne et al. 1969, 2000). Exposures on Macquarie Island comprise a complete, Penrose-type ophiolite including basalts (making up nearly twothirds of the exposures) with intercalated sedimentary rocks, sheeted dykes, gabbro and ultramafic rocks. The basaltic rocks display MORB and enriched MORB (E-MORB) chemistry, and plate motion reconstructions place the site of ophiolite generation at a mid-ocean ridge spreading centre (Varne et al. 2000).

Development of metamorphic soles and subduction initiation: a critical part of ophiolite emplacement Structure and evolution of metamorphic soles Initiation of subduction and formation of metamorphic soles have been linked to the ophiolite emplacement process. Some researchers have explicitly defined the inception of subduction and consequent development of a metamorphic sole beneath an ophiolite as emplacement, or at least the first stage of emplacement (e.g. Williams & Smyth 1973; Malpas 1979; McCaig 1983; Jamieson 1980, 1986; Hacker et al. 1996).

432

J. WAKABAYASHI & Y. DILEK

Subophiolitic metamorphic soles, or simply metamorphic soles, are thin (<500 m thick), faultbounded sheets of highly strained high-grade metamorphic rocks that structurally underlie many ophiolite complexes (e.g. Williams & Smyth 1973; Jamieson 1986). The higher-grade parts of the metamorphic soles are composed mainly of metabasic rocks of oceanic affinity, with minor metamorphosed pelagic sedimentary rocks. Many soles display inverted metamorphic field gradients and an inverted ocean crustal sequence. The highgrade parts of such soles appear to grade structurally downward from metagabbros to metabasalts to metamorphosed pelagic sedimentary rocks (Jamieson 1986). The pressure-temperature (P—T) conditions of metamorphism for subophiolitic soles are consistent with high-temperature metamorphism beneath hot suboceanic mantle. Metamorphic soles are thought to form at the inception of oceanic subduction beneath the hot sub-ophiolitic mantle of the hanging wall, as suggested by estimated P— T conditions of their metamorphism, their oceanic protoliths, and the presence of an ophiolite structurally above them (e.g. Williams & Smyth 1973; Malpas 1979; Nicolas & Le Pichon 1980; Spray 1984; Jamieson 1986) (Fig. 1). The inverted temperature anomaly responsible for the high-grade metamorphism of the sole decays quickly (<2 Ma) as subduction continues (Peacock 1988; Hacker 1990, 1991; Hacker et al. 1996). Thus the high-grade metamorphism of the sole can occur only at the inception of subduction because the hanging wall would be too cold to cause highgrade metamorphism thereafter. As a result of thermal insulation from continuing subduction, the metamorphic rocks of the sole cool rapidly through the blocking temperature of commonly applied isotopic dating methods such as 40Ar/39Ar (Wakabayashi & Dilek 2000). Because of this rapid cooling, the metamorphic age of the sole closely approximates the inception of subduction (e.g. Spray 1984; Peacock 1988). Pressures of metamorphism associated with metamorphic soles are higher than can be explained by the structural thickness of material found above them (Wakabayashi & Dilek 2000), indicating that: (1) the amount of underthrusting represented by metamorphic soles is considerable (burial depths range from 20 to 40km); (2) normal faulting must have occurred between the ophiolite and the sole after sole development, to exhume the sole to the present field relationship with the ophiolite. Alternatively, the relationship of the ophiolite structurally above the high-P metamorphic sole can be explained by multiple thrusting events instead of normal faulting (e.g. Cowan et al, 1989; Ring & Brandon 1994). How-

ever, thrusting, in contrast to normal faulting, requires the erosional removal of all material originally present between the ophiolite and sole, an enormous volume of ultramafic material. Large volumes of syn-exhumational ultramafic sediments have not been observed to be associated with ophiolites and their soles. The absence or scarcity of metaclastic rocks in the higher-grade (earliest formed) part of metamorphic soles indicates that the ocean floor at the site of subduction initiation lacked terrigenous sediment cover. This observation suggests either that sites of subduction initiation were far from a major landmass, or that sufficient submarine topography was present to shield the nascent subduction zone from terrigenous sediments. In addition to the inverted temperature gradient recorded in metamorphic soles, an inverted pressure gradient is also observed (Jamieson 1986; Gnos 1998), indicating that the sole is a composite of slices formed at different times, brought together by thrust faulting (Casey & Dewey 1984; Gnos 1998). The structurally lower (lower-grade) parts of soles were probably scraped off the subducting oceanic plate some time after the structurally higher (higher-grade) parts of the sole were formed. The inferred origin of metamorphic soles, as products of subduction initiation, suggests that they may offer insight into how subduction begins, at least in the cases that result in development of metamorphic soles and subsequent emplacement of ophiolites. The high temperature of metamorphism as reflected by mineral assemblages in sole rocks and the small age difference between soles and overlying ophiolites indicate that ocean crust was young (generally 5 Ma or younger) and hot at the inception of subduction (e.g. Spray 1984; Jamieson 1986; Hacker et al. 1996; Dilek et al. 1999). No soles have been found that predate the ophiolites found structurally above them, so if such ophiolites were of SSZ origin (as has commonly been interpreted), they must have been formed above an older subduction zone than the one that began with the formation of the sole (Wakabayashi & Dilek, 2000). Thus, SSZ ophiolites must have been emplaced above a separate, younger, subduction zone than the one they formed over. The inverted ocean crustal sequence exposed in the high-grade parts of soles is not compatible with an ordinary sequence of underplating or offscraping during subduction initiation that would produce a right-side-up ocean-floor stratigraphy within each thrust sheet. The inverted ocean-floor sequence found in metamorphic soles suggests that subduction might have started as the downbowing of young oceanic crust that developed into

EMPLACEMENT OF AN OPHIOLITE an overturned fold (Fig. 3). The inferred overturning of the flexure at the inception of subdue tion is consistent with the model of Mueller & Phillips (1991), who have suggested that foundering of dense oceanic lithosphere alone cannot initiate subduction; an external force is needed. The overturned limb of the progressively forming oceanic flexure becomes thinned by numerous thrust faults as the fold develops into a young subduction zone and the future ophiolite, as well as the sole, are left on the upper plate of this system (Fig. 3). Oceanic rocks in the overturned, fault-thinned limb are subject to high-temperature metamorphism beneath the hot suboceanic mantle; this configuration forms the high-grade metamorphic sole with an 'inverted' sequence of oceanic crust. The actual subduction break is the main structure of what may be a broader shortened zone in the oceanic lithosphere. Subsequent continuous subduction leads to offscraping under lower-temperature conditions, as the hanging wall rapidly cools. The structurally lower parts of the sole may include metaclastic

433

material that may indicate the approach of a continental margin in the lower plate (as in Tethyan ophiolites), or the development of an arc-trench depocentre in the upper plate (as in Cordilleran ophiolites). Anticlockwise P—T paths of metamorphism from both intact (Hacker & Gnos 1997) and dismembered (Wakabayashi 1990) metamorphic soles show evidence of a pressure increase with cooling. This can most easily be achieved if the upper plate of the nascent subduction zone is imbricated (tectonically thickened) after subduction has begun and the metamorphic sole has started to cool (e.g. Wakabayashi 1990).

Exhumation of metamorphic soles Following sole metamorphism, the higher-grade part of the sole is exhumed relative to the overlying ophiolite. This differential exhumation appears to be accommodated by a normal fault above (\ow-P ophiolite on higher-P sole) and by thrust faults below (inverted pressure gradient in

Fig. 3. Model for inception of subduction. The width of the subophiolitic sole is exaggerated in this view so that it is visible. Additional imbrication of the upper plate may occur, leading to the increasing burial with cooling noted in some metamorphic soles. The new subduction zone generally forms in young oceanic crust (near a spreading centre) and may exploit a pre-existing zone of weakness. There may be a density contrast across this zone of weakness with older, more dense material on the side that subducts. Such density contrast would be greatest across a fracture zone, but smaller contrasts may be present across zones of normal faulting in the oceanic crust.

434

J. WAKABAYASHI & Y. DILEK

the sole), suggesting an apparent extrusion of the higher-grade part of the sole. Such relations are analogous to those associated with high-pressure metamorphic rocks (blueschists and eclogites) (Wakabayashi & Dilek 2000). The exhumation fault (or faults) structurally above the sole does not necessarily coincide with the contact between the peridotite and the metamorphic sole; it could be somewhere between the sole and the crustal section of the ophiolite (Hacker & Gnos 1997; Wakabayashi & Dilek 2000). Timing of the exhumation of the high-grade part of the metamorphic sole relative to the ophiolite is poorly constrained in many cases. Geochronological and structural data from the Oman ophiolite (e.g. Hacker et al. 1996; Gregory et al. 1998; Gray et al. 2000) suggest that exhumation of the sole relative to the Oman ophiolite occurred probably less than 10 Ma after the metamorphism of the sole rocks. Such an exhumation event would have occurred prior to, or in the earliest stages of, the thrusting of the ophiolite over the Arabian continental margin. For Tethyan ophiolites in general, metamorphic soles were probably exhumed relative to the ophiolite prior to, or during the earliest stages of thrusting onto a continental margin, because the soles and ophiolites commonly are thrust over continental margin sequences as part of the same nappe system (Moores 1982; Searle & Cox 1999).

Subduction initiation models Mueller & Phillips (1991) showed that the blockage of a subduction zone by buoyant material (island arc, continental fragment, continental margin) is probably the only event capable of generating a large enough external force to initiate a new subduction zone. The conclusions of Mueller & Phillips (1991) are consistent with models of subduction initiation based on field relations in ophiolites (Casey & Dewey 1984), as well as the geodynamic history of the SW Pacific, where the clogging of subduction zones with buoyant material was followed by initiation of new subduction zones (Hall 1996). Although an external force apparently triggers subduction initiation, a material contrast and a zone of weakness in the oceanic lithosphere may determine the location of the nascent subduction zone (Casey & Dewey 1984). An oceanic spreading centre (i.e. mid-ocean ridge) has been suggested as a site of subduction initiation (e.g. Casey & Dewey 1984; Hacker et al. 1996). Initiation of subduction at a spreading centre is consistent with the indistinguishable ages of some ophiolites and their soles (Hacker et al. 1996).

Fracture zones separate oceanic lithosphere of differing age and density and are likely sites of subduction initiation (Casey & Dewey 1984; Hawkins et al. 1984; Stern & Bloomer 1992). Subduction may be initiating along two segments of the Macquarie Ridge, an oceanic transform, in the SW Pacific Ocean (Collot et al. 1995; Frohlich et al. 1997), along the Azores-Gibraltar transform fault, and along a fracture zone in the East Caroline Basin (Mueller & Phillips 1991). The age of oceanic crust on the upper plate of a subduction initiated at a fracture zone should become progressively older along the trend of the trench-line, away from the spreading centre. Such a relationship may result in a variation of age of tens of million years over the hundreds of kilometres of oceanic crust that form a future ophiolite. Some ophiolites in the North American Cordillera show large (tens of million years) ranges of igneous ages as well as lithological heterogeneity suggesting a similar scenario; their tectonic evolution is consistent with subduction initiation along a fracture zone (Saleeby 1990, 1992). In contrast, many ophiolite belts that are hundreds of kilometres long, such as the Coast Range ophiolite of California (Hopson et al 1981, 1996) and many Tethyan ophiolites (e.g. Dewey 1976; Juteau 1980; Dercourt et al. 1986; Dilek & Moores 1990), show a restricted (generally 5 Ma or less) age range. Such ophiolite belts are inconsistent with subduction initiation along an oceanic transform fault or fracture zone. Major contrasts in age and density of oceanic lithosphere are also found where new spreading began in older ocean crust (i.e. rift propagation in the Lau Basin; Parson & Wright 1996; Zellmer & Taylor 2001). Such contrasts in lithospheric age and density would be parallel to a spreading centre, as in the case of ridge-parallel normal faults (Dilek et al. 1988), which constitute preexisting zones of mechanical weakness. Initiation of subduction along such a ridge-parallel discontinuity between old and young crust would result in an ophiolite (on the upper plate of the subduction zone) that is older than its metamorphic sole, and that is of relatively consistent age along the trend of the trench-line.

Ophiolite emplacement mechanisms Proposed models and the problem of emplacing oceanic lithosphere over less dense rocks The existing ophiolite emplacement models generally fall into four categories (Fig. 1): (1) emplacement by partial subduction of a continental

EMPLACEMENT OF AN OPHIOLITE margin beneath the displaced fossil oceanic crust (e.g. Temple & Zimmerman 1969; Dewey & Bird 1970, 1971; Moores 1970); (2) emplacement by antithetic thrusting of oceanic crust from the subducting plate (e.g. Coleman 1971), referred to by some as flake tectonics (e.g. Oxburgh 1972); (3) emplacement by gravity sliding (e.g. Reinhardt 1969; Church & Stevens 1971; Smith & Woodcock 1976); (4) emplacement through intersection of a spreading ridge with a subduction zone (e.g. Forsythe & Nelson 1985; van Beukel & Wortel 1992; Lytwyn et al. 1997). The term 'obduction' was first defined by Coleman (1971) to explain ophiolite emplacement through antithetic thrusting along active continental margins. Dewey (1976) used obduction, however, to refer to any type of ophiolite emplacement mechanism, and others have followed this usage (e.g. Searle & Stevens 1984). How oceanic crust comes to be emplaced over the less dense continental margin material or subduction-accretion complex is the central problem of ophiolite emplacement. A viable emplacement model needs to include a mechanism that overcomes this adverse density contrast. Gravity sliding, in the absence of other processes, requires an unrealistic topographic high on the ocean floor and improbable transport distances necessary to emplace an ophiolite (Dewey 1976; Moores 1982). Gravity sliding might have played a partial role in emplacement of some ophiolites (Searle & Stevens 1984), particularly after collision-induced thrusting caused significant crustal uplift and topographic buildup; these processes might have then produced high gravitational potential energy in the upper-plate rocks that would have triggered downward sliding of ophiolitic packages onto the continental margin sequences in the lower plate (Gregory et al. 1998; Gray et al. 2000; Gray & Gregory 2003). However, it is unlikely that gravity sliding can be the sole agent or primary mechanism of ophiolite emplacement. Subduction zones are the only locations on Earth where less dense material is thrust beneath denser material on a large scale. Thus, ophiolite emplacement mechanisms must be spatially associated with subduction. Such a linkage is consistent with the occurrence of metamorphic soles beneath ophiolites. Underthrusting of buoyant material at subduction zones is a consequence of the attachment of such material to the dense downgoing oceanic lithosphere. This includes both passive continental margins or island arcs attached to a downgoing oceanic slab, and accretionary wedge materials that are scraped off the downgoing oceanic slab. Because the pull of the sinking oceanic lithospheric slab is such an important driving force in plate tectonics (e.g. Forsythe &

435

Uyeda 1975), ophiolite emplacement is best viewed as less dense material being dragged (by the descending slab) beneath an ophiolite, rather than the pushing of an ophiolite over less dense material. Consequently, an ophiolite emplaced as part of the upper plate of a subduction system (Temple & Zimmerman 1969; Dewey & Bird 1970, 1971; Moores 1970), is more plausible than emplacement from the downgoing slab (e.g. 'obduction' of Coleman 1971), because in the latter scenario there is no slab to drag the less dense material beneath the ophiolite (Fig. 1). An exception is the subduction of an active spreading centre, which may arrest the subduction of very young, buoyant, oceanic lithosphere and result in emplacement of ophiolites from the downgoing plate (e.g. Forsythe & Nelson 1985; van Beukel & Wortel 1992). The connection between subduction zones and ophiolite emplacement links ophiolites to the development of orogenic belts. Ophiolites make up the structural 'roof of palaeosubduction zones, and ophiolite-marked subduction sutures have been considered the most important first-order structures in orogenic belts (e.g. Moores 1970; Moores et al 1999).

Collisional emplacement of Tethyan-type ophiolites Emplacement of Tethyan ophiolites has been traditionally defined as the thrusting of an ophiolite over a continental margin and/or a crystalline complex of a microcontinent (e.g. Temple & Zimmerman 1969; Dewey & Bird 1970, 1971; Moores 1970; Coleman 1971) (Fig. 4). By this definition, the inception of oceanic subduction (and development of the metamorphic sole) beneath the ophiolite predates the terminal emplacement event (Fig. 4) (Moores 1982). However, some others have defined the inception of subduction as emplacement itself (e.g. Williams & Smyth 1973; Malpas 1979; Jamieson 1980; McCaig 1983). Alternatively, thrusting of an ophiolite over a passive continental margin has been considered but one step in a multi-stage emplacement process, the beginning of which may involve transform fault tectonics (Brookfield 1977) or the inception of subduction (e.g. Casey & Dewey 1984; Jamieson 1986; Hacker^ al. 1996). Collision of a passive continental margin leads to the arrest of subduction because the continental material is too buoyant to be subducted (e.g. Temple & Zimmerman 1969; Moores 1970). Subduction jump and a flip of subduction polarity may then follow, creating the field relations noted by Coleman (1971) in which the active subduction

436

J. WAKABAYASHI & Y. DILEK

Fig. 4. Emplacement of Tethyan and Cordilleran ophiolites (a) and ridge-trench ophiolites (b). It should be noted that if a continental margin is attached to the plate subducting beneath a Cordilleran ophiolite, such an ophiolite may eventually be thrust over a continental margin, 'converting' it to a Tethyan-type ophiolite. Similarly, if subduction continues after ridge-trench ophiolite emplacement, a subduction complex may develop structurally beneath such an ophiolite and it would effectively become a Cordilleran ophiolite, although rocks structurally above the ophiolite would exhibit higher-grade metamorphism than those associated with a typical Cordilleran ophiolite. Ridge-trench emplacement may also be followed by collision, converting the ophiolite into a Tethyan ophiolite; rocks structurally above the ophiolite would show a higher grade of metamorphism than a typical Tethyan ophiolite setting.

zone dips beneath the recently emplaced ophiolite. In such a scenario, emplacement of the ophiolite is facilitated by the previous subduction zone dipping away from the continental margin. Such subduction polarity flips have occurred in the SW Pacific, including the continuing subduction polar-

ity flip in eastern New Guinea (Cooper & Taylor 1987), and the c. 10-5 Ma polarity flip in northern Sulawesi (Hall 1996). Because the collision of a buoyant microcontinent or arc with a subduction zone results in the arresting of subduction, similar to the collision of a continent in the downgoing

EMPLACEMENT OF AN OPHIOLITE

437

Fig. 4. (continued)

plate with a trench (e.g. Cloos 1993), ophiolites that are thrust over microcontinents or island arcs (e.g. Hall 1996) are also considered as Tethyantype ophiolites in this treatment. We define this mechanism of emplacement as 'collisiona!'. The emplacement of collisional Tethyan-type ophiolites includes the following events from oldest to youngest: (1) initiation of intra-oceanic subduction and formation of metamorphic sole; (2) exhumation of metamorphic sole relative to the ophiolite; (3) thrusting of the ophiolite over continental margin via collision; (4) subaerial exposure of the ophiolite. Ages of metamorphic soles are commonly similar to or slightly younger (<2 Ma) than the igneous ages of crustal rocks in many Tethyan ophiolites, and most Tethyan ophiolites were emplaced onto continental crust within 10 Ma of their formation (Dewey 1976; Dilek et

al. 1999). A notable exception is the Cretaceous Troodos ophiolite, whose thrusting over the Eratosthenes Seamount was facilitated by the collision of this seamount with the Cyprus trench (and a north-dipping subduction zone) starting in the Pliocene (Robertson 1998). Oman ophiolite. The Tethyan-type Oman ophiolite, considered by many as the best-exposed and most complete ophiolite in the world, has been the subject of debate over its emplacement mechanism^). Some researchers believe that the Oman ophiolite was emplaced over a single subduction zone directed away from the Arabian continental margin (e.g. Dewey 1976; Hacker et al. 1996; Searle & Cox 1999; Searle et al. 2003). Other workers have concluded that an earlier episode of subduction, dipping beneath the continent, was

438

J. WAKABAYASHI & Y. DILEK

followed by gravitational collapse and possible formation of a new subduction zone dipping away from the continent (Gregory et al. 1998; Gray et al 2000; Gray & Gregory 2003). In this model, ophiolite emplacement is inferred to have been associated with the latter two events. The controversy involves different interpretations of the structures and, in particular, the geochronology of the high-P metamorphic rocks, including blueschists and eclogites, which occur structurally beneath the ophiolite. One research group interprets the metamorphic ages as reflecting a high-P, subduction-related metamorphic event that preceded ophiolite generation (Miller et al. 1999; Gray et al. 2000; Gray & Gregory 2003). Other researchers (e.g. Hacker et al. 1996; Searle et al. 2003) have concluded that Ar/Ar metamorphic ages from eclogites that exceed the ophiolite in age are a consequence of excess Ar that results in ages significantly older than the actual crystallization age of the metamorphic rocks. If the interpretation of older eclogites is correct, the Oman ophiolite may differ markedly in tectonic setting from all other ophiolites. This is because high-P metamorphic rocks are universally younger than the ophiolites that structurally overlie them (Wakabayashi & Dilek 2000). Accretionary emplacement of Cordillerantype ophiolites Similar to the emplacement of Tethyan ophiolites, the emplacement of Cordilleran ophiolites begins with the inception of subduction beneath the future ophiolite and the formation of a metamorphic sole. The ophiolite is not thrust over a passive continental margin as in collisional Tethyan ophiolites; instead, materials are added to subduction-accretion complex beneath the ophiolite by progressive tectonic accretion (Fig. 4). The history of subduction may also involve removal of previously accreted materials, known as subduction erosion (e.g. von Huene 1986). The subduction-accretion complexes beneath Cordilleran ophiolites include trench sediments, as well as the upper parts of seamounts, oceanic plateaux and aseismic ridges, and other topographic highs from the downgoing oceanic plate (e.g. Cloos 1993). The emplacement process of a Cordilleran ophiolite is gradual or cumulative, in contrast to the punctuated process of Tethyan ophiolite emplacement. The following events are common to all Cordilleran ophiolites: (1) initiation of subduction and formation of metamorphic sole; (2) exhumation of metamorphic sole relative to ophiolite; (3) progressive underthrusting of oceanic material

beneath the ophiolite following inception of subduction; (4) subaerial exposure of the ophiolite. Coast Range ophiolite. The emplacement history of the Coast Range ophiolite in California illustrates the stages of emplacement of a Cordilleran ophiolite (Fig. 5). The Coast Range ophiolite forms scattered exposures over a distance of 900km in western California, a distance that extends to 1300km when slip on the dextral San Andreas fault system is restored (Bailey et al. 1970; Hopson et al. 1981). The crustal sections of the ophiolite are 4km thick or less, and the exposures range from sheared ultramafic rocks with lenses of gabbro and mafic volcanic rocks, to nearly 'complete' Penrose-type sequences that include ultramafic rocks, cumulate and isotropic gabbros, sheeted intrusive rocks, and mafic and intermediate volcanic rocks (e.g. Point Sal ophiolite, Hopson et al. 1981). The Coast Range ophiolite structurally overlies the Franciscan subduction complex (e.g. Bailey et al. 1970) and is depositionally overlain by the forearc basin strata of the Great Valley Group (Dickinson 1970). The Coast Range ophiolite was formed at about 165-170 Ma (Mattinson & Hopson 1992; Hopson et al. 1996), possibly in a back-arc or nascent arc setting (e.g. Moores 1970; Schweickert & Cowan 1975; Dickinson et al. 1996; Ingersoll 2000; Wakabayashi & Dilek 2000) (Fig. 5). Alternatively, the ophiolite may have been formed in a forearc (Shervais 1990; Saleeby 1996) or a midocean ridge setting (Hopson et al. 1981, 1996). Any of the proposed settings of ophiolite genesis is broadly compatible with the emplacement events described below (starting with the inception of subduction beneath the ophiolite); the tectonic setting of ophiolite genesis places no constraint on the mechanisms of emplacement. The Coast Range ophiolite was placed in the upper plate of the east-dipping Franciscan subduction zone at 165-160 Ma, based on interpreted age of metamorphic sole formation of 159163 Ma (Wakabayashi & Dilek 2000). Continued subduction resulted in overprinting of the sole with high-P-low-r (HP-LT) metamorphic minerals (Wakabayashi 1990) (Fig. 6), and continued deformation broke up the metamorphic sole. Remnants of the sole currently occur mostly as blocks in Franciscan melanges, commonly referred to as 'high-grade' blocks (Wakabayashi 1990; Wakabayashi & Dilek 2000). Some of the pieces of the metamorphic sole and some of the structurally highest (and oldest) blueschist-facies rocks of the structurally underlying Franciscan Complex may have been exhumed by Tithonian time (151-144 Ma; Gradstein et al. 1995) based on the following observations:

EMPLACEMENT OF AN OPHIOLITE

439

Fig. 5. Tectonic history of the Coast Range ophiolite. The ophiolite is illustrated as being generated in a nascent arc setting, although other tectonic settings are compatible with the tectonics of emplacement. It should be noted that the width of the metamorphic sole is greatly exaggerated so that it shows on the diagram. The entire forearc region is submerged until the final frame (70-20 Ma). The last three frames are modified from Wakabayashi & Unruh (1995) and Wakabayashi (1999).

(1) high-grade blocks are present in basal (Tithonian and Valanginian) Great Valley Group (Carlson 1981; Phipps 1984); (2) a TithonianValanginian Franciscan sandstone (Moore 1984) contains block(s) of high-grade rock(s); (3) blueschist cobbles including rutile (in Franciscan metamorphic rocks found only in the high-grade blocks) are present in some Tithonian to Valanginian age Franciscan conglomerates (Moore & Liou 1980); (4) intergrown lawsonite and white mica (a texture limited to Franciscan high-grade blocks) are detrital clasts in some TithonianValanginian Franciscan sandstones (Crawford 1975; Brothers & Grapes 1989). Older blueschist belts in the Sierra Nevada to the east (east of the

forearc basin) lack the mineral assemblages or textures listed above. These observations collectively suggest that parts of the blueschist-overprinted metamorphic sole were exhumed prior to Tithonian-Valanginian redeposition into the trench and forearc basin. Much of the early exhumation of the sole may have occurred as blocks in a shear zone rather than as a coherent sheet, because many of the blocks have actinoliteand chlorite-bearing rinds suggesting reaction with surrounding ultramafic rocks at reasonably elevated temperatures (Coleman & Lanphere 1971). The high-grade blocks may have been exhumed as blocks in serpentinite diapirs, a setting similar to the occurrence of blueschist blocks in serpentinite

440

J. WAKABAYASHI & Y. DILEK

Fig. 6. Comparing two representative P-Tpaths from the high-grade blocks of the Franciscan Complex [P-T paths labelled W90 from Wakabayashi (1990)] with two representative P-T paths from the metamorphic sole of the Oman ophiolite (Hacker & Gnos 1997; labelled HG97).

mud volcanoes in the Marianas forearc (Fryer et al 2000) (Fig. 5). Franciscan subduction continued unabated for over 140 Ma, resulting in progressive accretion of units scraped off the downgoing plate (Wakabayashi 1992). At least 25% of the exposed Franciscan complex was metamorphosed under HP-LT, blueschist-facies conditions. Deposition of clastic sediments in the forearc basin on the future Coast Range ophiolite took place while the Franciscan subduction complex was forming structurally beneath the ophiolite (Fig. 5). The burial of the ophiolite beneath forearc strata did not result in significant burial metamorphism of the ophiolite. Subaerial exposure of the Coast Range ophiolite and some of the underlying accretionary complex may have locally occurred in the Eocene (Nilsen & McKee 1979), while subduction was still active, and was widespread by the Miocene, when a transform plate boundary replaced the subduction zone (Cole & Armentrout 1979). As illustrated in this Coast Range ophiolite example, emplacement of a Cordilleran ophiolite is gradual, with only the inception of subduction standing out as a welldefined event in the history of the ophiolite.

& Wortel 1992). It is difficult to envision why the oceanic flake would form from the landward side of the spreading centre because such a piece is attached to the downgoing plate, and because the strength of the oceanic plate should be lowest at the spreading centre (e.g. Mueller & Phillips 1991) (Fig. 4). It seems more plausible that the ophiolitic slice is derived from the seaward side of the spreading centre (Fig. 4) where the oceanic lithosphere, of zero age and buoyant, is separated from the downgoing oceanic slab and thus from slab pull forces by the spreading centre itself. If this scenario is correct, then this piece of young oceanic lithosphere stalls and does not subduct. A new subduction zone then forms outboard of the stalled piece of oceanic lithosphere (Fig. 4). Neartrench intrusions and volcanism occur as a result of thermal activity related to the slab-free window (e.g. Thorkelson 1996) and the initiation of new subduction (Fig. 4). Elevated geothermal gradients from the slab-free window result in low-P-high-J metamorphism (e.g. Sisson & Pavlis 1993; Brown 1998). An exhumed accretionary wedge with such high-temperature metamorphism and plutons may resemble an exhumed magmatic arc (Brown 1998). Ridge-trench intersections are common geological phenomena and this process has been suggested as a common ophiolite emplacement mechanism (van den Beukel & Wortel 1992). However, only two examples of this type of ophiolite have been identified thus far. This may be because the ophiolite subsides beneath sea level after subduction resumes (e.g. Collot et al. 1995; see discussion below). If a significant volume of accretionary wedge material were accreted structurally beneath the ophiolite as a consequence of continued subduction, then the ridge-trench intersection ophiolite would become a Cordilleran-type ophiolite.

Emplacement of the Macquarie Island ophiolite

Interpretations of the emplacement of the Macquarie Island ophiolite include emplacement over a subduction complex (Dewey & Bird 1971) and emplacement antithetic to subduction (Coleman 1971). More recent studies have suggested that the Emplacement ofRTI ophiolites Macquarie Island ophiolite has been thrust over Ridge-trench intersection ophiolites are emplaced oceanic crust as a consequence of transpression as a consequence of the subduction of an oceanic along a diffuse transpressional plate boundary spreading ridge (Forsythe & Nelson 1985; van den (Varne et al. 2000; Goscombe & Everard 2001). Beukel & Wortel 1992). Published models suggest In contrast, Frohlich et al. (1997) and Dazcko et that the emplacement process occurs by stranding al. (2002) suggested that there is no evidence for a piece of the oceanic plate from the landward underthrusting of oceanic crust beneath Macquarie side of the spreading centre (e.g. van den Beukel Island.

EMPLACEMENT OF AN OPHIOLITE The Macquarie Island ophiolite differs from all other known ophiolites in that no lower-density material structurally underlies the ophiolite. The ophiolite structurally overlies either oceanic crust or suboceanic mantle. If the ophiolite overlies suboceanic mantle instead of underthrust oceanic crust, a good argument could be made that the Macquarie Island ophiolite is not emplaced at this time. Subduction may be initiating along segments of the transform plate boundary both north and south of Macquarie Island (Ruff et al. 1989; Collot et al 1995; Frohlich et al 1997). If subduction begins and then progresses beneath the ophiolite, the upper plate of the subduction zone may subside, leading to submergence of the ophiolite. Submergence of islands on the upper plate of the subduction zone has occurred along the northern part of the plate boundary, where subduction has only recently initiated (Collot et al 1995). If the Macquarie Island ophiolite should experience such submergence in the future, it is likely that re-emergence of the ophiolite would occur only with significant underplating of a subduction complex beneath the ophiolite (converting the ophiolite to a Cordilleran-type) or collision of a passive continental margin (i.e. the western edge of the Campbell Plateau) with the subduction zone beneath the ophiolite (converting the ophiolite to a Tethyan-type).

Potential impact of complex plate interactions along irregular continental margins The discussion of different types of ophiolite emplacement mechanisms has focused on 2D, cross-sectional views of emplacement processes. Such cross-sectional views assume regular continental margins and a uniformity of processes along the strike of a subduction zone; these features do not reflect the actual complexity of interactions observed along modern convergent plate margins (e.g. Hall 1996). Synchronous ophiolite emplacement and new oceanic crust generation may occur at some convergent plate boundaries, where the collision of an irregular continental margin with a trench may result in ophiolite emplacement at promontories whereas in slab-rollback, forearc magmatism and young ocean crust formation within embayments. Actualistic examples exist along the collision zone between the Indo-Australia plate and the Sunda arc-trench system, where the Australian continental margin is colliding with a segment of the trench south of Timor, whereas subduction of the Indian Ocean floor is proceeding along the Sunda Trench farther west (Harris 2003). As a conse-

441

quence, emplacement of a single ophiolite may vary along-strike from collisional to accretionary. In addition to changes along the strike of a subduction zone, strike-slip faulting may play an important role in the juxtaposition of ophiolites with adjacent terranes (e.g. Hopson et al 1996) and/or in lateral translation of ophiolites and 'suspect terranes' for long distances along-strike of an orogenic belt (e.g. Cowan et al 1997).

Discussion: how should we define ophiolite emplacement? Existing definitions of ophiolite emplacement in the literature, developed mainly for Tethyan ophiolites, clearly cannot be applied to other ophiolite types. Two geological events are common to all types of ophiolites despite their differing tectonic histories: (1) initiation of subduction beneath the ophiolite; (2) subaerial exposure of ophiolite. An emplacement definition that would apply to all ophiolites, with the possible exception of the Macquarie Island ophiolite, would be the inception of subduction beneath the ophiolite. However, using the inception of subduction as the definition of emplacement may create confusion because it would contradict decades of published emplacement definitions for Tethyan ophiolites. Given that ophiolites are defined as con-land fossil oceanic crust', an argument could be made that subaerial exposure of an ophiolite characterizes its emplacement. To date, there are no submerged units of oceanic rocks that are called ophiolites (unless they are physically connected to a subaerial exposure). The only named ophiolite in the world that does not have a surface exposure is the 'Great Valley ophiolite' of California (Godfrey & Klemperer 1998; Coleman 2000; Godfrey & Dilek 2000), which is buried beneath several kilometres of sedimentary rocks of the Great Valley basin. Some may argue, however, that the Great Valley ophiolite does not truly fit the definition of an ophiolite because it is not exposed. Subaerial exposure of an ophiolite is generally fairly easy to define in the geological record, but it is not commonly associated with an important tectonic event in its history. Consequently, defining subaerial exposure as emplacement may not be as useful as the inception of subduction for the purposes of discussing ophiolite tectonic history. In addition, existing ophiolites can be submerged as a result of eustatic sea-level changes, subsidence, or technically induced burial by sediment and rock avalanches, leading to a potential for future 'unemplacement' of an ophiolite.

442

J. WAKABAYASHI & Y. DILEK

Another alternative is to consider emplacement as the entire process between subduction initiation and subaerial exposure of the ophiolite. Thus, geological events such as inception of subduction, thrusting over a continental margin (if applicable), and subaerial exposure become individual stages in the emplacement of an ophiolite. A viable alternative would be to classify the emplacement mechanism according to the four prototypes discussed in this paper. Classifying the emplacement mechanisms using these four prototypes would require little modification of existing definitions, but complications still exist. For example, the potential for along-strike changes from accretionary to collisional emplacement can potentially complicate the use of a single emplacement classification for an ophiolite. In addition, closure of wide ocean basins following a protracted subduction history may result in the thrusting of an ophiolite and underlying accretionary wedge over a continental margin. In essence, this may 'convert' an emplacement mechanism from an 'accretionary' to a 'collisional' one in time. Another drawback of classifying emplacement mechanisms according to the four prototypes is the potential confusion resulting from our mechanistic rather than geographical use of the terms 'Cordilleran' and 'Tethyan'. No single definition of emplacement is free of drawbacks in terms of the potential confusion that it may cause for readers of the ophiolite literature. Perhaps the best recommendation is for researchers to clearly specify what they mean by emplacement when they use the term, with the awareness that inception of subduction beneath an ophiolite is probably the only definition that would apply to all ophiolites. We thank E. Moores for years of stimulating discussions on the subject of ophiolite tectonics. Our studies of ophiolites have been supported by research grants from the National Science Foundation (EAR-9219064, EAR9796011, OCE-9813451 to YD.), the NATO Science Programme (CRG-970263, ESTCLG-97617), and the National Geographic Society. Constructive reviews by P. Robinson, R. Sedlock and E. Moores greatly improved the text and our understanding of processes and mechanisms of ophiolite emplacement, and are gratefully acknowledged.

References ALABASTER, T., PEARCE, J.A. & MALPAS, J. 1982. The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex. Contributions to Mineralogy and Petrology, 82, 168-183. BAILEY, E.H., BLAKE, M.C. JR & JONES, D.L. 1970. Onland Mesozoic ocean crust in California Coast Ranges. US Geological Survey, Professional Paper,

700,70-81. BROOKFIELD, M.E. 1977. The emplacement of giant ophiolite nappes I. Mesozoic-Cenozoic examples. Tectonophysics, 37, 247-303. BROTHERS, R.N. & GRAPES, R.H. 1989. Clastic lawsonite, glaucophane, and jadeitic pyroxene in Franciscan metagreywackes from the Diablo Range, California. Geological Society of America Bulletin, 101, 14-26. BROWN, E.H., WILSON, D.L., ARMSTRONG, R.L. & HARAKAL, J.E. 1982. Petrologic, structural, and age relations of serpentinite, amphibolite, and blueschist in the Shuksan Suite of the Iron Mountain-Gee Point area, North Cascades, Washington. Geological Society of America Bulletin, 93, 1087-1098. BROWN, M. 1998. Ridge-trench interactions and highT-low-P metamorphism, with particular reference to the Cretaceous evolution of the Japanese islands. In: TRELOAR, P.J. & O'BRIEN, P.J. (eds) What Drives Metamorphism and Metamorphic Reactions?. Geological Society, London, Special Publications, 138, 137-169. CANNAT, M. & BOUDIER, F. 1985. Structural study of intra-oceanic thrusting in the Klamath Mountains, northern California: implications on accretion geometry. Tectonics, 4, 435-452. CARLSON, C. 1981. Upwardly mobile melanges, serpentinite protrusions, and transport of tectonic blocks in accretionary prisms. Geological Society of America, Cordilleran Section, Abstracts with Programs, 13, 48. CASEY, J.F. & DEWEY, J.F. 1984. Initiation of subduction zones. In: GASS, I.G, LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 269-290. CASEY, J.F., DEWEY, J.F., Fox, P.J., KARSON, J.A. & ROSENCRANTZ, E. 1981. Heterogeneous nature of oceanic crust and upper mantle: a perspective from the Bay of Islands Ophiolite Complex. In: EMILIANI, C. (ed.) The Oceanic Lithosphere. Wiley, New York, VII, 305-338. CHURCH, W.R. & STEVENS, R.K. 1971. Early Paleozoic ophiolite complexes of the Newfoundland Appalachians as mantle-ocean crust sequences. Journal of Geophysical Research, 16, 1460-1466. CLOOS, M. 1993. Lithospheric buoyancy and collisional orogenesis: subduction of oceanic plateaus, continental margins, island arcs, spreading ridges, and seamounts. Geological Society of America Bulletin, 105, 715-737. COLE, M.R. & ARMENTROUT, J.M. 1979. Neogene paleogeography of the western United States. In: ARMENTROUT, J.M., COLE, M.R. & TERBEST, H. JR (eds) Cenozoic Paleogeography of the Western United States. Pacific Section, Society of Economic Paleontologists and Mineralogists Pacific Coast Paleogeography Symposium, 3, 297-324. COLEMAN, R.G. 1971. Plate tectonic emplacement of upper mantle peridotites along continental edges. Journal of Geophysical Research, 76, 1212-1222. COLEMAN, R.G. 2000. Prospecting for ophiolites along the California continental margin. In: DILEK, Y.D., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds)

E M P L A C E M E N T OF AN O P H I O L I T E Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 351-364. COLEMAN, R.G. & LANPHERE, M.A. 1971. Distribution and age of high-grade blueschists, associated eclogites, and amphibolites from Oregon and California. Geological Society of America Bulletin, 82, 2397-2412. COLLOT, J.-Y., LAMARCHE, G., DELTEIL, J., WOOD, R., SOSSON, M., COFFIN, M. & LEBRUN, J.-F. 1995. Morphostructure of an incipient subduction zone along a transform plate boundary: Puysegur Ridge and Trench. Geology, 23, 519-523. COOPER, P. & TAYLOR, B. 1987. Seismotectonics of New Guinea: a model for arc reversal following arc-continent collision. Tectonics, 6, 53—67. COWAN, D.S., BOYER, S.E. & TILL, A.B. 1989. Contraction—not extension—at convergent margins elevates high—P rocks and covers them with low-P lids. Geological Society of America, Abstracts with Programs, 21(6), 216. COWAN, D.S., BRANDON, M.T. & GARVER, J.I. 1997. Geologic tests of hypothesis for large coastline displacements—a critique illustrated by the Baja— British Columbia controversy. American Journal of Science, 297, 117-173. CRAWFORD, K.E., 1975. The geology of the Franciscan tectonic assemblage near Mount Hamilton, California. PhD thesis, University of California, Los Angeles. DACZKO, N.R., WERTZ, K.L., MOSHER, S. & COFFIN, M.F. 2002. Extension along the transpressional plate boundary near Macquarie Island. EOS Transactions, American Geophysical Union, Fall Supplement, 83(47), F1358. DERCOURT, J., ZONENSHAIN, L.P. & Ricou, L.E. ET AL. 1986. Geological evolution of the Tethys belt from the Atlantic to the Pamirs since the Lias. Tectonophysics, 123, 241-315. DEWEY, J.F. 1976. Ophiolite obduction. Tectonophysics, 31, 93-120. DEWEY, J.F. & BIRD, J.M. 1970. Mountain belts and the new global tectonics. Journal of Geophysical Research, 75, 2625-2647. DEWEY, J.F. & BIRD, J.M. 1971. Origin and emplacement of the ophiolite suite: Appalachian ophiolites in Newfoundland. Journal of Geophysical Research, 76, 3179-3206. DICKINSON, W.R. 1970. Relations of andesites, granites and derivative sandstones to arc-trench tectonics. Reviews of Geophysics and Space Physics, 8, 813-860. DICKINSON, W.R., SCHWEICKERT, R.A. & INGERSOLL, R.D. 1996. Coast Range ophiolite as backarc/ interarc basin limosphere. GSA Today, 6(2), 2-3. DILEK, Y. 2003. Ophiolite concept and its evolution. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 1-16. DILEK, Y. & MOORES, E.M. 1990. Regional tectonics of the eastern Mediterranean ophiolites. In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues.

443

Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 295-309. DILEK, Y., MOORES, E.M. & FURNES, H. 1998. Structure of modern oceanic crust and ophiolites and implications for faulting and magmatism at oceanic spreading centers. In: BUCK, R., KARSON, J., DELANEY, P. & LAGABRIELLE, Y. (eds) Faulting and Magmatism at Mid-ocean Ridges. American Geophysical Union Monograph, 106, 219-266. DILEK, Y., THY, P., MOORES, E.M. & GRUNDVIG, S. 1990a. Late Paleozoic-early Mesozoic oceanic basement of a Jurassic arc terrane in the northwestern Sierra Nevada, California. In: HARWOOD, D. & MILLER, M.M. (eds) Paleozoic and Early Mesozoic Paleogeographic Relations; Sierra Nevada, Klamath Mountains, and Related Terranes. Geological Society of America, Special Papers, 255,351-370. DILEK, Y., THY, P., MOORES, E.M. & RAMSDEN, T.W. 1990b. Tectonic evolution of the Troodos ophiolite within the Tethyan framework. Tectonics, 9, 811-823. DILEK, Y., THY, P. & MOORES, E.M. 1991. Episodic dike intrusions in the northwestern Sierra Nevada, California: implications for multistage evolution of a Jurassic arc terrane. Geology, 19, 180-184. DILEK, Y., THY, P., HACKER, B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and mafic dike intrusions (Turkey): implications for the Neo-Tethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. DUNCAN, R.A. & VARNE, R. 1988. The age and distribution of the igneous rocks of Macquarie Island. Papers and Proceedings of the Royal Society of Tasmania, 114, 45-50. ERNST, W.G. 1971. Metamorphic zonations on presumably subducted lithospheric plates from Japan, California, and the Alps. Contributions to Mineralogy and Petrology, 34, 43-59. ERNST, W.G. 1988. Tectonic history of subduction zones inferred from retrograde blueschist P-T paths. Geology, 16, 1081-1084. FORSYTHE, R.D. & NELSON, E.P. 1985. Geological manifestations of ridge collision: evidence from the Golfo de Penas-Taitao basin, southern Chile. Tectonics, 4, 477-495. FORSYTHE, D. & UYEDA, S. 1975. On the relative importance of the driving forces of plate motion. Geophysical Journal of the Royal Astronomical Society, 43, 163-200. FROHLICH, C., COFFIN, M.F. & MASSELL, C. ET AL. 1997. Constraints on Macquarie Ridge tectonics provided by Harvard focal mechanisms and teleseismic earthquake locations. Journal of Geophysical Research, 102, 5029-5041. FRYER, P., LOCKWOOD, J.P., BECKER, N., PHIPPS, S. & TODD, C.S. 2000. Significance of serpentine mud volcanism in convergent margins. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349,35-51. GASS, I.G. 1968. Is the Troodos massif of Cyprus a

444

J. WAKABAYASHI & Y. DILEK

fragment of Mesozoic ocean floor? Nature, 221, 926-930. GNOS, E. 1998. Peak metamorphic conditions of garnet amphibolites beneath the Semail ophiolite: implications for an inverted pressure gradient. International Geology Review, 40, 281-304. GODFREY, NJ. & DILEK, Y. 2000. Mesozoic assimilation of oceanic crust and island arc into the North American continental margin in California and Nevada: insights from geophysical data. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 365-382. GODFREY, NJ. & KLEMPERER, S.L. 1998. Ophiolitic basement to a forearc basin and implications for continental growth: the Coast Range/Great Valley ophiolite. Tectonics, 17, 558-570. GOSCOMBE, B.D. & EVERARD, J.L. 2001. Tectonic evolution of Macquarie Island: extensional structures and block rotations in oceanic crust. Journal of Structural Geology, 23, 639-673. GRADSTEIN, P.M. ET AL. 1995. A Triassic, Jurassic, and Cretaceous time scale. In: BERGGREN, W.A., KENT, D.V., AUBRY, M.-P. & HARDENBOL, J. (eds) Geochronology Time Scales and Global Stratigraphic Correlation. SEPM Special Publication 54, 95-126. GRAY, D.R. & GREGORY, R.T. 2003. Ophiolite obduction and the Samail Ophiolite: the behaviour of the underlying margin. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 449-465. GRAY, D.R., GREGORY, R.T. & MILLER, J.McL. 2000. A new structural profile along the Muscat-Ibra transect, Oman: implications for emplacement of the Samail ophiolite. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 513-523. GREGORY, R.T., GRAY, D.R. & MILLER, J.McL. 1998. Tectonics of the Arabian margin associated with emplacement of the Oman margin along the Ibra transect: new evidence from northeast Saih Hatat. Tectonics, 17, 657-670. HACKER, B.R. 1990. Simulation of the metamorphic and deformational history of the metamorphic sole of the Oman ophiolite. Journal of Geophysical Research, 95, 4895-4907. HACKER, B.R. 1991. The role of deformation in the formation of metamorphic field gradients: ridge subduction beneath the Oman ophiolite. Tectonics, 10, 455-473. HACKER, B.R. & GNOS, E. 1997. The conundrum of Samail: explaining the metamorphic history. Tectonophysics, 279, 215-226. HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman ophiolite: thermal and geochronologic constraints. Tectonics, 15, 1230-1247. HALL, R. 1996. Reconstructing Cenozoic Southeast Asia.

In: HALL, R. & BLUNDELL, D. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 153-184. HARRIS, R. 2003. Geodynamic patterns of Ophiolites and marginal basins in the Indonesian and New Guinea regions. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 481-505. HAWKINS, J.W., BLOOMER, S.H., EVANS, C.A. & MELCHIOR, J.T. 1984. Evolution of intra-oceanic arctrench systems. Tectonophysics, 102, 216-229. HOPSON, C.A., MATTINSON, J.M. & PESSAGNO, E.A. JR 1981. The Coast Range ophiolite, western California. In: ERNST, W.G. (ed.) The Geotectonic Development of California. Rubey Volume L Prentice-Hall,, Englewood Cliffs, NJ, 419-510. HOPSON, C.A., PESSAGNO, E.A. JR & MATTINSON, J.M. ET AL. 1996. Coast Range ophiolite as paleoequatorial mid-ocean lithosphere. GSA Today, 6(2), 3-4. INGERSOLL, R.V. 2000. Models for origin and emplacement of Jurassic ophiolites of northern California. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 395-402. IRWIN, W.P. 1977. Ophiolitic terranes of California, Oregon, and Nevada. In: COLEMAN, R.G. & IRWIN, W.P. (eds) North American Ophiolites. Oregon Department of Geology and Mineral Industries Bulletin, 19, 75-92. JAMIESON, R.A. 1980. Formation of metamorphic aureoles beneath ophiolites—evidence from the St. Anthony Complex, Newfoundland. Geology, 8, 150-154. JAMIESON, R.A. 1986. PT paths from high temperature shear zones beneath ophiolites. Journal of Metamorphic Geology, 4, 3-22. JENNER, G.A., DUNNING, G.R., MALPAS, J., BROWN, M. & BRACE, T. 1991. Bay of Islands and Little Port complexes, revisited: age, geochemical and isotopic evidence confirm supra subduction zone origin. Canadian Journal of Earth Sciences, 28, 1635-1652. JUTEAU, T. 1980. Ophiolites of Turkey. Ofioliti, 2, 199-238. KAEDING, M., FORSYTHE, R.D. & NELSON, E.P. 1990. Geochemistry of the Taitao ophiolite and neartrench intrusions from the Chile Margin Triple Junction. Journal of South American Earth Sciences, 3, 161-177. LAGABRIELLE, Y., LE MOIGNE, J., MAURY, R.C., COTTEN, J. & BOURGOIS, J. 1994. Volcanic record of the subduction of an active spreading ridge, Taitao Peninsula (southern Chile). Geology, 22, 515-518. LAGABRIELLE, Y., GUIVEL, C., MAURY, R., BOURGOIS, J., FOURCADE, S. & MARTIN, H. 2000. Magmatictectonic effects of high thermal regime at the site of active ridge subduction: the Chile triple junction model. Tectonophysics, 326, 255—268. LEMOIGNE, J., LAGABRIELLE, Y., WHITECHURCH, H., GIRARDEAU, J., BOURGOIS, J. & MAURY, R.C. 1996. Petrography and geochemistry of the Ophiolitic and volcanic suites of the Taitao peninsula-Chile triple

EMPLACEMENT OF AN OPHIOLITE junction area. Journal of South American Earth Sciences, 9, 43-58. LYTWYN, J., CASEY, J., GILBERT, S. & KUSKY, T. 1997. Arc-like mid-ocean ridge basalt formed seaward of a trench-forearc system just prior to ridge subduction: an example from subaccreted ophiolites in southern Alaska. Journal of Geophysical Research, 102, 10225-10243. MALPAS, J. 1979. The dynamothermal aureole of the Bay of Islands ophiolite suite. Canadian Journal of Earth Sciences, 16, 2086-2101. MARSHAK, R.S. & KARIG, D.E. 1977. Triple junctions as a cause for anomalously near-trench igneous activity between the trench and volcanic arc. Geology, 5, 233-236. MATTINSON, J.M. & HOPSON, C.A. 1992. U-Pb ages of the Coast Range ophiolite: a critical reevaluation based on new high-precision Pb-Pb ages. American Association of Petroleum Geologists Bulletin, 76, 425. McCAiG, A. 1983. P—T conditions during emplacement of the Bay of Islands ophiolite complex. Earth and Planetary Science Letters, 63, 459-473. MILLER, J.McL., GREGORY, R.T., GRAY, D.R. & FOSTER, D.A. 1999. Geological and geochronologic constraints on the exhumation of a high-pressure metamorphic terrane, Oman. In: RING, U., BRANDON, M.T., LISTER, G.S. & WILLETT, S.D. (eds) Exhumation Processes: Normal Faulting, Ductile Flow and Erosion. Geological Society, London, Special Publications, 154, 241-260. MOORE, D.E. 1984. Metamorphic history of a high-grade blueschist exotic block from the Franciscan Complex, California. Journal of Petrology, 25, 126-150. MOORE, D.E. & Liou, J.G. 1980. Detrital blueschist pebbles from Franciscan metaconglomerates of the northeast Diablo Range, California. American Journal of Science, 280, 249-264. MOORES, E.M. 1970. Ultramafics and orogeny, with models of the U.S. Cordillera and the Tethys. Nature, 228, 837-842. MOORES, E.M. 1982. Origin and emplacement of ophiolites. Reviews of Geophysics and Space Physics, 20, 735-760. MOORES, E.M. & DAY, H.W. 1984. Overmrast model for the Sierra Nevada. Geology, 12, 416-419. MOORES, E.M. & VINE, FJ. 1971. Troodos massif, Cyprus, and other ophiolites as oceanic crust: evaluation and implications. Philosophical Transactions of the Royal Society of London, Series A, 268, 443-466. MOORES, E.M., DILEK, Y. & WAKABAYASHI, J. 1999. California terranes. In: SLOAN, D., MOORES, E.M. & STOUT, D. (eds) Classic Cordilleran Concepts: a View From California. Geological Society of America, Special Papers, 338, 227-234. MOORES, E.M., KELLOGG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency': a resolution of the 'ophiolite conundrum'. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 3-12.

445

MUELLER, S. & PHILLIPS, RJ. 1991. On the initiation of subduction. Journal of Geophysical Research, 96, 651-665. NELSON, E., FORSYTHE, R., DIEMER, J., ALLEN, M. & URBINO, O. 1993. Taitao ophiolite: a ridge collision ophiolite in the forearc of southern Chile (46°S). Reviews of the Geology of Chile, 20, 137-165. NICOLAS, A. & LE PICHON, X. 1980. Thrusting of young lithosphere in subduction zones with special reference to structures in ophiolitic peridotites. Earth and Planetary Science Letters, 46, 397-406. NILSEN, T.H. & McKEE, E.H. 1979. Paleogene paleogeography of the western United States. In: ARMENTROUT, J.M., COLE, M.R. & TfiRBfiST, H. JR

(eds) Cenozoic Paleogeography of the Western United States. Pacific Section, Society of Economic Paleontologists and Mineralogists Pacific Coast Paleogeography Symposium, 3, 257-276. OXBURGH, E.R. 1972, Flake tectonics and continental collision. Nature, 239, 202-204. PARSON, L.M. & WRIGHT, I.C. 1996. The Lau-HavreTaupo back-arc basin; a southward propagating, multi-stage evolution from rifting to spreading. Tectonophysics, 263, 1-22. PEACOCK, S.M. 1988. Inverted metamorphic gradients in the westernmost Cordillera. In: ERNST, W.G. (ed.) Metamorphism and Crustal Evolution of the Western United States. Rubey Volume VII. PrenticeHall, Englewood Cliffs, NJ, 954-975. PEARCE, J.A. 1975. Basalt geochemistry used to investigate past tectonic environments on Cyprus. Tectonophysics, 25, 41-67. PEARCE, J.A., LIPPARD, SJ. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology: Volcanic and Associated Sedimentary and Tectonic Processes in Modern and Ancient Marginal Basins. Geological Society, London, Special Publications, 16, 77-96. Penrose Conference Participants 1972. Penrose field conference: Ophiolites. Geotimes, 17(12), 24-25. PHIPPS, S.P. 1984. Ophiolitic olistostromes in the basal Great Valley sequence, Napa County, northern California Coast Ranges. In: RAYMOND, L. (ed.) Melanges: their Nature, Origin, and Significance. Geological Society of America, Special Papers, 198, 103-125. PLATT, J.P. 1975. Metamorphic and deformational processes in the Franciscan Complex, California: some insights from the Catalina Schist terrane. Geological Society of America Bulletin, 86, 1337-1347. PLATT, J.P. 1986. Dynamics of orogenic wedges and the uplift of high-pressure metamorphic rocks. Geological Society of America Bulletin, 97, 1037-1053. RAUTENSCHLEIN, M., TENNER, G.A., HERTOGEN, J., HOFMANN, A.W., KERRICH, R., SCHMINCKE, H.-U. & WHITE, W.M. 1985. Isotopic and trace element composition of volcanic glasses from the Akaki Canyon, Cyprus: implications for the origin of the Troodos ophiolite. Earth and Planetary Science Letters, 75, 369-383. REINHARDT, B.M. 1969. On the genesis and em-

446

J. WAKABAYASHI & Y. DILEK

placement of ophiolites in the Oman Mountains geosyncline. Schweizerische Mineralogische und Petrographische Mitteilungen, 49, 1 -30. RING, U. & BRANDON, M.T. 1994. Kinematic data for the Coast Range fault and implications for exhumation of the Franciscan subduction complex. Geology, 22, 735-738. ROBERTSON, A.H.F. 1998. Mesozoic-Tertiary tectonic evolution of the easternmost Mediterranean area: integration of marine and land evidence. In: ROBERTSON, A.H.F., EMEIS, K.-C, RICHTER, C. & CAMERLENGHI, A. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 160. Ocean Drilling Program, College Station, TX, 723-782. ROBINSON, P.T. & MALPAS, J. 1990. The Troodos ophiolite of Cyprus: new perspectives on its origin and emplacement. In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues. Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 13-26. ROBINSON, P.T., MELSON, W.G., O'HEARN, T. & SCHMINCKE, H.-U. 1983. Volcanic glass compositions of the Troodos ophiolite Cyprus. Geology, 11, 400-404. RUFF, L.J., GIVEN, J.W., SANDERS, C.O. & SPERBER, C.M. 1989. Large earthquakes in the Macquarie Ridge Complex: transitional tectonics and subduction initiation. Pure and Applied Geophysics, 128, 72-129. SALEEBY, J.B. 1990. Geochronologic and tectonostratigraphic framework of Sierran—Klamath ophiolitic assemblages. In: HARWOOD, D.S. & MILLER, M.M. (eds) Paleozoic and Early Mesozoic Paleogeographic Relations; Sierra Nevada, Klamath Mountains and Related Terranes. Geological Society of America, Special Papers, 255, 93-114. SALEEBY, J.B. 1992. Petrotectonic and paleogeographic settings of U.S. Cordilleran ophiolites. In: BURCHFIEL, B.C. (ed.) The Cordilleran Orogen: Conterminous U.S.. Geological Society of America, Geology of North America, G-3, 632-682. SALEEBY, J.B. 1996. Coast Range ophiolite as parautochthonous forearc lithosphere. GSA Today, 6(2), 6-10. SCHWEICKERT, R.A. & COWAN, D.S. 1975. Early Mesozoic tectonic evolution of the western Sierra Nevada, California. Geological Society of America Bulletin, 86, 1329-1336. SEARLE, M.P. & Cox, J. 1999. Tectonic setting, origin, and obduction of the Oman ophiolite. Geological Society of America Bulletin, 111, 104-122. SEARLE, M.P. & STEVENS, R.K. 1984. Obduction processes in ancient, modern, and future ophiolites. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 303-320. SEARLE, M.P., WARREN, C.J., WATERS, D.J. & PARRISH, R.R. 2003. Subduction zone polarity in the Oman Mountains: implications for ophiolite emplacement. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 467-580.

SHERVAIS, J.W. 1990. Island arc and ocean crust ophiolites: contrasts in the petrology, geochemistry and tectonic style of ophiolite assemblages in the California Coast Ranges. In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C (eds) Ophiolites, Oceanic Crustal Analogues. Proceedings of the Symposium 'Troodos 1987':. Geological Survey Department, Nicosia, 507-520. SHERVAIS, J.W. & KIMBROUGH, D.L. 1985. Geochemical evidence for the tectonic setting of the Coast Range ophiolite: a composite island arc-oceanic crust terrane in western California. Geology, 13, 35-38. SISSON, V.B. & PAVLIS, T.L. 1993. Geologic consequences of plate reorganization: an example from the Eocene southern Alaska forearc. Geology, 21, 913-916. SMITH, A.G. & WOODCOCK, N.H. 1976. Emplacement model for some 'Tethyan' ophiolites. Geology, 4, 653-656. SPRAY, J.G. 1984. Possible causes and consequences of upper mantle decoupling and ophiolite displacement. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 255-268. STERN, RJ. & BLOOMER, S.H. 1992. Subduction zone infancy; examples from the Eocene Izu-BoninMariana and Jurassic California arcs. Geological Society of America Bulletin, 104, 1621-1636. TEMPLE, P.G. & ZIMMERMAN, J. 1969. Tectonic significance of Alpine ophiolites in Greece and Turkey. Geological Society of America, Abstracts with Programs, Part, 1, 221-222. THORKELSON, D.J. 1996. Subduction of diverging plates and the principles of slab window formation. Tectonophysics, 255, 47-63. UMINO, S., YANAI, S., JAMAN, A.R., NAKAMURA, Y. & IIYAMA, J.T. 1990. The transition from spreading to subduction: evidence from the Semail ophiolite, Northern Oman Mountains. In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues. Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 375-384. VAN DEN BEUKEL, J. & WORTEL, R. 1992. Ridge-trench interaction: a possible mechanism for ophiolite emplacement. Ofioliti, 17, 141-154. VARNE, R., GEE, R.D. & QUILTY, PJ. 1969. Macquarie Island and the cause of oceanic linear magnetic anomalies. Science, 166, 230-232. VARNE, R., BROWN, A.V. & FALLOON, T. 2000. Macquarie Island: its geology, structural history, and the timing and tectonic setting of its N-MORB to E-MORB magmatism. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 301-320. VON HUENE, R. 1986. To accrete or not to accrete, that is the question. Geologische Rundschau, 75, 1-15. WAKABAYASHI, J. 1990. Counterclockwise P-T-t paths from amphibolites, Franciscan Complex, California: relics from the early stages of subduction zone metamorphism. Journal of Geology, 98, 657-680.

EMPLACEMENT OF AN OPHIOLITE WAKABAYASHI, J. 1992. Nappes, tectonics of oblique plate convergence, and metamorphic evolution related to 140 million years of continuous subduction, Franciscan Complex, California. Journal of Geology, 100, 19-40. WAKABAYASHI, J. 1999. Subduction and the rock record: concepts developed in the Franciscan Complex, California. In: SLOAN, D., MOORES, E.M. & STOUT, D. (eds) Classic Cordilleran Concepts: a View From California. Geological Society of America, Special Papers, 338, 123-133. WAKABAYASHI, J. & DILEK, Y. 2000. Spatial and temporal relations between ophiolites and their subophiolitic soles: a test of models of forearc ophiolite genesis. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 53—64. WAKABAYASHI, J. & UNRUH, J.R. 1995. Tectonic wedging, blueschist metamorphism, and exposure of

447

blueschist: are they compatible? Geology, 23, 85-88. WILLIAMS, H. & SMYTH, R. 1973. Metamorphic aureoles beneath ophiolite suites and Alpine peridotites: tectonic implications with west Newfoundland examples. American Journal of Science, 273, 594-621. WIRTH, K.R., BIRD, J.M., BLYTHE, A.E., HARDING, DJ. & HEIZLER, M.T. 1993. Age and evolution of the Brooks Range ophiolites, Alaska: results from 40Ar/39Ar thermochronometry. Tectonics, 12, 410-432. YILMAZ, Y., TUYSUZ, O., YIGITBAS, E., GENC, S.C. & SENGOR, A.M.C. 1997. Geology and tectonic evolution of the Pontides. In: ROBINSON, A.G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. Memoirs, American Association of Petroleum Geologists, 68, 183-226. ZELLMER, K.E. & TAYLOR, B. 2001. A three-plate kinematic model for Lau Basin opening. Geochemistry, Geophysics, Geosystems, 2, Paper No. GC000106.

This page intentionally left blank

Ophiolite obduction and the Samail Ophiolite: the behaviour of the underlying margin D. R. GRAY 1 & R. T. GREGORY 2 i yjEps School of Earth Sciences, University of Melbourne, Vic. 3010, Australia (e-mail: drgray @unimelb. edu.au) 2 Department of Geological Science, Southern Methodist University, Dallas, TX 75275, USA Abstract: Samail Ophiolite emplacement has become a type-example for ophiolite obduction. Despite this, problems and controversy remain with respect to the age of high-P metamorphism, the vergence of structures in deformed rocks beneath the ophiolite, the suprasubduction character of the ophiolite and the P—T conditions of the metamorphic sole. The presence of major, regional-scale NE-facing isoclinal folds and SW- and west-dipping shear zones with topto-the-NE shear sense in Arabian margin rocks beneath the Samail Ophiolite nappe requires that the margin was not simply passively overridden during obduction of the ophiolite. The development of these folds at 72-76 Ma is at the time the ophiolite is emplaced finally onto the margin (80-70 Ma), accompanied by development of a major shear zone in Saih Hatat (the upper plate-lower plate discontinuity described by earlier workers) at c. 82-80 Ma. Structural scenarios that incorporate these folds and shear zones include lateral escape from a rising buoyant crustal slice along the former subduction interface, back-folding ('retrocharriage') associated with major oceanwards-directed underthrusting, or simple undermrasting of the margin by the oceanic realm. Previous models involving craton-directed overthrusting with domal culminations related to deep-seated footwall and lateral ramps are more applicable to the Tertiary structure and Tertiary evolution of the mountain range. Oman ophiolite obduction clearly involves ocean-vergent thrusting within the continental margin platform to slope facies sequences.

Emplacement of allochthonous oceanic crust onto continental margins (obduction, Coleman 1971) is part of the process of ocean closure in plate tectonics (Dewey 1976). How obduction takes place with respect to the nature of the forces involved and in what tectonic setting still remains problematical (e.g. Coleman 1971; Dewey & Bird 1971; Dewey 1976; Moores 1982; Nicolas 1989; Dilek et al 2000). The question also remains of how large, thick, dense ophiolite slabs can be emplaced over lower-density continental crust (e.g. Hacker et al. 1996). Tectonic settings have been speculated to range from Atlantic-type ocean basins with either collapse off the inflated spreading ridge (Fig. la) or intra-oceanic thrusting at or near the spreading centre (Fig. Ib) to cause the oceanic lithosphere to overrun the nearby continental margin. Other scenarios involve subduction, either with subduction of the continental margin causing the overriding oceanic plate to overrun the margin (Fig. Ic), or collision of the continental margin with an island arc (Fig. Id). Ophiolitic crust with suprasubduction affinities may require subduction beneath a spreading ridge system (Fig. le). To generate the forces necessary to rupture oceanic lithosphere, collision of an arc

(Gealey 1977,1979; Searle & Stevens 1984; Mitchell 1986) or subduction zone 'lock-up', as a result of attempted subduction of continental margins or continental crustal fragments, is necessary (see Mueller & Phillips 1991). From a structural viewpoint, Tethyan ophiolite obduction (Moores 1982) involves closure of a marginal ocean basin by thrusting of an external oceanic domain over and towards a passive continental margin (Fig. 2). Implicit in such models is the continent-vergent, piggy-back style, overthrusting of the overridden ocean basinal facies and margin. The Samail Ophiolite of the Sultanate of Oman on the Arabian Peninsula (Fig. 3) is the best-exposed example of allochthonous ocean lithosphere emplaced onto continental crust. It has become a type-example for obduction. This paper first reviews obduction models for the Samail Ophiolite nappe, and then provides structural data from rocks beneath the Samail nappe suggesting that the obduction process may not be as simple as originally defined. In Oman, there is clear evidence that the margin has been involved in oceanward-directed thrusting and overfolding at the time of ophiolite emplacement onto the margin.

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 449-465. 0305-8719/037$ 15 © The Geological Society of London 2003.

450

D. R. GRAY & R. T. GREGORY

Fig. 1. Schematic models of speculative obduction scenarios in various tectonic settings (modified from Dewey 1976, figs 4 and 5): (a) collapse from a spreading centre; (b) overthrusting at spreading centre; (c) subduction of continental margin; (d) arc-continent collision as a result of subduction of the margin; (e) combination of model scenarios (b) and (c) with subduction of continental margin triggering overthrusting at spreading centre. Black, oceanic crust; slashed regions, continental lithosphere; 'v's, volcanic arc.

Fig. 2. Thrust-stacking model of Tethyan-type obduction where the structurally highest unit (oceanic lithosphere) is the furthest travelled thrust-sheet (from Moores & Twiss 1995, fig. 9.25). (a) Pre-thrasting configuration, (b) Thrust stack with order dependent on original position in the palaeogeography.

The implications of this are investigated with respect to the general mechanism of obduction that requires continent-vergent overthrusting.

Background Structure of the Oman Mountains The Oman Mountains are a c. 750 km long arcuate mountain chain, parallel to the Batinah coast, with local elevations up to 3000 m. Made up primarily of

the obducted Cretaceous Samail Ophiolite nappe, the major Hawasina, Jabal Akhdar and Saih Hatat windows expose the overridden rocks beneath the ophiolite (Fig. 3). These include Permian-Mesozoic ocean floor and continental rise sedimentary rocks (Haybi and Hawasina units), which in turn overlie time-equivalent Permian-Mesozoic shelf and slope facies carbonates (Hajar and Sumeini Groups) (Fig. 4). In an eastward direction, the windows in turn expose deeper structural levels of the Oman Mountains. The westernmost Hawasina window shows fold structures within the deeper water Hawasina units immediately beneath the ophiolite, and slope carbonates of the margin. The central part, the Jabal Akhdar window, exposes the upper limb of a regional nappe structure in shelf carbonates that is fully exposed within the upper part of the Saih Hatat window (see Le Metour et al. 1990). The easternmost Saih Hatat window preserves the deepest structural levels within the Oman Mountains, where a major shear zone (upper plate-lower plate discontinuity) is exposed within the northern part of the Saih Hatat window (Gregory et al, 1998; Miller et al 1998, 1999, 2002). This shear zone separates the higher-P blueschist and eclogite of the HuwlMeeh and As Sifah sub-windows, respectively (Figs 5 and 6) from less deformed and metamorphosed upper-plate rocks involving nappes developed in the pre-Permian basement and Hajar Supergroup (see Gregory et al 1998; Miller et al 1998; Gray et al 2000). Topographically, the Oman Mountains are dominated by the Saih Hatat and Jabal Akhdar domal

OBDUCTION OF THE SAMAIL OPHIOLITE

451

Fig. 3. Geological map of the Oman Mountains (modified from Glennie et al. 1974), showing the distribution of the Samail Ophiolite, the positions of windows containing pre-Permian basement, and the locations of sections A-A' (Fig. 7) and B-B'(Fig. 10), and Saih Hatat map (Fig. 5).

Fig. 4. Pre-thrusting reconstruction of the Oman margin showing restoration of the various facies (from Robertson & Searle, 1990, fig. 11) deduced from assumptions implied in Figure 2.

452

D. R. GRAY & R. T. GREGORY

Fig. 5. Maps of Saih Hatat. (a) Geological map of the Saih Hatat domal culmination (modified from Le Metour et al. 1986), showing the axial trace of the Saih Hatat fold-nappe, the trace of the upper plate-lower plate discontinuity, the locations of the Huwl-Meeh and As Sifah subwindows, and the axial traces of both early and late folds, (b) Simplified geological map with superimposed key metamorphic indicator minerals or assemblages from Goffe et al. (1988, fig. 3).

OBDUCTION OF THE SAMAIL OPHIOLITE

453

Fig. 6. Landsat 5 image of NE Saih Hatat showing the Huwl-Meeh and As Sifah subwindows defined by domal culminations within the upper plate-lower plate discontinuity (bold line). Axial surface traces of major regional folds are shown (compare with Fig. 5a). Locations of Wadi Adai (see Fig. 8c) and Wadi Meeh (see Fig. 8b and h) are shown. Symbols are as for Figure 5a.

culminations, interpreted to be ramp-related anticlines and/or antiformal stacks in basement duplexes (Searle 1985; Goffe et al 1988, fig. 2; Cawood et al 1990; Mann & Hanna 1990; Mount et al 1998) (Fig. 7a and b). The dome structures were identified as Tertiary fold interference structures by Glennie et al. (1974), as these folds affect Tertiary limestone units in the NE corner of Saih Hatat and within the Ibra ophiolite block, as well as ophiolite and basement rocks. Apatite fission-track data from the Precambrian Mistal Formation of the Jabal Akdah window limit the uplift history of this part of the Oman Mountains to the Oligocene (Mount et al. 1998).

Previous ideas on obduction of the Samail Ophiolite Emplacement of the Samail Ophiolite nappe has been related to overthrusting and/or gravity-

induced emplacement mechanisms. Two contrasting tectonic settings involving spreading and/or subduction have been proposed (see Hacker et al. 1996): a 'ridge' model (Boudier & Michard 1981; Coleman 1981, fig. 6; Boudier et al 1982, 1985; Montigny et al 1988; Nicolas et al 1988; Hacker 1991, 1994; Michard et al 1991; El-Shazly et al 2001) where young, hot oceanic crust adjacent to the spreading centre was thrust over the subducting lithosphere onto the Arabian margin (e.g. Fig. la), versus an 'arc' model (Gealey 1977; Pearce et al 1981; Searle & Malpas 1980; Lippard et al 1986; Searle & Cox 1999) where the oceanic crust originated in a back-arc region (Fig. Id) or as an intraoceanic arc and was also thrust onto the margin over subducting oceanic lithosphere. Some have the intra-oceanic thrust initiating subduction of the continental margin (e.g. Searle & Cox 1999, fig. 13; El-Shazly et al 2001, fig. 8).

454

D. R. GRAY & R. T. GREGORY

Recognition of the differences between the metamorphic sole and the present basal fault to the ophiolite slab has led to a two-stage emplacement model for the Samail Ophiolite (e.g. Coleman 1981, fig. 6; Hopson et al. 1981; El-Shazly & Coleman 1990; Hacker et al 1996; Gregory et al 1998). The metamorphic sole reflects ductile flow under amphibolite-facies metamorphic conditions during the intra-oceanic thrusting stage, whereas brittle extensional faults characterize the highlevel final emplacement of the ophiolite nappe along the present basal fault (Boudier et al. 1985, 1988; Gray & Gregory 2000). Early stage emplacement of the ophiolite has largely been linked with overthrusting scenarios involving piggy-back emplacement of successively lower thrust slices from an external oceanic domain towards the continental interior (see Searle 1985; Lippard et al. 1986; Bechennec et al. 1990; Hanna 1990; Mann & Hanna 1990). Thrusting has generally been linked to a subduction zone dipping away from the continental margin (e.g. Gealey 1977, 1979; Lippard et al 1986; Chemenda et al 1996; Searle & Cox 1999). The direction of the subduction zone dip has been argued on suprasubduction affinities of the ophiolite (e.g. Searle & Cox 1999, fig. 2), as well as on structural geometry (e.g. Chemenda et al 1996, fig. 8; Mattauer & Ritz 1996). Gray et al (2000), however, have argued based on structural grounds for underthrusting towards the margin for at least part of the history. All the observed ductile structures in both Saih Hatat and the Hawasina window show NE-directed thrusting and folding. During ophiolite emplacement there needs to be attenuation of the ophiolite with thinning of a 5515 km pseudostratigraphy within the mantle and crust sequences down to a <5 km thick sheet. Thinning of the ophiolite sheet seems to be associated with internal development of both lowand high-angle normal faults (Gray & Gregory 2000). These faults are brittle in character and clearly overprint or truncate the ductile fabrics preserved within the ophiolite and relicts of the metamorphic sole, respectively. Scenarios involve: (1) a rapidly rising arc producing a regional surface slope to induce down-surface slope stresses to emplace ophiolite thrust sheets (Elliott 1976, p. 960), (2) 'culmination collapse' with gravity-induced sliding of ophiolite slabs away from structural domes (Hanna 1986, 1990, p. 355; Cawood et al 1990; Coffield 1990, p. 447; Cawood 1991); (3) gravity-induced sliding of ophiolite slabs away from the rising NE-directed Saih Hatat fold-nappe, now preserved within the Saih Hatat window and offshore along the Batinah coast as the Saih Hatat axis (Gray & Gregory 2000); (4) crustal loading, flexure, and 'isostatic

rebound' (Shelton & Egan 1991); (5) buoyancyinduced collapse (Chemenda et al 1996; Gregory etal 1998). Constraints on the obduction process The Samail Nappe Size. The Samail Ophiolite nappe was a once contiguous sheet extending c. <400 km in length and up to 150km in width. It is now broken by cross-strike faults or by imbricate low-angle faults into approximately a dozen blocks (Nicolas et al 1988). These blocks have marked structural thinning and taper at their leading and trailing edges (Boudier et al 1985, 1988). Geometry. The present-day Samail nappe is thin (<5 km) relative to the 10 and possibly 20km thickness dictated by the ophiolitic pseudostratigraphy (e.g. Hopson et al 1981; Gregory 1984). In the northern part of the Oman Mountains, the intact oceanic stratigraphy of the ophiolite is folded into a simple antiformal structure (Glennie et al 1974; Nicolas et al 1988). In the southeastern part, the ophiolite nappe shows dome and basin fold interference structures (Gray & Gregory 2000, fig. 2). Metamorphism. There is a general lack of any high-P, low-r (temperature) metamorphism in any part of the preserved Samail nappe. The only secondary-mineral assemblages are the result of sea-floor hydrothermal alteration (e.g. Gregory & Taylor 1981). Basal fault. The basal fault to the Samail Ophiolite nappe is flat-lying and only weakly folded relative to the ophiolite in the hanging wall. It truncates the ophiolite pseudo-stratigraphy, crosscuts the phyllitic contact aureole blocks of the metamorphic sole, and truncates the early fold structures in the ophiolite (Gregory et al 1998, fig. 3; Gray et al 2000, fig. 5). Formerly defined as the Samail thrust (Rheinhardt in Glennie et al 1974), this fault shows characteristics more typical of extensional faults associated with the core complexes of western North America (e.g. Gregory et al 1998; Gray & Gregory 2000). Downcutting or excision along this Samail basal fault has led to previous extensional interpretations (e.g. Cawood etal. 1990). Metamorphic sole. The metamorphic sole consists of high-J garnet amphibolite-facies metamorphic rocks now present as dismembered remnants scattered along the present, brittle basal fault. Temperature and pressure estimates from the sole

OBDUCTION OF THE SAMAIL OPHIOLITE rocks range from 450 to 850 °C and from 4 to 15 kbar, respectively (Ghent & Stout 1981; Gnos 1998; Searle & Cox 1999). The presence of kyanite in the metamorphic sole (Asimah, northern Oman) suggests pressures > 11 kbar (Gnos 1998) associated with peak metamorphic temperatures of 450-550 °C on the basis of magnesiohornblende + albite assemblages (Bucher 1991). Most recently, Searle & Cox (2002) have argued for maximum temperatures of 800-860 °C and pressures of 10.5 ± 1 kbar to 14.7 ± 3 kbar for granulite-facies metamorphic sole rocks at Bani Hamid (United Arab Emirates). At the same locality garnet-pyroxene-hornblende-plagioclase amphibolites give temperatures of 840 ± 70 °C and pressure of 11.6 ± 1.6 kbar (THERMOCALC average P-T mode). These metamorphic conditions have been interpreted to represent the intraoceanic thrusting (stage 1) of ophiolite emplacement (Searle & Malpas 1980; Ghent & Stout 1981; Montigny et al 1988; Hacker 1990, 1991; Hacker et al. 1996; Hacker & Gnos 1997; Hacker & Mosenfelder 1997; Gnos 1998). Structurally the metamorphic sole consists of typically 100m scale sections of metamorphic rocks welded to the basal section of the peridotite of the ophiolite. A narrow zone (metre scale) of strongly to intensely deformed garnet amphibolites commonly overlies fault-bounded slices of amphibolite and greenschists to give an apparent inverted metamorphic gradient (e.g. Coleman 1977; Gnos 1998). The sole rocks within these slices are isoclinally folded and show sheath-folding and internal boudins at the mesoscopic scale (e.g. Mahlah slice, Wadi Tayin: Gray & Gregory 2000). Displacement. The present c. 150km outcrop width of the ophiolite requires at least 150 km displacement, whereas reconstruction of spreading-centre geometry based on orientations of sheeted dykes suggests a c. 200 km width of the palaeo-ocean (Boudier et al. 1988). Michard et al. 1991 and Hacker et al. (1996) suggested that the Samail nappe has been transported some 400 km.

Structure of the rocks beneath the ophiolite Saih Hatat window. The platform succession beneath the ophiolite in the Saih Hatat window (Fig. 3) is folded into a regional-scale, NE-facing, antiformal fold nappe cored by pre-Permian Hatat Schist basement and carbonates of the Hajar Supergroup (Fig. 5c). Apart from work by Le Metour et al. (1986, 1990), structural relationships to define the Saih Hatat fold nappe (see Miller et al. 2002) have not been identified previously and have led to schematic profiles across Saih Hatat

455

that are more likely to reflect the Tertiary doming and uplift of the Oman Mountains (e.g. Fig. 7a and b). The trace of the major anticlinal nappe can be identified by large areas of completely overturned sequences in the centre of the Saih Hatat window (e.g. overturned Ordovician Amdeh at Arqi, Fig. 5a). Strain within the fold-nappe increases downwards towards a major shear zone (the upper plate-lower plate discontinuity of Gregory et al. 1998; Miller et al. 1998, 2002, fig. 3), with a strong increase in fabric development, increasing pressure of metamorphism and development of isoclinal, sheath-like folds with markedly attenuated fold limbs (Fig. 8a and b). Kinematic indicators, such as shear bands, tails on porphyroclasts, and pressure shadows around porphyroblasts, require top to the NE-directed shear (Michard et al. 1984; Le Metour et al 1986, 1990; Gregory et al. 1998; Miller et al 1998, 2002). The major regional NE-facing anticlinal closure that dominates the Saih Hatat window (Fig. 7c) is made up of a series of large asymmetric recumbent closures (Fig. 8a) exposed on Jabal Qirmadhil, west of Wadi Adai (Fig. 8c and d). Below these is a major, SW-facing recumbent syncline (Fig. 8b and e) whose form is extremely attenuated in Saiq 1 Limestones exposed in the southwestern extremity of Wadi Meeh (Fig. 8a and b). The right-way-up lower limb of this fold is truncated by the upper plate-lower plate discontinuity (Fig. 8a). The hinge zone exposed in Wadi Meeh gorge shows spectacular recumbent fold structures with juxtaposed, oppositely facing closures (Fig. 8h) indicative of high strain and macroscale sheath-like folding. Upper plate-lower plate discontinuity. The upper plate-lower plate discontinuity is a brittle fault (Fig. 8f) marked by localized occurrences of PbZn and Cu mineralization that defines two ovoidshaped windows, the Huwl-Meeh and As Sifah subwindows (Figs 5 and 6) The fault truncates both stratigraphic and structural features of the overlying (upper) and underlying (lower) plates (Fig. 6). Strain gradients in the upper plate towards the discontinuity suggest that it is a major ductile shear zone with top-to-the-NE shear sense, overprinted by late-stage, brittle fault movement. Fold structures in upper plate units directly above the discontinuity have classic recumbent isoclinal form (Fig. 8g and i) reflecting the higher strain. Huwl and As Sifah subwindows. Lower plate rocks exposed in the Huwl-Meeh and As Sifah subwindows are intensely and multiply deformed with a pervasive schistosity (L-S tectonite) and sheathfolds visible at all scales. The calc-schist, mafic

456

D. R. GRAY & R. T. GREGORY

Fig. 7. Profiles across the Saih Hatat domal culmination. Line of section is A-A' of Figure 3. (a) Schematic interpretative profile from Hanna (1990, fig. 11); (b) schematic profile from Goffe et al (1988, fig. 2); (c) structural profile from Gray et al, (2000, fig. 2).

schist and quartz-mica schist tectonostratigraphy is folded into a series of tight to isoclinal sheathlike folds (Fig. 6) whose closures are subparallel to the pronounced stretching lineation. Shear bands, asymmetrically sheared clasts and pressure shadows around porphyroblasts (Miller et al. 2002, fig. 12) all indicate that the lower plate units have undergone intense, NE-directed non-coaxial shear. Disruption of the mafic schist layer into isolated boudins and mega-boudins suggests the presence of higher strain zones (shear zones) within this pervasively deformed lower plate. This is supported by differences in metamorphic assemblages and grade in units bounded by these zones (Fig. 9). Hawasina window. The slope (Sumeini) and deepwater (Hawasina) successions beneath the ophiolite in the Hawasina window (Fig. 3) are folded into regional-scale, tight to isoclinal folds that are inclined to the west (Fig. 10). The Sumeini Group

limestones crop out in doubly-plunging, domal NE-facing antiformal closures that are underlain by west-dipping, high-strain zones in calc-schist and pelite. Kinematic indicators such as tails on porphyroclasts indicate top-to-the-NE movement sense in the high-strain zones (Fig. lOc). Earlier profiles by Villey et al. (1986) (Fig. lOa) and Searle (1985) (Fig. lOb) also show that the dominant structures in the window are verging to the east.

Metamorphism of the rocks beneath the ophiolite The nature and conditions of metamorphism vary from window to window along the ophiolite belt. The Hawasina and Jabal Akhdar windows generally show low-grade metamorphism, but little metamorphic work has been undertaken here. The structurally lowest rocks exposed in the lower

Fig. 8. The Saih Hatat fold—nappe, (a) Downplunge structural profile of the northeastern part of the fold-nappe (from Miller 1998; Miller et al. 2002, fig. 2b) showing locations of outcrop photographs, (b) Profile along the southern segment of Wadi Meeh with enlargement showing the detail in Wadi Meeh gorge (composite profile) (from Miller et al, 2002, fig. 9). Ha, Hatat Schist; Hi, Hijam Dolomite; Amqms, Amdeh Quartzite; Sql, Saiq 1 Limestone; Lqms, lower plate quartz-mica schist, (c) View of folds on Jabal Qirmadhil and hills on the west side of Wadi Adai. (d) Recumbent, box-like SW-facing closure above Wadi Adai (see position in (c)). (e) Structurally lowest, recumbent SWfacing closure above Wadi Meeh (north end), (f) Upper plate-lower plate discontinuity exposed in Wadi Huwl. The schistose nature of the lower plate rocks should be noted, (g) Recumbent isocline in cliffs above Wadi Meeh, north of Rija. (h) Oppositely facing recumbent fold closures in the hinge of the major SW-facing fold closure, southern part of Wadi Meeh gorge, (i) Recumbent isoclines in Amdeh Quartzite, southern margin of Huwl-Meeh window.

458

D, R. GRAY & R. T. GREGORY

Fig. 9. Saih Hatat fold—nappe schematic profile showing enlargement with the metamorphic mineral indicator localities of Goffe et al. (1988, fig. 2) projected onto the line of section. HS, hatat Schist; UP and LP, upper and lowerplate, respectively; S21, S22, and S23, shear zones.

plate of the major Saih Hatat domal culmination preserve eclogite assemblages as part of glaucophane eclogites in mafic megaboudins. These meta-basites record minumum pressures of 1012 kbar and peak temperatures of 500-580 °C (ElShazly et al 1990), although Searle et al (1994) and Wendt et al (1993) have argued for pressures as high as 20 kbar. In the Saih Hatat window, major faults and shear zones separate rocks that record different P—T conditions (Fig. 9). The Saih Hatat fold-nappe (or upper plate of Gregory et al 1998) contains carpholite-lawsonite assemblages indicating P-T conditions of 7-9 kbar and 315-435 °C (El-Shazly 1994, 1995) and carpholite-kaolinite assemblages indicating 8-10 kbar and 180-250 °C (Goffe et al 1988) (Fig. 6b). Underlying this, or below the upper

plate-lower plate discontinuity (SZ3, Fig. 9b), are rocks of the lower plate Huwl subwindow recording 7-9 kbar and 450-520 °C (Goffe et al 1988), although El-Shazly et al (1990) argued that pressures were never above 4.5-5.5 kbar. Differences in sodic amphibole composition (El-Shazly & Coleman 1990; El-Shazly et al 1990; Miller 1998; Miller et al 1998) and the general lack of garnet or pyroxene in the Hulw-Meeh lower plate window (Fig. 6b) suggest that peak metamorphic conditions were different for the two subdomains (Huwl-Meeh and As Sifah subwindows) within the lower plate. This change coincides with a zone of disrupted metabasites (SZ2, Fig. 9b). The lowest shear zone (SZ1, Fig. 9b) contains the eclogite megaboudins exposed in Wadi As Sifah and along the coast north of As Sifah village.

OBDUCTION OF THE SAMAIL OPHIOLITE

459

Fig. 10. Profiles across the Hawasina domal culmination. Line of section is B-B' of Figure 3. (a) Schematic profile from Bechennec et al. (1990, fig. 3). (b) Schematic interpretative profile from Searle (1985, fig. 2). The structure in the subsurface beneath the window rocks, and also beneath the ophiolite are speculative, (c) Schematic structural profile based on field structural relationships by Gray & Gregory (unpubl. data). Approximate scale only. Sum, Sumeini Group limestone; Hy, Haybi Volcanics; Haw, Hawasina unit; H, back-thrusted Hawasina.

Age constraints The most important age constraints are the following. (1) Crystallization age of the Samail Ophiolite: 94-97 Ma based on U-Pb ages on zircons from plagiogranites intruding the ophiolite sequence (Tilton et al. 1981). (2) Metamorphism age of the metamorphic sole: c. 94-93 Ma based on hornblende 40Ar-

39 Ar ages from the hornblende amphibolites of the metamorphic sole (Hacker et al. 1996). (3) Saih Hatat upper plate deformation: 8070 Ma based on 40Ar-39Ar geochronology on phengites defining the axial surface foliation to the recumbent folds (Miller et al. 1999). (4) Age of eclogite-facies metamorphism (lower plate, As Sifah subwindow): 40Ar-39Ar geochronology or thermochronology (Montigny et al.

460

D. R. GRAY & R. T. GREGORY

1988; El Shazly & Lanphere 1992; Searle et al 1994; Miller et al 1999) typically gives ages on white mica of 131 ±4, 122 ± 2.4, 106.6 ± 0.6, 96 ± 2, 893 ± 1.8 and 80.4 ± 2.5 Ma. This is in agreement with K-Ar ages from garnet-glaucophane schists of 80 ± 2 and 100-101 ± 4 Ma (Lippard 1983). Clearly, some of these ages are older than the Samail Ophiolite, which implies some of the high-pressure metamorphism predates the formation and obduction of the oceanic crust, although Searle et al (1994) and more recently El-Shazly et al (2001) have attributed the old ages to excess argon. (5) Age of deformational (7 kbar) overprint on lower plate: (a) 86-79 Ma for progressive drop in metamorphic grade associated with the NE stretching (recording partial exhumation of the lower plate); (b) 79-70 Ma for transposition of the upper and lower plate high-pressure fabrics during a NE-directed shearing event with regional-scale recumbent folding (recording final exhumation and docking with upper plate) (based on 40 Ar-39Ar thermochronology of phengites; Miller etal 1999). (6) Arrival of ophiolite onto margin: 78-71 Ma with the presence of igneous detritus within the Mid- to Late Campanian Juweiza Formation (Glennie et al 1974; Warburton et al 1990). (7) Final emplacement of the ophiolite: c. 68 Ma, as Late Maastrichtian (c. 68-67 Ma) autochthonous sediments unconformably overlie all sequences. (8) Uplift of the Oman Mountains: Tertiary, post-early Miocene (Le Metour et al 1995) or Oligocene based on fission-track data (Mount etal. 1998).

Oman obduction revisited Four issues remain that have major implications for the obduction-emplacement of the Samail Ophiolite. These are: (1) age of the high-P metamorphism; (2) P-T conditions of the metamorphic sole; (3) suprasubduction character of the ophiolite; (4) vergence of structures in rocks beneath the ophiolite.

Age ofeclogite metamorphism The age of eclogite metamorphism is crucial to any tectonic reconstruction. Most agree that the ages older than 120 Ma are due to excess argon (e.g. Searle et al 1994; Miller et al 1999), but 100-105 Ma groupings may have significance (Miller et al 1999). More recently, El-Shazly et al (2001) have argued that Ar-Ar ages older than 80 Ma from the As Sifah eclogites are due to excess argon in crystal lattices. They used Rb-Sr

geochronology of clinopyroxene-phengite and epidote-phengite pairs, essentially two-point isochrons consisting of a Rb-rich phengite phase and Rb-poor second phase (epidote or pyroxene), to support this contention and argued that the eclogite cooled through 500 °C at 78 ± 2 Ma. This interpretation is somewhat controversial, as Miller et al (1999) showed that Ar-Ar apparent ages of 80-78 Ma reflect top-to-the-NE shear zone development in the lower plate related to uplift and exhumation. Significant dynamic recrystallization and mica growth in these shear zones requires resetting of geochronological systems to give the 80-78 Ma ages. It is interesting to note that the highest-grade rocks with the oldest fabrics tend to yield the oldest Ar-Ar ages (Miller et al 1999). Depending on which interpretation of metamorphic ages one believes (e.g. Miller et al 1999 vs. El-Shazly et al 2001), the high-P metamorphism in rocks of the lower plate of the Saih Hatat window is either younger than or older than the ophiolite. If the high-P rocks are older than the ophiolite (i.e. pre-ophiolitic), then there must have been some active subduction before Oman ophiolite generation, thrust initiation and metamorphic sole formation beneath the ophiolite. If they are younger, then the interpretation of El-Shazly et al (2001) is permissable. In their scenario the high-P metamorphism is due to crustal stacking at the scale of the oceanic lithosphere (see El-Shazly etal. 2001, fig. 8).

P-T conditions of the metamorphic sole Recent geobarometric determinations from metamorphic assemblages in the metamorphic sole (Gnos 1998; Searle & Cox 2002) have produced pressures that are in excess of the pressure that can be explained by the thickness of the overlying ophiolite sheet (e.g. c. 6 kbar, Gregory 1984), calling into the question the relationship between the contact aureole rocks and the overlying peridotite. Inferred pressures in excess of the structural load imposed by the pseudostratigraphic thickness of the ophiolite may not be consistent with short-lived imbrication near the ridge crest (see Wakabayashi & Dilek 2000; Searle & Cox 2002), instead requiring a subduction setting (e.g. Searle & Cox 1999, 2002), where the high-f fabrics (Boudier et al 1985; Nicolas 1989) in the metamorphic sole indicate a subduction zone dipping away from the margin instead of an intraoceanic thrust system. Breaking the link between the overlying peridotite and the metamorphism of the sole adds another exhumation event to the geological history of the belt. Boudier & Coleman (1981) summarized the fabrics in the peridotites overlying the meta-

OBDUCTION OF THE SAMAIL OPHIOLITE morphic sole and concluded that the peridotite fabrics spatially associated with the Wadi Tayin metamorphic sole were consistent with much higher deviatoric stress and lower temperatures than the fabrics associated with hypersolidus flow at the spreading centre. Boudier & Nicolas and their coworkers linked the later high-stress-lowertemperature fabrics to the metamorphism and fabric development of the sole rocks at pressures at the inferred load of the ophiolite or lower. Clearly, breaking the link between the aureole rocks and the overlying peridotite relies on pressure estimates based upon thermodynamic models whose internal precisions may far exceed their geological accuracy when the complications of amphibole chemistry and polymetamorphism are included.

Suprasubduction character of the ophiolite A depleted normal mid-ocean ridge basalt (NMORB), suprasubduction geochemical signature for upper unit lavas in the Samail Ophiolite (Pearce et al. 1981) has been used to argue for a subduction zone dipping away from the Arabian margin (e.g. Searle & Cox 1999, fig. 2). These upper lavas (the Lasail unit) comprise clinopyroxene-phyric basalts overlain by mildly alkaline transitional within-plate basalts that have been interpreted as immature island-arc tholeiites (Pearce et al. 1981). The lack of a constructional arc pile preserved anywhere in the Samail Ophiolite has made this interpretation somewhat controversial.

Vergence of structures in rocks beneath the ophiolite The dominant structures in both the platform and slope-facies carbonate rocks exposed in the Saih Hatat and Hawasina windows, in particular, are folds and shear zones verging away from the craton. The folds are extensively developed, and have regional scale and extent. Ductile shear zones or high-strain zones within the Hawasina window (Fig. 10), as well as the lower plate and the major upper plate-lower plate discontinuity of the Saih Hatat window (Fig. 9), all show top-tothe-NE shear sense (i.e. vergence away from the margin towards the former Tethys ocean). Their presence indicates margin involvement, with 'thrusting' and overfolding of the margin to the NE as the ophiolite is emplaced over the top to the SW or west. Kinematics for development of the major NE-facing Saih Hatat antiformal foldnappe requires underthrusting beneath the pinned upper limb (para-autochthonous carbonate of the

461

margin); that is, underthrusting towards the margin. All of the margin rocks exposed in the windows of the Oman Mountains, not just the Sumeini Group (e.g. Glennie et al. 1974; Searle 1985) are therefore at least para-autochthonous, and probably allochthonous.

Tectonic scenarios for obduction There are a number of permissable geometrical scenarios for Oman obduction incorporating multi-tiered detachment systems showing movement of the margin towards the former Tethys ocean at mid-crustal levels (Fig. 11). All of the models involve some form of extrusion where high-P rocks move upwards relative to rocks both structurally above and below them. The first two model scenarios involve subduction of the Arabian continental margin under the overriding oceanic plate (Fig. lla and b), but involve different driving forces. The third requires underthrusting of oceanic lithosphere beneath the margin (Fig. lie): (1) 'buoyancy' driven wedge uplift (Chemenda model; e.g. Chemenda et al. 1996, fig. 8) where an escaping subducted crustal slice intrudes the subduction interface and pushes up and deforms the previously subducted sediments ahead of it (Chemenda et al 1996, fig. 8a) (Fig. lla); (2) 'retrocharriage'-driven uplift (Alpine scenario; e.g. Roeder 1973, 1977) where back-thrusting and back-folding are an integral part of a doubly vergent thrust system above the former oceanwards-dipping subduction interface (Fig. 1 Ib); (3) 'buoyancy' driven wedge uplift associated with a subduction system directed beneath the margin (Gregory, Gray & Miller 'paddle-board' model: Gregory et al. 1998; Gray et al. 2000) (Fig. lie).

Implications 'Back-folding', or more simply folds verging away from the margin contra to the direction of thrusting of the Samail Ophiolite nappe, occurs along half of the mountain range, and this restriction may only be due to limitations of exposure beneath the ophiolite. It would appear therefore that this behaviour of the Arabian margin is an important part of the obduction process. Most obduction models for Oman, and other parts of the world generally, have obduction related to a subduction system (or underthrusting) dipping away from the margin. The question remains as to whether subduction dipping away from the margin is important in the generation of oceanic crust (i.e. a suprasubduction environment), the generation of the metamorphic rocks that form the sole at the base of the ophiolite nappe, and in constraining

462

D. R. GRAY & R. T. GREGORY

the final emplacement of the ophiolite (e.g. Searle & Cox 1999, 2002; El-Shazly et al 2001). Such a model is reliant on the interpretation of the geochemistry of the Lasail lavas as well as the reliability and accuracy of P-T determinations on the metarnorphic sole rocks. Structural relationships in Sain Hatat and the Hawasina window necessitate some form of underthrusting toward the margin for at least part of the history. The critical relationship to determine between the tectonic scenarios will be the direction in which the shear zones root into the lower part of the continental lithosphere. Gray et al. (2000) have speculated that important shear zones in Saih Hatat may root into the lithosphere on the continent side of the margin. If this is the case then these structures can not be simple back-thrusts, and would indicate that craton-directed underthrusting has to be extremely important in the obduction process of the Samail Ophiolite. Furthermore, Samail Ophiolite obduction would be linked to, and constrained by, an early formed subduction system dipping beneath the Arabian continental margin. Whatever the case, vergence of the shear zones, regional-scale folds and foldnappes away from the margin towards the ocean in the margin rocks being 'overthrust' by the ophiolite nappe cannot be ignored (e.g. El-Shazly et al 2001, fig. 8; Searle & Cox 2002, fig. 7). Funding for the research was from Australian Research Council Grant A39601548 (awarded to D.R.G.) and National Science Foundation Grant EAR91-06016 (awarded to R.T.G.). Preparation of the manuscript was undertaken while D.R.G. was under the tenure of an Australian Professorial Research Fellowship. We wish to thank the Ministry of Petroleum and Minerals for their kind support during field work, particularly H. Al Azry (Director General of Minerals). Discussions with R. Coleman, I-P. Breton (BRGM, Muscat), and J. Filbrandt (Petroleum Development Oman) on this problem are gratefully acknowledged. Careful reviews by J. Wakabayashi, T. Peters and S. Wojtal led to development of some of the ideas in this paper.

References Fig. 11. Speculative schematic illustrations of permissable tectonic scenarios allowing a mid-crastal geometry where the margin rocks are back-folded and/or thrust towards the former Tethys ocean at the time of ophiolite emplacement, (a) Buoyancy-driven crustal slice inducing back-folding or thrusting (after Chemenda et al 1996, fig. 8). (b) 'Retrocharriage' model after the Alps (e.g. Roeder, 1977, fig. 3). (c) Underthrasting towards the margin incorporating tiered-detachments and buoyant ascent of lower plate blocks (after Gregory et al. 1998, fig. 8; Gray et al 2000, fig. 7) UP, upper plate; LP, lower plate. Black, oceanic lithosphere; grey, continental crust.

BECHENNEC, F., LE METOUR, J., RABU, D., BOURDILLON-DE-GRISSAC, CH., DE WEVER, P., BEURRIER, M. & VILLEY, M. 1990. The Hawasina Nappes: stratigraphy, palaeogeography and structural evolution of a fragment of the south-Tethyan passive continental margin. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 213-223. BOUDIER, F. & MICHARD, A. 1981. Oman ophiolite: the quiet obduction of oceanic crust. Terra Cognita, 1, 109-118. BOUDIER, F., NICOLAS, A. & BOUCHEZ, J.L. 1982.

OBDUCTION OF THE SAMAIL OPHIOLITE Kinematics of oceanic thrusting and subduction from basal sections of ophiolites. Nature, 296, 825-828. BOUDIER, F., BOUCHEZ, J.L., NICOLAS, A., CANNAT, ML, CEULENEER, G., MISSEEI, M. & MONTIGNY, R. 1985. Kinematics of oceanic thrusting in the Oman Ophiolite: model of plate convergence. Earth and Planetary Science Letters, 75, 215-222. BOUDIER, R, CEULENEER, G. & NICOLAS, A. 1988. Shear zones, thrusts and related magmatism in the Oman Ophiolite: initiation of thrusting on an ocean ridge. Tectonophysics, 151, 275-296. BUCHER, M. 1991. The metamorphic series associated with the Semail Ophiolite nappe of the Oman Mountains in the United Arab Emirates. MSc thesis, University of Berne. CAWOOD, P.A. 1991. Processes of Ophiolite emplacement in Oman and western Newfoundland. In: PETERS, TJ., NICOLAS, A. & COLEMAN, R. (eds) Ophiolite Genesis and Evolution of the Ocean Lithosphere. Kluwer Academic, Dordrecht, 501-516. CAWOOD, P.A., GREEN, RK. & CALON, TJ. 1990. Origin of culminations within the Southeast Oman Mountains at Jebel Ma Jhool and Ibra Dome. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 429-446. CHEMENDA, A.I., MATTAUER, M. & BOKUN, A.N. 1996. Continental subduction and a mechanism for exhumation of high-pressure metamorphic rocks: new modelling and field data from Oman. Earth and Planetary Science Letters, 143, 173-182. COFFIELD, D.Q. 1990. Structures associated with nappe emplacement and culmination collapse in the central Oman Mountains. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 447-458. COLEMAN, R.G. 1971. Plate tectonic emplacement of upper mantle peridotites along continental edges. Journal of Geophysical Research, 86, 1212-1222. COLEMAN, R.G. 1977. Ophiolites: Ancient Oceanic Lithosphere?. Springer, New York. COLEMAN, R.G. 1981. Tectonic setting for Ophiolite obduction in Oman. Journal of Geophysical Research, 76, 2497-2508. DEWEY, J.F. 1976. Ophiolite obduction. Tectonophysics, 31, 93-120. DEWEY, J.F. & BIRD, J.M. 1971. Origin and emplacement of of the Ophiolite suite: Appalachian ophiolites in Newfoundland. Tectonophysics, 76, 31793206. DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (EDS) 2000. Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 552 pp. ELLIOTT, D. 1976. The motion of thrust sheets. Journal of Geophysical Research, 81, 949-963. EL.-SHAZLY, A.K. 1994. Petrology of lawsonite-, pumpellyite- and sodic amphibole-bearing metabasites from NE Oman. Journal of Metamorphic Geology,

463

12, 23-48. EL-SHAZLY, A.K. 1995. Petrology of Fe-Mg-carpholitebearing metasediments from NE Oman. Journal of Metamorphic Geology, 13, 379-396. EL-SHAZLY, A.K. & LANPHERE, M.A. 1992. Two high pressure metamorphic events in NE Oman: evidence from 40Ar/39Ar dating and petrological data. Journal of Geology, 100, 731-751. EL-SHAZLY, A.K. & COLEMAN, R. 1990. Metamorphism in the Oman Mountains in relation to the Semail ophiolite emplacement. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 473-493. EL-SHAZLY, A.K., COLEMAN, R.G. & Liou, J.G. 1990. Eclogites and blueschists from northeastern Oman; petrology, P-T evolution. Journal of Petrology, 31, 629-666. EL-SHAZLY, A.K., BROCKER, M., HACKER, B. & CALVERT, A. 2001. Formation and exhumation of blueschists and eclogites from NE Oman: new perspectives from Rb-Sr and 40Ar/39Ar dating. Journal of Metamorphic Geology, 19, 233—248. GEALEY, W.K. 1977. Ophiolite obduction and geologic evolution of the Oman Mountains and adjacent areas. Geological Society of America Bulletin, 88, 1183-1191. GEALEY, W.K. 1979. Ophiolite obduction mechanism. In: PANAYIOTOU, A. (ed.) Proceedings of the International Ophiolite Symposium, Cyprus, 1979. Cyprus Geological Survey, Nicosia, 228-243. GHENT, E.D. & STOUT, M.Z. 1981. Metamorphism at the base of the Semail Ophiolite, southeastern Oman Mountains. Journal of Geophysical Research, 86, 2557-2572. GLENNIE, K.W., HUGHES CLARKE, M.W., BOEUF, M.G., PILAAR, W.F. & REINHARDT, B.M. 1974. Geology of the Oman Mountains. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootsschap, 31. GNOS, E. 1998. Peak metamorphic conditions of garnet amphibolites beneath the Semail Ophiolite: implications for an inverted pressure gradient. International Geology Review, 40, 281-304. GOFFE, B., MlCHARD, A., KlENAST, J.R. & Lfi MER, O.

1988. A case of obduction-related high pressure, low temperature metamorphism in upper crustal nappes, Arabian continental margin, Oman. Tectonophysics, 151, 363—386. GRAY, D.R. & GREGORY, R.T. 2000. Implications of the structure of the Wadi Tayin metamorphic sole, the Dasir-Ibra block of the Samail Ophiolite, and the Saih Hatat window for late stage extensional ophiolite emplacement, Oman. Marine Geophysical Researches, 21, 211-227. GRAY, D.R., GREGORY, R.T. & MILLER, J.McL. 2000. A new structural profile along the Muscat-Ibra transect, Oman: implications for the emplacement of the Samail Ophiolite. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 513-523. GREGORY, R.T. 1984. Melt percolation beneath a spread-

464

D. R. GRAY & R. T. GREGORY

ing ridge: evidence from Semail peridotite, Oman. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 14, 55-62. GREGORY, R.T. & TAYLOR, H.P. JR 1981. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail ophiolite, Oman: evidence for 618O buffering of the oceans by deep (>5 km) seawaterhydrothermal circulation at mid-ocean ridges. Journal of Geophysical Research, 86, 2737-2755. GREGORY, R.T., GRAY, D.R. & MILLER, J.McL. 1998. Tectonics of the Arabian margin associated with the emplacement of the Oman margin along the Ibra transect: new evidence from NE Saih Hatat. Tectonics, 17, 657-760. HACKER, B.R. 1990. Simulation of the metamorphic and deformational history of the metamorphic sole of the Oman ophiolite. Journal of Geophysical Research, 95, 4895-4907. HACKER, B.R. 1991. The role of deformation in the formation of metamorphic gradients: ridge subduction beneath the Oman ophiolite. Tectonics, 10, 455-473. HACKER, B.R. 1994. Rapid emplacement of young oceanic lithosphere: argon geochronology of the Oman Ophiolite. Science, 265, 1563-1565. HACKER, B.R. & GNOS, E. 1997. The conundrum of Samail: explaining the metamorphic history. Tectonophysics, 279, 215-226. HACKER, B.R. & MOSENFELDER, J.L. 1997. Metamorphism and deformation of the sole beneath the Samail Ophiolite. Tectonics, 10, 455-473. HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman Ophiolite: thermal and geochronological constraints. Tectonics, 15, 1230-1247. HANNA, S. 1986. The Alpine deformation (Late Cretaceous and Tertiary) tectonic evolution of the Oman Mountains: a thrust tectonic approach. In: EDITOR, A. (ed.) Symposium on the Hydrocarbon Potential of Intense Thrust Zones: Organisation of Arab Petroleum Exporting Countries Conference. OAPEC, Kuwait, 2, 125-174. HANNA, S. 1990. The Alpine deformation of the Central Oman Mountains. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 341-359. HOPSON, C.A., COLEMAN, R.G., GREGORY, R.T., PALLISTER, J.S. & BAILEY, E.H. 1981. Geological section through the Samail Ophiolite and associated rocks along a Muscat-Ibra transect, SE Oman mountains. Journal of Geophysical Research, 86, 2527-2544. LE METOUR, J., DE GRAMONT, X. & VILLEY, M. 1986. Geological map of Masqat and Quryat, sheets NF40-4A, NF40-4D, scale 1:100000. Explanatory notes. Ministry of Petroleum and Minerals, Directorate General of Minerals, Sultanate of Oman. LE METOUR, J., RABU, D., TEGYEY, M., BECHENNEC, E, BEURRIER, M. & VILLEY, M. 1990. Subduction and obduction: two stages in the Eo-Alpine tectonometamorphic evolution of the Oman Mountains. /W.ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C.

(eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 327-339. LE METOUR, J., MICHEL, J.C., BECHENNEC, F., PLATEL, J.P. & ROGER, J. 1995. Geology and Mineral Wealth of Oman, Ministry of Petroleum and Minerals, Muscat Oman, 285 pp. LIPPARD, SJ. 1983. Cretaceous high pressure metarnorphism in NE Oman and its relationship to subduction and ophiolite nappe emplacement. Journal of the Geological Society, London, 140, 97-104. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Geological Society of London, Memoir, 11. MANN, A. & HANNA, S.S. 1990. The tectonic evolution of pre-Permian rocks, Central and southeastern Oman Mountains. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 307-325. MATTAUER, M. & RITZ, J.-F. 1996. Arguments geologiques en faveur d'un modele de subduction continentale pour 1'exhumation du metamorphisme haute-pression d'Oman. Comptes Rendus de I'Academic des Sciences, Serie Ila, 322, 869-876. MlCHARD, A., BOUCHEZ, J.L. & OUAZZIANI-TOUHAMI,

O. 1984. Obduction-related planar and linear fabrics in Oman. Journal of Structural Geology, 6, 39—49. MICHARD, A., BOUDIER, F. & GOFFE, B. 1991. Obduction versus subduction and collision in the Oman case and other Tethyan settings. In: PETERS, TJ., NICOLAS, A. & COLEMAN, R. (eds) Ophiolite Genesis and Evolution of the Ocean Lithosphere. Kluwer Academic, Dordrecht, 447-467. MICHARD, A., GOFFE, B., SADDIQI, O., OBERHANSLI, R. & WENDT, A.S. 1994. Late Cretaceous exhumation of the Oman blueschists and eclogites: a two stage extensional mechanism. Terra Nova, 6, 404-413. MILLER, J.McL. 1998. The structure, metamorphism and 40 Ar/39Ar geochronology of northeast Saih Hatat, Oman: implications for the formation and preservation of high pressure rocks. Unpublished PhD thesis, Monash University, Clayton, Melbourne, Australia. MILLER, J.McL., GRAY, D.R. & GREGORY, R.T. 1998. Exhumation of high-pressure rocks in northeastern Oman. Geology, 26, 235-238. MILLER, J.McL., GRAY, D.R., GREGORY, R.T. & FOSTER, D.A. 1999. Geological and geochronological constraints on the exhumation of a high-pressure metamorphic terrane, Oman. In: RING, U., BRANDON, M.T., LISTER, G.S. & WILLET, S.D. (eds) Exhumation Processes: Normal Faulting, Ductile Flow and Erosion. Geological Society, London, Special Publications, 154, 241-260. MILLER, J.McL., GRAY, D.R. & GREGORY, R.T. 2002. Geometry and significance of internal windows and regional isoclinal folds in northeast Saih Hatat, Sultanate of Oman. Journal of Structural Geology^ 24, 359-386. MITCHELL, A.H.G. 1986. Ophiolite and associated rocks in four settings: relationships to subduction and collision. Tectonophysics, 125, 269-285.

OBDUCTION OF THE SAMAIL OPHIOLITE MONTIGNY, R., LE MER, O. & WHITECHURCH, H. 1988. K-Ar and Ar40/Ar39 study of metamorphic rocks associated with the Oman ophiolite: tectonic implications. Tectonophysics, 151, 345-362. MOORES, E.M. 1982. Origin and emplacement of ophiolites. Reviews of Geophysics, 20, 735-760. MOORES, E.M. & Twiss, R.J. 1995. Tectonics. W.H. Freeman, New York. MOUNT, V.S., CRAWFORD, R.I.S. & BERGMAN, S.C. 1998. Regional structural style of the central and southern Oman Mountains: Jabal Akhdar, Saih Hatat, and the Northern Ghaba Basin. GeoAmbia, 35 475-490. MUELLER, S. & PHILLIPS, R.J. 1991. On the initiation of subduction. Journal of Geophysical Research, 96, 651-665. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer Academic, Dordrecht. NICOLAS, A., CEULENEER, G., BOUDIER, F. & MISSERI, M. 1988. Structural mapping in the Oman Ophiolites: mantle diapirism along an oceanic ridge. Tectonophysics, 51, 27-56. PEARCE, J.A., ALABASTER, T., SHELTON, A.W. & SEARLE, M.P. 1981. The Oman ophiolite as a Cretaceous arc-basin complex: evidence and implications. Philosophical Transactions of the Royal Society of London, Series A, 300, 299-317. ROBERTSON, A.H.F. & SEARLE, M.P. 1990. The Oman Tethyan continental margin: stratigraphy, structure, concepts and controversies. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 3-25. ROEDDER, D. 1973. Subduction and orogeny. Journal of Geophysical Research, 78, 5005-5024. ROEDDER, D. 1977. Continental convergence in the Alps. Tectonophysics, 40, 339-350. SEARLE, M.P. 1985. Sequence of thrusting and origin of culminations in the northern and central Oman Mountains. Journal of Structural Geology, 1, 129-143. SEARLE, M.P. & Cox, J. 1999. Tectonic setting, origin and obduction of the Oman ophiolite. Geological Society of America Bulletin, III, 104-122. SEARLE, M.P. & Cox, J. 2002. Subduction zone metamorphism during formation and emplacement of the Semail ophiolite in the Oman Mountains. Geological Magazine, 139, 241-255. SEARLE, M.P. & MALPAS, J. 1980. Structure and

465

metamorphism of rocks beneath the Semail ophiolite of Oman and their tectonic significance in ophiolite obduction. Transactions of the Royal Society of Edinburgh, 71, 247-262. SEARLE, M.P. & STEVENS, R.K. 1984. Obduction processes in ancient, modern and future ophiolites. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society Special Publications, London, 13, 303-319. SEARLE, M.P., WATERS, D.J., MARTIN, H.N. & REX, D.C. 1994. Structure and metamorphism of blueschist-eclogite facies rocks from the NE Oman Mountains. Journal of the Geological Society, London, 151, 555-576. SHELTON, A.W. & EGAN, S.S. 1991. The obduction of the northern Oman ophiolite: crustal loading and flexure. In: PETERS, TJ., NICOLAS, A. & COLEMAN, R. (eds) Ophiolite Genesis and Evolution of the Ocean Lithosphere. Kluwer Academic, Dordrecht, 469-483. TILTON, G.R., HOPSON, C.A. & WRIGHT, I.E. 1981. Uranium-lead isotopic ages of the Semail Ophiolite, Oman, with applications to Tethyan ridge tectonics. Journal of Geophysical Research, 86, 2763-2775. VILLEY, M., BECHENNEC, F., BEURRIER, M., LE METOUR, J. & RABU, D. 1986. Geological map of Yanqul, Sheet NF40-2C, Scale 1:100000. Explanatory notes. Ministry of Petroleum and Minerals, Directorate General of Minerals, Sultanate of Oman. WAKABAYASHI, J. & DILEK, Y. 2000. Spatial and temporal relationships between ophiolites and their metamorphic soles; a test of models of forearc ophiolite genesis. In: MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling. Geological Society of America, Special Papers, 349, 53-64. WARBURTON, J., BURNHILL, T.J., GRAHAM, R.H. & ISAAC, K.P. 1990. The evolution of the Oman Mountains foreland basin. In: ROBERTSON, A.H.F., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 419-427. WENDT, A.S., D'ARCO, P., GOFFE, B. & OBERHANSLI, R. 1993. Radial cracks around a-quartz inclusions in almandine: constraints on the metamorphic history of the Oman mountains. Earth and Planetary Science Letters, 114, 449-461.

This page intentionally left blank

Subduction zone polarity in the Oman Mountains: implications for ophiolite emplacement M. P. SEARLE 1 , C. J. WARREN 1 , D. J. WATERS 1 & R. R. PARRISH 2 1

Department of Earth Sciences, Oxford University, Parks Road, Oxford OX1 3PR, UK (e-mail: [email protected]) 2 Department of Geology, University of Leicester, and Natural Environment Research Council, Isotope Geosciences Laboratory, British Geological Survey, Keyworth, Nottingham NG12 5GG, UK Abstract: Two end-member models have been proposed to account for the structure and metamorphism of rocks beneath the Semail ophiolite in the Oman mountains. Model A involves a single, continuous NE-directed subduction away from the continental margin during the late Cretaceous. The ophiolite and underlying thrust sheets of distal to proximal oceanic sediments were emplaced a minimum of 250 km SW onto the continental margin. Subduction of Triassic—Jurassic oceanic basalts to c. lOkbar (c. 39km depth) led to the accretion of amphibolite-facies rocks to the base of the ophiolite. Thrusting propagated towards the continental margin and ended with subduction of the thinned continental crust to c. 20 kbar (c. 78 km depth), choking the subduction zone. Buoyancy forces caused the rapid exhumation of eclogites, blueschists and carpholite-grade HP rocks along the NE margin of the continental plate. During the later phase of foreland-propagating thin-skinned thrusting in the SW, NEfacing backfolding and backthrusting occurred in the hinterland, with the final exhumation of the HP rocks. Model B follows recent suggestions that a nascent SW-dipping subduction zone, dipping beneath the continental margin, existed between 130 and 95 Ma, prior to formation and emplacement of the ophiolite. A major NE-facing fold-nappe structure in the pre-Permian basement rocks of Sain Hatat is interpreted as reflecting subduction beneath the margin. Two high-pressure metamorphic events have been suggested, the first predating ophiolite emplacement, the second caused by ophiolite loading. This model is untenable, being based on a misinterpretation of the NE-facing structures in northern Saih Hatat, and on some dubious older 40 Ar/39Ar cooling ages from the eclogite-facies rocks of As Sifah. We conclude that all structures in northern Oman and all the reliable geochronology point to a single emplacement obduction event lasting from Cenomanian-Turonian time (c. 95 Ma) when amphibolites were accreted along the metamorphic sole of the ophiolite, to Campanian time, when the continental margin was subducted to the NE producing blueschists and eclogites, to the final thin-skinned emplacement of all thrust sheets, which ended before the Late Maastrichtian, at c. 68 Ma.

Since Glennie et al. (1973, 1974) published the first detailed stratigraphie and structural results of mapping the entire Oman Mountain belt, it has been universally recognized that the mountains comprise a series of relatively thin-skinned thrust sheets emplaced from NE to SW onto the stable passive continental margin of Arabia (Fig. 1). These allochthonous sheets comprise (from lower to higher units): Middle Permian to Cenomanian shelf slope carbonates (Sumeini Group), proximal to distal sediments (Hawasina complex), deepwater trench sediments and Triassic seamounts (Haybi complex; Searle & Malpas 1980; Searle 1985), and the Semail ophiolite. The Sumeini, Hawasina and Haybi thrust sheets comprise timeequivalent units of differing palaeogeographical

facies with higher units being more distal to the shelf margin. Thrusting generally propagated from NE to SW with more distal thrust sheets emplaced onto more proximal units. The Semail ophiolite crustal sequence formed during the Cenomanian, from crystallization ages of plagiogranites in the lower-crustal sequence (U-Pb zircon ages of 93.5-97.9 Ma with a mean of c. 95 Ma; Tilton et al. 1981), and from early Cenomanian to early Turonian radiolaria in cherts from the volcanic sequence (Tippit et al. 1981). Amphibolites in the metamorphic sole rocks, accreted along the base of the peridotites of the Semail ophiolite, formed at peak P-T conditions of 840-870 °C and 11.6 ± 1.6 kbar equivalent to 45-50 km depth beneath oceanic crust hanging

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 467-480. 0305-8719/037$ 15 © The Geological Society of London 2003.

468

M. P. SEARLE ETAL.

Fig. 1. Geological map of the northern Oman Mountains.

wall (Gnos 1998; Searle & Cox 2002). 40Ar/39Ar cooling ages from hornblendes in the amphibolite sole are 95-92 Ma and record the timing at which these rocks cooled through 500 °C (Gnos & Peters 1993; Hacker 1994; Hacker et al 1996). Structural mapping, thermobarometry and geochronology clearly indicate that NE-directed oceanic subduetion of cold basaltic material to depths of 4550 km beneath the Semail ophiolite was occurring at c. 95 Ma (Searle & Malpas 1980, 1982; Lippard et al. 1986; Searle & Cox 1999, 2002). This NEdirected oceanic subduction system developed into thinner-skinned thrust tectonics with time as the Semail ophiolite, and the Haybi and Hawasina thrust sheets were progressively emplaced onto the passive margin of Arabia by a foreland-propagating 'piggy-back' style of thrusting (e.g. Searle 1985; Cooper 1988; Hanna 1990). In the southeastern part of the northern Oman mountains (Fig. 2) a high-pressure terrane is

centred around the village of As Sifah, where eclogite-facies metabasalts and metapelites crop out at the deepest structural levels exposed. The eclogites record P-T conditions of 540 ± 75 °C and c. 20kbar reflecting subduction of thinned continental crustal rocks to depths of about 7880km beneath a hanging wall that was always oceanic or ophiolitic (Searle et al. 1994; Searle & Cox 1999, 2002). Structurally overlying the eclogites are a series of blueschist- and carpholitegrade HP-LT rocks (Goffe et al 1988; El-Shazly & Coleman 1990; El-Shazly et al. 1990; El-Shazly & Liou 1991; Searle et al. 1994) showing an upwards decreasing thermal gradient. Most workers on Oman structural geology agree that subduction polarity throughout the emplacement process was NE-directed, associated with SW emplacement of oceanic units onto the continental margin (Model A, Fig. 3; e.g. Glennie et al. 1973, 1974; Searle 1985, 1988a, 1988b; Lippard et al. 1986;

SUBDUCTION ZONE POLARITY IN OMAN

469

Fig. 2. Map of the Muscat-Quriat and Saih Hatat region, after LeMetour et al. (1986) and Gray & Gregory (2000).

Searle & Cooper 1986; Bernoulli & Weissart 1987; Cooper 1988; Goffe et al. 1988; Searle et al 1990, 1994; Cawood et al 1990; Dunne et al 1990; Hanna 1990; LeMetour et al 1990; Hacker 1994; Hacker & Mosenfelder 1996; Hacker et al 1996; Searle & Cox 1999, 2002; El-Shazly et al 2001). Gregory et al (1998), Gray et al (2000) and Miller et al (2002) remapped northern Saih Hatat and proposed an entirely different tectonic model (Model B; Fig. 4). Their model involves two stages of subduction, an early, pre-ophiolite formation, 130-95 Ma SW-dipping subduction followed by a later NE-dipping subduction during emplacement, and two stages of HP metamorphism, a pre-ophiolite, Early-Mid-Cretaceous HP1,

which formed the eclogites, and an ophiolite emplacement related HP2 during the Late Cretaceous. Both these models are now critically assessed.

Model A: one NE-dipping subduction system Although some controversy still persists as to the original tectonic setting of the Semail ophiolite, either a mid-ocean ridge or suprasubduction zone setting (see Searle & Cox (1999, 2002) for a discussion), most workers agree that the emplacement of the ophiolite was related to NE-dipping, SW-verging thrusts emplacing thrust sheets of

470

M. P. SEARLE ETAL.

Fig. 3. Model A, after Searle et al. (1994). Tectonic model for the subduction-obduction evolution of the Oman ophiolite. (a) Palinspastically restored cross-section across the Arabian continental margin showing the positions of the major thrusts, the Hawasina Thrust (HT), Haybi Thrust (HyT) and Semail Thrust (ST). It should be noted that the shelf, slope (Sumeini) and basin (Hawasina) sediments are laterally time-equivalent units. The continental crust thins towards the margin in the NE. The bold-outlined boxes show the restored palaeostratigraphic positions of the protolith of the high-pressure metamorphic units along the continental margin in NE Oman. The crust beneath the distal Hawasina and Haybi units is dominantly alkali basaltic to transitional tholeiitic basaltic rocks of latest Permian—Triassic and Jurassic age. The sub-ophiolite amphibolites were formed at pressures around 7—10 kbar (28— 39 km depth) by accretion of Jurassic mid-ocean ridge basalt and Triassic Haybi-type basaltic crust to the base of the mantle sequence. Greenschists were subsequently accreted by underplating of Haybi complex sediments (pelites, quartzites and marbles) beneath the amphibolites and metamorphosed at pressures of 2-4 kbar (7-14 km depth). The Semail ophiolite plus metamorphic sole was then thrust onto the Haybi complex and thrusting propagated southwestwards with time, (b) During the later stages of the subduction-obduction process, the continental margin rocks were subducted to depths of around 80-90 km, where peak metamorphic pressures of c. 20 kbar are recorded by thermobarometry in the As Sifah unit eclogites. The sub-ophiolite thrust sheets (Haybi and Hawasina complexes) were emplaced onto the foreland, but are mainly missing in the internal (NE) parts of the Oman Mountains, where the ophiolite lies almost directly on the shelf carbonates. The subduction of continental margin material down to c. 80 km must have occurred in a phase of rapid plate motion during ophiolite obduction. Exhumation of the highpressure As Sifah unit rocks must also have occurred rapidly, back up to the same subduction zone, soon after peak metamorphism.

SUBDUCTION ZONE POLARITY IN OMAN

471

Fig. 4. Model B, after Gregory et al (1998) and Gray et al. (2000). In stages 1 and 2, a pre-Cenomanian platformdirected nascent subduction zone formed eclogites before ophiolite formation. Stage 3 is the intra-oceanic thrusting event when the ophiolite was formed and began emplacing towards the margin (95-80 Ma). During stage 4 the frontal part of the ophiolite overrode the nascent SW-dipping subduction zone. In stage 5 the present-day Semail 'thrust' formed, the folded ophiolite sheet was tectonically thinned and the HP rocks moved upward from south to north.

oceanic rocks onto the drowned continental shelf margin. Detailed structural mapping of individual culminations throughout the Oman mountains has shown that thrust sheets beneath the ophiolite have involved transport towards the Arabian continental

margin from east to west in the far north, NE to SW in the central mountains and NNE to SSW in the southeastern mountains. These include, for example, the Dibba zone in the far north (Searle 1988a, 1988b; Gnos 1998), the Sumeini Window,

472

M. P. SEARLE ETAL.

northern Oman (Searle & Malpas 1980, 1982; Searle & Cox 2002), the foreland fold-thrust belt in the Hamrat Duru range (Cooper 1988), the Asjudi and Hawasina Windows (Searle 1985; Searle & Cooper 1986), the Jebel Akhdar Window (Searle 1985; Bernoulli & Weissart 1987; Hanna 1990), the southeastern mountains (Mann & Hanna 1990) and culminations south of Saih Hatat (Cawood et al 1990). Seismic and well data beneath the foreland, both in north Oman and UAE (Dunne et al. 1990), and central Oman (Warburton et al. 1990) have confirmed the surface structural geology, i.e. that all thrust directions were towards the foreland. Detailed mapping of sub-ophiolite metamorphic sole rocks and imbricate structures beneath the Semail ophiolite confirms that thrusts and folds are related throughout to SW-directed emplacement. Imbricate faults and folds in the Hamrat Duru foreland thrust-fold belt also show that emplacement was SW-directed (Cooper 1988). Searle (1985) also described out-of-sequence thrusts, which truncate structures in the footwall on all scales from outcrop to the regional motion along the Semail 'thrust'. Many different units are in direct contact with the base of the ophiolite, and this geometry has been interpreted as late outof-sequence thrusting and/or late-stage extensional normal faulting. Although in most areas along the Oman mountains, SW-vergent folds and thrusts are evident, there are two areas where regional NE-verging thrusts and NE-facing folds have been documented, in the northern Hawasina Window (Searle & Cooper 1986) and in northern Saih Hatat (BRGM mapping; LeMetour et al. 1990; Gregory et al. 1998; Miller et al. 2002).

NE-vergent structures NE-vergent backthrusts and NE-facing folds were mapped by Searle & Cooper (1986) across the northern part of the Hawasina Window (Fig. 5). These folds are restricted to the central part of the northern Hawasina Window only. To the west in the Haybi region, and to the SE, at Jebel Akhdar, all structures show SW-directed thrusting and SWfacing folding (Searle 1985). The NE-facing folds affect the entire allochthonous tectonostratigraphy from Sumeini Group shelf edge carbonates up to the Semail ophiolite and the backthrusts cut earlier SW-directed thrusts in the footwall. NEdirected backfolding and backthrusting must therefore have formed at a late stage, after emplacement of all thrust sheets onto the margin. The internal geometry of the structures in the Hamrat Duru Group sedimentary rocks in the Hawasina Window shows an imbricate fan with SW-verging thrusts and SW-facing folds along the

southern margin of the window, and NE-vergent back thrusts and backfolds along the northern margin. These structures form a classic 'pop-up* structure related to divergent thrusts above a footwall ramp in the shelf margin. Searle and Cooper (1986) related the Hawasina Window backfolds to a local NE-facing promontory in the shelf edge along the basal decollement. These workers also described late out-of-sequence motion along the Semail 'thrust' around the Hawasina Window culmination, as the peridotites of the Semail mantle sequence rest on top of numerous different thrust sheets in the footwall. Minor extensional normal faults were also mapped, which had the effect of downthrowing the structurally higher ophiolite around the margins of the Window. In northern Saih Hatat, NE-facing folds attain very large proportions (LeMetour et al. 1986, 1990; Miller et al. 2002). These structures were noted by Glennie et al. (1974) during the initial mapping, and also by Bailey (1981) and during the BRGM mapping of the region (LeMetour et al. 1990), although their significance was not really publicized until Gregory et al. (1998), Gray et al. (2000) and Miller et al. (2002) described the structures in great detail. Our interpretation of the structures in Saih Hatat is shown in Figure 6, and the interpretation of Gregory et al. (1998), Gray et al (2000) and Gray & Gregory (2000) is shown in Figure 7. These structures form the main evidence for the model of SW-directed subduction (Model B) and the arguments against this will be laid out in the following section.

Timing of folding and thrusting The early SW-directed thrusting initiated with the accretion of HT amphibolites to the base of the ophiolite at 35-40 km depth beneath the ophiolite during the Cenomanian-Turonian (Searle & Malpas 1980, 1982; Gnos 1998). Thrusting then propagated SW with time, with successively outboard, more distal units thrust over more inboard, proximal units. As the shelf margin collapsed during the late Coniacian-early Santonian at 8884 Ma (Boote et al. 1990; Warburton et al. 1990) the Hawasina, Haybi and overlying Semail ophiolite thrust sheets were emplaced into the flexed Aruma basin along the continental margin. Final stages of the emplacement involved localized NEdirected backfolding in the northern Hawasina Window and along northern Saih Hatat. The entire process was over by the late Maastrichtian, when shallow marine rudist-bearing limestones (Simsima Formation) were deposited over all allochthonous units. Because no Tertiary cover rocks are exposed directly over the Hawasina Window, it is difficult

Fig. 5. Geological cross-sections across the Hawasina Window, after Searle and Cooper (1986), showing the late-stage localized backthrusting and backfolding to the NE. Sedimentary rocks of the Hamrat Duru Group include the Zulla Fm (Z), Guwayza sandstone (Gs) and limestone (Gl), Sidr chert (S) and Nayid limestone (N). The Sumeini group (Sm) shelf slope carbonates are the deepest structural unit and the most proximal palaeogeographically. Early SW-directed thrusts have been refolded around the fan structure and cut by later, steep NE-verging backthrusts.

474

M. P. SEARLE ETAL.

Fig. 6. Structural section across the Saih Hatat culmination and Ibra ophiolite block, showing our interpretation of the structural geometry. The HP units of As Sifah eclogites are the structurally deepest level exposed. The NE-facing backfolds of the Wadi Mayh region are upper-crustal retro-shears and backfolds, pinned to the 'autochthonous' shelf carbonates along the south side of Saih Hatat. Late-stage listric normal faults surround the Saih Hatat culmination, dropping the Ibra block ophiolite down to the south along the southern margin, and dropping the Muscat-Muttrah peridotite down to the north along the northern margin.

Fig. 7. Structural section across Saih Hatat, following the same section as shown in Figure 6, according to Gray & Gregory (2000). 1 and 2, Hawasina and Semail thrusts in the foreland belt; 3 and 4, interpreted crustal-scale shear zones; 5, Ibra dome; 6, Saih Hatat dome and regional-scale fold nappes; 7, normal faults off the coast of north Oman. Inset shows a structural profile across the Ibra block of the Semail ophiolite.

to precisely pin down the age of NE-direeted backfolding. However, it must be the youngest phase of late Cretaceous thrusting, because all earlier SW-directed thrusts and SW-facing folds have been refolded around the NE-verging backfolds. The style of folding in the PaleoceneEocene rocks of Jebel Abiad, south of the Hawasina Window, is very different, showing more open box-type folding, with upright or steep axial planes. Although shortening in the Tertiary jebels of Oman is minimal, only a few kilometres at most, uplift must have been substantial as Eocene-Oligocene limestones formed close to sea level have been uplifted to 2200 m on Jebel Abiad, SE of Quriat, south of Tiwi-Ash Shab in the

southeastern mountains (Fig. 2). Minor collapse type folding occurs along the flanks of these Tertiary mountain ranges, but Tertiary horizontal shortening is minimal.

Summary of main points (1) The NE-facing folds affect all units from deep-level eclogites up to the Permian Saiq Formation. They do not affect the highest thrust slices along the north coast from Muscat to Bandar Khuyran, that show south-directed thrusting and folding in the Triassic shelf units, Ruwi melange and ophiolite.

SUBDUCTION ZONE POLARITY IN OMAN (2) The major detachment termed the upper plate-lower plate discontinuity by Gregory et al. (1998) and Miller et al. (1998, 2002), carrying the large-scale NE-facing folds of northern Saih Hatat along the hanging wall, can be traced from surface geology down to only the upper Proterozoic basement, and not into the lower crust. (3) Metamorphic grade is similar across the upper plate-lower plate discontinuity, with carpholite-bearing metasedimentary rocks and crossite-bearing metabasalts indicating pressures of about 5-7 kbar. The major metamorphic break occurs between the Hulw unit and the As Sifah unit with a large pressure jump of around 8— 10 kbar. This contact is a major shear zone, but it is not a suture zone. (4) The timing of the NE-facing folds in Saih Hatat must have occurred prior to the deposition of the late Paleocene-early Eocene neo-autochthonous limestones, which unconformably cover all structures around the margin of northern Saih Hatat. (5) 40Ar/39Ar cooling ages of 82-79 Ma in the lower plate are associated with transposition fabrics and C' type shear bands during NE-directed shearing, and mica ages from the upper plate are 76-70 Ma (Miller et al. 1999). These cooling ages are consistent with late NE-directed backfolding associated with the final stages of Late Cretaceous culmination of Saih Hatat.

Model B: nascent SW-dipping subduction zone Gregory et al. (1998), Gray & Gregory (2000) and Gray et al. (2000) proposed a very different model involving initial subduction and high-pressure metamorphism of a small continental slab or microplate towards the Arabian continental margin, prior to ophiolite formation and emplacement (Fig. 4; stages 1 and 2). This 'nascent subduction zone' dipping SW beneath the continental margin resulted in the first stage of HP metamorphism at 130-95 Ma producing the eclogites at As Sifah. NE-directed shearing and exhumation of the eclogites occurred between 95 and 80 Ma at the same time as the ophiolite was thrusting towards the SW (Fig. 4; stage 3). A second stage of HP metamorphism of upper plate limestones and upper thrust sheets between 80 and 76 Ma resulted from structural thickening of the ophiolite (Fig. 4; stage 4), followed by gravitational collapse of the ophiolite onto the Arabian margin (Fig. 4; stage 5). Arguments against this model can be grouped into four categories, as described below.

475

Sedimentary evolution of the passive margin The Gray-Gregory model (Fig. 4) requires that the previously well-documented passive continental margin of Arabia (e.g. Glennie et al. 1973, 1974; Robertson & Searle 1990; Scott 1990) actually would have to be a destructive plate margin overlying a subduction zone, dipping SW beneath the continent. Normally, some geological evidence for this destructive margin in terms of calc-alkaline granites or volcanic rocks along an Andean-type margin would be expected. None exists throughout Oman. Indeed, the continental margin of Arabia throughout the Mid-Permian to Cenomanian was completely stable, with the sedimentation of highly fossiliferous, shallow marine carbonates, continuing through the Cretaceous up to Cenomanian times, when the Semail ophiolite crustal sequence was formed and began to emplace SW onto the margin. The sedimentarystratigraphic record suggests that it is inconceivable that the stable, shelf carbonates of the Kahmah and Wasia Groups of the Cretaceous shelf sequence were deposited anywhere other than along a passive, aseismic margin (Glennie et al. 1974; Searle et al. 1983; Rabu et al. 1990; Scott 1990). Extremely detailed biostratigraphy and chronostratigraphy of the Cretaceous shelf sequence throughout the Oman mountains shows that the Lekhwair, Kharaib, Shuaiba and Nahr Umr Formations were deposited on a stable, gently subsiding wide continental shelf (Scott 1990). The first indication of the foreland basin deposition comes in the Coniacian-Santonian (88.5-84 Ma) when the shelf edge began to collapse to accommodate the incoming thrust sheets of the Hawasina, Haybi and Semail allochthons (Warburton et al. 1990). Throughout the period 130-95 Ma, required for SW subduction beneath the margin by the GrayGregory model, there is no indication of any tectonic or magmatic activity along the shelf margin during that period whatsoever. Thick carbonate passive margins do not normally overlie active subduction zones. Size and shape of the subducting microplate The Gray-Gregory model (Fig. 4) requires a suture zone to be present along the top of their proposed 'microplate' during their stage 2 SWdirected subduction. The size of the microplate could only be the area of the HP rocks in the As Sifah unit in the lower plate, a present-day area of <18km 2 , because there is no continental crust outboard of the As Sifah rocks. How could a microplate this small descend to 80 km depth with no continental crust outboard of it? There is no indication of any continental crust offshore NE

476

M. P. SEARLE ETAL.

Oman today, and there never has been in the past. There is no evidence of any suture zone above the As Sifah unit. Indeed, the protoliths of the As Sifah eclogites are Permian lava flows within the Saiq Formation carbonates, and the protoliths of the quartz mica schists could be lateral equivalents of the Ordovician Amdeh Formation quartzites. The Gray-Gregory model also would require the subduction polarity to flip. From 130 to 95 Ma it supposedly dipped SW beneath the Arabian margin, then from 95 Ma onwards it would have to dip NE, during ophiolite obduction (Fig. 4). Some active subduction zones are known to steepen and even overturn with depth, but to flip directions is not easily done. The entire mantle convection system would have to suddenly reverse. There is no evidence in Oman or the surrounding oceans to support this reversal of dip direction. 40

Ar/39Ar geochronological evidence for pre95 Ma HP metamorphism El-Shazly & Coleman (1990), El-Shazly et al (1990) and El-Shazly and Lanphere (1992) proposed two HP metamorphic events in NE Oman, based on 40Ar/39Ar ages of white micas. Micas from the highest-grade blueschists and eclogites from As Sifah yielded variably discordant age spectra with weighted mean 'plateau' ages of 111-106 Ma. Conventional K-Ar and 40Ar/39Ar dating of whole rocks and white micas from structurally higher units including the Ruwi melange gave ages of between 80 and 72 Ma. ElShazly and Lanphere (1992) interpreted these ages to record an early Cretaceous HP event, the result of partial subduction of the continental margin beneath a microcontinent, followed by a second HP-LT event during the Late Cretaceous ophiolite emplacement stage. Goffe et al (1988), LeMetour et al (1990), Michard et al (1984) and Searle et al (1994) argued that phengites from the HP blueschists and eclogites incorporated excess Ar and therefore none of the older (pre-95 Ma) ages were meaningful. Excess Ar is usually considered to result in saddle-shaped spectra. Although some of the older 40 Ar/39Ar phengite ages reported from the HP rocks in Oman by Montigny et al (1988) and Searle et al (1994) did have apparent good 'plateaux', it has become apparent that excess Ar can still become incorporated without showing a disturbed age spectrum. Searle et al (1994) attributed all HP metamorphism to the ophiolite obduction phase in the Late Cretaceous. El-Shazly et al (2001) then attempted to date more eclogites from As Sifah, all of which yielded saddle-shaped,

hump-shaped or irregular spectra with uninterpretable isochrons. Clinopyroxene-phengite, epidote-phengite and whole-rock-phengite Rb-Sr isochrons for the same samples gave cooling ages of 78 ± 2 Ma, so those workers, in contradiction to their earlier arguments, finally agreed that none of the old Ar ages were meaningful, and attributed all HP metamorphism and cooling to the Late Cretaceous event associated with ophiolite obduction. NE-facing and -verging fold nappes Gregory et al (1998) and Gray et al (2000) proposed that at least two major, SW-dipping crustal-scale shear zones bounding the NE-facing Saih Hatat fold (3 and 4 in Fig. 5) root into the Moho (Gray et al 2000, p. 514). There is actually no evidence from structural field mapping that these shear zones penetrate any deeper than the top of the Proterozoic Hatat schists. The metamorphic grade is almost identical below and above the upper plate-lower plate discontinuity (4 in Fig. 7), at least in the northern part of the area, so there is no metamorphic evidence for any great vertical throw on this fault. There is no evidence at all that either shear zone extends down to the lower crust, let alone down to the Moho. An alternative interpreted cross-section to the Gray-Gregory section of Figure 7 is shown in Figure 6. This shows the same geological structures in the upper part with the large-scale NEvergent fold of northern Saih Hatat. The interpretation of the lower structural units of the Hulw and As Sifah Windows is based on down-plunge and along-strike projection. The margins of the Saih Hatat dome are bounded by late-stage normal faults, which drop the Jebel Dimh ophiolite block down to the south along the southern flank, and the Wadi Kabir normal fault, which drops the Muttrah-Muscat peridotite down to the north along the northern flank. The Wadi Kabir fault is definitely an early Tertiary structure because it offsets Paleocene rocks. The Wadi Tayyin fault is more difficult to pin down because no Tertiary rocks are directly in contact with it. It could either be the very last stage of the late Cretaceous culmination of Saih Hatat or an early Tertiary fault, similar to the Wadi Kabir fault. Most workers who have mapped in northern Saih Hatat agree that very large-scale NE-vergent fold nappes with sheath-like folds are present in the upper plate units (LeMetour et al 1986, 1990; Searle et al 1994; Gregory et al 1998; Miller et al 1998, 2002). Stretching lineations are consistently NNE-SSW in the basement rocks, upper and lower plates, and in the As Sifah eclogites. Structurally higher units above As Sifah show at

SUBDUCTION ZONE POLARITY IN OMAN least three major thrust-bounded sheets, which show SW vergence and facing directions in the Wadi Aday-Wadi Mayh, Al Khuyran, Yenkit and Muttrah-Ruwi thrust sheets (Searle et al. 1994). These shear zones and faults are thought to young structurally upwards, as the highest ones offset lower Tertiary limestones at the highest structural levels around Ruwi and Bandar Jissah region. The eclogites and their enclosing calc-schists and quartz mica schists show isoclinal folding on all scales, and a strong flattening fabric as well as the 'extensional crenulation schistosity' described by Searle et al (1994). These fabrics were imposed during HP metamorphism and are entirely consistent with SW-directed exhumation of deeper, HP footwall rocks during the late Cretaceous NE-directed subduction. Following peak eclogite metamorphism these rocks were exhumed rapidly back up the same subduction zone and then rolled over into domal culminations as the thrusts climbed up-section to the SW. The extensional crenulation schistosity was imposed by exhumation of deeper, higher-pressure footwall units relative to shallower, hanging-wall units, so they are 'extensional' structures developed in a purely compressive environment. Nothing in the microstructures or fabrics in the As Sifah eclogites supports deep continental subduction SW beneath the continental margin.

Discussion Many other orogenic belts show divergent thrust and fold structures like the southeastern part of the Oman mountains. In the Alps, soumeastwards subduction of European lower crust beneath the Adriatic lower crust has resulted in a fan structure (Schmid et al. 1996; Pfiffner et al. 2000). The upper-crustal Aar, Gotthard and Austro-Alpine nappes, and the Mesozoic-Cenozoic sedimentary rocks of the Helvetic and Penninic nappes are all NW-vergent. Tectonic wedging led to backthrusting along the Insubric mylonite belt and in the SE-vergent South Alpine thrust sheets. The basal thrust of the Helvetic nappes (Glarus thrust) was propagating NW into the foreland Swiss molasse basin at the same time as the backthrusting occurred along the Insubric line. The wedging of the lower crust contrasts strongly with the piling up of upper-crustal Alpine nappes, in a very similar fashion to the structures in Oman. In the late collisional stage the thickness of the European crust increased as more proximal units (underlying the Helvetic nappes) entered the subduction zone (Schmid et al. 1996). This led to detachment of the upper and lower crust and hints at a relatively strong lower crust. The situation is again very similar to Oman, where the deepest subducted

477

rocks (As Sifah eclogites) were part of the thinned continental crust. The upper-crustal rocks originally above the As Sifah unit choked the subduction zone and were piled up into a series of nappes, backthrust to the NE (Fig. 6). The northern margin of the Indian plate along the 2500 km strike of the Himalayan orogen also shows regional-scale backthrusting and backfolding, verging and facing to the north (e.g. Yin et al. 1999; Corfield & Searle 2000). During the collision of India and Asia, the Himalayan thrusts developed in sequence from the zone of collision in the north, the Indus-Tsangpo suture zone (c. 50 Ma) to the Main Central Thrust (c. 20-17 Ma) to the Main Boundary Thrust (<10Ma), the seismically active fault along the southern margin of the Himalaya. The north-dipping, south-vergent thrusts mainly developed in sequence, propagating towards the Indian foreland. During the later stages of orogeny, however, out-of-sequence thrusts cut through the previously stacked Tethyan Himalaya, and north-vergent backfolds and backthrusts developed along the northern margin of the Indian plate and across the suture zone. These divergent thrusts formed a very large-scale 'popup' structure, c. 50 km wide and extending the length of the Himalaya. Tectonic wedging resulted in Indian lower crust being subducted north beneath southern Tibet and Indian plate upper crust flaking off, forming large north-vergent nappes. In Oman, tectonic wedging resulted from NEdirected subduction of lower crust whereas the upper crust was flaked off, forming NE-vergent fold nappes. The Gray-Gregory model makes the assumption that NE-verging fabrics and sense of shear is indicative of SW-directed subduction, an assumption that is not correct, as shown from numerous orogenic belts, including the Alps and the Himalaya.

Conclusions We refute recent suggestions that the tectonic evolution of northern Oman involved an initial period of SW-directed subduction beneath northern Oman during the interval 130-95 Ma, before formation of the Semail ophiolite, as proposed by Gregory et al (1998), Gray & Gregory (2000) and Gray et al. (2000). Those workers proposed their model of a nascent Early Cretaceous subduction beneath the Oman continental margin (Model B) based mainly on two factors: older 40Ar/39Ar cooling ages, not all of which they claimed were a result of excess Ar, and early fabrics in the highpressure rocks indicating a south-over-north sense of shear. However, nothing in the entire geological record of the basement, shelf carbonates, Hawasi-

478

M. P. SEARLE ETAL.

na-Haybi allochthon or Semail ophiolite suggests that there was any tectonic activity during the period that the Gray-Gregory model requires deep subduction beneath the passive margin. There is no evidence of a suture zone at the site of this detachment (above the As Sifah eclogites). The sedimentary evolution of the shelf carbonate sequence during the early and mid-Cretaceous records a completely stable passive carbonate margin during that time (Scott 1990), which could not possibly have overlain a SW-dipping subduction zone. Pre-ophiolite formation 40Ar/39Ar cooling ages of white micas from the As Sifah region HP terrane have uninterpretable spectra that reflect excess Ar (Searle et al 1994; El-Shazly et al 2001). NE-vergent folds and NE-directed sense of shear kinematic indicators are real, but reflect two processes. NE-vergent 'extensional' fabrics in the deep-level eclogites reflect SW-directed footwall uplift of HP rocks relative to lower-pressure rocks above (in other words, the footwall moves up whereas the hanging wall remains stationary). The upper-crustal NE-directed sheath folds and nappes are high structural level retro-shears and backfolds during NE-directed subduction. The shortening within the upper plate NE-directed backfolds should be balanced by the shortening within the subducted and exhumed higher-pressure lower plate. We suggest that all structures and metamorphism in northern Oman are related to one extended period of ophiolite obduction, which lasted for c. 27 Ma from c. 95 to c. 68 Ma. Initial oceanic subduction of basaltic rocks beneath the ophiolite formed amphibolites of the metamorphic sole at the same time as the ophiolite crustal section was crystallizing. The timing constraints imply that the ophiolite must have formed in a suprasubduction zone position with the subduction zone dipping NE away from the continental margin. Thrusting then propagated southwestwards with time, accreting successively higher units to the base of the amphibolites (greenschists, then unmetamorphosed Haybi complex thrust sheets). Thrusting continued with distal Tethyan oceanic units emplaced over proximal units (Haybi complex over distal Hawasina complex over proximal Hamrat Duru Group). During the later stages of obduction, the thinned continental margin was subducted to depths of c. 80 km forming eclogites, which record pressures of up to c. 20kbar (Searle et al. 1994). Exhumation of the HP eclogites occurred back up the same zone with HP rocks moving up along the footwall of major shear zones, which record relative extensional microstructures. Tectonic wedging resulted in NE-facing nappes and NE-directed shears developing in the upper crust as the continental margin choked the subduction zone.

We thank Hilal al-Azri of the Ministry of Petroleum in Muscat, and David and Judy Willis for support. M.P.S. would like to thank Edwin Gnos, Dave Gray and Bob Gregory for vigorous discussions in the field during the Oman ophiolite conference in January 2001. We also thank the reviewers Ron Harris and John Wakabayashi for very helpful reviews. We also acknowledge NERC grant NER/K/S/2000/951 to M.P.S.

References BAILEY, E.H. 1981. Geologic map of the Muscat-Ibra area, Sultanate of Oman (pocket map). Journal of Geophysical Research, 86. BERNOULLI, D. & WEISSART, H. 1987. The Upper Hawasina nappes in the central Oman mountains; stratigraphy, palinspastics and sequence of nappe emplacement. Geodinamica Acta, 1, 47-58. BOOTE, D.R.D., Mou, D. & WAITE, R.I. 1990. Structural evolution of the Suneinah Foreland, Central Oman. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 397-418. CAWOOD, P.A., GREEN, F.K. & GALON, TJ. 1990. Origin of culminations within the Southeast Oman Mountains at Jebel MaJool and Ibra dome. /W.ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 429-446. COOPER, D.J.W. 1988. Structure and sequence of thrusting in deep-water sediments during ophiolite emplacement in the south-central Oman mountains. Journal of Structural Geology, 10, 473-485. CORFIELD, R.I. & SEARLE, M.P. 2000. Crustal shortening estimates across the north Indian continental margin. In: KHAN, M.A., TRELOAR, P.J., SEARLE, M.P. & JAN, M.Q. (eds) Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya. Geological Society, London, Special Publications, 170, 395-410. DUNNE, L.A., MANOOGIAN, P.R. & PIERINI, D.F. 1990. Structural style and domains of the Northern Oman Mountains (Oman and United Arab Emirates). In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 375-386. EL-SHAZLY, A.K. & COLEMAN, R.G. 1990. Metamorphism in the Oman Mountains in relation to Semail ophiolite emplacement. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 473-494. EL-SHAZLY, A.K. & LANPHERE, M.A. 1992. Two highpressure metamorphic events in NE Oman: evidence from 40Ar/39Ar dating and petrological data. Journal of Geology, 100, 731-751. EL-SHAZLY, A.K. & Liou, J.G. 1991. Glaucophanechloritoid-bearing assemblages from NE Oman: petrologic significance and a petrogenetic grid for high P metapelites. Contributions to Mineralogy

SUBDUCTION ZONE POLARITY IN OMAN and Petrology, 107, 180-201. EL-SHAZLY, A.K., COLEMAN, R.G. & Liou, J.G. 1990. Eclogites and blueschists from Northeastern Oman: petrology and P— revolution. Journal of Petrology, 31, 629-666. EL-SHAZLY, A.K., BROKER, W., HACKER, B. & CALVERT, A. 2001. Formation and exhumation of blueschists and eclogites from NE Oman: new perspectives from Rb—Sr and 40Ar/39Ar dating. Journal of Metamorphic Geology, 19, 233-248. GLENNIE, K.W., BOEUF, M.G., HUGHES-CLARKE, M.H.W., MOODY-STUART, M., PILAAR, W.F. & REINHARDT, B.M. 1973. Late Cretaceous nappes in the Oman mountains and their geologic evolution. AAPG Bulletin, 57, 5-27. GLENNIE, K.W., BOEUF, M.G., HUGHES-CLARKE, M.H.W., MOODY-STUART, M., PILAAR, W.F. & REINHARDT, B.M. 1974. Geology of the Oman Mountains. Verhandelingen Koninklijk Nederlands Geologisch Mijnbouwkundidg Genootschap, 31. GNOS, E. 1998. Peak metamorphic conditions in garnet amphibolites beneath the Semail ophiolite: implications for an inverted pressure gradient. International Geological Review, 40, 281-304. GNOS, E. & PETERS, TJ. 1993. K-Ar ages of the metamorphic sole of the Semail ophiolite: implications for ophiolite cooling history. Contributions to Mineralogy and Petrology, 113, 325-332. GOFFE, B., MlCHARD, A., KlENAST, J.R.

& LfiMER,

O. 1988. A case of obduction related high P low T metamorphism in upper crustal nappes, Arabian continental margin, Oman: P-T paths and kinematic interpretation. Tectonophysics, 151, 363-386. GRAY, D.R. & GREGORY, R.T. 2000. Implications of the structure of the Wadi Tayin metamorphic sole, the Ibra-Dasir block of the Semail ophiolite, and the Saih Hatat window for late stage extensional ophiolite emplacement, Oman. Marine Geological Researches, 21, 211-227. GRAY, D.R., GREGORY, R.T. & MILLER, J.M. 2000. A new structural profile along the Muscat-Ibra transect, Oman: implications for emplacement of the Semail ophiolite. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 513-523. GREGORY, R.T., GRAY, D.R. & MILLER, J.M. 1998. Tectonics of the Arabian margin associated with the emplacement of the Oman margin along the Ibra transect: new evidence from northeast Saih Hatat. Tectonics, 17, 657-670. HACKER, B.R. 1994. Rapid emplacement of young oceanic lithosphere: argon geochronology of the Oman ophiolite. Science, 265, 1563-1565. HACKER, B.R. & MOSENFELDER, J.L. 1996. Metamorphism and deformation along the emplacement thrust of the Samail ophiolite, Oman. Earth and Planetary Science Letters, 144, 435-451. HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman ophiolite: thermal and geochronological constraints. Tectonics, 15, 1230-1247.

479

HANNA, S. 1990. The alpine deformation of the Central Oman mountains. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 341-359. LEMETOUR, J., DEGRAMONT, X. & VILLEY, M. 1986. Geological Map of Masqat and Quryat and Accompanying Notes, Scale 1:100000. Ministry of Petroleum and Minerals, Sultanate of Oman. LEMETOUR, J., RABU, D., TEGYEY, M., BECHENNEC, F., BEURRIER, M. & VILLEY, M. 1990. Subduction and obduction: two stages in the Eo-Alpine tectonometamorphic evolution of the Oman mountains. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 327-339. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1986. The Ophiolite of Northern Oman. Blackwell Scientific, Oxford. MANN, A. & HANNA, S. 1990. The tectonic evolution of pre-Permian rocks, central and southeastern Oman Mountains. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 307-325. MlCHARD, A., BOUCHEZ, J.-L. & OUAZZANI-TOUHAMI,

M. 1984. Obduction-related planar and linear fabrics in Oman. Journal of Structural Geology, 6, 39-49. MILLER, J.M., GRAY, D.R. & GREGORY, R.T. 1998. Exhumation of high-pressure rocks in northeastern Oman. Geology, 26, 235-238. MILLER, J.M., GREGORY, R.T., GRAY, D.R. & FOSTER, D.A. 1999. Geological and geochronological constraints on the exhumation of a high-pressure metamorphic terrane, Oman. In: RING, U., BRANDON, M.T. & LISTER, G. (eds) Exhumation Processes; Normal Faulting, Ductile Flow and Erosion. Geological Society, London, Special Publications, 154,241-260. MILLER, J.M., GRAY, D.R. & GREGORY, R.T. 2002. Geometry and significance of internal windows and regional isoclinal folds in northeast Saih Hatat, Sultanate of Oman. Journal of Structural Geology, 24, 359-386. MONTIGNY, R., LEMER, O., THUIZAT, R. & WHITECHURCH, H. 1988. K-Ar and 40Ar/39Ar study of metamorphic rocks associated with the Oman ophiolite: tectonic implications. Tectonophysics, 151, 345-362. PFIFFNER, O.A., ELLIS, S. & BEAUMONT, C. 2000. Collision tectonics in the Swiss Alps: insight from geodynamic modeling. Tectonics, 19, 1065-1094. RABU, D., LEMETOUR, J., BECHENNEC, F., BEURRIER, M., VILLEY, M. & BOURDILLON-DE-GRISSAC, C. 1990. Sedimentary aspects of the Eo-Alpine cycle on the northeast edge of the Arabian platform. /W.ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 49-68. ROBERTSON, A.H.F. & SEARLE, M.P. 1990. The northern Oman Tethyan continental margin: stratigraphy,

480

M. P. SEARLE ETAL.

structure, concepts and controversies. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 3-26. SCHMID, S.M., PFIFFNER, O.A., FROITZHEIM, N., SCHONBORN, G. & KISSLING, E. 1996. Geophysicalgeological transect and tectonic evolution of the Swiss-Italian Alps. Tectonics, 15, 1036-1064. SCOTT, R.W. 1990. Chronostratigraphy of the Cretaceous carbonate shelf, southeastern Arabia. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 89-108. SEARLE, M.P. 1985. Sequence of thrusting and origin of culminations in the northern and central Oman Mountains. Journal of Structural Geology, 7, 129-143. SEARLE, M.P. 1988a. Thrust tectonics of the Dibba zone and the structural evolution of the Arabian continental margin along the Musandam Mountains (Oman and United Arab Emirates). Journal of the Geological Society, London, 145, 43-53. SEARLE, M.P. 1988b. Structure of the Musandam culmination (Sultanate of Oman and United Arab Emirates) and the Straits of Hormuz syntaxis. Journal of the Geological Society, London, 145, 831-845. SEARLE, M.P. & COOPER, DJ.W. 1986. Structure of the Hawasina Window culmination, central Oman mountains. Transactions of the Royal Society, Edinburgh, Earth Sciences, 77, 143-156. SEARLE, M.P. & Cox, J.S. 1999. Tectonic setting, origin and obduction of the Oman ophiolite. Geological Society of America Bulletin, III, 104-122. SEARLE, M.P. & Cox, J.S. 2002. Subduction zone metamorphism during formation and emplacement of the Semail Ophiolite in the Oman mountains. Geological Magazine, 139, 241—255. SEARLE, M.P. & MALPAS, J. 1980. The structure and metamorphism of rocks beneath the Semail Ophiolite of Oman and their significance in ophiolite

obduction. Transactions of the Royal Society, Edinburgh, Earth Sciences, 71, 247-262. SEARLE, M.P. & MALPAS, J. 1982. Petrochemistry and origin of sub-ophiolite metamorphic and related rocks in the Oman Mountains. Journal of the Geological Society, London, 139, 235-248. SEARLE, M.P., JAMES, N.P., CALON, TJ. & SMEWING, J.D. 1983. Sedimentological and structural evolution of the Arabian continental margin in the Musandam mountains and Dibba zone, United Arab Emirates. Geological Society of America Bulletin, 94, 1381-1400. SEARLE, M.P., COOPER, DJ.W. & WATTS, K. 1990. Structure of the Jebel Sumeini-Jebel Ghawil areas, Northern Oman. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 361-374. SEARLE, M.P., WATERS, D.J., MARTIN, H.N. & REX, D.C. 1994. Structure and metamorphism of blueschist-eclogite facies rocks from the northeastern Oman mountains. Journal of the Geological Society, London, 151, 555-576. TILTON, G.R., HOPSON, C.A. & WRIGHT, J.E. 1981. Uranium—lead isotopic ages of the Samail ophiolite, Oman with application to Tethyan ocean ridge tectonics. Journal of Geophysical Research, 86, 2763-2775. TIPPIT, P.R., PESSAGNO, E.A. & SHEWING, J.D. 1981. The biostratigraphy of sediments in the volcanic unit of the Samail Ophiolite. Journal of Geophysical Research, 86, 2756-2762. WARBURTON, J., BURNHILL, T.J., GRAHAM, R.H. & ISAAC, K.P. 1990. The evolution of the Oman Mountains foreland basin. In: ROBERTSON, A.F.H., SEARLE, M.P. & RIES, A.C. (eds) The Geology and Tectonics of the Oman Region. Geological Society, London, Special Publications, 49, 419-427. YIN, A., HARRISON, T.M. & MURPHY, M.A. ET AL. 1999. Tertiary deformation history of southeastern and southwestern Tibet during the Indo-Asian collision. Geological Society of America Bulletin, III, 1644-1664.

Geoch namic patterns of ophiolites and marginal basins in the Indonesian and New Guinea regions RON HARRIS Department of Geology, Brigham Young University, Provo, UT 84602-4606, USA (e-mail: [email protected]) Abstract: Analysis of spatial, temporal, geological and geochemical patterns of ophiolites in the Indonesian and New Guinea region indicates a strong correlation with marginal basin development and closure. The spatial distribution of ophiolites is mostly linked with marginal basin producing zones of oblique convergence and collision. Strain partitioning in these zones creates a series of ephemeral plate boundaries between several independently moving lithospheric blocks. Repeated disruption of the diffuse boundaries between the blocks by changes in plate motion and collision-induced mantle extrusion creates space that is rapidly filled by new ocean basins in the upper plate of subduction zones. Suprasubduction zone (SSZ) spreading of these basins is enhanced by episodic extrusion of asthenosphere escaping collisional suture zones. Various closure events and global plate motion changes are reflected in the temporal distribution of marginal basin and ophiolite ages. Most ophiolite slabs in the Indonesian and New Guinea region represent fragments of oceanic lithosphere with a subduction zone component, as indicated by the common refractory petrochemistry of the mantle sequence and occurrence of boninite. Age and compositional heterogeneity may indicate that some ophiolite bodies are composite terranes. Collisions with buoyant lithosphere transform parts of these ocean basins into ophiolites. The connection between ophiolites and marginal basins is strongest where parts of actively spreading SSZ basins are partially represented as ophiolites in collision zones.

Many of the various tectonic models proposed for the origin and tectonic evolution of ophiolites are supported by plate interactions in the Indonesian and New Guinea region (Dewey & Bird 1970; Silver & Smith 1983; Moores et al 1984; Searle & Stevens 1984; Dilek & Moores 1990; Harris 1992; Dickinson et al. 1996). However, the geological associations of these interactions remain poorly resolved, allowing different parts of the region to be used as modern analogues for almost any tectonic scenario. Simple explanations based on plate kinematics alone fail to account for the increasing levels of complexity observed in the Indonesian and New Guinea region. New ideas and supporting evidence for mantle extrusion mechanisms (e.g. Flower et al. 2001) also rely heavily on descriptions of plate interactions in this region (Doglioni et al. 1999), and raise new questions about how these processes may relate to the spatial and temporal pattern of ophiolites. This paper provides a synthesis of the temporal and spatial patterns, and tectonic associations of marginal basins and ophiolites throughout the Indonesian and New Guinea region. It explores the relations between various plate boundary processes and ophiolite genesis and emplacement, such as the role of strain partitioning, trench

rollback, subduction polarity asymmetry, asthenospheric flow and plate kinematics.

Tectonic evolution of the Indonesian and New Guinea region and its contribution to the ophiolite debate The Indonesian and New Guinea region is an active collisional amalgamation of the complex Asian and Australian continental margins with crustal heterogeneities of the western Pacific (Hamilton 1979; Hall & Blundell 1996; Hall 2002). Rapid changes in plate boundary location and function in this region are emblematic of incipient phases of continental collision manifest in ophiolite-bearing Cordilleran and Tethyan mountain systems. As in these regions, the Indonesian and New Guinea region has buffered the changing motion and boundaries of some of the Earth's largest plates. Convergence of the IndoAustralian plate from the SW has mostly been absorbed along the Sunda arc-trench system, whereas convergence of the Pacific plate from the east has progressed by sequential movement along an array of short-lived subduction zones and spreading centres. Similar plate boundary asym-

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 481-505. 0305-8719/037$ 15 © The Geological Society of London 2003.

482

R. HARRIS

metry is observed in many other parts of the world (Uyeda & Kanamori 1979; Doglioni et al. 1993). The triple junction between the Indo-Australian, Pacific and Asian plates is a complex repository of island arcs, marginal basins, continental fragments and ophiolites amalgamated by repeated plate boundary reorganizations. Many of the oceanic terranes in the mix were emplaced onto the edge of partially subducted continental margins that began arriving at the triple junction during the mid-Tertiary. The incipient continent-continent collision has progressed to the stage where the Sunda Shelf of Asia and the Sahul Shelf of Australia are now partially connected by a collage of island arcs and continental fragments separated by trapped and partially obducted ocean basin lithosphere (Fig. 1). An analogous stage of tectonic development to that in the Indonesian and New Guinea region may have existed in the Jurassic amalgamation of Alaska between North America and Siberia, or the Mediterranean region between Africa and Europe. However, in both of these locations, ophiolites are all that remains of what once was a complex plate boundary system that included ocean basins that have since been closed and partially obducted. The survival of many of these features in the Indonesian and New Guinea region offers a unique perspective into the connection between ophiolites and marginal basins. Although ocean basin closure along this plate boundary zone has already progressed to the stage of producing several classic ophiolites, many of the original tectonic features, such as west-verging subduction zones, that produced these ophiolites remain intact. These features provide a rare glimpse of the complexity associated with convergent triple junctions involving continents. From the Andaman Sea to New Guinea (Fig. 1), the marginal basins and suprasubduction zone (SSZ) ophiolites in the natural laboratory of the Indonesian and New Guinea region contribute in many ways to the continuing debate about the geodynamic meaning of ophiolites worldwide. Global positioning system (GPS) and earthquake studies throughout the region reveal many fragmentary and transitional plate boundary zones capable of simultaneous opening, closure and obduction of new oceanic lithosphere. The age and chemistry of oceanic crust in these ocean basins and the ophiolites that surround them provide a context for interpreting ophiolites in mountain systems where the original tectonic setting is obscured. Most ophiolites of the Indonesian and New Guinea region also have some component of refractory or SSZ tectonomagmatism, like most other ophiolites. In some cases the

SSZ history is superimposed onto or evolves into less refractory lithosphere or is sourced from mantle previously contaminated by subduction. Geophysical and geological studies of plate boundaries throughout the ING region demonstrate, at a range of temporal and spatial scales, that plate convergence is partitioned by an intricate network of faults that separate independently moving lithospheric blocks (Hamilton 1979; Bowin et al. 1980; Cardwell et al. 1980; McCaffrey et al. 1985; McCaffrey 1988; Curray 1989; Audley-Charles & Harris 1990; Genrich et al 1996; Hall & Blundell 1996; Kroenke 1996; Pubellier & Cobbold 1996; Simons et al 1999; Stevens et al 1999; Kreemer et al 2000). The rapid rates of movement and transitional nature of these microplate boundaries provide a variety of settings for SSZ ocean basin development. Eastward asthenospheric flow may also play an important role in steepening subduction zones and facilitating marginal basin development (e.g. Doglioni 1993). It is the combination of strain partitioning, plate boundary reorganization, asthenospheric flow, and collisional termination of subduction in the Indonesian and New Guinea region that produces and transforms many newly formed SSZ ocean basins into ophiolite-bearing collisional mountain systems. Each stage of this process is represented by the evolving nature of plate boundaries in the Indonesian and New Guinea region, where many ophiolites already abound.

Transitional nature of plate boundary reorganization The interaction of the three large plates that converge in the Indonesian and New Guinea region produces a wide zone of plate boundary deformation composed of several fault-bounded oceanic and continental fragments (Fig. 2). Most of the fault zones between these discrete lithospheric blocks are ephemeral and accommodate strain for only a few million years or less. The most stable and long-lived plate boundary is the north-directed Sunda subduction system (Fig. 2). However, this boundary has been intensely modified by collision with India in the west and Australia in the east. In both cases indentation of the convergent boundary has increased its obliquity and resulted in partitioning of strain away from the trench into the weak parts of the upper plate. The collision of the Indo-Australian and Asian plates is further modified by the rapid westerly convergence of a series of Pacific microplates linked by multiple subduction zones, spreading centres and large strike-slip faults (Figs

Fig. 1. Location map of ophiolites and marginal basins of the Indonesian and New Guinea region. Arrows indicate plate motion relative to the Asian plate. Locations of more detailed maps shown in boxes, urn, ultramafic rocks.

Fig. 2. Plate boundaries, active faults and marginal basin age of the ING region (Modified from Hall, 1996).

Fig. 3. Tectonic map of the eastern Indonesian region. Banda Sea marginal basins are shown in slanted stripes. Lines of section shown in Figure 4 are heavy black lines. (Modified from Hamilton, 1979)

486

R. HARRIS

2 and 3). The convergent triple junction of the three plates has been periodically disrupted by collision with island arcs, oceanic plateaux and continental fragments, and the accompanying extrusion of asthenosphere. It is in these collisional settings that most ophiolites are found.

Seismic tomography Tomographic images of the mantle to depths of 1400 km produced by Widiyantoro & van der Hilst (1997) and Hafkenscheid et al (2001) show that the entire Indonesian and New Guinea region is underlain by subducted lithosphere. Up to 5400km of convergence is estimated along the Sunda arc system, resulting in a high P-wave velocity anomaly that stretches from Java to the southern Philippines. The subducting plate begins to lose its seismic and tomographic expression below 600 km depth where it merges with a more diffuse zone of high-velocity mantle that underlies the entire region at depths of 600-1400 km. Wellimaged slabs of lithosphere feed into this highvelocity zone (slab graveyard) from the Pacific side in the New Guinea, northern Banda arc, Molucca and Philippine regions. Tomographic maps of P-wave velocities at 200 and 500km depth show that all of the actively spreading marginal basins of the Indonesian and New Guinea region are underlain by subducted lithosphere. West-directed slabs are generally steeply inclined, whereas east-directed slabs dip less than 45°. The influence of these slabs on the petrochemistry and stress history of the marginal basins above them is poorly resolved.

NE-directed subduction beneath the Asian continental margin Subduction along the southern and western margins of the Sunda craton is long-lived and dates back to at least the Early Cretaceous (Hamilton 1979). The most recent expression of subduction along this margin is the Late Neogene Sunda arc, which has formed on top of older subduction complexes accreted to the edge of the Sunda craton (Fig. 2). The Sunda arc-trench system stretches for nearly 6000km from Myanmar, where it is terminated by continental collision with India, to the Banda Sea, where it is transitional with the Banda arc-continent collision (Figs 1 and 2). The active accretionary wedge of the Sunda arc consists mostly of Late Paleogene to Recent cover sediments accreted from the subducting Indian Ocean sea floor (Hamilton 1979). Higher rates of accretionary influx exist in the NW because of the

influence of the Bengal Fan, and at the eastern end of the Sunda arc where the Australian passive margin collides with the trench. In both locations the crest of the accretionary wedge emerges to form a chain of forearc islands (Fig. 1) that reveal that little, if any, oceanic lithosphere has accreted from the lower, subducting plate. If the lower plate were a common source of ophiolites then there should be an extensive repository of them accreted to the edge of the Sunda craton. However, exposures both on forearc islands and of the older accretionary wedge have only small blocks of oceanic material embedded in melange, and no ophiolites are documented (Ketner et al. 1976; Hamilton 1979). The 7-8 cm a"1 of plate convergence measured between the Indo-Australian plate and the Sunda arc (Tregoning et al. 1994; Kreemer et al. 2000) is mostly taken up near the trench south of Java, where convergence is orthogonal (Stockmal 1983; McCaffrey 1996). Strain is increasingly partitioned away from the trench at its western and eastern ends, where collisional indentation increases the obliquity of convergence and resistance to subduction (Curray 1989; McCaffrey 1996). West of Java a major component of convergence is distributed to the Mentawai and Sumatra faults (Fig. 1). Obliquity of convergence also increases to the east in the Banda arc region, where plate boundary redistribution is more obscure (Harris 1991; Genrich et al. 1996). In both of these pericollisional settings of oblique convergence SSZ oceanic basins have formed within the Sunda arc-trench system. The Banda, Bali and Flores basins formed in the east, and the Andaman Sea basin is currently opening in the west (Fig. 1).

Andaman Sea The Andaman Sea is an actively spreading interarc basin with a tectonic history associated with ridge subduction (Eguchi et al. 1979), pericollisional extension or 'rip-off basin' development (Lee & Lawver 1994), and highly oblique convergence (Curray 1989). Partitioning of strain away from the trench in the region is manifest by the rightlateral Mentawai Fault in the forearc and the Sumatra Fault astride the volcanic arc (Fig. 2). These fault systems segment the NW Sunda arctrench system into three orogen-parallel sliver plates (Fig. 2) that have a component of northward motion with the subducting Indian plate relative to the Sunda craton (Curray 1989; Malod & Kemal 1996; Prawirodirdjo et al 1997). Episodic SSZ spreading between these sliver plates in the Andaman Sea region began in the Oligocene with the formation of NNE-SSWtrending grabens of the Mergui and North Sumatra

OPHIOLITES OF SE ASIA AND NEW GUINEA continental back-arc basins (Curray 1989). At 13 Ma NW-SE-directed extension began to form the present oceanic basin within the Mergui Shelf. An overall spreading rate of 4 cm a"1 is estimated from the combined motion along six short, eastwest-trending ridge segments linked by long (c. 100 km) transform faults (Guzman-Speziale & Ni 1993). The total amount of extension is estimated at around 460 km (Curray 1989). This estimate is similar to the amount of right-lateral strike-slip offset measured along the Sagaing Fault to the north (Hla-Maung 1987), which links the Andaman Sea to the eastern Himalayan syntaxis. Active volcanism in the Andaman Sea is found mostly around the Barren and Narcondam Islands. The sea-bed extent of these volcanoes forms a north-south ridge about 150km long that is bounded by the west Andaman Fault (Weeks et al. 1967). Volcanic rocks include basalt, dacite, andesite, and basaltic-andesite suites with little or no volcanic ash (Dasgupta et al. 1990). Although no petrochemical data are available from Andaman Sea oceanic basement, the active volcanic region yields major and trace element abundances that plot in both the island arc tholeiite (IAT) and midocean ridge basalt (MORE) fields of discriminate diagrams (Haldar et al. 1992). Before the onset of sea-floor spreading in the Andaman Sea region, the Sunda arc formed a continuous belt of magmatism that stretched from Sumatra northward into Myanmar (Mitchell 1993). The opening of the Andaman Sea transformed the linear trace of stratovolcanoes of the Sunda arc into a diffuse zone of SSZ spreading centres from the northern tip of Sumatra to the southern coast of Myanmar (Fig. 2). The 'disappearance' of the arc is a common characteristic of many Tethyan-rype ophiolites throughout the world (Mitchell 1983). Moores et al. (1984) related the segmented and discontinuous nature of Andaman Sea SSZ spreading centres to geological associations of eastern Mediterranean ophiolites. Parallels also exist between the tectonic setting of the Andaman Sea and reconstructions of the belt of Jurassic ophiolites found throughout the Cordillera of North America (e.g. see Dickinson et al. 1996). Emphasizing the ephemeral nature of the Andaman Sea SSZ ocean basin, Curray (1989) compared it with the Rocas Verdes ophiolite complex of the southern Andes (Dalziel 1981). The Late Cretaceous and early Eocene Andaman ophiolite forms the southernmost part of a 2500km long north-south belt of dismembered ophiolites extending from the Naga Hills near the eastern syntaxis of the Himalaya to the Andaman-Nicobar Islands (Sengupta et al. 1990). The ophiolites are interpreted as remnants of narrow, marginal ocean basins that were opened and

487

closed within a short time immediately preceding the collision of India (Acharyya et al. 1989). Two groups of volcanic rocks are found throughout this belt. The dominant high-TiOa type is interpreted as off-axis seamount basalt because of the mild alkaline affinity and within-plate trace and rare earth element (REE) characteristics (Jafri et al. 1990). The other group of volcanic rocks ranges in composition from basaltic andesite to more felsic differentiates that may be cogenetic with plagiogranite found intruding gabbro (Ray et al. 1988). These units have characteristics of both island arcs and MORB, and may be associated with a marginal basin similar to the present Andaman Sea (Ray et al. 1988; Jafri et al. 1990). Emplacement of the ophiolites occurred during the Early Eocene oblique collision of the Myanmar and Indian continents (Acharyya et al. 1989).

Banda Sea basins The Banda Sea is an interarc basin composed of a series of east-west-trending continental and volcanic ridges separated by anomalously deep oceanic basins (Fig. 3). These ocean basins accommodated the eastern expansion of the Sunda arc to form the Banda arc, which is currently in collision with the northern margin of Australia (Fig. 3). The Banda Sea has been interpreted as trapped fragments of old Pacific or Indian Ocean lithosphere (Lapouille et al. 1985), as an older continuation of the Celebes and Sulu Basins to the north (Lee & McCabe 1986), and as a young back-arc basin (Hamilton 1979). Dredge samples (Silver et al. 1985; Guilou et al 1998; Honthaas et al. 1998) and mapping of magnetic anomalies (Hinschberger et al. 2001) support the interpretation by Hamilton (1979) and suggest that the ocean basins opened during the Late Miocene to Pliocene, within the stretched eastern edge of the Asian continental margin. Separating the sub-basins are the Banda and Nieuwerkerk-Emperor of China (NEC)- Lucipara Ridges (Fig. 3). These ridges, along with the Wetar Ridge to the south, consist mostly of Neogene volcanic deposits mounted on much older volcanic and continental metamorphic basement that formed the structurally attenuated edge of the southern Sunda craton and possibly slivers of the Australian continent (Silver et al. 1985; Honthaas etal. 1998). Studies of dredge samples and magnetic anomalies in sub-basins of the Banda Sea are interpreted to show that the basins opened sequentially toward the SE behind the incipient Banda volcanic arc (Honthaas et al 1998; Hinschberger et al 2001), predicting that the South Banda Basin is the youngest feature. However, the magnetic surveys

488

R. HARRIS

cover only a small area, and although the proposed magnetic section used by Hinschberger et al. (2001) fits the Late Miocene to Pliocene section well, it does not show how many other parts of the time scale the section may also fit. Another problem is that the area around the Banda ridges (Fig. 3) is hot, bathymetrically high and associated with east-west-trending transform structures (Silver et al. 1985), whereas the South Banda Basin is deep, cold and buried by 1-2 km of Pliocene-Quaternary sediment. These data suggest a difference in both age and tectonic evolution between the North and South Banda Basins. Geochronological studies of the North Banda Basin (Fig. 3) suggest that it opened first at around 9.4-7.3 Ma, and separated the Banda Ridges from the SE part of Sulawesi (Honthaas et al. 1998). Dredge samples from horst blocks within the basin and along its margins are basalts and trachyandesites with negative Nb and Ta anomalies in the range of back-arc basin basalt (BABB) (Honthaas etal 1998). The South Banda Basin opened within the stretched continental borderlands of the southern Sunda Shelf, and is transitional with the Bali and Flores Basins. The NEC-Lucipara Ridges form the northern boundary of the South Banda Basin and the Wetar Ridge forms the southern boundary. Geochronology and magnetic anomaly studies are interpreted to indicate a spreading event at around 6-3 Ma (Honthaas et al. 1998; Hinschberger et al. 2001). Spreading rates of 3 cm a"1 are estimated along what are interpreted as several arc-parallel ridge segments (Hinschberger et al. 2001). Although no samples have been dredged from the floor of this basin, geochemical and geochronological analysis of samples dredged from the Wetar and NEC-Lucipara Ridges have strong affinities, suggesting that each of these ridges represents fragments of a single magmatic arc that was split by the opening of the South Banda Basin (Honthaas et al. 1998). Calc-alkaline andesites with high ratios of large ion lithophile elements to high field strength elements (LILE/HFSE), negative Nb anomalies, and some cordierite, are a dominant feature of both ridges. OIB-type transitional volcanic rocks are also found, as in the North Fiji back-arc basin (Honthaas et al. 1998). The southern portion of the South Banda Basin has been overthrust by the Wetar Ridge along the Wetar Thrust, which now takes up some of the convergence between the Australia and Asia plates (Silver et al. 1983; Harris 1991; Genrich et al. 1996). Loading associated with the Wetar Thrust may explain the anomalous depth vs. young age of the South Banda Basin. Fragments of the Wetar Ridge and southern Banda Sea, known as the Banda Terrane (Harris

1992), also occur as nappes thrust over partially subducted Australian continental margin units on Timor (Harris 1991). The nappes probably represent Sunda forearc basement that was detached from the forearc and uplifted during Late Miocene to Pliocene subcretion of Australian continental margin units to the edge of the Asian plate. The Banda Terrane consists mostly of a mix of pelitic and mafic metamorphic rocks overlain by remnants of a Cretaceous-Early Tertiary volcanic arc, and Oligocene-Miocene massive limestone. Radiometric age analysis of the metamorphic rocks yield a Permian Rb/Sr whole-rock age and a consistent 32-35 Ma 40Ar/39Ar cooling age from hornblende and biotite (Harris et al. 1998). These data indicate that the continental fragments may be parts of the easternmost Sunda craton that were separated from the craton during earlier phases of back-arc extension (Brown & Earle 1983), such as the extensional event that formed the Bali and Flores Basins along the eastern edge of the Sunda Shelf (Prasetyo 1989, 1994).

Ocussi ophiolite of Timor The northernmost part of the Banda Terrane nappe in Timor consists of a thick succession of basaltic andesite pillow lavas and sheet flows known as the Ocussi Volcanics. These steeply dipping volcanic units have 40Ar/39Ar ages (3-5 Ma) and SSZ geochemical characteristics similar to dredge samples from other parts of the southern Banda Sea Basin (Harris 1992). Ultramafic blocks in melange at the base of the Ocussi nappe are geochemically depleted with an estimated 20-25% partial melting. These blocks are the only ultramafic rocks throughout the Banda arc with any SSZ characteristics (Harris & Long 2000). Based on these relations, the Ocussi Volcanics are interpreted as the emergent tip of part of the SSZ ocean basin that formed within the eastern Sunda arc, like the southern Banda Basin. The southern edge of this intra-arc basin has been emplaced onto the edge of the Australian continental margin by arccontinent collision (Harris & Long 2000). In other words, the same geochemical and age relations found from dredge samples from the South Banda Sea Basin are also found in nappes of the Timor collision zone, where the Australian continent has underthrust, uplifted and exposed the edge of the Asian upper plate. Several lines of evidence indicate that the most recent phase of extension of the Banda Basins was synchronous with the arrival of the Australian continental margin at the Java Trench and the onset of arc-continent collision. For example, the first part of the Australian continental margin to accrete to the edge of the southern Banda Basin,

OPHIOLITES OF SE ASIA AND NEW GUINEA

489

the Aileu-Maubisse Complex, yields 40Ar/39Ar metamorphie cooling ages from hornblende of 58 Ma and from biotite of 3-5 Ma (Berry & McDougall 1986). These ages indicate that the distal part of the Australian continental margin had already accreted to the edge of the Asian plate and cooled from 600 °C to 350 °C around the same time as the Banda Basins were opening (38 Ma) and forming new oceanic lithosphere. The synorogenic sedimentary record of Timor also dates the initial emergence and erosion of the Banda arc-continent collision at around this time (Audley-Charles 1986; Fortuin & de Smet 1991).

were interpreted by Linthout & Helmers (1994) as the mantle component of a young ophiolite formed by transtension. One of the major problems with interpreting these and other Iherzolite bodies as fragments of ophiolites is the notable lack of the crustal component of the ophiolite sequence. The only volcanic units throughout the Banda arc that are geochemically compatible with the Iherzolite are Permian-Jurassic age intraplatetype pillow lavas of Australian affinity (Berry & Jenner 1982), which supports a continental margin origin for the Iherzolite.

Lherzolite bodies of the Banda arc

Geodynamics of the Banda arc region and ophiolite emplacement

Lherzolite bodies and some gabbro are commonly found along the suture of the Australian and Asian plates throughout the Banda arc-continent collision. In Timor these bodies were initially interpreted as parts of young mantle sequence of the Banda Sea Basin (Hamilton 1979; Brown & Earle 1983; Helmers et al 1989; Sopaheluwakan et al 1989; Linthout & Helmers 1994). However, detailed mapping and geochemical analyses indicate continental margin and pre-oceanic rift associations (Berry 1981; Harris 1992; Harris & Long 2000). The most chemically similar rocks to the Timor Iherzolite bodies are Red Sea-type peridotites (Bonatti et al. 1986) and Iherzolite dredged from the Australian (Nicholls et al. 1981) and Iberian (Boillot et al. 1980) distal passive continental margins. Based on these discoveries and comparisons, the Iherzolite bodies of Timor were reinterpreted as fragments of subcontinental mantle exposed by extensional denudation before being incorporated into a collision zone (Harris 1992; Harris & Long 2000). The distal part of the Australian continental margin, where similar Iherzolite bodies have been dredged, was the first to enter the subduction zone and be accreted as thrust units. As the arc-continent collision progressed, these thrust sheets were elevated by understacking of continental margin units more proximal to the shelf (Harris et al 2000). The Iherzolite bodies of the external Ligurides, part of the original 'Steinmann trinity' (Steinmann 1906), have been reinterpreted in a similar manner (Rampone & Piccardo 2000). These studies imply that many of the Iherzolite-type peridotite bodies found throughout the Tethys and other suture zones may be thrust-stacked fragments of the continental lower plate exhumed by extensional events associated with passive margin development. Lherzolite bodies of the northern Banda arc island of Seram have similar tectonic associations and geochemistry affinities to those of Timor, but

Initiation of collision in the Banda arc is marked by progressive uplift and emergence of the forearc ridge and eventually the accretion of the eastern Sunda arc to the Australian plate via partitioning of strain away from the trench and incipient subduction polarity reversal (Fig. 4). GPS measurements (Genrich et al. 1996) show that at decadal time scales the accreted materials of the collision behave as a number of independently moving blocks with poorly defined boundaries. A similar pattern is also revealed at time scales of 103-105 years by studies of uplifted coral terraces throughout the region (Merritts et al. 1998). These studies help quantify for the first time the processes of plate boundary reorganization that lead to the formation and obduction of ophiolites. The progressive accretion of the Banda arctrench system to the edge of Australia is uniquely displayed by the obliquity of arc-continent collision (Fig. 3). The arrival of the Australian rise and slope at the Java trench initiates a phase of rapid accretion and uplift of the forearc upper plate. The thin leading edge of the forearc basement becomes a roof thrust that is folded and detached from its roots by backthrusting and understacking of continental margin units (Harris 1991). In Timor these nappes mostly consist of a heterogeneous mixture of continental and oceanic material (see discussion of Banda Terrane above). However, along erogenic strike to the east the Weber forearc basin, which is at an earlier stage of tectonic emplacement onto the Australian continental margin, is mostly oceanic and represents a future ophiolitic terrane as large as any known. The arrival of the Australian continental shelf at the trench causes indentation of the deformation front and partitioning of strain to hinterland back thrusts at the rear of the orogenic wedge (Prasetyadi & Harris 1996). This arcward-verging part of the orogenic wedge is progressively thrust over much of the forearc basin until the entire

490

R. HARRIS

Fig. 4. Crastal cross-sections of different collisional settings for ophiolite emplacement in the eastern Indonesian region. Black - Oceanic crust and ophiolites, horizontal stripes - mantle lithosphere, stipples - island arc terranes, vertical lines - melange and sedimentary units, grey - continental crust. See Figure 3 for section locations. Horizontal scale = vertical scale.

OPHIOLITES OF SE ASIA AND NEW GUINEA volcanic arc itself is accreted to the lower plate. GPS measurements (Genrieh et al. 1996), seismicity (Abbott & Chamaluan 1981; McCaffrey & Nabelek 1984), and coral terrace studies (Merritts et al. 1998) indicate that at this stage of collision, much of the plate convergence is taken up along a back-arc thrust system (Wetar Thrust). The Wetar Thrust may develop into a new subduction zone with opposing dip. The eastern Sunda arc is inactive in this region and has accreted to the edge of Australia. A similar scenario of arc-continent collision occurs to the east in New Guinea, where the Australian continent collides with Pacific microplates converging from the east.

Convergence with the Philippine Sea and Pacific plates Ophiolites throughout the eastern Indonesian and New Guinea region are mostly interpreted as fragments of the southwestern Philippine Sea and Pacific plates that were emplaced onto continental and arc terranes of Asian and Australian affinity. Most of the ophiolites are Mesozoic in age and formed the basement of a Late Cretaceous to Paleogene volcanic arc. Neogene collision of the irregular-shaped Asian and Australian continental margins with these oceanic plates initiated a phase of ophiolite emplacement and plate boundary reorganization that stretches from Taiwan and the Philippines to the Moluccas and New Guinea. The Molucca arc-arc collision and the New Guinea arc-continent collision are two examples discussed here.

New Guinea ophiolites The New Guinea region represents the northernmost promontory of the Australian continent that initiated a series of arc-continent collisions with the southern boundary of the Pacific plate (Dewey & Bird 1970; Jaques & Robinson 1977; Hamilton 1979). This plate boundary was modified through the course of changes in plate motion into a series of microplates, including the present Caroline, Bismarck and Solomon plates (Figs. 2 and 5). The insertion of the Australian continental margin into this fragmentary plate boundary zone has resulted in the formation and emplacement of some of the largest ophiolites known. The ophiolite-bearing collisional mountains of New Guinea are an older continuation of the active Banda arc-continent collision to the west (Harris 1991). The collision-related phase of plate boundary reorganization and partial subduction polarity reversal currently initiating in the Banda arc region occurred in the NW New Guinea region

491

during the Latest Miocene (Pigram & Davies 1987). The propagation of collision along the northern margin of Australia provides a unique way to simultaneously observe various stages of ophiolite emplacement and mountain system development along orogenic strike. Collision in the New Guinea region began in the Early Oligocene, when the northern passive margin of Australia was pulled into a northdipping subduction zone at the edge of the Pacific-Philippine Sea-Caroline plate (PPSCP). Paleogene intra-oceanic arc terranes mounted on the edge of the PPSCP have since accreted to the margin of Australia via a combination of partial subduction polarity reversal and transvergence (Pigram & Davies 1987). The edge of the PPSCP had experienced a phase of Jurassic SSZ spreading prior to the construction of the Paleogene arc. Arc-continent collision ceased in western New Guinea around 3-5 Ma, but is still active in eastern New Guinea, where collision of the New Britain arc with the Australian continental margin gives rise to the Huon Peninsula (Silver et al. 1991). The Huon Peninsula is the onshore extension of the New Britain arc accretionary wedge where continental margin material is stacked beneath and uplifts pieces of the south New Britain plate. This is happening at the same time as a new SSZ ocean basin, the Bismarck Basin, is forming in another part of this plate. The present tectonic scenario of the New Guinea region is much like the one that existed earlier when the south-facing Paleogene arc terrane, mounted on the edge of the PPSCP, accreted to New Guinea. This arc terrane is bounded to the north and south by ophiolites. To the south, the Central and Papuan ophiolite belts are sandwiched between the arc terrane and fold and thrust continental margin units of Australia. These ophiolites are mostly Jurassic in age and restore to a forearc position of the Paleogene arc terrane. North of the arc terrane is the Cyclops ophiolite belt, which is Tertiary in age and probably represents pieces of a younger interarc basin lithosphere that formed in a similar manner to the present Bismarck Basin and was emplaced around 20 Ma (Monnier et al. 1999).

Central (Irian) ophiolite belt The Central ophiolite belt is exposed for 500 km along the north slope of the Central Range of western New Guinea (Dow et al. 1986). It has a minimum age of Late Cretaceous and occupies a suture zone between shortened Australian continental margin units to the south that have stackedup against accreted arc terranes of the PPSCP to the north (Dow & Sukamto 1984). The ophiolite

492

R. HARRIS

Fig. 5. Tectonic map of ophiolites and young marginal basins east of New Guinea. SSZ oceanic lithosphere of the Bismarck Sea plate is simultaneously opening and accreting to the edge of Australia. The Woodlark Basin has split a Tertiary arc and is currently propagating westward into continental crust. Previous episodes of subduction continue to influence the geochemistry of these new ocean basins. Details of the Papuan ophiolite are taken from Davies and Jacques (1984). The internal structure of the Woodlark Basin is modified from Johnson et al. (1987).

is structurally underlain by metabasites (amphibolite and blueschist facies) that yield Late Cretaceous and Eocene metamorphie cooling ages (Warren 1995; Weiland & Cloos 1996). These are structurally underlain by the inverted Ruffaer metamorphic complex, which formed along parts of the Australian continental margin that were stacked beneath the ophiolite as it collided with the distal Australian passive margin around 18 Ma (Warren 1995; Weiland & Cloos 1996). Reconnaissance studies of the composition and internal structure of the Central ophiolite belt indicate mostly a complete ophiolite sequence from residual mantle peridotite to basalt (Monnier et al. 2000). The mantle sequence is mostly harzburgite and dunite with rare Iherzolite, all with high-temperature porphyroclastic textures. Chrome spinel numbers (50-60) and REE pat-

terns of the residual peridotite indicate a high degree of partial melting (20-25%). Layered gabbros have adcumulate textures with abundant primary amphibole and labradorite typical of crystallization from aluminous, calc-alkaline magmas. Most volcanic units feature trace and REE abundances intermediate between N-MORB and calc-alkaline series, with Nb and Ta depletion indicative of subduction zone environments (Monnier et al 2000).

Papuan ophiolite belt Three major ophiolite complexes make up the Papuan ophiolite belt of eastern New Guinea. From west to east these are the April ultramafics, Marum ophiolite, and Papuan ophiolite (Fig. 1). The Papuan ophiolite is the largest and most

OPHIOLITES OF SE ASIA AND NEW GUINEA complete of the group (Davies & Smith 1971). It is exposed as thick thrust sheets rooted in the Solomon Sea oceanic basin and crops out for 400 km along the northern slope of the Owen Stanley Range (Fig. 5). The Marum ophiolite is less complete and consists of two very different thrust sheets (Jacques 1981). The larger upper sheet preserves a sequence of cumulus ultramafic and mafic plutonic units up to 3-4 km thick that are underlain by harzburgite. The thinner lower thrust sheet consists mostly of highly faulted basaltic units that are geochemically unrelated to the structurally overlying sheet. The April ultramafic complex further to the west comprises dismembered thrust sheets that have a maximum thickness of 2-3 km of predominantly highly strained Iherzolite with minor layered gabbro (Davies & Jaques 1984). The three complexes each occupy a similar structural position in the New Guinea mountain system but represent different tectonomagmatic settings. Differences in the tectonomagmatic origin of the three complexes are manifest in contrasting mantle sequence compositions. The Papuan Ophiolite preserves a 4-8 km thick section of residual peridotite (Davies & Jaques 1984) that consists almost entirely of highly refractory harzburgite. The Marum body has a less refractory mantle sequence (Jacques 1981), and the April ultramafics show evidence of only small amounts of partial melting (Davies & Jaques 1984). Differences in composition between the Papuan and Marum ophiolite are minor, and may be attributed to forearc heterogeneity. However, these bodies show little or no affinity to the April complex or the lower thrust sheet of the Marum ophiolite. Similar contrasts are also found in the Banda forearc nappe emplaeed over the Australian continental margin in Timor (Harris 1992). The Iherzolitic April complex and the thrust sheet of transitional MORB beneath the Marum ophiolite also have equivalents in the Banda arccontinent collision. Harris & Long (2000) showed that the lowest Iherzolite and basaltic (transitionalMORB) thrust sheets in the Timor region are slivers of the distal continental margin of Australia that were accreted to the base of the Banda forearc nappe. Davies & Jaques (1984) interpreted the lower thrust sheet of the Marum ophiolite, and the Emo metabasite beneath the Papuan ophiolite, in a similar way. Tectonic slivers of spilitic basalt are also found at the base of the Central ophiolite belt (Warren 1995). In each of these settings the subduction of the leading edge of the Australian continental margin has stacked forearc basement of a varying refractory nature over non-refractory transitional lithosphere of the lower plate. Examples of this stacking order abound in Tethyan-type

493

ophiolites throughout the world, such as the Haybi volcanics beneath the Oman ophiolite (Lippard et al 1984).

Tectonic emplacement of New Guinea ophiolites The timing and conditions of ophiolite emplacement in New Guinea are recorded in metamorphic rocks beneath the Central and Papuan ophiolite belts and in the timing of foreland basin development astride the fold belt. Metamorphic rocks of the Emo sheet grade upward from lower greenschist-facies units into amphibolites and granulites toward the base of the Papuan ophiolite thrust sheet (Lus et al 2001). Hornblende 40Ar/ 39 Ar ages of the Emo metamorphic rocks yield cooling ages of 60-58 Ma, which are near the same age as boninite of the Papuan ophiolite (Lus et al. 2001). Ages of metabasites beneath the Central ophiolite cluster at 68 Ma and 44 Ma (Weiland 1999). These ages are interpreted as a two-stage initiation of northward subduction of the Australian plate beneath the PPSCP plate (Weiland 1999). Jurassic parts of the ophiolite must represent an earlier incipient subduction event. Ages of metamorphic cooling (Weiland 1999) and incipient foreland basin development (Pigram & Symonds 1991) reveal that the distal New Guinea passive margin began to enter the subduction zone at the southern edge of the PPSCP around 25 Ma. However, the transition from carbonate shelf to siliciclastic synorogenic sedimentation did not reach western New Guinea until around 12 Ma (Quarles van Ufford 1996). Arc-continent collision ceased in western New Guinea at around 3 Ma as a result of continuing plate boundary reorganization involving partial subduction polarity reversal (Hamilton 1979; Cooper & Taylor 1987), delamination (Housh et al. 1994), and the development of left-lateral strike-slip fault systems (Sapiie et al. 1999).

Cyclops ophiolite Ophiolites found along the north coast of New Guinea, north of the Paleogene arc terrane, are referred to here as the Cyclops ophiolite belt. These bodies include an assortment of ophiolitic fragments interpreted as remnants of the Central ophiolite to the south (Dow et al. 1988). A complete ophiolite sequence is preserved near Jayapura, which includes a 10 km thick ultramafic unit and a 5 km thick crustal unit. The ultramafic unit consists of moderately refractory harzburgite with patches of dunite. High spinel Cr number

494

R. HARRIS

(58-63 for harzburgite and 72-90 for dunite) and low bulk and mineral abundances of A^Os (0.2— 0.4%) indicate 25-35% partial melting under hydrous conditions (Monnier et al. 1999). Zr and Hf enrichment may also indicate the crystallization of hydrated phases, which is common in SSZ ophiolites. Crustal rocks include layered and isotropic gabbro, massive and sheeted diabase, massive and pillow lavas, and a sedimentary cover of chert and marl overlain by boninites (Monnier et al. 1999). REE patterns of most basaltic rocks show high La/Nb ratios of 2-3 (Monnier et al 1999), which are typical of BABB. The occurrence of EnrichedMORB-like basalt may indicate a trapped fragment of Australian or Pacific oceanic crust in the arc complex. Variations in composition and age may indicate that the Cyclops ophiolite is a composite of several parts of an arc complex, which may also include slivers of the lower plate. Whole-rock K/ Ar analyses yield a bimodal age distribution of 43 Ma for boninites to 29 Ma for BABB (Monnier et al. 1999). The metamorphic sole (MacAuthor complex) beneath the Cyclops ophiolite consists of an arc terrane with an inferred metamorphic age of 20 Ma (Pieters et al. 1979).

Microplates of the New Guinea region The combination of changes in plate motion, increased coupling as a result of collision (Kroenke 1996) and a high obliquity of convergence (Sapiie et al. 1999) have caused fragmentation of, and renewed spreading within the western edge of the Pacific plate to form the Caroline, Solomon and Bismarck plates. The array of island arcs, and small, young ocean basins in this region has evolved through successive periods of convergence along different subduction zones into a megashear zone between the large Pacific and Australian plates. Left-lateral oblique convergence results in rapid changes in the function and position of plate boundaries. Repeated collisions of these boundaries with buoyant oceanic crust (e.g. Ontong-Java plateau) brought in from the east by the Pacific plate and continental fragments pushing in from the NE margin of Australia have resulted in the emplacement of ophiolites from New Guinea to New Zealand, subduction reactivation and polarity reversals, and terrane accretion (Silver et al. 1991; Kroenke 1996). The arc terranes and continental fragments embedded in the easternmost abyssal part of the Indo-Australian plate will make it difficult to subduct without great difficulty. Differential rigid-body rotation of microplates in the New Guinea megashear simultaneously

open and close young ocean basins such as the Ayu Trough (Fig. 2), and Bismarck and Woodlark Basins (Fig. 5). The Bismarck Basin opens behind the New Britain arc as it actively collides with and accretes to the northern margin of New Guinea. The Ayu Trough and Woodlark Basin open along microplate boundaries that are being subducted. Investigations of the geochemistry of Woodlark Basin volcanic rocks (Binns & Whitford 1987; Johnson et al. 1987; Perfit et al. 1987) reveal that portions of the basin are underlain by normal, depleted oceanic mantle, yet the majority of the dredge samples, including some from the ridge itself, are arc-like rocks that vary from BABB to boninites. Near where the ridge is being subducted, spatial variations show a pattern of increasing arc-type character toward the San Cristobal Trench (Fig. 5), which may indicate active mixing of N-MORB with arc components. Arc-like rocks were also dredged from seamounts within 30 km of the trench, on both the upper and lower plates. Other geochemical anomalies require recent conditioning of the mantle by previous subduction (Perfit et al. 1987). In summary, the Woodlark Basin, like much of the New Guinea region as a whole, demonstrates that igneous geochemical zoning is very difficult to explain by the present configurations of plate boundaries.

Molucca Sea region The Asian, Australian and Philippine Sea plates converge in the Molucca Sea region, where they are connected by an array of ephemeral subduction zones linked by strike-slip faults and localized rifts (Fig. 3). The major connecting boundary is the left-lateral Sorong Fault system. This fault slices the northern margin of Australia into a series of continental slivers that move west with the Philippine Sea plate until they collide with and are accreted to the Asian margin in the Sulawesi region. One of the most dramatic examples of this process is the collisional underthrusting of the Banggai-Sula Platform beneath the Eastern Sulawesi ophiolite (Figs 3 and 4). A similar scenario to the Sulawesi collision, but at an earlier stage of development, exists to the north, where amalgamated arc, ophiolite and continental terranes of the Halmahera region are moving westward with the Philippine Sea plate towards the Sangihe arc (Fig. 3). The intervening Molucca Sea slab has at least partly fragmented as the colliding arcs impinged upon it (Fig. 4). Incipient collision of the opposing arcs is manifest by recent rapid emergence of the Talaud Ridge as it is thrust eastward over the Halmahera arc and thrust westward over the Sangihe arc (Silver &

OPHIOLITES OF SE ASIA AND NEW GUINEA Moore 1978) and accretionary wedge (Moore et al. 1981). Seismic refraction profiles indicate that this doubly vergent collision complex that forms the Molucca Ridge is at least 60 km thick (Bader et al. 1999). The positive buoyancy of this feature probably influences the development of new subduction zones with reversed motion, such as the Cotobato Trench to the west and the southwardpropagating Philippine Trench to the east (Fig. 2). Closure of the Molucca Sea is creating a suture zone that juxtaposes ophiolites from two sources while technically obscuring the associated volcanic arcs. Ophiolites of eastern Halmahera and the nearby islands of Gebe, Waigeo, Obi, and Tapas probably have a Late Jurassic Pacific plate origin like the ophiolites of the eastern Philippines to the north (Hall et al. 1991) and New Guinea to the south. On the other hand, ophiolites emplaced onto the Asian margin, from Sulawesi through the Molucca Sea and into the western Philippines, are Eocene and Oligocene in age and have affinities with marginal basins that opened within the Asian continental margin (Silver & Rangin 1989). Most of the ophiolites from both tectonic settings record a polyphase history of incorporation into the forearc of a newly forming subduction zone, remelting of peridotite already depleted by extraction of MORB, collision with continental or arc terranes, and reshuffling by strike-slip faulting.

Halmahera ophiolite Halmahera forms the eastern part of the Molucca Sea arc-arc collision complex. The western part of the k-shaped island consists of an active intraoceanic island arc, whereas the eastern arms consist mostly of dismembered oceanic lithosphere (Hamilton 1979). Each unit of a complete ophiolite sequence is present (Hall et al. 1988). Although the sequence is structurally disrupted, its geochemistry is arguably cogenetic, with the exception of some units interpreted as seamounts and later magmatic pulses (Ballantyne & Hall 1990). The highly refractory nature of residual harzburgite, cumulate ultramafic and mafic sequences, and many of the volcanic units (including boninites) found throughout Halmahera require that the ophiolite had a shallow, subduction-related tectonomagmatic origin (Ballantyne & Hall 1990). Similar compositions and geological associations to those found in Halmahera are reported from forearc dredge samples of the nearby MarianaBonin arc system (Bloomer 1983). These similarities define Halmahera as an uplifted segment of the western edge of the Philippine Sea plate that formed during the Mesozoic in much the same way as the Mariana-Bonin arc system formed

495

along the eastern edge of the Philippine Sea plate during the Tertiary. Talaud ophiolite The Talaud Islands are the emergent expression of the Talaud Ridge that rises in the middle of the Molucca Sea arc-arc collision zone (Fig. 4). These islands expose an east-dipping ophiolite sequence (Silver & Moore 1978) interpreted as splintered vestiges of the Molucca Sea slab (Fig. 4) that were pushed beneath the colliding Sangihe and Halmahera arcs (Sukamto 1979; Cardwell et al. 1980; McCaffrey et al. 1985; Moore et al. 1980; Bader et al. 1999). Ultramafic and mafic rocks exposed on islands of the Talaud Ridge represent a disrupted cogenetic ophiolite suite (Moore et al. 1980) that is geochemically indistinguishable from MORB or BABB (Evans et al. 1983). Whole-rock and mineral analyses of Iherzolite and less abundant harzburgite indicate slightly smaller amounts of partial melting than average values for the Oman mantle sequence (Lippard et al. 1984). Cumulates and basalt also plot in MORB fields of the discriminate diagrams of Pearce et al. (1984) and Beard (1986). Chert and red limestone yield radiolaria of midEocene age, which is considered as the minimum age of the ophiolite. This is also the age of the nearby Celebes Sea marginal basin. Miocene arcrelated volcanic rocks and volcaniclastic sedimentary rocks are also found, but are of uncertain origin. Most interpretations of the origin of the Talaud ophiolite bodies agree that they are slivers of the Molucca slab, which is probably part of the Celebes Sea marginal basin. However, interpretations differ on the size of the slivers and their geometry at depth. Based on data from structural field mapping, Moore et al. (1980, 1981) showed the Talaud ophiolite as a rootless, east-dipping thrust sheet embedded in melange that was accreted to the eastern edge of the Snellius arc. Geophysical studies indicate that the Talaud ophiolite is a west-dipping sliver that connects at depth with the Molucca slab and Sangihe arc, respectively (McCaffrey 1991; Bader et al 1999). East Sulawesi ophiolite The East Sulawesi ophiolite is one of the largest and most complete, yet poorly known ophiolite bodies of the Indonesian and New Guinea region. It is exposed over more than 10000 km2 throughout the eastern and SE arms of the k-shaped island of Sulawesi (Fig. 3). The eastern arm exposes a complete ophiolitic sequence underlain by a metamorphic sole, melange, imbricate continental mar-

496

R. HARRIS

gin material and crystalline basement with a blueschist metamorphic overprint (Parkinson 1998). This sequence mostly faces east (Fig. 4), with sheeted dyke complexes consistently trending NNW-SSE (Silver et al 1983). The structural thickness of the ophiolite varies, but may exceed 15 km in places (Silver et al. 1983). The continental rocks upon which the Sulawesi ophiolite was emplaced were sliced from the northern Australian margin, translated westward and accreted to the eastern edge of the Asian plate (Silver et al. 1978; Hamilton 1979). The most recent manifestation of this process is the accretion of the Sula platform (Fig. 3), which has structurally dismembered the ophiolite into a series of eastward-verging imbricate thrust sheets. The age, internal structure and tectonomagmatic origin of the Sulawesi ophiolite remain poorly constrained by reconnaissance-style studies of limited parts of the body (Kundig 1956; Silver et al. 1978, 1983; Simandjuntak 1992; Mubroto et al. 1994; Monnier et al. 1995; Parkinson 1998). Seventeen K-Ar ages range from 93 ± 2 to 32 ± 2 Ma, with most ages clustering between 60 and 40 Ma. Sparse geochemical studies of the ophiolite show a broad range of compositions (Monnier et al. 1995). For example, only three whole-rock analyses from two locations that are more than 100km apart are reported for the entire mantle sequence. The two locations vary in A^Os and CaO abundances by an order of magnitude (0.22.0wt%). Mineral analyses from these samples have Cr numbers for spinel that vary from 18 to 84. These results may indicate that the mantle sequence consists of both fertile and depleted sections or that the ophiolite is a composite of different bodies. A composite origin is also consistent with the size, structural imbrication and broad apparent age range of the ophiolite. Volcanic rocks mostly from the NE arm of the Sulawesi ophiolite yield MORB-type REE and trace element patterns with the exception of a slight negative Nb anomaly (Monnier et al. 1995). Similar compositions are reported from the nearby Celebes Sea marginal basin immediately north of Sulawesi (Serri et al. 1991), which is Eocene in age (Weissel 1980). Several models exist for the origin of the Sulawesi ophiolite. One of the earliest considered it part of the Banda Sea that was thrust onto the edge of the Asian margin by collision with the westward-verging Sula block (Hamilton 1979). Reconnaissance geological mapping and marine geophysical investigations led Silver et al. (1983) to propose that the ophiolite is part of the Celebes Sea and Gorontalo Basin to the north. This model was also used by Monnier et al. (1995) to explain

geochemical similarities between the Sulawesi ophiolite and oceanic terranes to the north. Another model, based on poorly constrained paleomagnetic data, claims that the ophiolite originated as part of the Indian Ocean, 17° south of the equator (Mubroto et al. 1994). From studies of the metamorphic sole, Parkinson (1998) concluded that the ophiolite originated in an Andaman Seatype SSZ ocean basin that opened within the Philippine Sea plate during oblique subduction at the eastern edge of the Sunda arc. The age of emplacement is loosely constrained by K-Ar analysis of hornblende from the metamorphic sole, which yields cooling ages of 3326 ± 3 Ma (Parkinson 1998). Garnet peridotite fragments within strike-slip faults beneath the ophiolite yield Sm-Nd mineral ages of 27 Ma for peak metamorphism and 20 Ma cooling ages (Kadarusman 2001). These ages overlap the time when Australia arrived at the subduction zone along the southern edge of the Philippine Sea plate, which initiated the New Guinea collisional orogen and counter-clockwise rotation of the Philippine Sea plate (Hall 1996).

Origin of ophiolites in the Indonesian and New Guinea region Most ophiolites of the Indonesian and New Guinea region are a composite of new SSZ ocean basins that formed within pre-existing continental or oceanic material. Several examples of composite forearc slabs are provided here. Each has a component of refractory material resulting from high degrees of partial melting during incipient subduction and subsequent trench rollback (e.g. Robinson et al. 1983; Pearce et al. 1984; Bloomer et al. 1995). However, plate kinematic explanations alone do not adequately explain the temporal and spatial clustering (Fig. 6) of ophiolite and marginal basin development in the Indonesian and New Guinea region (e.g. Taylor & Karner 1983; Tamaki & Honza 1991). A strong correlation does exist in time and space between ophiolite genesis and collision (Edelman 1988) as suggested by the concept of pericollisional extension (Harris 1992; Royden 1993). The pericollisional extension model predicts that increased coupling along collisional promontories can pin a plate boundary at that point whereas adjacent parts of the lower plate are free to roll back and open small ocean basins adjacent to the collisional indenter (Vogt 1973; Miyashiro et al. 1982). Perhaps one of the best examples of this process is the deformation of the Izu-Bonin-Mariana trench by the Caroline Ridge (Fig. 1). Other examples in the Indonesian and

OPHIOLITES OF SE ASIA AND NEW GUINEA

497

Fig. 6. Spatial and temporal correlation between marginal basin and ophiolite genesis, and collisional events. Many of the New Guinea ophiolites were emplaced during the gap in time (gray)

New Guinea region may be the Banda Sea and associated back-arc basins that formed at the edge of the Sunda Shelf (Fig. 2). There is strong evidence that two other factors associated with collision, increased obliquity of convergence (transvergence) and mantle extrusion (Flower et al. 2001), also play a significant role in the marginal basin and ophiolite development in the ING region.

Transtension Strain partitioning associated with collisional indentation and oblique convergence strongly influences the spatial distribution of ophiolites throughout the Indonesian and New Guinea region (Figs 1 and 2). Clusters of ophiolites occur at highly oblique convergent margins where strain is partitioned away from the trench into the arc and back-arc regions, such as the Andaman and Bismarck Sea regions. These highly oblique configurations can arise from collisional modification of an existing arc-trench system, such as indentation of the Sunda arc by collision of India; or near triple junctions, such as the Sorong Fault system, which buffers changing motions between the Australian, Asian and Pacific plates. Ophiolites are also commonly found along the margins of the actively spreading ocean basins, which may indi-

cate an earlier phase of transtension that was followed by basin inversion.

Mantle dynamics and marginal basin development The influence of mantle flow on plate interactions may account for the episodic nature of temporal patterns of ophiolites in Indonesian and New Guinea. A global westward delay of lithosphere relative to the underlying mantle in the hotspot reference frame (Minster et al. 1974) is attributed to a westerly mantle flow associated with the Earth's rotation (Volpe et al. 1990; Doglioni 1993; Smith & Lewis 1999). The influence of this mantle flow on subduction zones may provide an explanation for the consistent asymmetry of orogens associated with steepening and rollback of west-directed slabs (Uyeda & Kanamori 1979; Doglioni et al. 1999). Extrusion of asthenosphere during various collisional episodes may also explain the spatial and temporal association of collisional events with rapid marginal basin development in the Indonesian and New Guinea region (Flower et al. 2001). According to this model, mantle is driven eastward by collisional constriction, such as the eastward extrusion of terranes toward the western Pacific by the collision of India. Although direct measurements of asthe-

498

R. HARRIS

nospheric flow remain elusive, considerable circumstantial evidence supports the notion, such as the asymmetry of subduction zones and associated orogens (Doglioni et al. 1999), and DUPAL-like mantle contamination patterns (Flower et al. 2001). The asymmetry of orogens in the Indonesian and New Guinea region is consistent with the global pattern of subduction zone asymmetry (Doglioni 1993). East-directed subduction zones generally have low slab dip angles; wide, doubly vergent thrust belts; and uplifted and highly deformed upper plates as a result of increased buoyancy and coupling induced by movement in the same direction as, and support of, asthenospheric mantle flow. In contrast, west-directed subduction zones generally have very steep dips, deep trenches, small and deep forearcs, and upper plate extension as a result of trench retreat with eastward-flowing asthenosphere. Plate kinematic explanations for this asymmetry, such as differing ages and velocities of subducting plates, do not account for observed differences in subduction zone behaviour in the Indonesian and New Guinea region or elsewhere. One example is the asymmetry of the doubly subducted Molucca plate. Although there is no significant difference in plate age or velocity, the slab dips steeply to the west and at a lower angle eastward (Fig. 4). The small Sangihe arc-trench system forms above the steep, west-directed limb. The lower-angle, east-directed limb of the plate supports the Halmahera arc, which exposes the oceanic basement upon which the arc is mounted. Not only is the forearc basement exposed, but it is also deformed into a thrust belt that exposes the Halmahera ophiolite (Hamilton 1979). Yet this ophiolite, and the Mesozoic ophiolites to the north in the Philippines and to the south in New Guinea, are probably a product of SSZ spreading above a west-directed subduction zone, such as the present Mariana system (Ballantyne & Hall 1990). The Banda arc may offer another example of mantle-flow induced subduction zone asymmetry. In this region the slab curves nearly 180°. Where it is west-directed, or influenced most by eastward asthenospheric flow, the slab dip angle is greatest, the forearc (Weber Basin) is more than 7 km deep and back-arc extension has formed the Banda Sea. East-directed subduction systems in the Indonesian and New Guinea region include most of the Sunda, Halmahera and New Britain arcs. Although most of the features of these subduction systems are similar to those predicted, particularly the low angle of slab dip and high coupling, localized upper plate extension is observed in the Andaman Sea and Bismarck Basin regions (Fig. 1). How-

ever, both of these SSZ spreading systems occur where the convergence angle is highly oblique, resulting in transtensional strain partitioning into the arc and back-arc. There also may be an influence in these regions of local mantle extrusion from the Himalayan and New Guinea collisions, respectively. These exceptions illustrate that the complexity of subduction systems cannot be adequately explained by any one simple mechanism. Three major pulses of mantle extrusion have been proposed by Flower et al. (2001) for the SE Asian region, each associated with major collision events. The first is connected with constriction of mantle by the Eocene collision of India, which may have contributed to eastward extension of the West Philippine Sea and Celebes Basins. This phase of extension is coeval with the genesis of ophiolites throughout the Molucca region, such as the East Sulawesi ophiolite. However, this was also a time when plate convergence was highly oblique in the Sulawesi region. The second phase of mantle extrusion is coincident in time with collision between the Australian continental margin and subduction zones along the southern edge of the Pacific and Philippine Sea plates. This phase of extrusion may have influenced the opening of the Japan, South China, Sulu and Makassar Seas, and the Parece Vela Basin of the IzuBonin-Mariana arc system. A third phase of extrusion associated with the opening of Okinawa and Mariana troughs, Andaman and Banda Seas, and microplate basins in the New Guinea region may be driven by local collisions in Taiwan, Sulawesi, the Banda arc and western New Guinea. Each of these phases of eastward asthenospheric extrusion is associated with episodic fusion of the Asian and Indo-Australian plates.

Changes in plate kinematics Strong temporal correlations also exist between events of major plate motion change and the genesis of marginal basins and ophiolites. Most ophiolites of the Indonesian and New Guinea region show evidence of residence in a forearc position at the time of emplacement. However, in many cases these bodies originated as marginal and back-arc basins with no close spatial relation to arcs. An active example of this situation is the current collision of the New Britain forearc, which was formerly part of the West Melanesian backarc that has been accreted onto the eastern edge of New Guinea to form the Finisterre terrane of the Huon Peninsula (Silver et al. 1991). Frequent changes in motion of one or more of the major plates converging in the Indonesian and New Guinea region are buffered by frequent plate

OPHIOLITES OF SE ASIA AND NEW GUINEA boundary reorganizations that may include phases of ophiolite genesis and emplacement. An example of this is the changes over the past 40 Ma in the complex plate boundary between the IndoAustralian, Pacific and Asian plates in the Sulawesi region (Hall 2002). Major plate boundary reorganization events in Indonesia correspond to documented changes of global plate motion (Fig. 6). For example, bends in the Hawaiian-Emperor seamount chain document changes in the motion of the Pacific plate through 55° at 43.1 ± 1.4 Ma (Duncan & Clague 1985), 10° at 28-23 Ma (Epp 1984), and 10-20° at 4 Ma (Cox & Engebretson 1985). Motion of the Indo-Australian plate also changed at around 50 Ma from NW to NNE, which is a shift of around 73° relative to the Pacific plate (Weiland 1999). Up to 35° of clockwise rotation of the Philippine Sea plate was also initiated at around 25 Ma (Hall 1996). Each of these plate motion changes coincides with major collisional events throughout SE Asia, which include the collision of India with the western Sunda arc at around 50 Ma, the arrival of the northern Australian continental margin at the southern plate boundary of the Pacific and Philippine Sea plates at around 30 Ma, and the collisions of Taiwan, the Moluccas and the Banda arc at around 5 Ma (Fig. 6). Temporal clustering of Mesozoic ophiolites of the Indonesian and New Guinea region with those in the Eastern Mediterranean and Cordillera indicate the global influence of plate motion changes on ophiolite genesis and emplacement. The clustering of Mid-Jurassic ophiolite ages throughout the world provides circumstantial evidence that there is a kinematic link between marginal basins on both sides of the Pacific and the Tethys. This global plate motion change event is probably associated with breakup of Pangaea and initiation of rapid sea-floor spreading in the central Atlantic. Many new west-directed subdue tion zones were initiated in the circum-Pacific as a result of these changes (Dickinson et al. 1996; Hall 2002). Another major phase of ophiolite genesis is recorded in the Tethyan and Indonesian and New Guinea regions during the Late Cretaceous (Fig. 6). Emplacement ages of ophiolites rarely cluster, because of local variations in the timing of collisions of oceanic plate edges with continental and other buoyant lithosphere. The Mesozoic ophiolites of Borneo, New Guinea, Halmahera and the Philippines were all emplaced at different times depending upon when the continental margins and arcs they overrode arrived at the leading edges of the trenches they occupied. The present configuration of marginal basins throughout the Indonesian and New Guinea region separated by

499

sinews of continental material presents a preemplacement view of what many ophiolite-bearing mountain systems may have looked like. The Halmahera ophiolite still has not been 'emplaced', but represents exposed forearc basement of an east-directed subduction zone similar to the occurrence of ophiolites in east-directed subduction systems of Cyprus and Macquarie Island.

Resolution of the 'ophiolite conundrum' Several features of Tethyan ophiolites are considered a 'conundrum' because they have an arcrelated petrochemistry, but a lack of 'western Pacific-like island arcs preserved' (Moores et al. 2000). Many of the ophiolites and marginal basins of the Indonesian and New Guinea region also lack this close association with island arcs, mostly as a result of the combined effects of strain partitioning away from the trench into the arc and back-arc regions, and rapid plate boundary reorganization. The Andaman Sea and Woodlark Basin provide examples that have been used as a model for the development of Tethyan ophiolites (Moores et al. 1984; Dilek & Moores 1990). Other examples of marginal basins with tenuous ties to any specific arc include the South China, Sulu and Celebes Seas. The Molucca Sea region demonstrates how arcs may be obscured during collision. One model proposed by Moores et al. (2000) to resolve the 'ophiolite conundrum' is the 'historical contingency' of mantle sources for ophiolites; in other words, the history of the mantle is not reflected in the current plate configuration above it, such as a spreading ridge above mantle that was chemically altered by a previous subduction event. The Woodlark Basin of the New Guinea region provides an active example of this disparity between the geochemical zoning of the spreading ridge and current plate configurations (Binns & Whitford 1987; Johnson et al. 1987; Perfit et al. 1987; Dilek & Moores 1990). Magmatism in this actively spreading ocean basin demonstrates that old slabs can continue to contribute volatiles to the asthenospheric wedge long after subduction has ceased, and that the older the slab, the longer the volatile retention time (Abbot & Fisk 1986). Seismic tomography also shows how the entire Indonesian and New Guinea region is underlain by subducted slabs (Hafkenscheid et al. 2001), which may have contaminated most of the mantle source regions for oceanic crust production. The combination of strain partitioning above an oblique convergent margin and the possible effects of asthenospheric flow provide a way to open many small ocean basins, like those of the New

500

R. HARRIS

Guinea region, that are tapping metasomatized mantle from previous episodes of subduction.

Conclusions (1) The ophiolites and marginal basins of the Indonesian and New Guinea region provide a glimpse of ophiolite-forming processes at various stages of development. (2) Marginal basins form mostly in zones of highly oblique convergence where strain is partitioned away from the trench into a wide plate boundary zone consisting of several independently moving blocks. (3) Periodic disruption of these wide plate boundary zones by global plate motion changes, collision and asthenospheric flow initiates rapid phases of near-simultaneous extension and shortening that can open and obduct new ocean basins in a relatively short time to form ophiolites. (4) Most ophiolites in the Indonesian and New Guinea region are heterogeneous in age and composition, but have some component of SSZ petrochemistry. (5) Some ultramafic bodies interpreted as ophiolites are probably thrust sheets of subcontinental mantle that was sheared off the lower plate during collision. Examples may include the ultramafic masses found throughout the Banda arc and in western New Guinea that are thrust up over the Australian shelf. (6) Temporal patterns of ophiolite genesis in the ING region, and perhaps worldwide, correspond to collision-related global plate motion changes, and possible phases of resulting mantle extrusion. (7) Eastward asthenospheric flow may explain the consistent asymmetry and behaviour of subduction zones throughout the Indonesian and New Guinea region. (8) The strong spatial, temporal, geological and geochemical correlations between marginal basins and ophiolites in the Indonesian and New Guinea region resolve many of the problems associated with the 'ophiolite conundrum'. (9) The Indonesian and New Guinea region demonstrates the diversity and complexity of tectonic scenarios that can result in exposure of oceanic lithosphere. E. Silver, H. Davies, Y. Dilek and J. Milsom graciously provided thorough reviews of this manuscript. M. Cloos and R. Hall kindly provided data from student theses. C. Ng is the computer graphics master.

References ABBOT, D. & FISK, M. 1986. Tectonically controlled origin of three unusual rock suites in the Woodlark

Basin. Tectonics, 5, 1145-1160. ABBOTT, MJ. & CHAMALAUN, F.H. 1981. Geochronology of some Banda Arc volcanics. In: BARBER, AJ. & WIRYOSUJONO, S. (eds) The Geology and Tectonics of Eastern Indonesia. Geological Research and Development Centre, Special Publications, 2, 253-268. ACHARYYA, S.K., RAY, K.K. & ROY, O.K. 1989. Tectono-stratigraphy and emplacement history of the ophiolites assemblage from the Naga Hills and Andaman Island arc, India. Geological Society of India Journal, 33, 4-18. AUDLEY-CHARLES, M.G. 1986. Rates of Neogene and Quaternary tectonic movements in the southern Banda arc based on micropaleontology. Journal of the Geological Society, London, 143, 161-175. AUDLEY-CHARLES, M.G. & HARRIS, R.A. 1990. Allochthonous terranes of the Southwest Pacific and Indonesia. Philosophical Transactions of the Royal Society of London, 331, 571-587. BADER, A.G., PUBELLIER, M. & RANGIN, C. 1999. Active slivering of oceanic crust along the Molucca ridge (Indonesia-Philippine). Implication for ophiolite incorporation in a subduction wedge. Tectonics, 18(4), 606-620. BALLANTYNE, D. & HALL, R. 1990. The petrology of the Halmahera Ophiolite, Indonesia. An early Tertiary forearc. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites; Oceanic Crustal Analogues; Proceedings of the Symposium Troodos 1987. Ministry of Agriculture and Natural Resources, Nicosia, 461-475. BEARD, J.S. 1986. Characteristic mineralogy of arcrelated cumulate gabbros. Implications for the tectonic setting of gabbroic plutons and for andesite genesis. Geology, 14(10), 848-851. BERRY, R.F. 1981. Petrology of the Hili Manu Iherzolite, East Timor. Journal of the Geological Society of Australia, 28, 453-469. BERRY, R.F. & TENNER, G.A. 1982. Basalt geochemistry as a test of the tectonic models of Timor. Journal of the Geological Society, London, 139, 593-604. BERRY, R.F. & MCDOUGALL, I. 1986. Interpretation of 40 Ar/39Ar and K/Ar dating evidence from the Aileu Formation, East Timor, Indonesia. Chemical Geology, 59, 43-58. BINNS, R.A. & WHITFORD, D.J. 1987. Volcanic rocks from the Western Woodlark Basin, Papua New Guinea. Proceedings, Pacific Rim Congress, 87, 525-531. BLOOMER, S.H. 1983. Distribution and origin of igneous rocks from the landward slopes of the Mariana Trench: implications for its structure and evolution. Journal of Geophysical Research, 88, 7411-7428. BLOOMER, S.H., TAYLOR, B., MACLEOD, C.J., STERN, R.J., FRYER, P., HAWKINS, J.W. & JOHNSON, L. 1995. Early arc volcanism and the ophiolite problem: a perspective from drilling in the western Pacific. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88, 1-30. BOILLOT, G., GRIMAUD, S., MAUFFRET, A., MOUGENOT, D., KORNPROBST, J., MERGOIL-DANIEL, J. & TORRENT, G. 1980. Ocean-continent boundary off the

OPHIOLITES OF SE ASIA AND NEW GUINEA Iberian margin. A serpentinite diapir west of the Galicia Bank. Earth and Planetary Science Letters, 48, 23-34. BONATTI, E., OTTONELLO, G. & HAMLYN, R. 1986. Peridotites from the island of Zabargad (St. John), Red Sea: petrology and geochemistry. Journal of Geophysical Research, 91, 599-631. BOWIN, C.O., PURDY, G.M., JOHNSTON, C, SHOR, G., LAWYER, L., HARTONO, H.M.S. & JEZEK, P. 1980. Arc-continent collision in the Banda Sea region. AAPG Bulletin, 64, 868-915. BROWN, M. & EARLE, M.M. 1983. Cordierite-bearing schists and gneisses from Timor, eastern Indonesia. P-T conditions of metamorphism and tectonic implications. Journal of Metamorphic Geology, 1, 183-203. CARDWELL, R.K., ISACKS, B.L. & KARIG, D.E. 1980. The spatial distribution of earthquakes, focal mechanisms solutions and subducted lithosphere in the Philippine and northeastern Indonesian islands. /W.HAYES, D.E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. Geophysical Monograph, American Geophysical Union, 23, 1-36. COOPER, C. & TAYLOR, B. 1987. Seismotectonics of New Guinea. A model for arc reversal following arc-continent collision. Tectonics, 6, 53-68. Cox, A. & ENGEBRETSON, D. 1985. The change in motion of the Pacific plate at 5 Myr BP. Nature, 313, 472-474. CURRAY, J.R. 1989. The Sunda Arc: a model for oblique plate convergence. Netherlands Journal of Sea Research, 24, 131-140. DALZIEL, I.W.D. 1981. Back-arc extension in the Southern Andes; a review and critical reappraisal. Philosophical Transactions of the Royal Society of London, Series A, 300(1454), 319-335. DASGUPTA, S., MUKHOPADHYAY, M. & NANDY, D.R. 1990. Magmatism and tectonic setting in the Burmese-Andaman Arc. Indian Journal of Geology, 62(2), 117-141. DA VIES, H.L. & JACQUES, A.L. 1984. Emplacement of ophiolite in Papua New Guinea. In: GASS, I.G., LIPPARD, S.J. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 341-350. DAVIES, H.L. & SMITH, I.E. 1971. Geology of Eastern Papua. Geological Society of America Bulletin, 82, 8299-8312. DEWEY, J.F. & BIRD, J.B. 1970. Mountain belts and the new global tectonics. Journal of Geophysical Research, 75, 2625-2647. DICKINSON, W.R., HOPSON, C.A. & SALEEBY, J.B. 1996. Alternate origins of the Coast Range Ophiolite (California): introduction and implications. GSA Today, 6, 1-10. DILEK, Y. & MOORES, E.M. 1990. Regional tectonics of the eastern Mediterranean ophiolites. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites; Oceanic Crustal Analogues; Proceedings of the Symposium Troodos 1987. Ministry of Agriculture and Natural Resources, Nicosia, 295-309. DOGLIONI, C. 1993. Geological evidence for a global

501

tectonic polarity. Journal of the Geological Society, London, 150,991-1002. DOGLIONI, C., HARABAGLIA, P., MERLINI, S., MONGELLI, F., PECCERILLO, A. & PIROMALLO, C. 1999. Orogens and slabs vs. their direction of subduction. Earth-Science Reviews, 45, 167-208. Dow, D.B. & SUKAMTO, R. 1984. Western Irian Jaya: the end product of oblique plate convergence in the late Tertiary. Tectonophysics, 106, 109-139. Dow, D.B., ROBINSON, G., HARTONO, U. & RATMAN, N. 1986. Geologic Map of Irian Jaya, Indonesia, 1:1000000 scale. Indonesian Geological Research and Development Center, Bandung. Dow, D. B., ROBINSON, G., HARTONO, U. & RATMAN, N. 1988. Geology of Irian Jaya. In: Sukamto, R. & Trail, D. S. (project managers) Irian Jaya Geological Mapping Project, Geoscientific Data Package. Indonesian Geological Research and Development Center and Australian Bureau of Mineral Resources, 298. DUNCAN, R.A. & CLAGUE, D.A. 1985. Pacific Plate motion recorded by linear volcanic chains. In: NAIRN, A.E., STEHLI, F.G. & UYEDA, S. (eds) The Ocean Basins and Margins, 7A—The Pacific Ocean. Plenum, New York, 89-121. EDELMAN, S.H. 1988. Ophiolite generation and emplacement by rapid subduction hinge retreat on a continent-bearing plate. Geology, 16, 311-313. EGUCHI, T., UYEDA, S. & MAKI, T. 1979. Seismotectonics and tectonic history of the Andaman Sea. Tectonophysics, 57, 35-51. EPP, D. 1984. Possible perturbations of the hotspot traces and implications for the origin and structure of the Line Islands. Journal of Geophysical Research, 89, 11273-11286. EVANS, C.A., HAWKINS, J.W. & MOORE, J.F. 1983. Petrology and geochemistry of ophiolitic and associated volcanic rocks on the Talaud island, Molucca Sea collision zone, northeast Indonesia. In: HILDE, T.W.C. & UYEDA, S. (eds) Geodynamics of the Western Pacific-Indonesian Region. Geodynamics Series, American Geophysical Union, 11, 159—172. FLOWER, M.F.J., Russo, R.M., TAMAKI, K. & HOANG, N. 2001. Mantle contamination and the IzuBonin-Mariana (IBM) 'high-tide mark': evidence for mantle extrusion caused by Tethyan closure. Tectonophysics, 333, 9-34. FORTUIN, A.R. & DE SMET, M.E.M. 1991. Rates and magnitudes of late Cenozoic vertical movements in the Indonesian Banda Arc and the distinction of eustatic effects. In: MACDONALD, D.I.M. (ed.) Sedimentation, Tectonics and Eustasy: Sea-level Changes at Active Margins. International Association of Sedimentologists, Special Publications, 12, 79-89. GENRICH, J.F., BOCK, Y., MCCAFFREY, R., CALAIS, E., STEVENS, S.W. & SUBARYA, C. 1996. Accretion of the southern Banda arc to the Australian plate margin determined by Global Positioning System measurements. Tectonics, 15, 288-295. GUILOU, H., ARNAUD, N. & HONTHAAS, C. ET AL. 1998. Neogene back-arc origin for the Banda Sea basins: geochemical and geochronological constraints from the Banda ridges (East Indonesia). Tectonophysics,

502

R. HARRIS

298,297-317. GUZMAN-SPEZIALE, M. & Ni, J.F. 1993. The opening of the Andaman Sea: where is the short-term displacement being taken up? Geophysical Research Letters, 20, 2949-2952. HAFKENSCHEID, S.J.H., BUITER, M.J.R., WORTEL, W. & SPAKMAN, H.B. 2001. Modelling the seismic velocity structure beneath Indonesia: a comparison with tomography. Tectonophysics, 333, 35-46. HALDAR, D., LASKAR, T., BANDYOPADHYAY, C., SARKAR, N.K. & BISWAS, J.K. 1992. Volcanic eruption of the Barren Island Volcano, Andaman Sea. Journal of the Geological Society of India, 39, 411-419. HALL, R. 1996. Reconstructing Cenozoic SE Asia. In: HALL, R. & BLUNDELL, D. (eds) Tectonics of Southeast Asia. Geological Society, London, Special Publications, 106, 153-184. HALL, R. 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computerbased reconstructions, model and animations. Journal of Asian Earth Sciences, 20, 353-431. HALL, R. & BLUNDELL, D. (EDS) 1996. Tectonics of Southeast Asia. Geological Society, London, Special Publications, 106. HALL, R., AUDLEY-CHARLES, M.G., BANNER, F.T., HIDAYAT, S. & TOEING, S. 1988. Basement rocks of the Halmahera region, eastern Indonesia; a Late Cretaceous—early Tertiary arc and forearc. Journal of the Geological Society, London, 145, 65-84. HALL, R., NICHOLS, G., BALLANTYNE, P., CHARLTON, T. & ALI, J. 1991. The character and significance of basement rocks of the southern Molucca Sea region. Journal of Southeast Asian Earth Sciences, 6, 249— 258. HAMILTON, W. 1979. Tectonics of the Indonesian Region. US Geological Survey Professional Papers, 1078. HARRIS, R.A. 1991. Temporal distribution of strain in the active Banda orogen: a reconciliation of rival hypotheses. Journal of Southeast Asian Earth Sciences, 6, 373-386. HARRIS, R.A. 1992. Peri-collisional extension and the formation of Oman-type ophiolites in the Banda Arc and Brooks Range. In: PARSON, L.M., MURTON, BJ. & BROWNING, P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 301-325. HARRIS, R.A. & LONG, T. 2000. The Timor ophiolite, Indonesia: model or myth? In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 321-330. HARRIS, R.A., SAWYER, R.K. & AUDLEY-CHARLES, M.G. 1998. Collisional melange development: geologic associations of active melange-forming processes with exhumed melange facies in the western Banda orogen. Indonesia. Tectonics, 17, 458-480. HARRIS, R.A., KAISER, J., HURFORD, AJ. & CARTER, A. 2000. Thermal history of Australian passive margin sequences accreted to Timor during Late Neogene arc—continent collision. Indonesia. Journal of Asian Earth Sciences, 18, 47-69. HELMERS, J., SOPAHELUWAKEN, J. & TJOKROSAPOETRO,

S. 1989. Tectonics of the Indonesian Region. US Geological Survey Professional Papers, 24, 357-371. HlNSCHBERGER, F., MALOD, J.-A., DYMENT, J., HON-

THAAS, C., REHAULT, J. & BURHANUDDIN, S. 2001. Magnetic lineation constraints for the back-arc opening of the Late Neogene south Banda Basin (eastern Indonesia). Tectonophysics, 333, 47-59. HLA-MAUNG 1987. Transcurrent movements in the Burma-Andaman Sea region. Geology, 15,911 -912. HONTHAAS, C., REHAULT, J. & MAURY, R. ET AL. 1998. A Neogene back-arc origin for the Banda Sea basins: geochemical and geochronological constraints from the Banda ridges (East Indonesia). Tectonophysics, 298, 297-317. HOUSH, T., JAMES, E. & MCMAHON, T. 1994. Isotopic constraints on Pliocene 'post-orogenic' magmatism in the Central Ranges of Irian Jaya (New Guinea), Indonesia. EOS Transactions, American Geophysical Union, 75, 739. JAQUES, A.L. 1981. Petrology and petrogenesis of cumulate peridotites and gabbros from the Marum ophiolite complex, northern Papua New Guinea. Journal of Petrology, 22, 1 -40. JAQUES, A.L. & ROBINSON, G. 1977. The continent/ island-arc collision in northern Papua New Guinea. BMR Journal of Australian Geology and Geophysics, 3, 289-303. JAFRI, S.H., BALARAM, V. & RAMESH, S.L. 1990. Geochemistry of Andaman-Nicobar island basalts: a case for a possible plume origin. Journal of Volcanology and Geothermal Research, 44, 339-347. JOHNSON, R.W., LANGMUIR, C.H. & PERFIT, M.R. ET AL. 1987. Ridge subduction and forearc volcanism: petrology and geochemistry of dredge rock samples from the western Solomon Arc and Woodlark Basin. In: TAYLOR, B. & EXON, N.F. (eds) Marine Geology, Geophysics and Geochemistry of the Woodlark Basin-Solomon Islands. Circum-Pacific Council for Energy and Mineral Resources, Houston, TX, 155-226. KADARUSMAN, A. 2001. Geochemistry and Sm-Nd dating of garnet peridotites from Central Sulawesi, and its implication to the Neogene collision complex in Eastern Indonesia. EOS Transactions, American Geophysical Union, 82(47), 1264. KETNER, K.B., KASTOWO, MODJO, SUBROTO, NAESER, C.W., OBRADOVICH, J., ROBINSON, K., SUPTANDAR, T. & WIKARNO, X. 1976. Pre-Eocene rocks of Java, Indonesia. US Geological Survey Journal of Research, 4, 605-614. KREEMER, C., HOLT, W.E., GOES, S. & GOVERS, R. 2000. Active deformation in eastern Indonesia and the Philippines from GPS and seismicity data. Journal of Geophysical Research, 105, 663-680. KROENKE, L.W. 1996. Plate tectonic development of the western and southwestern Pacific: Mesozoic to the present. In: KEAST, A. & MILLER, S. (eds) The Origin and Evolution of Pacific Island Biotas, New Guinea to Eastern Polynesia: Patterns and Processes. Academic Publishing, Amsterdam, 19-34. KUNDIG, E.H. 1956. Geology and ophiolite problems of East Celebes. Nederlandsch Geologische Mijn-

OPHIOLITES OF SE ASIA AND NEW GUINEA bouwk Genootsch, 16, 210-235. LAPOUILLE, A., HARYONO, H., LARUE, M. & LARDY, M. 1985. Age and origin of the seafloor of the Banda Sea. Oceanologica Acta, 8, 379-389. LEE, C.S. & McCABE, R. 1986. The Banda-CelebesSulu basin: a trapped piece of Cretaceous-Eocene oceanic crust? Nature, 322, 51-54. LEE, T.Y. & LAWYER, L.A. 1994. Cenozoic plate reconstruction of Southeast Asia. Tectonophysics, 251, 85-138. LINTHOUT, K. & HELMERS, H. 1994. Pliocene obducted, rotated and migrated ultramafic rocks and obduction-induced anatectic granite, SW Seram and Ambon, Eastern Indonesia. Journal of Southeast Asian Earth Sciences, 9, 95-109. LIPPARD, S.J., SHELTON, A.W. & GASS, I.G. 1984. The Ophiolite of Northern Oman. Geological Society of London, Memoir, 11. Lus, W.Y., GREEN, D., MCDOUGALL, I., EGGINS, S. & DA VIES, H. 2001. Papuan Ultramafic Belt (PUB) ophiolite: field mapping, petrology, mineral chemistry, geochemistry, geochronology, and experimental studies of the metamorphic sole. EOS Transactions, American Geophysical Union, 82(47), 1264. MALOD, J.A. & KEMAL, B.M. 1996. The Sumatra margin: oblique subduction and lateral displacement of the Accretionary prism. In: HALL, R. & BLUNDELL, DJ. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Publications, 106, 19-28. MCCAFFREY, R. 1988. Active tectonics of the eastern Sunda and Banda Arcs. Journal of Geophysical Research, 93(B12), 15163-15182. MCCAFFREY, R. 1991. Earthquakes and ophiolite emplacement in the Molucca Sea collision zone, Indonesia. Tectonics, 10(2), 433-453. MCCAFFREY, R. 1996. Slip partitioning at convergent plate boundaries of SE Asia. In: HALL, R. & BLUNDELL, D. (eds) Tectonics of Southeast Asia. Geological Society, London, Special Publications, 106, 3-18. MCCAFFREY, R. & NABELEK, J. 1984. The geometry of backarc thrusting along the eastern Sunda Arc, Indonesia: constraints from earthquake and gravity data. Journal of Geophysical Research, 89, 6171 — 6179. MCCAFFREY, R., SILVER, E.A. & RAITT, R.W. 1980. Crystal structure of the Molucca Sea collision zone, Indonesia. In: HAYES, D.E. (ed.) The Tectonic and Geological Evolution of Southeast Asian Seas and Islands, Geophysical Monograph of the American Geophysical Union, 23, 161-177. MCCAFFREY, R., MOLNAR, P., ROECKER, S. & JOYODIWIRYO, Y. 1985. Microearthquake seismicity and fault plane solutions related to arc-continent collision in the eastern Sunda arc, Indonesia. Journal of Geophysical Research, 90, 4511-4528. MERRITTS, D., EBY, R., HARRIS, R.A., EDWARDS, R.L. & CHENG, H. 1998. Variable rates of Late Quaternary surface uplift along the Banda Arc-Australian Plate Collision Zone, eastern Indonesia. In: STEWART, I.S. & VITA-FINZI, C. (eds) Coastal Tectonics. Geological Society, London, Special Publications, 146, 213-224.

503

MINSTER, J.B., JORDAN, T.H., MOLNAR, P. & HAINES, E. 1974. Numerical modeling of instantaneous plate tectonics. Geophysical Journal of the Royal Astronomical Society, 36, 541-576. MITCHELL, A.H.G. 1983. Where have all the old arcs gone? Philippine Geologist, 37, 20-41. MITCHELL, A.H.G. 1993. Cretaceous-Cenozoic tectonic events in the western Mayanmar (Burma) Assam region. Journal of the Geological Society, London, 150, 1089-1102. MIYASHIRO, A., AKI, K. & SENGOR, A.M.C. 1982. Orogeny. Wiley, Chichester. MONNIER, C., GlRARDEAU, J., MAURY, R.C. & GOTTEN,

J. 1995. Back arc basin origin for the East Sulawesi ophiolite (eastern Indonesia). Geology, 23, 851-854. MONNIER, C., GIRARDEAU, J., PUBELLIER, M., POLVE, M., PERMANA, H. & BELLON, H. 1999. Petrology and geochemistry of the Cyclops ophiolites (Irian Jaya, East Indonesia): consequences for the Cenozoic evolution of the north Australian margin. Mineralogy and Petrology, 65, 1-28. MONNIER, C., GIRARDEAU, J., PUBELLIER, M. & PERMANA, H. 2000. L'ophiolite de la chaine centrale d'Irian Jaya (Indonesie): evidences petrologiques et geochimiques pour une origine dans un basin arriere-arc. Earth and Planetary Sciences, 331, 691-699. MOORE, G.F., CURRAY, J.R., MOORE, D.G. & KARIG, D.E. 1980. Variation in geologic structure along the Sunda Fore arc, northeastern Indian Ocean. In: HAYES, D.E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. Geophysical Monograph, American Geophysical Union, 23, 145-160. MOORE, G.F., KADARISMAN, D., EVANS, C.A. & HAWKINS, J.W. 1981. Geology of the Talaud Islands, Molucca Sea collision zone, northeast Indonesia. Journal of Structural Geology, 3, 467-475. MOORES, E., KELLOG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency'; a resolution of the 'ophiolite conundrum'. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust; New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 3-12. MOORES, E.M., ROBINSON, T., MALPAS, J. & XENOPHONTOS, C. 1984. Model for the origin of the Troodos massif, Cyprus and other mideast ophiolites. Geology, 12, 500-503. MUBROTO, B., BRIDEN, J.C., MCCLELLAND, E. & HALL, R. 1994. Paleomagnetism of the Balantak ophiolite, Sulawesi. Earth and Planetary Science Letters, 125, 193-209. NICHOLLS, LA., FERGUSON, J., JONES, H., MARKS, G. & MUTTER, J.C. 1981. Ultramafic blocks from the ocean floor southwest of Australia. Earth and Planetary Science Letters, 56, 362-374. PARKINSON, C. 1998. Emplacement of the East Sulawesi Ophiolite: evidence from subophiolite metamorphic rocks. Journal of Asian Earth Sciences, 16, 13-28. PEARCE, J.A., LIPPARD, SJ. & ROBERTS, S. 1984. Characteristics and tectonic significance of supra-

504

R. HARRIS

subduction ophiolites. In: KOKELAAR, B. & HoWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PERFIT, M.F., LANGMUIR, C.H., BAEKISAPA, M., CHAPPELL, B., JOHNSON, R.W., STAUDIGEL, H. & TAYLOR, S.R. 1987. Geochemistry of volcanic rocks from the Woodlark Basin: addressing the questions of ridge subduction and a reversal in the polarity of subduction. In: TAYLOR, B. & EXON, N.F. (eds) Marine Geology, Geophysics and Geochemistry of the Woodlark Basin—Solomon Islands. Circum-Pacific Council for Energy and Mineral Resources, Houston, TX, TX, 113-154. PIETERS, E., RYBURN, RJ. & TRAIL, D.S. 1979. Geological reconnaissance in Irian Jay a, 1976—1977, 1979/19. PIGRAM, CJ. & DA VIES, H.L. 1987. Terranes and the accretion history of the New Guinea orogen. BMR Journal of Australian Geology and Geophysics, 10, 193-211. PIGRAM, CJ. & SYMONDS, A. 1991. A review of the timing of the major tectonic events in the New Guinea Orogen. Journal of Southeast Asian Earth Sciences, 6, 307-318. PRASETYADI, C. & HARRIS, R.A. 1996. Hinterland structure of the active Banda arc-continent collision, Indonesia: constraints from the Aileu complex of East Timor. In: Proceedings, 25th Convention of the Indonesian Association of Geology. Jakarta, Town, 144-173. PRASETYO, H. 1989. Marine Geology and Tectonic Development of the Banda Sea Region, Eastern Indonesia; a Model of an 'Indo-borderland' Marginal Basin. Marine Geological Institute of Indonesia, Special Publications, 1, 427. PRASETYO, H. 1994. The tectonics of the 'Sunda-Banda' forearc transition zone, eastern Indonesia. Bulletin of the Marine Geological Institute of Indonesia, 9(1), 23-47. PRAWIRODIRDJO, L., BOCK, Y. & MCCAFFREY, R. ET AL. 1997. Geodetic observations of interseismic strain segmentation at the Sumatra subduction zone. Geophysical Research Letters, 24, 2601-2604. PUBELLIER, M. & COBBOLD, R. 1996. Analogue models for the transpressional docking of volcanic arcs in the western Pacific. Tectonophysics, 253, 33-52. QUARLES VAN UFFORD, A. 1996. Stratigraphy, structural geology, and tectonics of a young forearc—continent collision, western Central Range, Irian Jaya (western New Guinea), Indonesia. PhD dissertation, University of Texas, Austin. RAMPONE, E. & PICCARDO, G.B. 2000. The ophiolites— oceanic lithosphere analogue: new insights from the Northern Apennines (Italy). In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust, New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 21-34. RAY, K.K., SENGUPTA, S. & VAN DEN HUL, H.J. 1988. Chemical characters of volcanic rocks from the Andaman ophiolite, India. Journal of the Geological Society, London, 145, 393-400. ROBINSON, P.T., MELSON, W.G., O'HEARN, T. &

SCHMINCKE, H.-U. 1983. Volcanic glass compositions of the Troodos ophiolite, Cyprus. Geology, 11, 400-404. ROYDEN, L.H. 1993. The tectonic expression of slab-pull at continental convergent boundaries. Tectonics, 12, 303-325. SAPIIE, B., NATAWIDJAYA, D.H. & CLOOS, M. 1999. Strike-slip tectonics of New Guinea: transform motion between the Caroline and Australian Plates. In: BUSONO, I. & ALAM, H. (eds) Developments in Indonesian Tectonics and Structural Geology. Proceedings of Indonesian Association of Geologists, 1, 1-12. SEARLE, M. & STEVENS, R.K. 1984. Obduction processes in ancient, modern and future ophiolites. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 303-320. SENGUPTA, S., RAY, K.K., ACHARYYA, S.K. & DE SMETH, J.B. 1990. Nature of ophiolites occurrences along the eastern margin of the Indian plate and their tectonic significance. Geology, 18, 439-442. SERRI, G. ET AL. 1991. Petrology of igneous rocks from the Celebes Sea basement. In: SILVER, E.A. & RANGIN, C. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 124. Ocean Drilling Program, College Stations, TX, 271-296. SILVER, E.A. & MOORE, J.C. 1978. The Molucca Sea collision zone, Indonesia. Journal of Geophysical Research, 83, 1681-1691. SILVER, E.A. & RANGIN, C. 1989. Celebes and Sulu marginal basins: constraints from ODP Leg 124. EOS Transactions, American Geophysical Union, 70, 1365. SILVER, E.A. & SMITH, R.B. 1983. Comparison of terrain accretion in Southeast Asia and the Mesozoic North American Cordillera. Geology, 11, 198— 202. SILVER, E.A., JOYODIWIRYO, Y. & MCCAFFREY, R. 1978. Gravity results and emplacement geometry of the Sulawesi ultramafic belt, Indonesia. Geology, 6, 527-531. SILVER, E.A., MCCAFFREY, R., JOYODIWIRYO, Y. & STEVENS, S. 1983. Ophiolite emplacement and collision between the Sula Platform and the Sulawesi island arc, Indonesia. Journal of Geophysical Research, 88, 9419-9435. SILVER, E.A., GILL, J.B., SCHWARTZ, D., PRASETYO, H. & DUNCAN, R.A. 1985. Evidence for a submerged and displaced continental borderland, north Banda Sea, Indonesia. Geology, 13, 687-691. SILVER, E.A., ABBOTT, L.D., KIRCHOFF-STEIN, K.S., REED, D.L., BERNSTEIN, T.B. & HILYARD, D. 1991. Collisional propagation in Papua New Guinea and the Solomon Sea. Tectonics, 10, 836-874. SIMANDJUNTAK, T.O. 1992. New data on the age of the ophiolite in Eastern Sulawesi. Bulletin of the Geological Research and Development Center, Bandung, 15, 38-44. SIMONS, W.M.F., AMBROSIUS, B.A.C. & NOOMEN, R. ET AL. 1999. Observing plate tectonics in SE Asia: geodetic results of the GEODYSSEA project. Geophysical Research Letters, 26, 2081-2084.

OPHIOLITES OF SE ASIA AND NEW GUINEA SMITH, A.D. & LEWIS, C. 1999. Differential rotation of lithosphere and mantle and the driving force of plate tectonics. Geodynamics, 28, 97-116. SOPAHELUWAKAN, J., HELMERS, H., TlOKROSAPOETRO,

S. & SURYA NILA, E. 1989. Medium pressure metamorphism with inverted thermal gradient associated with ophiolite nappe emplacement in Timor. Netherlands Journal of Sea Research, 24, 333-343. STEINMANN, G. 1906. Geologische Beobachtungen in den Alpenll, Die schard'tsche Uberfaltungstheorie und die geologischen bedeutung der Teefseeabatze und der ophiolitische Massengesteine. Berichte der naturforschenden Gesellschaft zu Freiburg im Breisgau. Wagner, Freiburg im Breisgau, 1 -66. STEVENS, C., MCCAFFREY, R. & BOCK, Y. ET AL. 1999. Rapid rotations about a vertical axis in a collision setting revealed by the Palu fault, Sulawesi, Indonesia. Geophysical Research Letters, 26, 2677-2680. STOCKMAL, G.S. 1983. Modeling of large-scale accretionary wedge deformation. Journal of Geophysical Research, 88, 8271-8287. SUKAMTO, R. 1979. Tectonic significance of melange on the Talaud Islands of northeastern Indonesia. Bulletin of the Geological Research and Development Center, Bandung, 2, 7-19. TAMAKI, K. & HONZA, E. 1991. Review of global marginal basins. Episodes, 14(3), 224-230. TAYLOR, B. & KARNER, G.D. 1983. On the evolution of marginal basins. Review of Geophysics and Space Physics, 21(8), 1727-1741. TREGONING, P., BRUNNER, F.K. & BOCK, R. ET AL. 1994. First geodetic measurement of convergence across the Java Trench. Geophysical Research Letters, 21, 2135-2138. UYEDA, S. & KANAMORI, H. 1979. Backarc opening and

505

the mode of subduction. Journal of Geophysical Research, 84, 1049-1061. VOGT, R. 1973. Subduction and aseismic ridges. Nature, 241, 189-191. VOLPE, A.M., MACDOUGALL, J.D., LUGMAIR, G.W., HAWKINS, J.W. & LONSDALE, F. 1990. Fine-scale isotopic variation in Mariana Trough basalts; evidence for heterogeneity and a recycled component in backarc basin mantle. Earth and Planetary Science Letters, 100, 251-264. WARREN, Q. 1995. Petrology, structure and tectonics of the Ruffaer Metamorphic Belt, west central Irian Jay a, Indonesia. MSc thesis, University of Texas, Austin. WEEKS, L.A., HARBISON, R.N. & PETER, G. 1967. Island arc system in Andaman Sea. AAPG Bulletin, 51(9), 1803-1815. WEILAND, RJ. 1999. Emplacement of the Irian ophiolite belt and unroofing of the Ruffaer metamorphic belt of Irian Jay a, Indonesia. PhD dissertation, University of Texas, Austin. WEILAND, RJ. & CLOOS, M. 1996. Pliocene-Pleistocene asymmetric unroofing of the Irian fold belt, Irian Jaya, Indonesia: apatite fission-track thermochronology. Geological Society of America Bulletin, 108, 1438-1449. WEISSEL, J.K. 1980. Evidence for Eocene oceanic crust in the Celebes basin. In: HAYES, D.E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. Geophysical Monograph, American Geophysical Union, 23, 37-47. WIDIYANTORO, S. & VAN DER HILST, R. 1997. Mantle structure beneath Indonesia inferred from highresolution tomographic imaging. Geophysical Journal International, 130, 167-182.

This page intentionally left blank

Forearc ophiolites: a view from the western Pacific JOHN MILSOM Department of Earth Sciences, University College London, London WC1E 6BT, UK (e-mail: [email protected]) Abstract: Many of the world's largest ophiolite masses are now being interpreted as remnants of oceanic forearcs stranded on continental margins in the course of arc-continent collision. This interpretation implies a former association between the ophiolite and a volcanic arc at a distance of 100-200 km. The New Caledonia region of the SW Pacific is one of the few areas where there is good evidence for the presence of such an arc. On New Caledonia itself, the Grand Massif du Sud ophiolite has been thrust over the Norfolk Ridge continental fragment, while the coral islands of the Loyalty Group, c. 100km to the NE, cap large sub-sea edifices with volcanic morphology. New Caledonia can be correlated geologically with New Guinea, despite the considerable width of open ocean that now separates the two. In central New Guinea, where ophiolites emplaced along the northern flank of the main mountain spine lie some 100 km south of exposures of arc-volcanic basement in the north coast ranges, the similarities include the relationship between New Caledonia and the Loyalty Islands. There is, however, no obvious comparable relationship in the case of the largest of the New Guinea ophiolites, the Papuan Ultramafic Belt of the eastern peninsula, which is backed to the north and east by the oceanic Solomon Sea. An associated volcanic arc can be recognized in this area only by assuming a complicated history of collision, post-collision arc-forearc separation and sea-floor spreading, followed by renewed contraction and a very recent and continuing second collision. The case for such a sequence of events can be made on a number of grounds. If the processes responsible are general in their nature, they could explain the apparent absence of arc-volcanic belts in association with many other supposedly forearc ophiolites.

The currently accepted definition of the term 'ophiolite' was formulated during a Penrose Conference in 1972, at which time the ophiolitic rock association was widely regarded as representative of normal oceanic crust. However, as more and more non-mid-ocean ridge basalt (non-MORB) characteristics were noted in the larger massifs (e.g. Pearce et al. 1984) and as the mechanical problems associated with thrusting ocean crust over continental crust came to be better appreciated, ophiolites came increasingly to be seen as examples of oceanic forearcs emplaced on continental margins in the final stages of ocean closure (e.g. Gealey 1980). This hypothesis is supported by the presence of arc moleiites and boninites, as well as some more acid rocks, in many ophiolites. Components with typical MORB composition, which are generally less common (Bloomer et al. 1995), can be regarded as samples of the ocean crust on which the arc was formed, and their presence does not conflict with the hypothesis. A more serious objection is that former volcanic arcs that can be associated with most of the world's largest ophiolites are hard to find. The Troodos massif in Cyprus and the

Semail nappe in Oman are two of the best studied, because most accessible and well-exposed, examples (Shervais 2001). Searle & Cox (1999), amongst others, have suggested the Lasail unit as a possible 'immature' volcanic arc remnant in Oman, but these rocks are volumetrically minor and areally restricted, and an arc is even less obviously present in Cyprus. Yet the ophiolite forearc equation can be accepted only if volcanic arcs can be identified in at least the majority of cases.

The Bonin (Ogasawara) Arc Discussion of the 'forearc ophiolite' hypothesis can usefully begin with the uncollided Bonin forearc in the Western Pacific (Fig. 1; inset). This is, logically enough, the type area for the boninites that have come to be considered hallmarks of suprasubduction zone (SSZ) ophiolites. Rocks recovered by dredging and deep drilling, and from exposures on the forearc islands, are almost exclusively either boninites or island arc tholeiites very similar to those found on Cyprus (Bloomer et al. 1995). However, the Pacific Ocean is still

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 507-515. 0305-8719/037$ 15 © The Geological Society of London 2003.

508

J. MILSOM

Fig. 1. New Caledonia and New Guinea: trenches, rises and ophiolites. The inset shows the Bonin Islands region to the same scale. Pecked line in inset indicates location of profile of Figure 2a. Major mafic or ultramafic complexes are shown in black. IJOC, Irian Jaya Ophiolite Complex; AU, April Ultramafics; MC, Marum Complex; PUB, Papuan Ultramafic Belt; CV, Cape Vogel. Light stipple indicates water depths of less than 2000 m, dark stipple water depths of more than 6000 m.

being actively subducted beneath the Bonin Arc and it will be a very long time before its forearc can be emplaced on any continental margin. Gravity measurements on the Bonin Islands and in the adjacent seas (Fig. 2) are even more remarkable than the very large positive Bouguer gravity anomalies observed over ophiolites in Cyprus (Gass & Masson-Smith 1963), Papua New Guinea (Milsom 1973, 1984) and Oman (Manghnani & Coleman 1981; Ravaut et al. 1997). A gravity high that may well be the largest anywhere on the Earth's surface is associated with the Bonin forearc ridge (Honza & Tamaki 1985; Milsom et al. 1996), reflecting the presence of high-density rocks in the forearc wedge and of oceanic mantle at shallow depths. Emplacement of this forearc onto a continental margin would be expected to reduce the amplitude of the high, because of the presence of lighter and thicker crust on the downgoing plate, but not to eliminate it altogether. Apart from the presence of boninites, the significance of the Bonin forearc as a model for a pre-emplacement ophiolite lies in the spatial relationships that it emphasizes. The profile in Figure 2 passes close to Ninoshima Island in the volcanic arc, which is marked by a small positive anomaly.

The distance between Ninoshima and the peak of the forearc gravity high is about 150-200 km and there is then a further gap of 100 km between the forearc ridge and the axis of the trench. It might be argued that the Bonin Arc is exceptional in this respect, as in so many others (for example, the volcanoes active today are very young and the modern forearc probably developed in association with remnant arcs further west), but even in those modern arcs that lack any form of forearc ridge, the distance between the volcanic arc and the trench axis is seldom less than 200km. This observation makes it difficult to agree with Shervais (2001) when he argues that many ophiolites 'will eventually evolve into mature island arc systems in which the older ophiolitic stage of formation is obscured by a carapace of younger volcanics, and by intrusion of later plutons'. Sites of steady-state arc volcanism lie above points where subducted slabs have reached depths of at least 70 km and are far removed from the forearc regions where boninites and arc moleiites were intruded and extruded during arc initiation. During and after collision there would be significant thrust shortening and the extreme frontal part of the forearc would be rapidly removed by erosion, but if an ophiolite is indeed a thrust forearc, then the

FOREARC OPHIOLITES, WESTERN PACIFIC

Fig. 2. Gravity profiles compared, for the Bonin Arc, New Caledonia and the Papuan Ultramafic Belt. In all cases Bouguer gravity is displayed onshore and free-air gravity offshore. The Bonin profile (after Milsom et al. 1996) is taken across the peak of the forearc gravity high and has been reversed to display the same polarity with respect to collision as the other two profiles, i.e. east is towards the left on the Bonin profile. The New Caledonia profile and interpretation, to the same scale, is after Collot et al. (1987). The profile location is shown in Figure 3. The location of the Papuan Ultramafic Belt profile, redrawn from Milsom (1973) with marine data from Abbott et al. (1994) is shown in Figure 4.

distance between a strong ophiolite-associated gravity high and the main line of arc volcanism would be expected to be of the order of 100 km. As in the case of the Finisterre and Adelbert ranges of NE New Guinea (discussed below), rocks of the volcanic arc association can be emplaced directly on a continental margin only if the corresponding forearc has first been removed.

New Caledonia The island of New Caledonia, some 5000 km to the south of the Bonin Arc, occupies a key position in the SW Pacific, roughly halfway between New Guinea and New Zealand (Fig. 1). The island is strongly elongated in a NW-SE direction, being over 300 km long but only 50 km across (Fig. 3), and is flanked to both the NE and SW by linear reef complexes forming the Eastern and Western lagoons. The presence of these lagoons indicates past and probably continuing subsidence, which is likely to have been driven by the loading of the crust by dense peridotites that have been thrust over a continental basement

509

(Lillie & Brothers 1970). The peridotites, which include the economically important Grand Massif du Sud (GMS), are frequently referred to as 'ophiolites' and it seems reasonable to do so, even though the upper (mafic and sedimentary), parts of the classic sequence are present either to only a very minor degree or not at all. The age of emplacement, determined both by stratigraphic methods and by dating of high-pressure metamorphic rocks beneath the overthrust, is Late Eocene (Collot et al. 1987). Radiometric (K/Ar) dates on dolerites intruded into the peridotites cluster either between 80 and 100 Ma or around 52 Ma, implying a Late Cretaceous minimum age for the surrounding mass (Collot et al. 1987). Bouguer gravity values on New Caledonia are relatively low, averaging about 4-50 mGal (Collot et al. 1987), and reflect the presence at depth of rocks less dense than the peridotites. There is a local high, of over +100 mGal, over a part of the GMS, but the root zone for the overthrust is marked by the higher values, of up to more than + 160 mGal, found along the whole length of the Eastern Lagoon. In Figure 2, a gravity profile across New Caledonia and the adjacent marine area is compared with a profile across the Bonin Arc and forearc. Although very large by all other standards, the New Caledonia anomaly is dwarfed by the anomaly over the Bonin Arc, as would be predicted for a forearc that is thrust over continental rocks rather than oceanic crust. There is a second strong peak on the New Caledonia profile in approximately the region where, on the Bonin model, the corresponding volcanic arc should lie. This peak is associated with the coral islands of the Loyalty chain, which cap the summits of discrete bathymetric highs with distinctive volcanic morphologies. Direct geological evidence on the origin of the Loyalty Ridge is ambiguous. Quaternary coral limestones cover most of the islands, and within-plate basalts dated to 9-11 Ma crop out on the most southeasterly of the three, but older (30-40 Ma) hyaloclastites have been dredged from a site further north (Collot et al. 1987), and Bitoun & Recy (1982) suggested a much earlier (pre-Eocene) date for the ridge on the basis of seismic reflection profiles. If the islands do, in fact, cap a volcanic arc associated with a 'Grand Massif du Sud' forearc, then the much stronger gravity anomaly, as compared with the Bonin volcanic arc, can be explained in part by a longer history of development and in part by the fact that the Loyalty Islands now occupy the crest of the pre-trench bulge in front of the New Hebrides Trench. Collot et al. (1987) and Auzende et al. (2000) have presented a variety of data that they interpreted as showing that the New Caledonia perido-

510

J. MILSOM

Fig. 3. Bathymetry, onshore gravity and ultramafic rocks in the New Caledonia region, SW Pacific Ocean after Collot et al. (1987). Bouguer gravity values in mGal. Light stipple indicates closed gravity lows. Bathymetry indicated by shading. Water depths exceed 2000 m in places in the Loyalty Basin, and exceed 3500 m in the trough SW of New Caledonia. Outlines of outcrops of ultramafic rocks indicated by pecked line (compare Figure 1). Bold dashed line shows location of profile of Figure 2b. GMS, Grand Massif du Sud.

tites are the overthrust leading edge of the Loyalty Basin. Surprisingly, the possibility that the Loyalty Islands mark the line of the volcanic arc to the New Caledonia forearc was suggested by neither group of workers, and the models proposed involve subduction beneath the SW coast of New Caledonia rather than immediately beneath the ophiolite. However, the basement rocks of New Caledonia do not constitute a plausible forearc. Described as continental by Lillie & Brothers (1970), they provide a rare glimpse of the basement of the Norfolk Ridge, which was rifted away from the east coast of Australia during the Paleocene initiation of the Tasman Sea. The deep marine trough that flanks New Caledonia to the SW (Fig. 3) is readily explained as a foreland basin to the onshore thrust belt. There is thus no substantive evidence for subduction SW of New Caledonia and nothing in the data published to date that conflicts with the hypothesized presence of a former subduction trace at the base of the ophiolite, with the Loyalty Islands as the former volcanic arc.

Ophiolites in New Guinea Attention has often been drawn to geological similarities that exist between New Caledonia and New Guinea (e.g. Lillie & Brothers 1970), even though the two islands are now separated by more than 1500km of open ocean. They would have been much closer to each other prior to the opening of the Tasman Sea and Coral Sea ocean basins in the Paleocene and Eocene respectively (Weissel & Watts 1979).

The geology of central New Guinea, i.e. of the part of New Guinea between 135°E and 145°E (Fig. 1), can be summarized in terms of four major units (see Dow 1977). In the south, a foreland basin underlies the plains of the Western and Gulf districts. The Central Ranges further north, which form the core of the New Guinea Mobile Belt (NGMB), reach heights of more than 5000m before dropping steeply to the broad alluvial Lake and Sepik Plains (Fig. 1) and Ramu and Markham valleys (Fig. 4). Beyond the plains, the north coast of the island is flanked by a mountain chain, which, although narrow and discontinuous, reaches heights of more than 4000 m above sea level in the Finisterre Range (Fig. 4). Ophiolites emplaced during the Paleogene occur in a fragmented belt along the northern edge of the NGMB (Fig. 1), and Paleogene volcanic rocks are important elements in the coast ranges. In the case of the Irian Jaya Ophiolite in the Indonesian province of Papua and the April Ultramafics in western Papua New Guinea, a forearc-volcanic arc linkage to the coast ranges has already been proposed by several workers (e.g. Hill et al. 1993). The collision of the entire allochthonous complex with the New Guinea margin would have taken a considerable time. There might well have been an interval of more than 1 Ma between 'soft' and 'hard' collisions at any one point, and a probability of diachronous collision along the entire belt over a period of 10 Ma or more. The setting of the Papuan Ultramafic Belt (PUB), which is the easternmost, largest and most intensively studied of the New Guinea Ophiolites, is rather different. The belt crops out over large

FOREARC OPHIOLITES, WESTERN PACIFIC

511

Fig. 4. NE Papua New Guinea. Black triangles indicate active and Recent volcanoes of the Bismarck volcanic arc. Bold dashed line shows location of profile of Figure 2c.

areas on the Papuan Peninsula east of 147°E (Figs 1 and 4), and forearc characteristics have been identified by Pearce et al. (1984), amongst others. Outcrops of boninitic lavas on Cape Vogel (Dallwitz et al. 1966) are particularly persuasive. Crystallisation dates on gabbros are Jurassic or Cretaceous but tonalite intrusions in the ophiolite are Paleogene (Davies 1971), as are the boninites and tholeiites on Cape Vogel (Smith & Davies 1971). The complete ophiolite sequence was thrust against and over the metamorphic rocks of the Owen Stanley Ranges in the Late Eocene or Oligocene (Davies & Jaques 1984) and is associated with an impressive positive Bouguer anomaly (Milsom 1973). Between 147°E and 148°E the regions 100-150 km to the NE of PUB gravity high, i.e. the regions where, on the Bonin-New Caledonia model, the associated magmatic arc should be found, are occupied by the oceanic Solomon Sea. Further east there is a more extended shelf, bounded by the Lusancay Rise (Fig. 4), which could conceal the rocks of a former volcanic arc, but onshore in the eastern region the thrust separating the PUB from the underlying metamorphic rocks is subhorizontal (Davies 1971). Also, the gravity peak, which in the west is

within the area of ophiolite exposure, moves steadily northwards and eventually out to sea in the east (Milsom 1973). Thus both geological and gravity evidence suggests that the Luscancay Rise is largely underlain by ophiolite. The Solomon Sea floor is now being subducted at the New Britain Trench (Fig. 4), beyond which lies New Britain itself and the subduction-related active volcanoes of the Bismarck Arc. The New Britain 'terrane' is extended by most workers (e.g. Hamilton 1979) to include the Adelbert and Finisterre Ranges of the mainland, as there are similarities both in the basement rocks and in the present-day tectonic setting in relation to the Bismarck volcanic arc (Fig. 4), which continues just north of the New Guinea coast as far as 145°E. The Finisterre Range in particular has attracted much recent attention as an example of very recent arc-continent collision (Abbott et al. 1994), a conclusion supported by continuing seismic activity associated with Benioff Zones that dip both north and south from the interpreted collision trace (Fig. 5) and mark the site of a now subducted ocean floor (Pegler et al. 1995). South of the Adelbert Range the Marum Complex (Figs 1 and 4), an ophiolite similar to, but much smaller

512

J. MILSOM

Fig. 5. The double subduction zone in NE Papua New Guinea, after Pegler et al. (1995). Rectangle shows swathe location.

than, the PUB, has been emplaced from the north along the front of the Central Ranges during the Paleogene (Jaques 1981). As with the PUB, the main mass of the Marum ophiolite is at least as old as Late Cretaceous, and emplacement occurred at some time between the Mid-Eocene and the early Miocene (Davies & Jaques 1984). Most of the recent palaeogeographical reconstructions of the New Guinea region (e.g. Hill et al. 1993; Struckmeyer et al. 1993) show the New Britain-Finisterre-Adelbert (NBFA) Terrane as originating far out in the Pacific and arriving at the New Guinea margin only within the last few million years. There are, however, aspects of the geology that are not consistent with this interpretation. Findlay et al. (1997) noted that coarse continentally derived conglomerates of the Sukurum Formation interfinger with the Paleogene Finisterre Volcanics but are very clearly not derived from them. They pointed out that, whereas it is hard to see how such sediments could have been deposited within a volcanic belt isolated

within an oceanic setting, suitable provenance areas can readily be identified along the former northern margin of Australasia, now exposed in central New Guinea. Moreover, the Finisterre Volcanics and the equivalent rocks of New Britain and the Adelberts have many features in common with the igneous basement of the north coast ranges further west. The latter have remained in contact with the New Guinea margin ever since their arrival there in the Eocene, at a time when the reconstructions cited above suggest that the NBFA Terrane was thousands of kilometres away. The fact that New Britain is now advancing across the Solomon Sea towards the PUB is also not as clear evidence for a far northern or eastern origin of its basement rocks as might at first sight appear. There has been no deep drilling in the oceanic basin, but various lines of evidence summarized by Honza et al. (1987) point to an early Miocene age. If this is correct, then either there was a second, now entirely subducted, ocean in the region or else New Britain lay alongside the

FOREARC OPHIOLITES, WESTERN PACIFIC

513

Fig. 6. Schematic illustration of phases in the development of the Marum Complex and the Papuan Ultramafic Belt. These may have been replicated, with variations, in the history of other SSZ ophiolites. (a) Approach. A typical passive margin converges on a 'Bonin-type' forearc consisting largely of autochthonous igneous rocks, with little material accreted from the lower plate, (b) Collision. The continent and the arc are juxtaposed and both are thickened by thrusting. A part of the forearc crust and underlying mantle is emplaced on the former continental margin. The present-day situation north of the Finisterre and Adelbert ranges suggests that arc-volcanic activity may continue for some time after collision, allowing interfmgering of volcanic rocks with sediments derived from the continental part of the orogen (e.g. Sukurum Formation, NE New Guinea), (c) Separation. Post-orogenic collapse reaches the point at which new ocean crust is developed in the vicinity of a line of weakness within the orogen and close to the original suture. In a small basin the spreading axis is unlikely to be associated with a topographic high that is sufficient to impede later subduction (compare the Woodlark Basin, eastern Papua New Guinea), (d) Return (as in the present-day Solomon Sea region). The ocean basin shrinks in response to subduction at its margins. The Solomon Sea has undergone subduction at each of its three margins, but such a complicated pattern seems unlikely to have been general in the destruction of all small oceans formed by post-orogenic collapse. It should be noted that the new subduction complexes are not Bonin-type. In the case of New Britain, the forearc now thrusting over the Solomon Trench consists largely of the igneous basement of the old volcanic arc. (e) Reunion (as in the present-day MarumFinisterre region). The ocean basin has been eliminated and the former volcanic arc and forearc have been restored to roughly their previous relative positions, but are separated by an accretionary complex rather than by a forearc basin.

514

J. MILSOM

PUB immediately before the Miocene. If the Adelberts and Finisterres were at this time similarly positioned with respect to the Marum Complex, by then already thrust over the Central Ranges, then the continentally derived Paleogene conglomerates are explained and an extended volcanic arc is identified to which the Marum and PUB could have been forearcs.

Conclusions Suprasubduction zone origins have been suggested, and disputed, for many large ophiolite bodies, including those in New Caledonia and eastern Papua New Guinea. Such origins imply the former existence of associated volcanic arcs, and in New Caledonia the Loyalty Islands are acceptable candidates for the role. The equivalent region NE of the Papuan Ultramafic Belt is, however, occupied by the oceanic crust of the Solomon Sea, while the much smaller but similar and probably related Marum Complex lies to the south of a site of very recent arc-continent collision. The absence of an obvious volcanic arc associated with these 'forearc' ophiolites can be explained if it is supposed that the Adelbert and Finisterre terranes share with New Britain a complex history involving Paleogene collision with a northern New Guinea passive margin, rifting away from that margin with the formation of a small ocean basin (Solomon Sea), and a subsequent resumption of convergence and collision (Fig. 6). During the rifting phase, the volcanic arc of the collided terrane (the NBFA Terrane) is assumed to have separated from the forearc, which remained attached to the New Guinea margin as the PUB and Marum complexes. Following rifting and separation, a second phase of convergence placed the Paleogene NBFA Terrane in a forearc position. The Adelberts and Finisterres have only recently arrived north of the Marum, and New Britain is now converging on the PUB.

Discussion The hypothesis outlined in this paper can be regarded merely as providing a possible solution to a purely local problem in SW Pacific geology. However, similar solutions may be more widely applicable. Collision between a non-accretionary (Bonintype) island arc and a continental margin may result in ophiolite emplacement. It will also, inevitably, lead to crustal thickening and mountain building. High mountains are inherently gravitationally unstable, and numerous recent papers (e.g. Jones et al 1996; Stuwe & Barr 2000, and references therein) have explored the consequences

of this instability, often quantified in terms of the excess potential energy stored in the thickened (and sometimes thinned) lithosphere. Gravitationally driven extension may not only return the land surface to sea level but may continue to, and even beyond, the point at which new oceanic crust is formed. Extension is likely to be localized within zones of weakness in the orogen, and the appropriate conditions may well exist in the vicinity of the former inter-arc gap (forearc basin). This seems, at any rate, to have been the case in eastern New Guinea. The new ocean may permanently separate the two components of the former arc or it may, as in eastern New Guinea, contract again after a relatively brief spreading phase. Even where such contraction leads to renewed juxtaposition of former volcanic arc and forearc elements, these are likely to be so modified that their original close relationship is obscured. Rifting and the generation of a new ocean basin are clearly not inevitable consequences of collision. They evidently did not occur to any significant extent west of the Marum, as there appears to be a much more direct relationship between the Irian Jaya Ophiolite Belt-April Ultramafics and the volcanic basements of the north coast ranges. They may, none the less, be common. If so, an explanation is provided for the difficulties often encountered in identifying volcanic arcs with which supposedly 'forearc' ophiolites can plausibly be associated. To the cycle of 'birth, death and resurrection' recently proposed by Shervais (2001) to describe the evolution of suprasubduction zone ophiolites, it seems necessary to add the possibility, and perhaps even probability, of divorce and remarriage.

References ABBOTT, L.D., SILVER, E.A. & GALEWSKY, J. 1994. Structural evolution of a modern are-continent collision in Papua New Guinea. Tectonics, 13, 1007-1034. AUZENDE, J.-M., VAN DE BEUQUE, S., REGNIER, M., LAFOY, Y. & SYMONDS, P. 2000. Origin of the New Caledonian ophiolites based on a French-Australian seismic transect. Marine Geology, 162, 225-236. BITOUN, G. & RECY, J. 1982. Origine et evolution du Bassin des Loyautes et de ses bordures apres la mise en place de la serie ophiolitique de Nouvelle Caledonie. Travuax et Documents ORSTOM, 147, 505-539. BLOOMER, S.H., TAYLOR, B., MACLEOD, C.J., STERN, R.J., FRYER, P., HAWKINS, J.W. & JOHNSON, L. 1995. Early arc volcanism and the ophiolite problem: a perspective from drilling in the Western Pacific. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western

FOREARC OPHIOLITES, WESTERN PACIFIC Pacific. Geophysical Monograph 88. American Geophysical Union, Washington, DC, 1-30. COLLOT, J.Y., MALAHOFF, A., RECY, J., LATHAM, G. & MISSEGUE, F. 1987. Overthrust emplacement of New Caledonia ophiolite: geophysical evidence. Tectonics, 6, 215-232. DALLWITZ, W.B., GREEN, D.H. & THOMPSON, I.E. 1966. Clinoenstatite in a volcanic rock from the Cape Vogel area, Papua. Journal of Petrology, 7, 375-403. DA VIES, H.L. 1971. Peridotite-Gabbro-Basalt Complex in Eastern Papua: an Overthrust Plate of Oceanic Material and Crust. Australian Bureau of Mineral Resources Bulletin, 128. DAVIES, H.L. & JACQUES, A.L. 1984. Emplacement of ophiolite in Papua New Guinea. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 341-349. Dow, D.B. 1977. A Geological Synthesis of Papua New Guinea. Australian Bureau of Mineral Resources Bulletin, 201. FINDLAY, R.H., ARUMBA, J., KAGI, J., NEKITEL, S., Mosusu, N., RANGIN, C. & PUBELLIER, M. 1997. Revision of the Markham 1:250,000 sheet, Papua New Guinea: What is the Finisterre Terrane? In: HANCOCK, G.E. (ed.) Proceedings of the Papua New Guinea Geology Mining and Exploration Conference. Australasian Institute of Mining and Metallurgy, Port Moresby, PNG, 87-97. GASS, I.G. & MASSON-SMITH, D. 1963. Geology and gravity anomalies of the Troodos massif, Cyprus. Philosophical Transactions of the Royal Society of London, Series A, 255, 417-467. GEALEY, W.K. 1980. Ophiolite obduction mechanisms. In: PANAYIOTOU, A. (ed.) Ophiolites. Geological Survey of Cyprus, Nicosia, 228-243. HAMILTON, W. 1979. Tectonics of the Indonesian Region. Special Publication 1078, US Geological Survey, Denver, CO. HILL, K.C., GREY, A., FOSTER, D. & BARRETT, R. 1993. An alternative model for the Oligo-Miocene evolution of northern PNG and the Sepik-Ramu Basins. In: CARMAN, GJ. & CARMAN, Z. (eds) Second PNG Petroleum Convention, Petroleum Exploration and Development in Papua New Guinea. Australasian Institute of Mining and Metallurgy, Port Moresby, PNG, 241-259. HONZA, E. & TAMAKI, K. 1985. The Bonin Arc. In: NAIRN, A.E.M., STEHLI, GF.G. & UYEDA, S. (eds) The Ocean Basins and Margins, Volume 7A, The Pacific Ocean. Plenum Press, New York, 459-502. HONZA, E., DAVIES, H.L., KEENE, J.B. & TIFFIN, D.L. 1987. Plate boundaries and evolution of the Solomon Sea region. Geo-Marine Letters, 7, 161-168. JAQUES, A.L. 1981. Petrology and petrogenesis of cumulate peridotites and gabbros from the Marum ophiolite complex, northern Papua New Guinea. Journal of Petrology, 22, 1-40. JONES, C.H., UNRUH, J.R. & SONDER, L.J. 1996. The roles of gravitational potential energy in active

515

deformation in the southwestern United States. Nature, 381, 37-41. LILLIE, A.R. & BROTHERS, R.N. 1970. The geology of New Caledonia. New Zealand Journal of Geology and Geophysics, 13, 145-183. MANGHNANI, M.H. & COLEMAN, R.G. 1981. Gravity profiles across the Samail ophiolite, Oman. Journal of Geophysical Research, 86, 2509-2525. MILSOM, J. 1973. Papuan Ultramafic Belt: gravity anomalies and the emplacement of Ophiolites. Geological Society of America Bulletin, 84, 2243-2258. MILSOM, J. 1984. The gravity field of the Marum ophiolite complex, Papua New Guinea. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 351-357. MILSOM, J., HALL, R. & PADMAWIDJAJA, T. 1996. Gravity fields in eastern Halmahera and the Bonin Arc: implications for ophiolite origin and emplacement. Tectonics, 15, 84-93. PEARCE, J.A., LIPPARD, S.J. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PEGLER, G., DAS, S. & WOODHOUSE, J.H. 1995. A seismological study of the eastern New Guinea and western Solomon Sea regions and its tectonic implications. Geophysical Journal International, 122,961-981. RAVAUT, P., BAYER, R., HASSANI, R., ROUSSET, D. & AL YAHYA'EY, A. 1997. Structure and evolution of the northern Oman margin: gravity and seismic constraints over the Zagros-Makran-Oman collision zone. Tectonophysics, 279, 253-280. SMITH, I.E. & DAVIES, H.L. 1971. Geology of the Southeast Papuan Mainland. Australian Bureau of Mineral Resources Bulletin, 165. SEARLE, M.P. & Cox, J. 1999. Tectonic setting, origin, and obduction of the Oman ophiolites. Geological Society of America Bulletin, III, 104-122. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, No. 2000GC000080. STRUCKMEYER, H.I.M., YEUNG, M. & PIGRAM, CJ. 1993. Mesozoic to Cainozoic plate tectonic and palaeogeographic evolution of the New Guinea Region. In: CARMAN, G.J. & CARMAN, Z. (eds) 2nd PNG Petroleum Convention, Petroleum Exploration and Development in Papua New Guinea. Australasian Institute of Mining and Metallurgy, Port Moresby, PNG, 261-290. STUWE, K. & BARR, T.D. 2000. On the relationship between surface uplift and gravitational extension. Tectonics, 19, 1056-1064. WEISSEL, J.K. & WATTS, A.B. 1979. Tectonic evolution of the Coral Sea basin. Journal of Geophysical Research, 84, 4572-4582.

This page intentionally left blank

Tethyan- and Cordilleran-type ophiolites of eastern Australia: implications for the evolution of the Tasmanides CATHERINE V. SPAGGIARI 1 ' 4 , DAVID R. GRAY 2 & DAVID A. FOSTER 3 1

School of Geosciences, PO Box 28E, Monash University, Clayton, Vic. 3800, Australia (e-mail: catherine@mail. earth, monash. edu. au) 2 School of Earth Sciences, University of Melbourne, Melbourne, Vic. 3010, Australia ^Department of Geological Sciences, University of Florida, Gainesville, FL 32611, USA ^Present address: Department of Applied Geology (Tectonics Special Research Centre), Curtin University, Bentley, WA 6102, Australia (e-mail: [email protected]) Abstract: The preservation of diverse ophiolitic rocks in the Tasmanides of eastern Australia reflects a period of complex oceanic crust formation off the east Gondwana margin from c. 560 to 495 Ma. This involved development of oceanic crust possibly at a spreading ridge (now preserved as the c. 560 Ma Marlborough ophiolite, New England Orogen), development of forearc crust in a suprasubduction zone between c. 530 and 515 Ma (Tyennan-Delamerian ophiolites), possible synchronous formation of suprasubduction zone crust further outboard (southern New England Orogen ophiolites), emplacement of Tyennan-Delamerian ophiolites onto the Gondwana margin at c. 515 Ma, followed by development of new forearc and backarc crust between c. 505 and 495 Ma (Lachlan Orogen). Tasmanide orogenesis has resulted in the production of both Tethyan-type and Cordilleran-type ophiolites. Tethyan types are represented by the Tyennan (Tasmanian) and possibly Delamerian Orogen ophiolites that were emplaced onto the passive continental margin. The Tasmanian ophiolite has been interpreted as having been emplaced as one or more thrust sheets, concurrent with the development of a metamorphic sole. In contrast, Lachlan Orogen ophiolites were emplaced within major fault zones by accretionary processes such as offscraping, underplating, and duplexing via underthrusting and basin closure during Late Ordovician to Silurian times. Cordilleran-type ophiolites, for example the Lachlan Orogen ophiolites, are dominated by disrupted upper oceanic crustal stratigraphies, whereas Tethyan-types tend to be dominated by mantle and gabbroic sequences, or preserve more complete stratigraphies.

Ophiolitic rocks of Neoproterozoic to Cambrian age occur throughout the Tasmanides, eastern Australia, in three north-south-trending Palaeozoic orogenic belts amalgamated to the Australian Craton (Fig. 1). These accretionary orogenic belts are distinguished by differences in their lithotectonic assemblages, and timing of orogenesis (Coney et al. 1990; Foster & Gray 2000). The majority of recognized ophiolitic assemblages appear to have formed in suprasubduction zone environments, in a long-lived, complex SW or Western Pacific style setting that evolved along the eastern Gondwana margin (e.g. Crawford & Keays 1987; Foster & Gray 2000). Their development reflects a period of major plate reorganization and tectonic events of Pan-African age, during assembly of the Gondwana supercontinent (Powell et al. 1993). Ophiolites of eastern Australia consist of both 'Tethyan type' (e.g. the Semail ophiolite of Oman) and 'Cordilleran type' (e.g. many Pacific-rim

ophiolites), where the distinction is made depending on their emplacement history and whether there is a clear structural relationship to continental basement (Moores 1982; Moores et al. 2000). The intention here is not to specifically classify these ophiolites, as both types may not be mutually exclusive, but to highlight the differences in emplacement styles and tectonic settings. Ophiolites now preserved near the western margin of the Tasmanides (Delamerian and Tyennan occurrences) are Tethyan type and were emplaced onto the Gondwana passive margin during the Early Cambrian (c. 515 Ma) (Berry & Crawford 1988; Meffre et al. 2000). In contrast, ophiolites in the central Tasmanides (Lachlan Orogen) are preserved in the frontal portions of large turbidite wedges, and were accreted and duplexed during underthrusting in a backarc setting during Late Ordovician to Silurian times (Gray & Foster 1998; Foster et al. 1999; Spaggiari et al. 2002a). Ophio-

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 517-539. 0305-8719/037$ 15 © The Geological Society of London 2003.

518

C. V. SPAGGIARI ETAL.

Fig. 1. Map of the major geological elements of Australia. Ophiolites occur within the Lachlan, New England, and Delamerian Orogens of the Tasmanides. (After Foster & Gray 2000.)

lites in the eastern Tasmanides (New England Orogen) have complex origin and emplacement histories, partly obscured by terrane or tectonic element accretion events that are less understood (e.g. Holcombe et al 1997a, 1997b). This paper provides an overview of ophiolites that occur throughout the Tasmanides of eastern Australia, and summarizes their geochemistry, rock associations, geochronology, and structural and metamorphic characteristics. It also includes new geochronological data that constrain the age of oceanic crustal development in the Lachlan Orogen, southeastern Australia (Fig. 1).

Delamerian-Tyennan ophiolites The Delamerian Orogen constitutes the southwestern part of the Tasmanides adjacent to the Proterozoic Archaean Australian Craton. It comprises a 550-530 Ma Jura-style, west-vergent, foreland fold and thrust belt in the west, and a polydeformed, metamorphic high-TMow-P complex in the east (Foster & Gray 2000, and references therein). Major orogenesis occurred

between 520 and 490 Ma following rifting and deposition of the c. 532-526 Ma Kanmantoo Group (Priess 1995). The Delamerian Orogen was once continuous with the Ross Orogen of Antarctica.

Tyennan (Tasmanian) Orogen occurrences Ophiolitic rocks in Tasmania are of Tethyan type and occur as allochthonous complexes that were emplaced onto the Late Neoproterozoic passive margin of east Gondwana during continentisland-arc collision at c. 515 Ma (Fig. 2) (Berry & Crawford 1988; Crawford & Berry 1992; Meffre et al. 2000). They are interpreted to have thrust sheet-like form, and have undergone disruption during post-emplacement deformation, analogous to the ophiolites of Oman and Newfoundland (Berry & Crawford 1988). A metamorphic sole at the base of the thrust sheet is characterized by magnesio-hornblende and plagioclase mylonites that have shear bands and rotated porphyroclasts indicative of transport to the west (Berry & Crawford 1988). The sole formed in peridotite and

OPHIOLITES OF EASTERN AUSTRALIA

519

Fig. 2. Map of western Tasmania, showing major structural and lithotectonic elements, locations of ophiolite complexes and metamorphic complexes. (After Meffre et al. 2000.)

the presence of metamorphic olivine, amphibole and hercynite spinel is indicative of high temperatures of formation (Meffre et al. 2000). Metamorphic soles are inferred to have formed as a result of overriding of hot mantle sequences either over the top of ophiolitic pelagic rocks and basalt or the underlying (usually passive) margin basement, producing strongly deformed metamorphic aureoles in these sequences (e.g. Bay of Islands Complex, Newfoundland, Fergusson & Cawood 1995; Semail ophiolite, Oman, Hacker et al. 1996; Tauride tectonic belt, Turkey, Dilek et al. 1999). The Tasmanian ophiolite differs in this respect in that the high-T shear zone formed in the mantle sequence. The shear zone is interpreted to relate to thrusting early in the emplacement history (Berry & Crawford 1988).

The mafic-ultramafic complexes consist of orthopyroxene-rich cumulates and very low-Ti boninitic lavas, interpreted to have formed in the forearc region of an island arc. The following description is from Crawford and Berry (1992) and references therein. The cumulates comprise three types, with the most distinctive being layered dunite-harzburgite found in the large Heazlewood River Complex, and also the Mount Stewart, Wilson River, and Adamsfield Complexes (Fig. 2). Olivine and orthopyroxene in the cumulates are rich in magnesium (Mg number 92-94), and chrome spinels are highly refractory (Cr number 87-94). In the Serpentine Hill, Heazlewood River, and Mclvor River Complexes, a layered peridotite-pyroxenite-gabbro succession is present. The same succession also occurs in the Adamsfield,

520

C. V. SPAGGIARI ETAL.

Huskisson River, and Wilson River Complexes with rafts of a layered pyroxenite-dunite succession. Both are dominated by orthopyroxene, and have similar cumulate mineral chemistry to the layered dunite-harzburgite succession, but greater CaO and A^Os contents. They are also slightly less refractory, with olivine and orthopyroxene Mg number in the range 85-90, and chromite with Cr number in the range 52-80. Lavas in the mafic-ultramafic complexes consist of lower greenschist-grade, pillowed or massive, low-Ti basalts, and boninites. The boninites are vesicular with phenocrysts of clinoenstatite, orthopyroxene, chromite, and possibly olivine in a glassy matrix. The chromites have Cr number of 85-93. Based on field relationships and geochemistry, the boninites are interpreted as extrusive equivalents of the layered dunite-harzburgite cumulate, and the basalts as counterparts of the layered peridotite-pyroxenite-gabbro and pyroxenite-dunite successions. The basalts are less magnesian than the boninites and more titanium-rich. They differ from typical mid-ocean ridge basalt (MORB) by their higher silica; lower Ti, Nb, and Zr contents; Ti/Zr ratios of 150-230 compared with 70-130 for MORB; lower Zr/Y, Zr/Sc, and Ti/V ratios; and depleted rare earth element (REE) patterns. A late-stage tonalite from the ophiolite sequence in the Heazlewood River Complex has been dated (sensitive high-resolution ion microprobe (SHRIMP), U-Pb in zircon) at 514 ± 5 Ma (Black et al. 1997). High-pressure rocks in the Franklin and Forth metamorphic complexes formed at c. 515 Ma and provide a minimum age for the ophiolite, as these metamorphic rocks formed during subduction of the passive margin just prior to or during ophiolite emplacement (Meffre et al. 2000). Calc-alkaline volcanic rocks (Mt Read Complex, Fig. 2) erupted at c. 500 Ma, during post-collisional extension (Crawford & Berry 1992; Perkins & Walshe 1993). These volcanic rocks interfinger with fossiliferous (Boomerangian c. 500 Ma) sandstone that contains detritus from the ophiolite (Crawford & Berry 1992).

Delamerian Orogen occurrences Mafic-ultramafic rocks occur along the eastern margin of the Delamerian Orogen, in western Victoria (Dimboola subzone) (Fig. 3). Outcrop is extremely poor and identification is mostly from drill-hole data (Maher et al. 1997). A large, NWSE-trending magnetic high corresponds to these occurrences, and their petrological and geochemical characteristics indicate that they may be part of the same forearc-derived ophiolitic sequence

preserved in western Tasmania (Maher et al. 1997). Greenschist-facies, ultramafic rocks consist of serpentinized pyroxenite-wehrlite and harzburgite with Cr-rich, highly refractory spinels. Upper greenschist-facies basalt and gabbro have also been intersected. Leucogabbro from the Ozenkadnook subzone (just west of the Dimboola subzone, Fig. 3) has been dated (SHRIMP, U-Pb in zircon) at 524 ± 9 Ma (Maher et al. 1997). Slivers of serpentinized ultramafic rocks (e.g. Hummocks serpentinite) occur c. 70 km east of the Dimboola subzone and are possible correlates (Morand et al. 2001). These occur within greenschist- or amphibolite-grade schists, and some have an alkalic composition. Fault-bounded, calc-alkaline volcanic rocks (Mt Stavely and Mt Dryden Belts) occur on the eastern margin of the orogen and share petrological and geochemical characteristics with the Mt Read volcanic rocks in Tasmania (Figs 2 and 3) (Crawford et al. 1996). Mt Dryden Belt rocks consist of medium-K, low-Ti andesites and dacites, and Mt Stavely Belt rocks consist of medium- to high-K andesite, dacite, rhyolite, and volcaniclastic rocks, plus small tonalitic intrusions and larger, massive diorite and granodiorite plutons (Crawford 1988; Crawford et al. 1996; Foster et al 1998). A sliver of sheared serpentinite (Williamsons Road Serpentinite) occurs on the eastern margin of the Mt Stavely Belt and may be part of the Dimboola subzone sequence (Maher et al. 1997). Dacite from the Mt Stavely Belt has been dated at 495 ± 5 Ma (U-Pb in zircon) and 500 ± 2 Ma (Ar/Ar in hornblende), and is therefore approximately the same age as the Mt Read volcanic rocks (Crawford et al. 1996; Foster et al 1998). Given the similarities in the evolution of this margin, it is possible that the two were broadly coeval and erupted during an extensional phase following ophiolite emplacement (see Crawford et al 1996).

Lachlan Orogen ophiolites The Lachlan Orogen occupies the central portion of the Tasmanides and consists of a Middle to Upper Cambrian, oceanic backarc-forearc crustal sequence conformably overlain by cherts and voluminous, deep-marine, quartz-rich turbidites. These rocks have been intruded by granitoids, and overlain by younger, continental cover rocks (Figs 1 and 3) (Foster & Gray 2000, and references therein). The Lachlan Orogen is divided into the western, central, and eastern provinces (Fig. 3) (Gray 1997). In the western and central provinces, the basal oceanic and turbiditic sequence has undergone marked structural thickening and shortening during Late Ordovician to Devonian orogen-

OPHIOLITES OF EASTERN AUSTRALIA

521

Fig. 3. Map of the Lachlan Orogen, and eastern Delamerian Orogen showing major structural and lithotectonic elements, western, central, and eastern provinces, and locations of ophiolite occurrences (see text for explanation). MFZ, Moyston Fault Zone; HFZ, Heathcote Fault Zone; GFZ, Governor Fault Zone; MWFZ, Mount Wellington Fault Zone; BBZ, Bendigo-Ballarat structural zone; MZ, Melbourne Zone; TZ, Tabberabbera structural zone. (Modified from Gray & Foster 1998.)

esis (Gray 1997; Gray & Foster 1998). Extension has been prominent in the eastern province, with voluminous granitoid intrusion and basin development as a result of roll-back of a long-lived subduction zone to the east (Foster et al. 1999; Foster & Gray 2000; Collins 2002). The ophiolites have suprasubduction zone affinities and are mostly preserved in major fault zones in the central and western provinces. These fault zones mark the boundaries between structural zones, and are interpreted to relate to a complex evolution of microplate interaction (Fig. 3) (Gray 1997; Foster etal 1999).

Heathcote Belt The Heathcote Belt (Fig. 3) is divided into northern, central, and southern segments based on

structural and lithological differences (Fig. 4). It is bounded to the east by a major west-dipping fault (the Mt William Fault) interpreted to have listric geometry and a complex history (Gray & Willman 1991). Most of the western margin is bounded by the west-dipping Heathcote Fault, where strongly deformed Lower Ordovician turbidites in the hanging wall mark a zone of high strain. These features, and the structural character of the Bendigo-Ballarat structural zone to the west, indicate that the ophiolitic section between the bounding faults (Heathcote Fault Zone) is the basal and frontal part of an imbricate fan thrust system (Gray & Willman 1991; Gray & Foster 1998). Along the western margin of the southern segment a conformable sequence of basalt, chert-shalevolcanogenic sandstone, and quartz-rich turbidite is preserved (VandenBerg 1991). The northern and

522

C. V. SPAGGIARI ETAL.

Fig. 4. (a) Simplified map of the northern and central segments of the Heathcote Belt. (Modified from Wohlt & Edwards 1999.) (b) Map showing outcrop extent of the Cambrian rocks in the Heathcote Belt. (Modified from VandenBerg et al. 2000.)

OPHIOLITES OF EASTERN AUSTRALIA southern sections both consist predominantly of tholeiitic basalt and dolerite, and Middle to Upper Cambrian sedimentary rocks (Crawford 1988; Edwards et al 1998; VandenBerg et al. 2000). The basalts are both pillowed and massive, and the dolerites occur predominantly as sills (Crawford 1988). The central segment consists of fault slivers or duplexes of Cambrian sedimentary rocks, andesite, boninite, diorite, and Ordovician turbidite and black shale, in part underlain by serpentinitematrix melange that includes blocks of variably deformed and metamorphosed boninite, andesite, dolerite, ultramafic rocks, chert, and volcanogenic sandstone (Gray & Willman 1991; Spaggiari et al. 2002b). Geophysical and drill-core data indicate that the melange may also underlie the tholeiitic section in the northern segment (Spaggiari et al. 2002a). The structurally higher, tholeiitic basaltdolerite sequence has been metamorphosed to prehnite-pumpellyite facies whereas the melange rocks show variable grade up to greenschist and blueschist facies (Spaggiari et al. 2002a). The boninitic lavas are either pillowed or massive, mostly fine-grained, amygdaloidal, and have wellpreserved quench textures. Cambrian sedimentary rocks comprise red, yellow, and grey chert, black silicified shale, chert breccia, volcanogenic sandstone, volcaniclastic rocks, and minor tuff (Edwards et al. 1998, and references therein). These contain a Mid- to Late Cambrian fauna, with one exception of possible late Early Cambrian fossils in tuff interbedded with volcanogenic sandstone and andesite, but these fauna are in need of revision (Crawford 1988; VandenBerg et al. 2000). Zircons from a litharenite, interpreted as tuffaceous, which forms part of the conformable sequence in the southern segment of the belt, have been dated (SHRIMP, U-Pb) at 503 ± 8 Ma (I. Williams, pers. comm. 2002). This places a minimum constraint of 495 Ma on the underlying basalt. The geochemistry and petrology of the Heathcote ophiolitic rocks indicates input from a subducting slab component and is consistent with their formation in a suprasubduction zone setting (Table 1) (e.g. Crawford & Keays 1987). The andesite has a boninitic composition and is interpreted as being derived from the same refractory mantle source as the Type B boninites, but by lower degrees of partial melting (Nelson et al. 1984; Crawford & Cameron 1985). The tholeiitic rocks have geochemical characteristics transitional between MORB and arc tholeiite, and are interpreted to have intruded the boninitic series as basin development progressed (Crawford & Keays 1987). Relative to the Mount Wellington Belt tholeiites (see below), the Heathcote Belt tho-

523

leiites show a very rapid increase in Ti, Zr, Y, and La with increasing differentiation, suggestive of higher eruption and magma supply rates (Crawford & Keays 1987).

Mount Wellington Belt The Mount Wellington Belt (Fig. 3) comprises segments of Middle to Upper Cambrian, suprasubduction zone ophiolitic rocks, similar to those in the Heathcote Belt, at Dookie, Tatong, Howqua, and Dolodrook (Fig. 5). These sequences are part of the Governor Fault Zone, a structurally complex boundary to the SW-vergent Tabberabbera structural zone that forms the western margin of the central province of the Lachlan Orogen (Fig. 3). The Tabberabbera Zone is linked to a high-Jlow-P metamorphic complex to the east (WaggaOmeo Complex) (Gray 1997). The Barkly River Belt comprises calc-alkaline, predominantly andesitic rocks preserved as fault slices in the adjacent Mount Wellington Fault Zone that in part has overthrust the sequences in the Governor Fault Zone (Crawford 1988; Spaggiari et al. 2002b). The boundary between the western and central provinces of the Lachlan Orogen occurs between the Mount Wellington Fault Zone and the Governor Fault Zone (Fig. 3). The geochemistry and petrology of ophiolitic rocks from the Mount Wellington Belt are consistent with their formation in a suprasubduction zone setting (Table 2). As at Heathcote, the tholeiitic rocks show MORB to arc tholeiite characteristics, and strong Fe-enrichment trends (Crawford & Keays 1987). Boninitic lavas are similar to those in the Heathcote Belt, but include more refractory types and none of andesitic composition (Crawford 1988). Chromite chemistry in peridotite at Dolodrook suggests that it was derived from the same refractory mantle source as the boninitic and tholeiitic rocks in the rest of the belt (Crawford 1988). Dookie. The Dookie segment is the most northern exposure of the belt (Fig. 3); however, aeromagnetic signatures indicate that the belt continues under cover to the NW, and truncates the Heathcote Belt. Unlike the rest of the belt, the exposed part of the Cambrian sequence at Dookie trends east-west and is bounded to the south by a north-dipping thrust fault (Dookie Fault) (Fig. 5a) (Tickell 1989). Footwall rocks are mostly SilurianDevonian turbidites of the Melbourne structural zone (Fig. 3) and the fault is probably a late or reactivated structure that does not necessarily relate to ophiolite emplacement. The sequence is dominated by tholeiitic basalt, dolerite and gabbro, including both massive and cumulate types,

C. V. SPAGGIARI ETAL.

524

Table 1. Summary of petrological and geochemical data from the Heathcote Belt Heathcote Belt

Relict primary mineralogy

Tholeiitic series

Major element chemistry

Trace element chemistry

Strong Fe enrichment trends, low K

Initial eNd c. 5 Flat (chondritic) REE patterns Immobile element ratios within the range of depleted MORB LREE-enriched

Basalt Dolerite-gabbro Boninitic series Type A

TypeB

Andesite

Aug-phyric or aphyric, PI, Fe-Ti oxides Aug, PI, Fe-Ti oxides Mostly opx-phyric

High Si, high Mg, lowTi

Low-Ca Pyx-Aug, Cr-Spl

Low-Ca Pyx, Cr-Spl, ±P1

Aug-Opx-Pl-phyric, Fe-Ti oxides

Cr number in chromite c. 0.85

Lower Ti than Type A

High Mg, low Ti

Ti/Zr 63 ± 4 LREE-enriched (La/Yb)N 2-3 HREE 5 X chondrite Slightly concave-up or flat REE patterns Initial eNd +5.8 Cr number in chromite 0.89-0.93 Low Ti, Y Ti/Zr 23 ± 3 (La/Yb)N >5 (greater LREE enrichment and lower HREE levels than Type A) HREE 2-3 X chondrite Initial eNd +3.8 Similar to Type B boninites (higher REE levels) HREE depleted LREE enriched (La/Yb)N 6.4-7.5 Ti/Zr 18 ±2 Low Ti, Y Initial eNd +3.3

Data from Crawford et al. (1984), Nelson et al. (1984), Crawford & Cameron (1985) and Crawford & Keays (1987).

some of which are intrusive into the sedimentary sequence (Glasson 1973, in Crawford 1988). The gabbros include particularly coarse-grained varieties, with plagioclase crystals up to 6 cm. The sedimentary sequence includes volcanogenic sandstone, conglomerate, tuff, and chert, some of which is hematite-rich (Crawford 1988; Tickell 1989). The mafic rocks have been metamorphosed at prehnite-pumpellyite to greenschist facies and show significant hydrothermal alteration. Epidote, calcite, quartz, (± axinite) veins are widespread, and barite veins occur locally. These tholeiites have similar geochemical and petrological characteristics to those from Heathcote and Howqua, and are also interpreted as transitional between MORB and arc tholeiite (Crawford & Keays 1987).

Conventional U-Pb zircon data from a hornblende gabbro in the Dookie complex are given in Table 3 and the results are shown in Figure 6. Of the five analyses three give similar 206pb/238U ages and the other two give ages that are significantly younger. A weighted mean age of the three older zircons gives an age of 502 ± 0.7 Ma, which we interpret to be the magmatic crystallization age of the gabbro. The two younger zircons probably lost minor radiogenic Pb. Tatong. The Tatong segment (Fig. 3) is poorly exposed and difficult to interpret structurally. It is dominated by basalt, dolerite, and minor gabbro similar to rocks at Howqua (see below) and Heathcote (VandenBerg et al. 2000) (Fig. 5b).

OPHIOLITES OF EASTERN AUSTRALIA

525

Fig. 5. (a) Simplified map of the Dookie region of the Mount Wellington Belt, showing the location of the dated gabbro. (Modified from Tickell 1989.) (b) Simplified map of the Tatong region of the Mount Wellington Belt. (Modified from Brown 1998.) (c) Simplified map of the Howqua region of the Mount Wellington Belt, (d) Simplified map of the Dolodrook region of the Mount Wellington Belt.

Table 2. Summary of petrological and geochemical data from Howqua and Dolodrook, Mount Wellington Belt Mount Wellington Belt

Relict primary mineralogy

Howqua Tholeiitic series

Major element chemistry

Trace element chemistry

Strong Fe enrichment trends, low K

Initial eNd +4.5 to + 5.0 Flat (chondritic) REE patterns Immobile element ratios within the range of depleted MORB LREE enriched

Basalt-hyaloclastite Dolerite-gabbro (cumulates) Boninitic series

Pyroxenite Dolodrook Harzburgite dunite

Aug ± Ol-phyric or aphyric, PI, Fe-Ti oxides Aug, PI, Fe-Ti oxides, ± Ol, ± Mg-Hbl Fo-clinoenstatite-phyric, High Si, very high Mg, Cr-Spl low Ti, low Al

Ol, clinoenstatite Ol, rare opx 01

Data from Nelson et al. (1984), Crawford & Keays (1987) and Crawford (1988).

Initial eNd+1.3 to -9.0 Low Zr, Y, very high Ni, Cr Low Ti/Zr LREE-enriched Low REE abundances Boninitic source Cr number in spinel 0.78-0.91 Cr number in spinel (dunite and chromitite) c. 0.58-0.72

C. V. SPAGGIARI ETAL.

526

Doleritic rocks near Samaria show weak development of blue-green amphiboles, indicative of blueschist-greenschist transitional assemblages. These occur adjacent to strongly deformed, veined black slates similar to those in melange at Howqua. Aeromagnetic imagery indicates that the ophiolitic sequence dips to the NE along its western margin. The Tatong sequence may represent a continuation of the Howqua segment, but with significant disruption by Late Devonian volcanism and plutonism that was accompanied by small-scale, intra-continental basin development. Howqua. The Howqua segment (Fig. 3) represents the most complete and best exposed section of ophiolitic rocks in the Mount Wellington Belt (Fig. 5c). The sequence consists of an internally imbricated, NE-younging section of tholeiitic pillow basalt, hyaloclasite, volcaniclastic rocks, dolerite, and gabbro, in fault contact with and underlain by mafic and ultramafic boninitic lavas and intrusive rocks, underlain by melange. The melange contains large fault slivers of both tholeiitic and boninitic rocks, blocks (or knockers) of these metamorphosed up to blueschist facies in talc schist matrix, and slivers of Ordovician phyllite, slate and minor sandstone (Spaggiari et al 2002a, 2002b). The structurally highest, tholeiitic sequence has been metamorphosed to prehnite-pumpellyite facies, and metamorphic pressure and temperature, as well as deformation intensity, increase down sequence into the melange. The metamorphic pattern is interpreted as having formed by accretionary processes during underplating of the melange, accompanied by duplexing of the upper sequences (Spaggiari et al. 2002a). Part of the tholeiitic pillow basalt sequence is conformably overlain by and interbedded with chert and silicified black shale, which in turn is conformably overlain by turbidites. Earliest Ordovician (Lancefieldian, La2 c. 490 Ma) graptolites occur within the upper part of the chert sequence and Latest Cambrian (Datsonian) conodonts occur

Fig. 5. (continued)

Table 3. U—Pb zircon data from gabbro pegmatite from the Dookie quarry and tip site Zircon(s)

Weight (mg)

U (ppm)

Pb (ppm)

1 (4 grains) 2 (1 grain) 3 (1 grain) 4 (1 grain) 5 (1 grain)

0.909 0.194 0.528 0.218 0.094

124 55 44 57 91

20 8 8 10 16

204

Pb/206Pb

0.056 0.0586 0.056 0.0561 0.0561

206

pb/238U

0.08096 0.08033 0.0797 0.08087 0.08106

Error %

0.241 0.174 0.269 0.21 0.256

207

Pb/235U

0.6439 0.6673 0.648 0.6703 0.6517

Error %

0.355 0.537 0.351 0.604 0.677

207

Pb/206Pb

0.05768 0.0602 0.05895 0.0601 0.0583

Error %

0.245 1.05 0.21 0.529 0.587

206

Pb/238U age (Ma)

Error (Ma)

501.87 498.13 494.58 501.41 502.45

1.2 0.8 1.3 1.1 1.3

Conventional isotope dilution thermal ionization mass spectrometry (ID-TIMS) analyses performed at the University of Florida following methods described by Mueller et al. (1988). Age calculated from the ratio corrected for common Pb (Stacey & Kramers 1975) and blank.

OPHIOLITES OF EASTERN AUSTRALIA

527

Fig. 5. (continued)

approximately midway (VandenBerg & Stewart 1992). These fauna provide a minimum age of 491 Ma for the basalts. The chert unit has a maximum thickness of c. 480m (Fergusson 1998) but may be thinner as the extent of structural thickening is not clear. This suggests that the basalts are likely to be approximately the same age as the Dookie gabbro, that is, c. 500 Ma. There are no large mantle sections and, where present, ultramafic rocks are of boninitic composition, including an olivine pyroxenite sill (Crawford

& Keays 1987). The boninitic lavas are mostly very fine-grained, amygdaloidal, often pillowed, and locally interbedded with chert. Within the tholeiitic sequence, gabbroic-doleritic sills with basal cumulates grade upwards to evolved, leucocratic sections (Crawford & Keays 1987). Dolodrook The Dolodrook inlier (Fig. 3) consists of a c. 5 km long, 1 km wide serpentinite body (Fig. 5d). It is completely fault-bounded and has an antiformal geometry, where only the top of the

528

C. V. SPAGGIARI ETAL.

Fig. 5. (continued}

body is exposed. Ultramafic rocks comprise harzburgite, dunite, orthopyroxenite, clinopyroxenite and podiform chromitite (Crawford 1988). A small sliver of very fine-grained pillow basalt is the only mafic rock exposed. The serpentinite is pervasively foliated, consists of lizardite-chrysotile, and metamorphic grade of the sequence, including the surrounding younger chert-shale and turbidite, is no higher than prehnite-pumpellyite facies. Small debris flows of serpenitinized breccia consisting of angular clasts of predominantly serpentinite, fine-grained basalt, and chert, occur throughout the body. Lenses of mafic sandstone, mudstone, gritstone and conglomerate that include large olistostromes of limestone with Mid- to Late Cambrian fauna (Garvey Gully Formation, Mindyallan-Idamean, c. 498-496 Ma; VandenBerg et al. 2000) flank the serpentinite body. This represents a proximal, medium- to high-energy deposit indicative of reef formation close to the serpentinite body, and provides a minimum age constraint. Serpentinization probably occurred during formation of the backarc-forearc crust, indicated by magmatic water signatures in O and H isotope data (average 618O is 7.12, average 6H is -84.75; Spaggiari et al. unpubl. data). Sedimentation patterns suggest that the serpentinite body may have been part of a topographic high during Cambrian to Mid-Ordovician times, possibly as a

seamount, or part of a transform fault zone. The sequence is interpreted to have been offscraped and accreted to the turbidite wedge (Tabberabbera structural zone) during Mid- to Late Silurian time.

Barkly River Belt Calc-alkaline rocks in the Barkly River Belt (Fig. 3) are preserved within a c. 150km long, 20km wide, ENE-vergent fault zone (Mount Wellington Fault Zone) (Gray & Foster 1998). Although these rocks have arc-like characteristics, they are also much like the mature-stage, suprasubduction zone rocks described by Shervais (2001). They are therefore possibly the remnants of a nascent island arc, representative of the mature stage of the sequences preserved in the Mount Wellington Belt. The sequence is best exposed in the Jamieson River and Licola regions and consists of andesitic to rhyodacitic lavas and minor pyroclastic rocks, voluminous volcanic breccias with interbedded volcanogenic sandstone, siltstone, lenses of black shale, and occasional limestone olistoliths (see also VandenBerg et al. 1995). The lavas are locally pillowed, or columnar, indicative of rapid cooling. No plutonic rocks, apart from quartz diorite dykes that are possibly Devonian in age, are present. This may be due to structural modification within the fault zone by imbrication of the upper volcanic parts of the sequence. The

529

OPHIOLITES OF EASTERN AUSTRALIA

Fig. 6. Concordia diagram (plotted using Ludwig 2000) and cumulative probability plot of 206pb/238U ages of zircons from pegmatitic gabbro from Dookie, Mount Wellington Belt. The weighted mean age of the three analyses that define the older age group is 502 ± 0.7 Ma. The two younger zircon analyses are interpreted to have experienced minor radiogenic Pb loss and were not included in the calculation.

volcanic rocks are mostly strongly deformed, except for some sections of massive lava, and metamorphosed to pumpellyite-actinolite or greenschist facies. Geochemical and petrological

data from the Barkly River Belt are limited but show strong fractionation trends from andesitic through to dacitic, and to rhyodacitic compositions, and island-arc affinities (Table 4) (VandenBergetal. 1995). Zircons from greenschist-facies andesite lava in the Licola region (Fig. 3) have been dated at 500 ± 8 Ma, using the evaporation technique (Table 5; Fig. 7). Six, individual filament loads of 510 of zircon crystals were step heated. The zircons selected were clear, pink, and roundish but prismatic. Back-scattered electron imagery showed they had no multiple growth rims. They were found to contain minimal common Pb, and have high Th/U ratios, consistent with a magmatic origin. The results are plotted on a graph, showing age against heating steps with the 'plateau' defining the crystallization age (Fig. 7). This age is in accord with Earliest Ordovician (Lancefieldian, La2, c. 490 Ma) graptolites that occur in a thin, black, pyritic shale lens in fault contact with underlying volcaniclastic rocks (VandenBerg et al. 1995). Lenses of the same shale also occur within the volcaniclastic rocks. These ages are indistinguishable from both radiometric and palaeontological data in ophiolitic rocks in the Mount Wellington Belt, supporting the interpretation that the Barkly River volcanic rocks may represent the mature stage of suprasubduction zone magmatism.

Occurrences of unknown connections Mafic rocks of presumed Cambrian age occur at Phillip Island, Barrabool Hills, and Waratah Bay in the southern Lachlan Orogen (Fig. 3). Greenschist-facies basalt, dolerite, and cumulate gabbro with low-K, tholeiitic affinities similar to other Lachlan Orogen tholeiites occur at Phillip Island on a c. 0.05 km2 coastal platform completely surrounded by Tertiary basalts (Henry & Birch 1992). The dolerite appears to be less evolved

Table 4. Summary of petrological and geochemical data from Jamieson and Licola, Barkly River Belt Barkly River Belt Jamieson Andesite Dacite-rhyodacite Licola Andesite

Relict primary mineralogy

Major element chemistry

Trace element chemistry (where available)

Cpx-Pl-phyric (some Cpx Medium K, high Mg, low Ti, Zr/Ti 0.01-0.04 dominated) low Al Qtz-sanidine-Pl ± Aug-phyric Zr/Ti 0.01-0.04 Cpx-Pl-phyric, Hbl-Pl-phyric, High K, high Mg, low Ti Hbl-Cpx-Pl-phyric

Zr/Ti 0.04-0.08 Enriched in Sr, Ba, and LREE

Data from Crawford (1988) and VandenBerg et al. (1995).

C. V. SPAGGIARI ETAL.

530

Table 5. Measured Pb isotope ratios and run errors, calculated radiogenic Pb ratio, and age for each heating step of zircon packets from Licola andesite 208pb/206pb Filament 1 0.255255 0.261078 0.258127 0.261928 Filament 2 0.267180 0.279058 Filament 3 0.261401 0.254852 Filament 4 0.271158 0.272015 Filament 5 0.263780 0.256614 Filament 6 0.309474

2a(l(T 4 )

207

Pb/206Pb 2a(l(r 4 )

204

pb/206Pb 2a(l(r 4 ) 207pb/206pb

Age

(Ma)

Th/u

0.000203 0.000053 0.000135 0.002515

0.057966 0.059038 0.057879 0.058483

0.000121 0.000039 0.000130 0.000703

0.000061 0.000146 0.000041 0.000065

0.000001 0.000001 0.000001 0.000002

0.0571 0.0569 0.0573 0.0575

472.93 ± 14.69 487.58 ±2.17 490.86 ± 9.66 494.17 ±5. 13

0.765 0.774 0.776 0.784

0.000079 0.000235

0.058126 0.058338

0.000231 0.000355

0.000077 0.000085

0.000002 0.000002

0.0570 494.91 ± 14.63 0.0571 495.92 ± 5.33

0.800 0.835

0.000149 0.000229

0.059105 0.058028

0.000602 0.000106

0.000104 0.000049

0.000004 0.000001

0.0576 501.85 ±5.27 0.0573 503. 10 ±4.52

0.778 0.765

0.000471 0.000584

0.057341 0.058968

0.000367 0.000649

0.000055 0.000068

0.000001 0.000003

0.0565 504.47 ± 10.06 0.815 0.0580 504.90 ± 14.42 0.813

0.000336 0.000260

0.058215 0.058407

0.000363 0.000230

0.000059 0.000073

0.000001 0.000003

0.0574 51 1.33 ±27.00 0.791 0.0574 513.48 ±24.03 0.768

0.000433

0.057493

0.000128

0.000025

0.000001

0.0571 528.30 ±25.39 0.933

The analyses were performed by TIMS on a Finnigan MAT 262 system at La Trobe University, Melbourne, following procedures described by Dougherty-Page & Bartlett (1999). The Cape Donnington Quartz Gabbro Norite Gneiss was used as a standard (1850 ± 2 Ma, n = 31; compared with 1848.7 ± 4.3 Ma (SHRIMP) and 1849.8 ± 1.1 Ma (conventional, TIMS) (Dougherty-Page & Bartlett 1999). Errors were calculated following Rock et al. (1987).

Fig. 7. Age vs. rank of heating step for zircons from andesite from Licola, Barkly River Belt. The plateaux age of 500 ± 8 Ma is determined from measured 206Pb/ 207 Pb ratios, plotted against individual runs of incremental heating of small zircon packets.

than the basalt, and has slightly lower Ti, Zr, Y, and V, and higher Ni and Cr (Henry & Birch 1992). Greenschist- to amphibolite-facies gabbro crops out sporadically at Barrabool Hills, mostly surrounded by Devonian granites (Crawford 1988). Field relationships are obscure, but the metamorphism is interpreted as syndeformational, rather than contact metamorphism (VandenBerg et al. 2000). Although this occurrence is roughly

along strike from the Heathcote Belt, it is unclear whether it can be correlated. Small, conjugate shear zones with mylonitic foliation trend NW and NE, approximately perpendicular to trends in the Heathcote Belt (VandenBerg et al. 2000). The Cambrian age is inferred, and no amphibolitegrade rocks occur in the Heathcote Belt. The Heathcote Fault Zone may also terminate north of Barrabool Hills, as the western margin of the southern segment is conformable, and the frontal fault appears to lose displacement (Gray & Willman 1991). The coastal platform at Waratah Bay (Fig. 3) exposes a sequence of basalt and gabbro, overlain by Lower Ordovician and Lower Devonian limestone, in fault contact with Devonian turbidites (Crawford 1988; Gray et al 1999; VandenBerg et al. 2000). Loose blocks of serpentinized peridotite occur on the beach (Crawford 1988). The age of these mafic rocks is unclear, but lithological and geochemical similarities to the tholeiitic rocks in the Mount Wellington and Heathcote Belts suggest they may be part of the same sequence. Unlike the other two belts, the sequence trends NE-SW and is not along strike from either. Weakly deformed pillow basalts dominate the exposure and are interbedded with and overlain by bedded cherts, not unlike those at Howqua. The gabbros are mostly hornblende- and pyroxene-phyric, and are partially serpentinized. Metamorphism of the mafic rocks is no higher than prehnite-pumpellyite facies.

OPHIOLITES OF EASTERN AUSTRALIA

531

The presence of a foliation in altered ?gabbro beneath relatively undeformed, Lower Ordovician limestone has been interpreted as representative of a major unconformity by Cayley et al. (2002). This fabric is restricted to a c. 100 m wide zone in heavily metasomatized rocks, and an alternative explanation is that it may have formed during transform or normal faulting of the oceanic crust during spreading. The mafic sequence appears to have occupied a topographic high until approximately Early Devonian times as there is a break in the sedimentation record following deposition of the Lower Ordovician limestone (VandenBerg et al. 2000). The presence of a basal conglomerate containing clasts of basalt and chert conformably beneath the Lower Devonian limestone sequence indicates erosion and exposure of these rocks at that time. A strong magnetic high to the south of the exposure indicates continuation of the mafic rocks offshore (Fig. 3). The sequence may have been part of a transform fault zone, series of seamounts, or an oceanic plateau.

New England Orogen ophiolites The New England Orogen comprises a collage of NNW-trending terranes or tectonic elements ranging in age from Early Palaeozoic to Late Triassic, that represent volcanic arcs, forearc and backarc basins, and subduction complexes (Figs 1 and 8a) (e.g. Murray et al. 1987; Holcombe et al. 1997a, 1997b). These have been interpreted to have formed, or been juxtaposed, along a convergent margin related to a west-dipping subduction zone during Carboniferous to Triassic times that underwent a major phase of shortening during the Late Permian to Mid-Triassic (e.g. Murray et al. 1987; Holcombe et al. 1997a, 1997b). Where constrained, ophiolitic rocks are of Late Neoproterozoic or Early Cambrian age and are faulted against younger arc and subduction complex rocks (Aitchison & Ireland 1995; Bruce et al. 2000). This, and the fact that most of the ophiolitic rocks occur in serpentinite-matrix melanges, has made reconstruction and correlation of ophiolitic stratigraphies problematic.

Northern New England occurrences The most complete ophiolite is the Marlborough Block (Fig. 8a), which is a large, fault-bounded, nappe sheet comprising serpentinized mafic-ultramafic rocks, and younger, unrelated greenschistto lower amphibolite-facies sedimentary rocks and granitoids juxtaposed by thrusting and imbrication (Holcombe et al, 1997a; Korsch et al. 1997). Chert-argillite from the sedimentary sequence is

Fig. 8. (a) Map of the New England Orogen showing the major structural and lithotectonic elements. (After Murray et al 1987; Aitchison & Ireland 1995; Holcombe et al 1997; Bruce et al 2000.) (b) Simplified ophiolitic stratigraphies of the northern and southern sections of the Peel-Manning Fault system. (After Yang & Seccombe 1997.)

interpreted to be equivalent to Carboniferous to Devonian accretionary rocks to the east, whereas the schistose granitoids are similar to Upper Carboniferous S-type granitoids to the south (Hoi-

532

C. V. SPAGGIARI ETAL.

combe et al. 1997a). The nappe was thrust over previously imbricated Devonian to Upper Permian rocks and is therefore unrelated to ophiolite emplacement (Holcombe et al. 1997a). The basal portions of the ultramafic-mafic thrust sheets are schistose, display east over west (toward the margin) shear sense indicators, and are metamorphosed from upper greenschist to amphibolite facies (Bruce et al. 2000). Ultramafic-mafic rocks make up approximately half of the block and consist predominantly of serpentinized massive harzburgite, plus minor dunite, pyroxenite, gabbros, and various mafic and felsic intrusive rocks (Bruce et al. 2000). The Marlborough ultramafic-mafic sequence has been interpreted as a fragment of a spreading centre, based on depleted MORB-like signatures in basalt, Nd isotopes typical of depleted mantle, Cr/(Cr + Al) ratios of c. 0.4, and no high field strength element negative anomalies (Bruce et al. 2000). A Sm/Nd isochron from mafic whole-rock samples indicates crystallization at 562 ± 22 Ma (Bruce et al. 2000). The North D'Aguilar block (Fig. 8a) contains mafic-ultramafic rocks and pelitic schists interpreted as imbricated and underplated oceanic crust, exposed within and above a metamorphic core complex (Little et al. 1995). From top to base, the complex consists of a thin sliver of phyllitic rocks, underlain by the Mount Mia serpentinite, which is underlain by the Rocksberg greenstone (Little et al. 1995). The Mount Mia serpentinite is described as a sheet of ophiolitic serpentinite-matrix melange containing blocks of serpentinite, mafic schist, metachert, metagabbro, peridotite (including Iherzolite, wehrlite, and clinopyroxenite), marble, garnet amphibolite, and blueschist (Little et al. 1995). The matrix contains schistose antigorite, and is of greenschist grade. The melange has a wedge shape with thickness varying from c. 4 km to less than 1 km. It is interpreted to be of sedimentary origin, formed by mass flow deposits during faulting on the sea floor (Little et al. 1993). The Rocksberg greenstone consists of metabasaltic rocks including titanaugite-rich volcaniclastic rocks, and basaltic and serpentinized breccias (Little et al. 1993). Lowgrade chert, slate, broken formation and pillow basalt occur above the Mt Mia Fault and locally contain radiolaria of Early Carboniferous age (Little et al. 1995, and references therein). Ophiolitic rocks of the North D'Aguilar block are interpreted to be the underplated remnants of ophiolitic debris and ophicalcite, and their serpentinized, largely ultramafic substrate that may have been in part derived from a seamount, or from a transform fault zone (Little et al. 1993). They share some similarities with low-grade pelagic

sedimentary rocks and associated basalts from the South D'Aguilar block (Fig. 8a) that contain Late Devonian to Early Carboniferous fauna (Holcombe et al. 1997a, and references therein). Dismembered ultramafic rocks also occur sporadically along the entire length of the Yarrol Fault (Fig. 8a) but very little information is available on these occurrences. They are similar to occurrences along the faulted western margin of the North D'Aguilar Block and some of the Peel-Manning Fault occurrences (see below), and may represent exhumed Lachlan Orogen basement (Korsch et al. 1991; R. Holcombe pers. comm. 2002).

Southern New England occurrences Dismembered ophiolitic rocks occur along a major fault boundary, the Peel-Manning Fault system, in ophiolitic melange (also known as the Great Serpentinite Belt, and Weraerai terrane) (Fig. 8a) (Aitchison et al. 1992; Offler et al. 1997; Yang & Seccombe 1997). This fault zone is c. 300km long, and separates Devonian to Carboniferous arc and forearc rocks (Tamworth Belt, including the Gamilaroi terrane) from Silurian to Lower Carboniferous accretionary-subduction complex rocks to the east (Aitchison & Ireland 1995; Offler et al 1997). Based on seismic and field data, the Peel Fault has been interpreted as a steeply eastdipping splay off a major west-dipping fault zone that extends to the Moho, where the fault-bounded serpentinite melange and ophiolitic rocks form part of the hanging wall (Korsch et al. 1997). Yang and Seccombe (1997) have recognized two ophiolitic associations from the northern and southern sections of the Peel-Manning Fault system, based on differences in petrology and geochemistry. Their simplified stratigraphies are shown in Fig. 8b. They concluded that ophiolites from the northern part were formed by single-stage melting in an open ocean or backarc setting as they have geochemical affinities with MORB or MORB to island-arc basalt signatures (e.g. little fractionation of the REE), whereas those from the southern part (south of Glenrock) are low-Ti, more refractory (high-Mg), and more likely to have been produced from second-stage melts in an island-arc or forearc setting. Possible genetic links between these ophiolites are unclear, in part because of structural complexities within the fault zone. These ophiolitic sequences are mostly of prehnite-pumpellyite metamorphic grade, and are technically thinned with a maximum thickness of c. 1 km (Aitchison & Ireland 1995, and references therein). In addition to the ophiolitic rocks, tectonic blocks include rocks derived from adjacent terranes, as well as rocks of variable metamorphic grade including amphibolite, 465-480 Ma blueschist, and rare eclogite (Aitchi-

OPHIOLITES OF EASTERN AUSTRALIA son & Ireland 1995; Fukui et al 1995). Plagiogranites that occur as blocks in serpentinite-matrix melange within the northern part of the PeelManning Fault system have been dated (SHRIMP, U-Pb in zircon) at 530 ± 6 Ma, 509 ± 30, and 535 ± 10 Ma (Aitchison et al. 1992; Aitchison & Ireland 1995). Based on their geochemistry, the plagiogranites are interpreted as differentiates from the same highly refractory magmas as the mafic ultramafic rocks, and therefore date the late stages of oceanic crustal formation (Aitchison et al. 1994; Aitchison & Ireland 1995). Gamilaroi terrane rocks (part of the Tamworth Belt) occur to the west of the Peel-Manning Fault system (Fig. 8a) and include fault slices of a Late Silurian-Devonian island arc (Aitchison et al. 1997). They comprise felsic to intermediate, calcalkaline volcanic rocks and volcaniclastic rocks, and tholeiitic lavas and shallow intrusive rocks. Further north, and just west of the Peel-Manning Fault system, resedimented limestone clasts in forearc basin conglomerate contain Mid-Cambrian to ?Early Ordovician fauna, along with low-K, lowTi andesitic, basaltic, and rhyolitic clasts, indicative of an early Palaeozoic, oceanic or island-arc substrate (Leitch & Cawood 1987). East of the Peel-Manning Fault system, ophiolitic rocks within the Ngamba subduction complex in the Port Macquarie area (Fig. 8a) occur in a serpentinite-matrix melange and comprise blocks of gabbro, diorite, wehrlite, harzburgite, pyroxenite, dolerite, pillow basalt, chert, volcaniclastic rocks, amphibolite, as well as glaucophane schist and rare eclogite that have K/Ar ages of 444 Ma and 469 Ma (Scheibner 1985; Watanabe et al. 1993; Aitchison et al. 1994). The other blocks are mostly of prehnite-pumpellyite to greenschist grade (Aitchison et al. 1994). The Baryulgil ultramafic rocks (Fig. 8a) occur as a c. 32 km long serpentinized body metamorphosed to amphibolite grade on its western margin (Aitchison et al. 1994). They comprise predominantly massive harzburgite, with minor dunite and pyroxenite, intruded by doleritic dykes that have both tholeiitic and boninitic affinities (Aitchison et al. 1994). Unconformably overlying Upper Permian volcanosedimentary rocks (Drake Volcanic series) provide a minimum age of emplacement, but protolith ages are unknown (Aitchison et al. 1994).

Discussion Tectonic setting and evolution oftheprotoTasmanides The main features of the major ophiolites documented here are summarized in Table 6. The preservation of diverse ophiolitic rocks, with ages

533

534

C. V. SPAGGIARI ETAL.

spanning from c. 560 to 495 Ma, reflects a period of complex oceanic crustal development off the east Gondwana margin. This involved development of oceanic crust at a spreading ridge (now preserved as the c. 560 Ma Marlborough ophiolite), development of forearc crust in a suprasubduction zone between c. 530 and 515 Ma (Tyennan-Delamerian occurrences) (Fig. 9a and b), possible synchronous formation of suprasubduction zone crust further outboard (southern New England Orogen occurrences), emplacement of Tyennan-Delamerian ophiolites onto the Gondwana margin at c. 515 Ma, followed by development of new forearc and backarc crust between c. 505 and 495 Ma (Lachlan Orogen) (Fig. 9c). Eruption of transitional alkaline volcanic rocks (Mt Arrowsmith volcanic series, Wonominta Block; Fig. 1) has been interpreted as indicative of continental margin rifting at c. 585 Ma (Crawford et ol. 1997). The same rifting event is possibly recorded by Neoproterozoic rift tholeiites preserved in the Rocky Cape Block of Tasmania (Crawford et at. 1997). This rifting potentially led to production of the proto-Marlborough ophiolite at a spreading centre outboard of the margin (proto-Pacific?). One or more island arcs may have developed within this ocean basin at around 530-520 Ma, and produced suprasubduction zone crust (Fig. 9a and b). Collision of the TyennanDelamerian forearc sequence with the passive Gondwana margin at c. 515 Ma led to partial subduction of the passive margin, ophiolite emplacement (Crawford & Berry 1992), followed by rapid exhumation of metamorphic complexes in Tasmania at 505 Ma, preceding development of the calc-alkaline Mt Read volcanic rocks at c. 500 Ma (Foster et at. 2002). This margin extension and volcanism may have been synchronous with renewed subduction, and roll-back, beneath the Gondwana margin to form the backarc and forearc sequences of the Lachlan Orogen, between c. 505 and 495 Ma (Foster et at. 2002) (Fig. 9c). As such, the Lachlan Orogen ophiolites are representative of a marginal basin formed within a broader Palaeo-Pacific ocean, with intervening topographic highs such as oceanic plateaux, seamounts, and possibly including fragments of relict arc or continental ribbons from the rifted margin (Fig. 9d). This complex proto-Tasmamde evolution may have been in part related to far-field plate tectonic processes during the assembly of Gondwana in Pan-African times. Geochronological data from Prydz Bay and the Prince Charles Mountains, Antarctica, indicate tectonic activity at c. 550490 Ma that is potentially related to a previously unrecognized suture within East Gondwana (Fitzsimons et at. 1997; Boger et at. 2001). Whether

this tectonism had a direct effect on the early development of the Tasmanides is unknown; however, it is evident that significant complexities occurred during this time of oceanic crustal development along the East Gondwana margin.

Cordilleran- and Tethyan-type ophiolites Tethyan-type ophiolites in the Tasmanides are represented by the Tyennan (Tasmanian) ophiolite, and possible correlates in the Delamerian Orogen. These are interpreted to have thrust sheet-like form, and, in the case of the Tyennan ophiolite, a well-developed metamorphic sole (Berry & Crawford 1988). Radiometric dating and field relationships suggest that this ophiolite was probably emplaced onto the continental margin while it was still young and warm, in keeping with many other examples of this type (e.g. Semail ophiolite; Hacker et at. 1996). In contrast, Lachlan Orogen ophiolites have no relationship to continental basement and were emplaced within major fault zones c. 50 m.y. after their formation (Foster et at. 1999; Spaggiari et at. 2002b). They are therefore considered to be Cordilleran type, and are interpreted to have become dismembered during underthrusting and basin closure in a predominantly oceanic (backarc) setting, inboard of a major subduction zone to the east. These ophiolites have metamorphic and deformational characteristics consistent with those formed by processes of accretion via offscraping or underplating, and duplexing during intraoceanic thrusting (Gray & Foster 1998; Spaggiari et at. 2002a, 2002b). This includes widespread and pervasive prehnite-pumpellyite metamorphism, and low-temperaturehigh-pressure assemblages. New England Orogen ophiolites are more difficult to interpret, in part because of obscure relationships of the various tectonic elements or terranes, and significant modification during later deformation. The lack of any continental basement suggests they are likely to be Cordilleran type, in particular, those in the North D'Aguilar Block that are interpreted to be the preserved remnants of underplated and accreted ophiolitic debris (Little et at. 1993). The formation of these types of ophiolites in the Tasmanides reflects a history that lacks any terminal continent-continent collision (see Coney 1992). Tethyan types emplaced onto continental margins appear to preserve larger mantle sections and a greater proportion of gabbro than dismembered Cordilleran types, and may even be dominated by these (e.g. peridotite-gabbro sections in Tasmania; Semail ophiolite, Oman; Tauride ophiolites, southern Turkey). Although in some cases complete pseudostratigraphies are preserved (e.g. Se-

OPHIOLITES OF EASTERN AUSTRALIA

535

Fig. 9. (a) Schematic tectonic diagram showing island-arc and related forearc crust, just prior to continent-arc collision, partial subduction of the margin, and ophiolite emplacement. (Modified from Crawford & Berry 1992.) (b) Schematic tectonic map of the east Gondwana margin for the time interval 530—515 Ma, including a tentative position for the location of the proto-New England ophiolites now preserved along the Peel-Manning Fault system, (c) Schematic illustration of the development of forearc and backarc crust in the Lachlan Orogen. In this model subduction initiated beneath the Gondwana margin following arc collision and obduction of c. 525 Ma ophiolites, which led to roll-back and formation of new forearc crust (proto Heathcote Belt). The subduction zone migrated eastward and stepped outboard of potentially older (c. 525 Ma?), rifted ?arc crust, forming new forearc crust (e.g. proto Mount Wellington Belt), and eventually mature-stage, calc-alkaline volcanic rocks (Jamieson-Licola and Barkly River Belt), (d) Schematic diagram showing Early Ordovician palaeogeography of the east Gondwana margin and proto-Lachlan Orogen.

536

C. V. SPAGGIARI ETAL.

Fig. 9. (Continued}

mail ophiolite, Oman) often the upper sections of ophiolite pseudostratigraphies (pelagic, basalt, and sheeted dyke sections) are missing. This is possibly due to post-emplacement erosion (e.g. Dilek et al, 1999), but may also be due to later deformation. The preferential preservation of complete or lower sections of Tethyan-type ophiolites is probably a function of their emplacement as minimally disrupted thrust sheets, often during partial subduction of the passive margin they are being emplaced onto. In contrast, Cordilleran types (e.g. Lachlan Orogen ophiolites; many Pacific-rim occurrences) are dominated by upper oceanic crustal sequences (gabbro-dolerite-basalt-pelagic rocks). The exception for the Lachlan Orogen is the Dolodrook serpentinite but this may have been part of the upper oceanic crust prior to emplacement, either as a seamount or preserved in a transform fault zone. The preferential preservation of upper-crustal stratigraphies is also probably a function of emplacement processes, where the top of the oceanic crust is disrupted during subduction, forming melange and fault slivers that are imbricated or duplexed and accreted within evolving accretionary wedges (see Kimura & Ludden 1995; Coleman 2000). Many of the Cordilleran ophiolites along the California margin are interpreted as

having been emplaced as technically underplated to the base of a subduction wedge, imbricated and accreted during convergence, or detached from oceanic fracture zones during orthogonal subduction (Coleman 2000). This work has been supported by the Australian Geodynamics Cooperative Research Center, Australian Research Councils Grant A39601548 (awarded to D.R.G.), and National Science Foundation Grant EAR 0073638 (awarded to D.A.E). M. Elburg is thanked for analysing the zircons from the Licola andesite, and A. Heatherington is thanked for analysing zircon from the Dookie gabbro. Helpful reviews by Y. Dilek and R. Holcombe improved the manuscript. This is Tectonics Special Research Center (TSRC) publication no. 251.

References AITCHISON, J.C. & IRELAND, T.R. 1995. Age profile of ophiolitic rocks across the Late Palaeozoic New England Orogen, New South Wales: implications for tectonic models. Australian Journal of Earth Sciences, 45, 11-23. AITCHISON, J.C., IRELAND, T.R., BLAKE, M.C. JR & FLOOD, P.G. 1992. 530 Ma age for ophiolite from the New England Orogen: oldest rocks known from eastern Australia. Geology, 20, 125-128. AITCHISON, J.C., BLAKE, M.C. JR, FLOOD, P.G. & JAYKO, A.S. 1994. Palaeozoic ophiolitic assem-

OPHIOLITES OF EASTERN AUSTRALIA blages within the southern New England orogen of eastern Australia: implications for growth of the Gondwana margin. Tectonics, 13, 1135-1149. AITCHISON, J.C., STRATFORD, J.M.C. & BUCKMAN, S. 1997. Geology of the upper Barnard region: evidence of Early Permian oblique-slip faulting along the Peel-Manning Fault System. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19, 188-196. BERRY, R.F. & CRAWFORD, A.J. 1988. The tectonic significance of Cambrian allochthonous mafic ultramafic complexes in Tasmania. Australian Journal of Earth Sciences, 35, 523-533. BLACK, L.P., SEYMOUR, D.B. & CORBETT, K.D. ET AL. 1997. Dating Tasmania's Oldest Geological Events. AGSO Record, 1997/15. BOGER, S.D., WILSON, CJ.L. & FANNING, C.M. 2001. Early Paleozoic tectonism within the East Antarctic craton: the final suture between east and west Gondwana? Geology, 29(5), 463-466. BROWN, M.C. 1998. The eastern boundary of the Melbourne Zone and the northern Mount Howitt province near Tatong. AGSO Record, 1998/2, 10-11. BRUCE, M.C., Niu, Y., HARBORT, T.A. & HOLCOMBE, R.J. 2000. Petrological, geochemical and geochronological evidence for a Neoproterozoic ocean basin recorded in the Marlborough terrane of the northern New England Fold Belt. Australian Journal of Earth Sciences, 47, 1053-1064. CAYLEY, R.A., TAYLOR, D.H., VANDENBERG, A.H.M. & MOORE, D.H. 2002. Proterozoic-early Palaeozoic rocks and the Tyennan Orogeny in central Victoria: the Selwyn block and its tectonic implications. Australian Journal of Earth Sciences, 49, 225-254. COLEMAN, R.G. 2000. Prospecting for ophiolites along the California continental margin. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 351-364. COLLINS, WJ. 2002. Nature of extensional accretionary orogens. Tectonics, 21, 1-12. CONEY, P.J. 1992. The Lachlan belt of eastern Australia and Circum-Pacific tectonic evolution. Tectonophysics,2U, 1-25. CONEY, P.J., EDWARDS, A., HINE, R., MORRISON, F. & WINDRUM, D. 1990. The regional tectonics of the Tasman Orogenic system, eastern Australia. Journal of Structural Geology, 125, 19-43. CRAWFORD, A.J. 1988. Cambrian. In: DOUGLAS, J.G. & FERGUSON, J.A. (eds) Geology of Victoria. Geological Society of Australia, Victoria Division, Melbourne, 37-62. CRAWFORD, AJ. & CAMERON, W.E. 1985. Petrology and geochemistry of Cambrian boninites and low-Ti andesites from Heathcote, Victoria. Contributions to Mineralogy and Petrology, 91, 93-104. CRAWFORD, AJ. & BERRY, R.F. 1992. Tectonic implications of Late Proterozoic-Early Palaeozoic igneous rock associations in western Tasmania. Tectonophysics, 214, 37-56.

537

CRAWFORD, A.J. & KEAYS, R.R. 1987. Petrogenesis of Victorian Cambrian tholeiites and implications for the origin of associated boninites. Journal of Petrology, 28, 1075-1109. CRAWFORD, A.J., CAMERON, W.E. & KEAYS, R.R. 1984. The association boninite-low-Ti andesite-tholeiite in the Heathcote greenstone belt, Victoria; ensimatic setting for the early Lachlan fold belt. Australian Journal of Earth Sciences, 31, 161-175. CRAWFORD, A.J., DONAGHY, A.G., BLACK, L. & STUART-SMITH, P. 1996. Mt Read Volcanics correlates in Western Victoria: a new exploration opportunity. Australian Institute of Geoscientists Bulletin, 20,97-102. CRAWFORD, AJ., STEVENS, B.PJ. & FANNING, C.M. 1997. Geochemistry and tectonic setting of some Neoproterozoic and Early Cambrian volcanics in western New South Wales. Australian Journal of Earth Sciences, 44, 831-852. DILEK, Y., THY, P., HACKER, B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and mafic dike intrusions (Turkey): implications for the Neotethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. DOUGHERTY-PAGE, J.S. & BARTLETT, J.M. 1999. New analytical procedures to increase the resolution of zircon geochronology by the evaporation technique. Chemical Geology, 153, 227-240. EDWARDS, J., WHOLT, K.E., SLATER, K., OLSHINA, A. & HUTCHINSON, D.F. 1998. Heathcote and Parts of Woodend and Echuca 1:100 000 Map Area Geological Report. Geological Survey of Victoria Report, 108. FERGUSSON, C.L. 1998. Cambrian-Silurian oceanic rocks, upper Howqua River, eastern Victoria: tectonic implications. Australian Journal of Earth Sciences, 45, 633-644. FERGUSSON, C.L. & CAWOOD, P.A. 1995. Structural history of the Bay of Islands Complex, western Newfoundland. Canadian Journal of Earth Sciences, 32, 533-544. FITZSIMONS, I.C.W., KINNY, P.D. & HARLEY, S.L. 1997. Two stages of zircon and monazite growth in anatectic leucogneiss: SHRIMP constraints on the duration and intensity of Pan-African metamorphism in Prydz Bay, East Antarctica. Terra Nova, 9, 47-51. FOSTER, D.A. & GRAY, D.R. 2000. Evolution and structure of the Lachlan Fold Belt (Orogen) of Eastern Australia. Annual Review of Earth and Planetary Sciences, 28, 47-80. FOSTER, D.A., GRAY, D.R., KWAK, T.A.P. & BUCHER, M. 1998. Chronology and erogenic framework of turbidite hosted gold deposits in the Western Lachlan fold belt, Victoria: 40Ar/39Ar results. Ore Geology Reviews, 13, 229-250. FOSTER, D.A., GRAY, D.R. & BUCHER, M. 1999. Chronology of deformation within the turbiditedominated Lachlan Orogen: implications for the tectonic evolution of eastern Australia and Gondwana. Tectonics, 18(3), 452-485. FOSTER, D.A., GRAY, D.R., HARTLEY, M. & SPAGGIARI, C.V. 2002. Dating metamorphism and exhumation of Forth metamorphic complex, Tasmania with

538

C. V. SPAGGIARI ETAL.

implications for Paleozoic tectonics of the east Gondwana margin. Geological Society of Australia Abstracts, 67, 116. FUKUI, S., WATANABE, T., TETSUMARU, I. & LEITCH, E.G. 1995. Middle Ordovician high PT metamorphic rocks in eastern Australia: evidence from K-Arages. Tectonics, 14, 1014-1020. GRAY, D.R. 1997. Tectonics of the southeastern Australian Lachlan Fold Belt: structural and thermal aspects. In: BURG, J.-P. & FORD, M. (eds) Orogeny through Time. Geological Society, London, Special Publications, 121, 149-177. GRAY, D.R. & FOSTER, D.A. 1998. Character and kinematics of faults within the turbidite-dominated Lachlan Orogen: implications for tectonic evolution of eastern Australia. Journal of Structural Geology, 20(12), 1691-1720. GRAY, D.R. & WILLMAN, C.E. 1991. Deformation in the Ballarat Slate Belt, central Victoria, and implications for the crustal structure across southeast Australia. Australian Journal of Earth Sciences, 38, 171-201. GRAY, D.R., JANSSEN, C. & VAPNIK, Y. 1999. Deformation character and fluid flow across a wrench fault within a Palaeozoic accretionary wedge: Waratah Fault Zone, southeastern Australia. Journal of Structural Geology, 21, 191-214. HACKER, B.R., MOSENFELDER, J.L. & GNOS, E. 1996. Rapid emplacement of the Oman ophiolite: thermal and geochronological constraints. Tectonics, 15(6), 1230-1247. HENRY, D.A. & BIRCH, W.D. 1992. Cambrian greenstone on Phillip Island, Victoria. Australian Journal of Earth Sciences, 39, 567-575. HOLCOMBE, R.J., STEPHENS, CJ. & FIELDING, C.R. ET AL. 1997a. Tectonic evolution of the northern New England Fold Belt: Carboniferous to Early Permian transition from active accretion to extension. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19, 66-79. HOLCOMBE, R.J., STEPHENS, CJ. & FIELDING, C.R. ET AL. 1997b. Tectonic evolution of the northern New England Fold Belt: the Permian-Triassic Hunter Bowen event. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19, 52-65. KIMURA, G. & LUDDEN, J. 1995. Peeling oceanic crust in subduction zones. Geology, 23, 217-220. KORSCH, R.J., JOHNSTONE, D.W. & WAKE-DYSTER, K.D.

1997. Crustal architecture of the New England orogen based on deep seismic reflection profiling. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19,29-51. LEITCH, E.G. & CAWOOD, P.A. 1987. Provenance determination of volcaniclastic rocks: the nature and tectonic significance of a Cambrian conglomerate from the New England Fold Belt, eastern Australia. Journal of Sedimentary Petrology, 57(4), 630-638.

LITTLE, T.A., HOLCOMBE, R.J. & SILWA, R. 1993. Structural evidence for extensional exhumation of blueschist-bearing serpentinite matrix melange, New England Orogen, Southeast Queensland, Australia. Tectonics, 12(2), 536-549. LITTLE, T.A., MCWILLIAMS, M.O. & HOLCOMBE, R.J. 1995. 40Ar/39Ar thermochronology of epidote blueschists from the North D'Aguilar block, Queensland, Australia: timing and kinematics of subduction complex unroofing. Geological Society of America Bulletin, 107(5), 520-535. LUDWIG, K.R. 2000. Isoplot/Ex Version 2.3. A Geochronological Toolkit for Microsoft Excel. Berkeley Geochronology Center Special Publication, la. MAKER, S., MOORE, D.H., CRAWFORD, A.J., TWYFORD, R. & FANNING, C.M. 1997. Test Drilling of the Southern Margin of the Murray Basin. Victorian Initiative for Minerals and Petroleum Report, 52. MEFFRE, S., BERRY, R.F. & HALL, M. 2000. Cambrian metamorphic complexes in Tasmania: tectonic implications. Australian Journal of Earth Sciences, 47, 971-985. MOORES, E.M. 1982. Origin and emplacement of ophiolites. Reviews of Geophysics and Space Physics, 20, 735-760. MOORES, E.M., KELLOGG, L.H. & DILEK, Y. 2000. Tethyan ophiolites, mantle convection, and tectonic 'historical contingency': a resolution of the 'ophiolite conundrum'. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 3-12. MORAND, V.J., CAYLEY, R.A., WOHLT, K.E. & TAYLOR, D.H. 2001. Ultramafic rocks in the Glenelg Zone of western Victoria: tectonic implications. Geological Society of Australia Abstracts, 64, 124-125. MUELLER, P., SHUSTER, R., GRAVES, M., WOODEN, J. & BOWES, D. 1988. Age and composition of a late Archean magmatic complex, Beartooth Mountains, Montana-Wyoming. Montana Bureau of Mines and Geology, Special Papers, 96, 23-42. MURRAY, C.G., FERGUSON, C.L., FLOOD, P.G., WHITAKER, W.G. & KORSCH, R.J. 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Journal of Earth Sciences, 34, 213-236. NELSON, D.R., CRAWFORD, A.J. & MCCULLOCH, M.T. 1984. Nd-Sr isotopic and geochemical systematics in Cambrian boninites and tholeiites from Victoria, Australia. Contributions to Mineralogy and Petrology, 88, 164-172. OFFLER, R., O'HANLEY, D.S. & LENNOX, P.G. 1997. Tectonic significance of veins and associated structures at Woodsreef, New South Wales. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19, 226-241. PERKINS, C. & WALSHE, J.L. 1993. Geochronology of the Mount Read Volcanics, Tasmania, Australia. Economic Geology, 88, 1176-1197. POWELL, C.M., Li, Z.X., MCELHINNY, M., MEERT, J.G. & PARK, J.K. 1993. Palaeomagnetic constraints on

OPHIOLITES OF EASTERN AUSTRALIA the timing of the Neoproterozoic break-up of Rodinia and the Cambrian formation of Gondwana. Geology, 21, 889-892. PRIESS, W.V. 1995. Delamerian Orogeny. In: DREXEL, J.F. & PREISS, W.V. (eds) The Geology of South Australia, Volume 2: The Phanerozoic. Geological Survey of South Australia Bulletin, 54, 45-60. ROCK, N.M.S., WEBB, J.A., MCNAUGHTON, N.J. & BELL, G.D. 1987. Nonparametric estimation of averages and errors for small datasets in isotope geoscience: a proposal. Chemical Geology, 66, 163-177. SCHEIBNER, E. 1985. Suspect terranes in the Tasman Fold belt system, Eastern Australia. In: HOWELL, D.G. (ed.) Tectonostratigraphic Terranes of the Circum-Pacific Region. Circum-Pacific Council for Energy and Mineral Resources, Earth Science, Series 1,493-514. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, No. 2000GC000080. SPAGGIARI, C.V., GRAY, D.R. & FOSTER, D.A. 2002a. Blueschist metamorphism during accretion in the Lachlan Orogen, southeastern Australia. Journal of Metamorphic Geology, 20, 711-726. SPAGGIARI, C.V., GRAY, D.R., FOSTER, D.A. & FANNING, C.M. 2002b. Occurrence and significance of blueschists in the southern Lachlan Orogen. Australian Journal of Earth Sciences, 49, 255-269. STAGEY, J. & KRAMERS, J. 1975. Approximation of terrestrial lead isotopic evolution by a two stage model. Earth and Planetary Science Letters, 26, 381-399.

539

TICKELL, SJ. 1989. Dookie 1:100000 Map and Geological Report. Geological Survey of Victoria Report, 87. VANDENBERG, A.H.M. 1991. Kilmore 1:50000 Map and Geological Report. Geological Survey of Victoria Report, 91. VANDENBERG, A.H.M. & STEWART, I.R. 1992. Ordovician terranes of the southeastern Lachlan Fold Belt: stratigraphy, structure and palaeogeographic reconstruction. Tectonophysics, 214, 159—176. VANDENBERG, A.H.M., WILLMAN, C.E., HENDRICKX, M., BUSH, B.D. & SANDS, B.C. 1995. The Geology and Prospectivity of the 1993 Mount Wellington Airborne survey Area. Victorian Initiative for Minerals and Petroleum Report, 2. VANDENBERG, A.H.M., WILLMAN, C.E. & MAKER, S. ET AL. 2000. The Tasman Fold Belt System in Victoria. Geological Survey of Victoria, Special Publications. WATANABE, T., LEITCH, E.C. & FUKUI, S. 1993. Early Ordovician high PIT metamorphic inclusions from serpentinite bodies in the southern New England Fold Belt. In: FLOOD, P.G. & AITCHISON, J.C. (eds) New England Orogen, Eastern Australia. University of New England, Armidale, 181-186. WOHLT, K.E. & EDWARDS, J., 1999. Heathcote 1:50000 Geological Map. Geological Survey of Victoria. YANG, K. & SECCOMBE, P.K. 1997. Geochemistry of the mafic and ultramafic complexes of the northern Great Serpentinite Belt, New South Wales: implications for first stage melting. In: ASHLEY, P.M. & FLOOD, P.G. (eds) Tectonics and Metallogenesis of the New England Orogen. Geological Society of Australia, Special Publications, 19, 197-211.

This page intentionally left blank

Ophiolites in China: their distribution, ages and tectonic settings Q. ZHANG 1 , Y. WANG 2 , G. Q. ZHOU 3 , Q. QIAN 1 & PAUL T. ROBINSON 4 ' 5 1

Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, RR. China 100029 (e-mail: [email protected]) 2 Department of Earth Sciences, The University of Hong Kong, Hong Kong, China ^Department of Earth Sciences, Nanjing University, Nanjing, PR. China 210083 A Department of Earth Sciences, Dalhousie University, Halifax, N.S., Canada B3H 3J5 5 Present address: Flemish Royal Academy of Belgium for Science and the Arts, Hertogsstraat 1, B-1000, Brussels, Belgium Abstract: Ophiolites are widespread and abundant in China, where they lie along suture zones delineating major tectonic blocks. They range in age from Proterozoic to Tertiary and generally belong to the Palaeo-Asian, Qinling-Qilian-Kunlun, Tethyan and Circum-Pacific systems. A few possible Proterozoic ophiolites may reflect rifting of Rodinia at c. 800-1000 Ma. The Palaeo-Asian ophiolites, which range in age from Cambrian to Carboniferous, crop out abundantly in the northern parts of the country whereas the Qinling-Qilian-Kunlun ophiolites occur in north-central China, along the boundaries of the Tarim, North China and Yangtze blocks. Tethyan ophiolites are confined to southwestern China and Tibet whereas those of the Circum-Pacific belt are found in Taiwan and northeastern China. Chinese ophiolites are typically technically disrupted melanges composed of isolated blocks of peridotite, gabbro and basalt. Sheeted dykes are rare or absent. Many of the ophiolites are compositionally complex, containing mixtures of island-arc tholeiite and boninite with lesser amounts of mid-ocean ridge basalt and ocean-island basalt. Most show evidence of having been formed or assembled in suprasubduction zone environments. However, Palaeo-Tethyan ophiolites generally lack suprasubduction zone signatures and may have formed in small, intracontinental basins. A rapidly expanding database of high-precision age dates and detailed geochemical analyses on Chinese ophiolites is providing new insight into the nature and timing of the tectonic events that shaped this part of Asia.

China and the surrounding regions are made up of a number of tectonic blocks, e.g. the North China, Yangtze, Tarim, Indian and Siberian Blocks (Fig. 1), and most of the ophiolites in the country occur along the margins of these tectonic units. Ophiolites are found primarily in the western, southwestern and northern parts of China (Fig. 2) and they fall into four principal age groups: Proterozoic, early Palaeozoic, late Palaeozoic and Mesozoic-Cenozoic (Zhang & Zhou 2001). A few older ophiolites have been reported (e.g. Kusky et al. 2001) but the true nature of these bodies is uncertain. The only ophiolites of likely Precambrian age are a few late Proterozoic bodies such as Songshugou and Kuanping of the North Qinling Mountains and an unnamed ophiolite in northeastern Jiangxi Province. Palaeozoic ophiolites are of Cambrian-Ordovician, Silurian-Devonian and Devonian-Permian age, and most of these lie in the Palaeo-Asian orogenic belt that separates the Siberian and North China Blocks (Fig. 2). Cambrian-Ordovician

ophiolites in this belt include Hongguleleng and Tangbale in West Junggar, Kumishi in the Tianshan Mountains, Beishan in Gansu Province, Ondor Sum in Inner Mongolia and numerous bodies in the Qinling-Qilian-Kunlun orogenic belt. Silurian-Devonian ophiolites include Mayile and Darbut in West Junggar, Changawuzi of the south Tianshan Mountains, and Solonshan and Hegenshan in Inner Mongolia. Devonian-Permian ophiolites are also common in the Palaeo-Asian orogenic belt and crop out at Darbut, Kelameile and Solonshan, as well as in the northern and southern Tianshan Mountains, Palaeo-Tethyan ophiolites are all DevonianTriassic in age and include Jinshajiang, Ailaoshan, Changning-Menglian and Ganzi-Litang, as well as several bodies in the Animaqin ophiolite belt. The Kaishantun ophiolite in NE China and the Tananao ophiolite in Taiwan may also belong to the Palaeo-Tethyan group. Neo-Tethyan ophiolites of Cretaceous-Tertiary age crop out chiefly in the Yarlung Zangbo and Bangong Lake-Nujiang

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 541-566. 0305-8719/037$ 15 © The Geological Society of London 2003.

542

Q. ZHANG ETAL.

Fig. 1. A generalized map showing the major tectonic blocks and erogenic belts of China.

suture zones of southern Tibet. The Lichi and Kenting ophiolites in Taiwan are also of Tertiary age but belong to the Circum-Pacific Series. In this paper, we briefly summarize the distribution, age, and composition of the most important and well-studied ophiolites in China and use these data to interpret the tectonic environments in which they formed. Because reliable age and geochemical data are sparse, our tectonic conclusions are preliminary. However, detailed ophiolite studies are advancing rapidly in China and, hence a review of current knowledge will help to focus attention on outstanding problems.

Precambrian ophiolites The oldest reported ophiolite in China is the Dongwanzi massif of northern China, which has been dated at 2500 Ma (Kusky et al 2001). However, identification of this body as an ophiolite has been disputed (Li et al 2002; Zhai et al 2002) and more work is needed to demonstrate that the various components of this complex are genetically related. Based on a review of the field and petrological data, we conclude that the Dongwanzi massif is probably not an ophiolite because: (1) the peridotite in the massif reported by Kusky

et al (2001) is actually amphibolite associated with a V-Ti magnetite deposit; (2) the gabbros have primary biotite, a feature unusual for ophiolites; (3) the so-called sheeted dykes have no chilled margins and are probably not diabase; (4) the chromites in Zhunhua reported by Li et al (2002) appear to be cumulate rather than podiform in character; (5) there are no detailed geochemical data available to show that the various units are related. Thus, we argue that no unequivocal ophiolite of Archaean age has yet been identified in China. A few ophiolites of probable Neoproterozoic age occur in the North Qinling Mountains and in NE Jiangxi Province. In these areas, numerous small ophiolitic blocks are scattered in flysch with well-developed schistosity. The ophiolitic blocks contain variable mixtures of serpentinized peridotite, gabbro, diabase, diorite, anorthosite, plagiogranite, basalt and andesite. Although the bodies have been extensively faulted, relationships among the various lithologies are locally preserved (Zhao et al 1995). U-Pb dating of zircons from jadeite-glaucophane anorthosite in NE Jiangxi Province yielded an age of 968 ± 23 Ma (Li et al 1994). Fifteen recalculated Sm-Nd dates from the same area yielded an isochron age of 955 ±

OPHIOLITES IN CHINA

543

Fig. 2. A generalized map showing the distribution and age of major ophiolites in China.

44 Ma with eNd(0 of +4.3 to +6.7. These data, together with published 40Ar/39Ar ages on the ophiolite, suggest that the NE Jiangxi tectonic belt developed between about 970 and 800 Ma (Li et al. 1994). Recent work on radiolarites in the Jiangxi ophiolite raised the possibility that this body may have formed in the late Palaeozoic rather than the Neoproterozoic (Li et al. 1996; Zhao et al. 1997). However, Li (2000) re-examined these radiolarian cherts and argued that they probably formed in a continental margin environment and that they are not genetically related to the ophiolite. This interpretation is in agreement with regional tectonic studies that argue against the presence of an ocean basin between the Yangtze and Cathaysia blocks in the late Palaeozoic. The Songshugou ophiolite in the North Qinling Mountains consists of mantle peridotite (mainly dunite) and mafic rocks, which have been metamorphosed to the amphibolite facies. The rocks include plagioclase amphibolite, garnet-bearing plagioclase amphibolite and amphibole schist with minor lenses of marble, all of which are interpreted as blocks of ophiolitic melange. The SmNd isochron age of the plagioclase amphibolite

ranges from 983 ± 140 Ma, with eNd(0 of +6.8 ± 0.9 (Li et al. 1991) to 1030 ± 46 Ma, with eNd(f) of +5.7 + 0.2 (Dong et al. 1997; Zhou et al. 1998). The mafic rocks of the Songshugou ophiolite have normal mid-ocean ridge basalt (NMORB) and enriched MORB (E-MORB) affinities and probably formed in a small ocean basin (Table 1, Figs 3a, b and 4).

Palaeo-Asian ophiolites It has generally been thought that the PalaeoAsian Ocean was located between the SiberianKazakhstan and Tarim-North China blocks (Dobretsov et al. 1995). However, we suggest on the basis of their age and distribution that the ophiolites of the Kunlun Mountains, the North Qilian Mountains and the northern Qinling Mountains (Figs 1 and 2) are also part of the Palaeo-Asian system. The Palaeo-Asian Ocean probably opened originally in the Mid- to Neoproterozoic (before 900 Ma) based on the presence of Neoproterozoic ophiolites in this system in Russia, Mongolia and Kazakhstan (Dobretsov et al. 1995). The late Proterozoic Kuanping and Songshugou ophiolites

544

Q. ZHANG ETAL.

Table 1. Major oxides and trace elements for selected ophiolites in China Rock: Sample: SiO2 TiO2 A12O3 Fe203 FeO MnO MgO CaO Na 2 O K2O P205 H20+ LOI Total Cr Ni Co Ba Rb Sr Y Hf Zr Nb Th Ta U La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Rock: Sample: SiO2 Ti02 A1203 Fe203 FeO MnO MgO CaO Na 2 O K20 P205 H2O+ LOI Total Cr Ni Co Ba Rb Sr Y Hf Zr Nb Th Ta U La Ce Pr Nd

1 Basalt LT16

Basalt T-03

3 Basalt 93Xs-3

44.01 1.64 15.46 2.87 10.36 0.24 8.63 12.58 1.68 0.32 0.16 1.56 0 99.51 248 55.8 65.5 80.2 3.45 168

47.97 1.21 14.5 2.64 9.13 0.21 8.38 11.87 2.74 0.36 0.04 0.99 0.01 100.05 293 80.6 42.1 35.7 4.5 139

43.89 0.91 13.55 3.37 12.41 0.35 7.44 14.57 0.92 0.28 0.02 1.57 0.78 100.06 347 62.3 74.5 201 9.1 358

3.17 81.7

1.78 65

2.63 92.3

0.55 0.47 0.58 4.83 13.8

0.27 0.33 0.29 4.16 9.42

0.40 0.32 0.42 4.55 7.25

0.53 0.231 0.65 3.31 7.9

10.9 3.81 1.67

6.56 2.39 0.94

7.37 2.26 0.81

5.19 1.51 0.514

0.96

0.69

0.68

0.31

2

4 Basalt T35

22 137 29.8 1.58 45.1 270 15.8 1.31 62.1

5 Basalt T-41-1

84 143 51.3 82 12.1 141 21.8 2.03 63.5

9 Basalt H-9A

10 Basalt CHA-8B

11 Basalt W5i

49.21 0.38 14.7 2.31 7.28 0.17 8.57 12.15 1.84 0.12 0.03 2.14 1.37 100.27 299 93

47.81 1.92 13.85 14.41

52.58 0.78 15.24 8.37

46.87 1.20 15.04 14.10

0.20 5.90 8.48 3.65 0.14 0.07

0.14 7.25 8.08 4.63 0.15 0.08

0.15 7.45 10.66 1.70 0.30 0.10

2.80 99.23

0.60 97.9

2.10 99.67

162 1 699 38 2.90 94 2.00

63 0 245 16 1.30 36 1.00

76 3 411 26 1.00 20 1.00

0.30 0.10 4.20 13.50 2.20 12.10 4.30 1.30

0.40 0.10 2.70 7.40 1.20 6.40 2.00 0.80

0.30 0.00 2.40 7.60 1.30 7.60 2.60 0.90

1.00 6.50 4.20 0.60 4.10 0.60

0.50 3.10 0.60 1.90 0.30 1.90 0.20

0.70 4.60 1.00 3.00 0.40 3.00 0.40

24 21 23 22 20 Diabase Grnschist Grnschist Grnschist Diabase GX7 ZH1-44 BB27 BB31 BB40

19 160 52.1 88 10.1 98 22.4 1.76 74.8

7 Basalt T14-3

71 43 50.3 83 10.7 101 22.3 1.95 64.2

0.33 0.315 0.42 4.17 10.5

0.30 0.319 0.36 4.32 10.5

7.78 2.6 0.813

8.97 3.16 0.987

8.35 2.98 0.968

0.682

0.792

0.665

0.27 0.16 0.32 3.99 10.1

3.98 0.64

3.15 0.49

4.37 0.72

1.18 0.175

2.81 0.437

15 Basalt Q5-8

16 Basalt Q5-9

17 Basalt Q5-10

18 Basalt BZ-41

19 Basalt DQ-99

49.81 1.84 16.22 6.22 4.30 0.18 5.47 7.27 3.44 2.00 0.19 2.36 0.84 100.14 154 69 76 100 40 115 44 4.33 155 10.00 0.53 0.36 0.30 4.6 15.5

53.09 1.18 14.78 3.39 4.30 0.16 6.48 6.71 4.59 0.46 0.12 2.88 1.38 99.52 214 61 47 85 6 111 25 2.55 90 5.10 0.18 0.27 0.30 1.76 7.94

48.42 1.03 15.57 5.68 4.26 0.17 7.23 10.25 2.83 0.99 0.09 2.28 0.69 99.49 181 67 68 117 24 117 29 1.96 74 5.20 0.36 0.25 0.59 2.28 7.36

53.1 1.13 16.8 8.5

50.6 1.67 15.9

100 214 94 30 264 12 265 26 2.78 109 36.00 1.46 2.28 0.55 8.23 17

14.90

11.70

6.76

10.12

0.12 8.8 7.2 3.95 0.28 0.14

8 Basalt 1

6 Basalt T14-2

12.3 0.19 8.3 7.6 2.84 0.30 0.23

100 45 34 30

2.93 0.448

2.87 0.437

40 8 111 11 32 5.00

3.11 8.11 1.27 6.01 2.31 1.36 3.61 0.64 3.87 0.91 2.47 0.38 2.34 0.47

47.58 1.97 13.44 2.48 9.36 0.18 8.12 8.76 3.00 0.15 0.19

4.00 99.23 130 341 40 45

42 30 3.47 68 5.80 1.06 0.37 0.36 9.06 20

112 39 3.57 152 3.40 0.20 0.22

13.45

12.00

5.22 14.6

36.9 56.3 27.6 5.7 0.8 78.8 50.8 7.70 153.0 2.70 0.48 0.06 0.27 5.2 19.0 3.7 20.25

136 91 38.0 10.7 0.8 110.5 29.4 5.30 81.3 1.28 0.29 0.39 0.18 3.07 10.13 1.97 10.99

78 68 42.7 8.7 2.0 89.5 31.3 4.90 76.1 2.15 0.34 0.83 0.22 3.04 10.45 1.95 10.61

12 14 13 Boninite Boninite Boninite Q 1-48-2 Ql-130 Ql-43

47.86 0.23 11.06 5.55 3.55 0.16 15.64 9.34 1.61 0.36 0.03 3.36 1.04 99.79 1360 636 68 56 8.65 62 9.1 0.62 25

51.82 0.26 13.68 2.45 3.88 0.16 11.28 11.33 0.10 0.11 0.04 4.01 0.72 99.84 887 292 41 13 6 64 8.6 0.40 28

53.51 0.34 11.76 3.50 4.43 0.14 12.22 5.85 3.05 0.09 0.04 3.88 0.69 99.5 934 254 49 10 1 38 10 0.44 28

0.21

0.16

0.24

0.48 1.72

1 2.45

1.05 2.72

1.54 0.57 0.25 0.15

1.95 0.75 0.34 1.17 0.21

0.22

0.33

0.11 0.76 0.14

0.16 1.02 0.16

2.12 0.81 0.44 1.20 0.23 1.65 0.36 1.10 0.17 1.10 0.17

25 Diabase GX8

26 Diabase GX10

27 Lava Ophl74

28 Lava NI458

56.26 0.22 11.85 7.39

56.45 0.22 12.10 7.45

60.51 0.22 13.89 5.52

49.32 0.84 15.26 8.99

48.20 1.10 14.70 9.31

0.11 11.22 9.30 1.83 0.25 0.04

0.10 10.73 9.60 2.00 0.21 0.05

0.07 6.72 6.75 4.15 0.28 0.05

0.15 7.94 6.99 5.13 0.23 0.16

0.16 5.51 11.00 3.70 0.80 0.00

98.47

98.91

98.16

5.42 100.43 372 84

5.78 100.26 253 81 32.0

41.8 7.4 72.7 5.4 0.60 18.7 0.81 0.52 0.08 0.20 2.16 4.93 0.68 3.17

38.1 4.9 76.7 5.3 0.64 19.9 0.85 0.46 0.07 0.25 2.14 4.89 0.70 3.06

61.3 9.2 91.1 4.6 0.72 20.4 0.88 0.54 0.09 0.12 2.58 5.65 0.74 3.06

128.0 24.0

61.0

1.10 4.30

15.1 120.0 29.0 2.50 94.0 2.20 0.00 0.09

9.40 10.90

OPHIOLITES IN CHINA Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Rock: Sample: Si02 TiO2 A1203 Fe2O3 FeO MnO MgO CaO Na2O K,0 P205 H2O+ LOI Total Cr Ni Co Ba Rb Sr Y Hf Zr Nb Th Ta U La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

34 Basalt B4

35 Basalt Y10

36 Basalt Y15

37 Basalt Y17

38 Basalt D2

39 Basalt D9

50.0 0.48 17.0 9.6

46.9 2.00 15.2 12.8

49.7 2.00 14.5 11.5

50.6 0.72 13.9 9.1

48.2 1.70 14.7 9.8

51.5 1.05 16.3 9.7

51.1 0.86 16.1 9.2

51.7 0.49 16.3 6.1

57.3 0.69 17.2 7.3

57.8 0.84 15.8 9.0

0.16 8.9 8.0 4.20 0.06 0.06

0.21 5.5 14.2 2.56 0.67 0.20

0.18 6.8 11.0 3.85 0.56 0.20

0.17 10.8 9.2 2.66 0.88 0.07

0.17 9.3 9.2 5.42 0.75 0.19

0.13 7.0 6.9 5.11 0.18 0.12

0.14 7.8 7.5 4.69 0.02 0.10

0.12 9.4 11.1 2.42 1.57 0.04

0.09 4.3 6.1 3.96 3.64 0.39

0.14 5.5 7.2 2.89 1.84 0.35

98.9 21 22

98.46 164 58

100.24

100.29

98.1 346 114

99.43 31 17

97.99 182 54

97.51 107 47

99.24 647 107

100.97 60 19

101.36 54 20

13 1.7 123 11 0.48 13 0.18 0.04 0.09 0.02 0.4 1.1 0.3 1.77 0.86 0.40 1.36 0.25 1.91 0.43 1.30 0.19 1.38 0.20

5 3.7 81 11 0.74 23 0.32 0.02 0.05 0.02 0.7 2.4 0.5 2.96 1.07 0.50 1.54 0.28 1.97 0.44 1.29 0.19 1.38 0.18

57 4.0 132 32 2.20 58 5.39 0.41 1.88 0.18 6.7 17.0 2.7 13.40 4.09 1.59 5.61 0.90 6.06 1.25 3.68 0.51 3.34 0.49

35 11.0 57 35 2.23 56 2.10 0.19 1.06 0.09 4.4 13.0 2.4 12.60 4.17 1.56 6.02 0.99 6.58 1.38 4.15 0.57 3.69 0.53

7 5.4 195 16 1.03 34 0.58 0.03 0.18 0.01 1.2 4.1 0.8 4.68 1.74 0.62 2.42 0.43 2.94 0.65 1.88 0.27 1.91 0.26

45 8.9 220 43 3.82 137 2.02 0.13 0.23 0.35 4.4 14.4 2.7 15.23 5.28 1.73 6.67 1.15 8.65 1.61 4.71 0.69 4.77 0.66

40 1.7 180 24 2.30 74 1.11 0.07 0.19 0.06 2.5 8.2 1.5 8.31 2.90 1.04 3.77 0.65 4.49 0.97 2.73 0.39 2.71 0.37

3 0.1 98 21 1.51 59 0.45 0.06 0.12 0.03 1.9 6.3 1.2 6.38 2.28 0.84 3.10 0.55 3.77 0.80 2.32 0.33 2.27 0.33

12 5.7 272 12 0.68 23 0.36 0.02 0.10 0.02 1.0 2.7 0.6 3.46 1.35 0.54 1.88 0.35 2.35 0.49 1.40 0.21 1.38 0.19

868 49.0 712 17 1.98 151 2.68 3.88 0.41 0.86 24.2 57.4 8.2 39.50 7.73 1.74 5.63 0.64 3.42 0.61 1.63 0.21 1.45 0.18

226 27.0 619 19 1.33 107 2.59 2.83 0.56 0.65 14.8 34.9 5.0 22.60 5.20 1.34 4.53 0.64 3.82 0.76 2.09 0.29 1.99 0.28

5.06 0.69

2.89 0.45

3.10 0.19

2.58 0.39

2.84 0.39

30 29 Diabase Diabase B219 B218

31 Basalt MRZ4

32 Basalt MRZ9

53.95 0.35 13.05 3.12 6.12 0.17 9.09 7.56 3.55 0.15 2.65 0.25 0.03 100.04 577 200

52.5 0.39 16.8 10.5 0.21 6.9 6.1 5.18 0.29 0.03

1.61 0.56 0.24

2.69 0.94 0.38

0.25

0.30

1.62 0.19

1.46 0.26

2.81 0.50

33 Basalt Bl

0.73

1.31 3.53

1.87 0.28

5.32 0.78

0.73

4.92 1.75 4.05 1.24

0.63 2.03

0.83

0.78 0.25 0.83 0.14 0.97 0.20 0.62 0.10 0.72 0.11

3.75 1.31

19.0

0.54

0.83 0.24 0.92 0.15 0.97 0.20 0.70 0.10 0.77 0.11

3.21 1.14

18.0

3.10 1.10

4.47 1.78 6.19 0.90 6.80 1.76 4.34 0.88 5.74 0.75

2.62 1.27 3.89 0.72

85.0 14.0

1.56 0.76

5.32 1.93 5.83 1.10 6.79 1.62 4.31 0.78 5.00 0.79

2.95 1.01 4.31 0.88

15.0 13.0

0.71 0.21 0.82 0.12 0.82 0.18 0.58 0.09 0.70 0.10

7.70 2.12 10.37 2.07 11.08 2.36 7.51 1.27 8.92 1.31

5.06 1.78 8.18 1.49

53.78 0.30 11.88 2.57 7.54 0.18 8.33 9.45 3.50 0.21 2.63 0.19 0.03 100.59 681 230

545

41 40 Andesite Andesite MZ8 MZ3

1-3, Songshugou ophiolite, Shanxi Province, Late Proterozoic (Zhou et al. 1998); 4-7, Tangbale ophiolite, West Junggar belt (Yang & San 1992); 8-10, Hegenshan ophiolite, Inner Mongolia (Robinson et al. 1999); 11, Kudi ophiolite, West Kunlun belt (Yang et al. 1996); 12-14, Dachadaban ophiolite, North Qilian belt (Zhang et al. 1997); 15-17, Jiugequan ophiolite, North Qilian belt (Zhang et al. 1997); 18 and 19, Deqin ophiolite; 20, Shuanggou ophiolite (Zhong 2000); 21-23, Babu ophiolite (Zhong et al. 1998); 24-26, Dingqing ophiolite, Bangong Lake-Nujiang suture zone (Liu et al. 2002); 27-30, Xigaze ophiolite (Zhang et al. 1982; Yang & Han 1992); 31 and 32, Zedang ophiolite, Yarlung Zangbo suture zone (Malpas et al. 2003); 33 and 34, Luobusa ophiolite, Yarlung Zangbo suture zone (Malpas et al. 2003); 35-37, Xigaze ophiolite, Yarlung Zangbo suture zone (Malpas et al. 2003); 38 and 39, Dazhuqu ophiolite, Yarlung Zangbo suture zone (Malpas et al. 2003); 40 and 41, Zedong andesites, Yarlung Zangbo suture zone (Malpas et al. 2003). Blank space means not determined.

of the North Qinling Mountains may also be part of this system. Early Palaeozoic ophiolites are common in the Chinese part of the Palaeo-Asian system, where they are associated with a few late Palaeozoic bodies (Fig. 5). The late Palaeozoic bodies represent the final stage of the Palaeo-Asian Ocean, which did not close completely until early Permian.

The most important Palaeo-Asian ophiolites occur in West Junggar, East Junggar, Tianshan, Biashan and Inner Mongolia.

West Junggar ophiolite belt Ophiolites in the West Junggar portion of the Palaeo-Asian system include Tangbale (of mid-

546

Q. ZHANG ETAL.

Fig. 3. Chondrite-normalized REE and MORB-normalized spider diagrams for ophiolites (chondrite and MORE values are after Sun & McDonough 1989). (a) and (b), amphibolites of Songsugou ophiolite, Shanxi Province, Late Proterozoic; (c) and (d), lavas of Tangbale from the Western Junggar ophiolite belt (Yang & Han 1992) and Kudi ophiolite from the Western Kunlun ophiolite belt (Yang et al. 1996); (e) and (f), lavas from Dachadaban and Jiugequan ophiolites from the North Qilian ophiolite belt (Zhang et al. 1997); (g) and (h), lavas and basalts from Deqin, Shuanggou (Zhong 2000) and Babu (Zhong et al. 1998) ophiolites (which all belong to the Palaeo-Tethyan belt); (i) and (j), diabases of Dingqing (Liu et al. 2002) and Xigaze (Zhang et al. 1982; Yang & Han 1992) ophiolites (Neo-Tethyan ophiolite belt).

Ordovician age), Mayile (mid-late Silurian), Darbut (early-mid-Devonian) and Honggulaleng (Ordovician) (Fig. 6). The Tangbale ophiolite, which is representative of this belt, consists of a melange of metaperidotite, cumulate ultramafic rocks, gabbro and pillow lava. Diabases and basalts in this body have widely varying TiC>2 contents (0.703.00wt%) and chondrite-normalized rare earth

element (REE) patterns, ranging from light REE (LREE) depleted (LaN/YbN <1) to LREE slightly enriched (LaN/YbN >2) to LREE strongly enriched (LaN/YbN >3) (Table 1; Figs 3c, d and 4). The LREE-depleted basalts are of N-MORB character, whereas those slightly enriched in LREE are depleted in high field strength elements (HFSE) (e.g. Th > Ta; Th/Ta - 2.7-3.9; Zhu et al 1987;

OPHIOLITES IN CHINA

547

Fig. 4. Ti-Zr (Pearce & Cann 1973) and Ta-Th-Hf (Wood 1980) diagrams. Legend as in Figure 3. OFB, ocean flow basalt; LKT, low-K tholeiite; CAB, calc-alkaline basalt; WPB, within plate basalt; WPT, within plate tholeiite; VAB, volcanic arc basalt.

Xiao et al. 1992) and are more characteristic of island-arc basalts. The LREE-enriched basalts also have high Ti and P, and are probably ocean island lavas (Xiao et al. 1992). It is generally agreed that the Tangbale ophiolite formed in a forearc or back-arc oceanic basin

(Xiao et al. 1992). This part of the ocean basin opened originally in the late Cambrian-early Ordovician and closed in the mid- to late Ordovician. The emplacement age of the Tangbale ophiolite is considered to be later than early Silurian (Xiao et al. 1992).

548

Q. ZHANG ET AL.

Fig. 5. A map illustrating the general geology and distribution of ophiolites in the Palaeo-Asian tectonic belt of northern China (after Tang & Shao 1996).

East Junggar ophiolite belt The East Junggar ophiolites include Kalamaili, Aermantai, Kekesentao-Qiaoxiahala and Kuerti (Fig. 7). The Kalamaili ophiolite crops out on the eastern margin of the Junggar Basin, and consists of serpentinite (altered mainly from harzburgite), gabbro and lava. The various rock units are in fault contact so there is no continuous section exposed. Basalts in this body have moderate to high TiO2 contents (1.31-2.73%). Chondritenormalized REE patterns show both LREE depletion and slight enrichment (Li 1991, 1995). The oceanic basin from which the Kalamaili ophiolite was derived probably formed in the early Devonian and closed in the early Carboniferous (Li 1991, 1995).

Tianshan Ophiolite Belt Four ophiolite sub-belts have been recognized in the Tianshan region (Gao et al. 1998), which from north to south consist of: (1) late Palaeozoic ophiolites of the north Tianshan (represented by the Bayingou ophiolite); (2) early Palaeozoic ophiolites in the northern part of the middle Tianshan; (3) early Palaeozoic ophiolites in the

southern part of the middle Tianshan; (4) late Palaeozoic ophiolites of the south Tianshan (Fig. 8). The Changawuzi ophiolite is representative of the early Palaeozoic ophiolites in the southern part of the middle Tianshan. It has been extensively metamorphosed under greenschist-, blueschistand eclogite-facies conditions. The metabasic rocks mostly have E-MORB compositions, although a few show LREE depletion characteristic of N-MORB, suggesting formation in a backarc basin (Tang et al 1995). 40Ar/39Ar ages of phengite and glaucophane from the blueschists range from 315 to 415 Ma, suggesting formation near the end of the late Silurian (Tang et al, 1995). Basalts in the late Palaeozoic ophiolites of the south Tianshan are markedly depleted in Ti, P and LREE, and have low Zr/Y ratios (2.0-2.5). Their eNd(0 values range from +6.8 to +7.6 (where t is 334 Ma) (Gao et al. 1998), indicating formation in a suprasubduction zone setting.

Inner Mongolian ophiolite belt There are two major ophiolite sub-belts in Inner Mongolia (Fig. 9): Ondor Sum-Kedanshan-Xilamulun River (early Palaeozoic) and Solonshan-

OPHIOLITES IN CHINA

549

Fig. 6. The distribution of West Junggar ophiolites (after Bai et al. 1995).

Hegenshan (late Palaeozoic) (Bao et al. 1994; Robinson et al. 1999; Nozaka & Liu 2002). Ondor Sum ophiolites occur in the lower part of the Ondor Sum Group and consist of metabasalt and metagabbro with minor diorite locally intruded by diabase dykes. A few lenses of metaperidotite occur in the gabbro. The ophiolites are associated with high-pressure glaucophane-lawsonite schists. The volcanic rocks in Kedanshan have TiO2 contents ranging from 0.75 to 2.20 wt% and averaging 1.49wt%. The lavas have both LREE-depleted and LREE-slightly enriched patterns, similar to those of N-MORB and E-MORB, respectively. The Hegenshan ophiolite is located along the boundary of the Siberian and North China Blocks in Inner Mongolia. The ophiolite has a Sm-Nd isochron age of 403 ± 27 Ma, with eNd(7) of +8.7 ± 0.6 (Bao et al. 1994) suggesting formation in the Devonian or early Carboniferous. The Hegenshan ophiolite consists of several blocks composed chiefly of serpentinized ultramafic rocks with lesser amounts of troctolite and gabbro and sparse lavas and dykes. The ultramafic rocks consist chiefly of depleted harzburgite and minor dunite and are interpreted as mantle tectonites. In

the Hegenshan block dunite is relatively abundant and is typically associated with podiform chromitite. Several blocks have well-layered cumulate sequences of gabbro and troctolite. Sheeted dykes are absent but small mafic dykes are common in some of the ultramafic sections (Robinson et al. 1999). Basaltic rocks from Hegenshan are chiefly tholeiitic in character (Table 1) with flat REE patterns showing slight LREE depletion. Nozaka & Liu (2002) suggested the ophiolite formed at a typical mid-ocean ridge whereas Robinson et al. (1999) postulated formation in a suprasubduction zone environment, probably a back-arc basin. A few basalts are enriched in LREE and may be ocean island lavas (Robinson et al. 1999).

Qinling-Qilian-Kunlun ophiolites West Kunlun ophiolites The Kudi ophiolite is typical of the west Kunlun belt. The ophiolite consists of peridotite, minor cumulate gabbro, diabase dykes and thick-bedded basalts and basaltic andesites, associated with chert, abyssal turbidites and sea-floor olistoliths (Jiang et al. 1992; Yang et al. 1996; Wang et al.

550

Q. ZHANG ETAL.

Fig. 7. The distribution of East Junggar ophiolites (after He et al 1990).

2001b). Yang et al. (1996) suggested that the Kudi ophiolite was formed during Sinian and Cambrian time, and that the ocean closed at the end of the Silurian. A slightly younger age (OrdovicianSilurian) is suggested by recently discovered radiolaria in the Kudi melange (Zhou et al. 1998) Spinels in the Kudi peridotite are relatively high in Cr (Cr number 0.57-0.67, Yang et al 1996). Basalts are low in HFSE, and plot in the islandarc field on a Y-Cr diagram (Jiang et al. 1992; Yang et al. 1996). They have MORB-normalized spider diagrams similar to those of basalts from Tonga, confirming a suprasubduction zone origin (Figs 3c, d and 4; after Yang et al. 1996). Boninite is also present in the Kudi ophiolite and is characterized by high Mg (MgO 8.57-10.58 wt%,

Mg number 0.62-0.72) and Cr (299-516 ppm) and low HFSE and heavy REE (HREE). TiO2 is <0.40wt%, P205 0.01-0.03 wt%, Y 7-11 ppm, Zr 15-32 ppm, and Yb is 4-7 times chondritic (Table 1) (Jiang ef al 1992). North Qilian ophiolites The North Qilian ophiolites are all early Palaeozoic in age. They occur in four sub-belts in the Qilian area of Qinghai Province and the Sunan area of Gansu Province, and include, from south to north, the Yushigou ophiolite, Bianmagou ophiolite, Baijing Temple ophiolite, Dachadaban ophiolite and Jiugequan ophiolite (Figs 10 and 11, Zhang etal 1997).

OPHIOLITES IN CHINA

551

Fig. 8. The general geology and distribution of ophiolites in the west Tianshan (after Gao et al. 1998). NMS, suture zone on the north side of the middle Tianshan; SMS, suture zone on the south side of the middle Tianshan; NTF, fault zone on the north side of the Tarim Basin. Ophiolite outcrops: 1, Gangou; 2, Mishigou; 3, Wusitegou; 4, Luweigou; 5, Bayingou; 6, Changawuzi; 7, Guluogou; 8, Yushugou; 9, Tonghuashan; 10, Liuhuangshan; 11, Misibulake; 12, Kule lake; 13, Dushanzi-Kuche Highway 965 km; 14, Serikeyayilake; 15, Keketiekedaban; 16, Tangbale.

The Dachadaban ophiolite is characterized by having both boninitic and tholeiitic lavas (Yang et al 1991 \ Zhang et al 1997, 1998). It is well exposed in the Dachadaban area of the North Qilian Mountains, where it consists, from the base upward, of meta-peridotite, gabbro-diabase and pillow lava. The ophiolite is considered to be early Palaeozoic in age and is unconformably overlain by coarse-grained clastic rocks in the NE. Many of the pillow lavas are now actinolite schists with very low FeO/MgO ratios, low TiO2, Zr and Y, and depleted HREE (Table 1), suggesting that the protoliths were boninites. The geochemistry suggests formation in an arc-forearc environment (Figs 3e, f and 4; Zhang et al 1997, 1998). The Jiugequan ophiolite (Figs 10 and 11) consists of meta-peridotite, gabbro-diabase and ba-

saltic pillow lava. The basalt is geochemically similar to N-MORB (Table 1) (Figs 3e, f and 4; Zhang etal. 1997).

Palaeo-Tethyan ophiolites The Tethyan Ocean was present in SW China from Devonian to Tertiary time. Some workers suggest that Tethys can be divided into two parts (Zhong 2000), but others have proposed a threepart division: Palaeo-Tethys, Meso-Tethys and Ceno-Tethys (Metcalfe 1999). In this paper, we adhere to a two-part division: Palaeo-Tethys and Neo-Tethys. Palaeo-Tethyan ophiolites occur on the NE and SE sides of the Tibet Plateau in the ChangningMenglian, Jinshajiang-Ailaoshan, Ganzi-Litang

552

Q. ZHANG ETAL.

Fig. 9. The general geology and distribution of ophiolites in Inner Mongolia (after Tang & Shao 1996).

and Animaqin-Mianlue ophiolite belts (Fig. 12). The Changning-Menglian and Jinshajiang-Ailaoshan ocean basins opened in late Devonianearly Carboniferous time and closed before the late Triassic. Late Triassic molasse deposits unconformably overlie the ophiolites (Zhang et al. 1998; Wang et al 2001a). The Ganzi-Litang oceanic basin was the latest to open and had the shortest duration, from early to late Triassic (Zhang et al 1992, 1994).

Jinshajiang-Ailaoshan ophiolite belt The Jinshajiang-Ailaoshan ophiolite belt lies between the Yangtze craton and the Simao block (Fig. 12). The Shuanggou ophiolite lies in the middle of the belt and consists of three main lithologies, which are, from the base upward, meta-peridotite (mainly plagioclase-bearing Iherzolite with minor spinel Iherzolite and harzburgite and rare dunite), gabbro-diabase and basalt (Fig.

OPHIOLITES IN CHINA

Fig. 10. The distribution of ophiolites in the North Qilian Mountains (after Zhang et al. 1997).

Fig. 11. Cross-section of North Qilian ophiolites (after Zhang et al. 1997).

553

554

Q. ZHANG ETAL.

Fig. 12. The distribution of Palaeo- and Neo-Tethyan ophiolites. Palaeo-Tethyan ophiolite zone: ALS, Ailaoshan ophiolite belt; JSJ, Jinshajiang ophiolite belt; CN-ML, Changning-Menglian ophiolite belt; GZ-LT, Ganzi-Litang ophiolite belt; MJ, Majiang ophiolite belt; MQ-ML, Maqing-Meanlue ophiolite belt; NH, Nanhe ophiolite belt; WDLW, Wendong—Laowu ophiolite belt. Neo-Tethyan ophiolite zone: BL-N, Bangong Lake-Nujiang ophiolite belt; NJS, Najiashan ophiolite belt; YLZB, Yarlung Zangbo ophiolite belt.

13). The basalt and diabase fall into two chemical types, one poor in Al (A^Os <16wt%) and depleted in LREE (LaN/YbN 0.51-0.79) (Table 1), the other enriched in Al (A12O3 >16%) and LREE (LaN/YbN 1.9-3.0), similar to N-MORB and EMORB, respectively (Figs 3g, h and 4). Their eNd(0 values are +9.3 to +7.9 (Zhong 2000). Jian et al (1998, 1999) obtained U-Pb ages of 362 ±41 Ma for gabbro and 328 ±16 Ma for plagiogranite in the Shuanggou ophiolite, suggesting a Devonian-Carboniferous age. They also reported ages of 340 ± 3 Ma and 294 ± 4 Ma for plagiogranite in the associated Jinshajiang ophiolite. Red conglomerates of the Yuweanshui Formation of late Triassic age unconformably overlie both of these ophiolites (Zhang & Zhou 2001).

The Babu ophiolite lies in the Malipo area along the boundary between Yunnan Province and Vietnam (Fig. 12). It is interpreted as a tectonic slice thrust northward from Vietnam (Wu et al. 1999; Zhong 2000). The ophiolite is composed mainly of metaperidotite, gabbro and basalt, with a few diabase dykes in the gabbro (Fig. 14). The Babu ophiolite is similar to the Shuanggou ophiolite in geochemistry (Table 1) and age (Wu et al. 1999; Zhong 2000).

Neo-Tethyan ophiolites Neo-Tethyan ophiolites occur mainly in SW China, particularly in southern Tibet, along the Yar-

OPHIOLITES IN CHINA

555

Fig. 13. Geological map and cross-section of the Shuanggou ophiolite section, Yunnan Province. For abbrevations see Fig. 12.

lung Zangbo (YLZB) and Bangong Lake-Nujiang suture zones (Fig. 12). Yarlung Zangbo ophiolite belt (YLZB) The Yarlung Zangbo belt marks the boundary between the Indian and Lhasa blocks in Tibet (Wu & Deng 1980). There are two major ophiolites in this belt, Xigaze in the west and Luobusa in the east, as well as a few small fragments such as that near Jedang (McDermid et al. 2002). The Xigaze is the largest and best-developed ophiolite in China. It forms a massif averaging about 10 km thick, which has been thrust over weakly metamorphosed sedimentary rocks of the Himalayan Tethyan province. To the north, it is

unconformably overlain by clastic rocks of the Congdui Group, which contain ophiolitic detritus. Based on early biostratigraphic and isotopic dating, the Xigaze ophiolite was estimated to have formed at c. 110-120 Ma and been emplaced at c. 40 Ma (Gopal et al. 1984; Pearce & Deng 1988; Wan et al. 1998). Recent sensitive high-resolution ion microprobe (SHRIMP) dating of zircon from the Dazhuqu massif indicates an age of formation of 126 ± 2 Ma (Malpas et al. 2003), which is in agreement with a Barremian biostratigraphic age reported by Zyabrev et al. (2002). The YLZB continues westward into the Indus Suture Zone in eastern Lakakh, where there is a late Jurassic (141 Ma) MORB-type ophiolite (the Nidar ophiolite).

556

Q. ZHANG ETAL.

Fig. 14. Geological map and cross-section of the Babu ophiolite, Yunnan Province.

The Xigaze ophiolite consists primarily of peridotite with minor cumulate rocks, gabbro, plagiogranite, sheeted dykes and pillow lavas. The sequence is overlain by red radiolarian chert and clastic rocks containing ophiolitic debris. The reconstructed lithological sequence is comparable with many well-known ophiolites such as Troodos and Semail. The sheeted dykes occur in the eastern part of the Xigaze ophiolite in the Decun area. This complex consists of nearly 100% dykes with only a few pillow screens. Individual dykes range from several centimetres to 4 m wide and have both symmetric and asymmetric chilled borders. Where pillows are present, the dykes intersect them at a

high angle (Fig. 15). The dykes consist of both island-arc tholeiite and boninite. The boninites have 48.9-54.0 wt% SiO2, low TiO2 (0.210.43 wt%), and high MgO (7.61-13.26 wt%), Cr (335-1020 ppm) and Ni (91-400 ppm). All are depleted in HFSE and HREE (Zr 5-15 ppm, Y 5-16 ppm) and have low Ti/V ratios (6-12) (Zhang et al 1982) (Figs 3i, j and 4). Lavas from the Xigaze ophiolite are arc tholeiites (Table 1) with concentrations of large ion lithophile elements up to 10 X MORE and clear depletion in Nb. These compositions are characteristic of suprasubduction zone volcanism. Wang et al (2000) and Hebert et al (2003) also postulated a suprasubduction origin for this ophio-

O P H I O L I T E S IN CHINA

557

Fig. 15. A cross-section sketch of the sheeted dyke swarm in the Xigaze ophiolite of Tibet (after Zhang et al. 1982).

lite based on the mineral chemistry of the mafic and ultramafic rocks. The Luobusa ophiolite, which lies about 200 km ESE of Lhasa, consists of a slab at least 1 km thick. It has been thrust northward onto the Tertiary Luobusa Formation and the Gangdese batholith and is technically overlain by Triassic flysch deposits on the south. The northern thrust contact dips southward at angles of 15-20° whereas the southern contact is much steeper. The sandstones and conglomerates of the Luobusa Formation are locally deformed, particularly along the thrust contact. Luobusa was originally thought to have formed in the late Cretaceous but a recent Sm/Nd wholerock age of 170 Ma (cited by Malpas et al. 2003) suggests that it is significantly older. Ar-Ar ages of metamorphic rocks in the Luobusa ophiolite suggest emplacement between 70 and 90 Ma (Malpas et al. 2003). From the base upward, the ophiolite contains a melange zone, a dunite transition zone sequence and a mantle sequence. The boundaries between the major units generally dip southward, suggesting that each unit is an individual thrust slice. The mantle sequence makes up the bulk of the ophiolite and is composed chiefly of harzburgite and clinopyroxene-bearing harzburgite with abundant lenses and pods of dunite and chromitite (Zhou et al. 1996;Hebert et al. 2003). Dunite is the most abundant rock type in the transition zone whereas the underlying melange zone consists of lenses of wehrlite, pyroxenite, layered and homogeneous gabbro, and pillow lava in a strongly serpentinized, ultramafic matrix. The sparse pillow lavas in the melange zone have been metamorphosed mostly to the greenschist facies but some are now amphibolites with rare relict volcanic textures. The lavas generally have an arc tholeiite composition (Table 1), similar to those elsewhere in the Yarlung Zangbo suture zone. However, the very high Cr content of the podiform chromitites in Luobusa indicates that they were precipitated from boninitic magmas (Zhou et al. 1996), suggesting a

two-stage magmatic evolution. Malpas et al. (2003) interpreted the Luobusa ophiolite as the basement of the mature arc sequence recognized in the adjacent Zedong terrane (McDermid et al. 2002).

Bangong Lake-Nujiang ophiolite belt The Bangong Lake-Nujiang suture runs through central Tibet and separates the Lhasa Block from the Qiangtang Block (Fig. 12). Ophiolites along this belt formed in the Jurassic and were emplaced at the end of the mid-Jurassic (Girardeau et al. 1984, 1985; Girardeau & Mercier 1988; Pearce & Deng 1988; Zhou et al. 1997). However, the oceanic basin did not close completely in the Bangong Lake area until late Cretaceous time. The Dingqing ophiolite, in the eastern part of the belt, is a north-dipping sheet that consists, from the base upward, of harzburgite, cumulate orthopyroxenite, quartz norite and dykes (Fig. 16). Radiolarians in chert overlying the ophiolite are early Jurassic to late Triassic in age (Zhang et al. 1994; Liu et al. 2002). A sequence of Middle Jurassic sandstone and conglomerate unconformably overlies the ophiolite. Thus, the Dingqing ophiolite was formed in the late Triassic-early Jurassic and emplaced before the middle Jurassic. Lavas within the Dingqing ophiolite have boninitic characteristics including enrichment of Si, Mg and LILE and depletion of HFSE (Ti, P, Zr, Y, Yb and Nb) (Table 1). Chondrite-normalized REE patterns of the gabbro and diabase are all U-shaped (Figs 3i, j and 4), implying that the boninites were derived from a depleted mantle source modified by slab-derived fluids (Zhang et al. 1994; Liu et al. 2002). The geochemical characteristics of these boninites are similar to those of Tertiary boninites in the Western Pacific (Meijer 1980;Hickey & Frey 1982; Crawford et al. 1989), indicating that the Dingqing ophiolite formed in a forearc setting. The Donqiao ophiolite in the central part of the Bangong Lake-Nujiang suture zone consists entirely of mantle peridotite with pods and lenses of

558

Q. ZHANG ETAL.

Fig. 16. Geological map and cross-section of the Dingqing ophiolite, Bangong Lake-Nujiang suture zone.

chromitite (Girardeau et al. 1984; Pearce & Deng 1988). Spinels in the peridotite have relatively high Cr numbers, suggesting a suprasubduction zone origin. The ophiolite is underlain by a metamorphic sole several metres thick, which suggests emplacement at the end of the midJurassic (Zhou et al 1997).

Circum-Pacific ophiolites East Taiwan ophiolite The east Taiwan ophiolite occurs in the Lichi melange belt of the Coastal Range of east Taiwan

(Fig. 17) and is the one of the youngest known ophiolites in the world. It represents relict ocean crust that formed in the South China Sea at c. 15 Ma and that was emplaced at c. 3-4 Ma by arc-continent collision (Chung & Sun 1992). The mantle peridotite is chiefly harzburgite with minor dunite and plagioclase-bearing Iherzolite (Liou et al. 1977). Gabbro is overlain directly by basalt with a chilled margin at the contact. Layers of red pelagic shale occur locally between the basalt and gabbro. The basalts contain abundant Mg-rich glass and have Pb isotopes with a DUPAL anomaly (Chung & Sun 1992). Chung & Sun (1992) considered the east Taiwan ophiolite to have

OPHIOLITES IN CHINA

559

Discussion The tectonic setting of ophiolites in China Here we attempt to identify the tectonic settings in which the various Chinese ophiolites were formed based on their geochemistry and regional geological relationships. This approach relies heavily upon comparisons of ophiolitic rocks with those in modern tectonic environments. Although not universally accepted, this approach has yielded many insights into the origin and emplacement of ophiolites (e.g. Shervais 1990, 2001) and is the best guide currently available. Most Chinese ophiolites appear to have been formed or assembled in suprasubduction zone environments, such as island arcs, forearcs, backarcs, or in intracontinental oceanic basins (Fig. 18). Many workers in China still interpret ophiolites as fragments of normal mid-ocean ridge lithosphere but that view is changing gradually as more detailed studies are undertaken.

Fig. 17. Major geological units and distribution of ophiolites in Taiwan (after Ho 1977).

formed at a normal slow-spreading axis where serpentinized peridotites and gabbros were exposed on the ocean floor.

Kaishantun ophiolite The Kaishantun ophiolite is exposed in the Yanbian area of Jilin Province (Fig. 2), and occurs as an ophiolite melange (Shao & Tang 1995). Blocks in the melange consist mainly of harzburgite and Iherzolite (with scattered pod-like bodies of chromitite), gabbro, anorthosite, diabase, basalt, picrite and chert, with lesser amounts of muddy sandstone and bioclastic limestone. The matrix of the melange includes muddy sandstone, sericitequartz schist, metasiltstone and greenschist (including actinolite-chlorite schist, actinolite-epidote-albite schist and amphibole schist). The entire melange has been intensely sheared and deformed.

Suprasubduction zone ophiolites in China. The ophiolites in the Palaeo-Asian, Qinling-QilianKunlun, Neo-Tethyan and Circum-Pacific systems all have suprasubduction zone (Pearce 1982) signatures although many also contain other material. The Tangbale, Mayile, Darbut and Hongguleleng ophiolites of west Junggar, and the Kalamaili and Aermantai ophiolites of east Junggar all have suprasubduction characteristics (Fig. 18). Of these, Tangbale is the most compositionally complex, with mixtures of N-MORB, island-arc tholeiite (IAT) and ocean-island basalt (OIB) lava. In most of these bodies, the cumulate rocks belong to the peridotite-pyroxenite-gabbro (PPG) series characteristic of island-arc assemblages. The occurrence of OIB lavas in Tangbale suggests the existence of seamounts. The Qiaoxiahala ophiolite in east Junggar has boninitic lavas whereas the Kuerti ophiolite contains N-MORB, with IAT and Mg-rich dacite. Thus, these are also interpreted as suprasubduction zone ophiolites that probably formed in a forearc (Fig. 18). In Inner Mongolia, Ondor Sum lavas are mainly E-MORB in character whereas Solonshan has boninitic lavas. The Hegenshan ophiolite has lavas and dykes with geochemical characteristics between MORB and IAT (Robinson et al 1999; Nozaka & Liu 2002), slightly depleted peridotites, and peridotite-troclolite-gabbro (PTG) type cumulates. The Inner Mongolia fold belt between the North China and Siberian blocks contains several belts of ophiolite melange (Fig. 5). These have been interpreted as island-arc or magmatic-sedimentary wedges accreted to an active continental

560

Q. ZHANG ET AL.

Fig. 18. A map showing the presumed tectonic environments of formation for the major ophiolites of China.

margin (Robinson et al. 1999). On the other hand, Nozaka & Liu (2002) postulated a mid-ocean ridge origin for the Hegenshan ophiolite. Adakites have been reported in the region, supporting the subduction zone model. Based on the available evidence, we suggest that all of the Inner Mongolian ophiolites formed in suprasubduction zones, probably in forearcs or back-arc basins (Fig. 18). The Kudi ophiolite in the Western Kunlun Mountains has MORB, IAT, calc-alkaline basalt (CAB) and boninitic lavas, clearly indicating formation in a variety of suprasubduction zone environments (Yang et al. 1996). Likewise, the North Qilian ophiolites contain both N-MORB and boninite lavas, suggesting formation in backarc and forearc settings (Zhang et al. 1997). This interpretation is consistent with the regional geology of the North Qilian erogenic belt, which consists of ophiolites, island-arc complexes with boninites and adakites, accretionary wedges, blueschists and eclogites. Farther south in the Bangong Lake-Nujiang belt, the Dingqing massif has typical boninitic lavas and is classified as a forearc ophiolite. The few other ophiolites in this belt with volcanic rocks have IAT, as well as boninite and MORB

lavas. Cumulate rocks in these bodies belong to the PPG series. Interpretations of the Yarlung Zangbo ophiolites are more complex. The Xigaze ophiolite contains mostly IAT and boninite lavas as well as PPG-type cumulates. We consider it to be a suprasubduction zone ophiolite that probably formed in an island arc. The Luobusa ophiolite farther east consists chiefly of harzburgite and clinopyroxene-bearing harzburgite with abundant podiform chromitites. The chromitites are high-Cr varieties believed to have formed from boninitic melts reacting with the harzburgite (Zhou et al. 1996). The ophiolite also contains a few IAT lavas and sparse PPG-type cumulate rocks. It is considered to be a suprasubduction zone ophiolite (Fig. 18) and may represent the basement of the volcanic arc in the adjacent Zedong terrane (Maipas et al. 2003). Cenozoic ophiolites in Lichi and Kenting from Taiwan have N-MORB and P-MORB lavas and are interpreted as fragments of oceanic lithosphere from the South China Sea. The modern South China Sea is a back-arc or marginal basin. Intracontinental oceanic basins. Palaeo-Tethyan ophiolites all appear to have been formed in intracontinental oceanic basins (Fig. 18), similar

OPHIOLITES IN CHINA to the Ligurian ophiolites in Italy (Ottonello et al. 1984; Tribuzio et al. 1996) and the Red Sea (Dupuy et al. 1991). In this paper, we use the term intracontinental ocean basin to refer to a basin floored only by MORB lithosphere in which there are no subduction zones. All of the lavas in the Palaeo-Tethyan ophiolites are MORB-like in character. For example, the Tongchangjie, Menglian, Babu and Animaqin ophiolites contain N-MORB and/or E-MORB (Table 1) (Fig. 3g and h) (Zhang et al. 1992, 1994). Island-arc tholeiites and boninites have not been found in any of these ophiolites, suggesting formation well away from subduction zones. Finally, all of the ophiolites are bounded on one or both sides by Precambrian basement (e.g. Changning-Menglian belt), or passive or active continental margins with Precambrian basement.

The temporal-spatial evolution of ophiolites in China A paucity of reliable age dates and palaeomagnetic data makes it difficult to constrain the tectonic and palaeogeographical evolution of China. However, increasingly detailed information on the age and origin of Chinese ophiolites allow us to make some first-order observations. The map shown in Figure 19 is a schematic representation of the various blocks that make up modern-day China (shown in red pattern) and the location of the ophiolite belts discussed in the text (bold black lines). Proterozoic ophiolites occur only in Qiling and southern Anhui-NE Jiangxi (Fig. 19a), where they probably represent early rifting of Rodinia. Very little work has been carried out on these ophiolites and it is not clear whether the two belts were linked in the Proterozoic. Dobretsov et al. (1995) and Amelin et al. (1997) suggested that the Palaeo-Asian Ocean extended over a large area in mid-Asia during the late Proterozoic and it was clearly extensive in Cambrian-Ordovician time (Fig. 19b). The Qinling-Qiling-Kunlun ocean between the Tarim and North China Blocks was also extensive at that time because there are numerous boninites, adakites and arc volcanic rocks of early Palaeozoic age in the western Kunlun and northern Qilian Mountains. Thus, it appears that the Tarim and North China Blocks were small continental fragments drifting within the large Palaeo-Asian Ocean. The available information indicates that the Qinling-Qilian-Kunlun Ocean mostly closed before the end of the Ordovician, leaving only a small remnant ocean in the Silurian. However, there are abundant arc volcanic rocks and asso-

561

ciated mineralization related to Silurian-Devonian subduction in the Palaeo-Asian Ocean (Fig. 19c), and the ocean persisted at least to the early Carboniferous (e.g. Bayingou ophiolite in Xinjiang). Beginning in the late Devonian the PalaeoTethyan Ocean opened in south China, and the Palaeo-Pacific Ocean extended to the vicinity of Jinlin and Helongjiang Provinces (e.g. Kaishantun and Yilan ophiolites). Thus, the Palaeo-Asian, the Palaeo-Tethyan and Palaeo-Pacific oceans may have been connected from the late Devonian to early Carboniferous (Fig. 19d). After the early Carboniferous, the Palaeo-Asian Ocean disappeared and the Palaeo-Tethyan and Palaeo-Pacific Oceans became separated. The Palaeo-Tethyan Ocean probably reached its maximum extent in the Permian and then closed before the late Triassic. The Ganzi-Litang oceanic basin was a very small feature probably representing the last remnant of the Palaeo-Tethyan Ocean (Fig. 19e). Beginning in late Jurassic-early Cretaceous time, the Bangong Lake-Nujiang and Yarlung Zangbo ocean basins opened in what is present-day Tibet (Fig. 19f). The Bangong Lake-Nujiang Basin closed in the late Cretaceous whereas the YarlungZangbo basin did not close until the Tertiary (4050 Ma, Fig. 19g). Closure of the Yarlung Zangbo basin and collision of the Indian and Eurasian Blocks marked the final amalgamation of the Chinese continent. The only younger ophiolites, such as those in Taiwan, lie in the Circum-Pacific belt (Fig. 19h).

Conclusions Unlike Europe and America, present-day China was formed by the convergence of a number of small continental blocks of widely varying ages. The largest of these are the Siberian, Kazakhstan, India, North China, Yangtze and Tarim blocks. Smaller bodies include the Yili, Qaidam, Qiangtang, Lhasa, Changdu-Simao, Yidun and ShanThai-Ma blocks. Chinese ophiolites occur along the margins of these blocks and thus are scattered throughout the country, particularly in the northern, western and southwestern regions. Most ophiolites in China are poorly preserved and occur as ophiolitic melange, rather than as complete sequences. Most also have relatively simple rock assemblages, consisting chiefly of mantle peridotite with lesser amounts of gabbro and pillow lava. Well-developed sheeted-like dyke swarms are rare, possibly implying a slow-spreading environment. The deformation and metamorphism that characterize Chinese ophiolites reflect tectonic overprinting related to collision of the various blocks over most of the Phanerozoic. For instance, the relatively recent collision of

562

Q. ZHANG ETAL.

OPHIOLITES IN CHINA India and Eurasia modified the tectonic framework in a large part of China. Chinese ophiolites are geochemically complex and two or more lava types commonly occur within a given ophiolite sequence. Most ophiolites in the Palaeo-Asian, Neo-Tethyan and CircumPacific belts contain IAT and/or boninite indicating a suprasubduction zone origin. MORE and OIB lavas in these bodies may have originated in back-arc basins or may represent relict fragments of normal ocean lithosphere. Palaeo-Tethyan ophiolites have mostly MORB lavas and lack typical suprasubduction zone components. Thus, they probably formed in intracontinental ocean basins lacking subduction zones. Ophiolites in China range from late Proterozoic to Tertiary in age. The earliest formed in the Proterozoic during rifting of Rodinia. Early to mid-Palaeozoic ophiolites formed in the PalaeoAsian Ocean in North China. Later ophiolites formed in the Palaeo-Tethyan and Neo-Tethyan Oceans in southern and southwestern China. There are no known remnants of the Palaeo-Pacific Ocean in China. The youngest ophiolite in China occurs in Taiwan and was formed in the South China Sea at c. 15 Ma. The manuscript benefited greatly from numerous discussions with J. S. Ren, Chinese Academy of Geological Sciences, D. L. Zhong, C. W. Jin and Y. L. Wang, Chinese Academy of Sciences, K. D. Tang, Shenyang Institute of Geology and Mineral Resources, X. M. Zhou, Nanjing University, J. A. Shao, Beijing University, and D. W. Zhou, Northwest University. Y. L. Wang kindly provided many references on ophiolites in the Xinjiang area and K. D. Tang generously provided his latest geological maps and manuscripts on the PalaeoAsian Ocean. S. L. Chung provided valuable information on the ophiolites in Taiwan. We thank G. Q. He, Beijing University, and X. X. Mo, Chinese University of Geosciences, for their many helpful suggestions on the manuscript. This work was supported by the National Natural Science Foundation of China and the Royal Flemish Academy of Belgium for Science and the Arts.

References AMELIN, Y.V., RITST, E.Y. & NEYMARK, L.A. 1997. Effects of interaction between ultramafic tectonic

563

and mafic magma on Nd-Pb-Sr isotopic systems in the Neoproterozoic Chaya Massif, Baikal-Muya ophiolite belt. Earth and Planetary Science Letters, 148,299-316. BAI, W., YANG, J. & ZHOU, M. 1995. Tectonic evolution of different age ophiolites in the western Junggar, Xinjiang. Acta Petrologica Sinica, 11, 62-72 (in Chinese with English abstract). BAG, Z.W., CHEN, S.H. & ZHANG, Z.T. 1994. Study on REE and Sm-Nd isotopes of Hegenshan ophiolite, Inner Mongolia. Geochimica, 23, 339-349 (in Chinese with English abstract). CHUNG, S. & SUN, S. 1992. A new genetic model for the East Taiwan Ophiolite and its implications for Dupal domains in the Northern Hemisphere. Earth and Planetary Science Letters, 109, 133-145. CRAWFORD, A.J., FALLOON, TJ. & GREEN, D.H. 1989. Classification, petrogenesis and tectonic settings of boninites. In: CRAWFORD, A.J. (ed.) Boninites. Unwin Hyman, London, 1-49. DOBRETSOV, N.L., BERZIN, N.A. & BUSLOV, M.M. 1995. Opening and tectonic evolution of the PalaeoAsian Ocean. International Geological Review, 37, 335-360. DONG, Y.P., ZHOU, D.W. & Liu, L. 1997. Sm-Nd isotopic ages of the Songshugou ophiolite from east Qinling and their geological significance. Regional Geology of China, 16, 217-221 (in Chinese with English abstract). DUPUY, C., MEVEL, C., BODINIER, J.-L. & SAVOYANT, L. 1991. Zabargad peridotite: evidence for multistage metasomatism during Red Sea rifting. Geology, 19, 722-725. GAO, J., Li, M.S., XIAO, X.C., TANG, Y.Q. & HE, G.Q. 1998. Paleozoic tectonic evolution of the Tianshan Orogen, northwestern China. Tectonophysics, 287, 213-231. GIRARDEAU, J. & MERCIER, J.C. 1988. Petrology and texture of the ultramafic rocks of the Xigaze Ophiolite (Tibet): constraints for mantle structure beneath slow-spreading ridges. Tectonophysics, 147, 33-58. GIRARDEAU, J., MARCOUX, J. & ALLEGRE, CJ. ET AL. 1984. Tectonic environment and geodynamic significance of the Nee-Cimmerian Donqiao Ophiolite, Bangong-Nujiang suture zone, Tibet. Nature, 307, 27-31. GIRARDEAU, J., MERCIER, J.C. & YOUGONG, Z. 1985. Origin of the Xigaze Ophiolite, Yarlung-Zangbo suture zone, southern Tibet. Tectonophysics, 119, 407-433. GOPEL, C., ALLEGRE, CJ. & Xu, R. 1984. Lead isotope study of the Xigaze ophiolite (Tibet): the problem

Fig. 19. Schematic temporal-spatial map of China showing the development of the major ophiolite belts through time (see explanation in text). AERJO, Aerjinshan ophiolite belt; ALSO, Ailaoshan ophiolite belt; ANMQO, Animaqin ophiolite belt; BLN, Banggong Lake-Nujiang ophiolite belt; BSO, Beishan ophiolite belt; CMO, Changning-Menglian ophiolite belt; ETWO, East Taiwan ophiolite belt; EZO, East Jungaer ophiolite belt; GLO, Ganzi-Litang ophiolite belt; IDO, Indus River ophiolite belt; JSJO, Jinshanjiang ophiolite belt; KSTO, Kaishantun ophiolite belt; NJSO, Nagaland Mountains ophiolite belt; NMO, Nei Mongol ophiolite belt; NQLO, North Qilian ophiolite belt; SSGO, Songsugou ophiolite belt; WKLO, West Kunlun ophiolite belt; WN-GDBO, south of AnhuiNE Jiangxi ophiolite belt; WZO, Western Junggar ophiolite belt; YLO, Yilan ophiolite belt; YLZB, Yarlung Zangbo ophiolite belt.

564

Q. ZHANG ETAL.

of the relationship between magmatism (gabbros, dolerites, lavas) and tectonites (harzburgites). Earth and Planetary Science Letters, 69, 301-310. HE, G., HAN, B. & YUE, Y. 1990. Tectonic division and crustal evolution of the Altay orogenic belt in China. Geoscience of Xinjiang, No. 2. Geological Publishing House, Beijing (in Chinese with English abstract), 9-20. HEBERT, R., HUOT, F., WANG, C. & Liu, Z. 2003. Yarlung Zangbo ophiolites, southern Tibet revisited: geodynamic implications from the mineral record. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, xx, xx-xxx. HICKEY, R.L. & FREY, F.A. 1982. Geochemical characteristics of boninite series volcanics: implications for their source. Geochimica et Cosmochimica Acta, 46,2099-2115. Ho, C.S. 1977. Melanges in the Neogene sequence of Taiwan. Memoir of the Geological Society of China, 2, 85-96. JIAN, P., WANG, X. & HE, L. 1998. U-Pb zircon dating of the Shuanggou ophiolite from Xingping County, Yunnan Province. Acta Petrologica Sinica, 14, 207-211 (in Chinese with English abstract). JIAN, P., WANG, X. & HE, L. 1999. U-Pb zircon dating of anorthosite and plagiogranite from the Jinshajiang ophiolite belt. Acta Petrologica Sinica, 15, 590-593 (in Chinese with English abstract). JIANG, C., YANG, J. & FENG, B. 1992. Opening and Closing Tectonics of the Kunlun Mountains. Geological Publishing House, Beijing (in Chinese with English abstract). KUSKY, T.M., Li, J.H. & TUCKER, R.D. 2001. The Archean Dongwanzi Ophiolite Complex, North China Craton: 2.505-billion-year-old oceanic crust and mantle. Science, 292, 1142-1145. Li, C., XUE, C. & ZHANG, K. 1996. The ophiolites in Northeastern Jiangxi Province and their geodynamical significance. In: ZHANG, Q. (ed.) Study on Ophiolites and Geodynamics. Geological Publishing House, Beijing, 213-217 (in Chinese with English abstract). Li, J. 1991. Early Paleozoic evolution of lithosphere plate, East Junggar, Xinjiang. Bulletin of the Chinese Academy of Geological Sciences, 23, 1-12 (in Chinese with English abstract). Li, J. 1995. The timing and spatial distribution of ophiolites in China. In: ZHANG, Q. (ed.) Study on Ophiolites and Geodynamics. Geological Publishing House, Beijing, 99-103 (in Chinese with English abstract). Li, J.H., KUSKY, T.M. & HUANG, S.N. 2002. Archean podiform chromitites and mantle tectonics in ophiolite melange, North China Craton: a record of early oceanic mantle processes. GSA Today, 12, 4-11. Li, S.G., CHE, Y.Z., ZHANG, G.W. & ZHANG, Z.Q. 1991. A 1 Ga alpine peridotite body emplaced into the Qinling Group: evidence for the existence of late Proterozoic plate tectonics in the north Qinling area. Geological Review, 37, 235-242 (in Chinese with English abstract). Li, X.H. 2000. Geochemistry of the late Palaeozoic radiolarian cherts within the NE Jiangxi ophiolite

melange and its tectonic significance. Science in China (D), 43, 617-624. Li, X.H., ZHOU, G.Q. & ZHAO, J.X. 1994. Shrimp ion microprobe zircon U-Pb age of the NE Jiangxi ophiolite and its tectonic implications. Geochimica, 23, 125-131 (in Chinese with English abstract). Liou, J.G., LAN, C.Y., SUPPE, J. & ERNST, W.G. 1977. The East Taiwan Ophiolite: its Occurrence, Petrology, Metamorphism and Tectonic Setting. Mining Research & Service Organization, Industrial Technology Research Institute, Taipei, Taiwan, Republic of China. Liu, W.B., QIAN, Q., YUE, G.L., Li, Q.S., ZHANG, Q. & ZHOU, M.F. 2002. The geochemical characteristics of a fore-arc ophiolite from the Dingqing area, Tibet. Acta Petrologica Sinica, 18, 392-400. MALPAS, J., ZHOU, M.-F., ROBINSON, P.T. & REYNOLDS, P.H. 2003. Geochemical and geochronological constraints on the origin and emplacement of the Yarlung Zangbo ophiolites, southern Tibet. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiiolites in Earth History. Geological Society, London, Special Publications, 218, 191-206. McDERMID,

I.R.C.,

AlTCHISON,

J.C.,

DAVIS,

A.M.,

HARRISON, T.M. & GROVE, M. 2002. The Zedang terrane: a Late Jurassic intra-oceanic magmatic arc within the Yarlung-Tsangpo suture zone, southeastern Tibet. Chemical Geology, 187, 267-277. MEIJER, A. 1980. Primitive arc volcanism and a boninite series: examples from western Pacific island arcs. In: HAYES, D.E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands. American Geophysical Union, Washington, DC, 269-282. METCALFE, L 1999. The ancient Tethys oceans of Asia: How many? How old? How deep? How wide? UNEAC Asia Papers, 1, 1-9. NOZAKA, T. & Liu, Y. 2002. Petrology of the Hegenshan ophiolite and its implication for the tectonic evolution of northern China. Earth and Planetary Science Letters, 202, 89-104. OTTONELLO, G., JORON, J.L. & PICCARDO, B. 1984. Rare earth and 3d transition element geochemistry of peridotitic rocks: II. Ligurian peridotites and associated basalts. Journal of Petrology, 25, 373-393. PEARCE, J.A. 1982. Trace element characteristics of lava from destructive plate boundaries. In: THORPE, R.S. (&d.)Andesites. Wiley, Chichester, 525-547. PEARCE, J.A. & CANN, J.R. 1973. Tectonics setting of basic volcanic rocks determined using trace element analyses. Earth and Planetary Science Letters, 19, 290-300. PEARCE, J.A. & DENG, W.M. 1988. The ophiolites of the Tibet Geotraverse, Lhasa to Golmud (1985) and Lhasa to Kathmandu (1986). In: CHANG, C.F., SHACKLETON, R.M., DEWEY, J.F. & JIN, J.X. (eds) The Geological Evolution of Tibet. The Royal Society, London, 215-238. ROBINSON, P.T., BAI, W., ZHOU, M.F., Hu, X.-F., REYNOLDS, P. & YANG, J.-S. 1999. Geochemical constraints on the origin of the Hegenshan Ophiolite, Inner Mongolia, China. Journal of Asian Earth Sciences, 11, 423-442. SHAO, J. & TANG, K. 1995. Terranes in Northeast China

OPHIOLITES IN CHINA and Evolution of Northeast Asia Continental Margin. Seismic Press, Beijing (in Chinese with English abstract). SHERVAIS, J.W. 1990. Island arc and ocean crust ophiolites: constraints in the petrology, geochemistry and tectonic style of ophiolite assemblages in the California Coast Ranges. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues. Cyprus Geological Survey Department, Nicosia, 507-520. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, No. 2000GC000080. SUN, S.S. & McDoNOUGH, W.F. 1989. Chemical and isotopic systematics of oceanic basalt: implications for mantle composition and processes. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 313-345. TANG, K. & SHAO, J. 1996. Some characteristics of ophiolites and ancient ocean evolution in Palaeoasia area. In: ZHANG, Q. (ed.) Study on Ophiolites and Geodynamics. Geological Publishing House, Beijing, 108-111 (in Chinese with English abstract). TANG, Y.Q., GAO, J. & ZHAO, M. 1995. The Ophiolites and Blueschists in the Southwestern Tianshan Orogenic Belt, Xinjiang, Northwestern China. Geological Publishing House, Beijing (in Chinese with English abstract). TRIBUZIO, R., MESSIGA, B., VANNUCCI, R. & BOTTAZZI, P. 1996. Rare earth element redistribution during high-pressure-low-temperature metamorphism in ophiolitic Fe-gabbros (Liguria, northwestern Italy): implications for light REE mobility in subduction zones. Geology, 24, 711-714. WAN, X., WANG, L., WANG, C. & JANSA, L. 1998. Discovery and significance of Cretaceous fossils from the Xigaze forearc basin, Tibet. Journal of Asian Earth Sciences, 16, 217-223. WANG, C., Liu, Z. & HEBERT, R. 2000. The YarlungZangbo palaeo-ophiolite, southern Tibet: implications for the dynamic evolution of the Yarlung-Zangbo Suture Zone. Journal of Asian Earth Sciences, 18, 651-661. WANG, X.F., METCALFE, I., JIAN, P., HE, L.Q. & QANG, C.S. 200la. The Jinshajiang-Ailaoshan Zone, China: Tectonostratigraphy, age and evolution. Journal of Asian Earth Sciences, 18, 675-690. WANG, Z., SUN, S. & Hou, Q.L. 2001b. Effect of meltrock interaction on geochemistry in the Kudi ophiolite (western Kunlun Mountains, Northwestern China): implication for ophiolite origin. Earth and Planetary Science Letters, 191, 33-48. WOOD, D.A. 1980. The application of a Th-Hf-Ta diagram to problems of tectonomagmatic classification and to establishing the nature of crustal contamination of basaltic lavas of the British Tertiary Volcanic Province. Earth and Planetary Science Letters, 50, 11-30. Wu, G., ZHONG, D., ZHANG, Q. & Ji, J. 1999. BabuPhu Ngu ophiolites: a geological record of Paleotethyan ocean bordering China and Vietnam. Gondwana Research, 2, 554-557.

565

Wu, H.R. & DENG, W.M. 1980. Basic geological features of the Yarlung-Zangbo ophiolite belt, Xizang, China. In: PANAYIOTOU, A. (ed.) Ophiolites. Cyprus Geological Survey, Nicosia, 462-472. XIAO, X., TANG, Y. & FENG, Y. 1992. Tectonic Evolution of Northern Xinjiang and its Adjacent Regions. Geological Publishing House, Beijing (in Chinese with English abstract). YANG, H.Y., Lu, J. & Wu, H.Q. 1997. Mineralogy and petrology of the ophiolite from the Namu Bridge in the North Qilian Mountains, northwest China. Journal of the Geological Society of China, 40, 499-526. YANG, J.S., ROBINSON, P.T., JIANG, C.F. & Xu, Z.Q. 1996. Ophiolites of the Kunlun Mountains, China and their tectonic implication. Tectonophysics, 258, 215-231. YANG, R.Y. & HAN, S. 1992. Applications of neutron activation analysis in geoscience. In: NEUTRON ACTIVATION ANALYSIS LABORATORY INSTITUTE OF HIGH ENERGY PHYSICS, ACADEMIA SINICA, (ed.) Applications of Neutron Activation Analysis in Environmental Science, Biology and Geoscience. Atomic Energy Press, Beijing, 198-267. ZHAI, M.G., ZHAO, G.C., ZHANG, Q., KUSKY, T.M. & Li, J.H. 2002. Is the Dongwanzi Complex an Archean Ophiolite? Science, 295, 923a. ZHANG, Q. & ZHOU, G.Q. 2001. Ophiolites of China. Science Press, Beijing (in Chinese with English abstract). ZHANG, Q., ZHOU, Y. & Li, D. 1982. Sheeted dyke swarms within ophiolites in the Xigaze-Baining district, Xizang. In: INSTITUTE OF GEOLOGY, ACADEMICA SINICA, (ed.) Petrological Research. Geological Publishing House, Beijing, 65-80 (in Chinese with English abstract). ZHANG, Q., Li, D. & ZHANG, K. 1992. Mafic-Ultramafic Rocks in the Hengduan Mountains Region. Science Press, Beijing (in Chinese with English abstract). ZHANG, Q., ZHOU, D. & ZHAO, D. 1994. Ophiolites of the Hengduan Mountains, China: characteristics and tectonic settings. Journal of Southeast Asian Geology, 9, 335-344. ZHANG, Q., SUN, X. & ZHOU, D. 1997. The characteristics of North Qilian ophiolites, environment of formation and their tectonic significance. Advances in Earth Science, 12, 366-393 (in Chinese with English abstract). ZHANG, Q., CHEN, Y. & ZHOU, D. 1998. Geochemical characteristics and genesis of Dachadaban ophiolite in the North Qilian area. Science in China (D), 41, 277-281 (in Chinese with English abstract). ZHAO, C.H., HE, K.Z. & TAI, D.Q. 1997.40Ar/39Ar ages of major igneous rocks in northeastern Jiangxi Province. Journal of China University of Geosciences, 22, 257-260 (in Chinese with English abstract). ZHAO, J.X., Li, X.H. & MCCULLOCH, M.T. 1995. Petrogenesis of ophiolites from south Anhui and northeast Jiangxi, and their tectonic implications: chemical and Sm-Nd isotopic constraints. Geochimica, 24, 311 -326 (in Chinese with English abstract). ZHONG, D. 2000. Palaeotethysides in West Yunnan and Sichuan, China. Science Press, Beijing, 248.

566

Q. ZHANG ETAL.

ZHONG, D., Wu, G., Ji, J. & ZHANG, Q. ET AL. 1998. Discovery of ophiolite in southeast Yunnan, China. Chinese Science Bulletin, 44, 36-41. ZHOU, D., DONG, Y. & Liu, L. 1998. Nd, Sr and Pb isotopic characteristics of the Proterozoic ophiolite from Songshugou area. Scientia Geologica Sinica, 33, 31-38 (in Chinese with English abstract). ZHOU, M.-F., MALPAS, J., ROBINSON, P.T. & REYNOLDS, P. 1997. The dynamothermal aureole of the Donqiao ophiolite, northern Tibet. Canadian Journal of Earth Sciences, 34, 59-65. ZHU, B., FENG, Y. & YE, L. 1987. Paleozoic ophiolites in west Junggar and their geological significance. In: SHENYANG INSTITUTE OF GEOLOGY AND

MINERAL RESOURCES, (ed.) Contribution to the Project of Plate Tectonics in Northern China (2). Geological Publishing House, Beijing, 19-28 (in Chinese with English abstract). ZHOU, M.-F., ROBINSON, P.T., MALPAS, J. & Li, Z. 1996. Podiform ehromitites in the Luobusa ophiolite (Southern Tibet): implications for mantle-melt interaction and chromite segregation. Journal of Petrology, 37, 3-21. ZYABREV, S.V., AITCHISON, J.C., BADENGZHU,, DAVIS, A.M., Luo, H. & Liu, J.B. 2002. Radiolarian biostratigraphy of supra-ophiolite sequences in the Xigaze area, Yarlung-Tsangpo suture, Southern Tibet (preliminary report). Radiolaria, 17, 13-19.

Mineral chemistry of ultramafic massifs in the Southern Uralides orogenic belt (Russia) and the petrogenesis of the Lower Palaeozoic ophiolites of the Uralian Ocean P. SPADEA 1 , A. ZANETTI 2 & R. VANNUCCI 2 ' 3 Dipartimento di Georisorse e Territorio, Universitd di Udine, Via Cotonificio 114, 1-33100 Udine, Italy (e-mail: [email protected]) 2 CNR-Istituto di Geoscienze e Georisorse, Sezione di Pavia, Via Ferrata 1, 1-27100 Pavia, Italy ^Dipartimento di Scienze della Terra, Universitd di Pavia, Via Ferrata 1, 1-27100 Pavia, Italy 1

Abstract: Ophiolites of the southern Uralides arc-continent collisional orogen include fertile mantle Iherzolites and minor harzburgites in the Nurali and Mindyak massifs located along the Main Uralian Fault suture of the East European craton margin and the Magnitogorsk island arc. We present the first in situ analyses of pyroxene from Nurali spinel ± plagioclase-bearing Iherzolites and harzburgites and Mindyak spinel Iherzolites and harzburgites. Based on the trace element signatures of pyroxene, the Nurali peridotites are divided into: Group 1, consisting of plagioclase-bearing spinel Iherzolites with moderately to extremely light rare earth element (LREE)-depleted clinopyroxenes, consistent with ^8% fractional melting followed by impregnation by incremental to mid-ocean ridge basalt (MORB)-like melts; Group 2, formed by a spinel peridotite with strongly LREE- to middle REE (MREE)-depleted to enriched clinopyroxenes that testify to re-equilibration with large volumes of melt of tholeiitic affinity; Group 3, consisting of amphibole-bearing spinel harzburgites that underwent pervasive percolation of alkali-enriched melts or fluids. Clinopyroxenes from the Mindyak peridotites are strongly depleted and re-equilibrated by reactive porous flow of infiltrating tholeiitic melts. Two alternative petrogenetic models are proposed. In Model 1 the peridotites derive from oceanic lithosphere generated by mid-ocean ridge processes and affected by refertilization via melt percolation. In Model 2 the peridotites were subcontinental lithospheric mantle that experienced deep-seated magmatism and sub-solidus re-equilibration prior to the opening of the Uralian Ocean, and interacted with melts generated in the asthenospheric mantle by extension-related decompression partial melting during the opening of the Uralian Ocean. In both models the final pre-orogenic events are related to the subduction of the Uralian oceanic lithosphere and to mantle wedge processes, notably intrusion of gabbro-diorite at c. 400 Ma into the Moho sections.

Mantle peridotite massifs occurring in orogenic belts world-wide represent an exceptional tool to unravel various stages associated with the evolution of a collisional orogen. For example, the peridotite massifs close to the Tertiary suture of the Alps have allowed researchers to document the processes related to the lithospheric break-up (Bodinier et al. 1991; Rampone et al. 1997) and the formation of Neotethyan oceanic crust (Rampone et al. 1996). On the other hand, these mantle rocks maintain records of older subduction events (Zanetti et al. 1999) and of a complex multistage evolution under subcontinental conditions (Rampone et al. 1993; Rivalenti et al. 1995; Mazzucchelli et al. 1999). Ultramafic mantle rocks from the ophiolites of the southern Uralides collisional orogen include

fertile mantle Iherzolites and minor harzburgites (Savelieva et al. 1997, 2002). These peridotites compose a series of massifs located along the Main Uralian Fault, which represents a Late Palaeozoic suture zone between the East European palaeomargin of Baltica and the Devonian island arc (Puchkov 1997a, 1997b; Brown et al. 1998). Extensive alteration of the rocks makes it difficult to decipher the nature of the petrogenetic processes responsible for the evolution of the Main Uralian Fault mantle peridotites and to evaluate the appropriate geodynamic environment of their formation through geochemical modelling of the conventional bulk chemistry. In this study we used in situ analyses for a series of geochemically relevant elements of fresh pyroxene from mantle lithologies with the following aims: (1) to

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 567-596. 0305-8719/037$ 15 © The Geological Society of London 2003.

568

P. SPADEA ETAL.

unravel the depletion and enrichment processes that the Uralian peridotites underwent; (2) to evaluate the petrogenetic processes and geological settings recorded by two representative peridotite massifs of the southern Urals, Nurali and Mindyak; (3) to provide geochemical constraints for geodynamic modelling. The in situ pyroxene analyses reported in this paper are the first trace element determinations carried out on fresh mineral phases from the Uralian mantle peridotites.

Geological background The Uralides are a fold-and-thrust belt that records a Late Palaeozoic arc-continent collisional event along the Eastern European palaeomargin of the Baltica plate (Zonenshain et al. 1990; Puchkov 1997a, 1997b; Brown et al 1998). The Main Uralian Fault is a suture zone along which an arc-continent collision took place. It represents the boundary of a continental domain (East European craton) that entered into a subduction zone and collided with an oceanic domain consisting of an intra-oceanic island arc system and dismembered ophiolites. This suture is marked by a melange zone that can be followed from the Polar Urals segment at 70-66°N to the southern Urals segment at 52-48°N and shows regional trends changing from NE-SW to NNE-SSW toward the south, where the melange reaches its broadest extension. The main events in the development of the Uralides include development in the Early Palaeozoic of the Eastern European passive margin (Puchkov 1997b) and the generation of a Uralian Ocean recorded by remnants of a MidOrdovician basaltic crust (Savelieva & Nesbitt 1996; Savelieva et al 1997). In Silurian-Devonian times, a subduction zone dipping away from the continent generated an intra-oceanic island arc, leading to the closure of the Palaeozoic Uralian Ocean (Savelieva & Nesbitt 1996; Spadea & Scarrow 2000; Savelieva et al 2002; Spadea et al 2002). This island arc complex is preserved in the Polar Urals (Savelieva & Nesbitt 1996; Saveliev et al 1999) and includes a large part of the Tagil zone in the central Urals, where the remnants of mid- and lower arc crust are preserved in the Platinum Belt (Seravkin 1997; Savelieva et al 2002). In the Southern Urals (Fig. 1) the island arc is known as the Magnitogorsk arc, and records the initial subduction in the Early Devonian, c. 410 Ma ago, evidenced by the occurrence of boninites in a forearc setting (Spadea et al 1998; Spadea & Scarrow 2000) and the development of a mature, mostly andesitic island arc that followed the formation of these boninites. In Late Devonian to Carboniferous time the Magnitogorsk arc was

accreted to the continental margin and an accretionary wedge was generated and emplaced over the subducting slab (Puchkov 1997a, 1997b; Brown et al 1998; Brown & Spadea 1999). The accretionary wedge is composed of Silurian to Mid-Devonian continental slope and platform sedimentary rocks that are overthrust by Late Devonian to Early Carboniferous syncollisional volcanoclastic turbidites derived from the Magnitogorsk arc (Brown et al 1998). These units are flanked to the east by blueschist- and eclogitefacies metamorphic rocks of continental provenance recording a peak metamorphic age of 380370 Ma (Matte et al 1993; Lennykh et al 1995; Hetzelef a/. 1998). The southern Urals ophiolites are different in size and tectonic position. Along the Main Uralian Fault between the accretionary wedge and the Magnitogorsk arc occur several dismembered fragments dominantly made of lenticular or wedge-shaped slabs of mantle ultramafic rocks, several hundred metres to 1 km in size (Fig. 1). These peridotite bodies are composed mainly of Iherzolite overlain by mantle restites and ultramafic cumulates representing a crust-mantle transition zone, and intruded by gabbro and diorite (Savelieva 1987; Garuti et al 1997; Savelieva et al 1997, 2002). The Nurali and Mindyak massifs are the best-studied bodies in this group. Fragments of oceanic rocks dispersed as blocks and chips in a mostly serpentinite matrix within the Main Uralian Fault melange have been interpreted as dismembered pieces of ophiolite sequences. They include serpentinites, ultramafic and mafic cumulates, layered and isotropic gabbros, basalts and diabases, and oceanic sedimentary rocks of Mid-Ordovician age (Gaggero et al 1997; Savelieva et al 1997, 2002; Saveliev et al 1998). The Kraka ophiolitic massif is a huge body compositionally similar to the Main Uralian Fault ophiolites mentioned above, which is made dominantly of mantle Iherzolites, mantle restites and ultramafic cumulates, and blocks of crustal rocks dispersed in a melange. The Kraka massif is distinct in its structural position in the orogen, in that it occurs as a nappe sheet west of the Main Uralian Fault and thrusts over the Early Palaeozoic sedimentary sequences (Savelieva 1987; Savelieva etal 1997). In the southernmost Urals east of the Main Uralian Fault, the Kempersay massif shows a complete ophiolite pseudostratigraphy and displays evidence for a complex evolutionary path from oceanic to supra-subduction environment (Melcher et al 1999). It consists of mantle tectonites, mostly harzburgite, overlain by lower cumulates, isotropic gabbro, a well-developed sheeted dyke complex, and basaltic pillow lavas

PALAEOZOIC OPHIOLITES, SOUTHERN URALS

569

Fig. 1. Geological map of the Southern Urals showing the location of the Main Uralian Fault suture zone with ophiolites, and the main Iherzolite massifs of Kraka, Nurali and Mindyak.

capped with Mid-Ordovician phtanites (Ivanov etal 1990; Saveliev & Savelieva 1991; Savelieva & Nesbitt 1996; Savelieva et al 1997, 2002). Radiometric dates reported and discussed by Savelieva et al. (2002) have confirmed a Mid-Ordovician age (487 ± 54 Ma) for early generation of the oceanic crust, and much younger, Early Devonian ages (in the 420-387 Ma range), for oceanic closure, obduction and late-stage magmatism. The geodynamic setting where the southern Urals ophiolites were generated and their role

during the complex sequence of events related to the Palaeozoic Uralishes orogeny are still a matter of debate. In fact, even though there is general agreement in identifying the ultramafic and mafic bodies as ophiolitic sensu lato, they have been attributed to different geodynamic environments, namely, mid-ocean ridge (Kempersay ophiolite, Savelieva et al 1997, 2002), or continental margin where rapid ascent of mantle diapir caused rifting and was accompanied by low degree of partial melting in the mantle (Kraka and Nurali ophiolite,

570

P. SPADEA^r^l.

Savelieva et al. 1997, 2002). The time span from ophiolite protolith generation to emplacement is bracketed between Mid-Ordovician and Early Devonian, and differing pre-orogenic evolutions in a oceanic sensu lato setting are inferred from the diversity of internal structure and primary petrological features. On the basis of petrological, geochemical and Nd-Sr isotopic data on the magmatic rocks of the Magnitogorsk arc and those occurring in the Main Uralian Fault melange, Spadea et al. (2002) argued that the Magnitogorsk arc was built on an older (Mid-Ordovician) oceanic crust of mid-ocean ridge type during the closure of the Uralian ocean basin in the Early Devonian. The geodynamic events from sea-floor spreading to oceanic closure are distinct, and the remnants of the Devonian island arc are genetically unrelated to the Main Uralian Fault mantle peridotites. Therefore, the Ordovician oceanic crust should have been trapped during Early Devonian intraoceanic subduction.

Field relations Nurali massif The Nurali massif, located about 10km to the south of Miass, is an ophiolite fragment, mostly peridotitic, elongated NNE-SSW and exposed in an area of about 100km2. It is subdivided into a southern body, which is wider and weakly deformed, and a northern body, which is smaller and internally deformed via thrust faulting. The southern body (Figs 2 and 3) shows the mantle section, which consists mostly of Iherzolite (about 400 m thick) and grades upward to a crust-mantle transition zone consisting of residual tectonites and ultramafic cumulates. The western margin of this body is marked by a steeply east-dipping contact, emplacing the ophiolite over the Precambrian quartzite and schist of the East European craton. To the east, the crust-mantle transition zone is intruded by diorite and gabbro and is locally affected by a west-dipping, low-angle thrust-fault zone that consists of slabs of volcanic and sedimentary rocks and tectonic breccia. The tectonic breccia includes blocks and chips of various igneous and sedimentary rocks, Silurian to Carboniferous in age, embedded in a serpentinite matrix. Further to the east, Ordovician basalts crop out near the town of Polyakovka, and volcanic rocks (lavas and tuffs mostly of basaltic andesitic composition) of the Magnitogorsk arc crop out along the Irendyk ridge east of Polyakovka. Along the Nurali ridge, which is the most elevated ridge, reaching about 750 m altitude, the structurally lower part of the mantle section consists mostly of plagioclase-bearing spinel Iherzo-

lite and minor spinel Iherzolite and spinel harzburgite exposed in a 1.5-1.8 km wide belt (Fig. 3). To the NW it becomes a 0.3-0.5 km wide belt of granular, pyroxene-rich Iherzolite. Along the eastern and southeastern slope of the Nurali Ridge the amount of harzburgite and dunite gradually increases, forming a 1.8km wide belt. To the east of the Nurali ridge, after a large depression overlain by serpentinite, a layered sequence composed of dunite-wehrlite-pyroxenite forms another NNE-SSW ridge of lower altitude. A detailed structural study by Savelieva (1987) has shown that the plagioclase Iherzolites have a nearly horizontal weak foliation, marked by plagioclase and pyroxene, that is associated with an almost horizontal lineation marked by centimetresized elongated aggregates of plagioclase and spinel, and less frequently pyroxene. Open folds from 2 m to a few hundred of metres wide, with well-marked hinges plunging 10-30° to the SE, and gently dipping limbs (up to 40°) fold layering. Similar structures continue at depth as observed in drill cores (Savelieva 1987). In the harzburgite-dunite sequence, layers of orthopyroxene and lineation defined by aggregates of orthopyroxene-spinel-diopside or spinel chains are the main structures. Folding is less pronounced and wider relative to that observed in the plagioclase Iherzolites. The foliation of the harzburgite-dunite sequence varies from nearly horizontal in the western part close to the boundary with the plagioclase Iherzolites, to dipping to NNE at an increasing angle, to almost vertical at the passage to the dunite-wehrlite-pyroxenite layered sequence. The latter, NE-SW striking and almost vertically dipping, is characterized by folds with axial planes dipping to the east or SE and hinges dipping 60-70° to the east and SE. As a whole, layering of the dunite-wehrlite-pyroxenite sequence is conformable with that of the harzburgite-dunite sequence and discordant with respect to that of the plagioclase-bearing Iherzolite. The dunite-wehrlite-pyroxenite sequence consists of the following lithological types upsection and cropping out from west to east (Savelieva 1987; Pertsev et al. 1997): (1) c. 10m of interlayered clinopyroxenite and olivine clinopyroxenite, overlain by fine-grained olivine clinopyroxenite with thin wehrlite and dunite lenses, coarse-grained clinopyroxenite schlieren, and, in the lower part of this interval, up to 10 cm thick chromitite layers (50 m); (2) olivine clinopyroxenite, wehrlite and dunite with layers and lenses of monomineralic diopsidite (Pertsev et al. 1997) followed by dunite with chromite interlayers (2-3 mm), and wehrlite and chromitite metre-sized lenses enveloped by serpentinite (140m); (3) interlayered coarse-

PALAEOZOIC OPHIOLITES, SOUTHERN URALS

571

Fig. 2. Geological map of the Nurali-Voshnezenka area (after Saveliev et al. 1998). In this area the Main Uralian Fault is marked by a melange of dismembered ophiolites, monogenetic and polygenetic tectonic breccias, and slices of diverse volcanic and sedimentary sequences, overthrust on Precambrian crystalline schists from the East European craton. The Polyakovka locality, where the Mid-Ordovician oceanic basalts capped with chert are exposed, is shown. The Irendyk Formation represents the upper unit of the Devonian Magnitogorsk arc, which collided in Late Devonian-Early Carboniferous time with the East European continental margin.

grained harzburgite (layers up to 1.5 m thick) and mostly schistose and serpentinized dunite (layers up to 1 -2 m thick) with parallel intrusions of banded hornblende ± clinopyroxene oxide gabbro and 1 -4 m thick hornblendite concordant with the harzburgite and dunite banding (60 m).

Hornblende gabbro and diorite are intruded into the transition zone with a sharp intrusive contact along its eastern edge. Farther east of this contact the gabbro includes ultramafic xenoliths from the transition zone, and hence providing further evidence of intrusive relationships. A U/Pb age of

572

P. SPADEA ETAL.

Fig. 3. Geological map and cross-section of the Nurali massif, southern body, The location of the peridotite samples from outcrops (N304, N310, N303, N50, N10, A403) and in boreholes (A403, A401, A402) analysed in this study is shown.

399 Ma for a gabbro-diorite was reported by Pertsev et al. (1997).

Mindyak massif The Mindyak massif, located 20 km to the south of Miass, is a dismembered ophiolite fragment, mostly composed of mantle peridotites, 20 km long by 8 km wide, internally thrust faulted, and elongated NNE-SSW. In the northern and central part (Fig. 4) the peridotites have tectonic boundaries dipping to the east, and locally to the NE as in a retrocharriage. The western margin of the massif has an overthrust contact with the Precambrian schists of the East European Craton, and/or a polygenetic tectonic breccia with a serpentinite matrix. The eastern margin of the massif is also bounded by a serpentinite tectonic breccia or tectonic slices of the Magnitogorsk arc lavas and volcanoclastic rocks (Irendyk Formation); a flysch of Late Devonian-Carboniferous age (Zilair Formation) overlies the peridotites (Seravkin 1997). The Mindyak peridotite is mostly composed of spinel and spinel ± plagioclase Iherzolites, which become more depleted upwards, while grading into harzburgites and dunites. Upsection, in the eastern part the peridotite body, the crust-mantle transition zone consists of a wehrlite-pyroxenitedunite association. In the central part of the massif, gabbro and diorite (Denisova 1984) in-

Fig. 4. Geological map and cross-section (redrawn from Denisova 1984) of the central and northern part of the Mindyak massif, and location of the peridotite samples analysed in this study.

trude the peridotites and the overlying wehrlitepyroxenite-dunite sequence. The mantle section of the northern part of the Mindyak massif consists of spinel Iherzolites grading upsection into harzburgites (analysed in

PALAEOZOIC OPHIOLITES, SOUTHERN URALS this study). In the northern part of the massive, near the contact with the crust-mantle transition zone, blocks of metabasites with garnet-pyroxene or garnet-amphibole assemblages are present within a serpentinite breccia (Scarrow et al. 2000; Saveliev et al. 2001). These rocks have been interpreted as oceanic gabbros that underwent seafloor alteration and high-pressure metamorphism. They have provided radiometric age constraints for magmatism (an Ordovician minimum age determined on inherited zircon cores) and peak metamorphism (an Early Devonian age of 410 ± 5 Ma determined on zircon populations by U/Pb method, Saveliev et al. 2001).

Sample selection for analysis For the Nurali massif, six samples among the freshest peridotites from the Nurali ridge and three from the harzburgite-dunite sequence have been selected (Fig. 3). Five spinel ± plagioclase Iherzolites have been collected from the southern sector of the Nurali ridge (N50, N303, N310, A401 and A402), whereas a spinel Iherzolite (H304) sample has been collected from the northern part of the massif. All the samples from the harzburgitedunite sequence are spinel harzburgites (A403, A404andN10). Samples from the Mindyak massif have been collected from the central part (Fig. 4), to document the transition from spinel Iherzolite s (GS505, GS505-2) to spinel harzburgite (GS5051, GS505-39). A synopsis of the distinctive features regarding petrography, mineralogy and geochemistry of the studied samples is presented in Table 1.

573

Petrography Nurali peridotites The spinel ± plagioclase Iherzolites from the Nurali ridge are composed of olivine (around 70%), 15-20% orthopyroxene, 5-10% clinopyroxene, 0.5-5% plagioclase and 1-2% spinel. The textures are coarse-granular to porphyroclastic. They are characterized by large (up to 5 mm in size) and deformed (as deduced by the occurrence of kink bands, subgrain boundaries, engulfing of grain edges, bending of elongated orthopyroxene) porphyroclasts of olivine and orthopyroxene, which are embedded in fine-grained (0.2-2 mm) aggregates. Clinopyroxenes make either interstitial grains or small isodiametric grains included in orthopyroxene. On the basis of the texrural relationships, two types of interstitial clinopyroxene (Cpxl and Cpx2) are recognized. Cpxl is represented by small (up to 1 mm) anhedral to subhedral (prismatic) clinopyroxenes with large exsolution lamellae (Fig. 5a); this type of clinopyroxene locally shows reactive relationships with plagioclase. Cpx2 is defined by larger (commonly close to 2 mm) anhedral patchy crystals, with no to rare exsolution lamellae, generally poikilitic on small rounded relics of olivine and contains rounded (frequently curved and elongated) plagioclase and secondary, anhedral, orthopyroxene (Fig. 5b). This secondary association of clinopyroxene + plagioclase 4- orthopyroxene can be interpreted as a coarse-grained symplectite formed by reaction between an interstitial melt and a primary, spinelfacies, mineral assemblage. Similarly, fine-grained aggregates of secondary orthopyroxene and plagi-

Table 1. Main petrographical and geochemical characteristics of the analysed Nurali and Mindyak mantle peridotites Petrography:

Sample: Bulk-rock Mg no. Bulk-rock LaN/SmN Bulk-rock LaN/YbN Xcr spinel Mg no. spinel Fo olivine A12O3 opx (wt%) A12O3 cpx (wt%) An plagioclase

Nurali massif

Mindyak massif

Spinel plagioclase Iherzolite

Spinel plagioclase Iherzolite

Spinel Iherzolite

Spinel harzburgite

Spinel Iherzolite

Spinel harzburgite

A401, A402, N303,N310

N50

N304

A403, A404, N10

GS505, GS505-2

GS505-1, GS505-3

93-91 0.16-0.29 0.03-0.01 38-44 68-55 91-90 1.4-4.4 1.4-4.4 85-28

91 0.20 0.11 19-21 68-55 91 3.2-3.4 3.9-4.4

93-91 0.60 0.14 35-38 68-66 90 2.5-2.6 3.1-3.3

92 0.34-1.55 0.25-1.62 38-51 68-62 92-91 2.3-2.6 2.6-3.6

92 0.31-0.54 0.69-0.72 27-32 75-68 92-90 1.4-3.7 2.2-5.3

91 1.24-1.45 0.49-0.74 31-41 66-63 91-90 2.7-3.0 2.7-3.5

574

P. SPADEAETAL.

Fig. 5. Petrographic features of spinel-plagioclase and spinel Iherzolites from Nurali and Mindyak massifs. Microphotographs with plane-polarized light (d) and crossed polars (a, b, c, e, f); field of view is 3 mm X 2 mm. (a) Spinel-plagioclase Iherzolite N50: primary clinopyroxene, with turbid core and dense exsolution lamellae (Cpxl). (b) Spinel-plagioclase Iherzolite N50: symplectite mineral assemblage constituted by Cpx2 (orange-yellow, centre), secondary orthopyroxene (grey, upper left) and plagioclase (black), partially replacing primary olivine as a result of its reaction with percolating melt, (c, d) Spinel—plagioclase Iherzolite A402: largely recrystallized spinel (black) overgrown by a symplectite formed by plagioclase (light grains showing twinning) and orthopyroxene. (e) Spinel— plagioclase Iherzolite N303: symplectite mineral assemblage formed by secondary orthopyroxene (grey-white) and plagioclase (white indicates preserved; black indicates altered) surrounding and replacing primary olivine. (f) Spinel Iherzolite GS505-2: clinopyroxene with moderate content of exsolution lamellae (centre) surrounded by strongly serpentinized olivine.

PALAEOZOIC OPHIOLITES, SOUTHERN URALS oclase overgrowing primary olivine can be interpreted as cpx-free symplectite (Fig. 5e). Spinel (brown reddish to dark brown) mostly occurs as skeletal crystals surrounded by and/or intimately interdigitated with relatively large isometric plagioclase (up to 0.6 mm) ± orthopyroxene. Subordinately, subhedral spinel can be found among or within olivine, whereas xenomorphic spinel (0.5-1 mm) is locally placed at the rim of tabular orthopyroxene. The association formed by relics of spinel, plagioclase ± orthopyroxene is present as patches or elongated lenses, 2-3 mm large, placed amongst olivine and orthopyroxene (Fig. 5c and d). Chains of elongated plagioclasespinel aggregates also occur within pyroxene bands, parallel to the elongation of orthopyroxene porphyryroclasts (e.g. sample N303). Most of the plagioclase grains are altered and only their primitive habitus can be recognized. Plagioclase makes rounded grains not connected with any pristine spinel, and mainly occurring within fine-grained neoblastic zones. Plagioclase grains are also found: (1) enclosed in secondary clinopyroxene (Cpx2 of cp 4- opx + pig symplectite); (2) enclosed in reaction zones, where orthopyroxene overgrows olivine (i.e. cpx-free symplectite); (3) along the contact between porphyroclasts. These rounded plagioclase grains commonly have a random distribution, although in some samples they define chains interpretable as pseudo-veins. Small prisms of colourless amphibole are found around chromite grains. The coarse-grained Iherzolites from the northeastern belt consist of 50-60% olivine, up to 30% orthopyroxene and 20% clinopyroxene, and 1-2% spinel, and have coarse-grained (5-10 mm) pyroxenite interlayers. The plagioclase-free Iherzolites contain up to 5% diopside and 2% spinel. In the spinel Iherzolite N304 (containing 5% clinopyroxene), spinel makes isometric and anhedral grains 0.1 -0.2 mm in size, commonly surrounding orthopyroxene, but also placed at the olivine-orthopyroxene boundary and included in olivine. Clinopyroxenes are always strongly exsolved. Also, the orthopyroxenes are exsolved and commonly include poikilitically small, rounded olivine grains. The harzburgites associated with the Iherzolites consist of 80-85% olivine, 10-15% orthopyroxene, 1-2% spinel and minor clinopyroxene (up to 1%). Deformation of olivine and orthopyroxene is similar to that observed in the Iherzolites, and recrystallization is moderate. Spinel is present in subhedral grains up to 1 mm in size at contact with orthopyroxene, or included as small grains (0.1-0,2 mm) within both orthopyroxene and olivine. The dunites within the harzburgites consist of coarse-grained olivine (0.6-0.9 mm), with up

575

to 2% chromite and 1% diopside; fine-grained orthopyroxene may be present. The Nurali peridotites are variously affected by serpentinization. Orthopyroxene is partly transformed to bastite, to a larger extent in harzburgite than in Iherzolite. Besides serpentine minerals, magnesite, talc, and carbonate compose the secondary minerals.

Mindyak peridotites The Mindyak Iherzolites (GS505, GS505-2) have protogranular to porphyroclastic textures, with grain size up to 0.9 mm but mostly averaging 0.6-0.2 mm. They consist of 75-80% olivine, 15% orthopyroxene, 5% clinopyroxene, and 1% reddish brown spinel. Orthopyroxene is generally exsolved and locally bent. Interstitial clinopyroxenes are strongly exsolved in sample GS505, whereas they are not exsolved to slightly exsolved in sample GS505-2. Spinel harzburgites (GS505-1, GS505-3) have porphyroclastic texture with orthopyroxene porphyroclasts up to 2 mm in length. Clinopyroxene is c. 1% by volume. However, sample GS505-3 is cut by a pseudo-vein formed by clinopyroxene of 1 mm size. The reddish spinel occurs as isodiametric, subhedral grains (O.lmm) included in olivine or as anhedral interstitial grains. The studied Iherzolite samples are less than 40% modal serpentinized, whereas the harzburgites are between 45 and 90% modal serpentinized, showing extensive bastite formation pseudomorphmg orthopyroxene.

Analytical techniques Bulk-rock XRF analysis was carried out on the PW2400 spectrometer at the Dipartimento di Mineralogia e Petrologia, Padova University (samples N10, N50, N303, N304, N310, A401, A402 and A404), and on the PW1400 spectrometer at the Dipartimento di Georisorse e Territorio, Udine University (samples GS505, GS505-1, GS505-2 and GS505-3). Major elements were measured on lithium borate glass discs prepared with a flux-tosample ratio of 10:1 to reduce matrix effects. Loss on ignition was determined by the gravimetric method. Bulk-rock trace element analyses (rare earth elements (REE) and V, Cr, Co, Ni, Cu, Zn, Y, Zr and Hf) were carried out by inductively coupled plasma-mass spectrometry (ICP-MS) at the Centre de Instrumentacion Cientifica, Granada University. Sample preparation involved digestion of 0.1 g of sample in HNOs + HF at high pressure and temperature, evaporation to dryness, and subsequent dissolution in 100ml of 4vol.% HNOs. Measurements were carried out in triplicate with a

576

P. SPADEA ETAL.

PE SCIEX ELAN-5000 spectrometer, with Re and Rh used as internal standards. Precision (2o) was about ±3 rel.% and ±8 rel.% for concentrations of 50 and 5 ppm, respectively. Minerals were analysed on the WDS Camebax SX50 Cameca at the CNR-Istituto di Geoscienze e Georisorse, Section of Padova. Operating conditions were 15 kV accelerating voltage, 2.2 nA beam current, and synthetic and natural standards were used. Trace elements in pyroxene were analysed at the CNR-Istituto di Geoscienze e Georisorse, Section of Pavia, by laser-ablation microprobe (LAM)- ICP-MS composed of a double focusing sector field analyser (Finnigan Mat Element) coupled with a Q-switched NdiYAG laser source (Quantel Brilliant). The fundamental emission of the laser source (1064 nm, in the near-IR region) was converted to 266 nm by two harmonic generators. Helium was used as carrier gas and was mixed with Ar downstream of the ablation cell. Spot diameter was varied in the range 40-60 uni. For quantification BCR2-g glass was used as external standard, with 44Ca as internal standard for clinopyroxene and amphibole and 29Si for orthopyroxene. Detection limits were in the range 100-500 ppb for Sc, Ti and Cr, 10-100 ppb for V, Rb, Sr, Zr, Cs, Ba, Gd and Pb, 1-10 ppb for Y, Mb, La, Ce, Nd, Sm, Eu, Tb, Dy, Er, Yb, Hf and Ta, and usually <1 ppb for Pr, Ho, Tm, Lu, Th and U. Precision and accuracy (both better than 10%) were assessed from repeated analyses of SRN NIST 612 standard. Full details of the analytical parameters and quantification procedures have been given by Tiepolo et al. (2003).

Chondrite-normalized REE patterns are shown in Figure 7a and b (C1 chondrite data after Anders & Grevesse 1989). Nurali plagioclase ± spinel Iherzolites are diverse in total REE abundance, with heavy REE (HREEN) up to 2.2, middle REE (MREEN) up to 1.6 and variably depleted light REE (LREE). The LaN/YbN ranges between 0.11 for sample N50 (which has the highest total REE content) and 0.02 for sample A401, whereas the LaN/SmN is between 0.29 for sample A402 and 0.16 for sample A401. The Nurali N304 spinel Iherzolite shows lower HREE content (YbN 0.77) than the plagioclase ± spinel Iherzolites, but it is less fractionated in the LREE region (LaN/YbN and LaN/SmN are 0.14 and 0.6, respectively). The Nurali harzburgites are even more markedly impoverished in HREE (YbN 0.11-0.29), their patterns being slightly enriched to strongly depleted in LREE (LaN/YbN 0.27-1.62; LaN/SmN 0.341.55). The overall refractory character shown by the major element composition of the Mindyak peridotites is confirmed by the trace element distribution. Spinel Iherzolites have a HREE content (HREEN 0.4-0.5) significantly lower than that in Nurali Iherzolites. On the other hand, their LREE-MREE content is higher than that in the N304 Iherzolite and the normalized patterns are flat (LaN/YbN 0.31-0.54; LaN/SmN 0.57-0.79). Harzburgites show a distinctive MREE-HREE depletion (0.1-0.3 times Cl), but the LREE content is similar to or even higher than that in Iherzolites. As a consequence, they show concavedownward patterns similar to that of the Nurali N404 harzburgite with LaN/SmN of c. 2 and LaN/ YbN of 1.24-1.45.

Bulk-rock major and trace element chemistry

Major element mineral chemistry

Bulk-rock major and trace element data for selected Nurali and Mindyak peridotites are presented in Table 2. Relevant major oxides (CaO, A^Os and MgO) and REE have been selected for chemical characterization and comparison among the peridotites of the two massifs. Figure 6 shows the Al2Os-MgO (a) and CaOMgO (b) covariation for the selected Nurali and Mindyak peridotites compared with the primitive mantle estimation by Hofmann (1988). The Nurali samples show evidence of linear depletion trends of CaO and A12O3 relative to MgO from plagioclase Iherzolite to harzburgite, and a gap of AliOs abundance, in the 1.5-2.5 wt% interval, from the plagioclase Iherzolite and the Iherzolite-harzburgite group. The Iherzolite and harzburgite samples from Mindyak show ranges of CaO-MgO and Al2O3-MgO comparable with those of the spinel harzburgites from Nurali, with distinct trends.

Olivine has forsteritic composition and shows no significant variation within samples. The Fo range is large in the olivine from the Nurali peridotites (Fig. 8c), in which the highest Mg content is shown by olivine from spinel harzburgite A403 (mean Fo 92.4), whereas the lowest is shown by olivine from plagioclase-bearing spinel Iherzolite N303 (mean Fo 89.8). Olivine from the Mindyak peridotites has Fo in the range of 90.5-91.3 (Fig. 8c). The NiO content is moderate (0.10-0.50 wt%). Clinopyroxene is chromium diopside in the analysed peridotites. Unlike olivine, the clinopyroxenes from some samples (e.g. spinel Iherzolite GS505) show significant compositional heterogeneity, in particular in the Al, Cr, Na and Ca contents. The Mg number value (expressed as 100 Mg/(Mg + Fe2+x)) is significantly correlated with the Fo content of the olivine. As a whole, the

Table 2. Bulk-rock major (XRF data) and trace element (ICP-MS data) composition ofNurali and Mindyak massif mantle peridotites Nurali massif Sample: Rock type: wt% SiO2 TiO2 A1203 Fe203

MnO MgO CaO Na2O

K2O P205 Total

LOI Mg no. ppm

V Cr Co Ni Cu Zn Y Zr Hf La Ce Pr Nd Sm Eu Gd Tb Dy

A401 Plagioclase spinel Iherzolite 44.08 0.04 2.50 8.77 0.13 42.44 1.62 0.17 0.02 0.00 99.77 6.55 0.9055 37.48 1698 100.9 2050 10.81 45.38 1.026 0.141 0.024 0.008 0.023 0.004 0.032 0.031 0.018 0.09 0.019 0.166

N303 Plagioclase spinel Iherzolite

N50 Plagioclase spinel Iherzolite

A402 Plagioclase spinel Iherzolite

N310 Plagioclase spinel Iherzolite

N304 Spinel Iherzolite

A403 Spinel harzburgite

44.14 0.05 3.14 8.98 0.13 40.63 2.85 0.23 0.03 0.00 100.18 4.13 0.8996

44.45 0.09 2.40 9.10 0.13 41.60 2.56 0.07 0.01 0.00 100.41 6.49 0.9005

44.72 0.03 2.71 8.74 0.13 41.30 2.57 0.17 0.01 0.00 100.38 6.48 0.9035

45.84 0.06 3.44 8.42 0.13 39.08 3.40 0.30 0.03 0.00 100.70 3.11 0.9019

44.20 0.03 1.34 9.43 0.14 42.97 2.04 0.05 0.01 0.00 100.21 4.89 0.9002

43.22 0.03 0.82 7.97 0.11 47.18 0.72 0.05 0.01 0.01 100.12 10.8 0.9214

54.69 1648 100.3 1970 27.84 38.35 1.731 0.38 0.049 0.019 0.063 0.012 0.125 0.066 0.039 0.176 0.036 0.312

56.75 1657 101.5 1946 29.81 36.89 2.073 2.285 0.119 0.043 0.181 0.038 0.296 0.136 0.066 0.278 0.053 0.407

47.73 1867 101.3 1929 5.59 46.33 1.411 0.127 0.024 0.022 0.072 0.015 0.094 0.047 0.023 0.116 0.027 0.226

62.41 1982 96.25 1847 12.83 43.09 2.209 0.273 0.05 0.022 0.074 0.016 0.096 0.067 0.041 0.208 0.046 0.379

58.94 2090 97.58 1938 4.85 45.26 0.699 0.188 0.023 0.025 0.075 0.008 0.052 0.026 0.01 0.062 0.014 0.112

26.94 2189 100.4 2199 6.47 50.31 0.332 0.253 0.022 0.017 0.051 0.01 0.07 0.031 0.013 0.056 0.01 0.071 (continued overleaf)

Table 2. (continued) Nurali massif Sample: Rock type:

Ho Er Tm Yb Lu Tm Yb Lu LaN/SmN Sm^/YoN LaN/YbN

A401 Plagioclase spinel Iherzolite

N303 Plagioclase spinel Iherzolite

N50 Plagioclase spinel Iherzolite

A402 Plagioclase spinel Iherzolite

N304 Spinel Iherzolite

0.077

0.092

0.062

0.096

0.03

0.016

0.15 0.026 0.175 0.029 0.026 0.175 0.029 0.16 0.20 0.03

0.237 0.04 0.245 0.043 0.04 0.245 0.043 0.18 0.30 0.05

0.263 0.043 0.262 0.046 0.043 0.262 0.046 0.20 0.57 0.11

0.198 0.032

0.301 0.052 0.325 0.054 0.052 0.325 0.054 0.21 0.23 0.05

0.106 0.017 0.125 0.021 0.017 0.125 0.021 0.60 0.23 0.14

0.052 0.007 0.047 0.008 0.007 0.047 0.008 0.34 0.73 0.25

0.2 0.037 0.032

0.2 0.037 0.29 0.26 0.08

Mindyak massif

A404 Spinel harzburgite

N10 Spinel harzburgite

GS505 Spinel Iherzolite

GS505-2 Spinel Iherzolite

43.66 0.03 1.23 8.44 0.16 45.32 0.87 0.01 0.00 0.02 99.77 9.53 0.9140

44.37 0.03 1.34 8.39 0.13 44.28 0.97 <0.01 <0.01 0.01 99.52 8.87 0.9127

GS505-3 Spinel harzburgite

GS505-1 Spinel harzburgite

wt% Si02 TiO2 A1203 Fe203

MnO MgO CaO Na2O

K2O P205 Total

LOI Mg-no.

A403 Spinel harzburgite

0.044

Nurali massif Sample: Rock type:

N310 Plagioclase spinel Iherzolite

42.25 0.01 0.70 8.53 0.12 48.08 0.35 0.04 0.01 0.00 100.09 10.11 0.9178

43.89 0.02 1.11 8.66 0.12 45.54 0.70 0.04 0.01 0.00 100.09 7.00 0.9124

42.50 0.03 0.73 9.47 0.14 46.58 0.18 <0.01 <0.01 0.01 99.64 11.56 0.91

44.00 0.02 0.99 9.14 0.14 44.31 1.10 <0.010 0.00 0.01 99.71 8.14 0.9057

ppm

V Cr Co Ni Cu Zn Y Zr Hf La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu LaN/SniN SmN/YbN LaN/YbN

18.61 2236 107.2 2218 5.69 40.36 0.08 0.604 0.022 0.042 0.08 0.012 0.057 0.017 0.006 0.019 0.003 0.02 0.004 0.014 0.003 0.018 0.004 1.55 1.04 1.62

22.95 1907 108.4 2196 6.03 51.4 0.239 0.228 0.013 0.018 0.053 0.008 0.043 0.017 0.008 0.029 0.005 0.043 0.011 0.035 0.006 0.047 0.009 0.66 0.40 0.27

34.0 2086

114 2340 40.8 0.84 0.547 1.421 0.045 0.053 0.134 0.023 0.122 0.042 0.016 0.059 0.011 0.082 0.021 0.069 0.011 0.068 0.011 0.54 0.78 0.72

39.0 2547

124 2374 43.4 1.00 0.546 0.831 0.028 0.034 0.085 0.017 0.088 0.042 0.017 0.063 0.012 0.094 0.021 0.067 0.011 0.075 0.012 0.31 0.49 0.69

21.5 1891

122 2511 43.9 0.66 0.273 1.014 0.035 0.075 0.152 0.022 0.098 0.024 0.007 0.025 0.004 0.038 0.01 0.029 0.004 0.042 0.008 1.24 1.54 0.49

28.0 2070

129 2316 46.9 0.81 0.214 1.026 0.031 0.063 0.141 0.019 0.076 0.019 0.004 0.027 0.005 0.036 0.007 0.02 0.004 0.03 0.005 1.45 1.81 0.74

580

P. SPADEA ETAL.

Fig. 6. Variation diagrams of A12O3 (a) and CaO (b) vs. MgO for Nurali and Mindyak representative peridotite samples selected for the chemical analysis of mineral phases.

Nurali and Mindyak clinopyroxenes show low to moderate Na2O content (0.3-0.7 wt% in the Mindyak, and 0.2-0.6 wt% in the Nurali peridotites; Fig. 8a), and low TiO2 (0.09-0.25 wt% in the Mindyak, and 0.03-0.56 wt% in the Nurali peridotites) and A12O3 (3.1-4.4 wt% in the Mindyak, and 2.6-4.1 wt% in the Nurali peridotites; Fig. 8b) contents. Conversely, the Cr2O3 content is consistently high (1.1 —1.35 wt% in the Mindyak, and 0.8-1.2 wt% in the Nurali peridotites; Fig. 8a). As shown in Figure 8a and b, these compositional features are typical of clinopyroxenes from depleted peridotites, namely, peridotites that have experienced significant episodes of melt extraction (e.g. abyssal peridotites (Johnson et al. 1990; Johnson & Dick 1992), and peridotites from extensional environments, such as the Internal Ligurides Unit in the Northern Apennines (Rampone et al. 1996) and supra-subduction environments (Bonatti & Michael 1989)). Orthopyroxenes have enstatite composition with moderately variable Al content (Fig. 8d). In some samples from the Mindyak and Nurali peridotites, orthopyroxenes show extremely variable CaO content (e.g. in spinel harzburgite GS505-3 the CaO

varies from 0.77 to 2.44 wt%). The highest CaO content is commonly associated with slightly higher A12O3 and lower MgO and FeO contents, but the Mg number values remain constant. On the other hand, the orthopyroxene with low CaO is statistically dominant. Similar CaO variations have been documented in the peridotites from the Internal Ligurides units (Rampone et al. 1996) and interpreted as the result of different T of equilibration. However, the local concentration of exsolved (cryptic) clinopyroxene lamellae cannot be disregarded in explaining the occurrence of the CaO-rich domains. The mean Mg number values are 89.8-92.5 in the Nurali orthopyroxenes, and 90.8-91.4 in the Mindyak orthopyroxenes. These values are roughly correlated with the Fo of the associated olivine, and closely correlated with the Mg number values of the clinopyroxenes, thus suggesting a temperature dependence of the observed Mg-Fe partitioning. Consistently with the clinopyroxenes, the Cr2O3 content of orthopyroxene is high (0.36-0.80 wt%), whereas the A12O3 content is rather low (2.3-3.6 wt%). Spinel is mostly characterized by relatively high Cr (Cr2O3 is 26.6-32.5 wt% in Mindyak perido-

PALAEOZOIC OPHIOLITES, SOUTHERN URALS

581

variability in anorthite content, with An in the range of 28-82 (N303), 54-78 (N310) and 43-85 (A402) (Fig. 9). This large variation of An is unusual in plagioclase peridotites. As shown by the Ab-An-Or diagrams of Figure 9, for example, plagioclases with different microstructural relations in peridotites from the Internal Ligurides show homogeneous Ca-rich composition (An 9194). On the other hand, An-poor plagioclase (An down to 46) has been found in granoblastic coronas around spinel from Zabargad peridotites, which have been interpreted as a result of subsolidus recrystallization at the transition from spinel- to plagioclase-facies conditions (Piccardo et al. 1988). It is noteworthy that the plagioclase from melt-impregnated peridotites from Zabargad has An in the range of 80-85, which is very similar to the most Ca-rich composition of the Nurali plagioclase.

Trace element mineral chemistry Clinopyroxene Fig. 7. Chondrite-normalized REE patterns for the Nurali (a) and Mindyak (b) peridotites. Normalization values after Anders & Grevesse (1989).

tites and 27.6-34.5 wt% in Nurali peridotites) and low Al (A12O3 is 33.6-41.6 wt% in Mindyak peridotites and 27.6-34.5 wt% in Nurali peridotites), with the significant exception of the spinel in the plagioclase ± spinel Iherzolite N50, which is more enriched in Al (A12O3 47.5 wt%; Cr2O3 17.7wt%). Most spinel compositions plot in the 'depleted area' of the field defined in the XCr (expressed as molar 100 X Cr/(Cr + Al)) vs. Mg number (expressed as molar 100 X Mg/(Mg + Fe2+)) diagram (not shown) for abyssal peridotites from the North Atlantic (Bonatti et al. 1993, and reference therein). The spinel from sample N50 is close to the field of fertile subcontinental peridotites (e.g. External Ligurides in the Northern Apennines; Rampone et al. 1993). It is noteworthy that the spinels from both Nurali and Mindyak peridotites plot within the olivine-spinel mantle array defined by Arai (1987) in the X& vs. Fo plot (Fig. 8c). Nevertheless, in the Cr-rich samples, the XCT increases with a decrease of Fo. This correlation is not consistent with the variations induced by different degrees of partial melting and suggests that other processes, such as sub-solidus reequilibration under plagioclase-facies conditions and/or metasomatism affect spinel compositions. Plagioclase has been analysed in the N303, N310 and A402 Iherzolites. It shows a large

Nurali peridotites. The trace element composition of clinopyroxenes from the Nurali peridotites (Table 3) is characterized by a large variability. On the basis of the geochemical affinity, three groups can be tentatively recognized and will hereafter be called Nurali Groups 1-3. Nurali Group 1 consists of the clinopyroxenes from the plagioclase ± spinel Iherzolites (N50, N303, N310 and A402), which show Cl-normalized patterns (Fig. lOa) that are strongly to moderately LREE depleted (CeN/YbN 0.002-0.2) and nearly flat in the HREE region (YbN 8.9-15). The LREE depletion is accompanied by low to very low amounts of the highly incompatible elements (Sr 0.09-2 ppm; Zr 0.42-30 ppm; Nb, Ta U and Th are nearly always below the analytical detection limits). The Cl-normalized patterns are also characterized by negative Sr and Ti anomalies. The most LREE-depleted samples show also negative Zr and Hf anomalies, whereas these elements are interpolated between the adjacent REE in the N50 clinopyroxenes. The REE distribution of the N50 clinopyroxenes is similar to that shown by the clinopyroxenes from fertile Iherzolites (e.g. Zabargad peridotites, Piccardo et al. 1993), but it displays some distinctive features, such as larger LREE depletion and large negative Sr anomaly. These latter characteristics are more typical of clinopyroxenes from peridotites that underwent either melt extraction, or melt percolation at low P (plagioclase-facies conditions) (e.g. Lanzo peridotites, Bodinier et al. 1991; Bracco peridotites, Rampone et al. 1997). An extremely LILE-depleted trace element signature has been

582

P. SPADEA ETAL.

Fig. 8. Major element variability of minerals from Nurali and Mindyak peridotites (P. Spadea, unpublished data), (a) and (b) show the clinopyroxene composition, and the compositional fields of reference clinopyroxene from: (1) subcontinental, fertile, spinel Iherzolites (Balmuccia massif, Western Alps, Italy, Rivalenti et al. 1995; Zabargad peridotite-pyroxenite association, Bonatti et al. 1986; Piccardo et al. 1988); (2) spinel peridotites recrystallized by impregnation of alkaline melts (Caussou massif; Eastern Pyrenees, France, Fabrics et al. 1989); (3) abyssal peridotites (Johnson et al. 1990); (4) ophiolites related to the Jurassic Ligurian—Piedmontese Ocean (Northern Apennines, Italy; Rampone et al. 1996); (5) erogenic massifs variably depleted by partial melting events, followed by metasomatic enrichments, in supra-subduction mantle-wedge environment (Finero massif, Western Alps, Zanetti et al. 1999). (c) and (d) document the relationships of spinel composition with Fo content of olivine and Al content of orthopyroxene. The plotted values of the Nurali and Mindyak minerals are the average composition of several spinel, olivine and orthopyroxene analyses for each sample. Only the composition of orthopyroxene with CaO < 1 wt% has been used for (d).

Fig. 9. Major element composition of plagioclase from Nurali Iherzolites (P. Spadea, unpublished data). For comparison, we show the composition of plagioclase from: (1) melt-impregnated ophiolites related to the Jurassic Ligurian-Piedmontese Ocean (Italy and Corsica, Rampone et al. 1997); (2) melt-impregnated transitional peridotites from Zabargad Island (near the African margin of the Red Sea, Piccardo et al. 1988); (3) Zabargad spinel Iherzolites with plagioclase coronas around spinel, as a result of incipient sub-solidus recrystallization at the spinel- to plagioclase-facies transition (Piccardo et al. 1988).

found in clinopyroxenes from residual oceanic peridotites (Johnson et al. 1990; Johnson & Dick 1992) and from peridotites belonging to ophiolitic sequences (e.g. Internal Ligurides, Northern Apen-

nines, Italy, Rampone et al. 1996). On the other hand, similar features are also displayed by clinopyroxenes from orogenic peridotites commonly ascribed to the subcontinental realm (e.g. Baldis-

Table 3. Trace element compositions (ppm element, except TiOi as wt%) of minerals from Nurali and Midyak mantle peridotites determined with laser-ablation microprobe ICPMS Nurali massif Sample: Mineral: Site: Sc TiO2 V Cr Rb Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Pb Th U

N303 Plagioclase-spinel Iherzolite

N50 Plagioclase-spinel Iherzolite Cpx D6

Cpx D2

Cpx D5

Opx D9

Cpx 2corel

Cpx 2core2

Cpx IBrim

Cpx lAcore

Opx ASA

63 0.66 288 5118 n.d. 1.55 26.8 30 n.d. n.d. 0.18 1.60 0.51 3.9 2.16 0.90 3.3 0.62 4.4 0.90 2.57 0.38 2.02 0.28 0.82 n.d. n.d. n.d. n.d.

77 0.66 324 5919 n.d. 1.22 23.7 27 n.d. n.d. 0.17 1.43 0.45 3.7 2.05 0.66 3.5 0.62 4.1 0.85 2.36 0.33 1.87 0.26 1.17 n.d. n.d. n.d. n.d.

75 0.65 323 6107 n.d. 2.01 22.4 28 n.d. n.d. 0.15 1.38 0.44 3.0 1.88 0.75 3.3 0.57 4.3 0.87 2.20 0.33 1.83 0.25 1.36 n.d. n.d. n.d. n.d.

22 0.16 114 2704 n.d. n.d. 1.94 1.65 n.d. n.d. n.d. n.d. n.d. 0.002 n.d. n.d. n.d. 0.033 0.21 0.060 0.25 0.049 0.36 0.060 0.09 n.d. n.d. n.d. n.d.

71 0.32 291 5149 n.d. 0.40 17.4 2.51 0.04 0.07 0.011 0.13 0.079 0.76 0.96 0.37 1.72 0.38 3.0 0.70 1.79 0.25 1.60 0.24 0.22 0.009 <0.01 0.001 0.001

77 0.34 312 5104 n.d. 0.46 20.8 2.79 0.05 0.05 0.016 0.17 0.073 0.97 0.99 0.39 2.00 0.43 3.1 0.73 2.40 0.32 1.99 0.27 0.34 0.002 0.02 0.001 0.0001

72 0.35 382 5825 n.d. 0.31 19.2 2.27 0.06 <0.03 0.031 0.19 0.086 0.89 1.07 0.43 1.81 0.44 3.6 0.76 2.23 0.33 2.36 0.27 0.26 <0.002 <0.02 0.004 0.003

78 0.40 442 7167 n.d. 0.52 20.6 2.22 0.08 0.05 0.020 0.24 0.066 1.10 1.15 0.44 1.99 0.48 3.7 0.79 2.56 0.31 2.10 0.37 0.20 0.009 <0.02 0.002 0.002

29 0.073 133 3958 n.d. n.d. 1.51 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.004 n.d. n.d. n.d. 0.17 0.054 0.24 0.050 0.36 0.079 n.d. n.d. n.d. n.d. n.d.

Opx A5B 31 0.087 141 3899 n.d. n.d. 1.92 0.18 n.d. n.d. n.d. n.d. n.d. 0.002 n.d. n.d. n.d. 0.011 0.19 0.062 0.30 0.061 0.56 0.066 n.d. n.d. n.d. n.d. n.d.

Opx B7A 27 0.086 129 3564 n.d. n.d. 1.68 0.19 n.d. n.d. n.d. n.d. n.d. n.d. 0.005 n.d. n.d. n.d. 0.20 0.054 0.28 n.d. 0.41 0.066 0.015 n.d. n.d. n.d. n.d.

Opx B8

Opx B9A

Opx B9C

28 0.108 142 4489 n.d. n.d. 1.81 0.27 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.015 0.20 0.075 0.34 0.056 0.55 0.057 0.020 n.d. n.d. n.d. n.d.

25 0.095 130 4139 n.d. n.d. 1.54 0.23 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.011 0.15 0.050 0.23 n.d. 0.40 0.066 n.d. n.d. n.d. n.d. n.d.

26 0.104 127 4053 n.d. n.d. 1.50 0.26 n.d. n.d. n.d. n.d. n.d. 0.003 n.d. 0.005 n.d. 0.014 0.15 0.070 0.18 0.041 0.47 0.087 n.d. n.d. n.d. n.d. n.d.

Nurali massif Sample:

N310 Plagioclase-spinel Iherzolite

A402 Plagioclase-spinel Iherzolite

Mineral: Site

Cpx B15

Cpx B16

Cpx B13

Cpx Bll

Opx B14

Cpx B23

Cpx B18

Cpx B20

Opx B17

Sc TiO2 V

72 0.28 296

73 0.34 333

62 0.29 283

83 0.38 353

26 0.08 126

64 0.16 233

72 0.20 271

71 0.19 268

25 0.06 109

N304 Spinel Iherzolite Opx B24 22 0.05 97

Cpx C3 83 0.14 270

Cpx C4A 85 0.15 270

Cpx C4B 93 0.16 276

Cpx C8 85 0.16 305

(continued overleaf)

Table 3. (continued} Nurali massif Sample: Mineral: Site: Cr Rb Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Pb Th U

N303 Plajjioclase- spinel Iherzolite

N50 Plagioclase-spinel Iherzolite Cpx D6

7701 n.d. 0.35 17.4 1.82 n.d. n.d. n.d. 0.060 0.046 0.73 0.95 0.37 1.94 0.41 2.7 0.67 1.80 0.28 1.79 0.26 0.28 n.d. n.d. n.d. n.d.

Cpx D2

6919 n.d. 0.25 19.8 2.48 n.d. n.d. n.d. 0.062 n.d. 0.71 0.89 0.44 2.02 0.43 3.1 0.77 2.14 0.30 2.05 0.30 0.32 n.d. n.d. n.d. n.d.

Cpx D5

Opx D9

Cpx 2corel

7276 n.d. 0.39 17.3 1.85 n.d. n.d. n.d. 0.067 0.050 0.71 0.97 0.36 1.67 0.37 2.84 0.63 1.92 0.27 1.79 0.25 0.26 n.d. n.d. n.d. n.d.

7552 n.d. 0.36 25.5 3.00 n.d. n.d. 0.012 0.070 0.062 0.93 1.04 0.43 2.29 0.51 4.0 0.94 2.59 0.41 2.44 0.31 0.34 n.d. n.d. n.d. n.d.

4721 n.d. n.d. 1.91 0.23 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.020 0.18 0.065 0.30 0.046 0.45 0.10 0.045 n.d. n.d. n.d. n.d.

Cpx 2core2

7850 n.d. 0.09 13.86 0.42 n.d. n.d. n.d. 0.010 0.011 0.24 0.49 0.26 1.36 0.28 2.22 0.55 1.51 0.21 1.45 0.19 0.09 n.d. n.d. n.d. n.d.

Cpx IBrim

Cpx lAcore

Opx A5A

7974 n.d. 0.10 16.89 0.57 n.d. n.d. n.d. 0.010 0.015 0.24 0.56 0.24 1.61 0.35 2.75 0.60 1.85 0.26 1.60 0.22 0.11 n.d. n.d. n.d. n.d.

8028 n.d. 0.09 16.49 0.56 n.d. n.d. n.d. 0.010 0.013 0.25 0.54 0.28 1.62 0.35 2.40 0.55 1.65 0.21 1.45 0.23 0.09 n.d. n.d. n.d. n.d.

4102 n.d. n.d. 1.36 n.d. n.d. n.d. n.d. n.d. n.d. 0.003 n.d. n.d. n.d. n.d. 0.12 0.045 0.20 0.052 0.35 0.059 0.012 n.d. n.d. n.d. n.d.

Opx A5B

3705 n.d. 0.11 1.13 n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.006 n.d. n.d. 0.009 0.11 0.033 0.14 0.040 0.29 0.059 0.023 n.d. n.d. n.d. n.d.

Opx B7A

6514 n.d. 3.0 7.4 1.25 n.d. n.d. 0.030 0.16 0.045 0.42 0.33 0.12 0.71 0.16 1.14 0.27 0.82 0.13 0.68 0.11 0.040 n.d. n.d. 0.001 0.001

Mineral: Site: Sc TiO2 V Cr Rb Sr Y Zr Nb Ba La Ce

Cpx B25B

Opx B26A

Opx B26B

Cpx B36

60 0.26 225 12322 n.d. 10.9 9.9 7.1 n.d. n.d. 0.11 0.75

62 0.24 226 11415 n.d. 10.9 10.1 7.3 n.d. n.d. 0.12 0.72

20 0.08 85 3941 n.d. n.d. 0.82 0.45 n.d. n.d. n.d. 0.006

20 0.08 88 4167 n.d. 0.12 0.70 0.46 n.d. n.d. n.d. 0.003

73 0.06 193 9926 n.d. 22 3.0 28 0.11 n.d. 0.70 2.72

Opx B9C

6697 n.d. 2.57 6.8 1.12 n.d. n.d. 0.023 0.12 0.044 0.35 0.26 0.14 0.68 0.18 1.15 0.26 0.75 0.13 0.74 0.11 0.10 n.d. n.d. 0.004 0.001

6261 n.d. 3.0 7.0 1.31 n.d. n.d. 0.051 0.18 0.039 0.41 0.30 0.16 0.76 0.17 1.11 0.26 0.78 0.13 0.71 0.085 0.076 n.d. n.d. 0.001 0.001

7283 n.d. 2.43 7.1 1.10 n.d. n.d. 0.020 0.13 0.041 0.38 0.31 0.13 0.73 0.18 1.09 0.25 0.76 0.13 0.74 0.10 0.076 n.d. n.d. 0.0003 <0.001

GS505 Spinel Iherzolite

A404 Spinel harzburgite

A403 Spinel harzburgite Cpx B25A

Opx B9A

Mindyak massif

Nurali massif Sample:

Opx B8

Cpx B34A

Cpx B34B

Opx B30

Amph B33

Cpx D15

Cpx D11A

Cpx D10A

CpxB D10

CpxB Dll

85 0.07 221 9542 n.d. 23 2.98 29 0.42 n.d. 0.82 2.71

94 0.07 227 9240 n.d. 26 3.4 30 0.28 n.d. 0.79 3.07

19 0.02 67 4249 n.d. 0.19 0.21 1.26 n.d. n.d. n.d. n.d.

64 0.11 203 8051 0.37 18 2.45 24 15 13 0.59 2.14

51 0.21 223 10455 n.d. 29 10.1 13 <0.03 <0.26 0.76 2.65

55 0.22 226 10911 n.d. 28 9.7 15 <0.05 <0.25 0.73 2.30

54 0.22 214 9551 n.d. 28 10.2 15 <0.03 <0.15 0.67 2.46

53 0.21 210 9960 n.d. 27 10.3 15 <0.03 0.12 0.70 2.49

57 0.22 221 10033 n.d. 30 10.2 16 <0.04 0.24 0.81 2.58

Opx D12 18 0.07 88 5050 n.d. 0.16 0.79 1.14 n.d. n.d. n.d. 0.023

Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Pb Th U

0.30 2.58 1.49 0.55 2.16 0.35 2.15 0.38 1.04 0.13 0.74 0.10 0.57 n.d. n.d. 0.006 0.003

0.28 2.39 1.36 0.60 2.25 0.33 1.98 0.36 0.90 0.12 0.66 0.09 0.43 n.d. n.d. 0.006 <0.004

n.d. 0.002 0.018 n.d. n.d. 0.018 0.14 0.031 0.08 0.017 0.16 0.029 0.012 n.d. n.d. n.d. n.d.

n.d. 0.005 n.d. n.d. n.d. n.d. 0.127 0.039 0.080 0.016 0.121 0.038 n.d. n.d. n.d. n.d. n.d.

0.57 3.1 0.93 0.33 0.76 0.10 0.70 0.13 0.29 0.05 0.38 0.05 0.88 0.043 n.d. 0.030 0.013

0.59 3.1 0.88 0.36 0.85 0.12 0.67 0.13 0.32 0.04 0.34 0.05 0.92 0.085 n.d. 0.038 0.008

0.55 3.7 1.06 0.35 0.88 0.13 0.72 0.14 0.31 0.04 0.20 0.06 0.87 n.d. n.d. 0.035 0.011

n.d. 0.004 n.d. n.d. n.d. n.d. 0.018 n.d. 0.025 n.d. n.d. 0.016 0.06 n.d. n.d. n.d. n.d.

0.40 2.40 0.68 0.21 0.64 0.094 0.48 0.082 0.30 n.d. 0.29 0.045 0.62 0.73 n.d. 0.018 <0.015

Mindyak massif Sample:

GS505-3 Spinel harzburgite

GS505-1 Spinel harzburgite

Mineral: Site:

Cpx D26B

Sc TiO2 V Cr Rb Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Pb Th U

50 0.07 165 9194 n.d. 27 2.64 1.80 n.d. n.d. 0.35 1.23 0.22 1.14 0.34 0.13 0.46 0.08 0.51 0.11 0.33 0.05 0.34 0.05 0.07 n.d. n.d. <0.003 0.003

n.d., below detection limit.

Cpx D27 54 0.07 158 7785 n.d. 24 3.1 1.65 n.d. n.d. 0.31 1.18 0.21 1.03 0.30 0.12 n.d. 0.06 0.42 0.10 0.29 n.d. 0.26 n.d. 0.05 n.d. n.d. <0.008 0.002

Cpx D22

Opx D24

Opx D25B

50 0.07 162 8910 n.d. 23 3.0 1.82 n.d. n.d. 0.23 1.06 0.20 1.06 0.36 0.10 0.42 0.08 0.54 0.12 0.38 n.d. 0.46 0.07 <0.10 n.d. n.d. <0.004 0.001

19 0.03 74 5467 n.d. 0.23 0.32 0.24 n.d. n.d. n.d. 0.011 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.082 0.018 0.13 0.027 0.01 n.d. n.d. n.d. n.d.

18 0.03 68 5306 n.d. 0.30 0.29 0.15 n.d. n.d. n.d. 0.011 n.d. n.d. n.d. n.d. n.d. n.d. 0.038 0.010 0.076 n.d. 0.11 0.019 n.d. n.d. n.d. n.d. n.d.

Cpx D16 49 0.19 187 8716 n.d. 16 6.5 7.1 <0.05 <0.25 0.25 1.03 0.23 1.55 0.74 0.27 0.86 0.17 1.07 0.24 0.59 0.08 0.59 0.08 0.36 <0.04 <0.3 0.004 n.d.

Opx D17

Opx D19

Opx D20

21 0.07 78 4027 n.d. <0.14 0.53 0.57 <0.06 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.029 0.100 n.d. 0.179 0.026 <0.07 n.d. n.d. n.d. n.d.

20 0.08 79 4638 n.d. <0.09 0.61 0.54 <0.04 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.027 0.098 n.d. 0.188 0.029 0.04 n.d. n.d. n.d. n.d.

18 0.08 77 4559 n.d. <0.09 0.55 0.52 <0.03 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. 0.009 n.d. 0.021 0.099 n.d. 0.171 0.020 0.01 n.d. n.d. n.d. n.d.

0.49 2.91 1.12 0.49 1.55 0.26 1.76 0.39 1.01 n.d. 0.91 0.14 0.47 0.02 n.d. 0.017 0.003

0.43 2.48 1.08 0.41 1.56 0.26 1.44 0.34 0.89 n.d. 0.95 0.11 0.50 <0.03 n.d. 0.016 0.015

0.44 2.68 1.04 0.47 1.60 0.27 1.60 0.35 1.05 0.16 0.95 0.13 0.73 <0.02 n.d. 0.013 0.004

0.44 2.57 1.08 0.49 1.56 0.29 1.76 0.38 1.02 0.14 1.00 0.12 0.59 <0.03 n.d. 0.011 0.002

0.50 2.83 1.14 0.48 1.46 0.27 1.69 0.38 1.02 0.14 0.92 0.12 0.61 <0.03 n.d. 0.006 0.010

0.007 0.046 n.d. n.d. n.d. n.d. 0.041 0.032 0.10 0.02 0.15 0.03 0.06 n.d. n.d. 0.003 0.0004

586

P. SPADEA£r^I.

Fig. 10. Cl-normalized (Anders & Grevesse 1989) REE patterns (a, c, e) and extended spider diagrams (b, d, f) for pyroxene and amphibole of Nurali peridotites.

sero massif and Balmuccia massif; Rivalenti et al. 1995; Mazzucchelli et al. 1999). Nurali Group 2 consists of the clinopyroxenes from the N304 spinel Iherzolite and the A403 spinel harzburgite. The REE patterns of the clinopyroxenes from these two samples (Fig. lOc) are significantly different in terms of LREEMREE abundance and fractionation (CeN/Yt>N 0.06-0.28 SmN/YbN 0.46-2.27; Sm 0.31.43ppm), but they are both characterized by similar low HREE content (c. 4-7 times Cl), the absence of significant Sr anomaly, moderately negative Ti and Zr(Hf) anomaly, and similar Th(U)/La ratios (Fig. lOd). Johnson et al. (1990) reported trace element fractionation similar to that shown by the clinopyroxenes from spinel Iherzolite N30 for clinopyroxenes from hotspot-proximal oceanic peridotites. Nurali group 3 is represented by the clinopyroxenes from the A404 spinel harzburgite. They show LREE-MREE-enriched convex-upward patterns

(CcN/YbN 2-4; SmN/YbN 2.7-5.7; YbN 1-2) (Fig. lOe), with no significant Zr and Hf anomaly, slightly negative Sr anomaly and negative Ti anomaly (Fig. lOf). The relative enrichment in LREE-MREE is accompanied by significant contents of even the most incompatible elements (Nb, Ta, Th and U). Consistent characteristics are shown by the associated amphibole, which is particularly enriched in Nb (15ppm) and Ta (0.7ppm) with respect to the clinopyroxenes. However, the amphibole/clinopyroxene compositional ratios for Sr (0.8), Ti (1.8), Y (0.8) and REE (minimum value 0.6 for Eu) are low for pairs from mantle ultramafic rocks (see, e.g. the compilation reported by Vannucci et al. 1995). This suggests that chemical equilibrium was not attained between the two minerals and/or their crystallization occurred under particular thermochemical conditions. The LREE/HREE, Nb(Ta)/ LREE and Ba/LREE ratios of the clinopyroxene and amphibole from the spinel harzburgite A404

PALAEOZOIC OPHIOLITES, SOUTHERN URALS (Fig. lOe and f) are strictly similar to their magmatic analogues in equilibrium with alkaline melts (Irving & Frey 1984; Zanetti et al 1995, 1996) and have been also documented in clinopyroxenes belonging to spinel harzburgite xenoliths from the Massif Central, France (Xu et al. 1998) and Pali Aike Volcanic Field, South Patagonia, Argentina (Vannucci et al. 2002). Mindyak peridotite. The trace element content of the clinopyroxenes from the Mindyak peridotites is broadly correlated with their fertility, being higher in the spinel Iherzolite GS505 than in the spinel harzburgites (GS505-1 and GS505-3). On the other hand, the Mindyak clinopyroxenes analysed in this study are homogeneously characterized by slightly to moderately LREE-depleted patterns (CeN/YbN 0.47-0.89), which are nearly flat to slightly convex-upward in the MREE— HREE region (SmN/YbN 0.58-1.4; YbN 2.2-5.8) (Fig. 11 a). The Sr content is in the range of 1630 ppm, and results in slightly negative to positive anomalies with respect to the adjacent REE (Fig. 1 Ib). Zr and Hf contents are low in the GS505 spinel Iherzolite (average 15 and 0.6 ppm, respectively), to very low in the GS505-1 harzburgite (average 1.8 and 0.06 ppm, respectively); as a result, they display negative anomalies, which

587

deepen with a decrease in REE content. It is noteworthy that also the content of incompatible to compatible elements, such as Ti, V and Cr, is positively correlated with the REE variation. The clinopyroxenes from the Mindyak Iherzolites show some geochemical affinity in trace element composition to Nurali Group 2, namely low Ti, Y and HREE content and Zr(Hf)/LREE ratio, along with Sr/LREE close to one. On the other hand, the clinopyroxenes from the GS505-1 harzburgite have Ti, Y and HREE contents very similar to those of Nurali Group 3. Orthopyroxene Orthopyroxene shows the typical linearly fractionated LREE-MREE-depleted patterns (Figs 10 and 11). Consistent with the associated clinopyroxene, it has very low concentrations of the highly incompatible trace elements (e.g. Th, Nb and LREE), frequently close to or below the detection limits. Therefore, variations of these elements cannot be monitored adequately. On the contrary, the content of moderately incompatible to compatible elements (such as HREE, Y, V, Sc and Cr) is appreciable and can be compared with that of the associated clinopyroxenes. As a whole, all these elements are positively correlated for the two

Fig. 11. Cl-normalized (Anders & Grevesse 1989) REE patterns (a) and extended spider diagrams (b) for pyroxene of Mindyak peridotites. Literature data are plotted in (c) and (d) for clinopyroxene from: (1) abyssal peridotites (Prot5:29-26 and 1011/76:56-10; Johnson et al. 1990); (2) ophiolitic peridotites (cpx ER-Fl/2-Cpxplc is from Internal Ligurides, Northern Apennines, Rampone et al. 1996); (3) transitional realm (cpx BR-2; Bracco unit, Liguride Alps; Rampone et al. 1997; cpx L213, Lanzo massif, Western Alps, Bodinier et al. 1991; cpx Z2037; plagioclase Zabargad peridotite, Red Sea, A. Zanetti, unpublished data).

588

P. SPADEA£T^L.

pyroxenes. V, Y and Ti show the strongest correlation, but HREE and Zr are also well correlated. A slight spread in correlation is displayed by Cr, Sc and Hf (with the last strongly biased by its low concentration in the orthopyroxene).

Petrogenesis of the Nurali and Mindyak peridotites T and P estimates The significant variability of texture, mineral assemblage and mineral composition reveals that the Nurali and Mindyak peridotites experienced a multistage evolution under largely different T and P conditions. According to Gasparik (1984) and Brey & Kohler (1990), if the Ca-richest compositions found in the orthopyroxene of the Mindyak peridotites actually document early stages of the thermal evolution of the massif, this latter would have experienced equilibrium T close to 1350°C. This corresponds to the anhydrous spinel-facies peridotite solidus at PC. 1.5 GPa (Kinzler 1997). Similar T can be estimated on the basis of the Carichest orthopyroxene compositions found in the N310 plagioclase ± spinel Iherzolite of the Nurali massif. On the other hand, the most abundant population of pyroxene by far consists of Ca-poor orthopyroxene (CaO <1 wt%) and Ca-rich clinopyroxene (CaO >23 wt%). The equilibrium T estimated for these pyroxenes is in the range 830950 °C using the geothermometers based on the enstatite-diopside solvus (Wells 1977; TBKN, Brey & Kohler 1990), whereas a higher T range, 9601070°C, can be estimated on the basis of Ca partitioning (7Ca, Brey & Kohler 1990). Consistent T (850-1000 °C) is also found for the spinelfacies peridotites using the geothermometer of Witt-Eickschen & Seek (1991), thus suggesting the presence of chemical equilibrium between spinel and low-T" pyroxene. These temperatures are probably related to the steady-state equilibrium under mantle (both spinel- and plagioclasefacies) conditions. The textural features of the plagioclase in the Nurali Iherzolites (N50, N303, N310, A402 and A401) indicate that they suffered an extensive recrystallization at low P (<0.8 GPa). The maximum P of the spinel-facies peridotites from Mindyak, estimated on the basis of the spinel composition (Webb & Wood 1986), is in the range of 2.2-2.3 GPa. These values are very similar to those estimated for spinel-facies peridotites from Nurali (2.3-2.5 GPa), but are not consistent with the low Ca contents of the associated olivine, which suggest low equilibrium P (=^1.0 GPa; geobarometer by Brey & Kohler 1990).

Information on the equilibrium T can be also obtained by the application of the empirical equations of Seitz et al. (1999), which are based on the orthopyroxene-clinopyroxene partitioning for V, Sc and Cr. The T range estimated on the basis of V partitioning (TV) is by far the smallest (assuming P is 0.8 GPa, TV is 1120-1150°C for the Mindyak pairs, 1090-1120°C for the Nurali Group 1-Group 2 pairs, and 1040°C for Nurali Group 3 pairs). The TSc values correlate significantly with the corresponding TV (assuming P is 0.8 GPa, TSC is 1060-1120 °C for the Mindyak pairs, 1030-1090 °C for the Nurali Group 1Group 2 group pairs and 950 °C for A404 pairs), but they are generally c. 45 °C higher than TV- The T values calculated on the basis of Cr partitioning (Tcr) are much more variable, encompassing the ranges of both TV and TSc The large TCr variability is associated with a significant variation of the Cr content in the pyroxene, which probably reflects local incomplete sub-solidus equilibrium. As a whole, the T values obtained on the basis of V and Sc partitioning among pyroxenes are significantly higher than those calculated on the basis of major elements, thus suggesting different closure T for the relative exchange reactions.

The origin of plagioclase The presence of plagioclase in mantle peridotites has been attributed to sub-solidus recrystallization of spinel-facies assemblages at P lower than 0.8 GPa and to the crystallization of interstitial melts, either of exotic origin or entrapped after partial melting of the host peridotite. The variable textural features and major element mineral chemistry of the plagioclase from the Nurali peridotites reveal that both of these plagioclase-forming processes probably occurred. In particular, the high-An (80-85) interstitial plagioclase randomly distributed or aligned in pseudo-veins is typical of peridotites impregnated by strongly LREEdepleted melts (Rampone et al. 1997), and can be therefore related to such a process. Conversely, some large low-An (<50) plagioclases overgrowing the spinel relics are probably derived by spinel breakdown at low P, possibly in the presence of a Na-rich (late?) metasomatic agent. Further petrographical and chemical investigations are needed to understand whether or not the two processes were genetically related (e.g. the sub-solidus recrystallization of spinel could have been triggered by the presence of interstitial melts). Noticeably, the extensive recrystallization that the spinel underwent and the evolved nature of the plagioclase-bearing aggregates after spinel are consistent with advanced re-equilibration of these peridotites under plagioclase-facies conditions.

PALAEOZOIC OPHIOLITES, SOUTHERN URALS This represents a distinctive feature with respect to subcontinental peridotites involved in oceanization processes that contain plagioclase forming fine-grained, narrow coronas around spinel (e.g. fertile Iherzolites from Zabargad peridotites, External Ligurides Units; Rampone et al. 1993). In contrast, the spinel-plagioclase relations of the plagioclase-bearing Nurali peridotites provide evidence for a long-term residence under low-P mantle conditions, before their emplacement at the surface.

Whole-rock vs. pyroxene chemistry It has been well documented and deeply debated in previous papers (Salters & Shimizu 1988; Rampone et al 1991, 1996; McDonough et al 1992) that the content and the fractionation of LREE-REE and Sr of bulk-rock composition of spinel-facies mantle peridotites can be confidently estimated on the basis of clinopyroxene composition. In contrast, the HREE, Zr, Ti, Sc and V budget is significantly influenced by the orthopyroxene contribution, and Cr mass-balance calculation requires the involvement of spinel. The comparison between the trace element composition of the pyroxenes from the Nurali and Mindyak peridotites and the respective bulk-rock composition show significant positive correlation for MREE-HREE, Y, Ti, Sc and V The correlation for Cr is positive but somewhat spread, confirming the important role played by spinel in its budget. It is noteworthy that clinopyroxenes and bulk rock show very poor to no correlation for LREE and Sr content. This mismatch can be of various origins. In particular, the higher LREE and Sr content, and the larger LREE-HREE value shown by the bulk rock with respect to the clinopyroxenes could be explained by the presence of interstitial plagioclase (given that this mineral shows a typical LREE-enriched pattern, e.g. Cortesogno et al 2000) and later interstitial components. Some plagioclase-free, spinel Iherzolites and spinel harzburgites from both Mindyak and Nurali massif (e.g. samples GS505-1, GS505-3, N304) also display a higher LREE-HREE value with respect to the clinopyroxenes. Conversely, their LREE and Sr content is always markedly lower than that of clinopyroxenes. The lack of petrographical evidence of the presence of minerals stable in the spinel-facies assemblage and able to host significant amount of LREE (e.g. apatite) could relate these LREE enrichments to late, lowr, alteration processes. In particular, the role of secondary alteration in determining U-shaped REE patterns in the bulk rock has been previously assessed for peridotites from the Trinity ophiolite complex (Gruau et al 1998).

589

Evidence for partial melting and melt flow processes Mindyak peridotites. The low bulk-rock contents of moderately incompatible elements such as Al, Ca, Na, HREE, Y and Ti, together with high Mg contents, suggest that all the Mindyak peridotites analysed in this study underwent large degrees of partial melting. This is confirmed by the pyroxene composition, which displays the same characteristics as the bulk rock, particularly a high Cr content. The bulk-rock HREE contents of the Mindyak Iherzolites and harzburgites are consistent with 11% and 14-15% of fractional melting, respectively. Similarly, the HREE content of clinopyroxene indicates 9-12% of partial (fractional) melting for the Iherzolites and 15% for the GS505-1 harzburgite. Conversely, the slightly LREE-depleted to nearly flat REE patterns shown by the clinopyroxenes and the overall Fe-enriched composition of the spinel harzburgite minerals cannot be explained by simple partial melting processes, but require the addition of LREE-MREE-enriched component during or after the melting events. An alternative working hypothesis is that both mineralogical and chemical differences between harzburgites and Iherzolites could be a direct result of reactive porous flow percolation of melts in different sectors of the mantle column. Consistent with the LREE-HREE fractionation, the Ti/Zr ratios of the Mindyak clinopyroxenes are too low to be explained by partial melting (see, e.g. the modelling of the variation of Ti/Zr ratio during different regimes of partial melting reported by Johnson et al. 1990). This indicates that percolating melts reset also the Zr (Hf) composition (Kelemen et al 1992, 1995). The trace element composition of the hypothetical melts percolating the Mindyak peridotites can be estimated on the basis of the clinopyroxene composition and appropriate clinopyroxene-liquid partition coefficients for basaltic systems (Z)cpx/L). The melts calculated in equilibrium with the spinel Iherzolites clinopyroxenes exhibit LREEenriched patterns (LaN/SmN 1.2-3; LaN/YbN 24.7), slightly negative to slightly positive Sr anomaly, zero to positive Zr anomaly (£)cPx/Basalt from Hart & Dunn 1993; Skulski et al 1994) and low ThN/LaN ratios (c. 0.4, with a z>rhcpx/Basalt of 0.02 calculated on the basis of A1IV content of the Mindyak clinopyroxenes using the empirical equations proposed by Lundstrom et al (1994)). In particular, the low ThN/LREEN can be diagnostic for constraining the geochemical nature of the percolating melts. In fact, ThN/LaN is typically 0.4 in normal mid-ocean ridge basalt (N-MORB; Hofmann 1988), with most alkaline magmas hav-

590

P. SPADEA ETAL.

ing ThN(UN)/LaN c. 1 (e.g. Halliday et al 1995) and calc-alkaline magmas having values from slightly less than one to significantly greater than one (e.g. Conrey et al 1997; Smith et al 1997). This ratio should increase during the melt-peridotite interaction, owing to the larger La compatibility with respect to Th in the mantle peridotites. Consistently, ThN/LaN values from slightly less than one to significantly greater than one, as recognized in most clinopyroxenes from metasomatized peridotites related to the Patagonian mantle wedge that interacted with slab-derived components (Ciuffi et al 2001; Laurora et al 2001) and to the lithospheric mantle beneath the Morocco continental rifting that interacted with alkaline, highly evolved melts (Raffone et al 2001). As a consequence, the low ThN/LaN must be an inherited character of the percolating melt and cannot be acquired by reactive porous flow. In particular, lavas of tholeiitic to transitional geochemical affinity show low ThN/LaN along with LREEN/HREEN >1 (e.g. Iceland plume, Hemond et al 1993). In general, the low abundance of Nb and Ta (below the detection limits), along with the low Ba content, may provide further evidence in favour of a tholeiitic nature of the infiltrating melts. Nurali Group 1 peridotites. The extreme REE fractionation shown by the clinopyroxenes of the Nurali N303, N310 and A402 Iherzolites has been documented for clinopyroxenes of Iherzolites from the oceanic realm (e.g. Johnson et al 1990), ophiolitic sequences (Rampone et al 1996; Rampone et al 1997) and subcontinental mantle sectors (Rivalenti et al 1995; Mazzucchelli et al 1999). Clinopyroxenes with such geochemical signatures have been interpreted as a refractory phase after fractional or incremental melting processes (Johnson et al 1990), or as a newly formed phase, crystallized in equilibrium with migrating (Rampone et al 1997) or trapped (Seyler et al 2001) melts, either single-step or incremental. On the other hand, the coarse, interstitial and/or symplectitic plagioclase of the Nurali Iherzolites is typical of peridotites involved in oceanization processes sensu lato, namely, abyssal peridotites, ophiolitic sequences and thinned subcontinental lithosphere from transitional settings (e.g. the peridotite-pyroxenite association of Zabargad Island, Red Sea, Piccardo et al 1988; Galicia margin, Cornen et al 1996; Seifert & Brunotte 1996; Charpentier 2000). In contrast, to our knowledge, such plagioclase has never been observed in truly subcontinental peridotites. In addition, the Nurali Iherzolites share the following specific features with the abyssal and ophiolitic peridotites having clinopyroxenes with

large LREE-HREE fractionation: (1) a relatively high modal clinopyroxene content (5-10vol.%), in spite of their residual character; (2) a low to very low amount of highly incompatible to moderately incompatible elements, such as Sr, Zr, Ti and Na; (3) a lower depletion of moderately incompatible elements, such as Ca, Al and V Therefore, it can be concluded that all the mineralogical, petrographic and geochemical features suggest that the LREE-depleted, plagioclasebearing Nurali Iherzolites were involved in petrogenetic processes related to the formation of oceanic lithosphere. In this frame, the microtextural features of Cpxl and Cpx2 are regarded as critical points to constrain the specific petrological processes affecting the evolution of these peridotites. In particular, the Cpxl-bearing mineral assemblages can be regarded as residual after partial melting events. In contrast, the Cpx2-bearing mineral assemblage characterized by symplectitic textures and the presence of pseudo-veins of plagioclase can be interpreted as the reaction product between peridotite and impregnating melt (Rampone et al 1997). Petrographic evidence indicates that the effects of impregnation are subordinate in the N401 and N402 Iherzolites (where Cpxl is abundant), intermediate in the N310 Iherzolite, and dominant in the N303 Iherzolite. Theoretical modelling of the fractional melting process (Johnson et al 1990) of a fertile spinel Iherzolite indicates that the bulk-rock HREE content and fractionation of the N401 and N402 Iherzolites are consistent with 6—8% melting of a fertile source under spinel-facies conditions. Instead, lower degrees of partial melting (3-5%) can be estimated for the N303 and N310 Iherzolites. The bulk-rock LREE/HREE value is too large to be related to melting processes, and probably indicates the addition of relatively LREE-enriched melts or late fluids to the system. Interestingly, the HREE composition and fractionation of the clinopyroxenes are not strictly consistent with the degree of partial melting as deduced by the bulk rock. In particular, the LREE-HREE fractionation of the clinopyroxenes is indicative of relatively large melting degrees, whereas the HREE content argues in favour of limited melt extraction. Moreover, the LREE content of the clinopyroxenes provides melting estimates similar to those obtained on the basis of bulk-rock composition. This apparent inconsistency can be explained by sub-solidus redistribution of REE between plagioclase and clinopyroxene, mainly resulting in the increase of HREE content in clinopyroxene. Further support to this interpretation is provided by the large negative Sr anomaly of the clinopyroxenes, which

PALAEOZOIC OPHIOLITES, SOUTHERN URALS is not shown by the bulk-rock composition. The petrographic and geochemical considerations reported above permit us to assume as realistic the degrees of partial melting estimated for the N401 and N402 Iherzolites, before the later impregnation event. However, the lower degrees of partial melting estimated for the N303 and 310 Iherzolites are probably an artefact and reflect the chemical signatures of the impregnating melts. These latter are possibly regarded as incremental melts rather than MORE in the light of the extreme LREE-HREE fractionation shown by the clinopyroxenes from the N303 and N310 Iherzolites. Bulk-rock and mineral major element chemistry suggest that the N50 Iherzolite experienced depletion to a limited extent (c. 3%). Similarly to the N303 Iherzolite, this sample shows diffuse Cpx2bearing symplectite and interstitial plagioclase. These features, along with the fertile REE composition of the clinopyroxenes and their strong negative Sr anomaly, suggest reaction with large volumes of impregnating melt and sub-solidus reequilibration in the plagioclase stability field. The overall trace element signatures of the clinopyroxenes indicate that the refertilization processes were operated by melts of MORB-like composition, as is the case for other occurrences elsewhere. Examples include: (1) the plagioclaserich subcontinental peridotites of Zabargad Island (Piccardo et al. 2002), which experienced fracturing, melt impregnation and uplift, as a consequence of the oceanization process of the Red Sea area; (2) the plagioclase-bearing peridotites of the southern body of the Lanzo Massif (Western Alps), which also represents a section of the subcontinental lithosphere that underwent a polyphase history of partial melting and melt impregnation during continental rifting related to the opening of the Jurassic-early Cretaceous Piedmont-Ligurian ocean (Bodinier et al. 1991; Piccardo et al. 2002); (3) the Internal Ligurides peridotites from the Bracco Unit (Rampone et al. 1997). Nurali Group 2 peridotites. The overall bulk-rock geochemical features of Group 2 peridotites, namely low HREE, Y, Ti and V, in association with LREE/HREE value significantly higher than in Group 1 peridotites, are uncommon for abyssal peridotites and ophiolitic sequences. Trace element compositions of peridotite clinopyroxene from hotspot zones similar to those of the N304 spinel Iherzolite clinopyroxenes have been interpreted by Johnson et al. (1990) as the result of a complex multistage evolution, involving: (1) relatively high degrees of fractional melting in the garnet stability field leaving resi-

591

dual garnet; (2) decompression reaction of garnet to form two-pyroxenes + spinel; (3) additional melting in the spinel stability field and exsolution of clinopyroxene from orthopyroxene. If the partial melting of the A403 spinel harzburgite started under garnet-facies conditions, then the REE distribution of its clinopyroxene is consistent with low degrees of fractional melting (<5%). However, to avoid the redistribution of HREE to clinopyroxene during garnet breakdown and to preserve the convex-upward pattern of clinopyroxene, garnet is required to be completely exhausted during melting. Although possible, particularly at low pressure, this scenario is regarded as unlikely. On the other hand, several mineralogical and geochemical features suggest that Group 2 peridotites underwent large melt percolation after or during a depletion processes. For example, the large modal clinopyroxene content of the N304 spinel Iherzolite (15% by volume) cannot be reconciled with simple clinopyroxene formation via orthopyroxene exsolution. As for the A403 spinel harzburgite, the highly residual character of its bulk-rock major and trace element composition (e.g. the large Mg and the low Al, Ca and HREE concentrations) is not consistent with the relatively large MREE content of the clinopyroxenes. A working alternative explanation is that the present mineralogical and geochemical features of Group 2 peridotites are inherited by pervasive melt percolation under spinel-facies conditions. The relationship between melt migration and partial melting processes is difficult to constrain because of the limited number of samples and their significant variability. In any case, the bulkrock HREE composition of the A403 harzburgite suggests a relatively high degree of partial (fractional) melting (c. 13%), whereas the HREE and Ti contents of bulk rock and clinopyroxenes of the N304 spinel Iherzolite are consistent with a lower degree (c. 9%) of fractional melting under spinelfacies conditions. The liquids calculated in equilibrium with the clinopyroxene of the N304 spinel Iherzolite (£>cPx/L vaiues from Hart & Dunn 1993) are LREE depleted (LaN/YbN c. 0.15; LaN c. 1.6), with ThN/ LaN c. 0.32. Similar compositions have been documented for olivine tholeiites related to the Iceland plume, which were generated by melting of a refractory mantle source (Hemond et al. 1993). In contrast, the hypothetical melts in equilibrium with clinopyroxene from the A403 spinel harzburgite have convex-upward patterns with a maximum in the MREE region and LaN/YbN c. 1 with LaN c. 10. These patterns are strictly similar to the integrated fractional liquids produced by

592

P. SPADEA ETAL.

10% fractional melting in the garnet field plus 10% in the spinel field of a fertile source (Johnson et al. 1990). The melts so produced must have a tholeiitic composition. This character is also confirmed by the very low TliN/LaN ratio of clinopyroxene and its theoretical equilibrium liquid. In conclusion, the data discussed above indicate that the Group 2 peridotites experienced refertilization processes under spinel-facies conditions as a result of the uprise of melts generated in deeper mantle levels having different fertility. Alternatively, if the melts percolating the Group 2 peridotites were genetically related, it might be considered that the melt in equilibrium with the clinopyroxene of the N304 spinel Iherzolite derived from a primitive melt similar in composition to the hypothetical melts in equilibrium with clinopyroxene of the A403 spinel harzburgite via reaction with a strongly refractory ambient peridotite. Nurali Group 3 peridotite. The relative enrichment in LREE-MREE, Nb, Ta, Th, U and Zr displayed by clinopyroxene and amphibole from the A404 spinel harzburgite argue in favour of pervasive percolation of melts or fluids with alkaline geochemical signatures. The bulk-rock and mineral compositions of this harzburgite are characterized by very low amounts of highly fusible major elements, such as Al, Ca and Ti, thus suggesting a high degree of partial melting and/or pyroxene (± spinel or garnet) dissolution during the injection of the alkaline component. Xu et al. (1998) and Vannucci et al. (2002) proposed that spinel harzburgite having similar clinopyroxene composition could form the lowermost part of a lithospheric mantle column extensively re-equilibrated with large volumes of ocean-island basalt (OIB) melts from a deeper source.

Inferences on the geodynamic setting of the Nurali and Mindyak peridotites The Nurali and Mindyak Iherzolites possess some critical mineralogical, petrographic and geochemical features, which are objectively anomalous for an abyssal peridotite lithosphere sensu stricto. The main distinctive features are: (1) the mostly fertile composition of the peridotites (Iherzolitic rather than harzburgitic); (2) the nature of the internal zoning of the peridotite mode, which varies from Iherzolite to dunite through harzburgite; (3) the presence of an anomalous crust-mantle transition zone, which includes amphibole-bearing, plagioclase-free ultramafic cumulates; (4) the lack of evidence of a crustal section related to the peridotites, and in particular of a gabbroic lower crust;

(5) the intrusion of gabbro-diorite experienced by the upper part of the crust-mantle transition zone at about 400 Ma, that is, the time of closure of the Uralian Ocean by intra-oceanic subduction. The petrographical and geochemical data presented in this paper show that the investigated peridotites of the Nurali and Mindyak massifs underwent multistage events of reactive porousflow possibly accompanied, or preceded, by pyroxene dissolution and/or partial melting. On the other hand, the only melt percolation event that can be straightforwardly related to an oceanization process is the plagioclase crystallization, which is well documented in oceanic and transitional ocean-continent environments. In contrast, the lithological zoning recorded by the upper part of both massifs and their geochemical gradients are typical of subcontinental to continent-ocean transition mantle (Western Alps massifs, Zabargad, Lanzo) rather than of mantle sectors from a truly oceanic lithosphere. Although largely speculative because of the limited sampling and the lack of isotope-based time constraints, two models (hereafter called Model 1 and Model 2) can be proposed to account for the anomalous (for abyssal peridotites) features of the Nurali and Mindyak massifs and to integrate the results of this geochemical study with the previous knowledge about these ultramafic massifs. In Model 1, the anomalous features were acquired after the oceanization process. Alternatively, in Model 2 they are considered to have preceded the plagioclase crystallization. Accordingly, these models offer alternative scenarios for the geological settings and the geodynamic evolution of the Nurali and Mindyak ultramafic massifs. In Model 1, the sequence of events is tentatively reconstructed as follows. (1) The Nurali and Mindyak peridotites belonged to a 'normal' oceanic lithosphere generated by mid-ocean ridge processes of the Uralian basin. These processes could be recorded in the spinel and spinel ± plagioclase peridotites of the Nurali ridge that show evidence of a moderate partial melting event (^9%) associated with, or followed by, refertilization processes via melt percolation. (2) Later, the upper part of the mantle ultramafic sequences experienced the percolation of large melt volumes, which generated the Nurali harzburgite-dunite zone and which are recorded by the Mindyak spinel Iherzolite to spinel harzburgite transition. These magmatic events had tholeiitic to alkaline geochemical signatures and were related to intra-plate or, more likely, to island-arc magmatism. The dunite-wehrlite-pyroxenite transition zone of Nurali could be a result of these magmatic episodes rather than of mid-ocean ridge

PALAEOZOIC OPHIOLITES, SOUTHERN URALS processes. The lack of conduits or channels clearly connected to the magma percolation of the Nurali and Mindyak upper zones suggests that this zone was fed laterally. (3) The final, pre-orogenic events recorded by the Nurali and Mindyak massifs are related to the subduction of the Uralian oceanic lithosphere and to the processes that occurred in sectors of the mantle wedge. During the subduction event dated at about 400 Ma the Moho sections of Nurali and Mindyak sequences were intruded by magmas that generated the gabbro-diorite sequence. For Model 2, a more complicated series of events is considered. (1) The Nurali and Mindyak peridotites were part of a subcontinental lithospheric mantle before the opening of the Uralian Ocean. The peridotites at this stage were probably variably depleted Iherzolites that experienced deep-seated magmatism and sub-solidus re-equilibration. (2) During the precursor phases of the extensional process related to the opening of the Uralian Ocean, sectors of the asthenospheric mantle underwent decompression partial melting and large volumes of melts migrated upward, interacting with the overlying peridotite lithospheric mantle. The melt-peridotite interaction was progressively more intense towards the mantle-crust transition, thus forming dominant harzburgite and dunite bodies upsection, essentially via pyroxene dissolution, and resulting in an underplating process. A similar process has been invoked to describe the intrusion of the Basic Complex of the Ivrea-Verbano Zone (Western Alps) related to the extensional regime preceding the opening of the Tethys Ocean during Jurassic time. At Nurali, this stage could be recorded by the recrystallization of the spinel Iherzolite and spinel harzburgite. The geochemical signature of the rising melts was dominantly tholeiitic. The variation of the trace element composition of the spinel-facies clinopyroxene indicates that geochemical gradients developed during the melt percolation and/or the peridotites were intruded by variously fractionated melts. (3) During the break-up of the continental lithosphere the mantle peridotites were partially intruded at shallower depth by tholeiitic melts. This event is marked by plagioclase crystallization; the absence of exsolution lamellae in the coexisting pyroxene formed by the percolating melt that produced plagioclase suggests that the massifs rapidly froze after this melt injection. (4) The final events recorded by the Nurali and Mindyak massifs are closely related to subduction, as described in point (3) of Model 1. In conclusion, Model 1 is an attempt to reconcile our petrographic and geochemical data with

593

the common geodynamic interpretation of the Nurali and Mindyak sequences (namely, oceanic peridotite, possibly involved in island-arc processes). In this scenario, the occurrence of spinel ± plagioclase peridotites overlain by spinel peridotites depleted in fusible elements and with evidence of significant melt percolation suggests that these sequences were probably located in a transitional area between oceanic and sub-arc lithosphere. An alternative scenario is proposed by Model 2, which is aimed at providing new matter for debate on the interpretation of the trace element fingerprints of the southern Urals ultramafic bodies. Important contributions to a more precise assessment of the geological setting and the geodynamic evolution might be furnished by further research aimed at characterizing geochemical gradients occurring in the Nurali harzburgite-dunite sequence and in the Mindyak spinel Iherzolite to harzburgite transition. Such research would provide precious insights into the primitive composition of the percolating melts, the physico-chemical parameters ruling the melt-peridotite reactions and the sequence of the events. We thank G. Savelieva for donation of Nurali samples. She, the unforgettable colleague and friend A. A. Saveliev and A. N. Pertsev have helped with fieldwork, exchange of ideas, suggestions, and useful discussions. We are grateful to F. Bea for producing high-quality trace element data. We thank R. Carampin for assistance during the microprobe chemical analyses of minerals in Padova, and M. Dini and J. H. Scarrow for help with chemical analyses of minerals. We thank Y. Dilek, E. Konstantinovskaya, S. Arai and an anonymous reviewer, who provided frank, constructive reviews. Financial support from the European Commission (TMR-URO Programme), the European Science Foundation (EUROPROBE Programme), and the Italian Ministry of Education, University and Scientific Research (Progetto Nazionale MURST-Cofin98) is acknowledged. The Italian National Research Council has made possible the microprobe work at the University of Padova.

References ANDERS, E. & GREVESSE, N. 1989. Abundances of the elements: meteoric and solar. Geochimica et Cosmochimica Acta, 53, 197-214. ARAI, S. 1987. An estimation of the least depleted spinel peridotites on the basis of olivine-spinel mantle array. Neues Jahrbuch fur Mineralogie, Monatshefte, 347-357. BODINIER, J.-L., MENZIES, M.A. & THIRLWALL, M.F. 1991. Continental to oceanic mantle transition— REE and Sr-Nd isotopic geochemistry of the Lanzo Iherzolite massif. Journal of Petrology, Lherzolite Special Issue, 191-210. BONATTI, E. & MICHAEL, PJ. 1989. Mantle peridotites from continental rifts to ocean basins to subduction

594

P. SPADEA ETAL.

zones. Earth and Planetary Science Letters, 91, 297-311. BONATTI, E., OTTONELLO, G. & HAMLYN, P.R. 1986. Peridotites from the Island of Zabargad (St. John), Red Sea: petrology and geochemistry. Journal of Geophysical Research, 91, 599-631. BONATTI, E., SEYLER, M. & SUSHEVSKAYA, N. 1993. A cold suboceanic mantle belt at the Earth's equator. Science, 261, 315-320. BREY, G.P. & KOHLER, T. 1990. Geothermobarometry in four-phase Iherzolites II. New thermobarometers, and practical assessment of the existing thermobarometers. Journal of Petrology, 31, 1353-1378. BROWN, D. & SPADEA, P. 1999. Processes of forearc and accretionary complex formation during arc-continent collision in the southern Urals. Geology, 27, 649-652. BROWN, D., JUHLIN, C., ALVAREZ-MARRON, J., PEREZESTAUN, A. & OSLIANSKI, A. 1998. Crustal-scale structure and evolution of an arc-continent collision zone in the southern Urals, Russia. Tectonics, 17, 158-171. CHARPENTIER, S. 2000. La zone de transition continentocean de la marge continentale passive ouestiberique: etudes petrologique et geochimique des roches magmatique and mantelliques. PhD thesis, University Blaise Pascal, Clermont-Ferrand. CIUFFI, S.I.A., ZANETTI, A., MAZZUCCHELLI, M., RIVALENTI, G. & CINGOLANI, C.A. 2001. The back-arc mantle lithosphere of the Andean volcanic front: xenoliths from Tres Lagos and Cerro Desconecido (Patagonia, Argentina). EUG 11, Strasbourg, France, 8-12 April. Journal of Conference Abstracts, 6, 415. CONREY, R.M., SHERROD, D.R., HOOPER, P.R. & SWANSON, D.A. 1997. Diverse primitive magmas in the Cascade arc, northern Oregon and southern Washington. Canadian Mineralogist, 35, 367—396. CORNEN, G., BESLIER, M.-O. & GIRARDEAU, J. 1996. Petrologic characteristics of the ultramafic rocks from the continent/ocean transition in the Iberia Abyssal Plain. In: WHITMARSH, R.B., SAWYER, D.S., KLAUS, A. & MASSON, D.G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 377-395. CORTESOGNO, L., GAGGERO, L. & ZANETTI, A. 2000.

Rare earth and trace elements in igneous and hightemperature metamorphic minerals of oceanic gabbros (MARK area, Mid-Atlantic Ridge). Contributions to Mineralogy and Petrology, 139, 373-393. DENISOVA, E.A. 1984. Internal structure of Mindyak ultramafic massif (Southern Urals). Reports, Academy of Sciences, 274, 382-387 (in Russian). FABRIES, J., BODINIER, J.-L., DUPUY, C., LORAND, J.-P. & BENKERROU, C. 1989. Evidence of modal metasomatism in the erogenic spinel Iherzolite body from Caussou (Northeastern Pyrenees, France). Journal of Petrology, 30, 199-228. GAGGERO, L., SPADEA, P., CORTESOGNO, L., SAVELIEVA, G.N. & PERTSEV, A.N. 1997. Geochemical investigation of the igneous rocks from Nurali ophiolite melange zone, Southern Urals. Tectonophysics, 276, 139-161.

GARUTI, G., FERSHTATER, G., BEA, F., MONTERO, P., PUSHKAREV, E.V. & ZACCARINI, F. 1997. Platinumgroup elements as petrological indicators in mafic ultramafic complexes of the central and southern Urals: preliminary results. Tectonophysics, 276, 181-194. GASPARIK, T. 1984. Two-pyroxene thermobarometry with new experimental data in the system CaO-MgOAl2O3-SiO2. Contributions to Mineralogy and Petrology, 87, 87-97. GRUAU, G., GRIFFITHS, J.B. & LECUYER, C. 1998. The origin of U-shaped rare earth patterns in ophiolite peridotites: assessing the role of secondary alteration and melt/rock reaction. Geochimica et Cosmochimica Acta, 62, 3545-3560. HALLIDAY, A.N., LEE, D.-C, TOMMASINI, S., DAVIES, G.R., PASLIK, C.R., FITTON, J.G. & JAMES, D.E. 1995. Incompatible trace elements in OIB and MORB and source enrichment in the sub-oceanic mantle. Earth and Planetary Science Letters, 133,379-395. HART, S.R. & DUNN, T. 1993. Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology, 113, 1-8. HEMOND, C., ARNDT, N.T., LICHTENSTEIN, U., HOFMANN, A.W., OSKARSSON, N. & STEINTHORSSON, S.

1993. The heterogeneous Iceland plume: Nd-Sr-O isotopes and trace element constraints. Journal of Geophysical Research, 98, 15833-15850. HETZEL, R., ECHTLER, H.P., SEIFERT, W., SCHULTE, B.A. & IVANOV, K.S. 1998. Subduction- and exhumation-related fabrics in the Paleozoic highpressure/low-temperature Maksyutov Complex, Antingan area, southern Urals, Russia. Geological Society of America Bulletin, 110, 916-930. HOFMANN, A.W. 1988. Chemical differentiation of the Earth: the relationships between mantle, continental crust, and oceanic crust. Earth and Planetary Science Letters, 90, 297-314. IRVING, AJ. & FREY, F.A. 1984. Trace element abundances in megacrysts and their host basalts: constraints on partition coefficients and megacrysts genesis. Geochimica et Cosmochimica Acta, 48, 1201-1221. IVANOV, K.S., PUCHKOV, V.N. & BABENKO, V.A. 1990. Finds of conodonts and graptolites in metamorphic units of South Urals. Doklady Academiya Nauk SSSR, 310, 676-679. JOHNSON, K.T.M. & DICK, H.J.B. 1992. Open system melting and temporal and spatial variation of peridotite and basalts at the Atlantis II Fracture Zone. Journal of Geophysical Research, 97(B6), 9219-9241. JOHNSON, K.T.M., DICK, H.J.B. & SHIMIZU, N. 1990. Melting in the oceanic upper mantle: an ion microprobe study of diopsides in abyssal peridotites. Journal of Geophysical Research, 95, 2661-2678. KELEMEN, P.B., DICK, H.J.B. & QUICK, J.E. 1992. Formation of harzburgite by pervasive melt/rock reaction in the upper mantle. Nature, 358, 635-641. KELEMEN, P.B., SHIMIZU, N. & SALTERS, V.J.M. 1995. Extraction of mid-ocean-ridge basalt from the upwelling mantle by focused flow of melt in dunite channels. Nature, 375, 747-753.

PALAEOZOIC OPHIOLITES, SOUTHERN URALS KINZLER, RJ. 1997. Melting of mantle peridotite at pressures approaching the spinel to garnet transition: application to mid-ocean ridge petrogenesis. Journal of Geophysical Research, 102(B1), 853-874. LAURORA, A., MAZZUCCHELLI, M., RIVALENTI, G., VANNUCCI, R., ZANETTI, A., BARBIERI, M.A. & CINGOLAMI, C.A. 2001. Metasomatism and melting in carbonated peridotite xenoliths from the mantle wedge; the Gobernador Gregores case (Southern Patagonia). Journal of Petrology, 42, 69-87. LENNYKH, V.I., VALIZER, P.M., BEANE, R. & ERNST, W.G. 1995. Petrotectonic evolution of the Maksyutov complex, Southern Urals, Russia: implications for ultrahigh-pressure metamorphism. International Geology Review, 37, 584-600. LUNDSTROM, C.C., SHAW, H.F., RYERSON, F.J., PHINNEY, D.L., GILL, J.B. & WILLIAMS, Q. 1994. Compositional controls on the partitioning of U, Th, Ba, Pb, Sr, and Zr between clinopyroxene and haplobasaltic melts: implications for uranium series disequilibria basalts. Chemical Geology, 128, 407-423. MATTE, PH., MALUSKI, H., GABY, R., NICHOLAS, A., KEPEZHINSKAS, P. & SOBOLEV, S. 1993. Geodynamic model and 39Ar/40Ar dating for generation and emplacement of high-pressure metamorphic rocks in SW Urals. Comptes Rendus de I'Academic des Sciences, Serie 2, 317, 1667-1674. MAZZUCCHELLI, M., RIVALENTI, G., ZANETTI, A., VANNUCCI, R. & CAVAZZINI, G. 1999. Origin and significance of late noritic dykes in the Baldissero peridotite massif (Ivrea-Verbano Zone). Ofioliti, 24, 129-130. McDoNOUGH, W.F., STOSCH, H.G. & WARE, N.G. 1992. Distribution of titanium and rare earth elements between peridotitic minerals. Contributions to Mineralogy and Petrology, 110, 321-328. MELCHER, F., GRUM, W., THALHAMMER, T.V. & THALHAMMER, O.A.R. 1999. The giant chromite deposits of Kempirsai, Urals: constraints from trace elements (PGE, REE) and isotope data. Mineralium Deposita, 34, 250-272. PERTSEV, A.N., SPADEA, P., SAVELIEVA, G.N. & GAGGERO, L. 1997. Nature of the transition zone in the Nurali ophiolite, southern Urals. Tectonophysics, 276, 163-180. PICCARDO, G.B., MESSIGA, B. & VANNUCCI, R. 1988. The Zabargad peridotite-pyroxenite association: petrological constraints on its evolution. Tectonophysics, 150, 135-162. PICCARDO, G.B., RAMPONE, E., VANNUCCI, R., SHIMIZU, N., OTTOLINI, L. & BOTTAZZI, P. 1993. Mantle processes in the subcontinental lithosphere: the case study of the rifted sp-lherzolites from Zabargad (Red Sea). European Journal of Mineralogy, 5, 1039-1056. PICCARDO, G.B., RAMPONE, E., ZANETTI, A., ROMAIRONE, A. & BRUZZONE, S. 2002. Melt percolation and impregnation in the Lanzo South peridotite: preliminary field, petrographic and mineral chemistry results. Annual Meeting of the Swiss Academy of Natural Sciences (SANW) 2002, 19-20 September, Davos (Switzerland), Abstracts. 36-38.

595

PUCHKOV, V.N. 1997a. Structure and geodynamics of the Uralian orogen. In: BURG, J.-P. & FORD, M. (eds) Orogeny through Time. Geological Society, London, Special Publications, 121, 201-236. PUCHKOV, V.N. 1997b. Tectonics of the Urals: modern concepts. Geotectonics, 31, 294-312. RAFFONE, N., ZANETTI, A., CHAZOT, G., DENIEL, C. & VANNUCCI, R. 2001. Geochemistry of mantle xenoliths from Ibalrhatene (Mid Atlas, Morocco): insights into the lithospheric mantle evolution during continental rifting. Eleventh Annual V. M. Goldschimdt Conference, May 20-24, Hot Springs, VA. Lunar and Planetary Institute Contribution. 1088, Abstract 3323. RAMPONE, E., BOTTAZZI, P. & OTTOLINI, L. 1991. Complementary Ti and Zr anomalies in orthopyroxene and clinopyroxene from mantle peridotites. Nature, 354, 518-521. RAMPONE, E., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L. 1993. Subsolidus reactions monitored by trace element partitioning: the spinelto plagioclase-facies transition in mantle peridotites. Contributions to Mineralogy and Petrology, 115, 1-17. RAMPONE, E., HOFMANN, A.W., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L. 1996. Trace element and isotope geochemistry of depleted peridotites from an N-MORB type ophiolite (Internal Ligurides, N. Italy). Contributions to Mineralogy and Petrology, 123, 61-76. RAMPONE, E., PICCARDO, G.B., VANNUCCI, R. & BOTTAZZI, P. 1997. Chemistry and origin of trapped melts in ophiolitic peridotites. Geochimica et Cosmochimica Acta, 61, 4557-4569. RIVALENTI, G., MAZZUCCHELLI, M., VANNUCCI, R., HOFMANN, A.W., OTTOLINI, L., BOTTAZZI, P. & OBERMILLER, W. 1995. The relationship between websterite and peridotite in the Balmuccia peridotite massif (NW Italy) as revealed by trace element variations in clinopyroxene. Contributions to Mineralogy and Petrology, 121, 275-288. SALTERS, V.J.M. & SHIMIZU, N. 1988. World-wide occurrences of HFSE-depleted mantle. Geochimica et Cosmochimica. Acta, 52, 2177-2182. SAVELIEV, A.A. & SAVELIEVA, G.N. 1991. The Kempersay ophiolite massif: main features of the structure and composition. Geotectonika, 6, 57-65. SAVELIEV, A.A., ASTRAKHANTSEV, O.V., KNIPPER, A.L., SHARASKIN, A.YA. & SAVELIEVA, G.N. 1998. Structure and deformation phases of the Northern Terminus of the Magnitogorsk Zone, Urals. Geotectonika, 3, 38-50. SAVELIEV, A.A., SHARASKIN, A.YA. & D'ORAZIO, M. 1999. Plutonic to volcanic rocks of the Voykar ophiolite massif (Polar Urals): structural and geochemical constraints on their origin. Ofioliti, 24, 21-30. SAVELIEV, A.A., BIBIKOVA, E.V., SAVELIEVA, G.N., SPADEA, P., PERTSEV, A.N., SCARROW, J.H. & KIRNOZOVA, T.I. 2001. Garnet pyroxenites of the Mindyak massif, South Urals: age and environment of their formation. Bulletin MOIP, 76, 22-29 (in Russian). SAVELIEVA, G.N. 1987. Gabbro-Ultrabasite Assem-

596

P. SPADEAETAL.

blages of Uralian Ophiolites and their Analogues in the Recent Oceanic Crust. Nauka, Moscow (in Russian). SAVELIEVA, G.N. & NESBITT, R.W. 1996. A synthesis of the stratigraphic and tectonic setting of the Uralian Ophiolites. Journal of the Geological Society, London, 153, 525-537. SAVELIEVA, G.N., SHARASKIN, A.YA., SAVELIEV, A.A., SPADEA, P. & GAGGERO, L. 1997. Ophiolites of the Southern Uralides adjacent to the East European continental margin. Tectonophysics, 276, 117-137. SAVELIEVA, G.N., SHARASKIN, A.YA., SAVELIEVA, A.A., SPADEA, P., PERTSEV, A.N. & BABARINA, I.I. 2002. Ophiolites and zoned mafic-ultramafic massifs of the Urals: a comparative analysis and some tectonic implications. In: BROWN, D., JUHLIN, C. & PUCHKOV, V. (eds) Mountain Building in the Uralides: Pangea to the Present. Geophysical Monograph, American Geophysical Union, 132, 135-153. SCARROW, J.H., SPADEA, P., CORTESOGNO, L., SAVELIEVA, G.N. & GAGGERO, L. 2000. Geochemistry of garnet metagabbros from the Mindyak ophiolite massif, Southern Uzals. Ofioliti, 25, 103-115. SEIFERT, K. & BRUNOTTE, D. 1996. Geochemistry of serpentinized mantle peridotite from Site 897 in the Iberia Abyssal Plain. In: WHITMARSH, R.B., SAWYER, D.S., KLAUS, A. & MASSON, D.G. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 149. Ocean Drilling Program, College Station, TX, 471-488. SEITZ, H.-M., ALTHERR, R. & LUDWIG, T. 1999. Partitioning of transition elements between orthopyroxene and clinopyroxene in peridotitic and websteritic xenoliths: new empirical geothermometers. Geochimica et Cosmochimica Acta, 63, 3967-3982. SERAVKIN, I.B. 1997. Southern Urals: tectono-magmatic zoning and position among orogenic systems of the Uralo-Mongolian Belt. Geotektonika, 1, 32-47 (in Russian). SEYLER, M., TOPLIS, M.J., LORAND, J.-P., LUGUET, A. & CANNAT, M. 2001. Clinopyroxene microtextures reveal incompletely extracted melts in abyssal peridotites. Geology, 29, 155-158. SKULSKI, T., MINARIK, W. & WATSON, E.B. 1994. Highpressure experimental trace-element partitioning between clinopyroxene and basaltic melts. Chemical Geology, 111, 127-147. SMITH, I.E.M., WORTHINGTON, T.J., PRICE, R.C. & GAMBLE, J.A. 1997. Primitive magmas in arc-type volcanic associations: examples from the southwest Pacific. Chemical Geology, 35, 257-273. SPADEA, P. & SCARROW, J.H. 2000. Early Devonian boninites from the Magnitogorsk Arc, Southern Urals (Russia): implications for early development of a collisional orogen. In: DILEK, Y., MOORES, E., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 461-472. SPADEA, P., KABANOVA, L.YA. & SCARROW, J.H. 1998. Petrology, geochemistry and geodynamic significance of Mid-Devonian boninitic rocks from the

Baimak-Buribai area (Magnitogorsk Zone, Southern Urals). Ofioliti, 23, 17-36. SPADEA, P., D'ANTONIO, M., KOSAREV, A., GOROZHANINA, Y. & BROWN, D. 2002. Arc-continent collision in the Southern Urals: petrogenetic aspects of the forearc-arc complex. In: BROWN, D., JUHLIN, C. & PUCHKOV, V. (eds) Mountain Building in the Uralides: Pangea to the Present. Geophysical Monograph, American Geophysical Union, 132, 101-134. TlEPOLO, M., BOTTAZZI, P., PALENZONA, M. & VAN-

NUCCI, R. 2003. A laser probe coupled with ICPdouble focusing sector field mass spectrometer for in-situ analysis of trace elements in the geological samples and U-Pb dating of zircons. Canadian Mineralogist, 41, 259-272. VANNUCCI, R., PICCARDO, G.B. & RIVALENTI, G. ET AL. 1995. Origin of LREE-depleted amphiboles in subcontinental mantle. Geochimica et Cosmochimica Acta, 59, 1763-1771. VANNUCCI, R., ZANETTI, A., KEMPTON, P., CIUFFI, S., MAZZUCCHELLI, M. & CINGOLANI, C. 2002. The chemical history of young lithospheric mantle in a back-arc region: the spinel + garnet peridotite xenoliths from Pali Aike (South Patagonia). In: CINGOLANI, C.A. ET AL. (eds) Acta del XV Congreso Geologico Argentina. Asociation Geologica Argentina, Buenos Aires, Tomo III, 71-74. WEBB, S.A.C. & WOOD, B.J. 1986. Spinel-pyroxenegarnet relationships and their dependence on Cr/Al ratio. Contributions to Mineralogy and Petrology, 92,471-480. WELLS, P.R.A. 1977. Pyroxene thermometry in simple and complex systems. Contributions to Mineralogy and Petrology, 62, 129-139. WITT-EICKSCHEN, G. & SECK, H.A. 1991. Solubility of Ca and Al in orthopyroxene from spinel peridotite: an improved version of an empirical geothermometer. Contributions to Mineralogy and Petrology, 106,431-439. Xu, Y.-G., MENZIES, M.A., BODINIER, J.-L., BEDINI, R.M., VROON, P. & MERCIER, J.-C.C. 1998. Melt percolation and reaction atop a plume: evidence from the poikiloblastic peridotite xenoliths from Boree (Massif Central, France). Contributions to Mineralogy and Petrology, 132, 65-84. ZANETTI, A., VANNUCCI, R., OBERTI, R. & DOBOSI, G. 1995. Trace-element and crystal-chemistry of mantle amphiboles from the Carpatho-Pannonian Region. Acta Vulcanologica, 7, 265-276. ZANETTI, A., VANNUCCI, R., BOTTAZZI, P., OBERTI, R. & OTTOLINI, L. 1996. Infiltration metasomatism at Lherz as monitored by systematic ion-microprobe investigations close to a hydrous dike. Chemical Geology, 134, 113-133. ZANETTI, A., MAZZUCCHELLI, M., VANNUCCI, R. & RIVALENTI, G. 1999. The Finero phlogopite-peridotite massif: an example of subduction-related metasomatism. Contributions to Mineralogy and Petrology, 134, 107-122. ZONENSHAIN, L.P., KAZMIN, M.L. & NATAPOV, L.M.

1990. Geology of the USSR: a Plate-Tectonic Synthesis. American Geophysical Union, Geodynamics Series, 21.

Petrological diversity and origin of ophiolites in Japan and Far East Russia with emphasis on depleted harzburgite AKIRA ISHIWATARI 1 , SERGEI D. SOKOLOV 2 & SERGEI V. VYSOTSKIY 3 1

Department of Earth Sciences, Faculty of Science, Kanazawa University, Kanazawa 9201192, Japan (e-mail: [email protected]) 2 Geological Institute, Russian Academy of Sciences, Pyzhevsky 7, Moscow 109017, Russia 3 Far East Geological Institute, Russian Academy of Sciences, Prospect 100 letiya 159, Vladivostok 690022, Russia Abstract: Ophiolites are divided into Iherzolite-type (L-type) and harzburgite-type (H-type) by the lithology of their mantle peridotites. Rare depleted harzburgite-type (DH-type) is distinguished from the normal H-type by the more refractory nature of its mantle peridotite and the occurrence of orthopyroxene-type cumulate rocks including iron-rich harzburgite and orthopyroxenite. The Shelting (Sakhalin) and Krasnaya (Koryak Mountains) ophiolites in Far East Russia, which have both depleted harzburgite and orthopyroxene-type cumulate rocks, belong to this newly defined DH-type. The ophiolites in SW Japan-Primorye, NE JapanSakhalin, and the Koryak Mountains in the northwestern Pacific margin have diverse ophiolite types ranging from L- to DH-types. The wide petrological diversity, the common occurrence of DH-type, and the presence of thick crustal sections in these ophiolites suggest regionally inhomogeneous, commonly very high degrees of mantle melting over subduction zones, as in the modern Mariana forearc environment. The ophiolites of Japan and Far East Russia range in age from Early Palaeozoic to Cenozoic and are tectonically underlain by younger blueschists and accretionary complexes. The spatial association of these ophiolites with blueschists is analogous to the ophiolite-blueschist assemblages recovered from the Mariana forearc. This association might have formed in a period of non-accretion at an oceanic subduction zone that was followed by voluminous accretion of sediments, facilitating subsequent uplift of the ophiolites and blueschists.

The concept of ophiolites as assemblages of mafic and ultramafic igneous rocks formed in deep ocean was first suggested by Steinmann (1927), and was further developed into the current model of obducted fossil oceanic crust-mantle by Moores (1969), Coleman (1971) and Moores & Vine (1971). This model helped to explain spreading processes at constructive plate boundaries and tectonic emplacement processes at destructive plate boundaries as the theory of plate tectonics was formulated. Ophiolites are currently used to help identify suture zones and ancient continental collision zones where they rest tectonically on older continental crust. Coleman (1986) called these ophiolites 'Tethyan-type' and distinguished them from those in Circum-Pacific belts, which were labelled 'Cordilleran-type'. In general, Cordilleran-type ophiolites are incomplete, dismembered and metamorphosed; however, they are none the less useful for unravelling global tectonic problems. The first purpose of this paper is to show that the Cordilleran-type ophiolites in the northwestern Pacific margin are emplaced over

young continental crust (or an accretionary complex younger than the ophiolite), are associated with blueschists, and represent remnants of fossil oceanic subduction zones rather than continental collision zones. Boudier & Nicolas (1985) classified ophiolites into Iherzolite- and harzburgite-types (designated LOT and HOT, respectively, by Nicolas (1989)). They postulated that HOT form at mid-ocean ridges with relatively fast spreading rates. However, recent studies on contrasting structural features of the Cyprus and Oman ophiolites suggest that these two bodies, both HOT, could be correlated with slow- and fast-spreading ridges, respectively (Dilek et al. 1998). Petrological and geochemical studies have revealed that at least some Tethyan ophiolites were generated from suprasubduction zone (SSZ) magmas rather than mid-ocean ridge basalt (MORE) (e.g. Cyprus: Miyashiro 1973; Robinson et al. 1983; Hebert & Laurent 1990; Oman: Alabaster et al. 1982; Umino et al. 1990; Lachize et al. 1996; Ahmed & Arai 2002; Bay of Islands: Edwards 1995; Suhr &

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 597-617. 0305-8719/03/$ 15 © The Geological Society of London 2003.

598

A. ISHIWATARI £T,4Z,

Edwards 2000). Pearce et al (1984) classified most HOT as SSZ-type and considered all LOT as MORB-type. Ishiwatari (1985a) divided harzburgite-type ophiolites into ordinary harzburgite (H) and depleted harzburgite (DH) categories (Fig. 1), and pointed out that DH-type ophiolites include olivine (Ol)-orthopyroxene (Opx) cumulates. In contrast, L- (Iherzolite) and H-type ophiolites have olivine-plagioclase (PI) and olivine-clinopyroxene (Cpx) cumulates, respectively. Although the crystallization order of minerals may change with pressure, all ophiolitic cumulates are believed to have crystallized at low, plagioclase-stable pressures, and compositional variations of cumulus minerals suggest that the crystallization sequence

was not pressure dependent (Ishiwatari 1985a). The associated volcanic rocks also typically show similar chemical variations. The residual peridotite of ophiolitic mantle sections is generally homogeneous over a scale of kilometres, if we exclude samples from the Moho transition zone and from dykes or veins of dunite, pyroxenite or gabbro. The mineral chemistry of such residual peridotite is the most reliable measure of the degree of melting of the mantle section. Variations in the Cr-number (Cr/(Al + Cr)) of spinel in the residual peridotite correspond closely to variations in lithology and mineral chemistry. The Cr-number is 0.3-0.5 (or less) in L-type, 0.5-0.7 in H-type, and 0.7-0.9 in DH-

Fig. 1. Petrological types of ophiolites, after Ishiwatari (1991), and examples of each type in the northwestern Pacific margin. See Ishiwatari (1985a, 1991) for references. * Lewis Hills massif of the Bay of Islands ophiolite shows a more depleted nature (Edwards 1995).

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA type ophiolites (Fig. 1), although the associated dunite and chromitite may have more Cr-rich spinel. An increase in spinel Cr-number with degree of melting was demonstrated experimentally by Jaques & Green (1980) and was related to peridotite compositions by Dick & Bullen (1984). Although the melting process may be complicated by hydrous remelting of already depleted mantle (e.g. Bloomer & Hawkins 1987; Sobolev & Danyushevsky 1994) and later reaction with percolating melts (e.g. Cannat et al. 1990; Arai et al. 1996), the spinel Cr-number of regionally homogeneous residual peridotite probably reflects the degree of melting of the primary fertile mantle. Therefore, we consider the spinel Cr-number of the residual peridotite as a significant parameter for classifying ophiolites in this paper. The second purpose of this paper is to report new examples of DH-type ophiolites in the northwestern Pacific region, where L- and H-types are also abundant, and to consider the geological significance of the extreme diversity of the ophiolitic mantle in this region. The geological and petrological data on the Russian ophiolites presented in this paper are mostly based on co-operative field studies in Primorye (1993, 1998, 1999 and 2002) and the Taigonos Peninsula (1995 and 1997) as well as on

599

a 1990 field workshop in the Mainits Zone, Koryak Mountains (Bryan 1991). Some of the important chemical data and geological maps reproduced here are from publications in Japan and Russia.

Ophiolites in the northwestern Pacific margin The northwestern Pacific margin, which extends from Japan to Russia, contains numerous ophiolitic bodies commonly associated with blueschists (Fig. 2). These ophiolites are highly variable in age, lithology and chemical composition, even in an apparently continuous, 'single' belt. The ages of the ophiolites and associated blueschists are summarized in Table 1.

Ophiolite belts in SW Japan and Primorye The Palaeozoic-Mesozoic accretionary complexes and associated ophiolites and blueschists of southwestern Japan may have been linked with the Sikhote-Alin Mountains of Russian Primorye before the Miocene opening of the Japan Sea (Ishiwatari & Tsujimori 2003). Jurassic accretionary complexes in Japan (Tamba, Mino and Ashio

Fig. 2. Location of major ophiolite belts and ophiolite complexes in the northwestern Pacific margin. A possible western extension of the Palaeozoic ophiolite-blueschist belt of the SW Japan-Primorye area is also shown (see Ishiwatari & Tsujimori (2003) for discussion).

600

A. ISHIWATARI ETAL zones) are believed to correlate with the Samarka-Nadanhada terranes in Primorye and northeastern China (Kojima 1989). Palaeozoic ophiolites in southwestern Japan include the Ordovician Oeyama ophiolite and the Permian Yakuno ophiolite. The hornblende K-Ar age of the Oeyama ophiolite is 450 Ma (Nishimura 1998), whereas that of the Yakuno ophiolite is 280 Ma (Shibata et al 1977). The latter age is in agreement with zircon U-Pb dates from the same body (Herzig et al. 1997). The Oeyama ophiolite is composed of Iherzolitic mantle peridotite (Cr-number 0.3) in the eastern Chugoku area (Kurokawa 1985) and of clinopyroxene-bearing harzburgite (Cr-number 0.5) in the western Chugoku area. Podiform chromitites (Cr-number 0.5) encased in dunite bodies are abundant in the western area (Arai 1980; Matsumoto & Arai 1997) (Fig. 3a). In both areas the mantle peridotites commonly contain vermicular spinel-pyroxene aggregates. The mantle peridotites are thrust over the 320-280 Ma Renge blueschist (Tsujimori & Itaya 1999), the Permian Akiyoshi accretionary complex, and the Permian Yakuno ophiolite. Small mantle peridotite bodies of analogous nature are also present in the Hida marginal belt to the east, where they are associated with the Renge blueschist and eclogite in the Omi area (Tsujimori et al. 2000). Orbicular chromitite with high Cr-number (0.76-0.85) is present in peridotites in the Omi area (Yamane et al. 1988). The high Cr-number suggests formation from highly refractory melts, although the degree of melting of the mantle peridotite is generally low (Fig. 3a). This ophiolite includes minor gabbroic rocks and ultramafic cumulates (Kurokawa 1985), and granulite-facies spinel metagabbro recrystallized to kyanite-bearing epidote amphibolite has been reported from the Oeyama massif (Tsujimori & Ishiwatari 2002). The Permian Yakuno ophiolite has a complete succession with MORB-type basalt, clinopyroxene-type cumulates, and mantle peridotite consisting of relatively depleted harzburgite with a variable composition (Cr-number 0.4-0.8) (Ishiwatari 1985a). Although regional variations in the mantle peridotite cannot be confirmed because of limited exposure, the gabbroic and basaltic rocks range from T-MORB in the eastern part to island arc tholeiite in the western part (Ishiwatari et al. 1990). In gabbros the Mg-number of clinopyroxene varies from 0.85 in the eastern part to 0.7 in the western part (Fig. 4a) although plagioclase has a relatively uniform composition (An80) throughout. Another peculiar feature of the Yakuno ophiolite is the occurrence of granulite-facies metacumulates (e.g. spinel-two-pyroxene metagabbro) at the Moho level, which suggests unu-

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA

601

Fig. 3. Compilation of spinel Cr-number and A12O3 content of coexisting orthopyroxene in residual mantle peridotite of ophiolites in (a) SW Japan-Primorye area, (b) NE Japan-Sakhalin area, (c) Taigonos Peninsula and (d) Mainits Zone, Koryak Mountains. References: (a) Arai (1980), Ishiwatari (1985a, 1985b), Kurokawa (1985), Yamane et al. (1988), Matsumoto & Arai (1997), Shcheka et al. (2001); (b) Ishizuka (1985, 1987), Ozawa (1988), Takahashi (1991), Tamura et al (1999), Vysotskiy et al. (2000); (c) Saito et al. (1999), Bazylev et al. (2001); (d) Dmitrenko et al. (1990) and this study. HMLS, main harzburgite-therzolite suite; BDH, banded durite-harzburgite.

sually thick (15-30 km) oceanic crust (Ishiwatari 1985b). In Primorye, the Sergeevka, Kalinovka and Bikin ophiolitic complexes lie along the western margin of the Mesozoic accretionary complexes near the boundary of the Khanka crystalline massif (Fig. 2). These ophiolites lack mantle peridotite and are composed solely of volcanic rocks, gabbros and mafic-ultramafic cumulates.

On the other hand, the Cambrian Khanka ophiolite rests on a Cambrian limestone in a small rift zone (Spassk zone) in the Khanka massif. Middle Cambrian conglomerate covering the body contains abundant chromian spinel grains derived from the ophiolite (Shcheka et al. 2001). This ophiolite contains serpentinized harzburgite with relatively chromian, unusually Mn-rich spinel (Crnumber 0.6-0.7) (Shcheka et al. 2001); it is

602

A. ISHIWATARI ETAL

suggest that the Yakuno ophiolite may also correlate with the Kalinovka ophiolite, which includes spinel-bearing troctolite and garnet-bearing metagabbro (Ishiwatari & Tsujimori 2003).

Ophiolite belts in NE Japan and Sakhalin

Fig. 4. Relationship between anorthite (An) content of plagioclase and Mg-number (= Mg/(Fe + Mg)) of coexisting clinopyroxene in the Yakuno ophiolite in SW Japan (a) and among the Elistratova ophiolite and Kengevayam gabbro body, Taigonos Peninsula, NE Russia (b). Fields for island arc basalts (IAB) and midocean ridge basalts (MORE) are from Ishiwatari et al. (1990). Beard (1986) and Hebert & Laurent (1990) proposed a similar discrimination between the two suites.

considered to be an H-type ophiolite transitional to DH-type (Fig. 3a). The residual peridotite spinels of the Khanka and Oeyama ophiolites suggest that the early Palaeozoic mantle of the SW Japan-Primorye belt was inhomogeneous and was locally relatively depleted. Dobretsov et al, (1994) correlated the Oeyama ophiolite with the Kalinovka ophiolite; however, age and structural data indicate that it is part of the Sergeevka terrane, a huge metagabbro body covered by Devonian and Permian sedimentary rocks. The Sergeevka terrane was later thrust onto a Jurassic accretionary complex (Samarka terrane) containing 250 Ma blueschist. The Sergeevka terrane, however, does not contain any mantle peridotite, which is abundant in the Oeyama ophiolite. The Bikin ophiolite, which consists of granulitefacies two-pyroxene metagabbro (Vysotskiy 1994) may be the Russian counterpart of the Permian Yakuno ophiolite. Recent geochronological data

The early Palaeozoic Miyamori (and Hayachine) ophiolite is present in the Kitakami Mountains, NE Honshu (Ozawa 1994). This ophiolite and its overlying Silurian to Jurassic sedimentary rocks were thrust over the Jurassic accretionary complex of the North Kitakami-SW Hokkaido zone (Tazawa 1988). The ophiolite consists mainly of moderately depleted harzburgite (spinel Cr-number 0.40.75) and clinopyroxene-type cumulates. The pervasive occurrence of hornblende (and some Tipoor phlogopite) in the mantle peridotite indicates that it was derived from a hydrous mantle wedge above a subduction zone (Ozawa 1988). The harzburgite contains layered harzburgite-wehrlite zones, which formed by melt-mantle interaction (Ozawa 1994). The Miyamori ophiolite is associated with the Motai blueschist of Late Palaeozoic age (Maekawa 1988). In Central Hokkaido, ophiolitic rocks occur in several zones. In the Sorachi-Yezo belt, the Jurassic Horokanai ophiolite is thrust over the Cretaceous Kamuikotan metamorphic belt, and dismembered equivalents of this ophiolite are distributed throughout the belt. The Horokanai ophiolite has a well-preserved succession, which is composed, from bottom to top, of harzburgite, orthopyroxenite-dunite cumulate rocks, metagabbros, amphibolite, MORB-type pillow lavas and an Upper Jurassic radiolarian chert. The mantle harzburgite is strongly depleted (spinel Cr-number 0.69-0.77 in harzburgite and 0.81-0.93 in dunite) (Fig. 3b) (Ishizuka 1985, 1987). In the Kamuikotan belt, the Takadomari harzburgite body near Horokanai is similarly highly depleted, but the Iwanai-dake body contains common harzburgite and the southernmost Nukabira body consists of Iherzolite (Furue et al. 1997; Tamura et al. 1999) (Fig. 3b). Tamura et al. (1999) also reported dunite with highly chromian spinel and magnesian olivine in the Nukabira Iherzolite (Table 2). The depleted harzburgite of the Horokanai and Takadomari complexes has been thrust onto the Kamuikotan metamorphic rocks, which include typical jadeite-bearing blueschists (Ishizuka 1985, 1987; Sakakibara & Ota 1994). In the Hidaka belt, to the east of the Kamuikotan belt, the Poroshiri ophiolite has a relatively complete succession (Miyashita 1983). This body contains abundant troctolite-anorthosite cumulates (Miyashita & Hashimoto 1975), whose olivine-plagioclase crystallization trend closely

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA follows that of MORJB, and which is thought to be an L-type ophiolite. Based on geological data, Miyashita & Yoshida (1988) postulated a Cretaceous age for the ophiolites in the Hidaka zone. The Horoman complex is a well-layered peridotite-gabbro body in the southern part of the Hidaka belt. The main harzburgite-lherzolite complex has spinel with Cr-number 0.2-0.6 and contains layers of spinel-rich dunite-wehrlite, gabbro, and minor banded dunite-harzburgite (spinel Cr-number 0.8-0.9) (Takahashi 1991). The pervasive occurrence of spinel-pyroxene symplectite after garnet suggests a deep mantle origin. Rare, large corundum crystals in the metagabbro suggest that the peridotite body is a fragment of recycled oceanic crust-mantle, which has been subducted and then emplaced as a diapir (Morishita & Arai 2001). L-type residual mantle and mafic-ultramafic cumulate rocks constitute the bulk of this peridotite body, but the occurrence of highly depleted harzburgite indicates that a preexisting DH-type ophiolite was also incorporated into the diapir. Dobretsov et al. (1994) compared ophioliteblueschist belts in central Hokkaido and Sakhalin and correlated the West and Central (LangeriSusunai) Sakhalin Zone with the Sorachi-Yezo and Kamuikotan belts in Hokkaido. Vysotskiy et al. (2000) described the boninite-bearing Shelting ophiolite from central-eastern Sakhalin (Fig. 5). This ophiolite, composed of dunite-harzburgite, websterite-orthopyroxenite and gabbronorite units, occurs as a tectonic slice in fault contact with Jurassic-Cretaceous volcaniclastic rocks containing boninite (the Rakitinskaya suite). The nearby Berezov massif (Fig. 5) has the same character and lithological sequence. Spinel in the dunite and harzburgite of this ophiolite is unusually chromian (Cr-number 0.86-0.89) (Fig. 3b, Table 2). The highly depleted nature of the mantle peridotite and the presence of the adjacent orthopyroxene-type cumulate rocks help define this ophiolite as a DH-type. The associated boninite has 25-30 modal % orthopyroxene phenocrysts, which show reverse zoning with an iron-rich core (Mg-number 0.72-0.75) and a magnesian rim (Mg-number 0.84-0.89). The spinel microphenocrysts also show reverse zoning with iron-rich cores and iron-poor rims (Cr-number 0.80-0.83). Vysotskiy et al. (2000) suggested that crystallization of the boninitic magma under more and more reducing conditions was due to the introduction of a hydrogen-rich fluid.

Ophiolite belts in the Koryak Mountains The Koryak Mountains extend from northern Kamchatka to the Bering Straits (Fig. 2). Fujita &

603

Newberry (1982) described the general geology of this area and Palandzhjan (1986) studied the ultramafic rocks. Fujita & Newberry (1982) proposed that the Koryak ophiolites were emplaced in the Jurassic. Pushcharovsky et al. (1988) and Sokolov (1992) outlined the geological structure of the Koryak Mountains, in which Early Palaeozoic ophiolites of the Penzhina zone (also called the Talovsko-Pekulneyskaya zone or the UstBelskaya zone) are thrust over the Koryak nappes, which contain abundant Mesozoic and rare Palaeozoic ophiolites. Stavsky et al. (1990) proposed a plate-tectonic model for accretion in the Koryak Mountains, and Tilman & Bogdanov (1992) produced a comprehensive geotectonic map of the area. Dobretsov (1999) compiled chronological and petrological data from blueschists in this area.

Penzhina zone and Taigonos Peninsula The Penzhina zone comprises the innermost part of the Koryak orogenic belt, where ophiolites and blueschists of various ages are exposed (Fig. 2). The Pekulney Range in the northern Penzhina zone has a large Proterozoic(?) metamorphosed mafic-ultramafic complex and Jurassic ophiolites (Dobretsov 1999). The Ust-Belaya ophiolite of Early Palaeozoic age (560 Ma K-Ar age on gabbro) consists of tectonic slices of pillow lava with chert, dunite-peridotite with gabbro-anorthosite layers, and gneissose metagabbro with eclogitic rocks (Dobretsov 1999). The entire complex was thrust onto the Late Jurassic-Early Cretaceous accretionary complexes of the Koryak nappe system (Sokolov 1992). The Ganychalan ('Kharitoninskii' of Dobretsov) nappe in the southern Penzhina zone also has a complete ophiolite sequence of Early Palaeozoic age (430 Ma). The Kuyul ophiolite in the southern Penzhina zone, mid-Triassic to Jurassic in age, has a complete succession, whose chemistry and mineralogy have an SSZ affinity (Khanchuk & Panchenko 1994; Luchitskaya 1996). The mineral chemistry of the harzburgite indicates that this is an H-type ophiolite. According to Dobretsov (1999), blueschists in the Koryak Mountains show chronological peaks at 330 Ma (Penzhina), 300 Ma (Penzhina), 180 Ma (Penzhina and Taigonos) and 150 Ma (Penzhina, Pekulnei and melange blocks). Associated ophiolites have ages of 380450 Ma, 208-290 Ma and > 150 Ma, and the blueschists of each period are slightly younger than the adjacent ophiolites. The Taigonos Peninsula constitutes the southernmost part of the Penzhina Zone (Figs 2 and 6). Several ophiolite bodies are present along the eastern coast of the peninsula, such as the Elistratova ophiolite, the Kengeveem metagabbro

Table 2. Representative mineral analyses of residual peridotites from ophiolite complexes in Hokkaido and northeastern Russia

Sample: Mineral: Wt% oxide Si02 TiO2 A1203 Cr203 Fe203 FeO MnO MgO CaO Na20 Total Cation number O= Si Ti Al Cr Fe3+ Fe2+ Mn Mg Ca Na Total Mg-no. Cr-no. Fe3+-no. Ol Mg-no.

Opx

Cpx

56.03 0.04 3.83 0.92

53.08 0.15 3.84 1.06

5.52 0.13 31.34 2.64 0.00 100.45

2.21 0.14 16.45 23.85 0.00 100.78

6 1.930 0.001 0.155 0.025

6 1.913 0.004 0.163 0.030

0.159 0.004 1.608 0.097 0.000 3.979 0.910 0.139

0.067 0.004 0.884 0.921 0.000 3.986 0.930 0.156

0.905

OU09-2 Hz

NB2643 Du

NB2811 Lh Spinel

Shelting

Takadomari

Nukabira

Massif:

Spinel

0.00 40.33 28.88 0.72 13.78 0.17 15.96

0.13 11.11 57.52 2.24 19.31 0.23 9.38

99.84

99.92

4

4

0.000 1.341 0.644 0.015 0.325 0.004 0.671

0.003 0.433 1.505 0.056 0.534 0.006 0.463

3.000 0.674 0.324 0.008

3.000 0.464 0.777 0.028 0.929

Southern Taigonos Peninsula [3]

Central -Eastern Sakhalin [2]

Kamuikotan Belt, Hokkaido [1]

Area:

Opx

Cpx

58.14 0.07 1.24 0.27

54.24 0.01 0.98 0.55

4.96 0.08 34.70 0.38 0.00 99.84

1.57 0.05 16.89 25.14 0.00 99.43

6 1.991 0.002 0.050 0.007

6 1.980 0.000 0.042 0.016

0.142 0.002 1.770 0.014 0.000 3.979 0.926 0.127

0.048 0.002 0.919 0.983 0.000 3.990 0.950 0.273

0.928

OU17-2 Du Spinel

Spinel

0.06 17.13 55.23 0.58 16.89 0.00 12.15

0.18 6.17 65.23 1.26 16.89 0.00 10.76

102.04

100.49

4

4

0.001 0.627 1.356 0.014 0.439 0.000 0.563

0.005 0.242 1.718 0.031 0.470 0.000 0.534

3.000 0.562 0.684 0.007

3.000 0.532 0.877 0.016 0.939

Dunite

Harzburgite Opx

Cpx

58.07 0.00 0.21 0.31

55.11 0.00 0.27 0.30

6.27 0.08 33.94 0.73 0.00 99.61

2.57 0.00 18.06 24.05 0.08 100.44

6 2.008 0.000 0.009 0.008

6 1.992 0.000 0.011 0.009

0.181 0.002 1.748 0.027 0.000 3.984 0.906 0.497

0.078 0.000 0.972 0.931 0.003 3.995 0.926 0.427

0.905

Spinel

Spinel

0.06 4.48 55.83 10.05 24.27 0.74 5.14

0.12 6.23 63.05 4.22 13.44 1.74 11.84

100.57

100.64

4

4

0.002 0.185 1.547 0.265 0.711 0.022 0.268

0.003 0.243 1.647 0.105 0.371 0.049 0.583

3.000 0.274 0.893 0.133

3.001 0.611 0.871 0.053

Povorotny

Nablyudeny

1104BLh

1004B*

Opx

Cpx

54.43 0.05 4.61 0.59

52.42 0.14 4.09 0.88

6.41 0.13 32.31 0.81 0.00 99.34

2.66 0.14 16.36 22.87 0.00 99.56

6 1.896 0.001 0.189 0.016

6 1.912 0.004 0.176 0.025

0.187 0.004 1.677 0.030 0.000 4.000 0.900 0.079

0.081 0.004 0.889 0.893 0.000 3.984 0.916 0.126

0.905

Spinel

Spinel

0.13 53.56 15.04 0.44 11.73 0.05 18.80

0.13 10.52 56.54 2.11 23.27 0.34 6.57

99.75

99.48

4

4

0.003 1.670 0.315 0.009 0.260 0.001 0.742

0.003 0.421 1.518 0.054 0.661 0.010 0.333

3.000 0.741 0.159 0.005

3.000 0.335 0.783 0.027 0.932

Northern Taigonos Peninsula [4]

Area:

Mainits Zone, Koryak Mountains

Massif:

Elistratova South

Elistratova North

Tamvatney [5]

Yagel Melange [6]

Sample:

1432 Hz

21 13 Hz

P307/2 Lh

21FLh

Mineral: Wt% oxide SiO2 TiO2 A12O3 Cr203 Fe2O3 FeO MnO MgO CaO Na 2 O Total Cation number O = Si Ti Al Cr Fe3+ Fe2+ Mn Mg Ca Na Total Mg-no. Cr-no. Fe3+-no. Ol Mg-no.

Opx

Cpx

56.52 0.12 3.49 0.70

52.63 0.08 4.27 1.33

5.96 0.16 32.28 1.00 0.00 100.23

2.17 0.13 15.87 24.14 0.18 100.80

6 1.943 0.003 0.141 0.019

6 1.899 0.002 0.182 0.038

0.171 0.005 1.654 0.037 0.000 3.973 0.906 0.119

0.065 0.004 0.853 0.933 0.006 3.983 0.929 0.173

0.907

Spinel

0.00 41.14 27.89 0.85 14.06 0.00 15.99

99.93 4 0.000 1.362 0.620 0.018 0.330 0.000 0.670

3.000 0.670 0.313 0.009

Opx

Spinel

56.84 0.20 1.69 0.47

0.14 20.57 49.95 0.84 16.48 0.17 12.39

5.40 0.23 34.34 0.95 0.00 100.12

100.54

6 1.955 0.005 0.069 0.013 0.155 0.007 1.760 0.035 0.000 3.999 0.919 0.157

0.909

4 0.003 0.751 1.223 0.020 0.427 0.004 0.572

3.000 0.573 0.620 0.010

Opx

Cpx

54.39 0.06 5.20 0.65

51.40 0.17 5.33 0.85

5.98 0.10 32.55 1.32 0.07 100.32

2.50 0.14 16.30 23.12 0.36 100.17

6 1.876 0.002 0.211 0.018

6 1.864 0.005 0.228 0.024

0.172 0.003 1.673 0.049 0.000 4.003 0.907 0.077

0.076 0.004 0.881 0.898 0.000 3.980 0.921 0.097

0.901

Spinel

0.05 51.25 15.46 2.57 10.25 0.18 19.21

98.97 4

0.001 1.619 0.327 0.052 0.230 0.004 0.767

3.000 0.769 0.168 0.026

Opx

Cpx

54.95 0.00 4.30 1.10

53.72 0.10 3.89 1.36

5.27 0.03 32.12 2.23 0.00 100.00

1.87 0.07 16.52 23.10 0.00 100.63

6 1.901 0.000 0.175 0.030

6 1.929 0.003 0.165 0.039

0.152 0.001 1.655 0.083 0.000 3.997 0.916 0.147

0.056 0.002 0.884 0.889 0.000 3.966 0.940 0.190

0.916

Krasnaya [6] 28BHz Spinel

0.10 40.74 28.92 0.54 13.57 0.00 16.40

100.27 4

0.002 1.345 0.640 0.011 0.318 0.000 0.684

3.000 0.683 0.322 0.006

Opx

Cpx

59.00 0.10 0.91 0.39

54.32 0.07 1.02 0.36

5.69 0.22 33.77 1.06 0.00 101.14

1.56 0.09 17.19 24.54 0.00 99.15

6 2.004 0.003 0.036 0.010

6 1.984 0.002 0.044 0.010

0.162 0.006 1.709 0.039 0.000 3.970 0.914 0.223

0.048 0.003 0.936 0.960 0.000 3.987 0.952 0.191

0.910

Spinel

Srednaya [6] 28C Du

17K Du

Spinel

Spinel

0.06 10.76 58.55 2.32 20.08 0.00 9.16

0.16 9.64 58.27 2.65 17.76 0.00 10.13

0.09 6.87 62.97 1.07 19.97 0.00 8.57

100.93

98.61

99.54

4

4

4

0.001 0.417 1.522 0.057 0.552 0.000 0.449

0.004 0.381 1.544 0.067 0.498 0.000 0.506

0.002 0.275 1.693 0.027 0.568 0.000 0.434

3.000 0.448 0.785 0.029

3.000 0.504 0.802 0.034

3.000 0.433 0.860 0.014

0.922

0.915

Data sources: [1] Tamura et at. (1999); [2] Vysotskiy et al. (2000); [3] thi s study (analyst: D. Saito); [4] Saito et al. (1999); [5] Dmitrenko et al. (1990) ; [6] this study (analyst: Y. Furuhashi). Du, dunite; Lh, Iherzolite; Hz, harzburgite; Ol, olivine; Opx, orthopyroxene; Cpx , clinopyroxene. ^Sample 1004B is olivine— talc rock after harzburgite. Analyses for [1], [3], [4] and [6] were performed by the same method using;; the Akashi-30A SEM-EDAX 9100 system of Kanazaiwa University. (See Lopez & Ishiwatari (2002) for further details of these analyses.)

606

A. ISHIWATARI ETAL

Fig. 5. Geological map of the DH-type Shelting Cape ophiolite in central eastern Sakhalin (after Vysotskiy et al. 2000). (See Fig. 2 for location.)

complex, the Nablyudeny complex, and the Povorotny melange. Although the Kengeveem metagabbro complex does not have mantle peridotite, the other three complexes do (Saito et al. 1999; Bazylev et al. 2001). The Nablyudeny harzburgite has the most chromian spinel (Cr-number 0.8), whereas the harzburgites of the Elistratova ophiolite have moderately chromian varieties (Crnumber 0.3-0.7), and the Povorotny Iherzolite has aluminous spinel (Cr-number 0.2) (Table 2). Some spinels in the Povorotny complex are chromian in composition with Cr-number up to 0.7 (Bazylev et al. 2001). Here again, an extreme diversity of residual mantle peridotite is present among these ophiolites (Fig. 3c; Table 2). The Elistratova ophiolite consists of a central cumulate gabbro body, locally cut by sheeted dykes, whereas the northern and southern ultramafic bodies are mostly composed of residual mantle peridotite (Belyi & Akinin 1985; Saito et al. 1999) (Fig. 6). Disseminated spinel in peridotite of the southern body is moderately aluminous (Cr-number 0.30-0.50), whereas that in the northern body is more chromian (Cr-number 0.40-0.65) (Fig. 3c). The gabbroic body is

intrusive into the southern ultramafic body with a clear-cut chilled igneous contact. Many dykes, which contain ultramafic xenoliths, also intrude into the ultramafic rocks. Saito et al. (1999) interpreted the northern ultramafic body and the central gabbroic body as an intact island arc ophiolite. The southern ultramafic body is less depleted than the northern one and may have served as a subvertical wall for the gabbroic magma chamber. The gabbroic rocks vary upward from olivine gabbronorite through gabbronorite to hornblende gabbronorite. The coexistence of extremely calcic plagioclase (about An9o) with relatively iron-rich clinopyroxene (about Mg-number 0.80) clearly indicates an island arc basalt affinity for the magma (Fig. 4b). Podiform chromitite from the northern melange has a Cr-number of 0.70, and the chromite grains are characterized by Fe3+-poor cores containing many hydrous inclusions (hornblende and phlogopite), suggesting the involvement of a hydrous, reducing fluid in podiform chromitite formation (Tsujimori et al. 1999). The Kengeveem metagabbro complex is distinct from the gabbroic section of the Elistratova ophiolite. The metagabbros are almost devoid of

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA

607

Fig. 6. Geological map and schematic cross-section of the Elistratova ophiolite in the Taigonos Peninsula (after Belyi & Akinin 1985; Saito et al. 1999; and our data). An intact ophiolite sequence composed of the northern ultramafic body (normal harzburgite) and central gabbro body, with ultramafic cumulates between them, and sheeted diabase dyke complexes in the gabbro, intrudes into the southern ultramafic body (less depleted harzburgite).

orthopyroxene, which is abundant in the Elistratova gabbros, but contain abundant clinopyroxene (Mg-number 0.80) and plagioclase (Ango) (Fig. 4b), indicating a MORE affinity. On its northern side, the Kengeveem gabbro body is associated with Ordovician sedimentary rocks, and thus they may be significantly older than the Mesozoic ophiolites of the Taigonos Peninsula. This body may possibly be the equivalent of the Early Palaeozoic Ganychalan ophiolite in the Penzhina zone.

Mainits zone The Mainits zone in the central Koryak Mountains is characterized by the occurrence of many ophiolites associated with Jurassic-Cretaceous accretionary complexes (Stavsky et al. 1990). Within an area of 40 km X 100 km, there are several ophiolitic bodies such as the Tamvatney Iherzolite massif, the Yagel serpentinite melange, the Krasnaya Mountain harzburgite nappe, and the Sred-

naya Mountain dunite body, all of which are associated with island arc type gabbro and volcanic rocks (Fig. 7). The Tamvatney body contains mostly Iherzolite (spinel Cr-number 0.20-0.30) (Table 2) with minor harzburgite (spinel Cr-number 0.35-0.50) (Dmitrenko et al. 1990), and contains some eclogitic inclusions. The Yagel serpentinite melange consists of blocks of Iherzolite, harzburgite (spinel Cr-number 0.4), dunite, gabbro, picritic sheeted dykes and pillow basalt. On the other hand, the Krasnaya Mountain ultramafic complex, which occurs as a subhorizontal nappe thrust onto the Jurassic-Cretaceous accretionary complexes and the Yagel melange, contain highly chromian spinel (Cr-number 0.8) and highly magnesian olivine (Fo90) (Fig. 3d, Table 2). The geological map of Dmitrenko et al. (1985) shows that the southern half of the complex consists mostly of depleted harzburgite, whereas the northern half is composed of ultramafic cumulates and dunite with orthopyroxenite veins (Fig. 8). The ultramafic cumulate includes iron-

608

A. ISHIWATARI ETAL

Fig. 7. Ophiolitic rock complexes of the Mainits Zone in the Koryak Mountains, NE Russia (simplified from Stavsky et al. 1990). (See Fig. 2 for location.)

rich harzburgite, dunite, orthopyroxenite and websterite. The Krasnaya complex has a lithological assemblage typical of the DH-type ophiolite, although gabbroic and volcanic rocks are missing. The Serdnaya Mountain dunite body includes dunite and chromitite. Disseminated spinel in the dunite is highly chromian (Cr-number 0.85) (Fig. 3d, Table 2). Although the Mainits Zone is a small area, both L-type and DH-type ophiolites are present, reflecting wide petrological diversity in the residual mantle peridotite.

Discussion Multiple nappe pile of ophiolites of various ages Irwin (1981) first described multiple nappe piles of ophiolites from the Klamath Mountains of the western USA, where the Jurassic Josephine ophiolite is technically overlain by an Upper Palaeozoic-Triassic ophiolite, which is in turn structurally overridden by the Lower Palaeozoic Trinity ophiolite, with intervening blueschists and accretionary complexes. Emplacement of older ophiolites over younger accretionary complexes and ophiolites is a regular rule not only for the Klamath Mountains but also for SW Japan (the Upper Palaeozoic Yakuno ophiolite is technically overlain by the Lower Palaeozoic Oeyama ophiolite), NE Japan (the Lower Palaeozoic Miyamori ophiolite is thrust over the Jurassic accretionary

complex and Jurassic-Cretaceous ophiolites in central Hokkaido), and northeastern Russia (Lower Palaeozoic ophiolites of the Penzhina zone are thrust over the Koryak nappe system containing abundant Mesozoic ophiolites). The Jurassic(?) Mikabu ophiolite and the Tertiary Mineoka (Setogawa) ophiolite are both present on the Pacific side of SW Japan, and one of the youngest ophiolites on Earth is present in East Taiwan (further review and references have been given by Ishiwatari (1991, 1994)). This is consistent with the long-lasting (Phanerozoic) subduction and accretion of the circum-Pacific erogenic belts.

Petrological diversity of the northwestern Pacific margin ophiolites Nicolas & Jackson (1972) demonstrated that Iherzolite is the dominant mantle peridotite in the western Mediterranean ophiolites, whereas harzburgite dominates in the eastern Mediterranean. This implies that the mantle composition is relatively homogeneous for more than 1000 km along the ophiolite belt, although small-scale heterogeneities may occur in the boundary area (e.g. in the Balkan Peninsula). Harzburgite represents residual mantle left after higher degrees of melting of Iherzolite, and the degree of melting is represented by various mineralogical parameters such as increasing Cr-number of spinel, increasing Fo content of olivine, and decreasing A^Os content of

Fig. 8. Geological map of the DH-type Mt. Krasnaya ultramafic complex (possibly the lower part of an ophiolite) after Dmitrenko et al. (1985). (See Fig. 7 for location.)

610

A. ISHIWATARI ETAL

orthopyroxene (Ishiwatari 1985a). DH-type ophiolites have not been found in the eastern Mediterranean area. However, the degree of depletion may vary significantly within a single 'Tethyan-type' ophiolite as in the Bay of Islands (Edwards 1995), and high-Cr spinel forms chromitite ore in many ophiolites far from the Pacific Rim, such as in Albania (Bulqiza massif: Cina et al. 1987) and the southern Urals (e.g. Kempirsai massif: Melcher et al. 1997), where boninitic volcanic rocks with high-Cr spinel microphenocrysts are also present (Spadea & Scarrow 2000). In these massifs, the high-Cr spinel is generally restricted to chromitite ores, and associated dunites and disseminated spinel in the surrounding harzburgite are more aluminous (e.g. Cr-number 0.41 in Kempirsai). Mantle sections composed totally of highly depleted harzburgite (spinel Cr-number >0.7) are characteristic of the western Pacific ophiolite belts. Such harzburgite is associated with orthopyroxene-type cumulates, which may have crystallized from boninitic or island arc tholeiitic melts, suggesting an SSZ origin of these ophiolites. The Shelting ophiolite in Sakhalin and the Mt. Krasnaya ultramafic complex in the Koryak Mountains are new examples of the DH-type ophiolites. Previously known DH-type ophiolites include Horokanai-Takadomari (Hokkaido; Ishizuka 1985, 1987; Tamura et al. 1999), Papua (England & Davies 1973; Jaques & Chappell 1980) and Adamsfield (Tasmania; Varne & Brown 1978). Highly depleted harzburgite (spinel Cr-number 0.77) and dunite (spinel Cr-number 0.81) were drilled from a conical serpentinite seamount in the Mariana forearc. Relatively fertile harzburgite (Crnumber 0.36) and intermediate varieties were also recovered from this site (Ishii et al. 1992), and the spinel Cr-number of all mantle peridotites from this seamount averages 0.61 (Fig. 9a). Harzburgite samples from the nearby landward walls of the Mariana Trench also have Cr-rich spinel (Crnumber 0.55-0.69) (Bloomer & Hawkins 1983), and a forearc seamount in the Izu islands has similar rocks. However, Iherzolite samples with spinel Cr-number of 0.27 were recovered from the Mariana Trough, an active back-arc rift zone (Ohara et al. 2002). Mantle peridotites from the extinct Parece Vela rift zone and southern Mariana Trench walls near Yap Island contain spinels of intermediate composition (Cr-number 0.43-0.52) (Bloomer & Hawkins 1983; Ohara et al. 1996) (Fig. 9a). On the other hand, peridotites from fracture zones of the slow-spreading South Atlantic Southwest Indian ridges such as Islas Orcadas, Vulcan and Bullard have aluminous spinels (Crnumber 0.15-0.30) (Dick 1989). The Bouvet

fracture zone has peridotites with spinel Cr-number of 0.34-0.55, similar to peridotites of fracture zones along the fast-spreading East Pacific Rise (such as the Garrett fracture zone) and the Hess Deep (Cr-number 0.35-0.45) (Cannat et al. 1990; Arai et al. 1996; Edwards & Malpas 1996). Spinels from forearc peridotites are clearly more Cr rich and more depleted when compared with those from mid-oceanic ridge peridotites. Boninites from the Mariana and Tonga forearcs have highly chromian spinel (Cr-number 0.70-0.90) and are believed to have formed by hydrous melting of a depleted mantle (Bloomer & Hawkins 1987; Sobolev & Danyushevsky 1994). Primitive magnesian andesite, containing chromian spinel (Cr-number >0.74) and magnesian orthopyroxene (Mg-number 0.88), has also been reported from Oligo-Miocene island arc volcanic rocks of Japan in association with tholeiitic basalt and calc-alkaline andesite (Lopez & Ishiwatari 2002). The occurrence of diverse mantle peridotites including highly depleted harzburgite (spinel Cr-number >0.70) in the Mariana forearc suggests that the ophiolite belts of the northwestern Pacific margins also originated in intra-oceanic SSZ environments extending from landward trench walls to back-arc basins. It should be noted, however, that the variation in Os isotopic composition among ophiolitic chromites, including those from the northwestern Pacific margins, is significantly less than has been reported for oceanic peridotites and MORB. Thus, there is little evidence to suggest modification of the mantle's original Os isotopic composition via radiogenic melts or fluids derived from subducting slabs (Walker et al. 2002). This may partly reflect the extremely low concentration of platinumgroup elements in such fluids but essentially suggests that simple, high-temperature melting of homogeneous, depleted MORB mantle (DMM) is the major ophiolite-forming process.

Occurrence of the ophiolites with thick oceanic crust Granulite-facies, two-pyroxene spinel metagabbro (sometimes with garnet) occurs at the Moho level of some circum-Pacific ophiolites in Japan, eastern Russia and Alaska (Ishiwatari 1985b; DeBari & Coleman 1989; Vysotskiy 1994; Tsujimori & Ishiwatari 2002). This observation suggests a relatively thick oceanic crust for these ophiolites, which are believed to have formed in island arc (DeBari & Coleman 1989), marginal basin (Ishiwatari 1985b), and/or oceanic plateau (Isozaki 1997) environments. Granulite-facies metagabbros (mostly spinel and garnet free) have also been

Fig. 9. (a) Tectonic framework of the Western Pacific area (with cross-section profiles marked). Spinel Cr-number data of mantle peridotites dredged and drilled from the IzuMariana island arc system are also shown. Average and standard deviations are indicated (av. ± s.d.) if many analyses are available (number in parentheses), but only a value or range is shown if three or fewer analyses are available. Data sources: [1] Ohara et al. (2002); [2] Ohara et al. (1996); [3] Ishii et al. (1992); [4] Bloomer & Hawkins (1983). (b) Schematic cross-section of the Nankai Trough, a typical accreting subduction zone with well-developed accretionary complexes (based on Taira 1985). (c) Schematic crosssection of the Mariana Trench (based on Bloomer & Hawkins 1983; Bloomer & Fisher 1987; Maekawa et al. 1993, 1995), a typical non-accreting subduction zone with outcrops of ophiolitic rocks on the landward trench slope and blueschist occurrences in a forearc serpentinite seamount.

612

A. ISHIWATARI ETAL

recovered from mid-oceanic ridges, but Gaggero & Cortesogno (1997) estimated that the pressure of formation for these rocks was 0.3 GPa (c. 10km depth). This is much lower than the pressure (0.5-1.1 GPa) estimated for ophiolitic granulite-facies metagabbros (Ishiwatari 1985&; DeBari & Coleman 1989; Vysotskiy 1994; Tsujimori & Ishiwatari 2002). Coffin & Eldholm (2001) pointed out that sections of oceanic crust in large igneous provinces (LIPs) are two or five times thicker than those of 'normal' oceanic crust and postulated that some ophiolites are of LIP origin. Xenoliths of granulite-facies, two-pyroxene spinel metagabbro have been reported in alkali basalt from the Kerguelen Archipelago, a typical LIP (Gregoire et al. 1998). However, analogous spinel metagabbro xenoliths are also known from continental margins (e.g. Francis 1976), although higher-pressure garnet granulites may comprise typical lower continental crust (e.g. Loock et al. 1990). Based on their petrology, mineralogy and geochemistry, circum-Pacific ophiolites with relatively thick mafic crust represent SSZ lithosphere (DeBari & Coleman 1989; Ishiwatari et al. 1990). Many are technically underlain by blueschist and younger accretionary complexes, suggesting that they represent the hanging wall of the subduction zone (i.e. mantle wedge and overlying crust). The Os isotope character of chromitite in the Yakuno ophiolite does not support an origin in a superplume-related LIP, which would be characterized by an isotopically distinct HIMU or enriched mantle (Walker et al. 2002).

Origin of the ophiolite-blueschist assemblage Japanese ophiolites commonly have metamorphic soles composed of blueschist and are tectonically underlain by younger, sediment-rich accretionary complexes, which contain greenstones of OIB and MORB origin (Table 1). For example, in southwestern Japan, the Ordovician Oeyama ophiolite (>450Ma) is underlain by the 320 Ma Renge blueschist and the Upper Permian (250 Ma) Akiyoshi accretionary complex. This spatial relationship suggests tectonic erosion or non-accretion during the intervening Siluro-Devonian time. A similar gap exists between the Lower Permian (280 Ma) Yakuno ophiolite and the underlying Jurassic Tamba accretionary complex (150 Ma). The accretionary complex is characterized by 'oceanic plate stratigraphy' composed of greenstone, chert, limestone, mudstone and sandstone in a younging order (Isozaki 1997). The basal greenstone commonly includes tholeiitic and alkaline

seamount basalt (OIB) with high Ti and Nb concentrations, but the ophiolite itself is almost free of OIB. In the present-day western Pacific, the IzuMariana and Tonga subduction zone environments are characterized by the presence of ophiolite outcrops on the trench slopes (Bloomer & Hawkins 1983; Bloomer & Fisher 1987) and blueschists (Maekawa et al. 1993, 1995), by the absence or scarcity of accretionary complexes, and by the currently active back-arc spreading. In contrast, areas off northeastern Honshu and Hokkaido are characterized by the development of vast accretionary complexes (Taira 1985) without submarine ophiolite or blueschist outcrops and without active back-arc spreading (Fig. 9). These different environments may have been repeated in any segment of the Japanese orogenic belt throughout Phanerozoic time. Periods of oceanic island arc and marginal basin development (ophiolite formation) and tectonic erosion (blueschist metamorphism) might have alternated with periods of normal subduction, during which accretionary complexes were developed. The ophiolite-blueschist association is well documented in the Japan-Primorye area, e.g. the Oeyama ophiolite-Renge blueschist (Tsujimori & Itaya 1999) and Sergeevka ophiolite-Shaiginskiy blueschist (Kovalenko & Khanchuk 1991; Zakharov et al. 1992). In the NE Japan-Sakhalin belt, the Palaeozoic Miyamori ophiolite-Motai blueschist pair (Maekawa 1988; Ozawa 1988, 1994) and the Mesozoic Horokanai ophiolite-Kamuikotan blueschist pair (Ishizuka 1985, 1987; Sakakibara & Ota 1994) are well documented. Although a major blueschist belt is absent in the Koryak Mountains, many blueschist blocks occur in the Palaeozoic and Mesozoic accretionary complexes (Stavsky et al. 1990; Dobretsov 1999). It is likely that periods of accretion and nonaccretion, as represented by the present-day Nankai Trough and Mariana Trench, respectively, have been repeated many times in different segments of the Japan-NE Russia accretionary orogenic belts in the past. Periods of ophiolite-blueschist formation and tectonic erosion at subduction zones might have been followed by periods of massive accretion. Tectonic underplating of accreted sediments beneath the mantle wedge might have facilitated the uplift of overlying ophiolite-blueschist assemblages. This idea is compatible with the geochemical signatures of ophiolitic rocks showing SSZ affinities.

Conclusions The northwestern Pacific margin extending from Japan to Russia has many ophiolites of widely

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA varying ages, different petrological characteristics and distinctive tectonic histories. The following geological features suggest that these ophiolites probably formed in island arc environments in intra-oceanic settings: extremely diverse degree of melting in the residual mantle peridotite up to clinopyroxene disappearance and spinel Cr-number >0.70; the common occurrence of hydrous minerals and various metasomatic features in the mantle section; the common association with blueschist rocks; the presence of unusually thick oceanic crust. The modern Mariana and Tonga trenches, where ophiolitic rocks including highly depleted harzburgite and typical blueschist have been dredged from the sea floor, may be the modern analogues. The orogenic belts from Japan to NE Russia may have evolved through repeated stages of non-accretion, in which SSZ ophiolites and blueschists formed, and accretion, in which accretionary complexes mainly composed of clastic and volcaniclastic rocks developed. The association of highly depleted mantle harzburgite and orthopyroxene-type cumulate rocks is reinforced by reported occurrences of DH-type ophiolites from NE Russia (Shelting and Krasnaya). These ophiolites have only been reported so far from the western Pacific margins such as Hokkaido (Horokanai), Papua, and Tasmania (Adamsfield). The association of some DH-type ophiolites with boninitic volcanic rocks (Shelting, Papua, and possibly Mariana and Tonga) suggests that the depleted harzburgite is a residuum after boninitic melt production, although boninite is also reported from some ophiolites with less depleted peridotite (e.g. Robinson et al. 1983; Spadea & Scarrow 2000). Some primitive island arc tholeiite and magnesian andesite magmas could also coexist with the depleted harzburgite. The depleted harzburgite may form by either hightemperature dry melting of primary mantle or hydrous melting of previously depleted mantle. However, Os isotope studies of ophiolitic chromitites do not support much involvement of slabderived fluids in mantle melting. The Os isotope data are also inconsistent with an oceanic plateau (or LIP) origin postulated for some ophiolites with thick crustal sections. These instead may represent robust magmatic activity in SSZ environments, where they would be associated with highly depleted harzburgite massifs. We are grateful for T. Tsujimori, D. Saito, S. Miyashita, A. P. Stavsky, O. Morozov, S.A. Shcheka, A. I. Khanchuk, S. G. Byalobzhesky, W. B. Bryan and J. Hourigan for their help with fieldwork in Far East Russia. A.I. thanks S. A. Palandzhjan for valuable information on the ophiolites of the Koryak Mountains, and Y. Dilek for his encouragement in writing this paper. S. Maruyama is acknowledged for his help in arranging financial support

613

for our fieldwork. A.I. also acknowledges Grant-in-Aid for Scientific Research (C)-(2)-10640462 and -14540447 by the Ministry of Education, Japan. S. Arai is thanked for several discussions. Y. Furuhashi and D. Saito contributed the mineral analyses.

References AHMED, A.H. & ARAI, S. 2002. Unexpectedly high-PGE chromitite from the deeper mantle section of the northern Oman ophiolite and its tectonic implications. Contributions to Mineralogy and Petrology, 143, 263-278. ALABASTER, T., PEARCE, J.A. & MALPAS, J. 1982. The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex. Contributions to Mineralogy and Petrology, 81, 168-183. ARAI, S. 1980. Dunite-harzburgite-chromitite complexes as refractory residue in the Sangun-Yamaguchi zone, western Japan. Journal of Petrology, 21, 141-165. ARAI, S. ET AL. 1996. Petrology of the gabbrotroctolite-peridotite complex from Hess Deep, Equatorial Pacific: implications for mantle-melt interaction within the oceanic lithosphere. In: MEVEL, C. (ed.) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 135-155. BAZYLEV, B.A., PALANDZHJAN, S.A., GANELIN, A.V., SlLANTYEV, S.A., ISHIWATARI, A. & DMITRENKO,

G.G. 2001. Petrology of peridotites from the ophiolite melange in Cape Povorotny, Taigonos Peninsula, NE Russia: mantle processes beneath a subduction zone. Petrology, 9(2), 142-160. BEARD, J.S. 1986. Characteristic mineralogy of arcrelated cumulate gabbros: implications for the tectonic setting of gabbroic plutons and for andesite genesis. Geology, 14, 848-851. BELYI, V.F. & AKININ, V.V. 1985. Geologic Structure and Ophiolites in the Elistratova Peninsula, Volumes 1 and 2. Academy of Sciences of the USSR, Far East Science Centre, Magadan (in Russian). BLOOMER, S.H. & FISHER, R.L. 1987. Petrology and geochemistry of igneous rocks from the Tonga Trench—a non-accreting plate boundary. Journal of Geology, 95, 469-495. BLOOMER, S.H. & HAWKINS, J.W. 1983. Gabbroic and ultramafic rocks from the Mariana Trench: an island arc ophiolite. In: HAYES, D.E. (ed.) The Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Part 2. Geophysical Monograph, American Geophysical Union, 27, 294-317. BLOOMER, S.H. & HAWKINS, J.W. 1987. Petrology and geochemistry of boninite series volcanic rocks from the Mariana trench. Contributions to Mineralogy and Petrology, 97, 361-377. BOUDIER, F. & NICOLAS, A. 1985. Harzburgite and Iherzolite subtypes in ophiolitic and oceanic environments. Earth and Planetary Science Letters, 76, 84-92. BRYAN, W.B. 1991. Exploring Pacific seafloor ashore: Magadan Province, USSR. Oceanus (Woods Hole Oceanographic Institution), Summer 1991, 48-61.

614

A. ISHIWATARI ETAL

CANNAT, M., BIDEAU, D. & HEBERT, R. 1990. Plastic deformation and magmatic impregnation in serpentinized ultramafic rocks from the Garrett transform fault (East Pacific Rise). Earth and Planetary Science Letters, 101, 216-232. £INA, A., TASHKO, A. & TERSHANA, A. 1987. The Bulqiza and Gomsiqe ultrabasic massifs, ophiolites of Albanides: a geochemical comparison. Ofioliti, 12, 219-236. COFFIN, M.F. & ELDHOLM, O. 2001. Large igneous provinces: progenitors of some ophiolites? In: ERNST, R.E. & BUCHAN, K.L. (eds) Mantle Plumes: Their Identification through Time. Geological Society of America, Special Papers, 352, 59-70. COLEMAN, R.G. 1971. Plate tectonic emplacement of upper mantle peridotites along continental edges. Journal of Geophysical Research, 76, 1212-1222. COLEMAN, R.G. 1986. Ophiolites and accretion of the North American Cordillera. Bulletin de la Societe Geologique de France, 1986, 961-968. DEBARI, S.M. & COLEMAN, R.G. 1989. Examination of the deep levels of an island arc: evidence from the Tonsina ultramafic-mafic assemblage, Tonsina, Alaska. Journal of Geophysical Research, 94(B4), 4373-4391. DICK, H.J.B. 1989. Abyssal peridotites, very slow spreading ridges and ocean ridge magmatism. In: SAUNDERS, A.D. & NORRY, MJ. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 71-105. DICK, H.J.B. & BULLEN, T.B. 1984 Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DILEK, Y., MOORES, E.M. & FURNES, H. 1998. Structure of modern oceanic crust and ophiolites and implications for faulting and magmatism at oceanic spreading centers. In: BUCK, W.R., DELANEY, P.T., KARSON, J.A. & LAGABRIELLE, Y. (eds) Faulting and Magmatism at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union, 106, 219-265. DMITRENKO, G.G., MOCHALOV, A.G., PALANDZHJAN, S.A. & GORYACHEVA, E.M. 1985. Chemical Composition of the Rock-forming and Accessory Minerals of the Koryak Highland Alpine-type Ultramafics. Part 1. Rock-forming Minerals. Academy of Sciences of the USSR, Far East Science Centre, Interdisciplinary Research Institute, Magadan (in Russian). DMITRENKO, G.G., MOCHALOV, A.G. & PALANDZHJAN, S.A. 1990. Petrology and Platinum Occurrence in Lherzolite Massifs of the Koryak Mountains. Academy of Sciences of the USSR, Far East Science Centre, Interdisciplinary Research Institute, Magadan (in Russian). DOBRETSOV, N.L. 1999. Systematic periodicity of glaucophane metamorphism: an illusion or an important geologic regularity. Petrology, 7, 410-439. DOBRETSOV, N.L., WATANABE, T., NATAL'IN, B. & MIYASHITA, S. 1994. Comparison of ophiolites and blueschists of Sakhalin and Hokkaido. Ofioliti, 19, 157-176. EDWARDS, SJ. 1995. Boninitic and tholeiitic dykes in the Lewis Hills mantle section of the Bay of Islands

ophiolite: implications for magmatism adjacent to a fracture zone in a back-arc spreading environment. Canadian Journal of Earth Science, 32, 2128-2146. EDWARDS, SJ. & MALPAS, J. 1996. Melt-peridotite interactions in shallow mantle at the East Pacific Rise: evidence from OOP Site 895 (Hess Deep). Mineralogical Magazine, 60, 191-206. ENGLAND, R.N. & DA VIES, H.L. 1973. Mineralogy of ultramafic cumulates and tectonites in eastern Papua. Earth and Planetary Science Letters, 17, 416-425. FRANCIS, D. 1976. Corona-bearing pyroxene granulite xenoliths and the lower crust beneath Nunivak Island, Alaska. Canadian Mineralogist, 14, 291-298. FUJITA, K. & NEWBERRY, J.T. 1982. Tectonic evolution of northeastern Siberia and adjacent regions. Tectonophysics, 89, 337-357. FURUE, R., NIIDA, K. & KATOH, T. 1997. Origin of ophiolitic chromitites in the Kamuikotan Belt, Hokkaido. Bulletin of Hobetsu Museum, 13, 9-26 (in Japanese with English abstract). GAGGERO, L. & CORTESOGNO, L. 1997. Metamorphic evolution of oceanic gabbros: recrystallization from subsolidus to hydrothermal conditions in the MARK area (OOP Leg 153). Lithos, 40, 105-131. GREGOIRE, M., COTTIN, J.Y., GIRET, A., MATTIELLI, N. & WEIS, D. 1998. The meta-igneous granulite xenoliths from Kerguelen Archipelago: evidence of a continent nucleation in an oceanic setting. Contributions to Mineralogy and Petrology, 133, 259-283. HEBERT, R. & LAURENT, R. 1990. Mineral chemistry of the plutonic section of the Troodos ophiolite: new constraints for genesis of arc-related ophiolites. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crust Analogues. Proceedings of the Troodos '87 Symposium. Geological Survey Department, Nicosia, 149-163. HERZIG, C.T., KIMBROUGH, D.L. & HAYASAKA, Y. 1997. Early Permian zircon uranium-lead ages for plagiogranites in the Yakuno ophiolite, Asago district, Southwest Japan. Island Arc, 6, 396-403. IRWIN, W.P. 1981. Tectonic accretion of the Klamath Mountains. In: ERNST, W.G. (ed.) The Geotectonic Development of California (Rubey Volume 1). Prentice Hall, Englewood Cliffs, NJ, 29-49. ISHII, T., ROBINSON, P.T., MAEKAWA, H. & FISKE, R. ET AL. 1992. Petrological studies of peridotites from diapiric serpentinite seamounts in the IzuOgasawara-Mariana forearc. In: FRYER, P., PEARCE, J. & STOKKING, L.B. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 445-485. ISHIWATARI, A. 1985a. Igneous petrogenesis of the Yakuno ophiolite (Japan) in the context of the diversity of ophiolites. Contributions to Mineralogy and Petrology, 89, 155-167. ISHIWATARI, A. 1985b. Granulite-facies metacumulates of the Yakuno ophiolite, Japan: evidence for unusually thick oceanic crust. Journal of Petrology, 26, 1-30. ISHIWATARI, A. 1991. Time-space distribution and

OPHIOLITES IN JAPAN AND FAR EAST RUSSIA petrologic diversity of Japanese ophiolites. In: PETERS, TJ., NICOLAS, A. & COLEMAN, R.G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Kluwer, Dordrecht, 723-743. ISHIWATARI, A. 1994. Circum-Pacific Phanerozoic multiple ophiolite belts. In: ISHIW ATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Circum-Pacific Ophiolites (Proceedings of the 29th IGC, Kyoto, Part D). VSP, Utrecht, 7-28. ISHIW ATARI, A. & TSUJIMORI, T. 2003. Paleozoic ophiolites and blueschists in Japan and Russian Primorye in the tectonic framework of East Asia: a synthesis. Island Arc, 12, 190-206. ISHIWATARI, A., IKEDA, Y. & KOIDE, Y. 1990. The Yakuno ophiolite, Japan: fragments of Permian island arc and marginal basin crust with a hotspot. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crust Analogues. Proceedings of the Troodos '87 Symposium. Geological Survey Department, Nicosia, 497-506. ISHIZUKA, H. 1985. Prograde metamorphism of the Horokanai ophiolite in the Kamuikotan Zone, Hokkaido, Japan. Journal of Petrology, 26, 391-417. ISHIZUKA, H. 1987. Igneous and metamorphic petrology of the Horokanai ophiolite in the Kamuikotan zone, Hokkaido, Japan: a synthetic thesis. Memoirs of Faculty of Science, Kochi University, Series E, 8, 1-70. ISOZAKI, Y. 1997. Contrasting two types of orogen in Permo-Triassic Japan: accretionary versus collisional. Island Arc, 6, 2-24. JAQUES, A.L. & CHAPPELL, B.W. 1980. Petrology and trace element geochemistry of the Papuan ultramafic belt. Contributions to Mineralogy and Petrology, 75, 55-70. JAQUES, A.L. & GREEN, D.H. 1980. Anhydrous melting of peridotite at 0-15 kb pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology, 73, 287-310. KHANCHUK, A.I. & PANCHENKO, I.V. 1994. Mineral assemblages of the Gankuvayam ophiolitic section in the Kuyul ophiolitic terrane of the Russia Far East. Ofioliti, 19, 257-268. KOJIMA, S. 1989. Mesozoic terrane accretion in Northeast China, Sikhote-Alin and Japan regions. Palaeogeography, Palaeoclimatology, Palaeoecology, 69, 213-232. KOVALENKO, S.V. & KHANCHUK, A.I. 1991. First find of glaucophane schist in Sikhote Alin. Doklady Akademii Nauk SSSR, 318, 692-694 (in Russian). KUROKAWA, K. 1985. Petrology of the Oeyama ophiolitic complex in the Inner Zone of Southwest Japan. Science Reports, Niigata University, Series E, Geology and Mineralogy, 6, 37-113. LACHIZE, M., LORAND, J.P. & JUTEAU, T. 1996. Calcalkaline differentiation trend in the plutonic sequence of the Wadi Haymiliyah section, Haylayn massif, Semail ophiolite, Oman. Lithos, 38, 207-232. LOOCK, G., STOSCH, H.-G. & SECK, H.A. 1990. Granulite facies lower crustal xenoliths from the Eifel, West Germany: petrological and geochemical aspects.

615

Contributions to Mineralogy and Petrology, 105, 25-41. LOPEZ, J.C. & ISHIWATARI, A. 2002. Petrogenesis of the tholeiitic basalt, calc-alkaline basaltic andesite and high magnesian andesite lava succession of the Oligo-Miocene Anamizu Formation in northeastern Noto Peninsula, central Japan. Journal of Mineralogical and Petrological Sciences, 97, 85-113. LUCHITSKAYA, M.V. 1996. Plagiogranites of the Kuyul ophiolite massif (northeastern Kamchatka, Koryak Upland). Ofioliti, 21, 131-138. MAEKAWA, H. 1988. High PIT metamorphic rocks in Northeast Japan: Motai-Matsugadaira zone. Earth Science (Chikyu Kagaku), 42, 212-219 (in Japanese with English abstract). MAEKAWA, H., SHOZUI, M., ISHII, T., FRYER, P. & PEARCE, J.A. 1993. Blueschist metamorphism in an active subduction zone. Nature, 364, 520-523. MAEKAWA, H., FRYER, P. & OZAKI, A. 1995. Incipient blueschist-facies metamorphism in the active subduction zone beneath the Mariana forearc. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the Western Pacific. Geophysical Monograph, American Geophysical Union, 88,281-289. MATSUMOTO, I. & ARAI, S. 1997. Characterization of chromian spinel as a tool of petrological exploration for podiform chromitite. Resource Geology, 47, 189-199. MELCHER, F., GRUM, W., SIMON, G., THALHAMMER, T.V. & STUMPEL, E.F. 1997. Petrogenesis of the ophiolitic giant chromite deposits of Kempirsai, Kazakhstan: a study of solid and fluid inclusions in chromite. Journal of Petrology, 38, 1419-1458. MIYASHIRO, A. 1973. The Troodos ophiolitic complex was probably formed in an island arc. Earth and Planetary Science Letters, 19, 218-224. MIYASHITA, S. 1983. Reconstruction of the ophiolite succession in the western zone of the Hidaka Metamorphic Belt, Hokkaido. Journal of the Geological Society of Japan, 89, 69-86. MIYASHITA, S. & HASHIMOTO, S. 1975. The layered basic complex of Mt. Poroshiri, Hokkaido, Japan. Journal of Faculty of Science, Hokkaido University, Series IV, 16, 421-452. MIYASHITA, S. & YOSHIDA, A. 1988. Pre-Cretaceous and Cretaceous ophiolites in Hokkaido, Japan. Bulletin de la Societe Geologique de France, 1988, 251-260. MOORES, E.M. 1969. Petrology and Structure of the Vourinos Ophiolitic Complex of Northern Greece. Geological Society of America, Special Papers, 118. 74 p. MOORES, E.M. & VINE, F.J. 1971. The Troodos massif, Cyprus and other ophiolites as oceanic crust: evaluation and implications. Philosophical Transactions of the Royal Society of London, Series A, 268, 443-466. MORISHITA, T. & ARAI, S. 2001. Petrogenesis of corundum-bearing mafic rock in the Horoman Peridotite complex, Japan. Journal of Petrology, 42, 1279-1299. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht.

616

A. ISHIWATARI ETAL

NICOLAS, A. & JACKSON, E.D. 1972. Repartition en deux provinces des peridotites des chaines alpines longeant la Mediterranee: implications getectoniques. Schweizerische Mineralogische und Petrographische Mitteilungen, 52, 479-495. NISHIMURA, Y. 1998. Geotectonic subdivision and areal extent of the Sangun belt, Inner Zone of Southwest Japan. Journal of Metamorphic Geology, 16, 129-140. OHARA, Y., KASUGA, S. & ISHII, T. 1996. Peridotites from the Parece Vela rift in the Philippine Sea: upper mantle material exposed in an extinct backarc basin. Proceedings of Japan Academy, Series B, 72, 118-123. OHARA, Y., STERN, J.R., ISHII, T., YURIMOTO, H. & YAMAZAKI, T. 2002. Peridotites from the Mariana Trough: first look at the mantle beneath an active back-arc basin. Contributions to Mineralogy and Petrology, 143, 1-18. OZAWA, K. 1988. Ultramafic tectonite of the Miyamori ophiolite complex in the Kitakami Mountains, Northeast Japan: hydrous upper mantle in an island arc. Contributions to Mineralogy and Petrology, 90, 159-175. OZAWA, K. 1994. Melting and melt segregation in the mantle wedge above a subduction zone: evidence from the chromite-bearing peridotites of the Miyamori ophiolite complex, northeastern Japan. Journal of Petrology, 35, 647-678. PALANDZHJAN, S.A. 1986. Ophiolite belts in the Koryak upland, northeast Asia. Tectonophysics, 127, 341-360. PEARCE, J.A., LIPPARD, S.J. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 79-95. PUSHCHAROVSKY, YU.M., RUZHENTSEV, S.V. & SOKO-

LOV, S.D. 1988. Tectonic thrust sheets and geologic mapping. Geotectonics, 22, 1-8 (original Russian paper published in 1969). ROBINSON, P.T., MELSON, W.G., O'HEARN, T. & SCHMINCKE, H.-U. 1983. Volcanic glass compositions of the Troodos ophiolite, Cyprus. Geology, 11, 400-404. SAITO, D., ISHIWATARI, A., TSUJIMORI, T., MIYASHITA, S. & SOKOLOV, S.D. 1999. Elistratova ophiolite in Taigonos peninsula, Far East Russia: an island arc ophiolite intruding into oceanic mantle. Memoirs of the Geological Society of Japan, 52, 303-316 (in Japanese with English abstract). SAKAKIBARA, M. & OTA, T. 1994. Metamorphic evolution of the Kamuikotan high-pressure and lowtemperature metamorphic rocks in central Hokkaido, Japan. Journal of Geophysical Research, 99, 22221-22235. SHCHEKA, S.A., ISHIWATARI, A. & VRZHOSEK, A.A. 2001. Geology and petrology of Cambrian Khanka ophiolite in Primorye (Far East Russia) with notes on its manganese-rich chromian spinel. Earth Science (Chikyu Kagaku), 55, 265-274. SHIBATA, K., IGI, S. & UCHIUMI, S. 1977. K-Ar ages of hornblendes from gabbroic rocks in Southwest

Japan. Geochemical Journal, 11, 57-64. SOBOLEV, A.V. & DANYUSHEVSKY, L.V. 1994. Petrology and geochemistry of boninites from the north termination of the Tonga Trench: constraints on the generation conditions of primary high Ca-boninite magmas. Journal of Petrology, 35, 1183-1211. SOKOLOV, S.D. 1992. Accretion Tectonics in the Koryak- Chukotka Segment of the Pacific Belt. Nauka, Moscow (in Russian). SPADEA, P. & SCARROW, J.H. 2000. Early Devonian boninites from the Magnitogorsk arc, southern Urals (Russia): implications for early development of a collisional orogen. In: DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 461-472. STAVSKY, A.P., CHEKHOVICH, V.D., KONONOV, M.V. & ZONENSHAIN, L.P. 1990. Plate tectonic and palinspastic reconstructions of the Anadyr-Koryak region, Northeast Russia. Tectonics, 9, 81-101. STEINMANN, G. 1927. Die ophiolitischen Zonen in den Mediterranen Kettengebirgen. In: JACOB, C.H. (ed.) Geologic de la Mediterranee. Proceedings of the 14th International Geological Congress, Madrid. Graficas Reunidas, Madrid, 2, 638-667. SUHR, G. & EDWARDS, S.J. 2000. Contrasting mantle sequences exposed in the Lewis Hills massif: evidence for the early, arc-related history of the Bay of Islands ophiolite. In: DILEK, Y. ET AL. (eds) Ophiolites and Oceanic Crust: New Insights From Field Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 433-442. TAIRA, A. 1985. Sedimentary evolution of Shikoku subduction zone: the Shimanto belt and Nankai trough. In: NASU, N. ET AL. (eds) Formation of Active Ocean Margins. Terra, Tokyo, 835-851. TAKAHASHI, N. 1991. Evolutional history of the uppermost mantle of an arc system: petrology of the Horoman peridotite massif, Japan. In: PETERS, TJ, NICOLAS, A. & COLEMAN, R.G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Kluwer, Dordrecht, 195-205. TAMURA, A., MAKITA, M. & ARAI, S. 1999. Petrogenesis of ultramafic rocks in the Kamuikotan belt, Hokkaido, northern Japan. Memoirs of the Geological Society of Japan, 52, 53-68 (in Japanese with English abstract). TAZAWA, J. 1988. Paleozoic-Mesozoic stratigraphy and tectonics of the Kitakami Mountains, northeast Japan. Earth Science (Chikyu Kagaku), 42, 165178 (in Japanese with English abstract). TILMAN, S.M. & BOGDANOV, N.A. 1992. Tectonic Map of Northeast Asia, 1:5,000,000, with Explanatory Notes. Institute of the Lithosphere, Moscow, 000. TSUJIMORI, T. & ISHIWATARI, A. 2002. Granulite facies relics in the Early Paleozoic kyanite-bearing ultrabasic metacumulate in the Oeyama belt, the Inner Zone of southwestern Japan. Gondwana Research, 5, 823-835. TSUJIMORI, T. & ITAYA, T. 1999. Blueschist-facies metamorphism during Paleozoic orogeny in southwestern Japan: phengite K-Ar ages of blueschist-

O P H I O L I T E S IN JAPAN AND FAR EAST RUSSIA facies tectonic blocks in a serpentinite melange beneath early Paleozoic Oeyama ophiolite. Island Arc, 8, 190-205. TSUJIMORI, T., SAITO, D., ISHIWATARI, A., MIYASHITA, S. & SOKOLOV, S.D. 1999. Electron microprobe element image of zoned chromian spinel with hydrous mineral inclusions in a chromitite from Elistratova ophiolite, Far East Russia. Science Reports ofKanazawa University, 43, 1-11. TSUJIMORI, T., ISHIWATARI, A. & BANNO, S. 2000. Eclogitic glaucophane schist from the Yunotani valley in Omi Town, the Renge metamorphic belt, the Inner Zone of southwestern Japan. Journal of the Geological Society of Japan, 106, 353-362. UMINO, S., YANAI, S., JAMAN, A.R., NAKAMURA, Y. & IIYAMA, J.T. 1990. The transition from spreading to subduction: evidence from the Semail ophiolite, Northern Oman Mountains. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Proceedings of Troodos '87 Symposium. Geological Survey Department, Nicosia, 375-384. VARNE, R. & BROWN, A.V. 1978. The geology and petrology of the Adamsfield ultramafic complex,

617

Tasmania. Contributions to Mineralogy and Petrology, 67, 195-207. VYSOTSKIY, S.V. 1994. High and low pressure cumulates of Paleozoic ophiolites in Primorye, eastern Russia. In: ISHIWATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Circum-Pacific Ophiolites. Proceedings of the 29th IGC, Kyoto, Vol. D. VSP, Utrecht, 145-162. VYSOTSKIY, S.V., GOVOROV, G.L, KEMKIN, I.V. & SAPIN, V.I. 2000. Boninitic ophiolite assemblage on eastern Sakhalin Island: geology and petrogenesis. Geology of Pacific Ocean, 15, 1067-1090. WALKER, R.J., PRICHARD, H.M., ISHIWATARI, A. & PIMENTEL, M. 2002. The osmium isotopic composition of convecting upper mantle deduced from ophiolite chromites. Geochimica et Cosmochimica Acta, 66, 329-345. YAMANE, M., BAMBA, M. & BAMBA, T. 1988. The first finding of orbicular chromite ore in Japan. Mining Geology, 38, 501-508. ZAKHAROV, Yu.D., PANCHENKO, I.V. & KHANCHUK, A.I. 1992. A Field Guide to the Late Paleozoic and Early Mesozoic Circum-Pacific Bio- and Geologic Events. Russian Academy of Sciences, Far East Geological Institute, Vladivostok.

This page intentionally left blank

Ophiolites in accretionary complexes along the Early Cretaceous margin of NE Asia: age, composition, and geodynamic diversity S. D. SOKOLOV 1 , M. V. LUCHITSKAYA 1 , S. A. SILANTYEV 2 , O. L. MOROZOV 1 , A. V. G A N E L I N 1 , B. A. BAZYLEV 2 , A. B. OSIPENKO 3 , S. A. P A L A N D Z H Y A N 4 & I. R. K R A V C H E N K O - B E R E Z H N O Y 1 1

Geological Institute of the Russian Academy of Sciences, 7 Pyzhevsky per., Moscow, 109017, Russia (e-mail: [email protected]) 2

Vernadsky Institute of Geochemistry and Analytical Chemistry, 19 Kosygina St., 119991, Moscow, Russia ^Vernadsky State Geological Museum, 11/2 Mokhovaya St., Moscow, 103009, Russia

A

Institute of the Lithosphere of Marginal and Inland Seas, Russian Academy of Sciences, 22 Staromonetny per., Moscow, 109180, Russia

Abstract: The existing published data, combined with our own new field, petrographic, and geochemical observations and data show that ophiolites of the West Koryak fold system originated in a variety of tectonic environments. This fold system stretches along the boundary shared by two of NE Asia's largest tectonic units, the Verkhoyansk-Chukotka and KoryakKamchatka foldbelts. The fold system abounds in Palaeozoic and Mesozoic ophiolites and sedimentary and volcanic island-arc assemblages. The ophiolites are Palaeozoic and Mesozoic in age. The variety of geological and geochemical signatures implies ophiolite origin in diverse tectonic settings. The Early Palaeozoic ophiolites of the Ganychalan accreted terrane and Devonian(?) ophiolites of the Ust-Belaya accreted terrane are fragments of the Panthalassan oceanic lithosphere. Serpentinite melange in the Ust-Belaya terrane contains some blocks of island-arc provenance. They are probably Late Palaeozoic-Early Mesozoic in age as determined by K-Ar measurements, which require validation by other techniques. Mesozoic, chiefly Late Jurassic-Early Cretaceous ophiolites of the Beregovoi and Kuyul accreted terranes, originated in a suprasubduction-zone (SSZ) setting (ensimatic island arc and back-arc basin). Among the Mesozoic ophiolites, one finds blocks of oceanic assemblages in serpentinite melanges as well. Basalt and chert blocks of clearly oceanic derivation are viewed as detached fragments of the upper part of the oceanic lithosphere. The ophiolites have experienced a variety of accretionary scenarios. Palaeozoic ophiolites docked onto the Koni-Taigonos island arc (of Late Palaeozoic-Early Mesozoic age), probably in the Late Palaeozoic or Early Mesozoic, whereas Mesozoic ophiolites accreted onto the Uda-Murgal island arc (of Late Jurassic-Early Cretaceous age) in the terminal Early Cretaceous. Sedimentary deposits, whose base is late Albian in age, make a post-accretionary sequence. These island arcs portray the overall history of the convergent boundary between the North Asian continent and NW Pacific. Ophiolites of the Ganychalan and Ust-Belaya terranes consist of thrust sheets and, jointly with Yelistratov Peninsula ophiolites, make up the basement to the forearc of the Uda-Murgal island arc, ophiolites of Cape Povorotny and Kuyul terrane being incorporated in accretionary prisms of the same arc. Ophiolites and associated metamorphic, volcanic, and sedimentary rocks of Palaeozoic-Early Cretaceous age underwent three deformation phases, each reflecting a different stage in the evolution of the NE Asian continental margin and readily correlative with principal tectonic events in the northern Circum-Pacific region.

In NE Asia, ophiolites have been reported from the Verkhoyansk-Chukotka and Koryak-Kamchatka foldbelts (Fig. 1). The ophiolites span a broad Early Palaeozoic to Mesozoic age interval, The Verkhoyansk-Chukotka belt, except for the Kolyma loop (Fig. 1), displays a structural grain dominated by northwesterly trends resulting from

collisional processes (Pushcharovsky et al. 1992; Bogdanov & Tilman 1992; Parfenov et al. 1993). Accreted terranes found in the belt represent fragments of microcontinents, such as Chukotka, Omolon, Okhotsk, etc. Ophiolites occur sporadically within collisional piles of the Chersky Range and South Anyui suture, where they make up

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 619-664. 0305-8719/037$ 15 © The Geological Society of London 2003.

620

S. D. SOKOLOV ETAL.

Fig. 1. Tectonic map of NE Asia (by S. D. Sokolov and G. Ye. Bondarenko). Black areas are ophiolites: 1.1, Cape Povorotny; 1.2, Yelistratov Peninsula; 2.1, Kuyul terrane; 2.2, Ganychalan terrane; 3.1, Ust-Belaya terrane.

small, undeformed slices associated with greenschist- and amphibolite-facies metamorphic rocks (Parfenov et al. 1993; Nokleberg et al 1994; Oxmanetal. 1995). The Koryak-Kamchatka foldbelt is located east of the Okhotsk-Chukotka Volcanic Belt and stretches north-south to NE-SW (Fig. 1). It is a typical example of an accretionary continental margin formed through successive docking onto the Asian continent of outboard terranes having a variety of ages and geodynamic settings and arriving from the Pacific (Zonenshain et al. 1990; Bogdanov & Til'man 1992; Pushcharovsky et al.

1992; Sokolov 1992; Parfenov et al. 1993; Nokleberg et al. 1994). The terranes are of the following types: island arc, ophiolite, back-arc and turbidite basins, oceanic crust, and accretionary prism. Ophiolites are widespread, making up major bodies and entire terranes (Markov et al. 1982; Peyve 1984; Palandzhyan 1992; Sokolov 1992; Nokleberg et al. 1994). No consensus exists regarding the age, composition, or provenance of NE Asian ophiolites. Some workers (e.g. Fujita & Newberry 1982; Parfenov 1984; Zonenshain et al. 1990) view the ophiolites as fragments of a large oceanic basin, a

OPHIOLITES OF NORTHEAST ASIA former constituent of the Palace-Pacific. Others believe them to be relics of minor oceanic basins or rifts (Lychagin et al. 1991; Bogdanov & Til'man 1992; Oxman et al. 1995), and still others suggest that both Palace-Pacific and back-arc basin fragments come into play (Peyve 1984; Sokolov 1992; Nokleberg et al 2001). This controversy is due primarily to poor knowledge of the ophiolite assemblages; hence the critical importance of the new structural and compositional data presented here. Dating the ophiolites is crucial to unravelling their history. According to earlier workers (e.g. Coleman 1984; Ishiwatari 1994; Dilek et al. 1999; Searle & Cox 1999; Shervais 2001), ophiolites originate from a variety of tectonic environments, that obviates any palaeotectonic reconstructions or formative scenarios for the NE Asian margins without first identifying geodynamic affinities of the ophiolites. In addition, there are virtually no English-language publications on NE Asian ophiolites; hence, one more objective of this paper is to fill in this informational gap. The ophiolites discussed in this paper occur in the West Koryak fold system, which is located at the junction of the Verkhoyansk-Chukotka and Koryak-Kamchatka foldbelts (Fig. 1). In recent years, we have acquired new data on field relationships and evolution of this major tectonic unit, which incorporates accreted ophiolitic assemblages of various ages (Sokolov et al. 1999;

621

Silantyev et al. 2000). The paper focuses on new data and issues related to the tectonic setting, inner structure, composition, and age of the ophiolites.

Geological framework In NE Asia, four major tectonic units (Fig. 1) with distinctive structural grains and geological histories are recognized: (1) the Siberian craton; (2) the Verkhoyansk-Chukotka foldbelt; (3) the Koryak-Kamchatka foldbelt; (4) the OkhotskChukotka continental-margin volcanic belt (OCVB) (Fig. 1). These units represent continuous, albeit discrete, accretion onto the Siberian craton of geodynamically diverse terranes and microcontinents (Fig. 2). The West Koryak fold system lies along the boundary between the Koryak-Kamchatka and Verkhoyansk-Chukotka foldbelts. Most of the West Koryak fold system is overlain by OCVB volcanic and sedimentary rocks, and the latter also unconformably overlap the Verkhoyansk-Chukotka foldbelt. The OCVB is a Late Cretaceous Andean-type continental-margin volcanic belt. It was initiated after a major mid-Cretaceous (Aptian-Albian) phase of accretion (Fig. 2) onto the Asian continent (Sokolov 1992). The West Koryak fold system incorporates numerous island-arc assemblages and ophiolites that were brought together at the end of the Early

Fig. 2. Reconstruction showing continental growth of NE Asia (after S. D. Sokolov 1992).

622

S. D. SOKOLOV ETAL.

Cretaceous (Markov et al 1982; Parfenov 1984; Zonenshain et al 1990; Sokolov 1992; Parfenov etal 1993). The island-arc volcanic and sedimentary assemblages are of calc-alkaline affinity and range in age from Carboniferous to Early Cretaceous. Parfenov (1984) attributed these rocks to a continuous Late Palaeozoic to Mesozoic KoniMurgal island arc. However, Filatova (1988) viewed the Late Jurassic to Early Cretaceous volcanic and sedimentary sequences as part of the Uda-Murgal island arc. Zonenshain et al. (1990) identified the Koni-Murgal volcanic belt as a separate system, which they interpreted as an agglomeration of island-arc assemblages of various ages that were joined together in Mid-Cretaceous time. According to those workers, the original position of these assemblages is unknown, although they are believed to have been formed a considerable distance away from Siberia's continental margin (Zonenshain et al. 1990). These considerations drew on the pioneering palaeomagnetic data pinpointing the Omolon terrane in the Late Palaeozoic and Early Mesozoic at more southerly latitudes. These data, however, were at odds with palaeobiogeographical conclusions on the boreal faunas and Angara floras (Shapiro & Ganelin 1988). Sokolov (1992) postulated two convergent boundaries of contrasting ages in the region; one of Late Palaeozoic to Early Mesozoic age, during which the Koni-Taigonos island arc existed, and the other of Late Jurassic to Early Cretaceous age, composed of the Uda-Murgal island-arc system (Fig. 1). The volcanic and sedimentary assemblages of the Koni-Taigonos island arc are best exposed and most thoroughly studied in the Koni-Pyagina and Taigonos peninsulas (Nekrasov 1976; Zaborovskaya 1978). These areas provide a stage for reconstructing, in Permian to Mid-Jurassic times, the volcanic arc proper and the North Taigonos back-arc basin. These assemblages are also present in the Penzhina District (Khudoley & Sokolov 1998), in the Pekulnei Range, and in Chukotka (Morozov 2001). In the Penzhina segment, Carboniferous island-arc assemblages are exposed in the Kharitonya terrane and in thrust sheets within the Upupkin terrane; in this locality, they consist of coarse andesitic pyroclastic rocks and tuffaceous epiclastic rocks of Permian and Triassic ages (Khudoley & Sokolov 1998; Sokolov et al. 1999). In the Pekulnei Range and in Chukotka, the Late Palaeozoic to Early Cretaceous island-arc sequence includes metavolcanic and metasedimentary rocks, layered gabbros, and Early Mesozoic granitic rocks (Morozov 2001). Unfortunately, numerous aspects of the KoniTaigonos island arc are still unclear. These include

both the arc's polarity and basement composition, as well as the origin of its various segments. Faunal and floral data point to rock formation at high latitudes (Shapiro & Ganelin 1988; Sokolov 1992), which, in combination with structural data (Sokolov et al. 1999) and spatial position, suggest that the arc originated along a convergent boundary between the Asian continent and the NW Pacific. Various outboard terranes that arrived from the Palaeo-Pacific were accreted onto the arc (Zonenshain et al. 1990; Parfenov et al. 1993). These terranes are best exposed in the Penzhina segment, where they include the Ganychalan composite terrane and metamorphic rocks of the Upupkin terrane. Fragments of these assemblages have also been reported from the pre-arc basement of the Taigonos segment of the Uda-Murgal arc and as Ordovician deposits and ophiolites (Fig. 1; see also Fig. 15, below). Upper Jurassic to Lower Cretaceous volcanic and sedimentary rocks of the Uda-Murgal islandarc system are traceable for about 3000 km, from the Mongolia-Okhotsk foldbelt in the south along the Sea of Okhotsk coastline (via the Koni, Pyagina, and Taigonos peninsulas) as far northeastward as the Chukchi Peninsula (Fig. 1). The volcanic and sedimentary lithologies and the character and age of basement vary from place to place along the arc. In the southern segment, the only identifiable features are the volcanic portion of the island-arc system and some constituents of its associated back-arc basin. Island-arc assemblages rest on heterogeneous basement that includes fragments of the Asian continent (Siberian craton, Verkhoyansk complex, Okhotsk microcontinent), and the Koni-Taigonos Late Palaeozoic to Early Mesozoic island arc. Hence, the Late Jurassic to Early Cretaceous convergent boundary was located at an angle to the pre-existing structural grain. Throughout the study area, volcanic arc assemblages were located along the continental margin, strongly suggesting the existence of a continental-margin belt. The Taigonos and Penzhina segments provide evidence for reconstructing a lateral succession: volcanic arc-forearc-accretionary prism-trenchoceanic plate (Fig. 3). Basement to the island arc was provided by the pre-existing Koni-Taigonos arc with its accreted terranes, including the Early Palaeozoic ophiolites of the Ganychalan terrane. Within these segments, island-arc deposits were also formed in a continental margin setting. Further NE, however, the back-arc region was the locus of marine deposition, and the continentalmargin belt gave way to an ensialic arc (Sokolov etal 1999). In the Pekulnei segment, island-arc assemblages rest on heterogeneous basement that incorporates

OPHIOLITES OF NORTHEAST ASIA

623

Fig. 3. Taigonos segment of the Uda-Murgal island arc (Late Jurassic-Early Cretaceous).

fragments of both lower continental crust and oceanic lithosphere (Sokolov et al. 1999; Morozov 2001). A back-arc basin floored with oceanic crust was situated behind the arc and was probably linked to the Anyui palaeo-ocean. The northeastern Chukotka branch (Fig. 1) of the convergent boundary had heterogeneous basement that incorporated, among other things, ancient sialic crust. Consumption of oceanic crust along the Chukotka branch was not extensive, probably because of the strike-slip nature of plate interaction within this segment (Morozov 2001). Ophiolites of the West Koryak fold system (Fig. 1) occur either in forearc basement (type 1) or within accretionary prisms of the Uda-Murgal island arc (Fig. 1) (type 2). Type 1 ophiolites were accreted in the Late Palaeozoic-Early Mesozoic (Parfenov 1984; Sokolov et al. 1999) onto the Koni-Taigonos island arc. Type 2 ophiolites were accreted in Late Jurassic and, mainly, in Early Cretaceous times onto the frontal part of the UdaMurgal island-arc system (Parfenov 1984; Khanchuk et al 1990; Sokolov et al 1999). Ophiolites have been reported from the Taigonos, Penzhina, and Ust-Belaya segments of the island arc. Ophiolites are well exposed along the SE coast of the Taigonos Peninsula (Fig. 1). The largest ophiolitic outcrops occur within the accretionary pile exposed at Cape Povorotny and in pre-arc basement on the Yelistratov Peninsula (Belyi & Akinin 1985; Ishiwatari et al. 1998). In the Penzhina segment, two large ophiolitic terranes, the Ganychalan and Kuyul terranes, are well documented (Markov et al. 1982; Palandzhyan 1992; Ganelin & Peyve 2001; Nekrasov et al. 2001).

Analytical techniques Major and trace element analyses were carried out in various laboratories using a range of methods. Mineral compositions from Cape Povorotny rocks were measured in polished sections on an automated CAMEBAX-Microbeam four-channel

wavelength-dispersive electron probe at the Vernadsky Institute (GEOKHI). Whole-rock major element analyses from peridotites and gabbros were performed by X-ray fluorescence (XRF) on a Philips PW-1600 XRF automated multichannel spectrometer, and REE contents were determined by instrumental neutron activation analysis (INAA) at GEOKHI. All analytical investigations of the volcanic rocks were carried out at the Analytical Centre of the Geological Institute, Russian Academy of Sciences (GIN RAS) by INAA and inductively coupled plasma mass spectrometry (ICP-MS). Mineral chemistry of Yelistratov Peninsula ophiolitic peridotites and Ganychalan terrane plutonic rocks was analysed by N. N. Kononkova on a CAMECA CAMEBAX electron microprobe at GEOKHI at an accelerating voltage of 15 kV and beam current of 35 nA. Natural and synthetic minerals were used as standards. Rock chemistry was analysed on a PLASMA QUAD PQZ+Turbo (VG Instruments) mass spectrometer at the Institute of Mine Geology, Petrography, Mineralogy, and Geochemistry, Moscow. Routine sample preparation included dissolution in concentrated HF + HC1O4 mixture, followed by precipitation using HNOs. Analytical reproducibility was controlled using certified F, W, rare earth element (REE) + 25 ppb standard solutions and AGV-1 standard. Major, trace, and REE analyses on plagiogranites from blocks in the Main Melange unit of Cape Povorotny, Ganychalan terrane, and Kuyul terrane ophiolites, as well as ultramafic and mafic rocks of Ganychalan ophiolites were performed at the GIN RAS Analytical Centre. Major elements were measured by wet chemistry, and trace elements by XRF on a Russian-made ARF-6 quantographer in the concentration range 0.0001% to n%. REE were analysed by INAA in the range 0.000001% to n%. Ion microprobe measurements on minerals from Ust-Belaya peridotites were performed at the Northeastern Interdisciplinary Research Institute,

624

S. D. SOKOLOV ETAL.

Far East Division of the Russian Academy of Sciences (SVKNII DVO RAN), Magadan, on a CAMEBAX instrument (analysts E. M. Goryacheva and G.A. Merkulov). Major elements in ultramafic and mafic rocks were analysed at the X-ray Spectral Analysis Laboratory (SVKNII) and by gravimetric analysis at the Central Laboratory, Geological Survey, Magadan. Whole-rock K-Ar measurements were carried out by A. D. Lyuskin, at the Laboratory of Isotope Geochronology, SVKNII. REE in clinopyroxenes from the Ganychalan terrane plutonic rocks were measured on a Cameca IMS 4f ion microprobe, at the Institute of Microelectronics (IMAN), Russian Academy of Sciences, Yaroslavl. Major elements and V, Cr, Co, Ni, Cu, and Ba from Kuyul terrane peridotites were determined by wavelength-dispersive XRF using routine techniques at the Karpinsky Geological Institute (St. Petersburg, Russia). Other trace elements, including REE, were analysed by ICP-MS at the Institute of Geochemistry (Irkutsk, Russia). Mounts of 0.1 ± 0.001 g were digested with HF and HNO3 mixture in Teflon bombs for 24 h, evaporated until dry, taken up in HNO3, and once again evaporated until dry. Further HC1 was added and the product again evaporated until dry to assure quantitative removal of HF and chlorides. The samples were redissolved with deionized water. No undissolved spinels were detected in the Teflon bombs, and the fact that Cr values were comparable with those obtained by XRF suggests that all of the spinels went into solution. The samples were run on a VG Elemental Plasmaquad with long peak dwell times (320 ms per mass unit). Calibration was carried out using a set of high-Mg laboratory standards. Estimated precision is less than 10% for all of the determined elements. Microprobe analyses of minerals from Kuyul terrane peridotites were carried out at the Institute of Volcanology (Petropavlovsk-Kamchatsky, Russia) using a CAMECA CAMEBAX system equipped with a KEVEX energy-dispersive spectrometer with an accelerating voltage of 15 kV and a sample current of 15nA (counting time 100 s). Precision is estimated to be better than about 2% for all main components. Radiometric ages are taken from a number of publications, where their interpretation is provided as well.

Ophiolites in the Cape Povorotny accretionary complex Geological setting Five tectonic units are recognized in the Taigonos Peninsula (Fig. 4): (1) the Avekov terrane, which

is composed of Precambrian and Lower Palaeozoic metamorphic sequences; (2) the Pylgin suture zone, which incorporates metamorphosed Mesozoic volcanic and sedimentary rocks; (3) the Central Taigonos terrane, made up of Upper Permian-Lower Cretaceous island-arc strata; (4) the East Taigonos granite-metamorphic belt; (5) the Beregovoi terrane, which is composed of prearc complexes and the accretionary prism of the Uda-Murgal volcanic arc (Sokolov et al. 1999; Silantyev et al. 2000). A broad spectrum of igneous and metamorphic rocks are hosted by serpentinite melange in the Cape Povorotny accretionary complex and make up the following succession of tectonic units, from south to north (Fig. 5): (1) the Povorotny serpentinite melange with blocks of sheeted dykes, ultramafic rocks, and gabbro; (2) the Median serpentinite melange, with small blocks and fragments of ultramafic rock, gabbro, volcanic and terrigenous rocks, and chert; (3) the Main Melange unit, a serpentinite melange with blocks of peridotite, garnet-free and garnet-bearing amphibolites, greenschists, island-arc volcanic and sedimentary rocks, oceanic basalts and chert, and gabbro-diabase with plagiogranite veins (Fig. 6).

Petrography and geochemistry of metamorphic and igneous rocks Amphibolites occur as disrupted blocks in serpentinite melanges exposed on Cape Povorotny (Fig. 5). They consist of: (1) massive melanocratic rocks composed almost wholly of hornblende and minor plagioclase or garnet-hornblende rocks; (2) albite-hornblende schists. Judging by the characteristic mineral assemblages and mineral and bulk-rock compositions, the protoliths were made dominantly of plutonic rocks and subordinate volcanic rocks (Silantyev et al. 2000). Geochemical signatures of high-grade amphibolites have been detailed by Silantyev et al. (2000), indicating that volcanic protoliths of these rocks ranged in affinity from within-plate basalt (WPB) or enriched mid-ocean ridge basalt (E-MORB) to normal MORB (N-MORB). Mafic plutonic rocks are widespread as tectonic blocks in the Povorotny and Median melanges. Gabbro s typically occur as isolated boudins and small tectonic slices in the serpentinite matrix. Plutonic rock compositions from the Cape Povorotny ophiolite melange have a wide range, implying the existence of different gabbroic series of several geochemical types. Silantyev et al. (2000) presented chemical data indicating that gabbros chemically similar to boninite plutonic suites are abundant in the Cape Povorotny serpentinite mel-

OPHIOLITES OF NORTHEAST ASIA

625

Fig. 4. Tectonic map of the Taigonos Peninsula (after Sokolov et al. 1999).

anges. These boninite-like gabbros are low in TiC>2, their REE patterns being typical of boninites and their plutonic equivalents (low total REE contents, concave chondrite-normalized REE patterns, and considerable light REE (LREE) variations). The

Cape Povorotny boninite gabbros are associated with N-MORB and within-plate or E-MORB gabbros and dolerites. This group of plutonic rocks, including hornblende-bearing gabbro, is moderate to relatively high in TiC>2, FeO, and PiOs at

626

S. D. SOKOLOV ETAL.

Fig. 5. Map showing tectonic units of the Cape Povorotny accretionary complex (after Sokolov et al. 1999).

moderately high REE totals (analytical data have been presented by Silantyev et al. (2000)). Felsic veins in gabbro-diabase from blocks in the Main Melange unit (Fig. 6) are composed of plagiogranite and, sporadically, tonalite. The plagiogranites have magmatic textures with euhedral plagioclase crystals partly intergrown with quartz albite granophyre, suggestive of crystallization at shallow depths. The plagiogranites are composed of quartz, saussuritized plagioclase, epidote, chlorite, and magnetite. The tonalites differ from the plagiogranites in having smaller amounts of

quartz, and in that they contain light green amphibole and andesine plagioclase is andesine. The plagiogranites and tonalites have low A1203 (11.12-13.7%), K20 (0.03-0.06%) and K/ Rb ratios (0.01-0.03), and relatively high Y (3744ppm) (Table 1). Rb (7ppm) and Zr (120, 160ppm) contents of the plagiogranites (Table 1) are similar to those of the average Mid-Atlantic Ridge plagiogranites at latitude 2-3°N (Rikhter 1997). Chondrite-normalized REE patterns of the plagiogranites are nearly flat to slightly LREE en-

OPHIOLITES OF NORTHEAST ASIA

Fig. 6. Network of plagiogranite veins in gabbrodiabase from blocks in the Main Melange zone.

riched (Lan/Ybn = 1.01-1.95) and show distinct negative Eu anomalies (Eun/Eu* = 0.49-0.63, Fig. 7). The increase in total REE content from gabbro-diabase to plagiogranite and similarity of the REE patterns suggest that the rocks are cogenetic (Fig. 7). Comparison between the Cape Povorotny plagiogranites and those from the Bay of Islands ophiolites, Newfoundland (Elthon 1991), shows that both have negative Ta and Ti anomalies (Fig. 8) indicative of suprasubduction zone origin (Pearce & Norry 1979; Saunders et al 1980; Shervais 1982; Elthon 1991). Negative Ta and Nb anomalies are also shown by the Cape Povorotny plagiogranites when plotted on the ocean-ridge granite (ORG)-normalized (Pearce et al. 1984) patterns (Fig. 9). Chondrite-normalized REE patterns of plagiogranites and trondhjemites from the Maqsad ophiolite, Oman, considered by Amri et al. (1996) to have formed at a mid-ocean ridge, differ from those of Cape Povorotny plagiogranites in having lower REE totals and REE patterns with both negative and positive Eu anomalies (Fig. 10). It should be noted that Cox et al. (1999) and Searle & Cox (1999) assumed that Oman ophiolitic crust and its plagiogranite were generated above an intra-oceanic subduction zone. Basalts are the dominant volcanic rocks in the Main and Median melanges of the Cape Povorotny accretionary complex (Fig. 5). Basalts make up isolated tectonic blocks and flow units within terrigenous-tuffaceous sequences. Based on petrographic and geochemical evidence, these rocks are divided into the following groups (Sokolov et al.

627

1999; Silantyev et al. 2000): (1) boninites with low TiO2, extremely low middle REE (MREE) and heavy REE (HREE), and high MgO contents; (2) aphyric and phyric calc-alkaline basalts and low-K tholeiitic basalts; (3) pillowed and massive tholeiitic basalts with marked N- and E-MORB geochemical features. Cape Povorotny peridotites (Bazyler et al. 2000) occur in dismembered ophiolite sequences at various localities confined to NE-trending tectonic units. From SE to NW, these units crop out in the Povorotny, Median, and Main (including Greben and Beregovoi massifs) serpentinite melanges (Fig. 5). The central zone of the Greben massif is composed of spinel Iherzolite. The outer zone of the Greben massif, the Beregovoi massif, and small peridotite blocks of the Main and Median melanges are composed of harzburgite proper and Cpx-bearing harzburgite. Peridotites from the Povorotny melange are also harzburgites, but more depleted, judging from spinel compositions (Table 2). Compositions of other mineral phases have been given by Bazylev et al. (2001). The Median melange is dominated by cumulate peridotites, including pyroxene-bearing dunites, chromitites, wehrlites, and plagioclase harzburgites, which are also found in the Main Melange. Spinels from residual peridotites (Iherzolites and harzburgites) have Cr number (Cr/(Cr + Al)) ranges as wide as 0.18-0.70 (Bazylev et al. 2001), suggesting SSZ provenance for at least some of these rocks (Dick & Bullen 1984). Representative spinel compositions from the peridotites are given in Table 2. The Mg number (100Mg/(Mg + Fe)) of olivines and orthopyroxenes from the residual spinel peridotites does not correlate with the spinel Cr number (varying in the range 89.6-91.7), indicating their origin by open-system melting or melt-rock interaction, rather than by simple partial melting (Bazylev et al. 2001). Silicate mineral compositions from peridotites have been reported by Bazylev et al. (2001). Spinel compositions from the cumulate peridotites have high Cr number (0.30-0.79), elevated iron oxidation degrees, and low Ti contents, further supporting an SSZ rather than a MOR affinity for these rocks (Arai 1992). REE patterns from all the peridotite varieties including wehrlites are also LREE enriched, some of the spectra being U-shaped. They also have significant negative Nb and Zr anomalies, some samples also having negative Ti anomalies (Fig. 11). The data on rock geochemistry have been given by Bazylev et al. (2001). These features were explained by Bazylev et al. as resulting from open-system melting (Ozawa & Shimizu 1995) of mantle material accompanied by melt influx.

S. D. SOKOLOV ETAL.

628

Table 1. Major (wt%) and trace (ppm) element contents of gabbro-diabases and plagiogranites in blocks from the Main Melange zone Sample

c-2415

c-2340

c-2415/1

c-2340/2

c-2340/4

c-2340/1

Si02 TiO2 A12O3 Fe203

51.53 0.76 13.12 3.84 6.52 8.96 7.35 0.08 4.07 0.04 0.07 3.30 99.64

51.8 0.94 14.68 2.60 4.60 12.94 5.91 0.06 3.98 0.35 0.02 2.46 100.34

53.71 1.03 14.07 4.91 6.60 5.97 5.05 0.09 4.68 0.32 0.11 3.01 99.56

63.84 0.60 13.7 1.59 2.93 6.22 3.20 0.04 5.83 0.10 0.01 2.31 100.36

73.04 0.69 12.22 0.53 0.50 3.77 1.30 0.02 5.66 0.30 0.09 1.02 99.74

75.51 0.43 12.56 1.34 0.58 1.61 1.03 <0.01 6.15 0.20 0.02 0.84 100.33

FeO CaO MgO MnO Na2O

K20 P205

LOI Total

Th Zr Hf Nb Ta Y Rb Sr Ba Cs Sc Co La Ce Nd Sm Eu Tb Yb Lu Lan/Ybn Lan/Smn Eun/Eu*

-

_

-

1.70 5.30 4.90 1.80 0.35 0.51 2.00 0.32 0.57 0.84 0.63

4.60 11.00 9.50 3.40 0.95 0.83 3.20 0.50 0.96

1.1 0.89

0.9 120 4.4 2.7 0.06

44 7 21 0.3 20 2.2 11.00 28.00 19.00 6.00 1.10 1.20 4.60 0.71 1.60 1.03 0.49

11.00 26.00 18.00 5.60 1.40 1.30 4.60 0.73 1.60 1.11 0.63

1.5 160 5.9 3.9 0.10

37 7 130 0.3 13 1.2 5.60 17.00 11.00 3.20 0.93 0.92 3.70 0.65 1.01 0.87 0.63

c-2340/3 77.03 0.53 11.12 1.37 0.65 3.50 0.20

0 5.47 0.06 0.02 0.40 100.35

14.00 31.00 19.00 5.70 1.40 1.30 4.80 0.82 1.95 1.19 0.56

c-2415 to c-2415/1, gabbro-diabase; c-2340/2 to c-2340/3, tonalites md plagiogranites. Major elements were measured by wet chemistry, trace elements by XRF, and REE by INAA.

As the data reported above and by Bazylev et al. (2001) preclude a mid-oceanic ridge origin for the Cape Povorotny Iherzolites, the earlier geodynamic interpretation of these rocks (Palandzhyan & Dmitrenko 1999; Silantyev et al. 2000) should be revised.

Origin of igneous and metamorphic rocks Fig. 7. Chondrite-normalized (Sun & McDonough 1989) REE patterns for gabbro-diabase and plagiogranites from blocks in Main Melange zone.

The available geochemical data collectively suggest the principal rock assemblages that make up the ophiolite melange in the Cape Povorotny accretionary pile to be (Silantyev et al. 2000):

OPHIOLITES OF NORTHEAST ASIA

Fig. 8. Chondrite-normalized (Sun & McDonough 1989) REE patterns for plagiogranites from Cape Povorotny, Taigonos Peninsula, and Bay of Islands (Newfoundland).

Fig. 9. ORG-normalized (Pearce et al 1984) patterns for plagiogranites from Cape Povorotny, Taigonos Peninsula.

Fig. 10. Chondrite-normalized (Sun & McDonough 1989) REE patterns for plagiogranites from Cape Povorotny, Taigonos Peninsula, and trondhjemites and plagiogranites from Oman ophiolite, Maqsad region.

(1) high-grade amphibolites derived from plutonic and, rarely, volcanic rocks of N- and E-MORB affinities; (2) plutonic rocks including boninitic and normal gabbro and diabase with N-MORB and within-plate or E-MORB affinities, and plagiogranites; (3) MORB-like basalt, WPB, and various products of SSZ magmatism, including boninite; (4) peridotites, including spinel Iherzo-

629

lite, spinel harzburgite, dunite, and wehrlite of SSZ affinity. Thus, the Cape Povorotny ophiolite sequence probably contains lithospheric fragments of two types. Type 1 consists of spinel Iherzolite, spinel harzburgite, dunite, and wehrlite. Spinel Iherzolites and harzburgites originated through various degrees of mantle melting above a subduction zone. Judging by data reported by Silantyev et al. (2000), the low-Ti volcanic rocks of island-arc affinity (low in TiO2, Cr, Nb, and with high Mg number) from the Main Melange and the Cape Povorotny spinel harzburgites might be cogenetic, both representing products of subduction magmatism. The compositions of spinel from the cumulate peridotites, particularly elevated degrees of iron oxidation and low Ti contents, confirm an SSZ rather than oceanic-basin origin for these rocks (Arai 1992). Plutonic rocks of boninite affinity, as well as boninites and calc-alkaline volcanic rocks and low-K basalts, occur in the same lithospheric fragments (Silantyev et al. 2000). The plagiogranites may have originated through anatexis of a gabbroic parent or as interstitial melt crystallized during mafic magma fractionation. The similarity of chondrite-normalized REE patterns for the gabbro-diabase and plagiogranite supports the latter interpretation. Negative Ta, Nb, and Ti anomalies in the plagiogranites point to their origin in an SSZ setting (Figs 8 and 9) (Pearce & Norry 1979; Saunders et al 1980; Shervais 1982; Elthon 1991; Saunders et al 1991; Pearce 1992). Type 2 lithospheric fragments in the Cape Povorotny ophiolitic melange have oceanic geochemical and mineralogical signatures. These fragments are observed in the Main Melange (Fig. 5) and are represented by protoliths of the high-grade amphibolites, gabbros, and dolerites, which show N-MORB and within-plate (or E-MORB) affinities, as well as MORB and WPB volcanic rocks. According to Shervais (2001), Type 1 magmatic and metamorphic assemblages can be interpreted as products of a complete life cycle of SSZ ophiolites that includes (1) birth: low-K tholeiites, basaltic andesite, and spinel harzburgites; (2) youth: boninite s (volcanic and plutonic) and cumulate peridotites; (3) maturity: basaltic andesite, andesite, hornblende gabbro, and plagiogranite; (4) death: formation of high-J metamorphic soles (amphibolite and garnet amphibolite). P—T estimates for metamorphic rocks suggest two principal types of metamorphism for Cape Povorotny ophiolites (Silantyev et al. 2000). Type 1 reflects relatively high-P (8 kbar) and medium- T (500-700 °C) conditions and is unique to the high-grade amphibolites, and type 2, which post-

Table 2. Spinel composition (wt%) from Cape Povorotny peridotites, Taigonos Peninsula No.: Sample: Rock: Phase: Points: SiO2 TiO2 A1203 FeO MnO MgO Cr203 NiO V203 ZnO Total

No.: Sample: Rock: Phase: Points: Si02 TiO2 A12O3 FeO MnO MgO Cr203 NiO V203 ZnO Total

1 T40/1 LH Spll 3

2 T41/1 LH Spll 4

3 C2312 LH Spll 8

4 C2356 LH/H2 Spll 4

5 C2362/1 HI Spll 1

6 T4/2 HI Spll 3

7 C94011 HI Spll 4

8 C94011 HI FCR 2

9 T49/1 HI Spll 2

0.02 0.02 50.60 13.91 0.07 17.92 18.18 0.22 n.a. n.a. 100.94

0.01 0.01 47.54 15.39 0.07 18.01 20.13 0.25 n.a. n.a. 101.41

0.09 0.07 50.20 14.04 0.12 18.16 17.13 n.a. n.a. 0.11 99.93

0.12 0.12 47.50 14.63 0.09 18.07 19.29 n.a. n.a. n.a. 99.80

0.01 0.08 40.61 14.80 0.17 16.79 27.44 n.a. n.a. 0.26 100.15

n.a. 0 39.06 15.60 0.19 16.08 29.07 0.15 n.a. n.a. 100.15

0.25 0.03 33.97 17.55 0.25 14.81 31.86 0.16 0.13 0.37 99.38

3.35 0.03 2.94 55.30 3.61 4.75 26.46 0 0.15 0.42 97.01

n.a. 0 30.89 19.63 0.34 12.98 35.56 0.07 n.a. n.a. 99.47

14 C9413/4 HI Spll 2

15 C9413/4 HI FCR 3

16 C2492/2 CP1 Spll* 1

17 C2492/2 CP1 Spl2 1

18 C2492/2 CP1 Spl2 3

19 C2492/2 CP1 Spl2 2

20 C2501/1 GDI Spll 2

21 C9425/8 CP1 Spll 7

22 C9425/8 CP1 FCR 2

0.25 0.05 33.08 16.46 0.20 15.46 33.43 0.15 0.26 0.34 99.69

5.36 0.11 1.66 54.55 2.94 6.25 27.56 0.11 0.15 0.38 99.06

0.12 0.20 30.67 26.09 0.22 12.72 29.63 n.a. n.a. n.a. 99.60

0.24 0.33 25.10 35.42 0.56 10.54 27.39 n.a. n.a. n.a. 99.58

0.07 0.36 16.13 38.76 0.52 8.38 35.12 n.a. n.a. n.a. 99.33

0.06 0.09 11.15 39.05 0.42 6.93 40.56 n.a. n.a. 0.28 98.54

0.06 0.21 28.33 19.50 0.25 14.24 36.45 n.a. n.a. 0.07 99.11

0.11 0.29 25.45 26.41 0.26 12.38 34.13 0.20 0.18 0.19 99.60

2.40 0.22 0.61 67.28 0.51 2.20 20.59 0.08 0.08 0.07 94.04

10 C2352/1 HI Spll 2

0.05 0.10 29.84 16.73 0.17 14.99 37.40 n.a. n.a. 0.12 99.39 23 C2315 GDI Spll 4

0.07 0.13 25.10 18.18 0.21 13.28 41.83 n.a. n.a. 0.12 98.92

11 C9413/6 HI Spll 4

0.10 0.06 28.95 18.84 0.22 13.70 37.18 0.11 0.24 0.21 99.60 24 C2501/14 CWH1 Spll 9

0.16 0.15 23.29 24.16 0.24 11.21 40.46 0.12 0.25 0.25 100.28

12 C9413/6 HI FCR 2

4.30 0.17 2.60 51.39 3.06 5.87 28.98 0.05 0.16 0.74 97.33 25 T19/5 H2 Spll 2

n.a. 0 22.27 22.97 0.26 11.50 43.27 0.01 n.a. n.a. 100.28

13 C2314 HI Spll 2

0.06 0.05 27.92 18.49 0.19 13.33 39.09 n.a. n.a. 0.15 99.29 26 C2330/4 H2 Spll 2

0.07 0.03 22.12 19.80 0.24 11.47 44.67 n.a. n.a. 0.26 98.68

OPHIOLITES OF NORTHEAST ASIA

631

dates type 1, is associated with low-J recrystallization of all igneous and metamorphic rocks, including the pre-existing high-grade amphibolites, under greenschist- and zeolite-facies conditions (350-380 °C and <2 kbar) (Silantyev et al 2000). Silantyev et al. (2000) proposed that discrete blocks of high-grade amphibolite from Cape Povorotny ophiolites reflect dismemberment of an inverted subduction-related metamorphic aureole. Judging by the P-T path of different metamorphic events recorded in the amphibolites, Cape Povorotny high-grade amphibolites were exhumed from a depth of c. 20km (Silantyev et al. 2000). The available geochemical and petrological data suggest that the Cape Povorotny ophiolite developed as follows: (1) generation of MORB-like and within-plate magmatic assemblages (gabbro, diabases, and volcanic rocks) in an oceanic basin; (2) partial subduction of these rocks into a relatively shallow, warm, and young subduction zone with partial detachment of the subducting oceanic lithosphere and its tectonic incorporation into the SSZ pile; (3) formation of an SSZ magmatic complex, including spinel Iherzolite and harzburgite, cumulate peridotites, and plagiogranite; (4) exhumation of high-grade amphibolites and residual peridotites followed by their tectonic incorporation into the SSZ pile.

Ophiolites in the Yelistratov Peninsula Geological setting In the Yelistratov Peninsula, ophiolites make up forearc basement to the Uda-Murgal island arc (Sokolov et al. 1999) and are unconformably overlain by Lower Cretaceous clastic deposits (Parfenov 1984; Filatova 1988). A preliminary geological description of the ophiolite was given by Belyi & Akinin (1985). Our study shows that ophiolitic rocks occur in a number of tectonic slices (Fig. 12). On the north of the peninsula, the serpentinite melange dips steeply NW, under the Jurassic-Lower Cretaceous sequence of calc-alkaline volcanic rocks of the Uda-Murgal island arc (Fig. 12). In the southern part, the melange is overlain by slices of cumulate gabbro and Berriasian to Hauterivian (Rosenkrantz 1986) tuffaceous and terrigenous deposits (Belyi & Akinin 1985; Ishiwatari et al. 1998). The melange contains harzburgite and dunite blocks of various sizes showing different degrees of serpentinization, and sheeted dykes, metabasalt, and radiolarite. The gabbro sheet forms a gentle synform and, in the south, is underlain by a slice of harzburgite slice. The sheeted dyke complex is

632

S. D. SOKOLOV ETAL.

Fig. 11. Chondrite-normalized (Anders & Grevesse 1989) concentrations of REE (a) and other trace elements (b) in peridotites from Cape Povoromy. Analytical data have been given by Bazylev et al. (2001).

composed of a differentiated series from basalt to dacite. The age of the ophiolite is uncertain, but microfossils from radiolarites encountered in the Northern melange zone range from Mid- to Late Jurassic (Belyi & Akinin 1985). Relevant mineralogical data were recently published and discussed by Ishiwatari et al. (1998), Palandzhyan & Dmitrenko (1999) and Saito et al. (1999). This review draws generally on these papers, as well as on our own mineralogical and chemical data, partly presented below.

Petrography and geochemistry of mafic and ultramafic rocks The rock suite of the Yelistrarov Peninsula ophiolite assemblage includes both residual and cumulate peridotites, cumulate plagioclase Iherzolite, pyroxenite and gabbronorite, gabbroic screens be-

tween the diabase dykes, and discrete blocks of foliated metabasalt. Mantle peridotites are dominated by spinel harzburgites with subordinate pyroxene-bearing dunites. Overall, their serpentinization degree ranges from 85 to 100%. Primary, low-Ti spinels in the residual peridotites are moderately to highly chromian with Cr numbers ranging from 0.41 to 0.66 (Table 3). Those in dunites have Cr numbers up to 0.75. Saito et al. (1999) reported an even wider range of spinel Cr number in the peridotites (0.29-0.72). The oxidation degree (Fe3+/(Cr + Al + Fe3+) of spinels in harzburgite ranges from 0.021 to 0.044, and in dunites it is up to 0.087 (Table 3). Olivines in these peridotites are highmagnesian with Mg numbers in the range of 90.6-91.9 and have elevated nickel contents as compared with other ultramafic lithologies (0.36— 0.40 wt% NiO) (Table 4). Residual peridotites exhibit a negative correlation between spinel Cr number and olivine Mg number, such that rocks

633

OPHIOLITES OF NORTHEAST ASIA

Fig. 12. Map showing schematic geological structure of the Yelistratov Peninsula.

Table 3. Representative spinel compositions from the peridotites of the Yelistratov Peninsula ophiolite complex Sample: Rock: Points:

C2518/1 Hz 3

C2529/4 Hz 5

C2524/1 Hz 3

C2524/3 Hz 4

C2524/2 D 3

C2526 0 5

C2520/5

SiO2 TiO2 A1203 FeO MnO MgO Cr203 NiO V203 ZnO Total Crno. Mg no. Fe no.

0.05 0.07 34.03 14.14 0.14 16.64 34.71 0.22 0.21 0.12 100.32 0.406 0.716 0.028

0.21 0.02 26.97 17.65 0.21 13.54 41.57 0.14 0.20 0.22 100.72 0.508 0.607 0.026

0.09 0.07 22.63 18.54 0.18 12.74 45.60 0.09 0.35 0.10 100.39 0.575 0.585 0.031

0.07 0.08 17.85 19.86 0.22 12.22 50.76 0.10 0.19 0.09 101.43 0.656 0.569 0.044

0.08 0.10 11.67 24.37 0.30 10.36 53.05 0.07 0.20 0.11 100.32 0.753 0.507 0.087

0.08 0.48 16.43 31.33 0.32 8.44 42.55 0.17 0.28 0.28 100.35 0.635 0.408 0.130

3

C2524/4 PW 3

C2518/7 PL 8

0.17 0.34 17.53 39.16 0.44 6.57 34.13 0.11 0.35 0.37 99.17 0.566 0.324 0.205

0.17 0.97 18.34 44.31 0.37 6.26 29.07 0.26 0.61 0.07 100.43 0.515 0.298 0.253

0.11 0.72 15.00 49.94 0.38 5.48 27.40 0.17 0.65 0.18 100.03 0.551 0.269 0.332

oc

Rock types: Hz, spinel harzburgite; D, dunite; O, orthopyroxenite; OC, olivine clinopyroxenite; PW, plagioclase websterite; PL, plagioclase Iherzolite. All Fe is in the form FeO. Cr number = Cr/(Cr + Al); Mg number = Mg/(Mg + Fe2+); Fe number = Fe3+/(Cr + Al + Fe3+), where Fe3+ is calculated from stoichiometry.

S. D. SOKOLOV ETAL.

634

Table 4. Representative olivine composition from the peridotites of the Yelistratov Peninsula ophiolite complex Rock: Sample: Points:

Hz C2518/1 4

Hz C2529/4 5

Hz C2524/1 5

Hz C2524/3 4

0 C2526 5

OC C2520/5 4

PW C2524/4 4

PL C2518/7 4

SiO2 FeO MnO MgO CaO NiO Total Mg no.

40.80 7.92 0.12 50.63 0.05 0.38 99.89 91.9

41.87 8.40 0.13 50.25 0.04 0.39 101.08 91.4

41.02 9.14 0.13 49.63 0.04 0.38 100.34 90.6

41.46 8.88 0.15 50.20 0.05 0.37 101.10 91.0

40.20 12.62 0.19 47.85 0.05 0.28 101.18 87.1

40.11 13.54 0.18 46.41 0.05 0.14 100.44 85.9

39.81 16.13 0.24 44.81 0.06 0.24 101.29 83.2

39.73 16.79 0.26 44.34 0.05 0.18 101.34 82.5

Rock types: Hz, spinel harzburgite; O, orthopyroxenite; OC, olivine clinopyroxenite; PW, plagioclase websterite; PL, plagioclase Iherzolite. All Fe is in the form FeO. Mg number = 100Mg/(Mg + Fe)

with the least chromian spinels contain the most magnesian (91.4-91.9) olivines. The cumulate rock suite includes Ol-bearing and Ol-free gabbronorite and hornblende gabbronorite (Saito et al. 1999), as well as wehrlite, olivine orthopyroxenite, clinopyroxenite, olivine websterite, and plagioclase-bearing pyroxenite and Iherzolite. A few isolated bodies of hornblendebearing plagioclase peridotite are also present. The degree of serpentinization in ultramafic rocks of the cumulate sequence ranges from 15 to 50%, being significantly higher only in olivine-rich lithologies. The cumulate peridotites are distinctly less magnesian than the residual peridotites and have olivines with Mg numbers of 79.4-87.1 and reduced NiO contents of 0.14-0.28 wt% (Table 4). Primary spinel is present in most of the cumulate peridotites, except for some plagioclase Iherzolites with the least magnesian silicates. The spinel has moderate titanium contents (0.34— 0.97 wt% TiO2), moderate Cr numbers (0.450.65), and high oxidation indices (0.130-0.330) (Table 3). Clinopyroxenes from the cumulate peridotites are relatively low in aluminium (1.82.9 wt% A12O3) and titanium (0.10-0.24 wt% TiO2) but show somewhat elevated sodium contents (0.16-0.27 wt% Na2O) as compared with mantle peridotites, whose clinopyroxenes have 0.05-0.23 wt% Na2O. In the cumulate peridotites, plagioclase is An92-93 anorthite. Compositions of spinel harzburgites and orthopyroxene dunites from the northern and southern outcrops are similar and have U-shaped REE patterns with chondrite-normalized contents of Ce 0.13-1.8, Tb 0.03-0.41, and Yb 0.06-0.57 (Table 5, Fig. 13). Such patterns for peridotites (as well as for clinopyroxenes from these rocks) are widely accepted as being typical of SSZ magmatism (Parkinson & Pearce 1998; Batanova & Sobolev 2000; Bizimis et al 2000).

Mafic screens between dykes consist of amphibole gabbro, Cpx-bearing amphibole gabbro, and amphibole gabbronorite. Mafic rocks of the sheeted dyke complex include both aphyric and porphyritic diabases, commonly with an aphanitic groundmass. These rocks consist of variable proportions of mostly albitized plagioclase, amphibole, clinopyroxene, orthopyroxene, and sporadic olivine. Small blocks of foliated metabasalts occur in the Northern zone melange (Fig. 12). They typically lack primary minerals and consist of carbonate, actinolite, chlorite, epidote, and albite. The most mafic rocks of the sheeted dyke complex are strongly depleted and fractionated low-K tholeiites with island-arc signatures, such as characteristic REE patterns, high field strength element (HFSE) depletion, and relatively elevated contents of large ion lithophile elements (LILE), including La (Fig. 14, Table 6). Metabasalts from the Northern melange zone and lavas from a breccia of Berriasian to Valanginian age have similar chemical signatures (Fig. 14, Table 6). The sheeted dyke complex contains some rocks (Table 4) closely resembling high-Ca boninites.

Origin of mafic and ultramafic rocks A number of mineralogical and chemical indicators can be used to elucidate the geodynamic setting of peridotites. The ranges of Cr number in residual spinels from the Yelistratov peridotites are typical of SSZ settings rather than mid-ocean ridges, whose typical ranges are 0.2-0.6 (Arai 1994). High iron oxidation degrees in cumulate spinels coupled with the moderate titanium contents indicate an island-arc setting for magmatism (Arai 1992). The decrease of olivine Mg number with increasing Cr number in spinel (Tables 3 and 4) suggests extensive melt-rock interaction during a magmatic process and is inconsistent with a

OPHIOLITES OF NORTHEAST ASIA

635

Fig. 13. Chondrite-normalized REE patterns for peridotites, Yelistratov ophiolite. U-shaped patterns for residual peridotites suggest suprasubduction-zone setting for magmatism.

Fig. 14. Chondrite-normalized (Sun & McDonough 1989) (a) and N-MORB-normalized (Saunders & Tarney 1984) (b) patterns for basic rocks of dyke and lava complexes from the Yelistratov Peninsula.

simple partial melting of mantle (Jaques & Green 1980). The olivine-spinel compositional relationships of this type were described from mantle peridotites of SSZ origin (Bazylev et al 1993, 2001; Sobolev & Batanova 1995), but they do not appear in the mid-ocean ridge mantle peridotites (Dick & Bullen 1984; Arai 1994).

S. D. SOKOLOV ETAL.

636

Table 6. Major (wt%) and trace element (ppm) contents in dykes and lavas of the Yelistratov Peninsula ophiolite Sample: SiO2 TiO2 A1203 Fe203 FeO MnO MgO CaO Na2O K20 P205 LOI Total Cr Co Ni V Sc Nb Rb Sr Zr Y La Ce Nd Sm Eu Tb Yb Lu

M-97-37/3

C-2505/2

C-2505/6

C-2533/3

C-2533/7

C-2516/4

C-2527/3

M-97-23/1

51.35 0.43 13.22 3.41 6.48 0.16 10.09 9.01 3.17 0.18 0.05 1.76 99.31 513 248 1.3 1.52 123.7 26.3 15.08 1.3 2.3 2.5 0.96 0.36 0.3 1.5 0.24

55.38 0.72 15.2 5.42 4.22 0.07 5.33 6.28 4.95 0.08 0.05 2.1 99.8 33.3 22.2 21.6 383 38.5 1 2.6 200 55 19 2.5 6.2 5.6 2.1 0.61 0.58 2 0.36

51.33 0.46 14.22 4.86 5.03 0.27 10.9 4.84 2.61 1.42 0.05 3.88 99.87 52.1 29.8 46.3 274 37.9 1 27 150 29 12 2 4.6 3.7 1.2 0.43 0.38 1.3 0.19

45.71 0.33 13.78 4.99 8.77 0.09 14.25 4.26 1.88 0.25 0.05 5.57 99.93 359 69.5 181 265 42 1 2.4 95 22 9.8 1 2.6 2.4 0.87 0.31 0.26 1.1 0.18

54.2 0.37 13.45 3.73 4.87 0.08 8.42 7.16 2.91 0.91 0.05 3.82 99.97 65.1 34.3 40.7 298 38.2 1 7.6 200 33 13 1.5 4.3 3.9 1.4 0.47 0.48 1.7 0.26

53.03 0.55 16.15 3.94 4.26 0.16 9.81 2.2 3.78 2.06 0.07 3.65 99.66 81.3 30.5 45.8 201 26 1.1 17 82 58 17 3.6 9.1 6.5 2.1 0.7 0.64 1.8 0.3

48.39 1.12 13.49 6.14 5.72 0.52 10.09 5.59 2.75 2.01 0.08 3.98 99.88 259 51.9 86.7 282 47.3 1.8 33 87 47 19 2.5 6.3 5.4 2 0.63 0.6 2.1 0.32

41.24 1.01 10.82 4.55 8.31 0.2 7.91 20.79 0.2 0.01 0.08 4.24 99.36 135 281 1.4 1 32 44 19 2.5 6.2 5.8 2.4 0.74 0.68 2.6 0.41

1-6, rocks of dyke complex; basalt from volcanic breccias Klb-v; 8 and 9, metabasalts from Northern melange zone. Major elements were determined by wet chemistry, trace elements by XRF, and REE by INAA.

The subdivision of Yelistratov residual peridotites by Ishiwatari et al (1998) and Saito et al (1999) into the Northern ultramafic body (with more depleted peridotites showing spinel Cr number (0.39-0.72) of presumable SSZ origin) and the Southern ultramafic body (with less depleted peridotites showing spinel Cr number (0.29-0.49) of presumable mid-ocean ridge origin) is highly questionable in view of the following. The residual spinel Cr number of 0.49 is out of the ranges typical for slow-spreading mid-ocean ridges (Dick 1989), and the spinel Cr number of 0.29 is out of ranges typical both for fast-spreading mid-ocean ridges and for geochemically anomalous segments of slow-spreading ridges (Dick & Natland 1996; Bazylev & Silantyev 2000). Therefore, the ranges of residual spinel Cr number of 0.29-0.49 are not consistent with any particular mid-ocean ridge setting. Additionally, the large difference between the minimum and maximum values of the residual spinel Cr number found within a body implies a large difference in the mantle partial melting

degrees (e.g. Jaques & Green 1980) between neighbouring parts of this body. Such a difference cannot be achieved by a decompression mechanism of mantle melting below a mid-oceanic ridge (Langmuir et al. 1992), but invokes another mantle melting mechanism and another geodynamic setting for the magmatic process. The reported trace element data for the residual peridotites are consistent more with the conclusion of Palandzhyan & Dmitrenko (1999) that all the harzburgites of the Yelistratov complex, as well as the ultramafic-mafic cumulate rocks, are of SSZ origin. It can also be concluded that the volcanic rocks and sheeted dykes were generated in an ensimatic island arc. Ophiolites of the Ganychalan terrane Geological setting The Ganychalan terrane is situated within the Penzhina segment of the Uda-Murgal island arc (Fig. 15).

OPHIOLITES OF NORTHEAST ASIA

Fig. 15. Geological map of the Penzhina segment (after Sokolov et al. 1996).

637

638

S. D. SOKOLOV ETAL.

On the NW, it is separated from the Kharitonya terrane by a high-angle normal fault boundary. The Kharitonya terrane is a fragment of the Koni-Taigonos volcanic arc and consists of an Early Carboniferous terrigenous assemblage intercalated with intermediate volcanic rocks in its lower part and with carbonaceous shales and coal beds in the upper (Khanchuk et al. 1992; Sokolov 1992). On the SE, the Ganychalan terrane rests techtonically over the Upupkin thrust pile along a thrust plane dipping from 30° to 80°. The internal structure of the Ganychalan terrane is a large closed or tight antiform of complicated shape with a thick normal NW limb and tectonically reduced overturned SE limb (Fig. 15c). The Ganychalan terrane consists, from bottom upward, of four tectonic slices: Ilpenei, Mrachnaya, Khinantynup, and Elgeminai (Fig. 16) (Khudoley & Sokolov 1998; Ganelin & Peyve 2001). The Ilpenei slice is composed of basalt, tuff, shale, chert, limestone, and quartzite, all affected by metamorphism of greenschist and glaucophane schist facies (Dobretsov 1974; Silantyev et al. 1994; Vinogradov et al. 1995). Metabasalts in this slice are MORE- and ocean island basalt (OIB)like (Khanchuk et al. 1992; Silantyev et al. 1994). The cherts yielded deformed unidentifiable radiolarians, but sporadic finds of poorly preserved conodonts suggest a Palaeozoic age (V A. Aristov, pers. comm.). The radiometric age of greenschist metamorphism is 327 ± 5 Ma (whole-rock Rb/Sr determination,Vinogradov et al. 1995). The meta-

Fig. 16. Tectonostratigraphic units of the Ganychalan terrane.

morphic rocks are viewed as a subduction complex associated with the Koni-Taigonos island arc. The Mrachnaya, Khinantynup, and Elgeminai slices (Figs 16 and 17) exhibit ultramafic-mafic

Fig. 17. Geological map of the central Ganychalan terrane (by V G. Batanova & A. V Ganelin).

OPHIOLITES OF NORTHEAST ASIA magmatic assemblages that comprise a dismembered ophiolite suite of Early Palaeozoic age, constrained by radiometric and palaeontological data (Markov et al 1982; Khanchuk et al. 1992; Ganelin & Peyve 2001; Nekrasov et al 2001). The Mrachnaya slice is composed of a serpentinite melange consisting of chaotically oriented blocks of ultramafic and gabbroic rocks. The most widespread lithologies are plagioclase wehrlite, pyroxenite, troctolite, and olivine gabbro associated with minor olivine-free normal gabbro and plagiogranite. Dunite and harzburgite are extremely rare and are totally serpentinized. The Khinantynup slice is a plutonic assemblage dominated by mafic lithologies, including leucocratic medium-grained layered gabbro, coarsegrained isotropic gabbro, gabbronorite, and ferrogabbro. Felsic rocks include a gneissosed tonalite-plagiogranite body, 50-70 m thick, which is concordant with textures in gabbroic host rocks, and small concordant (0.1-0.5m) lenses and veins of quartz diorite, tonalite, and plagiogranite. The base of the gabbroic section contains layers of zoisite-bearing amphibolite and quartzgarnet-amphibole schists (Markov et al. 1982). A hornblende Ar/Ar age of the gabbroic rocks is Early Ordovician (559 Ma; Khanchuk et al 1992). Toward the top of the slice, gabbro gives way to gabbro-diabase and diabase, which are chilled against the host gabbro. Gabbro-diabases are crosscut by a 20-30 m thick tonalite-plagiogranite body. Felsic rocks contain abundant gabbrodiabase xenoliths. Irregular zones of pegmatitic gabbro are found along the contact of gabbrodiabases and tonalite-plagiogranite body. This disrupted dyke system can be viewed as a deformed sheeted dyke complex that occurs along the tectonic contact with volcanic and sedimentary rocks of the Elgeminai slice. These rocks include albitized basalts, spilites, amygdaloid pillowed basalts, and subordinate intercalations of chert and siliceous tuff. These are interpreted as the upper portion of the ophiolite assemblage, overlain by terrigenous and carbonate strata. Basalts yield whole-rock K-Ar ages of 490-480 Ma (Dobretsov 1974). Cherts contain conodonts of Arenig to Llanvirn age (485-464 Ma).

that they are ultramafic cumulates (Nekrasov et al 2001). Plagioclase wehrlite, troctolite, and olivine gabbro of the Mrachnaya slice are fragments of a cumulate complex. They show variable proportions of mafic minerals and plagioclase with ensuing broad variations in major element oxides, especially CaO, A12O3, and MgO (Table 7). This group is the lowest in TiO2, Zr, Y, and REE among Ganychalan ophiolite plutonic rocks. The rocks have REE (La + Sm + Yb) in the range 0.14-0.87 ppm and sawtooth REE patterns (Fig. 18), some of which show positive Eu anomalies. They contain high-Mg, high-Cr clinopyroxenes (Mg number 87.3-88.4; 0.53-1.15 wt% Cr2O3), which are low in TiO2 and REE (Table 8). Petrographically and geochemically, the gabbroic rocks of the Khinantynup slice fall into several varieties. Leucocratic gabbro is high in A12O3 (22.7-24.8 wt%) and low in TiO2 and other

Petrography and geochemistry of the plutonic and volcanic rocks In serpentinized peridotites, the only surviving primary mineral is chrome-spinel. The rocks have high Mg number of 88.6-90.2, but are rather low in Cr (850-1600 ppm) and Ni (900-1300 ppm). Their high Cr number of 65.6-87.4 and relatively high-Mg spinels (Mg number 36.4-57.4) suggest

639

Fig. 18. Chondrite-normalized (Sun & McDonough 1989) REE patterns of mafic plutonic rocks from the Ganychalan terrane. Grey field indicates gabbro from Ocean Drilling Program Site (ODP) 894, Hess Deep (Pedersen et al. 1996).

Table 7. Major (wt%) and trace (ppm) element contents ofmagmatic rocks of the Ganychalan terrane Sample:

Si02 TiO2 A1203 Fe2O3 FeO MnO MgO CaO Na2O K20 P205 LOI Total Rb Ba Nb Sr Zr Y La Ce Nd Sm Eu Tb Yb Lu

1 Bm 19-8

2 Bm 19-11

3 Bm 19-14

4 Bm 23-2

5 Bm 17-3

6 Bh 3-3

7 Bh 15-1

8 Bh 6-2

9 Bh 6-9

10 Bh 7-1

11 Bh 18-1

12 P 108-1

13 P 129-1

14 mh65/2

36.33 0.48 5.64 5.77 5.37 0.17 31.16 3.95 0.41 0.07 0.01 10.3 99.66

36.25 0.4 4.88 8.25 3.45 0.17 31.15 3.2 0.09 0.01 0.02 11.66 99.53

40.15 0.28 21.32 2 3.29 0.06 12.66 12.86 1.51 0.04 0.01 5.27 99.45

43.9 0.48 11.54 1.19 4.74 0.09 19.43 13.74 0.98 0.24 0.02 2.9 99.25

44.29 0.28 22.76 2.29 1.2 0.07 7.1 17.83 1.28 0.22 1.79 99.11

45.67 0.71 16.8 1.99 3.52 0.12 9.9 15.73 1.92 0.16 0.01 3.21 99.74

47.12 0.68 17.58 2.33 4.89 0.11 8.26 13.65 2.42 0.09 0.01 2.9 100.04

46.33 0.61 17.18 3.08 3.34 0.11 4.09 17.24 2.7 0.68 3.1 98.46

42.43 4.58 17.36 5.72 7.84 0.17 5.23 10.87 3.77 0.45 0.02 2.3 100.74

48.1 1.35 15.92 5.78 4.82 0.12 9.55 8.91 4.86 0.47 0.13 1.2 101.21

47.82 1.95 17.63 4.46 4.84 0.22 7.62 10.63 4.32 0.36 0.14 1.35 101.34

51.97 1.03 19.38 1.89 5.05 0.14 5.92 9.16 4.49 0.85 0.11 1.92 101.91

49.67 1.23 16.78 2.7 6.36 0.21 8.11 9.83 4.23 0.7 0.18 1.32 101.32

62.82 0.94 13.59 2.75 2.26 0.05 0.97 7.07 6.03 0.42 0.07 2.00 100.19

18 10 11 0.09 0.34

10 10 10 0.06 0.23

170 10 11 0.06 0.19

77 10 10 0.08 0.43

160 10 10 0.09 0.27

_ _ _ _

230 10 14 0.42 2.95

180 27 12 0.34 1.55

_ 10 14 0.61 1.84

165 94 26 5.9 16.57

84 98 24 4.54 11.36

190 140 51 4.33 13.41

140 82 37 3.87 11.5

_ _ _

0.20 0.10 0.05 0.29 0.05

0.1 0.05 0.03 0.2 0.03

0.05 0.11 0.03 0.03 0.04

0.34 0.15 0.12 0.45 0.07

0.2 0.15 0.07 0.23 0.03

_ _ _ _

1.22 0.81 0.37 1.22 0.18

1 0.59 0.32 1.24 0.2

1.02 0.57 0.34 1.13 0.16

4.66 1.66 1.14 3.83 0.59

2.89 1.01 0.68 2.27 0.36

3.92 1.55 1.13 3.3 0.52

3.76 1.36 1.15 3.35 0.42

_ _ -

-

15 mh65/1

16 mh67/1

71.47 68.32 0.61 0.79 12.92 15.05 1.73 0.81 2.20 2.87 0.07 0.10 0.32 1.05 3.24 3.31 5.19 5.43 0.37 0.40 0.04 0.08 1.99 2.22 100.43 100.15 <10 <10 62 73 <10 <10 100 170 54 57 12 8 4.50 3.70 8.40 7.80 4 80 4 80 1.00 1.30 0.54 0.53 0.18 0.21 0.69 0.93 0.13 0.16

17 mh71/12

18 mh71/5

19 mh72/1

72.16 0.47 12.75 1.52 2.39 0.03 0.46 2.67 6.15 0.12 0.03 1.15 99.91

73.09 0.62 12.29 1.54 2.12 0.07 0.89 2.62 4.66 0.51 0.00 1.92 100.37 <10 51 <10 130 51 <10 3.30 6.80 0 00 0.92 0.53 0.13 0.47 0.07

72.32 0.51 11.71 2.79 2.51 0.04 0.16 3.48 4.55 0.49 0.00 1.63 100.08

_ _ _ _ _ -

20 mh66/6

68.30 0.76 15.47 0.79 2.30 0.07 0.50 3.30 6.40 0.34 0.05 2.15 100.43 <10 43 <10 _ 110 210 40 24.00 3.70 52.00 7.30 0 00 3 80 4.30 0.75 1.40 0.51 0.77 0.09 0.68 4.20 0.12 0.70

21 mh72/12

22 mh66/4

23 mh72/13

71.02 0.56 14.04 2.06 2.30 0.07 0.05 2.58 6.09 0.42 0.01 0.90 100.25

74.56 0.31 12.30 1.77 1.66 0.02 0.08 1.68 6.45 0.28 0.00 0.90 100.26

75.41 0.30 12.24 1.70 1.78 0.04 0.14 1.56 6.21 0.32 0.00 0.66 100.45

_ 31.00 63.00 32 00 5.30 1.40 0.83 4.60 0.79

_ _ 29.00 60.00 28 00 5.00 0.86 0.71 3.70 0.63

_ _ _ _ _ -

1 and 2, plagioclase wehrlite; 3, troctolite; 4, olivine gabbro; 5, leucogabbro; 6 and 7, coarse-grained gabbro; 8, gabbro-norite; 9, ferrogabbro; 10 and 11, diabase; 12 and 13, basalts; 14-19, tonalites, plagiogranites of the Khinantynup slice; 20-23, plagiogranites of the Elgeminay slice. Major elements determined by wet chemistry, trace elements by XRF, and REE by INAA.

OPHIOLITES OF NORTHEAST ASIA

641

Table 8. Major (wt%) and trace (ppm) element contents of clinopyroxenes from plutonic ophiolitic rocks of the Ganychalan terrane Sample: n: Si02 TiO2 A1203 Cr203 FeO MnO MgO CaO Na2O Total MAG n Ti Sr Zr La Ce

Nd Sm Eu Dy Er Y Yb

1 4 2 3 Bm 19-8 Bm 19-11 Bm 19-14 Bm 23-2 4 6 3 2

51.42 0.49 3.65 1.15 3.91 0.13 15.77 22.8 0.49 99.8 87.79 1 3699 1.73 21.12 0.33 1.74 3.31 1.74 0.48 3.99 2.65 14.81 2.49

51.56 0.37 3.16 0.97 4.21 0.11 16.36 22.75 0.36 99.85 87.35 2 2015 0.92 4.82 0.21 0.81 1.38 0.73 0.35 1.98 1.38 6.98 1.15

52.34 0.38 2.86 0.53 3.97 0.11 16.25 22.71 0.29 99.46 87.91 -

52.44 0.39 3.01 0.81 3.83 0.11 16.39 22.89 0.42 100.26 88.4 2 2505 1.06 6.88 0.31 1.3 1.6 0.95 0.32 2.04 1.38 7.16 1.02

5 Bm 17-3 2

6 Bh3-3 2

7 Bh 15-1 2

8 Bh6-2 2

9 Bh6-9 4

52.98 0.25 2.21 0.09 5.31 0.17 16.24 23.6 0.25 101.07 84.5 2 1381 0.72 2.89 0.39 0.88 0.93 0.45 0.24 1.35 0.93 5 0.74

52.05 0.84 2.97 0.49 4.94 0.14 15.69 21.12 0.43 98.64 84.99 1 2911 1.06 11.42 0.4 1.55 2.58 1.26 0.39 2.63 1.73 9.06 1.36

52.12 0.68 2.25 0.03 8.62 0.21 14.56 20.71 0.32 99.45 75.06 3 5054 1.22 24.2 0.56 2.93 5.41 2.96 0.87 6.3 4.38 23.08 3.56

51.77 0.66 3.55 0.08 8.86 0.25 15.41 19.36 0.38 100.23 75.61 1 6420 1.39 21.35 1.46 2.28 4.73 2.48 0.9 5.79 3.41 19.23 2.94

48 2.02 6.77 0.02 10.27 0.26 14.34 17.24 1.48 100.36 71.5 1 4332 2.11 48.6 1.34 5.6 8.79 3.84 1.65 8.05 4.68 24.14 3.91

Trace elements were determined by secondary ion mass spectrometry. MAG is the molecular ratio 100Mg/(Mg + Fe). Mineral compositions were analysed by electronic microprobe. n, number of spots analysed.

incompatible elements (La -f Sm + Yb of 0.52 ppm). These features suggest comagmatic origin of the cumulates (olivine-bearing gabbroids and wehrlites) and leucogabbros, a conclusion that is further supported by the clinopyroxene composition, the Mg number being as high as 84.5, and the low incompatible element abundances (Table 7). The coarse-grained gabbros, gabbronorites, and ferrogabbros, despite their petrographic diversity, make up a chemically uniform group (Table 7). They have low Mg number of 68-41.7 and elevated TiO2, up to 4.5 wt%, as a result of a considerable proportion of titanomagnetite. The average REE (La -f Sm + Yb) content in rocks within this group is c. 2.7 ppm, which is the highest value of all the plutonic complexes. These samples are LREE depleted (La/Yb = 0.3-0.5). Clinopyroxenes from coarse-grained gabbros and gabbronorites display moderate and uniform Mg number (75.6) and TiO2 contents (0.6 wt%), and are low in Cr2O3 (0.03-0.08 wt%) and high in REE (Table 8). Clinopyroxenes from ferrogabbroic

samples have the lowest Mg number of 71.5 and Cr2Oa content of 0.02 wt%, whereas their incompatible element concentrations are the highest. Diabases and basalts are chemically uniform and constitute a moderately differentiated rock series. Alkalinity is somewhat elevated owing mainly to spilitization-induced high Na2O content (Na2O + K2O of 3.15-5.62 wt%). Although the FeO*/(FeO* + MgO) ratio shows meaningless variations, from 0.94 to 1.70, MgO content ranges from 5.73 to 9.13wt%; TiO2 from 1.00 to 1.89 wt%, and A12O3 from 15.2 to 18.7 wt%. The REE total (La + Sm -f Yb) ranges from 9.4 to 13.9 ppm. REE patterns are flat and slightly LREE enriched (La/Yb = 1.16-2.00). These characteristics are similar to N-MORB and similar to TMORB (Fig. 18). Tonalites in the Khinantynup slice have gneissic textures with relics of subhedral granular texture. Plagioclase crystals are partly bent and have undulatory extinction. Quartz forms lens-shaped grains with undulatory extinction or aggregates of small grains with sutured boundaries. Primary

642

S. D. SOKOLOV ETAL.

mafic minerals, green hornblende and brown biotite, occur as relics in aggregates of secondary chlorite, sericite, and fine scaly green biotite. Plagiogranites also have gneissic texture, a lower amount of mafic minerals (1-2%), and as much as 30-40% quartz. Accessory minerals are zircon, sphene, and apatite. Tonalites and plagiogranites from the Elgeminai slice contain quartz-albite granophyric intergrowths. They are composed of quartz, plagioclase, amphibole, biotite, zircon, apatite, magnetite, and epidote. In plagiogranites, quartz accounts for 30-45%, and minor K-feldspar (<1%) is present. Plagiogranites from both slices fall dominantly in the trondhjemite field on the Ab-An-Or diagram (O'Connor 1969) (Fig. 19). They are low in A12O3 (11.71-15.47 wt%), K2O (0.120.51 wt%), and Rb and Nb (<10 ppm) (Table 7). Chondrite-normalized REE patterns for tonalites and plagiogranites from the two slices differ considerably (Fig. 20, Table 5). Those from the Khinantynup slice have low REE totals, a slightly fractionated REE distribution (Lan/Ybn = 2.66^.7; Lan/Smn = 1.25-1.41), and positive Eu anomalies (Eun/Eu* = 0.98-1.29) (Fig. 20). REE patterns from the Elgeminai tonalites and plagiogranites show higher REE totals, are partly more fractionated (Lan/Ybn = 3.2-5.4; Lan/Smn = 1.21-1.29), and have negative Eu anomalies (Eun/Eu* = 0.27-0.48).

Origin ofgabbroic and volcanic rocks and plagiogranites of the Ganychalan terrane The above compositional data suggest that the entire spectrum of ultramafic-mafic plutonic ophiolitic rocks from the Ganychalan terrane were

Fig. 19. Ab-An-Or diagram (O'Connor 1969) for plagiogranites and tonalites from the Khynantinup and Elgeminai slices, Ganychalan terrane ophiolites.

Fig. 20. Chondrite-normalized (Sun & McDonough 1989) REE patterns for plagiogranites and tonalites from the Khynantinup and Elgeminai slices, Ganychalan terrane ophiolites (data from Table 5).

produced by crystal fractionation of a basaltic melt. This is evidenced by (1) TiC>2, Zr, Y, and REE increasing consistently from olivine-bearing cumulates to ferrogabbros, while Mg number generally decreasing; (2) parallelism of the REE patterns. The overall degree of REE enrichment of the gabbros matches that of MOR gabbros (Fig. 18). To decide if the Mrachnaya and Khinantynup plutonic rocks are complementary to the Elgiminay basalts, Ganelin & Peyve (2001) calculated compositions of hypothetical parental melts in equilibrium with clinopyroxenes from the plutonic rocks. Using melt-clinopyroxene partition coefficients from Hart & Dunn (1993), they showed that hypothetical parental melt compositions underwent varying degrees of fractionation (Fig. 2la and b). REE abundances from melts in equilibrium with clinopyroxenes from the Mrachnaya olivine-bearing rocks and the Khinantynup noncumulate gabbros display a moderate degree of fractionation consistent with the upper limit of the N-MORB compositional range or with a somewhat enriched type, similar to T-MORB (Fig. 2la). This type matches closely the diabases and pillow basalts from the Elgeminai slice. The hypothetical melts in equilibrium with clinopyroxenes from coarse-grained gabbros, gabbronorites, and ferrogabbros from the upper part of the Khinantynup slice are very strongly fractionated, their REE enrichment being fivefold for clinopyroxenes from coarse-grained gabbro and gabbronorite and tenfold for clinopyroxene from ferrogabbro (Fig. 21b). Parental melt compositions similar to the calculated melts have been proposed by various workers for the upper gabbroic complexes of a variety of ophiolites and modern ocean

O P H I O L I T E S OF NORTHEAST ASIA

643

developed yet. The difference in REE totals and the character of chondrite-normalized REE patterns imply different origins for tonalites and plagiogranites of the two slices (Table 7, Fig. 20). Low REE contents in the tonalites and plagiogranites from the Khinantynup slice render them similar to their host coarse-grained gabbros and ferrogabbros (Figs 18 and 20), suggesting that these rocks might be fractionates of gabbroic magma, but positive Eu-anomaly in plagiogranites is at odds with this process. Unfortunately, we have no isotopic data on gabbros or plagiogranites to prove or disprove their kinship. Another possible explanation of different LREE enrichment degrees and REE totals of the Elgeminai and Khynantynup tonalites and plagiogranites is their origin from different mafic sources as a result of partial melting or different degrees of such melting.

Ophiolites of the Kuyul terrane Geological setting Fig. 21. Chondrite-normalized (Sun & McDonough 1989) REE patterns for hypothetical melts parental to clinopyroxenes from plutonic rocks of the Ganychalan terrane (rock names are given in Table 6). In (a), grey field indicates diabases and pillow basalts from the Elgeminai slices; in (b), grey field indicates melts in equilibrium with clinopyroxenes from the upper gabbro from ODP Site 894, Hess Deep (Gillis 1996).

crust assemblages (Pallister & Hopson 1981; Dick & Natland 1996; Kelemen et al 1997; Tiepolo et al. 1997). All these workers attributed the considerable REE enrichment of liquids parental to highlevel plutonic rocks to a strong, 40-50%, fractionation of primary N- and/or T-MORB-like melts of mantle derivation. Therefore, our data suggest that Ganychalan ophiolites are a geodynamically coherent assemblage generated in a MOR setting. Based on the presence of high-Ti hornblende, Khanchuk et al. (1992) proposed that the Ganychalan terrane ophiolite originated in an oceanic island setting. Nekrasov et al. (2001) used Ti, Zr, and Y variations to discriminate several plutonic series in the Ganychalan terrane. According to Nekrasov et al. (2001), these series may have resulted from stepwise melting of an enriched mantle source, suggestive of an oceanic plateau setting for ophiolite petrogenesis. This is not inconsistent with the above data from the ophiolitic reference section; however, identifying individual geochemical series would require a large number of samples from various portions of the Ganychalan terrane to be analysed. No petrogenetic model for the Khinantynup and Elgeminai tonalites and plagiogranites has been

The Kuyul thrust pile (Sokolov 1992) brings together three terranes, the Upupkin, Ainyn, and Kuyul terranes, that are thrust over each other successively from NW to SE (Fig. 15). The Upupkin terrane is composed of Devonian shallowwater organic limestone, Upper Carboniferous to Lower Permian clastic rocks, and Permian and Triassic tuffaceous epiclastic deposits. The Upper Jurassic to Lower Cretaceous strata comprise turbidites deposited in the forearc of the Uda-Murgal arc (Sokolov et al. 1999). In places, these turbidites contain equant bodies of serpentinized ultramafic material enclosed in sedimentary serpentinites. According to Sokolov et al. (2000), these rocks strongly resemble the Mariana arc serpentinite diapirs (Fryer 1992; Lagabrielle et al. 1992). The Ainyn terrane is composed of Upper Jurassic to Lower Cretaceous turbidites. The terrane is highly deformed and displays a broken formation, sedimentary melange composed of terrigenous material (type I and type IV melanges of Cowan (1985)) and numerous duplexes (Khudoley & Sokolov 1998). The depositional environment is interpreted as the landward slope of an accretionary prism (Khudoley & Sokolov 1998; Sokolov etal. 1999). The Kuyul terrane is traceable for 140km northeastward from the Mametchinsky Peninsula to the eastern side of the Talovka River in discrete outcrops 10-20 km wide. To the SE, the terrane is overlapped by Paleogene-Quaternary sedimentary rocks (Fig. 15b). The ophiolite is disrupted into a serpentinite melange containing blocks of ultramafic, gabbroic, and metamorphic (greenschists, blueschists and amphibolites) rocks, as well as

S. D. SOKOLOV ETAL.

644

basalts, cherts, limestones, and terrigenous rocks. Detailed mapping has identified a number of melange slices distinguished by clast lithologies (Sokolov et al. 1996). The characteristic lithologies and geodynamic interpretations of tectonostratigraphic assemblages are listed in Table 9. The most complete ophiolite fragment is the Gankuvayam sequence (Khanchuk et al. 1990), or slice (Sokolov et al. 1996). The Gankuvayam ophiolite is a dismembered suite, truncated at several levels by serpentinite melange zones and folded into a synform fold. Khanchuk et al. (1990) have reconstructed the following ophiolite sequence (Fig. 22): (1) serpentinized harzburgite overlain by dunite, 470 m thick; (2) a gabbro troctolite-wehrlite complex, 340 m thick; (3) plagiogranite, 40 m thick; (4) a dyke complex, compositionally differentiated from basaltic to dacitic, 400 m thick; (5) a pillow lava sequence ranging from basalt to dacite, 300 m thick. The age of ophiolite is estimated on the basis of Bathonian to Early Tithonian radiolarians in cherts from interpillow spaces within basalts of the differentiated series (Grigoriev et al. 1995). The Gankuvayam ophiolite has been interpreted to have originated from a variety of geodynamic settings, such as: (1) a Galapagos-type spreading centre (Khanchuk et al. 1990); (2) an intraoceanic island arc (Palandzhyan 1992); (3) a suprasubduction-zone setting (Sokolov et al. 1996). Resolving this issue called for follow-up studies on Gankuvayam ultramafic rocks from various localities in the Kuyul terrane, including the Mamet Peninsula. Results from this sampling exercise are presented below.

Petrography and geochemistry of the ultramafic rocks The ultramafic rocks are represented by tectonized residual peridotites and cumulative ultramafic

Fig. 22. Ophiolite stratigraphy from the Gankuvayam section.

rocks of the dunite-orthopyroxenite series. Overall, the massive ultramafic rocks are composed of serpentinized spinel peridotites (harzburgites and diopside harzburgites). Less frequently, dunites form lenses and bands in peridotites, clino- and orthopyroxenites, wehrlites, and websterites. The modal mineral composition of the ultramafic rocks and chemical composition of the main rock-forming minerals are shown in Figure 23. Most of the ultramafic rocks are significantly altered. Their serpentinization degree is 60-

Table 9. Tectonostratigraphic units of the Kuyul terrane Tectonic sheet

Rocks

Age

Geodynamic setting

Gankuvayam

Peridotite, cumulate complex, gabbro, plagiogranite, sheeted dykes, basaltandesite-dacite complex Basalt-limestone-chert, peridotite, cumulate, gabbro Basalt-chert Peridotite, cumulate complex, gabbro Peridotite, basalt Sandstone, siltstone, mudstone, chert, olistostrome Amphibolite, greenschist, blueschist

J 2 -J

Suprasubduction zone

P, Tr-J2, Jj

Oceanic

Tr2-J* Mz? ?

Oceanic Oceanic Within-plate Trench sediment

Veselaya Vstrechny Unnavayam Talovka Tylpyntyhlavaam Udachny

JS-KI

134, 122, 82 Ma (K/Ar method) Subduction zone

OPHIOLITES OF NORTHEAST ASIA

645

Fig. 23. Mineral composition features of ultramafic rocks from the Kuyul ophiolite terrane. Amph, amphibole; Cpx, clinopyroxene; Cr-Sp, chrome spinel; Ol, olivine; Opx, orthopyroxene; Spt, serpentine.

100%. Our study is therefore focused on Crspinel, whose chemistry best portrays the petrogenesis of the ultramafic rocks (e.g. Irvine 1967; Dick & Bullen 1984; Arai 1994), as it is resistant to serpentinization. Representative analyses of Cr-spinels from various ultramafic rocks from the Gankuvayam sequence are shown in Table 10. The Cr number, used as a measure of depletion in ultramafic rocks, is highly variable and largely overlaps the range typical of Cr-spinel compositions from oceanic and island-arc peridotites (Fig. 24). The Cr number of Cr-spinel correlates negatively with the Mg number, which is typical of residual mantle peridotites (Dick & Bullen 1984; Barnes & Roeder 2001). Cr-spinel compositions from the harzburgites indicate strong depletion and suggest a residual origin for these rocks (Dick & Bullen 1984; Ishii et al. 1992; Arai 1994) (Fig. 24). There are no systematic differences in Cr number between Cr-spinels from the dunites, orthopyroxenites, and chromitites. This invokes tight genetic links among these rocks, which could all be members of a dunite-ormopyroxenite-chromitite suite (e.g. Arai 1994). High Fe oxidation degrees in spinels from dunites and chromitites further support this conclusion. Major element contents of the harzburgites vary in a narrow range (Table 9) and correspond to non-isochemically recrystallized residual mantle-

derived peridotites (Bazylev et al. 1999). This is further supported by a good correlation (Fig. 24) between Cr2O3 and A12O3 of the host ultramafic rocks and their accessory Cr-spinels. The harzburgites are appreciably depleted in A12O3 and alkalis; remarkably, A12O3 and Cr2O3 contents, immobile under metamorphic conditions, are closely similar to one another, which is typical of residual SSZ-type peridotites (e.g. Bazylev et al. 1993, 1999; Rampone et al. 1995; Parkinson & Pearce 1998; Pearce et al. 2000; Takazawa et al. 2000). The harzburgites are slightly enriched in TiO2 as compared with typical restite peridotites (Barnes & Roeder 2001). Trace element contents of the ultramafic rocks are given in Table 11 and Figure 25. Cu, Zn, Sc, V and Ga show a negative correlation with MgO, and Ni, Cr, and Co a positive correlation with MgO. A similar relationship exists for most mantle-derived peridotites, in agreement with models for partial melting of initial fertile ultramafic rocks (e.g. Niu 1997). Peridotites of the Gankuvayam sequence are very strongly REE depleted (Fig. 25), harzburgites having no more than 0.2-0.5 times chondritic values (Fig. 25, Table 11). Chondrite-normalized REE patterns for the dominant ultramafic rocks, Di-free harzburgites, have asymmetric U-shaped forms ((La/Sm)N = 2.76 ± 0.54; (Sm/Yb)N = 1.15 ±0.19) (Fig. 25). Similar REE patterns and

Table 10. Representative microprobe analyses ofCr-spinels in ultramafic rocks from the Gankuvayam sequence Rock: Sample:

tz

552

504

546 c

r

0.00 0.03 0.05 Ti02 11.99 12.53 11.14 A1203 FeO 16.20 18.70 13.17 2.33 Fe2O3 1.34 0.67 0.33 MnO 0.33 0.24 11.45 MgO 9.90 13.26 NiO 0.00 0.05 0.15 ZnO 0.10 0.11 0.06 58.04 Cr2O3 57.50 60.93 100.44 Total 100.49 99.67 Cations recalculated per 32 oxygens Ti 0.000 0.007 0.010 Al 3.851 3.391 3.657 2+ Fe 4.078 2.843 3.506 3+ Fe 0.454 0.263 0.131 Mn 0.072 0.074 0.053 Mg 3.847 5.102 4.416 Ni 0.000 0.011 0.031 Zn 0.022 0.012 0.020 Cr 11.878 11.855 12.437 Cations 24.007 24.009 24.003 Crno. 0.76 0.75 0.79 Mg no. 0.49 0.64 0.56 Fe3 no. 0.03 0.02 0.01

Dunites

Di-bearing harzburgites

Di-free harzburgites

c

0.05 16.69 15.29 1.02 0.24 12.55 0.00 0.15 54.20 100.19 0.009 4.962 3.228 0.193 0.052 4.722 0.000 0.027 10.813 24.005 0.69 0.59 0.01

r

c

tz

583

553 r

0.00 16.37 15.02 1.35 0.27 12.44 0.01 0.02 53.55 99.03

0.03 30.71 14.60 1.04 0.24 14.29 0.12 0.09 38.41 99.53

0.05 31.02 13.99 0.78 0.18 14.54 0.12 0.15 37.64 98.47

0.04 31.37 14.10 0.55 0.19 14.63 0.11 0.14 37.94 99.07

0.000 4.926 3.207 0.259 0.058 4.733 0.003 0.005 10.810 24.000 0.69 0.60 0.02

0.005 8.583 2.896 0.186 0.049 5.051 0.022 0.017 7.201 24.008 0.46 0.64 0.01

0.008 8.720 2.790 0.141 0.037 5.169 0.023 0.026 7.097 24.011 0.45 0.65 0.01

0.007 8.758 2.793 0.098 0.037 5.164 0.021 0.025 7.107 24.010 0.45 0.65 0.01

c

0.07 12.76 19.29 0.98 0.36 9.48 0.03 0.23 57.10 100.30 0.014 3.937 4.222 0.194 0.080 3.698 0.006 0.045 11.816 24.011 0.75 0.47 0.01

tz

r

0.08 12.82 18.75 0.56 0.37 9.72 0.02 0.14 57.20 99.66

0.06 12.93 18.71 0.19 0.33 9.75 0.00 0.15 57.35 99.47

0.015 3.967 4.118 0.110 0.083 3.805 0.005 0.027 11.876 24.007 0.75 0.48 0.01

0.012 4.006 4.112 0.038 0.073 3.820 0.000 0.030 11.916 24.006 0.75 0.48 0.00

c

tz

r

0.03 11.38 15.76 1.28 0.30 11.65 0.04 0.11 59.62 100.17

0.07 11.39 16.23 0.59 0.33 11.41 0.08 0.13 60.25 100.48

0.04 11.11 17.03 0.02 0.31 10.79 0.05 0.15 60.62 100.12

0.007 3.484 3.422 0.251 0.067 4.510 0.008 0.021 12.237 24.064 0.78 0.57 0.02

0.013 3.482 3.520 0.115 0.073 4.410 0.017 0.025 12.355 24.009 0.78 0.56 0.01

0.009 3.425 3.725 0.004 0.069 4.204 0.010 0.029 12.533 24.008 0.79 0.53 0.00

c

0.24 15.04 11.71 3.27 0.22 14.73 0.13 0.02 54.33 99.69 0.045 4.462 2.464 0.620 0.047 5.526 0.026 0.004 10.812 24.006 0.71 0.69 0.04

Chromitites

Websterite

536

584

tz

r

0.21 15.06 11.76 2.91 0.22 14.72 0.16 0.07 54.84 99.95

0.22 15.31 11.73 3.17 0.25 14.7 0.15 0.04 54.11 99.68

0.040 4.457 2.471 0.549 0.046 5.512 0.033 0.013 10.889 24.010 0.71 0.69 0.03

0.041 4.537 2.466 0.600 0.053 5.511 0.031 0.008 10.761 24.008 0.70 0.69 0.04

c

tz

r

0.08 7.74 18.06 1.29 0.36 10.16 0.09 0.01 64.23 102.02

0.07 7.80 18.76 0.78 0.35 9.70 0.00 0.07 64.41 101.94

0.05 7.55 17.97 1.25 0.35 10.17 0.05 0.07 64.50 101.96

0.016 2.391 3.960 0.255 0.081 3.969 0.020 0.002 13.311 24.004 0.85 0.50 0.02

0.013 2.418 4.126 0.154 0.078 3.805 0.000 0.015 13.394 24.002 0.85 0.48 0.01

0.011 2.334 3.945 0.247 0.079 3.980 0.010 0.014 13.384 24.005 0.85 0.50 0.02

Samples 548 and 583 are from the Mametchinsky Peninsula, the rest are from the Gankunvayam sequence. Fe2Os and FeO calculated from spinel stoichiometry. c, grain cores; r, rims; tz, transitional zones. Cr number = Cr/(Cr + Al), Mg number = Mg/(Mg + Fe2+), Fe3 number = Fe3+/(Fe3+ + Cr + Al). Analysed on CAMEBAX microprobe in the Institute of Volcanology, PetropavlovskKamchatsky (analyst V M. Chubarov).

OPHIOLITES OF NORTHEAST ASIA

Fig. 24. Compositions of rock-forming and accessory Cr-spinels from ultramafic rocks of the Gankuvayam sequence. Fields of Cr-spinel compositions in abyssal peridotites (dotted line), SSZ-type harzburgites (dashed line), and SSZ-type dunites (continuous line) are shown. Compositional fields are after Dick & Bullen (1984) and Ishiietal. (1991).

extremely low REE contents are inherent in residual spinel peridotites of typical SSZ ophiolite complexes (e.g. Parkinson & Pearce 1998; Pearce et al 2000). Hence, the ultramafic compositions detailed above suggest an SSZ setting for their genesis. The initial ultramafic rocks underwent extensive partial melting and reaction with primitive islandarc melts, resulting in cumulate veins and chromite mineralization in the ultramafic rocks.

Petrography and geochemistry of the plagiogranites Plagiogranites form a separate slice between the gabbroic rocks and sheeted dyke complex in the Gankuvayam thrust sheet (Fig. 22). Their lower contact with gabbro is sharp and faulted, whereas the upper part of the plagiogranite slice contains fragments of sheeted dykes. The plagiogranites are strongly tectonized and brecciated along both contacts.

647

The plagiogranites are divided into plagiogranites proper (65-75 wt% SiO2) and quartz diorites and tonalites (62-65 wt% SiO2). The latter are the least differentiated members, their volume being negligible. The plagiogranites have subhedral granular to granophyric texture. They contain 30-40% modal quartz, 40-60% plagioclase, and 5-15% amphibole. Plagioclase, An25-35, is euhedral (0.31.2 mm across), partly saussuritized, and often zoned. Quartz is anhedral. Amphibole forms green euhedral prisms. Accessory minerals include zircon, apatite, sphene, and magnetite. Secondary alteration minerals are epidote, chlorite, albite, and prehnite. The tonalites and quartz diorites contain the same minerals but have lower percentages of quartz (20-25 modal %), higher percentages of plagioclase (55-70%) and amphibole (15-25%), and more calcic plagioclase. On the Ab-An-Or diagram, the felsic rocks plot with tonalites and trondhjemites. Their low K2O (0.1-0.8%) (Luchitskaya 1996) places them with oceanic plagiogranites (Coleman & Donate 1979). On Harker diagrams, plagiogranites, intermediate to felsic dykes, and gabbros from the upper part of gabbroic section form a coherent fractionation trend (Luchitskaya 1996). Chondrite-normalized REE patterns of plagiogranites are slightly LREE enriched, with an almost horizontal HREE portion (Lan/Ybn = 0.8-1.37) and a negative Eu anomaly (Eun/Eu* = 0.45-0.90). The similarity of these rocks to those of the Samail ophiolite is evident (Fig. 26). The similarity of REE patterns for plagiogranites, dacites, andesites, and basalts from the sheeted dyke complex suggests that the rocks are cogenetic (Fig. 27). On spidergrams of Pearce et al. (1984), the plagiogranites are slightly enriched in Rb and depleted in HFSE with respect to ORG (Luchitskaya 1996). Their ORG-normalized patterns are similar to those of ocean-ridge granites as given by Pearce et al (1984) (Troodos plagiogranites) and volcanic-arc granites (granites of Lower Intrusive complex, Oman (Luchitskaya 1996)). On an Nb-Y diagram (Pearce et al. 1984), plagiogranites plot in the ORG field (Luchitskaya 1996).

Origin of the ophiolite The harzburgites of the Gankuvayam unit are restites, and the lower gabbro-wehrlite complex represents basal cumulates formed in a magma chamber at the crust-mantle interface, with gabbronorites and plagiogranites forming coeval intrusions. The initial ultramafic rocks underwent extensive partial melting and reaction with primi-

Table 11. Major (wt%) and trace (ppm) element contents ofultramafic rocks of the Gankuvayam sequenc Sample: Si02 Ti02 A1203 Cr203 Fe203 FeO MnO MgO CaO Na20 K2O H2CT LOI Total V Cr Ni Rb Sr Y Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Th

571

504

546

581

515

583/1

548

552

547

583

553

36.96 0.04 1.12 0.31 4.11 3.75 0.12 39.59 0.90 0.05 0.01 0.83 11.56 99.35 26.0 2104 1445 0.25 25 1.4 5.1 1.2 12.5 0.06 0.09 0.01 0.03 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.10

37.64 0.06 1.92 0.27 3.11 4.45 0.11 37.82 1.01 0.09 0.03 0.70 12.05 99.26 17.4 1842 1532 0.84 29 0.9 2.7 0.8 19.5 0.04 0.06 0.01 0.03 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.12

36.56 0.03 1.03 0.24 3.95 3.05 0.10 38.22 0.83 0.04 0.03 0.91 13.96 98.95 4.3 1632 1985 0.35 74 0.8 7.1 0.7 24.1 0.06 0.12 0.02 0.03 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.41

37.50 0.04 1.42 0.32 3.77 3.43 0.11 37.50 0.90 0.06 0.03 1.47 12.55 99.10 21.5 2204 1025 0.61 51 1.0 1.5 2.6 14.2 0.04 0.10 0.02 0.03 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.10

38.44 0.04 2.24 0.23 3.99 1.22 0.12 37.09 0.96 0.04 0.01 1.45 13.40 99.23 14.2 1586 893 0.68 26 3.1 3.2 3.7 16.2 0.05 0.08 0.02 0.03 0.01 0.00 0.02 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.24

36.72 0.04 0.81 0.25 3.86 3.13 0.11 37.74 0.71 0.04 0.01 1.61 14.07 99.10

38.14 0.06 1.72 0.21 3.28 4.08 0.11 36.85 1.35 0.06 0.03 1.06 12.55 99.50 24.5 1429 787 0.25 84 4.5 4.1 4.2 31.2 0.10 0.18 0.03 0.04 0.01 0.00 0.01 0.00 0.02 0.01 0.01 0.00 0.01 0.00 0.14

36.08 0.05 2.34 0.27 5.84 3.41 0.11 36.06 1.23 0.10 0.02 1.20 14.20 100.91 31.2 1867 950 0.32 95 3.2 12.1 0.5 24.1 0.06 0.16 0.03 0.05 0.01 0.01 0.01 0.00 0.03 0.01 0.01 0.00 0.01 0.00 0.32

31.20 0.02 1.03 0.37 4.82 3.59 0.11 41.37 0.45 0.05 0.02 1.26 15.26 99.55 17.0 2531 1565 0.14 24 0.5 8.2 2.1 8.2 0.16 0.26 0.05 0.05 0.02 0.01 0.01 0.01 0.03 0.01 0.02 0.01 0.02 0.01 0.14

32.56 0.01 1.23 0.31 4.11 2.70 0.10 41.13 0.56 0.07 0.01 1.81 14.67 99.27 12.6 2118 1485 0.41 18 0.9 6.3 2.6 15.4 0.13 0.21 0.06 0.05 0.03 0.01 0.01 0.01 0.04 0.01 0.02 0.01 0.21 0.01 0.41

31.98 0.01 1.12 0.34 6.33 0.91 0.11 41.29 0.41 0.03 0.01 1.06 15.61 99.21 13.5 2345 1782 0.39 16 1.0 14.1 0.6 22.0 0.10 0.21 0.05 0.04 0.02 0.01 0.00 0.01 0.04 0.01 0.02 0.01 0.17 0.01 0.42

571-583/1, Di-free harzburgites; 548 and 552, Di-bearing harzburgites; 547-553, dunites. Major elements analysed by XRF in the Karpinsky Geological Institute, St. Petersburg; trace elements by ICPMS in the Institute of Geochemistry, Irkutsk (G. A. Samdimirova, analyst).

OPHIOLITES OF NORTHEAST ASIA

649

Fig. 25. Chondrite-normalized (Anders & Grevesse 1989) REE patterns for the Gankuvayam sequence ultramafic rocks.

tive island-arc melts, resulting in cumulate veins and chromite mineralization in the ultramafic rocks. The plagiogranites may have formed by fractional crystallization of a basic magma combined

with some filter-pressing (squeezing of remaining interstitial acid melt) mechanism. They are believed to have formed in the upper part of the ophiolite crustal section above a subduction zone (for details, see Luchitskaya (1996, 2001));

650

S. D. SOKOLOV £T,4L.

Fig. 26. Chondrite-normalized (Sun & McDonough 1989) REE patterns for plagiogranites from the Kuyul terrane, Troodos ophiolites (Kay & Senechal 1976) and Samail ophiolites (Pallister & Knight 1981; Coleman & Donate 1979).

the Gankuvayam ophiolite originated in an SSZ setting. Study of volcanic (Gankuvayam type) and chert-volcanic (Kingiveyem type) assemblages indicates that the Kuyul terrane comprises tectonically juxtaposed assemblages of distinctive ages and origins (Grigoriev et al. 1995) (Table 9). Gankuvayam extrusive rocks and dykes are composed of calc-alkaline basalt to dacite (Grigoriev et al. 1995; Sokolov et al. 1996). The Kingiveyem type of stratigraphy (Table 9, Veselaya and Vstrechny slices) includes chertbasalt-limestone and chert-basalt associations of Late Triassic to early Tithonian age. Chemically, the basalts are N-MORB-like (Khanchuk et al. 1990; Grigoriev et al 1995). In addition, blocks of WPB, in association with limestones of Permian age (Grigoriev et al. 1995), occur in the serpentinite melange. N-MORB and WPB hosted in melange and tectonic juxtaposition of geodynamically contrasting assemblages (Grigoriev et al. 1995; Sokolov et al. 1996) suggest that the Kuyul melange may yet yield less depleted ultramafic rocks and gabbros of oceanic provenance.

Ophiolites of the Ust-Belaya segment Tectonic setting

Fig. 27. Chondrite-normalized (Sun & McDonough 1989) REE patterns for plagiogranites, and basalts, andesites and dacites from the sheeted dyke complex from the Kuyul ophiolite terrane.

It thus follows that the additional studies in ultramafic compositions and geochemical data on the plagiogranites further corroborate the inference (Palandzhyan 1992; Sokolov et al 1996) that

The Ust-Belaya terrane consists of a complex package of allochthonous units, thrust over the Middle Jurassic to Valanginian terrigenous, chert, and volcanic deposits of the Algan terrane (Palandzhyan & Dmitrenko 1996; Nekrasov et al. 2001) (Fig. 28a). The base of the overlap assemblage is composed of late Albian to early Senonian marine terrigenous deposits (Markov et al. 1982; Filatova 1988; Sokolov 1992), which represent sediment fill of a forearc basin in front of the Okhotsk-Chukotka continental-margin volcanic belt. Eocene to Oligocene strata comprise on-land volcanic deposits of the western KamchatkaKoryak continental-margin volcanic belt, and its coeval terrigenous deposits. The allochthonous package (Fig. 28b) is divided into three large nappes (Palandzhyan 2000). The Lower, or Utyosiki, thrust sheet incorporates slices of island-arc volcanic and clastic deposits of postulated Late Palaeozoic and Mesozoic age. Foliated serpentinites occur between some slices. The thrust unit has a total thickness of 1.6 km. The Middle, Otrozhnaya, nappe consists of two thrust units. The lower unit consists of: (1) a Middle Palaeozoic ophiolite assemblage, comprising heavily deformed serpentinized harzburgites, Iherzolites, and dunites overlain by a deformed slice c. 1 km thick of amphibolized gabbro; (2) a

OPHIOLITES OF NORTHEAST ASIA

651

Fig. 28. (a) Geological map of the Ust-Belaya terrane (after S. A. Palandzhyan); (b) succession of the nappe units. In (a): A, mountain peak (E, Eldernyr; Ot, Otrozhnaya; Zm, Zmeyevik; Pk, Porfiritovyi Kamen; Sp, Sprut). (b) For symbols, see (a).

652

S. D. SOKOLOV ETAL.

sequence, some 0.7 km thick, of MORB-like basaltic rocks; (3) cherty and tuffaceous clastic rocks with Mid- to Late Devonian faunas; (4) Early Carboniferous terrigenous rocks with serpentinite, diabase, and chert clasts and Cr-spinel grains. The gabbros and basalts are cut by diabase and porphyritic microplagiogranite dykes. The microplagiogranite dykes have whole-rock K-Ar ages of 218-178 Ma (Palandzhyan 1997). The second thrust unit, Udachny, consists of c. 1.1 km thick Upper Jurassic to Valanginian flysch, interpreted to have been deposited in a forearc basin in front of the Uda-Murgal arc, and containing fragments of tuffaceous terrigenous material of Mid-Jurassic age, and carbonate and terrigenous sequences of Mid-Late Devonian and Early Carboniferous age. The upper, Ust-Belaya nappe, over 5 km in thickness, is mainly composed of peridotites and gabbros. At the base of the nappe is a serpentinite melange up to 0.5 km thick. Blocks in the melange are composed not only of the overlying rocks (Iherzolite, harzburgite, cumulate ultrabasic rocks, gabbro, basalt, siliceous-terrigenous rocks, organic limestone of Mid- and Late Devonian ages), but also of rocks lacking from all three nappes mentioned above, such as depleted harzburgite, high-Cr chromitite, schist, garnet amphibolite, and plagiogranite.

Petrography and geochemistry of the ophiolite The peridotites show a variety of metamorphic fabrics, including protogranular, which is commonly preserved in Iherzolites. Porphyroclastic and mosaic fabrics and mylonites are typical in the shear zones and at the base of the mantle suite. The peridotites are abundant in antigorite, a distinctive feature of the ultramafic rocks of the Ust-Belaya terrane, whereas in the other ultramafic massifs lizardite and chrysotile are abundant. Stratigraphically higher, in the eastern part of Ust-Belaya unit, Iherzolite gives way to Cpxbearing harzburgites and dunites and then to a thinly banded (1-3 cm) sequence of alternating dunite, cortlandite (two-pyroxene hornblende peridotite), wehrlite, pyroxenite, and hornblendite, intruded by microgabbroic dykes (Table 12, Fig. 29a). According to Nekrasov et aL (2001), the least depleted Iherzolites of the Ust-Belaya massif are compositionally close to pyrolite. Major element compositions of coexisting minerals from the Iherzolites and harzburgites are given in Fig. 30 and Table 13. Although the compositional range is wide, most peridotites are slightly to moderately depleted, as indicated by pyroxene and Cr-spinel

OPHIOLITES OF NORTHEAST ASIA

653

Fig. 29. Plots for mafic rocks showing (a) TiO2 v. MgO/(MgO + FeO*) and (b) TiO2 v. FeO*/MgO, wt% in whole rock. Compositional fields in (a) are after Miyashiro & Shido (1980); those in (b) are after (Glassely 1974).

Fig. 30. Plot of Cr/(Cr + Al) in Cr-spinel v. A12O3 in Opx and Na/Cr (crystallochemical formula units) in Cpx for metamorphic peridotites. Fields of mantle peridotite compositions from different geodynamic settings, after Kornprobst et al. (1981), Bonatti & Michael (1989) and Palandzhyan (1992). 1, Subcontinental; 2, passive margin; 3, mid-ocean ridge (dotted contours delineate extremely slow-spreading MOR peridotites), 4, suprasubduction zone.

Table 13. Major element contents (wt%) of coexisting minerals from ophiolitic peridotites of the Ust-Belaya terrane 520/1, LZ

532/L, LZ

Sample: Mineral: n:

Ol 4

Spl 7

Si02 TiO2 A1203 Cr203 Fe2O3 FeO* MnO MgO NiO CaO Na20 Total Cr Al Na Cr* Ft

40.90 n.d. 9.15 0.17 49.38 0.37 _ n.d. 99.97 _ _ 9.39

55.43 n.d. 0.05 0.10 4.05 51.48 0.53 15.98 n.d. 1.87 6.28 11.24 0.10 0.13 33.37 18.63 0.12 0.29 0.82 n.d. n.d. 99.72 100.76 0.014 0.338 0.164 1.621 _ _ 7.87 17.23 9.55 25.28

Opx 2

Cpx 3

52.69 0.09 4.45 0.83 n.d. 2.19 0.10 16.31 0.07 22.16 0.76 99.66 0.024 0.191 0.054 11.10 7.00

Ol 3

40.48 0.02 0.03 n.d. 9.00 0.11 50.60 0.40 0.01 n.d. 100.65 _ _ _ 9.06

520/2 , LZ

523/1 , LZ Cpx 3

Spl 5

Opx 2

n.d. 0.04 52.69 14.71 1.85 11.02 0.12 18.84 0.28 n.d. n.d. 99.55 0.309 1.652 _ 15.77 24.71

52.89 54.79 0.12 0.07 4.87 3.67 0.81 0.31 n.d. n.d. 2.07 6.37 0.08 0.15 15.75 33.90 0.06 0.02 22.94 0.50 0.95 0.03 99.82 100.54 0.023 0.009 0.207 0.150 _ 0.066 11.00 5.66 6.86 9.54

Ol 5

Spl 6

Opx 5

54.83 n.d. 40.16 0.04 0.02 0.01 3.97 51.10 0.52 0.04 16.97 n.d. 1.85 n.d. 5.83 8.56 11.19 0.13 0.11 0.15 34.19 50.58 18.72 0.05 0.24 0.38 0.48 n.d. 0.01 0.02 n.d. n.d. 99.89 100.20 100.06 _ 0.014 0.358 0.162 1.605 _ _ _ _ 8.79 16.55 8.73 8.67 25.11

Cpx 5

51.94 0.12 5.09 1.06 n.d. 1.97 0.10 15.75 0.04 22.50 1.11 99.69 0.031 0.219 0.078 12.25 6.56

Ol 3

40.51 0.01 0.03 n.d. 8.75 0.16 50.93 0.36 0.01 n.d. 100.76 _ _ _ 8.81

Spl 6

Opx 4

584/1, DHZ

454/3, LZ Cpx 3

53.64 n.d. 55.06 0.12 0.03 0.05 5.06 3.72 48.00 1.28 0.64 19.63 n.d. n.d. 2.18 1.97 5.59 11.61 0.11 0.14 0.14 15.16 34.37 18.07 0.02 0.11 0.23 22.01 n.d. 0.55 1.37 n.d. 0.01 99.91 100.23 100.75 0.036 0.421 0.017 0.214 0.143 1.533 _ _ 0.095 14.40 10.63 21.53 6.78 26.49 8.36

Ol 6

Spl 6

39.79 n.d. 0.01 0.01 53.74 13.66 n.d. 1.96 11.86 9.06 0.11 0.13 50.23 18.53 0.29 0.27 0.01 n.d. n.d. n.d. 99.50 100.16 _ 0.286 1.674 _ _ _ 14.57 9.18 26.36

Opx 7

Cpx 4

Ol 2

54.72 0.02 3.57 0.36 n.d. 6.31 0.13 34.37 0.05 0.29 0.04 99.87 0.010 0.146 _ 6.36 9.34

52.73 0.11 4.53 0.81 n.d. 2.15 0.08 16.17 0.02 22.62 0.69 99.93 0.023 0.194 0.049 10.70 6.94

40.98 _

Spl 5

Cpx 3

n.d. 53.18 0.02 0.02 3.34 41.14 0.08 26.84 0.95 n.d. 2.66 n.d. 9.13 13.12 1.79 0.16 0.10 0.16 50.15 16.44 16.55 0.20 0.35 0.01 _ n.d. 24.20 0.29 n.d. n.d. 100.85 100.59 100.45 _ 0.027 0.592 0.142 1.352 _ _ 0.020 _ 30.44 15.98 5.72 9.27 30.93

LZ, Iherzolite; DHZ, diopside harzburgite; Ol, olivine; Spl, chromian spinel; Opx, orthopyroxene; Cpx, clinopyroxene; n, number of analysed grains (mean values are shown for several analyses for each mineral); n.d., not determined. FeO*, total Fe as FeO. In chromian spinel, FeO and Fe2O3 were calculated using the compound electroneutrality principle. 454/3, 584/1, Ust-Belaya massif; others, Eldernyr massif. All analyses from cores of mineral grains. *Cr= 100XCr/(Cr + Al). t p = 100XFe/(Fe + Mg).

OPHIOLITES OF NORTHEAST ASIA compositions being similar to peridotites of modern passive margins and ultra slow-spreading centres. The least depleted Iherzolites of the UstBelaya nappe, based on their mineral composition, are similar to subcontinental Iherzolites (Fig. 30). Nekrasov et al (2001) showed that the UstBelaya Iherzolites have nearly chondritic REE contents with a slight negative Eu anomaly. These data suggest that the rocks of the Iherzolite unit of the Ust-Belaya terrane are oceanic lithospheric fragments of the marginal part of a large oceanic basin (Palandzhyan 1992). Nekrasov et al. (2001), on the other hand, proposed that the Iherzolites resemble xenoliths of mantle rocks of the Ontong-Java Plateau. The Iherzolite-gabbro contact, exposed over a considerable length, is marked by a thin (0.10.2 km) transition zone, composed of dunite, wehrlite, and Pi-bearing dunite. The gabbroic complex is composed chiefly of banded olivine gabbro and troctolite with lens-like bodies of Pibearing ultramafic rocks in the lower part. Most of the banded gabbro is metamorphosed to amphibolite. Petrographic observations in gabbros with primary banded fabrics surviving among gabbroamphibolites show that plagioclase crystallized before clinopyroxene and the rocks lack orthopyroxene, which places the cumulates with the A-l type as described by Ohnenstetter (1985). These characteristics suggest that the ophiolite may have originated in a mid-ocean ridge or in a continental-margin basin, rather than in an SSZ setting. Basalts in the Otrozhnaya nappe are similar to MORB-type tholeiites both in terms of major oxides (Fig. 29b) and trace elements, particularly REE, Zr, Y, and Nb (for details, see Nekrasov et al. 2001).

Origin of the ophiolite Thrust faulting in the Ust-Belaya terrane has juxtaposed two ophiolite associations. The Lower to Middle(?) Devonian ophiolites, which make up the Ust-Belaya and Otrozhnaya nappes, have slightly depleted peridotites, primitive crustal magmatic lithologies, and MORB-like volcanic rocks. Major and trace element composition of restite peridotites and gabbros confirm oceanic affinity for these ophiolites (Palandzhyan 1992; Nekrasov et al. 2001). The serpentinite melange contains fragments of an SSZ-type ophiolite association that includes strongly depleted peridotite, high-Cr chromitites, and differentiated volcanic rocks. These ophiolites are tentatively dated as Late Palaeozoic to Early Mesozoic based on whole-rock K-Ar ages: 262 Ma from a dacite, 164 and 178 Ma from

655

amphibolites, and 172, 186, and 218-178 Ma from plagiogranites (Palandzhyan 1997).

Discussion The above summary illustrates the diversity of structural settings, ages, compositions, and origins of ophiolites in the West Koryak fold system (Fig. 31, Table 14). They are all parts of a continentalmargin assemblage that was added to the North Asian continent at the end of the Early Cretaceous (Sokolov 1992; Parfenov et al. 1993; Khudoley & Sokolov 1998). The ophiolites are all deformed and unconformably overlain by upper Albian strata. The ophiolites come in two age groups, Palaeozoic and Mesozoic. Palaeozoic ophiolites of the Ganychalan and Ust-Belaya terranes are oceanic in type. They can be viewed as being fragments of the Panthalassa Ocean. Counterparts to Ganychalan ophiolites may be those of the Livengood terrane, Central Alaska. Exotic, possibly North American, provenance has been proposed for the Ordovician deposits overlying ophiolites of the Ganychalan terrane (Khanchuk et al. 1992; Sokolov et al. 1997). Other workers (Nekrasov et al. 2001), by contrast, believe that the Ganychalan and Ust-Belaya ophiolites were formed in an oceanic plateau near the Asian continent. Correlating the Ust-Belaya ophiolites will be possible only after their age has been properly constrained. Currently, they are dated to the Devonian based on the Mid-Late Devonian age of their overlapping sediments (ui in Fig. 31). However, radiometric datings from the peridotites and gabbros of their structurally overlying MORB-like basalts are lacking. Not inconceivably, the ophiolites are older, Early Palaeozoic in age, representing fragments of the same oceanic plate as the Ganychalan ophiolites. Provided that their Devonian age is validated, they could be correlated with the Brooks Range ophiolites of North Alaska, which are interpreted as fragments of the Angachuyam Ocean (Nokleberg et al. 1994, 2001; Wirthetal. 1994). The serpentinite melange of the Ust-Belaya nappe (u2 in Fig. 31) hosts occasional blocks of island-arc-derived lithologies: strongly depleted peridotites, plagiogranites, and differentiated volcanic rocks. The previous K/Ar measurements suggest Late Palaeozoic and Early Mesozoic age for the island-arc assemblage. This is in keeping with the reconstructions for the Koni-Taigonos island arc, which once marked the convergent boundary between the North Asian continent and NW Pacific (Parfenov 1984; Sokolov 1992; Sokolov et al. 1999; Parfenov et al. 1993). However,

656

S. D. SOKOLOV ETAL.

Fig. 31. Temporal and spatial distribution of the ophiolites. ui, Otrozhnaya nappe; u2, Ust-Belaya nappe; ki, Unnavayam unit; k2, Gankuvayam unit; ei, Northern unit; 62, Southern unit.

the K/Ar age determination ought to be validated by new measurements using other isotope systems. Mesozoic ophiolites are dominated by those of SSZ origin (Fig. 31, Table 14). The most complete fragment of an ophiolite suite has been reported from the Gankuvayam unit (k2 in Fig. 31) of the Kuyul terrane (Khanchuk et al. 1990; Sokolov et al. 1996) and in the Yelistratov Peninsula (e2 in Fig. 31), where relics survive of a magma chamber composed of ultramafic-mafic cumulates and gabbro (Ishiwatari et al. 1998; Saito et al. 1999). In addition, among Mesozoic ophiolites one finds rock fragments of oceanic provenance. These include the Unnavayam unit (ki in Fig. 31) of the Kuyul terrane (Grigoriev et al. 1995; Sokolov et al. 1996) and the Southern ultramafic body

from the Yelistratov ophiolite (Ishiwatari et al. 1998; Saito et al. 1999). Previously, serpentinite melanges occurring in the Cape Povorotny accretionary prism were believed to contain both oceanic Iherzolites and SSZ harzburgites (Palandzhyan & Dmitrenko 1999; Sokolov et al. 1999), but more recent geochemical data (Bazylev et al. 2001), including those presented in this paper, suggest an SSZ origin for these rocks. On the other hand, because the accretionary pile contains some basalt-chert associations of oceanic origin and ensimatic volcanic rocks, as well as oceanic and SSZ gabbro, it is still possible that more detailed study may detect oceanic ultramafic rocks there as well. The fact that SSZ assemblages are lacking in Early Palaeozoic ophiolites and, by contrast, are

Table 14. Age, composition, structural location and geodynamic setting ofNE Asia ophiolites Ophiolites

Composition (from bottom to top)

Age

Structural location

Geodynamic setting

1 . Ganychalan terrane

Mrachnaya slice: serpentinite melange with blocks of plagioclase wehrlite, pyroxenite, troctolite, olivine gabbro Khynantynup slice: layered gabbro, isotropic gabbro, gabbronorite, ferrogabbro, gabbro-diabase, diabase, plagiogranite Ilgeminai slice: basalt, tuff, chert Otrozhnaya nappe: serpentinized harzburgite, Iherzolite, dunite, amphibolitized gabbro; MORB-like basaltic rocks; cherty and tuffaceous clastic rocks Ust-Belaya, nappe: serpentinite melange; peridotite and gabbro

Early Palaeozoic

Forearc basement

Fragment of oceanic crust

Late Palaeozoic

Forearc basement

Fragment of oceanic crust

Accretionary prism

Suprasubduction

Forearc basement

Suprasubduction

Accretionary prism

Oceanic (Unnavayam unit) and Suprasubduction (Gankuvayam unit)

2. Ust-Belaya terrane

3. Beregovoi terrane 3.1. Cape Povorotny 3.2. Elistratov Peninsula 4. Kuyul terrane

Serpentinite melange with blocks of ultramafic rocks (Iherzolite, Mesozoic harzburgite), gabbro, sheeted dykes, plagiogranite, amphibolite, greenschist, island-arc volcanic and sedimentary rocks, oceanic basalt and chert Northern unit: serpentinite melange with blocks of harzburgite, dunite, Mesozoic sheeted dykes, metabasalts, radiolarite; Southern wmY:harzburgite, cumulate complex, gabbro, sheeted dyke complex differentiated from basalt to dacite Unnavayam unit: serpentinite melange with blocks of harzburgite, Mesozoic cumulates, basalt and chert Gankuvayam unit: harzburgite, dunite, cumulate complex (gabbro, wehrlite, troctolite), isotropic gabbro; plagiogranite, differentiated complexes from basalt to dacite: sheeted dykes and pillow lavas

658

S. D. SOKOLOV ETAL.

widespread in Mesozoic ophiolites might be critical to the general evolution of the Palaeo-Pacific ocean. This inference, however, should be tested against a more representative dataset covering other Circum-Pacific regions as well. Ophiolites and their associated metamorphic, volcanic, and sedimentary rocks of PalaeozoicEarly Cretaceous age (including early Albian) show evidence of at least three deformation phases: DI, D2, and D3 (Khudoley & Sokolov 1998). Fabrics and structures related to Dj are clearly pronounced in the Ganychalan terrane. Khudoley & Sokolov (1998) proposed an Early Carboniferous age for the DI event. Rb/Sr ages obtained from Ganychalan greenschists (Vinogradov et aL 1995) imply that D! deformations were accompanied by some greenschist metamorphism at 321 ±5 Ma. We interpret this deformational event as resulting from amalgamation of previously separated blocks, now represented by the Ilpenei, Mrachnaya, Khinantynup, and Elgeminai thrust sheets in the unique Ganychalan composite terrane. The D2 deformation event is well documented by palaeontological data (Khudoley & Sokolov 1998). The youngest rocks affected by D2 folds and faults contain early Albian faunas, whereas the oldest unconformably overlying undeformed units contain late Albian faunas (Sokolov 1992). Based on regional geological considerations and terrane analysis, the latest Jurassic-Early Cretaceous tectonism is usually interpreted as the main stage of accretion of Pacific-related terranes to the Asian continental margin (Sokolov 1992; Parfenov et aL 1993; Nokleberg et al 1994). D3 faults cross-cut all terranes and their overlap sequence, including Paleogene rocks (Bondarenko & Sokolov 1996; Khudoley & Sokolov 1998). The D3 deformation event is interpreted as being related to major sinistral strike-slip displacement. These deformation events reflect various phases of evolution of the continental margin of NE Asia and can be readily correlated with principal tectonic events in the north Circum-Pacific region. Thus, DI stage and island-arc volcanism in the Taigonos and Penzhina segments were related to evolution of the Palaeozoic convergent boundary of Asia and the NW Pacific (Zonenshain et al. 1990; Sokolov 1992; Parfenov et al. 1993). Strain analysis shows that the direction of maximum extension is parallel to the regional trend (Khudoley & Sokolov 1998) and is typical of accretionary belts of the Pacific Rim (Toriumi 1985). The event produced D2 structures throughout all the terranes and correlates well with synchronous tectonic processes in southern Alaska (Nokleberg et al. 1994). The D3 event is approximately synchronous with Late Cretaceous to Eocene dextral strike-slip displace-

ments and transpression in Alaska and the northern Cordillera and is typical of post-accretionary tectonics (Coney et al. 1980; Gabrielse et al. 1991). The degree of disruption of the ophiolites differs from locality to locality. Deformation is strongest and serpentinite melanges are present in Cape Povorotny and Kuyul terrane ophiolites. The latter retains only one fragment (Gankuvayam unit), interpretable as a dismembered ophiolite. Ust-Belaya terrane and Yelistratova Peninsula ophiolites, alongside serpentinite melange, preserve large intact fragments of gabbro, cumulates, and sheeted dykes. Ganychalan terrane ophiolites are dismembered, with serpentinite melange zones being here encountered only locally, mainly in the base of the ophiolitic allochthon. Therefore, the degree of disruption is least, strangely enough, in the older, Early Palaeozoic ophiolites of the Ganychalan terrane. This may be due to three circumstances: (1) the mode of ophiolite incorporation into continental margin; (2) dissimilarity of formative geodynamic settings of the ophiolites; (3) post-accretionary history of the accreted terranes. The first factor is favoured by the fact that overlap sediments are lacking from Mesozoic and available in Palaeozoic ophiolites. Ganychalan and Ust-Belaya ophiolites have preserved depositional contacts with strata of terrigenous provenance (Fig. 31). These were accumulated on top of the ophiolites as the oceanic plate converged with and resided near the continental margin. The age of sedimentary cover constrains the lower time limit for ophiolite accretion. In Mesozoic ophiolites, stratigraphic contacts survive only in post-accretionary deposits. Also, Cape Povorotny and the Kuyul terrane show extensive strike-slip motions (Bondarenko & Sokolov 1996; Khudoley & Sokolov 1998), probably suggestive of oblique subduction during the docking of Mesozoic ophiolite assemblages. Palaeozoic accreted terranes are made chiefly of oceanic ophiolites, and Mesozoic terranes of SSZ ophiolites, which lends support to the second factor. The third factor being in operation is evidenced by strike-slip offsets related to D3 deformation (see below). Not only do the ophiolites have different origins, but they also have different accretionary histories. Palaeozoic ophiolites of the Ganychalan and Ust-Belaya accreted terranes docked onto the Koni-Taigonos island arc, which marked the convergent boundary of the Asian continent and NW Pacific in Late Palaeozoic to Early Mesozoic times. The SSZ ophiolites from the serpentinite melange of the Ust-Belaya nappe may be relics of the Koni-Taigonos island arc, whose fragments have recently been reported from the Penzhina District (Sokolov et al. 1999) and from the

OPHIOLITES OF NORTHEAST ASIA Pekulney Range (Morozov 2001). No fragments of an ancient accretionary prism have been identified to date. They may have survived and might still be detected among Palaeozoic deposits of the Beregovoi terrane (Carboniferous flysch) in the Taigonos Peninsula and in the Upupkin terrane (Permian and Triassic tuffaceous-terrigenous and clastic deposits). At the same time, subductionrelated glaucophane-greenschist metamorphic rocks of Carboniferous age are well developed in the Ganychalan terrane (Ilpenei unit) of Penzhina District (Dobretsov 1974; Silantyev et al. 1994). The exact timing of accretion and original location of the Ganychalan and Ust-Belaya terranes remain unknown. In this context, Ganychalan ophiolites should be classed with 'suspect terranes' as described by Coney et al. (1980). The Carboniferous and Permian deposits within the Koni-Taigonos arc and in terrigenous sequences associated with the ophiolites were clearly formed at high latitudes, because they contain boreal faunas and Angaran floras (Sokolov 1992). In addition, Carboniferous strata in the Ust-Belaya terrane contain ophiolitic detritus (Markov et al. 1982; Palandzhyan 2000), suggesting that this terrane was previously amalgamated with or even accreted onto the Eurasian plate. No reliable age data for the Yelistratov Peninsula ophiolites are available. They are overlain by Berriasian strata, deposited in the forearc of the Uda-Murgal island arc. Serpentinite melange contains occasional blocks of basalt and radiolarite of Mid- to Late Jurassic age. Conceivably, they could be Mesozoic in age, like the Cape Povorotny and Kuyul ophiolites. However, the Yelistratov ophiolite was accreted to the Uda-Murgal arc somewhat earlier than the Cape Povorotny ophiolite. The latter is overlain by Valanginian-Hauterivian strata of the Vitayetglia unit (Fig. 5), which were deposited in a forearc basin associated with the Uda-Murgal island arc and contain ophiolitic detritus (Sokolov et al. 1999). It is not yet possible to constrain the timing of accretion of the Kuyul ophiolites. These ophiolites form part of the accretionary wedge (Ainyn terrane) exposed in the Penzhina segment of the Uda-Murgal island-arc system. The first ophiolitic clasts to appear in the overlap sequence and which are clearly derived from the Kuyul terrane are dated as late Albian. On the other hand, ophiolite fragments have been reported from Hauterivian strata of the Ainyn and Upupkin terranes as well (Nekrasov 1976; Markov et al. 1982; Palandzhyan 1992). However, our own studies (Grigoriev et al. 1995; Sokolov et al. 1996) have not confirmed the presence of Mesozoic ophiolitic clasts in the Ainyn terrane, and serpentinite conglomerates and sandstones in the Ma-

659

metchinsky Peninsula are unfossiliferous. These coarse clastic rocks may be late Albian in age (K. A. Krylov, pers. comm.). Fragments of the more ancient Ganychalan ophiolite have been reported from Hauterivian strata of the Upupkin terrane (Sokolov et al. 2000). In Late Jurassic to Early Cretaceous time, a new convergent boundary took shape, with a younger Uda-Murgal island-arc system stretching along it (Sokolov et al. 1999). Along this boundary the Pacific plates (Izanagi and Farallon) were subducting with the result that ophiolites of Cape Povorotny, the Yelistratov Peninsula, and the Kuyul terrane were emplaced in the forearc. Accretionary prisms of Cape Povorotny and the Penzhina District contain both oceanic rock assemblages and fragments of ensimatic arcs, with outboard chertbasalt associations of Cape Povorotny, from the Izanagi plate (Bazhenov et al. 1999). The original positions of the ensimatic islandarc complexes and their relationships with the convergent boundary of the North Asian continent remain as yet unclear. In unravelling these issues, the greatest promise seems to be offered by the Yelistratov Peninsula ophiolites, because they are extremely well exposed and their stratigraphy is relatively coherent, such that primary intrusive contacts of SSZ gabbro with its surrounding oceanic upper-mantle rocks are preserved intact today (Ishiwatari et al. 1998; Saito et al. 1999). The accretionary prisms of Cape Povorotny and Ainyn terrane contain terrigenous melanges (Khudoley & Sokolov 1998; Sokolov et al. 1999). By and large, these melanges are very similar to those described from the Shimanto Belt of southeastern Japan (Kano et al. 1991) and from southern Alaska (Fisher & Byrne 1987). The existence of two island-arc systems of different ages (Koni-Taigonos and Uda-Murgal) has determined the evolution of the NE Asian continental margin. Sequential growth of the North Asian continent (Fig. 2) is evidenced by ophiolites becoming increasingly younger oceanward, a pattern typical of Circum-Pacific ophiolites (Ishiwatari 1994). This is especially readily apparent in the Penzhina segment, where Mesozoic ophiolites of the Kuyul accreted terrane occur SE of the Early Palaeozoic ophiolites of the Ganychalan terrane. In association with ophiolites one finds glaucophane schists, also younging oceanward. In the Ganychalan terrane, subduction-related metamorphic schists are dated at 327 ± 5 Ma, and in the Kuyul terrane, at 139 ± 6 Ma (Rb/Sr method, four isochron points) with an initial Rb/Sr ratio of 0.711 (Vinogradov et al. 1995), and 92 ± 10 Ma (Rb/Sr method, two isochron points) with an initial Rb/Sr ratio of 0.7324 (Vinogradov et al. 1995).

660

S. D. SOKOLOV£T^L.

The long-lived convergent boundary between the North Asian continent and the NW Pacific, initiated in the Late Palaeozoic, is evidenced by the ages of SSZ ophiolites, glaucophane and greenschist rocks, and island-arc assemblages. This reconstruction places additional constraints on migration of the Omolon and Okhotsk microcontinents. Previous notions (Parfenov 1984; Zonenshain et al. 1990) on Pacific provenance of the microcontinents and their arrival from southern latitudes require revision. On the other hand, palaeomagnetic and palaeobiogeographical data imply that both Palaeozoic and Mesozoic ophiolites and their associated chert and basalt assemblages are rather far travelled (Heiphetz et al. 1994; Bazhenov et al. 1999; Harbert et al. 2003). In this respect, Pacific ophiolites differ significantly from their Tethyan counterparts.

Conclusions Lithostratigraphic, petrographic, mineralogical, geochemical, and temporal characteristics of the ophiolites and associated volcanic rocks of the West Koryak fold system are extremely variable and suggest a spectrum of formative tectonic settings: ocean basin, back-arc basin, volcanic arc, and within plate. West Koryak ophiolites are very diversified, as in other orogenic belts (e.g. the Brooks Range, Klamath Mountains, Japan). Ages of reliably dated ophiolites of the West Koryak fold system are consistent with temporal distribution of ophiolites in worldwide orogenic belts. According to Ishiwatari (1994), distribution of Phanerozoic ophiolites shows two peaks, Ordovician and Jurassic-Cretaceous. The oceanward and downward younging of the ophiolites through accretionary piles is very readily demonstrable in the Penzhina segment. Such a pattern has been reported from Japan (Ishiwatari 1994) and the Klamath Mountains (Coleman 1986), and it is characteristic of accretionary continental margins. In most accreted terranes (Ust-Belaya, Kuyul, Beregovoi), ophiolites of contrasting ages and geodynamic types are tectonically juxtaposed. Thus, both SSZ and oceanic (MOR, BABB) ophiolites are present in the Ust-Belaya and Kuyul terranes, where they form individual thrust sheets. The ophiolites have distinctive tectonic positions. The Palaeozoic ophiolites occur as nappes in the Ganychalan and Ust-Belaya terranes, and Mesozoic ophiolites and melanges occur among accretionary piles of the Beregovoi and Kuyul terranes. Some of them (Ganychalan, Yelistratov, Ust-Belaya) are incorporated in pre-arc basement

of the Uda-Murgal island-arc system, whereas others (Cape Povorotny, Kuyul) are part of accretionary prisms of the same system. The above considerations support the model for ophiolite generation in a large ocean basin with active margins (e.g. the Pacific Ocean), followed by their tectonic juxtaposition within a single orogenic belt (West Koryak fold system). Thanks are due to our reviewers P. T. Robinson, R. Hebert, and C. Buehan. We thank Y. Dilek for his fruitful criticism and invaluable suggestions. This work was supported by the Russian Foundation for Basic Research (project nos 02-05-64217, 01-05-64469, 00-0564165) and by INTAS (project no. 96-1880).

References AMRI, I., BENOIT, M. & CEULENEER, G. 1996. Tectonic setting for the genesis of oceanic plagiogranites: evidence from a paleo-spreading structure in the Oman ophiolite. Earth and Planetary Science Letters, 139, 177-194. ANDERS, E. & GREVESSE, N. 1989. Abundances of the elements: meteoritic and solar. Geochimica et Cosmochimica Ada, 53, 197-214. ARAI, S. 1992. Chemistry of chromian spinel in volcanic rocks as a potential guide to magma chemistry. Mineralogical Magazine, 56, 173-184. ARAI, S. 1994. Characterization of spinel peridotites by olivine—spinel compositional relationships: review and interpretation. Chemical Geology, 113, 191204. BARNES, S.J. & ROEDER, P.L. 2001. The range of spinel compositions in terrestrial mafic and ultramafic rocks. Journal of Petrology, 42, 2279-2302. BATANOVA, V.G. & SOBOLEV, A.V. 2000. Compositional heterogeneity in subduction-related mantle peridotites, Troodos massif, Cyprus. Geology, 28, 55-58. BAZHENOV, M.L., ALEKSUYTIN, M.V., BONDARENKO, G.YE. & SOKOLOV, S.D. 1999. Mesozoic paleomagnetism of the Taigonos Peninsula, the Sea of Okhotsk: implications to kinematics of the continental and oceanic plates. Earth and Planetary Science Letters, 173, 113-127. BAZYLEV, B.A. & SILANTYEV, S.A. 2000. Geodynamic interpretation of the subsolidus recrystallization of mantle spinel peridotites: 1. Mid-oceanic ridges. Petrology, 8, 201-213. BAZYLEV, B.A., MAGAKYAN, R., SILANTYEV, S.A., IGNATENKO, K.I., ROMASHOVA, T.V. & XENOPHONTOS, C. 1993. Petrology of peridotites from the Mamvonia complex, Southwestern Cyprus. Petrology, 1, 302-330. BAZYLEV, B.A., ZAKARIADZE, G.S., ZHELYAZKOVAPANAYOTOVA, M.D., KOLCHEVA, K., OBERKHANSLI, R. & SOLOV'EVA, N.V. 1999. Petrology of ultramafic rocks from the ophiolite association in the crystalline basement of the Rhodope massif. Petrology,!, 191-212. BAZYLEV, B.A., SILANTYEV, S.A. & GANELIN, A.V. 2000. Different ultramafic types at Povorotny Cape ophiolite melange, Taigonos Peninsula, NE Russia.

OPHIOLITES OF NORTHEAST ASIA Geosciences 2000. Confererence abstracts, University of Manchester, 25. BAZYLEV, B.A., PALANDZHYAN, S.A., GANELIN, A.V., SlLANTYEV, S.A., ISHIWATARI, A. & DMITRENKO,

G.G. 2001. Petrology of peridotites from the ophiolite melange in Cape Povorotny, Taigonos peninsula, NE Russia: mantle processes beneath a subduction zone. Petrology, 9, 142-160. BELYI, V.F. & AKININ, V.V. 1985. Geological Structure and Ophiolites in Yelistratov Peninsula, Volumes 1 and 2. NE Scientific Centre of the USSR Academy of Science, Magadan (in Russian). BIZIMIS, M., SALTERS, V.J.M. & BONATTI, E. 2000. Trace and REE content of clinopyroxenes from supra-subduction zone peridotites. Implications for melting and enrichment processes in island arcs. Chemical Geology, 165, 67-85. BOGDANOV, N.A. & TIL'MAN, S.M. 1992. Tectonics and Geodynamics of Northeast Russia. Institute of the Lithosphere, Moscow (in Russian). BONATTI, E. & MICHAEL, P.E. 1989. Mantle peridotites from continental rifts to ocean basins to subduction zones. Earth and Planetary Science Letters, 91, 297-311. BONDARENKO, G.E. & SoKOLOV, S.D. 1996. Strike-slips in the structure of southwestern part of Taigonos Peninsula. Pacific Geology, 2, 99-106. COLEMAN, R.G. 1984. The diversity of ophiolites. Geologic en Mijnbouw., 63, 267-283. COLEMAN, R.G. 1986. Ophiolites and accretion of the North American Cordilliera. Bulletin de la Societe Geologique de France, 6, 961-968. COLEMAN, R.G. & DONATO, M.M. 1979. Once more about oceanic plagiogranites. In: BARKER, F. (ed.) Trondhjemites, Dacites and Related Rocks. Mir, Moscow, 118-130. CONEY, P.J., JONES, D.L. & MONGER, J.W.H. 1980. Cordilleran suspect terranes. Nature, 288, 329-333. COWAN, D.S. 1985. Structural style in Mesozoic and Cenozoic melanges in the western Cordillera of North America. Geological Society of America Bulletin, 96, 451-462. Cox, J., SEARLE, M. & PEDERSEN, R. 1999. The petrogenesis of leucogranite dykes intruding the northern Semail ophiolite, United Arab Emirates: field relationships, geochemistry and Sr/Nd isotope systematics. Contributions to Mineralogy and Petrology, 137, 267-287. DICK, H.J.B. 1989. Abyssal peridotites, very slow spreading ridges and ocean ridge magmatism. In: SAUNDERS, A.D. & NORRY, MJ. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 71-105. DICK, H.J.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DICK, H.J.B. & NATLAND, J.H. 1996. Late-stage melt evolution and transport in the East Pacific Rise. In: GILLIS, M.C., ALLAN, K.M. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 103-134. DILEK, Y., MOORES, E.M., ELTHON, D. & NICOLAS, A.

661

1999. Ophiolites and oceanic crust: new insights from field studies and Ocean Drilling Program. GSA Today, 9, 30-32. DOBRETSOV, N.L. 1974. Glaucophane Schist and Eclogite-Glaucophane Schist Complexes of the USSR. Nauka, Novosibirsk (in Russian). ELTHON, D. 1991. Geochemical evidence for formation of the Bay of Islands ophiolite above a subduction zone. Nature, 354, 141-143. FILATOVA, N.I. 1988. Perioceanic Volcanic Belts. Nedra, Moscow (in Russian). FISHER, D.M. & BYRNE, T. 1987. Structural evolution of underthrusted sediments, Kodiak Islands, Alaska. Tectonics, 6, 775-793. FRYER, P. 1992. A synthesis of leg 125 drilling of serpentinite seamounts on Mariana and Izu-Bonin forearcs. In: FRYER, P. ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 593-614. FUJITA, K. & NEWBERRY, T. 1982. Tectonic evolution of Northeastern Siberia and adjacent regions. Tectonophysics, 89, 337-357. GABRIELSE, H., MONGER, J.W.H., YORATH, C.J. & DODDS, C. 1991. Transcurrent faults. In: GABRIELSE, H. & YORATH, C.J. (eds) Geology of the Cordilleran Orogen in Canada. Geological Society of America, The Geology of North America, G-2, 651-660. GANELIN, A.V. & PEYVE, A.A. 2001. Geodynamic setting of formation of the Ganychalan terrane ophiolite (Koryakya Upland). In: KARPOV, A.G. (ed.) Petrology and Metallogeny of Basic-Ultrabasic Complexes of Kamchatka. Scientific World, Moscow, 215-230 (in Russian). GILLIS, C.M. 1996. Rare-earth element constraints on the origin of amphibole in gabbroic rocks from site 894, Hess Deep. In: GILLIS, M.C., ALLAN, K.M. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 135-155. GLASSLEY, W. 1974. Geochemistry and tectonics of Crescent volcanic rocks, Olympic Peninsula, Washington. Geological Society of America Bulletin, 85, 785-794. GRIGORIEV, V.N., SOKOLOV, S.D., KRYLOV, K.A., GOLOZUBOV, V.V. & PRALNIKOVA, I.E. 1995. Geodynamic types of the Triassic-Jurassic volcanicsiliceous complexes in the Kuyul terrane (Koryak Upland). Geotectonics, 29, 248-258 (in Russian). HARBERT, W., SOKOLOV, S., KRYLOV, K., ALEXUTIN, M., GRIGORIEV, V. & HEIPHETZ, A. 2003. Reconnisance paleomagnetism of Late Triassic blocks, Kuyul region, northern Kamchatka Peninsula, Russia. Tectonophysics, 361, 215-227. HART, S.R. & DUNN, T. 1993. Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology, 113, 1-8. HEIPHETZ, A., HARBERT, W. & LAYER, P. 1994. Preliminary reconnaissance paleomagnetism of some Late Mesozoic ophiolites, Kuyul region, Koryak superterrane, Russia. In: THURSTON, D.K. & FUJITA, K. (eds) Proceedings of the International Conference on Artie Margisn MMS 94-0040.

662

S. D. SOKOLOV £T^L.

US Department of Interior Minerals Management Service Alaska Outer Continental Shelf Region, Anchorage, 229-234. IRVINE, T.N. 1967. Chromium spinels as a petrogenetic indicator. II. Petrologic applications. Canadian Journal of Earth Sciences, 4, 71-103. ISHII, T., ROBINSON, P.T., MAEKAWA, H. & FISKE, R. ET AL. 1992. Petrological studies of peridotites from diapiric serpentinite seamounts in the Izu-Ogasawara-Mariana forearc. In: FRYER, P. (ed.) Proceeding of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station, TX, 445-486. ISHIWATARI, A. 1994. Circum-Pacific Phanerozoic multiple ophiolite belts. In: ISHIWATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Proceedings of the 29th International Congress. Part D. Circum-Pacific Ophiolite. Utrecht, Netherland: VSP, 7-28. ISHIWATARI, A., SOKOLOV, S.D., SAITO, D., TSUJIMORI, T. & MIYASHITA, S. 1998. Geology and petrology of Yelistratov ophiolite in Taigonos Peninsula, Northeastern Russia: an island-arc ophiolite intruded into oceanic mantle. In: HANSKI, E. & VUOLLO, J. (eds) International Ophiolite Symposium and Field Excursion—Generation and Emplacement of Ophiolites through Time. Geological Survey of Finland, Special Papers, 26, 29-30. JAQUES, A.L. & GREEN, D.H. 1980. Anhydrous melting of peridotite at 0-15 kbar pressure and the genesis of tholeiitic basalts. Contributions to Mineralogy and Petrology, 73(3), 287-310. KANO, K.L, NAKAJI, M. & TAKEUCHI, S. 1991. Asymmetrical melange fabrics as possible indicators of the convergent direction of plates: a case study from Shimanto Belt of the Akaishi Mountains, central Japan. Tectonophysics, 185, 375-388. KAY, R.W. & SENECHAL, R.G. 1976. The rare earth geochemistry of the Troodos ophiolite complex. Journal of Geophysical Research, 81(5), 964-970. KELEMEN, P.B., KOGA, K. & SHIMIZU, N. 1997. Geochemistry of gabbro sills in the crust-mantle transition zone of the Oman ophiolite: implications for the origin of the oceanic lower crust. Earth and Planetary Science Letters, 146, 475-488. KHANCHUK, A.L, GRIGORYEV, V.N. & GOLOZUBOV, V.V. ET AL. 1990. Kuyul Ophiolite Terrane. Academy of Science Publications, Vladivostok (in Russian). KHANCHUK, A.L, GOLOZUBOV, V.V., PANCHENKO, A.V., IGNATIEV, A.V. & CHUDAEV, O.V. 1992. Ganychalan terrane of the Koryak Upland. Pacific Geology, 3, 82-93 (in Russian). KHUDOLEY, A.K. & SOKOLOV, S.D. 1998. Structural evolution of the northeastern Asian continental margin: an example from the western Koryak foldand-thrust belt (northeast Russia). Geological Magazine, 135,311-330. KORNPROBST, L, OHNENSTETTER, D. & OHNENSTETTER, M. 1981. Na and Cr contents in clinopyroxenes from peridotites: a possible discriminant between 'sub-continental' and 'sub-oceanic' mantle. Earth and Planetary Science Letters, 53, 241-254. LAGABRIELLE, Y., KARPOFF, A.M. & COTTON, J. 1992. Mineralogical and geochemical analyses of sedi-

mentary serpentinite from conical seamount (hole 778A): implications for the evolution of serpentinite seamount. In: FRYER, P., PEARCE, LA., STOCKING, L.B., ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific results, 125. Ocean Drilling Program, College Station, TX, 325-342. LANGMUIR, C.H., KLEIN, E. & PLANK, T. 1992. Petrological systematics of Mid-Ocean ridge basalts: constraints on melt generation beneath ocean ridges. In: LANGMUIR, C.H. ET AL. (eds) Mantle Flow and Melt Generation at Mid-Ocean Ridges. Geophysical Monograph, American Geophysical Union, 71, 183-280. LUCHITSKAYA, M.V. 1996. Plagiogranites of the Kuyul ophiolite massif, Northeastern Russia. Ofioliti, 21, 131-138. LUCHITSKAYA, M.V. 2001. Tonalite-Trondhjemite Complexes of Koryakya-Kamchatka Region (Geology, Geodynamics). Transactions of the Geological Institute of the Russian Academy of Sciences, Moscow, GEOS, 123. LYCHAGIN, P.P., BYALOBZHESSKY, S.G., KOLYASNIKOV, Yu.A., KORAGO, YE.A. & LIKMAN, V.B. 1991. The magmatic history of the South Anyui folded zone. In: SIDOROV, A.A. & MILOV, A.P. (eds) Geology of Continent-Ocean Transition Zone in Northeast Asia. SVKNII RAN, Magadan, 140-157 (in Russian). MARKOV, M.S., NEKRASOV, G.E. & PALANDJYAN, S.A. 1982. Ophiolites and melanocratic basement of the Koryak Upland. In: PUSHCHAROVSKY, Y.M. & TIL'MAN, S.M. (eds) Essays of Tectonics of the Koryak Upland. Nauka, Moscow, 30-70. MIYASHIRO, A. & SHIDO, F. 1980. Differentiation of gabbros in the Mid-Atlantic Ridge near 24°N. Journal of Geochemistry, 14, 145-154. MOROZOV, O.L. 2001. Geology and Structure of Central Chukotka. Transactions of the Geological Institute of the Russian Academy of Sciences, Moscow, GEOS, 200 (in Russian). NEKRASOV, G.E. 1976. Tectonics and Magmatism of Taigonos and Northwestern Kamchatka. Nauka, Moscow (in Russian). NEKRASOV, G.E., ZABOROVSKAYA, N.B. & LYAPUNOV, S.M. 2001. Pre-Late Paleozoic ophiolites of the western part of the Koyak Upland: fragments of an oceanic plateau. Geotectonics, 2, 41-63 (in Russian). Niu, Y. 1997. Mantle melting and melt extraction processes beneath ocean ridges: evidence from abyssal peridotites. Journal of Petrology, 38, 10471074. NOKLEBERG, W.L, PARFENOV, L.M. & MONGER, LW.H.

ET AL. 1994. Circum-North Pacific Tectono-Stratigraphic Terrane Map. US Geological Survey OpenFile Report, 94-714. NOKLEBERG, W.L, PARFENOV, L.M. & MONGER, LW.H. ET AL. 2001. Phanerozoic Tectonic Evolution of the Circum-North Pacific. US Geological Survey Professional Paper IG2G. O'CONNOR, J.T. 1969. A classification for quartz-rich igneous rocks based on feldspar ratios. US Geological Survey, Professional Papers, 525-B, 79-84. OHNENSTETTER, M. 1985. Classification petrographique

OPHIOLITES OF NORTHEAST ASIA et structurale des ophiolites, de la dynamique des zones transition croute-manteau. Incidence sur la nature et la disposition des corps de chromite associes. Comptes Rendus de I'Academic des Sciences, Serie II, 301(20), 1413-1418. OZAWA, K. & SHIMIZU, N. 1995. Open-system melting in the upper-mantle—constraints from the Hayachine-Miyamori Ophiolite, Northeastern Japan. Journal of Geophysical Research, 100, 2231522335. OXMAN, V.S., PARFENOV, L.M. & PROKOFIEV, A.V. ET AL. 1995. The Chersky Range ophiolite belt, Northeast Russia. Journal of Geology, 103, 539-556. PALANDZHYAN, S.A. 1992. Classification of Mantle Peridotites by Geodynamic Settings of their Formation. SVKNII DVO RAN, Magadan (in Russian). PALANDZHYAN, S.A. 1997. Granitoids of the Ophiolite Belts of Koryak Foldbelt. In: BYALOBZHESSKY, S.G. (ed.) Magmatism and Mineralization of Northeast Russia. SVKNII DVO RAN, Magadan, 204-224 (in Russian). PALANDZHYAN, S.A. 2000. Ophiolites of the Ust-Belaya terrane: a Middle Paleozoic oceanic-type association in the West Koryak fold-and-thrust belt. In: Magmatism and Metamorphism in the Northeast Asia. North East Interdisciplinary Research Institute, Magadan, 180-184 (in Russian). PALANDZHYAN, S.A. & DMITRENKO, G.G. 1996. Ophiolitic Complexes and Associated Rocks in the UstBelaya Mountains and Algan Ridge, Russian Far East. US Geological Survey, Open-File Report, OF 92-20-1. PALANDZHYAN, S.A. & DMITRENKO, G.G. 1999. Peridotites of the Taigonos Pribrezhny Belt: chemical compositions of minerals as indicators of geodynamic conditions of ophiolites formation. In: KARYAKIN, Yu.V. (ed.) Tectonics, Geodynamics and Magmatic and Metamorphic Processes. Proceedings of 32nd Tectonic Conference. GEOS, Moscow, 53-57 (in Russian). PALLISTER, J.S. & HOPSON, S.A. 1981. Samail Ophiolite plutonic suite: field relations, phase variation, cryptic variation and layering, and a model of spreading ridge magma chamber. Journal of Geophysical Research, 86, 2593-2644. PALLISTER, J.S. & KNIGHT, R.J. 1981. Rare-earth element geochemistry of the Semail ophiolite near Ibra, Oman. Journal of Geophysical Research, 86, 2673-2697. PARFENOV, L.M. 1984. Continental Margins and Island Arcs in the Mesozoides of Northeastern Asia. Nauka, Novosibirsk (in Russian). PARFENOV, L.M., NATAPOV, L.M., SOKOLOV, S.D. & TSUKANOV, N.V. 1993. Terrane analysis and accretion in northeast Asia. Island Arc, 2, 35-54. PARKINSON, I.J. & PEARCE, J.A. 1998. Peridotites from the Izu-Bonin-Mariana forearc (OOP Leg 125): evidence for mantle melting and melt-mantle interaction in a suprasubduction zone setting. Journal of Petrology, 39, 1577-1618. PEARCE, J.A. 1992. An element recycling. Nature, 360, 629-630. PEARCE, J.A. & NORRY, M.J. 1979. Petrogenetic implications of Ti, Zr, Y, and Nb variations in volcanic

663

rocks. Contributions to Mineralogy and Petrology, 69, 33-47. PEARCE, J.A., HARRIS, N.B.W. & TINDLE, A.G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. PEARCE, J.A., BARKER, P.P., EDWARDS, S.J., PARKINSON, I.J. & LEAT, P.T. 2000. Geochemistry and tectonic significance of peridotites from the South Sandwich arc-basin system, South Atlantic. Contributions to Mineralogy and Petrology, 139, 36-53. PEDERSEN, R.B., MALPAS, J. & FALLOON, T. 1996. Petrology and geochemistry of gabbroic and related rocks from Site 894, Hess Deep. In: MEVEL, C., ALLAN, J.F. & MEYER, P.S. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 147. Ocean Drilling Program, College Station, TX, 3-19. PEYVE, A.A. 1984. Structure and Structural Position of the Koryaksky Range Ophiolites. Nauka, Moscow (in Russian). PUSHCHAROVSKY, Yu.M., SOKOLOV, S.D., TlLMAN, S.M.

& KRYLOV, K.A. 1992. Tectonics and geodynamics of northwestern rim of the Pacific Ocean. In: BORUKAEV, CH.B. & VRUBLEVSKY, A.A. (eds) Issues of Tectonics and Mineral and Energy Resources of Northwestern Pacific. DVO ITIG AN USSR, Khabarovsk, 1, 128-137 (in Russian). RAMPONE, E., HOFMANN, A.W., PICCARDO, G.B., VANNUCCI, R., BOTTAZZI, P. & OTTOLINI, L.

1995.

Petrology, mineral and isotope geochemistry of the External Liguride peridotites (Northern Apennines, Italy). Journal of Petrology, 36, 81-105. RIKHTER, YA.A. 1997. New data on the gabbro-tonalite assemblage of the Mid-Atlantic ridge equatorial segment. Doklady Russian Academy of Sciences, 355, 506-508 (in Russian). ROSENKRANTZ, A.A. (ed.) 1986. Geological Map of USSR. New Series. Scale 1:1000000. Sheets P-58, 59. Mingeo USSR, VSEGEI, Leningrad. SAITO, D., ISHIWATARI, A., TSUJIMORI, T., MIYASHITA, S. & SOKOLOV, S.D. 1999. Yelistratov ophiolite in Taigonos peninsula, Far-East Russia: an island arc ophiolite intruding into oceanic mantle. Memoirs of Geological Society of Japan, 52, 303-316 (in Japanese with English abstract). SAUNDERS, A.D. & TARNEY, J. 1984. Geochemical characteristics of basaltic volcanism within back-arc basin. In: KOKELAAR, B.P. & HOWELLS, M.F. (eds) Marginal Basin Geology: Volcanic and Associated Sedimentary and Tectonic Processes in Modern and Ancient Marginal Basins. Geological Society, London, Special Publications, 16, 59-76. SAUNDERS, A.D., TARNEY, J. & WEAVER, S.D. 1980. Transverse geochemical variations across the Antarctic Peninsula: implications for the genesis of calc-alkaline magmas. Earth and Planetary Science Letters, 46, 344-360. SAUNDERS, A.D., NORRY, M.J. & TARNEY, J. 1991. Fluid influence on the trace element compositions of subduction zone magmas. Philosophical Transactions of the Royal Society of London, 335, 377-392. SEARLE, M.P. & Cox, J.S. 1999. Tectonic setting, origin and obduction of the Oman ophiolite. Geological

664

S. D. SOKOLOV ETAL.

Society of America Bulletin, 111, 104-122. SHAPIRO, M.N. & GANELIN V.G. 1988. Paleogeodynamic interrelations in the large blocks of the Mesozoic of the North-Eastern USSR. Geotectonics, 5, 94-104 (in Russian). SHERVAIS, J.W. 1982. Ti-V plots and the petrogenesis of modern and ophiolitic lavas. Earth and Planetary Science Letters, 59, 101-118. SHERVAIS, J.W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, 200GC 000080. SlLANTYEV, S.A.,

SOKOLOV, S.D.

& POLUNIN,

G.V.

1994. New data on the composition of metamorphic rocks of the Talovka River basin (Penzhina region, Northern Russia). Geotectonics, 2, 169-176. SILANTYEV, S.A., SOKOLOV, S.D., BONDARENKO, G.YE, MOROZOV, O.L., BAZYLEV, B.A., PALANDZHYAN, S.A. & GANELIN, A.V. 2000. Geodynamic setting of the high-grade amphibolites and associated igneous rocks from the accretionary complex of Povorotny Cape, Taigonos Peninsula, Northeastern Russia. Tectonophysics, 325, 107-132. SOBOLEV, A.V. & BATANOVA, V.G. 1995. Mantle Iherzolites of the Troodos ophiolite complex, Cyprus: clinopyroxene geochemistry. Petrology, 3, 440-448. SOKOLOV, S.D. 1992. Accretionary tectonics of the Koryak—Chukotka Segment of the Pacific belt. Transactions of Geological Institute, Russian Academy of Sciences, 182 (in Russian). SOKOLOV, S.D., GRIGORIEV, V.N., PEYVE, A.A., BATANOVA, V.G., KRYLOV, K.A., LUCHITSKAYA, M.V. & ALEKSUTIN, M.V. 1996. The structural and compositional regulation in serpentinite melanges. Geotectonics, 1, 3-16. SOKOLOV, S.D., GRIGORIEV, V.N., ARISTOV, V.A., PEYVE, A.A. & SHTERENBERG, L.E. 1997. Ordovician deposits of the Ganychalan terrane. Stratigrafiya i Geologicheskaya Korrelyatsiya, 5, 73-84 (in Russian). SOKOLOV, S.D., BONDARENKO, G.YE., MOROZOV, O.L. & GRIGORIYEV, V.N. 1999. Transition zone of the Asian continent-Northwestern Pacific in Late Jurassic and Early Cretaceous. In: GAVRILOV, Yu. O.

& KURENKOV, S.A. (ed.) Theoretic and Regional Problems of Geodynamics. Nauka, Moscow, 30—84 (in Russian). SOKOLOV, S.D., LAGABRIELLE, Y., GERARD, J.-K. & BAZYLEV, B.A. 2000. Geodynamic setting of ultramafic bodies of Mt. Dlinnaya (Penzhina region, Northeast Russia). Byulleten' Moskovskogo Obshchestva Ispytateley Prirody, Seria Geologicheskaya, 75, 51-55 (in Russian). SUN, S.-S. & McDoNOUGH, W.F. 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: SAUNDERS, A.D. & NORRY, M.J. (eds) Magmatism in the Ocean Basins. Geological Society, London, Special Publications, 42, 313-345. TAKAZAWA, E., FREY, F.A., SHIMIZU, N. & OBATA, M. 2000. Whole rock compositional variations in an upper mantle peridotite (Horoman, Hokkaido, Japan): are they consistent with a partial melting process? Geochimica et Cosmochimica Acta, 64, 695-716. TIEPOLO, M., TRIBUZIO, R. & VANNUCCI, R. 1997. Mgand Fe-gabbroids from Northern Apennine ophiolites: parental liquids and igneous differentiation process. Ofioliti, 22, 57-69. TORIUMI, M. 1985. Two types of ductile deformation/ regional metamorphic belts. Tectonophysics, 113, 307-326. VlNOGRADOV, V.L, YURKOVA, R.M.,

SOKOLOV, S.D.,

BUYAKAYTE, M.I. & VoRONiN, B.I. 1995. Rb-Sr dating of dynamometamorphosed rocks from the Penzhinskiy Ridge, Kamchatka. Geotectonics, 28, 424-430. WIRTH, K.R., BIRD, J.M. & WESSELS, J.N. 1994. The diversity of accreted oceanic lithosphere in the Brooks Range, Alaska. In: ISHIWATARI, A., MALPAS, J. & ISHIZUKA, H. (eds) Proceeding of the 29th International Congress. Part D. CircumPacific Ophiolite. VSP, Utrecht, 89-108. ZABOROVSKAYA, N.B. 1978. Inner Zone of the OkhotskChukotka Belt, Taigonos. Nauka, Moscow (in Russian). ZONENSHAIN, L.P., KUZMIN, M.I. & NATAPOV, L.M.

1990. Tectonics of Lithospheric Plates of the USSR Territory, Book 2. Nedra, Moscow (in Russian).

Rocas Verdes ophiolites, southernmost South America: remnants of progressive stages of development of oceanic-type crust in a continental margin back-arc basin CHARLES R. STERN 1 & MAARTEN J. DE WIT 2 Department of Geological Sciences, University of Colorado, Boulder, CO 80309-0399, USA (e-mail: Charles.Stern@colorado. edu) 2 Department of Geological Sciences, University of Cape Town, Rondebosch 7700, South Africa Abstract: The Mesozoic Rocas Verdes are a group of mafic igneous complexes in the southernmost Andes. They consist of pillow lavas, dykes and gabbros, interpreted as the upper parts of ophiolites formed along mid-ocean-ridge-type spreading centres that rifted the southwestern margin of the Gondwana continental crust, during the onset of spreading in the South Atlantic, to form the mafic igneous part of the floor of a back-arc basin behind a contemporaneous convergent plate boundary magmatic arc. Mafic dykes and gabbros intrude older continental lithologies along both flanks of the Rocas Verdes, and these same leucocratic country rocks are engulfed in the Rocas Verdes mafic complexes. These relations indicate that the Rocas Verdes ophiolites formed in place and are autochthonous. Zircon U/Pb ages, as well as both chemical and lithostructural characteristics of these ophiolite complexes, suggest that the Rocas Verdes basin formed by 'unzipping' from the south to the north, with the southern part beginning to form earlier, and developing more extensively, than the northern part of the basin. The Rocas Verdes ophiolites contain a wealth of information about progressive stages of continental rifting during back-arc basin formation, magmatic and metamorphic processes along mid-ocean-ridge-type spreading centres, and as analogues to Archaean greenstone belts.

The Mesozoic Rocas Verdes (Spanish for green rocks, or greenstones as they were described by Katz (1964)) consist of a group of igneous complexes located between 50°S and 56°S in the Andes of southernmost South America (Fig. 1; Katz 1964, 1972, 1973; Dalziel et al. 1974). These complexes consist predominantly of mafic pillow lavas and breccias, sheeted dykes (Fig. 2a) and gabbros, interpreted as the upper parts of ophiolites formed along mid-ocean-ridge-type spreading centres (Dalziel et al. 1974; de Wit & Stern 1976, 1978, 1981; Stern et al. 1976; Stern & de Wit 1980). Mafic dykes and gabbros cut older Palaeozoic (pre-Andean) and Jurassic (Andean) lithologies along the flanks of the Rocas Verdes (Fig. 2b), and some of these same leucocratic country rocks are engulfed in the Rocas Verdes mafic complexes, suggesting that these ophiolites are essentially autochthonous. The Rocas Verdes mafic complexes floored part of a Late Jurassic and Early Cretaceous marine basin that developed NE of a convergent plate boundary magmatic arc (Figs 3 and 4; Dalziel et al. 1974; Suarez & Pettigrew 1976; de Wit 1977; Suarez 1977, 1979; Tanner & Rex 1979;

Dalziel 1981). They are interpreted as the remnants of the mafic igneous part of the floor of this ensialic back-arc (marginal) basin, generated along mid-ocean-ridge-type spreading centres that rifted the western continental margin of Gondwana-South America just to the north and east of a contemporaneous magmatic arc. Sediments derived from both this magmatic arc on the SW, and the continental platform on the NE, were deposited in the Rocas Verdes basin (Fig. 4b; Dott et al. 1977; Winn 1978; Winn & Dott 1978). Closure, uplift and deformation of the basin took place during the mid-Cretaceous (Fig. 4c; Halpern 1973; Dalziel et al. 1974; Bruhn & Dalziel 1977; Dott et al. 1977; Bruhn 1979; Herve et al. 1981; Nelson 1982; Dalziel 1986; Grunow et al. 1992; Cunningham 1994, 1995). Zircon U/Pb ages of 150 ± 1 Ma for the Larson Harbor Formation (Mukasa & Dalziel 1996), a southern extension of the Rocas Verdes on South Georgia (Storey et al. 1977; Storey & Mair 1982), and 139 ± 2 Ma for the Sarmiento complex, the northernmost in the Rocas Verdes belt (Fig. 1; Stern et al. 1992), are consistent with the Late Jurassic and Early Cretaceous age range implied

From: DILEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 665-683. 0305-8719/037$ 15 © The Geological Society of London 2003.

666

C. R. STERN & M. J. DE WIT

Fig. 1. Simplified geological map of the southernmost Andes, modified after Dalziel et al. (1974). Inset shows the location of the area in the context of the Scotia Arc.

by stratigraphic controls (Bruhn et al. 1978; Fuenzalida & Covacevich 1988). These ages suggest that the basin may have opened by unzipping from the south to the north. Field and petrochemical data also suggest that the Rocas Verdes backarc basin widened, from possibly <50 km in the north to possibly > 100 km in the south (de Wit 1977; de Wit & Stern 1978, 1981), and towards the south its mafic igneous floor became more oceanic in character (Saunders et al. 1979; Stern 1979, 1980, 1991; Alabaster & Storey 1990; Storey & Alabaster 1991). The mafic rocks of the Tortuga complex, the southernmost in the Rocas Verdes belt, are chemically similar to mid-ocean ridge basalt (MORE; Stern 1979, 1980, 1991), and include high-MgO (>10wt% MgO) dykes (Elthon 1979; Elthon & Ridley 1980), identified as probable parental magmas for this mafic complex. Further to the south, the Rocas Verdes basin may have been linked to the even wider oceanic

environment of the proto-Weddell Sea (de Wit 1977; Grunow 1993a, 1993b; Mukasa & Dalziel 1996). These relations suggest that the Rocas Verdes igneous complexes, along with the associated mafic dykes and gabbros that intrude older preAndean and Andean lithologies along their flanks, are remnants of progressive stages of development of oceanic-type crust in a continental back-arc extensional tectonic environment. The Rocas Verdes formed during the initial stages of Gondwana breakup, and their origin may reflect processes associated with subduction (Bruhn et al. 1978; de Wit & Stern 1981; Alabaster & Storey 1990), reduction in plate boundary forces associated with changes in plate configuration and absolute motion (Dalziel 1981, 1986; de Wit & Stern 1981; Storey & Alabaster 1991), and/or mantle plumes involved in the opening of the South Atlantic (Cox 1978, 1988; de Wit &

ROCAS VERDES OPHIOLITES, SOUTH AMERICA

667

Fig. 2. (a) Sheeted dykes of the Sarmiento ophiolite. Geological hammer provides an approximate scale. Individual dykes are 0.5-1 m wide, and north-south-oriented sheeted dykes occur across an area 3—5 km wide in Lolos and Encuentro Fjords, southernmost Chile (Fig. 5; de Wit & Stern 1981). (b) Multiple north-south-oriented mafic dykes (dark rocks) cutting leucocratic continental basement (light rocks) along the flanks of the Sarmiento ophiolite. Field notebook in the centre of the photograph provides an approximate scale. Similar features are common along both flanks of the Rocas Verdes ophiolites and may be traced across strike into the sheeted dyke units of these ophiolites (de Wit & Stern 1981).

Fig. 3. Map showing the major lithotectonic units of southernmost South America during the Early Cretaceous. Sequential lithotectonic sections across XY during the Mid-Jurassic, Early Cretaceous and Late Cretaceous are shown in Figure 4.

Ransome 1992; Storey 1995; Storey & Kyle 1997; Dalziel et al 2000; Storey et al 2001). The Rocas Verdes and their associated rocks are remarkably well exposed (Fig. 2), and contain a wealth of information about progressive stages of continental rifting during back-arc basin formation (Dalziel et al. 1974; de Wit & Stern 1978, 1981; Saunders et al. 1979; Stern 1979, 1980), as well as petrological and metamorphic processes along MOR-type spreading centres (de Wit & Stern 1976, 1978; Stern et al. 1976; Elthon & Stern 1978; Elthon 1979; Stern 1979, 1980, 1991; Stern & Elthon 1979; Elthon & Ridley 1980; Stern & de Wit 1980; Elthon et al. 1984), and also as analogues to Archaean greenstone belts (Tarney et al. 1976; Stern & de Wit 1997); however, they have not received as much scientific attention as many other ophiolite complexes worldwide, perhaps because of their remote location and the logistic difficulties involved in their study. This

668

C. R. STERN & M. J. DE WIT

Fig. 4. Sequential lithotectonic sections (across XY in Fig. 3) during the Mid-Jurassic to Late Cretaceous, illustrating the formation and collapse of the volcano-tectonic rift zone and the Rocas Verdes back-arc basin. Figure modified after Dalziel et al. (1974), Bruhn et al (1978) and de Wit & Stern (1978, 1981).

paper reviews the major geological and petrochemical features of the Rocas Verdes ophiolite complexes and discusses some of the implications of these features for the process of back-arc basin development and the generation of oceanic-type crust in a continental margin setting.

Regional geology The formation of the Rocas Verdes basin in southernmost South America was immediately preceded by the development of a Jurassic volcano-tectonic rift zone (Fig. 4a; Bruhn et al 1978). The relatively narrow Rocas Verdes basin formed along the western edge of this much broader rift zone, which was characterized by extensive bimodal volcanism

associated with horst and graben tectonics. The voluminous silicic volcanic rocks produced by this rifting range in thickness up to >2 km. They are referred to as the Tobifera Formation in Magallanes, Chile (Bruhn et al. 1978; Hanson & Wilson 1991), and more broadly as the Chon Aike Formation in southern South America (Gust et al. 1985), that formed between 153 and 188 Ma (Riley & Knight 2001). Peraluminous granitoids in Tierra del Fuego, considered to be possible plutonic equivalents of the Tobifera Formation silicic volcanic rocks (Nelson et al 1988; Suarez et al 1990), have been dated at 157 ± 8 Ma by an Rb-Sr isochron for whole-rock samples (Herve et al 1981) and 164.1 ± 1.7 Ma by zircon U-Pb systematics (Mukasa & Dalziel 1996).

ROCAS VERDES OPHIOLITES, SOUTH AMERICA The Tobifera and Chon Aike Formations are the Patagonian portion of an even larger and longerlived Gondwana continental silicic volcanic province that produced the Upper Palaeozoic and Lower Mesozoic Choiyoi granites and rhyolites in central and northern Chile and Argentina (Kay et al. 1989; Mpodozis & Kay 1990), and Middle Jurassic rhyolites and granites of the eastern Antarctic Peninsula dated at 167-189 Ma (Dalziel & Elliot 1982; Dalziel et al. 1987; Elliot 1992; Riley & Knight 2001). This extensive silicic volcanic province, which persisted from the Late Palaeozoic into the Jurassic, or even Early Cretaceous (Kirstein et al. 2001), extended as far east as southern Africa, and south into the proto-Weddell Sea, before and during the opening of the South Atlantic (de Wit 1977; Dalziel et al. 1987; de Wit & Ransome 1992; Elliot 1992; Grunow 1993a, 1993b). The Jurassic Tobifera Formation silicic volcanic rocks unconformably overlie the eroded Palaeozoic (pre-Andean) metamorphic crystalline basement of Patagonia (Fig. 4a; Halpern 1973; Forsyth 1982; Herve et al. 1991). Rifting associated with the formation of the Tobifera volcanic rocks thinned the lithosphere below Patagonia to less than 80 km, c. 20 km thinner than its current 100 km thickness, as indicated by geotherms obtained from mantle xenoliths found within the Quaternary Patagonian plateau basalts (Skewes & Stern 1979; Stern et al 1989, 1999). The mafic igneous protoliths for granulite xenoliths, also found in the plateau basalts, probably were intruded into the base of the South American crust during the formation of this volcano-tectonic rift zone (Selverstone & Stern 1983). As Late Jurassic and Early Cretaceous rifting became focused within the Rocas Verdes basin to the SW, and the South Atlantic spreading ridge to the east, southernmost Patagonia became a stable (cratonic-like) continental platform (Figs 3 and 4b). At the same time, the southwestern margin of South America remained an active convergent plate boundary magmatic arc, as indicated by the fact that a number of I-type calc-alkaline granitic plutons of the southern Patagonian batholith, which represent the roots of this arc south and west of the Rocas Verdes belt, date between 151 and 138 Ma (Halpern 1973; Stern & Stroup 1982; Herve et al. 1984; Nelson et al. 1988; Weaver et al. 1990; Bruce et al. 1991). These plutons, which intrude the same Palaeozoic metamorphic basement underlying the Tobifera Formation volcanic rocks, are thus contemporenous with the magmatic activity that formed the Rocas Verdes mafic igneous complexes (Fig. 4). These relations are consistent with the suggestion that the Rocas Verdes basin was a back-arc

669

(marginal) basin that rifted the continental crust NE of a contemporaneous magmatic arc active along the southwestern continental margin of Gondwana, and subsequently South America, during the Gondwana continental breakup and the onset of spreading in the South Atlantic (Dalziel et al. 1974; Suarez & Pettigrew 1976; Suarez 1979; Tanner & Rex 1979; Dalziel 1981). Sediments derived from both the convergent plate boundary magmatic arc on the SW and the continental platform on the NE were deposited in the Rocas Verdes basin (Fig. 4b; Dott et al. 1977; Winn 1978; Winn & Dott 1978). Arc volcanic rocks and associated coarse volcanogenic sediments are interbedded with mafic pillow lavas along both margins of the Rocas Verdes complexes, and overlie these complexes within the centre of the basin, where they are interbedded with chert and fine-grained deep-water turbidites (Dott et al. 1977; Suarez 1979). The Rocas Verdes ophiolite complexes consist of 2-3 km thick units of predominantly mafic submarine extrusive rocks, including pillow lavas and breccias, cut by dykes and overlying a sheeted dyke complex (Figs 5 and 6). Below the 300500 m thick sheeted dyke complex (Fig. 2a), the lower contact of which may be either sharp or gradational (Fig. 6), c. 1 km of massive diabases and coarse-grained gabbros are exposed. North-tosouth variations in eruptive products (de Wit & Stern 1978), ophiolite pseudostratigraphy (de Wit & Stern 1978, 1981), and petrochemistry (Stern 1979, 1980; Saunders et al. 1979) have been shown to result from smaller degrees of extension in the northern part and greater degrees of extension in the southern part of the Rocas Verdes basin (Fig. 7). Mafic dykes and gabbros of similar composition cut older lithologies on the flanks of the Rocas Verdes ophiolites (Fig. 2b). Here we describe first the mixed mafic-felsic outcrops that flank the Rocas Verdes ophiolites, followed by the Sarmiento ophiolite complex at the northern, narrow end of the Rocas Verdes belt, and then the Tortuga complex at the southern, wider end of this belt (Fig. 1). These igneous complexes are interpreted as representing remnants of the progressive stages of development of the igneous floor of the extensional back-arc Rocas Verdes basin, which ranged from intermediate between continental and oceanic to typically oceanic in character (Fig. 7).

Mixed mafic-felsic terranes flanking the ophiolites Flanking the belt of Rocas Verdes ophiolites are mixed mafic-felsic outcrops in which basaltic

670

C. R. STERN & M. J. DE WIT

Fig. 5. Geological maps of the regions around the Sarmiento and Tortuga ophiolites, at the northern and southern extremes, respectively, of the Rocas Verdes belt, southern Chile.

dyke swarms (Fig. 2b) and large diabase and gabbro sills and stocks intrude remnants of the pre-Cretaceous continental crust, including the Palaeozoic metamorphic basement and Jurassic silicic volcanic and plutonic rocks. Similar mafic intrusions seldom occur within the Lower Cretaceous basin sedimentary rocks, which overlie both the ophiolite complexes and these mixed mafic felsic rock units. The dyke swarms within these mixed mafic-felsic blocks have regular orientations, which parallel both the strike of the ophiolite complexes and the orientation of sheeted dykes within these complexes. Volumetrically, these mafic intrusions gradually increase towards and merge with the intrusive components of the ophiolite complexes. The basaltic igneous rocks in these mixed mafic-felsic rock units are greenstones or spilites, altered by hydrothermal metamorphism without formation of schistosity, similar to the metamorphic overprint observed within the Rocas Verdes ophiolite complexes (Stern et al. 1976; Elthon & Stern 1978; Stern & Elthon 1979). The major and immobile trace element (Ti, Zr, Y and rare earth elements (REE)) compositions of the metamorphosed mafic dykes, sills and stocks in these mixed mafic-felsic terranes flanking the

Rocas Verdes ophiolites are comparable with those of basaltic dykes and lavas from within the ophiolites (Fig. 8; Stern 1979, 1980; de Wit & Stern 1981). In particular, the REE contents of these dykes and sills are similar to those of dykes and lavas from the Sarmiento ophiolite complex, and both have similar normalized light REE (LREE; La) to heavy REE (HREE; Yb) ratios, with (La/Yb)N >1 (Fig. 9; Stern 1980). Both the field and geochemical data are consistent with the suggestion that the basaltic dykes and sills in the mixed mafic-felsic terranes flanking the ophiolites are a phase of the same igneous activity that formed the Rocas Verdes ophiolite complexes. The sheeted dyke complexes within the ophiolites imply 100% extension along oceanic type spreading centres, and the mixed maficfelsic terranes that flank the ophiolites also imply extension during intrusion of mafic magmas, but less than 100%. The boundary between the mafic ophiolite complexes and the flanking mixed mafic-felsic blocks is diffuse. The variation from 100% pre-Cretaceous continental rocks to 100% mafic rocks may occur across a narrow or relatively wide distance of several tens of kilometres. The pre-Cretaceous continental rocks are progressively more disrupted with the volumetric increase

ROCAS VERDES OPHIOLITES, SOUTH AMERICA

671

Fig. 6. Schematic sections across the Sarmiento and Tortuga ophiolite complexes, modified after de Wit & Stern (1978, 1981). (Note the abrupt igneous contact between the sheeted dykes and plagiogranites in the Sarmeinto complex and the continuous transition between dykes and massive diabase in the Tortuga complex, where plagiogranites are absent.) Acid (trondhjemite) crustal xenoliths also occur in the Sarmiento complex, but are absent in the Tortuga complex.

of mafic rocks above 60-70%. When mafic rocks dominate by >75%, they cause melting and remobilization of the host country rocks, leading to the reconstitution of continental crust through the formation of hybrid rock types (de Wit & Stern 1981). Field relations representing different stages of this process are best preserved along the flanks of the Sarmiento ophiolite complex, in the narrow northern part of the original basin (Fig. 5). Along the western flank of the Sarmiento complex, on Young Island, basaltic dyke swarms and gabbros intrude horizontally bedded silicic volcanic rocks of the Jurassic Tobifera Formation (de Wit & Stern 1981). In situ brecciation of the silicic rocks occurs near large mafic bodies and in areas of high concentrations of mafic dykes, creating raggedly outlined blocks of silicic lavas set in a new pale brown-green hybrid igneous matrix. This hybrid matrix is clearly a mixture of mafic and felsic components, although the exact physical process of mixing, whether by partial melting and/ or mechanical brecciation, remains obscure. However, the hybrid matrix is observed to reintrude as an independent magma, on a large scale, the entire

rock sequence of silicic rocks and mafic dykes and stocks (de Wit & Stern 1981).

Sarmiento ophiolite complex A vertical section 1-3 km thick of the Sarmiento ophiolite complex, the northernmost in the Rocas Verdes belt, is exposed in the Lolos and Encuentro fjords (Fig. 5; Dalziel et al 1974; de Wit & Stern 1978, 1981) of southern Chile. In these two fjords, coarse-grained gabbros and ferro-gabbros cut by basaltic dykes occur at sea level (Fig. 6). Dyke concentration increases vertically, as do the volumetric proportions of finer-grained diabases and leucocratic rocks. A sheeted dyke complex, consisting of 100% dykes (Fig. 2a), occurs at c. 1 km above sea level. The base of the sheeted dyke complex in the Sarmiento complex is an abrupt igneous contact marked by the intrusion of leucocratic rocks into the sheeted dykes by magmatic stoping. Approximately 300-500 m above this contact the first screen of extrusive pillow lavas and breccias is observed within the dyke complex. The volume per cent of these screens increase upwards, and the dyke complex is overlain by an

Fig. 7. Schematic cross-sections through three parts of the Rocas Verdes basin in the Early Cretaceous, modified after de Wit & Stern (1981): (a) in the northernmost part (north of 50°S) of the basin, where the igneous floor of the basin was continental in character; (b) in the area of the Sarmiento complex, where the igneous floor of the basin was intermediate between continental and oceanic in character; (c) in the southern part of the basin, in the area now preserved as the Tortuga complex, where the igneous floor of the basin was oceanic in character.

ROCAS VERDES OPHIOLITES, SOUTH AMERICA

Fig. 8. Titanium (Ti) v. zirconium (Zr) concentrations for the major Mesozoic igneous rock suites from southernmost South America, modified after de Wit & Stern (1981). Mafic dykes and sills (A; Stern 1980) intruding pre-Cretaceous continental crust flanking the Rocas Verdes ophiolite complexes (Fig. 2) plot along the same fractionation trend as do basalts from the ophiolites (bold line; Stern 1979), and also within the field of ocean-floor basalts (OFB; Pearce & Cann 1973). Silicic country rocks from Young Island (•; de Wit & Stern 1981) plot in the field of Jurassic Tobifera silicic volcanic rocks (Bruhn et al. 1978). Trondhjemites from within the Sarmiento complex (•; de Wit & Stern 1981) plot within the fields of both Tobifera silicic volcanic rocks and silicic plutons from the Patagonian batholith (Bruhn et al. 1978; Stern & Stroup 1982). In contrast, plagiogranites from the Sarmiento complex have significantly higher Zr contents (Stern 1979).

estimated 2 km of extrusive rocks, themselves cut by numerous mafic dykes (de Wit & Stern 1978). Two chemically and petrologically distinct types of leucocratic rocks may be distinguished in the Sarmiento complex (Saunders et al. 1979; Stern 1979; de Wit & Stern 1981; Elthon et al. 1984; Stern et al. 1992). One type, termed plagiogranites (Figs 6 and 8), is fine grained, occurs as both the massive body that intrudes the base of the sheeted dyke complex and also as dykes within this complex, and has relatively high incompatible trace element (Zr, Y, REE) concentrations compared with the mafic rocks of the complex (Figs 8 and 9), but with similar (La/Yb)N >1 and large negative Eu anomalies (Fig. 9). These plagiogranites have been shown to be derived from the mafic rocks of the complex by a process of closed-system crystal-liquid fractionation involving separation of mineral phases found in ferrogabbros (>20 wt% FeO; Stern 1979), which occur within the gabbro units of the Sarmiento complex. Ferro-basalts (>15wt% FeO and >2 wt% TiO2) and FeO- and TiO2-rich (>8 wt% FeO and >lwt% TiO2) intermediate (55-65 wt% SiO2) icelandites, which are products of less extensive

673

Fig. 9. Rare earth element (REE) concentrations, normalized to the REE content of a chondrite meteorite, of dykes and sills flanking the Sarmiento (•) and Tortuga (A) ophiolite complexes, as well as dykes and lavas from these ophiolites (Stern 1980).

fractionation, also occur as dykes in the sheeted dyke complex and cutting both gabbros and extrusive rocks. Approximately 30% of the dykes in the sheeted dyke complex of the Sarmiento complex consist of these evolved compositions. The second type of leucocratic rocks found within the Sarmiento complex is coarse grained and occurs as blocks, ranging from several centimetres to tens of metres across, within the gabbro unit of the complex ('Acid xenoliths' in Fig. 6). These rocks are referred to as trondhjemites (Fig. 8). They have distinctly lower Zr (Fig. 8), Y and HREE contents, and thus higher (La/Yb)N (de Wit & Stern 1981, p. 251, fig. 12), than plagiogranites in the Sarmiento complex. The trondhjemites are chemically similar to both the silicic volcanic rocks of the Jurassic Tobifera volcanic rocks and granitic plutons from the Patagonian batholith (Fig. 8), and they are interpreted as xenoliths of country rocks within the ophiolite. Trondhjemite xenoliths have been cut by dykes and are in places brecciated and remelted, producing hybrid magmas that have reintruded the gabbros of the Sarmiento complex. These relations resemble those in the remobilized country rocks flanking the ophiolite complexes.

674

C. R. STERN & M. J. DE WIT

The Sarmiento complex underwent hydrothermal 'ocean-floor' metamorphism, characterized by the growth of secondary minerals without the development of schistosity, prior to its uplift and exposure in the Andes (de Wit & Stern 1976; Stern et al, 1976; Elthon & Stern 1978; Stern & Elthon 1979; Elthon et al. 1984). The metamorphic overprint of this metamorphism on the pseudostratigraphy of the ophiolite exhibits a steep vertical gradient, passing from zeolite to amphibolite facies in 2 km (Fig. 10), and with equally abrupt variations in the extent of metamorphic replacement of igneous minerals and textures. Metamorphic facies boundaries are irregular and disequilibrium retrograde effects are common. The extent of metamorphic replacement is most intense just above and within the sheeted dyke complex, but decreases markedly within the gabbro unit of the ophiolite, probably because of restricted access of circulating sea water in the deeper plutonic levels of this ophiolite complex. Oxygen isotopes confirm that the extrusive rocks are enriched in 18O relative to fresh basalts as a result of interaction with sea water at relatively low temperatures (<500 °C). In contrast, deeper in the ophiolite, sheeted dykes, dykes cutting gabbros, and the gabbros themselves, are depleted in 18 O as a result of relatively high-temperature (>500 °C) interactions with sea water. Water-rock

ratios involved in this metamorphism decrease from between 15 and 90 in the extrusive section of the Sarmiento complex, to < 1 in the plutonic portion of the complex (Elthon et al. 1984). However, in detail, the effects of metamorphism are locally controlled by variations in permeability related to fault zones and the complex thermal history associated with the generation of the ophiolites. Metamorphism resulted in large-scale migration of KaO, Na2<3, CaO, Sr and Rb, but was nearly isochemical with regard to other elements such as SiO2, FeO, MgO, TiO2, Zr, Y and REE (Stern & Elthon 1979). The basalts of the Sarmiento complex exhibit a tholeiitic differentiation trend, and the least differentiated basalts are olivine normative. (La/Yb)N ratios of both basalts and more differentiated rocks are >1 (Fig. 9; Saunders et al. 1979; Stern 1980), similar to mafic magmas produced during the early stages of evolution of the Larson Harbour complex, a southern extension of the Rocas Verdes on South Georgia (Alabaster & Storey 1990; Storey & Alabaster 1991). The observed chemical variations within the Sarmiento complex, which includes ferro-basalt and intermediate icelandite dykes, as well as plagiogranites, are best explained by limited open-system magma chamber behaviour, followed by closed-system crystalliquid igneous fractionation (Stern 1979).

Fig. 10. Histograms illustrating the distribution of metabasalts and metagabbros, from the various stratigraphic units of the Sarmiento complex, among the metamorphic facies defined in the inset in the upper right of the figure (Elthon & Stern 1978; Stern & Elthon 1979). Filled bars represent relative abundance of rocks without relict higher temperature facies minerals, and diagonally ruled bars are rocks with such relict minerals. The left side of the diagram summarizes the observed petrographic variations in the extent of metamorphic replacement of original igneous minerals and textures.

ROCAS VERDES OPHIOLITES, SOUTH AMERICA Tortuga ophiolite complex The Tortuga ophiolite complex, the southernmost in the Rocas Verdes belt, is best exposed on Navarino and Milne Edwards islands (Fig. 5; de Wit & Stern 1978, 1981; Elthon & Ridley 1980). In the Tortuga complex, neither silicic plagiogranites nor trondhjemites have been observed, and the lower contact of the sheeted dyke complex is gradational. Dykes grade downwards into medium-grained diabase cut by later dykes (Fig. 6) and finally cumulate gabbros, which include both plagioclase-rich and olivine-bearing varieties. The igneous rocks of the Tortuga complex exhibit a more restricted chemical range than those of the Sarmiento complex, and no ferrobasalts, icelandites or silicic plagiogranites occur in the Tortuga complex. However, high-MgO (>10wt% MgO) dykes occur cutting the deeper gabbro and massive diabase level of this complex, below the sheeted dyke complex. These highMgO dykes are similar in composition to liquids derived by high degrees (25-30%) of partial melting of mantle Iherzolite at 20 kbar pressure, and they contain phenocrysts of highly aluminous picotite spinel consistent with high-pressure formation (Elthon 1979). They are interpreted as the parental mantle-derived magmas from which the more evolved basalts in the Tortuga complex formed. The observed chemical variations among the mafic dykes and lavas of the Tortuga complex may be modelled by open-system magmatic differentiation (Stern 1979). This model involves the periodic input of mantle-derived high-MgO basaltic magmas into the base of a magma chamber within which gabbros are forming along the floor and walls and from which differentiated (lower MgO) basalts are being erupted through the roof (Stern 1979; Elthon & Ridley 1980; Stern & de Wit 1980). Based on the determination of liquids in equilibrium with the cumulate minerals in the gabbros of the Tortuga complex, Elthon & Ridley (1980) concluded that the lack of systematic trends in FeO/MgO in vertical transects through the gabbros is consistent with open-system differentiation within a periodically refilled magma chamber. The relatively high FeO/MgO of the exposed gabbros implies that a significant volume of more magnesian cumulates must be associated with the Tortuga complex, but are not yet exposed. The tholeiitic basalts of the Tortuga complex have (La/Yb)N <1 (Fig. 9), similar to MORE, and they must have formed from melting of a mantle depleted in incompatible large ion lithophile elements (LILE), including LREE (Suarez 1977; Stern 1980). This LILE-depleted mantle source may have been the global asthenospheric source

675

of MORE, or alternatively depleted mantle produced by a process of extraction of successive melts during the formation of the mafic floor of the Rocas Verdes back-arc basin (Stern 1980). The former possibility is preferred, as it is consistent with the MORB-like initial Sr and Nd isotopic ratios of relatively fresh plagioclase and clinopyroxenes in cumulate gabbros (Fig. 11), when these ratios are corrected for the small amount of sea water that interacted with these gabbros (<25 wt% of the rock; Stern 1991). Dykes also have similar Nd isotopic composition to MORE, but significantly higher measured Sr isotopic ratios (Fig. 11) because of the extensive interaction and exchange of Sr with sea water that they experienced during the ocean-floor metamorphism of the Tortuga complex. Closure and deformation of the Rocas Verdes basin Closure and deformation of the Rocas Verdes basin took place during the mid-Cretaceous (Fig. 4c; Halpern 1973; Dalziel et al 1974; Bruhn & Dalziel 1977; Dott et al 1977; Bruhn 1979; Nelson et al. 1980; Dalziel 1981, 1986; Herve et al 1981; Nelson 1982; Stern et al 1992;

Fig. 11. Sr v. Nd initial isotopic composition of cumulate plagioclase and clinopyroxene in gabbros and a dyke from the Tortuga ophiolite compared with MORB, Patagonian mantle lithosphere based on analysis of mantle peridotite xenoliths in Cenozoic plateau basalts (Stern et al 1989, 1999), and plutons from the Patagonian batholith (Weaver et al 1990). Figure modified from Stern (1991). The original isotopic composition of the mafic magmas that formed the Tortuga complex, prior to sea-water alteration, was most similar to that of the cumulate gabbro plagioclase, and therefore MORB (Stern 1991).

676

C. R. STERN & M. J. DE WIT

Grunow 1993a, 1993b; Cunningham 1994, 1995). The style of deformation associated with the closure of the basin varies considerably from north to south. In the region of the Sarmiento ophiolite complex, where the Andean Cordillera trends north-south, the basin was uplifted and shortened, but the Rocas Verdes remained essentially autochthonous. Heterogeneous deformation took place along major vertical shear zones localized in the basement, batholith, and the mafic igneous complexes that floored the Rocas Verdes basin. More homogeneous deformation, involving folding and cleavage formation, occurred in the sedimentary rocks that filled the basin overlying the mafic complexes. Along the eastern margin of the Rocas Verdes ophiolites, the back-arc basin evolved into the foreland Magellan basin (Fig. 4c; Wilson 1991). Syntectonic sedimentation filled the Magellan basin with as much as 9 km of poorly sorted clastic sediments (Winn & Dott 1977; Winslow 1983; Soffia & Harambour 1988; Wilson 1991). The structural style of the late MesozoicCenozoic foreland fold and thrust belt that subsequently deformed and uplifted the Magellan basin is characterized by megascopic folds associated with bedding plane detachment faults as well as late cross-cutting thrusts. Mafic igneous rocks of the floor of the Rocas Verdes basin may have, in places, become thrust east across the Magellan basin during the late Mesozoic-Cenozoic deformation of this foreland basin, but the amount of thrusting was not substantial. In the region of the Tortuga ophiolite complex, where the strike of the Andes trends east-west, uplifting and thrusting of the Rocas Verdes basin across the foreland basin has been much more extensive. Collapse and uplift of this region during normal and strike-slip faulting in the Late Cretaceous has exposed a high-grade metamorphic window (Cordillera Darwin) similar to a metamorphic core complex (Nelson et al. 1980; Herve et al. 1981; Nelson 1982; Dalziel & Brown 1989; Kohn et al. 1993; Grunow et al. 1992; Cunningham 1993, 1994, 1995). Deformation and foreland basin development directly to the north of this southernmost east-west-striking extremity of the Andes (Klepeis 1994a), and also in both Isla de los Estados (Dalziel & Palmer 1979) and in South Georgia (Dalziel et al. 1974), involved a significant component of strike-slip faulting (Cunningham 1993, 1995; Klepeis 1994b), but was otherwise similar to that in the Magellan basin further to the north in the area east of the Sarmiento complex. The earliest documented post-extensional magmatic rocks are mid-Cretaceous (106 Ma) shoshonites erupted along shear zones flanking the eastern margin of the Rocas Verdes basin (Stern

et al. 1991). Younger plutons of the Patagonian batholith also intrude the uplifted and deformed sedimentary rocks of both the Rocas Verdes and Magallanes basins (Halpern 1973; Stern & Stroup 1982; Nelson et al. 1988; Weaver et al. 1990; Bruce et al. 1991). These relations are consistent with the closure, uplift and deformation of the basin having resulted from flattening of the angle of east-dipping subduction below the western continental margin of South America (Fig. 4c), as a result of either ridge subduction or a global increase in spreading and plate convergence rates (de Wit 1977; Dalziel 1986; Stern et al. 1991). Gealey (1980), in contrast, proposed that the Rocas Verdes basin was considerably wider than the present width of the ophiolite belt, and that west- and south-dipping subduction of the southwestern portion of mafic floor of this basin occurred during closure of the basin. In Gealey's model, when most of the marginal basin floor had been consumed, the continental margin of the Patagonian platform rode down the subduction zone and collided with the incoming magmatic arc. This may explain deepening of the Magellan forearc basin to the NE, the delivery of extremely thick conglomerate and marine turbidite sequences into this developing foreland basin, as well as the overthrusting of the Rocas Verdes along east- and north-verging structures into the foreland fold and thrust belt that deformed and uplifted this basin. However, there is no magmatic evidence for the existence of this south- and/or west-dipping Late Cretaceous subduction zone.

Discussion The Rocas Verdes ophiolite complexes preserve geological and geochemical information related to the progressive stages of rifting of continental crust in an extensional back-arc (marginal) basin setting, and petrological processes along midocean-type spreading centres. Here we summarize the most important implications for these processes and discuss the genesis of the Rocas Verdes ophiolites within the broader context of the breakup of the Gondwana supercontinent and the onset of opening of the South Atlantic (Fig. 12).

Progressive stages of continental rifting Observed north-to-south variations in the structure and petrochemistry of the Rocas Verdes ophiolite complexes imply a back-arc basin floor with crustal characteristics ranging from intermediate between continental and oceanic in nature in the north to more typically oceanic in the south (Fig. 7). These variations suggest that in the Early Cretaceous the width of the original Rocas Verdes

ROCAS VERDES OPHIOLITES, SOUTH AMERICA

677

Fig. 12. Selected igneous and tectonic elements associated with the fission process of southern Gondwana and the onset of spreading in the South Atlantic, modified after de Wit & Ransome (1992). The early formation of the Rocas Verdes (150 Ma) coincided with the early separation, by sea-floor spreading, between East Gondwana (AntarcticaIndia and Madagascar-Australia) and West Gondwana (Africa-South America), in a direction parallel to the Davie Fracture Zone (DFZ) to form the proto Indian Ocean and Weddell Sea (Reeves & de Wit 2000). The final emplacement of the Rocas Verdes (139 Ma) occurs close to the onset of the opening of the South Atlantic as expressed by the Parana-Etendeka flood basalts and seaward-dipping reflectors (SDR), interpreted as associated volcanic wedges, in a direction parallel to the Agulhas-Falkland Fracture Zone (AFFZ). It should be noted that, in general, the spatial relationship between the large igneous provinces along the convergent margin of southern Gondwana (inset: 1, Central Atlantic Magmatic Province (CAMP) (200 Ma); 2, St. Helena (120-130 Ma); 3, Tristan da Cunha (135 Ma); 4, Bouvet (180 Ma); 5, Kerguelen (120 Ma)) suggests that there appears to be a geodynamic relationship between the subduction process, back-arc lithosphere stretching, and the focusing of mantle plumes, which occur at regular spacing and similar distances behind the convergent plate margin. Reconstructed at 125 Ma (orthocentre: -40, 37.5) using Atlas Software (Cambridge, UK). The reconstruction does not show the position of the microplates now embedded in Antarctica and the Falkland Plateau, nor is the precise position of the Antarctic Peninsula known.

basin was possibly <50 km in the north and > 100 km in the south (Figs 7 and 12; de Wit 1977; de Wit & Stern 1978, 1981). We interpret the structural and petrochemical variations in the Rocas Verdes ophiolites as representing different stages of evolution of a back-arc basin, which formed as a result of the subtle interplay between back-arc mantle convection and the release of stress across the continental plate boundary (Fig.

7; de Wit & Stern 1981). Prior to the release of stress, heat transferred from mantle diapirs to the base of the crust caused widespread bimodal mafic and silicic volcanism across a >300km wide region of southern South America (Fig. 7a). With the release of stress, mantle-derived melts erupted to the surface along structural pathways, resulting in extensive basaltic volcanism and plutonism in a linear belt directly behind a convergent plate

678

C. R. STERN & M. J. DE WIT

boundary volcanic arc and the cessation of silicic volcanism. In the early stages of development of the original Rocas Verdes basin, mafic melts intruded continental crust over a diffuse area, causing extensive remobilization and reconstitution of the sialic continental crust (Figs 2b and 7b). Progressive localization of the zone of intrusion of mafic mantle-derived magmas eventually resulted in the development of oceanic-type spreading centres (Fig. 7c). Mantle-derived tholeiitic basalts were generated by melting of progressively more LILE-depleted (lower La/Yb), possibly deeper regions of the subcontinental mantle as the basin developed. Extreme alternative theories for the development of extensional back-arc basins call for, on the one hand, separation of an island arc from the continent by extension and formation of new oceanic crust along oceanic-type spreading ridges (Watts & Weissel 1975; Hawkins 1976), and, on the other hand, crustal subsidence and basification of sialic continental crust (oceanization) as a result of intrusion, contamination and replacement by mafic magmas over a diffuse zone (Katz 1973). The variability observed within the Rocas Verdes basin suggests that these extremes of sialic crustal reconstitution and sea-floor-type extension might operate separately, in sequence, and/or simultaneously during the development of a back-arc basin. Extension, geochemical reconstitution of pre-existing continental crust and magmatic stoping all played significant roles in the early stages of formation of the floor of the Rocas Verdes basin. Together these processes constitute a complex crustal dilation mechanism that is markedly different from sea-floor spreading occurring at mid-ocean ridges. However, in the later stages of the development of the basin, extension was clearly associated with oceanictype spreading centres. The various lithologies observed in and around the Rocas Verdes ophiolites might also be expected to form during the progressive stages of opening of major ocean basins and to currently underlie Atlantic-type continental margins. Variability in horizontal stress might be one of the important factors in determining the regionaltemporal response of continental crust to subduction and thermally or mechanically induced backarc mantle convection, and hence the mechanism of emplacement into the crust of mafic mantlederived magmas (de Wit & Stern 1981). The arcuate nature of the then Gondwana continental margin (de Wit 1977) was possibly one significant factor in causing north-to-south variations in stress across this continental margin, resulting in northto-south differences in the extent of development of the Rocas Verdes back-arc basin floor.

Implications for mid-ocean ridge spreading centres First-order features of the Rocas Verdes ophiolite complexes that have significant implications for understanding the structure, petrology and genesis of oceanic crust along mid-ocean-ridge spreading centres include the range of rock types and the difference in the ranges of rock types, observed in the Sarmiento and Tortuga ophiolites (Elthon 1979; Stern 1979; Elthon & Ridley 1980; Stern & de Wit 1980), the different pseudostratigraphy of each of these ophiolites (Fig. 6; Stern & de Wit 1980), and the nature of the hydrothermal 'ocean floor' metamorphic overprint in each of these ophiolites (Fig. 10; de Wit & Stern 1976; Stern et al 1976; Elthon & Stern 1978; Stern & Elthon 1979; Elthon et al. 1984). The data from the Chilean ophiolites illustrate the significance within spreading centre magma chambers of crystal-liquid igneous fractionation within both open and closed magma chambers (Stern 1979). Differentiation was much more extensive, and the resultant abundance of ferrogabbros, ferro-basalts, icelandites and plagiogranites is greater in the Sarmiento complex, within which more nearly closed-system crystal-liquid fractionation occurred, compared with the Tortuga complex, within which open-system fractionation occurred (Stern 1979; Elthon & Ridley 1980; Stern & de Wit 1980). These differences in the processes of igneous fractionation and the resulting range of rocks types that occur in each ophiolite complex also lead to the development of significantly different pseudostratigraphic structures observed in the Sarmiento and Tortuga ophiolite complexes, with a sharp igneous contact between plagiogranites and sheeted dykes in the former and a continuous transition between dykes, diabases and gabbros in the latter (Fig. 6). Along mid-oceanic spreading ridges, such differences may depend on spreading rates and magma supply (Stern & de Wit 1980; Dilek et al. 1998). Another significant observation from the Tortuga ophiolite complex is the presence of high-MgO dykes cutting the deeper diabase and gabbro units below the sheeted dyke complex (Elthon 1979; Elthon & Ridley 1980). Similar high-MgO magmas are the probable parental magmas for all the Chilean ophiolite complexes (Stern 1979), and Elthon (1979) demonstrated that similar highMgO magmas must be the parental liquid for other ophiolite complexes worldwide, as ophiolites typically contain large volumes of ultramafic cumulates overlying depleted mantle harzburgites from which high percentages of mantle melt have been removed. The high-MgO basalts in the Tortuga complex do not occur in the sheeted dyke

ROCAS VERDES OPHIOLITES, SOUTH AMERICA or extrusive units of this complex, indicating that these mantle-derived compositions mixed in the open-system spreading centre magma chamber around which the complex formed, but only more differentiated basalts were erupted through the roof of this chamber. The Rocas Verdes ophiolite complexes experienced 'ocean-floor' metamorphism prior to their uplift and exposure in the Andean cordillera. The overprint of this hydrothermal metamorphism, which is characterized by the growth of secondary minerals without the development of schistosity, exhibits a steep vertical metamorphic gradient although the extent of metamorphic replacement decreases markedly within the gabbro unit of the complex, probably because of restricted access of circulating sea water at deeper levels of the complex (Fig. 10; de Wit & Stern 1976; Stern et al 1976; Elthon & Stern 1978; Stern & Elthon 1979; Elthon et al. 1984). The common presence of disequilibrium textures in metabasalts from the Chilean ophiolites, produced by a sequence of metamorphic reactions occurring with falling temperature, complicates recognition of facies divisions. Detailed textural, mineralogical and geochemical studies indicate that ocean-floor metamorphism results from the interaction between the various rock units that make up the igneous pseudostratigraphy of the ocean floor and sea water circulating close to the spreading centre at which the ophiolites were generated, and that this interaction is strongly dependent on thermal gradient and permeability.

Causes of basin formation Within the larger tectonic context of the breakup of Gondwana and the opening of the South Atlantic (Fig. 12), the origin of the Rocas Verdes basin may be associated with subduction and thermally or mechanically induced mantle convection in the mantle wedge above the subducted slab (Bruhn et al 1978; de Wit & Stern 1981; Alabaster & Storey 1990), reduction in plate boundary forces associated with changes in plate configuration and absolute motion (Dalziel 1981, 1986; Storey & Alabaster 1991), and/or mantle plumes involved in the opening of the South Atlantic (Cox 1978, 1988; de Wit & Ransome 1992; Storey 1995; Storey & Kyle 1997; Dalziel et al. 2000; Storey et al. 2001). Bruhn et al. (1978) and de Wit & Stern (1981) suggested that the silicic volcanic rocks of the Jurassic Chon Aike and Tobifera Formations, which formed over a > 3 00 km wide region behind the convergent plate boundary magmatic arc (Figs 4 and 12), resulted from the first stage of crust-mantle interaction associated with mantle convection either

679

thermally or mechanically induced by subduction. This silicic volcanic field persisted from 188 to 153 Ma (Riley & Knight 2001). The subsequent development of structural pathways for mantlederived magmas to form the mafic floor of the Rocas Verdes basin, between at least 150 and 139 Ma (Stern et al. 1991; Mukasa & Dalziel 1996), might have resulted from release of stress across the continental margin as a result of either changing age and density of the oceanic crust being subducted or changes in plate convergence direction associated with spreading ridge subduction (Dalziel 1981, 1986; de Wit & Stern 1981; Storey & Alabaster 1991). Regionally, the back-arc spreading that led to the opening of the Rocas Verdes basin overlaps in time with other major volcano-tectonic events forming large igneous provinces (LIPs) in SW Gondwana (Fig. 12). These all relate to breakup of the supercontinent around the southern South Atlantic and Indian Oceans, which was clearly linked to deep mantle plumes and related elevated heat flow (Cox 1978, 1988; de Wit & Ransome 1992; Storey 1995; Storey & Kyle 1997; Hawkesworth et al. 1999; Storey et al. 2001). The Karoo LIP at c. 183 Ma, now the Bouvet (or Crozet) hotspot, and the Parana-Etendeka LIP at 138129 Ma, now the Tristan da Cunha (or Walvis) hotspot, are the best examples (Fig. 12). Both of these LIPs created flood basalts and extensive related silicic volcanic fields. In the case of the Parana LIP, the associated silicic volcanic province can be traced as far south as Uruguay (Kirstein et al. 2001). These LIPs are also temporally associated with the bimodal volcano-tectonic province forming the Tobifera and Chon Aike silicic volcanic rocks between 188 and 153 Ma, which in turn can be traced to the south coast of South Africa and to the eastern margins of the Antarctic peninsula (Fig. 12; Dalziel et al. 1987; de Wit & Ransome 1992; Elliot 1992; Storey 1995). The Parana-Etendeka igneous activity overlaps in time with the final emplacement of the Rocas Verdes, and the beginning of spreading in the South Atlantic at c. 132 Ma. Magnetic anomaly data for regions off the continental margins of Argentina and southern Africa imply high halfspreading rates of 2.4-3.5 cm a"1 for Mil (c. 132 Ma) to M2 (c. 124 Ma) South Atlantic oceanic crust (Nurnberg & Muller 1991). These magnetic anomalies are flanked on both margins by extensive volcanic wedges as indicated by seaward-dipping reflectors (SDRs in Fig. 12; Hinz et al. 1999; Bauer et al. 2000). These SDRs extend into the contemporaneous Parana-Etendeka flood basalts and are thus probably all expressions of the same Tristan da Cunha plume magmatism.

680

C. R. STERN & M. J. DE WIT

Clearly, the margins of this very large extensional volcanic-tectonic province rapidly developed more focused lines of spreading that, in the case of the South Atlantic, may have been channelled by subcrustal outward flow from the Tristan da Cunha plume (Ebinger & Sleep 1998). Dalziel et al. (2000) suggested that plumes rising from the deep mantle may have caused flattening of the subducting lithosphere below southern Gondwana, generating the early Mesozoic Gondwanide fold belt prior to penetrating the subducting slab and initiating the breakup of the supercontinent. Alternatively, the geographical distribution of the LIPs, which formed at regular spacing and approximately similar distances behind the convergent plate margin along southern Gondwana (inset in Fig. 12), suggests that subduction processes may have induced the formation of the inferred mantle plumes that initiated the opening of the southern oceans (Burke 2001) and the contemporaneous emplacement of the Rocas Verdes. How all these geodynamic processes are ultimately related remains unresolved. Between 1973 and 1977, I. Dalziel provided both funding and logistic support within the framework of his larger research programme in the Scotia Arc. Others who collaborated in the field include K. Palmer, A. Skewes, J. Stroup, D. Elthon, R. Bruhn, E. Godoy, M. Dobbs and R. Fuenzalida, and in the laboratory include S. Mukasa, J. Lawrence and R. Kay. Reviews by I. Dalziel, A. Saunders and Y. Dilek helped improve the final manuscript. Financial and logistic support for this work was also provided by the Empresa Nacional del Petroleo (ENAP) of Chile, and the National Science Foundation. Bob Marley provided the music.

References ALABASTER, T. & STOREY, B.C. 1990. Modified Gulf of California model for South Georgia, north Scotia Ridge, and implications for the Rocas Verdes backarc basin, southern Andes. Geology, 18, 497-500. BAUER, K., NEBEN, S. & SCHRECKENBERGER, B. ET AL. 2000. Deep structure of the Namibian continental margin as derived from integrated geophysical studies. Journal of Geophysical Research, 105, 25829-25853. BRUCE, R.M., NELSON, E.P., WEAVER, S.G. & Lux, D.R. 1991. Temporal and spatial variations in the southern Patagonian batholith; constraints on magmatic arc development. In: HARMON, R.S. & RAPELA, C.W. (eds) Andean Magmatism and its Tectonic Setting. Geological Society of America, Special Papers, 265, 1-12. BRUHN, R.L. 1979. Rock structures formed during backarc basin deformation in the Andes of Tierra del Fuego. Geological Society of America Bulletin, 90, 998-1012. BRUHN, R.L. & DALZIEL, I.W.D. 1977. Destruction of the early Cretaceous marginal basin in the Andes of

Tierra del Fuego. In: TALWANI, M. & PITMAN, W.C. Ill (eds) Island Arcs, Deep Sea Trenches and Back-Arc Basins. American Geophysical Union, Maurice Ewing Series, 1, 395-405. BRUHN, R.L., STERN, C.R. & DE WIT, MJ. 1978. Field and geochemical data bearing on the development of a Mesozoic volcano-tectonic rift zone and backarc basin in southernmost South America. Earth and Planetary Science Letters, 41, 32-46. BURKE, K. 2001. Since 200 Ma nearly all large igneous provinces (LIPs) have been emplaced above deep mantle low velocity volumes (DMLWs). EOS Transactions, American Geophysical Union, 82(47), F1134. Cox, K.G. 1978. Flood basalts, subduction and the break-up of Gondwanaland. Nature, 274, 47-49. Cox, K.G. 1988. The Karoo province. In: MACDOUGALL, J.D. (ed.) Continental Flood Basalts. Kluwer, Dordrecht, 239-271. CUNNINGHAM, W.D. 1993. Strike-slip faults in the southernmost Andes and the development of the Patagonian orocline. Tectonics, 12, 169-186. CUNNINGHAM, W.D. 1994. Uplifted ophiolitic rocks on Isla Gordon, southernmost Chile: implications for the closure history of the Rocas Verdes marginal basin and the tectonic evolution of the Beagle Channel region. Journal of South American Earth Sciences, 7, 135-147. CUNNINGHAM, W.D. 1995. Orogenesis at the southern tip of the Americas: the structural evolution of the Cordillera Darwin metamorphic complex, southernmost Chile. Tectonophysics, 244, 197-229. DALZIEL, I.W.D. 1981. Back-arc extension in the southern Andes: a review and critical reappraisal. Philosophical Transactions of the Royal Society of London, Series A, 300, 319-335. DALZIEL, I.W.D. 1986. Collision and cordilleran orogenesis: an Andean perspective. In: COWARD, M.P. & RIES, A.C. (eds) Collision Tectonics. Geological Society, London, Special Publications, 19, 389-404. DALZIEL, I.W.D. & BROWN, R.L. 1989. Tectonic denudation of the Darwin metamorphic core complex in the Andes of Tierra del Fuego southernmost Chile: implications for Cordilleran orogenesis. Geology, 17, 699-703. DALZIEL, I.W.D. & ELLIOT, D. 1982. West-Antarctica— problem child of Gondwanaland. Tectonics, 1, 3-19. DALZIEL, I.W.D. & PALMER, K.F. 1979. Progressive deformation and orogenic uplift at the southern extremity of the Andes. Geological Society of America Bulletin, 90, 259-280. DALZIEL, I.W.D., DE WIT, MJ. & PALMER, K.F. 1974. Fossil marginal basin in the southern Andes. Nature, 250, 291-294. DALZIEL, I.W.D., STOREY, B.C., GARRETT, S.W., GRUNOW, A.M., HERROD, L.B.D. & PANKHURST, R.J. 1987. Extensional tectonics and the fragmentation of Gondwanaland. In: COWARD, M.P., DEWEY, J.F. & HANCOCK, P.L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 433-441. DALZIEL, I.W.D., LAWVER, LA. & MURPHY, J.B. 2000.

ROCAS VERDES OPHIOLITES, SOUTH AMERICA Plumes, orogenesis, and super-continental fragmentation. Earth and Planetary Science Letters, 178, 1-11. DE WIT, MJ. 1977. The evolution of the Scotia Arc as a key to the reconstruction of Gondwanaland. Tectonophysics, 37, 53-81. DE WIT, MJ. & RANSOME, I.G.D. 1992. Regional inversion tectonics along the southern margin of Gondwana. In: DE WIT, MJ. & RANSOME, I.G.D. (eds) Inversion Tectonics of the Cape Fold Belts, Karoo and Cretaceous Basins of Southern Africa. Balkema, Rotterdam, 15-21. DE WIT, MJ. & STERN, C.R. 1976. Ocean floor metamorphism, seismic layering and magnetism. Nature, 264, 615-619. DE WIT, MJ. & STERN, C.R. 1978. Pillow talk. Journal of Volcanology and Geothermal Research, 4, 55-80. DE WIT, MJ. & STERN, C.R. 1981. Variations in the degree of crustal extension during the formation of a back-arc basin. Tectonophysics, 72, 229-260. DILEK, Y., MOORES, E.M. & FURNES, H. 1998. Structure of modern oceanic crust and ophiolites and implications for faulting and magmatism at oceanic spreading centers. In: BUCK, W.R. ET AL. (eds) Faulting and Magmatism at Mid-ocean Ridges. Geophysical Monograph, American Geophysical Union, 106, 219-265. DOTT, R.H., WINN, R.D., DE WIT, MJ. & BRUHN, R.L. 1977. Tectonic and sedimentary significance of Cretaceous Tekenika Beds of Tierra del Fuego. Nature, 266, 620-622. EBINGER, CJ. & SLEEP, N.H. 1998. Cenozoic magmatism throughout East Africa resulting from impact of a single plume. Nature, 395, 788-791. ELLIOT, D.H. 1992. Jurassic magmatism and tectonism associated with Gondwanaland break-up: an Antarctic perspective. In: STOREY, B.C., ALABASTER, T. & PANKHURST, RJ. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publications, 68, 165-184. ELTHON, D. 1979. High magnesia liquids as the parental magma for ocean floor basalts. Nature, 278, 514-517. ELTHON, D. & RIDLEY, WJ. 1980. The petrology of the Tortuga ophiolite complex, southern Chile. In: PANAYIOTOU, A. (ed.) Ophiolites, Proceedings International Ophiolite Symposium, Cyprus. Cyprus Geological Survey, Nicosia, 507-513. ELTHON, D. & STERN, C.R. 1978. Metamorphic petrology of the Sarmiento ophiolite complex, Chile. Geology, 6, 464-468. ELTHON, D., LAWRENCE, J.R., HANSON, R.E. & STERN, C.R. 1984. Modelling of oxygen isotope data from the Sarmiento ophiolite complex, Chile. In: GASS, I., LIPPARD, S. & SHELTON, A. (eds) Ophiolites and Oceanic Lithosphere. Blackwell Scientific, London, 185-197. FORSYTH, R.D. 1982. The late Paleozoic to early Mesozoic evolution of southern South America: a plate tectonic interpretation. Journal of the Geological Society, London, 139, 671-682. FUENZALIDA, R. & COVACEVICH, V. 1988. Volcanismo y bioestratigrafia del Jurasico y Cretacico Inferior en la

681

Cordillera Patagonica, Region de Magallanes, Chile. In: Chile V Congreso Geologico Chileno. Santiago, 3,H159-H183. GEALEY, W.K. 1980. Ophiolite obduction mechanisms. In: PANAYIOTOU, A. (ed.) Ophiolites, Proceedings International Ophiolite Symposium, Cyprus. Cyprus Geological Survey, Nicosia, 228-243. GRUNOW, A.M. 1993a. Creation and destruction of Weddell Sea floor in the Jurassic. Geology, 21, 647-650. GRUNOW, A.M. 1993b. New paleomagnetic data from the Antarctic Peninsula and their tectonic implications. Journal of Geophysical Research, 98, 13815-13833. GRUNOW, A.M., DALZIEL, I.W.D., HARRISON, T.M. & HEIZLER, M.T. 1992. Structural geology and geochronology of subduction complexes along the margin of Gondwanaland: new data from the Antarctica Peninsula and southernmost Andes. Geological Society of American Bulletin, 104, 1497-1514. GUST, D.A., BIDDLE, K.T., PHELPS, D.W. & ULIAN, M.A. 1985. Associated middle to late Jurassic volcanism and extension in southern South America. Tectonophysics, 116, 223-253. HALPERN, M. 1973. Regional geochronology of Chile south of 50°S latitude. Geological Society of America Bulletin, 84, 2407-2422. HANSON, R.E. & WILSON, TJ. 1991. Submarine rhyolitic volcanism in a Jurassic proto-marginal basin, southern Andes, Chile and Argentina. In: HARMON, R.S. & RAPELA, C.W. (eds) Andean Magmatism and its Tectonic Setting. Geological Society of America, Special Papers, 265, 13-28. HAWKESWORTH, C., KELLEY, S., TURNER, S., LE ROEX, A. & STOREY, B.C. 1999. Mantle processes during Gondwana break up and dispersal. Journal of African Earth Sciences, 28, 239-261. HAWKINS, J.W. JR 1976. Petrology and geochemistry of basaltic rocks of the Lau basin. Earth and Planetary Science Letters, 28, 283-297. HERVE, F., NELSON, E., KAWASHITA, K. & SUAREZ, M. 1981. New isotopic ages and the timing of orogenic events in the Cordillera Darwin, southernmost Chilean Andes. Earth and Planetary Science Letters, 55, 257-265. HERVE, F., LOSKE, W.P., MILLER, H. & PANKHURST, RJ. 1991. Chronology of provenance, deposition and metamorphism in the southern limb of the Scotia Arc. In: THOMSON, M.R.A., CRANE, J.A. & THOMSON, J.W. (eds) Geologic Evolution of Antarctica. Cambridge University Press, Cambridge, 429-435. HERVE, M., SUAREZ, M. & PUIG, A. 1984. The Patagonian batholith south of Tierra del Fuego, Chile: timing and tectonic implications. Journal of the Geological Society, London, 141, 909-917. HINZ, K., NEBEN, S., SCHRECKENBERGER, B., ROESSER, H.A., BLOCK, M., GONZALEZ DE SOUZA, K. & MEYER, JH. 1999. The Argentine continental margin north of 48°S: sedimentary successions and volcanic activity during break-up. Marine Petrology and Geology, 16, 1-25. KATZ, H.R. 1964. Some new concepts on geosynclinal

682

C. R. STERN & M. J. DE WIT

development and mountain building at the southern end of South America. In: 22nd International Geological Congress, India, Proceedings. New Delhi, 4, 242-255. KATZ, H.R. 1972. Plate tectonics and orogenic belts in the south-east Pacific. Nature, 237, 331-332. KATZ, H.R. 1973. Contrasts in tectonic evolution of orogenic belts in the southwest Pacific. Journal of the Royal Society of New Zealand, 3, 333-351. KAY, S.M., RAMOS, V.A., MPODOZIS, C. & SRUODGA, P. 1989. Late Paleozoic to Jurassic silicic magmatism at the Gondwana margin: analogy to the Middle Proterozoic in North America? Geology, 17, 324-328. KIRSTEIN, L.A., KELLEY, S., HAWKSWORTH, C., TURNER, S., MANTOVANI, M. & WIJBRANS, J. 2001. Protracted felsic magmatic activity associated with the opening of the South Atlantic. Journal of the Geological Society, London, 158, 583-592. KLEPEIS, K.A. 1994a. Relationship between uplift of the metamorphic core of the southernmost Andes and shortening in the Magallanes foreland fold and thrust belt, Tierra del Fuego, Chile. Tectonics, 13, 882-904. KLEPEIS, K.A. 1994b. The Magallanes and Deseado fault zones: major segments of the South AmericanScotia transform plate boundary in southernmost South America, Tierra del Fuego. Journal of Geophysical Research, 99, 22001-22014. KOHN, M.J., SPEAR, F.S. & DALZIEL, I.W.D. 1993. Metamorphic P—T paths from the Cordillera Darwin, a core complex in Tierra del Fuego, Chile. Journal of Petrology, 34, 519-542. MPODOZIS, C. & KAY, S.M. 1990. Provincias magmaticas acidas y evolucion tectonica de Gondwana, Andes Chilenos (28°-31°S). Revista Geologica de Chile, 17, 143-180. MUKASA, S.B. & DALZIEL, I.W.D. 1996. Southern Andes and South Georgia Island, North Scotia Ridge: zircon U/Pb and muscovite 40Ar/39Ar age constraints on tectonic evolution of southwestern Gondwanaland. Journal of South American Earth Sciences, 9, 349-365. NELSON, E.P. 1982. Post tectonic uplift of the Cordillera Darwin orogenic core complex: evidence from fission track geochronology and closing temperature time relationships. Journal of the Geological Society, London, 139, 755-761. NELSON, E.P., DALZIEL, I.W.D. & MIMES, A.G. 1980. Structural geology of the Cordillera Darwin: collisional style orogenesis in the southernmost Chilean Andes. Eclogae Geologica Helvetia, 73, 727-751. NELSON, E.P., BRUCE, R.M., ELTHON, D., KAMMER, D. & WEAVER, S. 1988. Regional lithological variations in the Patagonian batholith. Journal of South American Earth Sciences, 1, 239-247. NURNBERG, D. & MULLER, R.D. 1991. The tectonic evolution of the South Atlantic from the Jurassic to the present. Tectonophysics, 191, 27-53. PEARCE, J.A. & CANN, J. 1973. Tectonic setting of basic rocks using trace element analysis. Earth and Planetary Science Letters, 19, 290-300. REEVES, C. & DE WIT, M.J. 2000. Making ends meet in Gondwana: retracing the transforms of the Indian Ocean and reconnecting continental shear zones.

Terra Nova, 12, 272-280. RILEY, T.R. & KNIGHT, K.B. 2001. Age of pre-break up Gondwana magmatism. Antarctic Science, 13, 99-110. SAUNDERS, A.D., TARNEY, J., STERN, C.R. & DALZIEL, I.W.D. 1979. Geochemistry of Mesozoic marginal basin floor igneous rocks from southern Chile. Geological Society of America Bulletin, 90, 237-258. SELVERSTONE, J.E. & STERN, C.R. 1983. Petrochemistry and recrystallisation history of granulite xenoliths from the Pali-Aike volcanic field Chile. American Mineralogist, 68, 1102-1112. SKEWES, M.A. & STERN, C.R. 1979. Petrology and geochemistry of alkali basalts and ultramafic inclusions from the Pali-Aike volcanic fields in southern Chile and the origin of the Patagonian plateau lavas. Journal of Volcanology and Geothermal Research, 6, 3-25. SOFFIA, J.M. & HARAMBOUR, P. 1988. Estructuras en el cinturon plegado y fallado de Ultima Esperanza, Magallanes, Chile. In: V Congreso Geologico Chileno. Santiago, A417-A431. STERN, C.R. 1979. Open and closed system igneous fractionation within two Chilean ophiolites and the tectonic implications. Contributions to Mineralogy and Petrology, 68, 243-258. STERN, C.R. 1980. Geochemistry of Chilean ophiolites: evidence of the compositional evolution of the mantle source of back-arc basin basalts. Journal of Geophysical Research, 85, 955-966. STERN, C.R. 1991. Isotopic composition of late Jurassic and early Cretaceous mafic igneous rocks from the southernmost Andes: implications for sub-Andean mantle. Revista Geologica de Chile, 18, 15-23. STERN, C.R. & DE WIT, M.J. 1980. The role of spreading centre magma chambers in the formation of Phanerozoic oceanic crust: evidence from Chilean ophiolites. In: PANAYIOTOU, A. (ed.) Ophiolites, Proceedings International Ophiolite Symposium, Cyprus. Cyprus Geological Survey, Nicosia, 497-506. STERN, C.R. & DE WIT, M.J. 1997. The Rocas Verdes 'Greenstone belt', southernmost South America. In: DE WIT, M.J. & ASHWAL, L.D. (eds) Greenstone Belts. Clarendon Press, Oxford, 791-801. STERN, C.R. & ELTHON, D. 1979. Vertical variations in the effects of hydrothermal metamorphism in Chilean ophiolites their implications for ocean floor metamorphism. Tectonophysics, 55, 179-213. STERN, C.R. & STROUP, J. 1982. The petrochemistry of the Patagonian batholith, Ultima Esperanza, Chile. In: CRADDOCK, C. (ed.) Antarctic Geoscience. University of Wisconson Press, Madison, 131-142. STERN, C.R., DE WIT, M.J. & LAWRENCE, J.R. 1976. Igneous and metamorphic processes associated with formation of Chilean ophiolites and their implications for ocean floor metamorphism, seismic layering and magnetism. Journal of Geophysical Research, 81, 4370-4380. STERN, C.R., SKEWES, M.A., SAUL, S.L. & FUTA, K. 1989. Garnet peridotite xenoliths of the Pali-Aike alkali basalts of southernmost South America. In: Ross, J. (ed.) Kimberlites and Related Rocks—Their Mantle and Crustal Settings. Geological Society of

ROCAS VERDES OPHIOLITES, SOUTH AMERICA Australia Special Publications, 14, 735-744. STERN, C.R., MOHSENI, P.P. & FUENZALIDA, R. 1991. Petrochemistry and tectonic significance of Lower Cretaceous Barros Arrana Formation basalts, southernmost Chilean Andes. Journal of South American Earth Sciences, 4, 331-342. STERN, C.R., MUKASA, S.B. & FUENZALIDA, R. 1992. Age and petrogenesis of the Sarmiento ophiolite complex of Southern Chile. Journal of South American Earth Sciences, 6, 97-104. STERN, C.R., KILIAN, R., OLKER, B., HUARI, E.H. & KYSER, T.K. 1999. Evidence from mantle xenoliths for relatively thin (<100km) continental lithosphere below Phanerozoic crust of southernmost South America. Lithos, 48, 217-235. STOREY, B.C. 1995. The role of mantle plumes in continental break-up. Nature, 377, 301-308. STOREY, B.C. & ALABASTER, T. 1991. Tectonomagmatic controls on Gondwana break-up models: evidence from the proto-Pacific margin of Antarctica. Tectonics, 10, 1274-1288. STOREY, B.C. & KYLE, P.R. 1997. An active mantle mechanism for Gondwana breakup. South African Journal of Geology, 100, 283-290. STOREY, B.C. & MAIR, B.F. 1982. The composite floor of the Cretaceous back-arc basin of South Georgia. Journal of the Geological Society, London, 139, 729-737. STOREY, B.C., MAIR, B.F. & BELL, C.M. 1977. The occurrence of Mesozoic ocean floor and ancient continental crust on South Georgia. Geological Magazine, 114, 203-208. STOREY, B.C., LEAT, P.T. & FERRIS, J.K. 2001. The location of mantle plume centres during the initial stages of Gondwana breakup. In: ERNST, R. & BUCHAN, K.L. (eds) Mantle Plumes: Their Identification Through Time. Geological Society of America, Special Papers, 352, 71-80. SUAREZ, M. 1977. Aspectos geoquimicos de complejo ofiolitico Tortuga en la Cordillera Patagonica del sur, Chile. Revista Geologica de Chile, 4, 3-14. SUAREZ, M. 1979. A late Mesozoic island arc in the southern Andes, Chile. Geological Magazine, 116, 181-190. SUAREZ, M. & PETTIGREW, T.H. 1976. An upper Mesozoic island-arc-back-arc basin system in the southern Andes and South Georgia. Geological Magazine, 113, 304-328.

683

SUAREZ, M., NARANJO, J.A. & PUIG, A. 1990. Mesozoic 'S-like' granites of the central and southern Andes: a review. In: KAY, S.M. & RAPELA, C.W. (eds) Plutonism from Antarctica to Alaska. Geological Society of America Special Papers, 241, 27-32. TANNER, P.W.G. & REX, D.C. 1979. Timing of events in an Early Cretaceous island arc-marginal basin system on South Georgia. Geological Magazine, 116, 167-179. TARNEY, J., DALZIEL, I.W.D. & DE WIT, M.J. 1976. Marginal basin Rocas Verdes complex from southern Chile: a model for Archaean greenstone belt formation. In: WINDLEY, B.F. (ed.) The Early History of the Earth. Wiley, London, 131-146. WATTS, A.B. & WEISSEL, J.K. 1975. Tectonic history of the Shokoku marginal basin. Earth and Planetary Science Letters, 25, 239-250. WEAVER, S.G., BRUCE, R.M., NELSON, E.P., BRUECKNER, H.K. & LEHURAY, A.P. 1990. The Patagonian batholith at 48°S latitude, Chile; geochemical and isotopic variations. In: KAY, S.M. & RAPELA, C.W. (eds) Plutonism from Antarctica to Alaska. Geological Society of America, Special Papers, 241, 33-50. WILSON, T.J. 1991. Transition from back-arc to foreland basin development in the southernmost Andes: stratigraphic record from the Ultima Esperanza District, Chile. Geological Society of America Bulletin, 103,98-111. WINN, R.D. 1978. Upper Mesozoic flysch of Tierra del Fuego and South Georgia Island: a sedimentological approach to lithospheric plate restoration. Geological Society of America Bulletin, 89, 533-547. WINN, R.D. & DOTT, R.H. 1977. Large-scale tractionproduced structures in deep-water fan-channel conglomerates in Southern Chile. Geology, 5, 41-44. WINN, R.D. & DOTT, R.H. 1978. Submarine-fan turbidites and resedimented conglomerates in a Mesozoic arc-rear marginal basin in Southern South America. In: STANDLEY, D.J. & KELLING, G. (eds) Sedimentation in Submarine Canyons, Fans and Trenches. Dowden, Hutchinson and Ross, Stroudsburg, PA, 362-373. WINSLOW, M.A. 1983. Clastic dike swarms and the structural evolution of the foreland fold and thrust belts of the southern Andes. Geological Society of America Bulletin, 94, 1073-1080.

This page intentionally left blank

Proterozoic ophiolites of the Arabian Shield and their significance in Precambrian tectonics YILDIRIM DILEK 1 & ZULFIQAR AHMED 2 1

Department of Geology, 116 Shideler Hall, Miami University, Oxford, OH 45056, USA (e-mail: [email protected]) ^Department of Earth Sciences, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia Abstract: Neoproterozoic ophiolites, ranging in age from c. 870 Ma to c. 627 Ma, occur in several discrete suture and/or fault zones within the Arabian Shield and display a record of riftdrift, sea-floor spreading and collision tectonics during the evolution of the East African Orogen. The ophiolites within the Yanbu and Bir Umq suture zones in the west are among the oldest (870-740 Ma) in the Shield, locally show a Penrose-type complete pseudostratigraphy, and have chemical compositions typical of modern forearc oceanic crust. They are spatially associated with coeval and younger volcanic arc assemblages and were incorporated into the Arabian Shield during a series of collisional events that amalgamated these ensimatic arc terranes. The ophiolites of the Hulayfah-Ruwah suture zone in the central Arabian Shield are coeval with and/or slightly younger (c. 843-821 Ma) than the ophiolites in the west and probably developed in a rifted ensimatic arc system that evolved as a volcanic archipelago near the Afif continental plate. Younger ophiolites (c. 694 Ma) of the Halaban and Al Amar suture zones in the eastern Arabian Shield were incorporated into a subduction-accretion complex that evolved at the Andean-type active margin along the eastern edge (in present coordinate system) of the Afif continental plate. The Halaban suture zone ophiolites represent forearc oceanic crust, whereas the Al Amar suture zone ophiolites are scraped-off fragments of Mozambique ocean floor, seamounts and/or ocean island(s); the Abt Schist between them corresponds to a Franciscan-type accretionary prism of the 'Halaban' subduction zone. The incorporation of these ophiolites and the continental plates (Afif and Ar Rayn) into the Arabian Shield during 640-620 Ma marks a major shift in the direction of convergence (from northerly to westerly) during the assembly of the Shield and distinct episodes of continental collisions during closure of the Mozambique Ocean. The ophiolites of the Nabitah-Hamdah fault zone within the Asir terrane are the youngest (c. 627 Ma) in the Shield, post-collisional in origin, display mid-ocean ridge basalt chemical affinity, and represent Ligurian-type oceanic crust developed in an intracontinental pararift zone. The ophiolite tectonics of the Arabian Shield indicates an eastward progression of continental growth through time as the East African Orogen was built during the late Neoproterozoic, following the breakup of Rodinia.

The Arabian-Nubian Shield is a composite mosale of ancient oceanic crust, volcanic arc terranes and pre-lOOOMa continental crust that was assembled during the Pan-African period in the aftermath of the breakup of the supercontinent Rodiniac. 900-800 Ma (Gass 1981; Kroner 1985; Shackleton 1986; Brown et al. 1989). Ophiolites in this mosaic occur in curvilinear belts (Fig. 1) with distinctive lithotectonic assemblages, age distributions and chemical fingerprints, indicating discrete pulses of oceanic crust generation and its incorporation into the continental margins of Afro-Arabia throughout the late Proterozoic (Coleman 1984). These ophiolite-bearing belts have been identified as suture zones that mark the sites of tectonic amalgamation of disparate terranes

predominantly of volcanic arc origin (Berhe 1990; Stern et al 1990). However, the tectonic setting of origin and the root zones of these Proterozoic ophiolites are uncertain, mainly because of the reworking of the Shield crust as a result of multipie deformational episodes (i.e. Church 1988). The apparent lack of a Penrose-type pseudostratigraphy in many Precambrian ophiolites has led some researchers to suggest that either 'true ophiolites' did not form or fundamentally different plate tectonic processes operated in the Earth during the Precambrian (see Helmstaedt & Scott 1992 for discussion and related references), Recent studies on ophiolites of the ArabianNubian Shield have shown, however, that the Proterozoic ophiolites are extensive there and that

From: DlLEK, Y. & ROBINSON P. T. (eds) 2003. Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 685-700. 0305-8719/037$ 15 © The Geological Society of London 2003.

686

Y. DILEK & Z. AHMED

Fig. 1. Simplified geological map of the Arabian-Nubian Shield showing the distribution of Neoproterozoic ophiolites, volcanic arc terranes (juvenile Neoproterozoic crust), and pre-Neoproterozoic crust. Data sources: Vail (1985); Pallister et al. (1988); Berhe (1990); Kroner et al. (1992); Johnson & Kattan (2001); Reischmann (2000); Stern (2002). Abbreviations for ophiolite names are as follows. Egypt: H, Allaqi-Heiani; Br, Barramiya; F, Fawkir; G, Gerf; Gh, Ghadir. Eritrea: Hg, Hagar terrane. Ethiopia: Bd, Baruda; DT, Daro Tekli belt; TD, Tulu Dimtu; Zg, Zager belt. Saudi Arabia: A, Arjah; AA, Al'Ays; AB, Al Bijadiyah; BF, Bir Fuqayr; BT, Bir Tuluha; Bu, Bir Umq; DZ, Darb Zubaydah; HB, Halaban; JB, Jabal Bitran; JE, Jabal Ess; JG, Jabal Ghurrab; JM, Mogheira; JTh, Jabal Thurwah; NB, Nabitah; TL, Tathlith. Sudan: AD, Atmur-Delgo; HS, Hamisana; I, Ingessana; K, Keraf; Kb, Kabus; M, Meritri; OS, Oshib; OSH, Onib-Sol Hamed.

they are structurally, chemically and isotopically similar to both Phanerozoic ophiolites and the modern oceanic crust of subduction zone environments (i.e. Stern 1981, 2002; Stoeser & Camp 1985; Pallister et al. 1988; Berhe 1990; Quick 1990). Thus, it is clear that the nature of post-900 Ma oceanic crust is similar to that of Phanerozoic ophiolites, unlike the Pre-Rodinian

(pre-1 Ga) ophiolites, which have much thicker crustal units (Moores 2002). Nevertheless, structural field observations and systematic geochemical, geochronological and isotopic data from the Arabian Shield ophiolites are severely limited, hampering a better understanding of the crustal and mantle processes involved in the evolution of Neoproterozoic oceanic lithosphere. A strict inter-

PROTEROZOIC ARABIAN SHIELD OPHIOLITES pretation of ophiolites in the sense of the 1972 Penrose definition (Anonymous 1972) has also hindered our progress in Precambrian ophiolite research. In this paper we present a systematic overview of the occurrence of Proterozoic ophiolites within the Arabian Shield as exposed in Saudi Arabia, and define their structure and tectonic origin in view of the newly developed classification scheme for ophiolites (Dilek 2003; Table 1). We then discuss the significance of the Arabian Shield ophiolites and their evolution in the context of Precambrian tectonics and continental growth. A more comprehensive treatment of the Arabian Shield ophiolites, including their geochemistry and palaeogeography, will be presented elsewhere (Ahmed & Dilek, in preparation).

687

of the Andean-type magmatic arc complex at the southern edge of the terrane starting around 680 Ma. The Jabal Ess and Al'Ays massifs constitute the main ophiolite occurrences within the Yanbu suture zone (Table 1); a minor ophiolite fragment occurs at Bir Fuqayr between these two ophiolites (Fig. 2; Ahmed 2001). The Yanbu suture zone ophiolites continue to the SW across the Red Sea and connect with the Onib-Sol Hamid ophiolite of the Onib suture zone in the Nubian Shield (Fig. 1; Sultan et al. 1993). The Onib suture zone separates the Gerf terrane in the north from the Gebeit terrane in the south in Sudan (e.g. Nassief et al 1984; Al-Shanti & El-Mahdy 1989; Reischmann 2000).

Jabal Ess ophiolite. The Jabal Ess ophiolite occurs as an east-west-trending massif in the northern part of the Yanbu suture zone. The ophiolite is Precambrian ophiolites in Saudi Arabia occur underlain by the Farri Group, which consists of a along NE- and NNW-trending linear belts that sequence of schists derived from andesite, dacite, have been identified as suture zones separating rhyolite, shale, sandstone, greywacke and concomposite terranes with different lithostratigraphic glomerate protoliths (Delfour 1979a), and it is units and evolutionary histories (e.g. Abdel-Gawad unconformably overlain by volcanic-volcaniclas1970; Bakor et al 1976; Nassief et al 1984; Fig. tic rocks of the Al'Ays Group. The ophiolitic 1). This distribution of the Precambrian ophiolites sequence includes harzburgite, dunite, serpentiand crystalline basement rocks has been affected nite, wehrlite, pyroxenite, layered gabbro, metaby the NW-trending Najd fault system, particularly gabbro, plagiogranite, sheeted dyke complex, in the central part of the Arabian Shield (Fig. 2; metabasalt, dark shale and chert. Shanti & Roobol Quick 1991). We describe below the occurrence (1979) noted the presence of a 100-250 m thick of suture zones and major ophiolites from west to discontinuous melange zone beneath the ophiolite east across the Arabian Shield. that is composed of volcaniclastic rocks in a serpentinite matrix. Serpentinized peridotites, about 1.2 km thick, Ophiolites of the Yanbu suture zone occur in the lower part of the ophiolite. Remnants The NE-trending Yanbu suture zone occurs in the of harzburgite contain accessory clinopyroxene northwestern part of the Arabian Shield and is and some clinopyroxene lamellae in orthopyroxoffset by en echelon splays of the sinistral Najd ene, which is mostly replaced by bastite. Serpentifault system (Fig. 2). It separates the Midyan nized dunite hosts rare chromite lenses affected by terrane to the north from the Hijaz terrane to the postmagmatic deformation. Chromite grains in the south (Fig. 2). The Midyan terrane is composed dunite are anhedral to subhedral and commonly mainly of calc-alkaline volcanic rocks with ages exhibit pull-apart fractures. Few patches or fracof 725 Ma and younger (Stoeser & Camp 1985). ture fillings of cryptocrystalline magnesite are also Similar rocks probably continue into eastern present. Cr-spinels from harzburgites and chromiEgypt as the Dokhan volcanic succession, which tites are compositionally similar to those of forms the northernmost part of the Nubian Shield. modern forearc peridotites (Pallister et al. 1988). The Hijaz terrane to the south contains three Gabbroic intrusions in the serpentinized peridodistinct igneous sequences composed of a primi- tites are extensively rodingitized. Ultramafic cutive island arc complex at the bottom (c. 805 Ma mulates are overlain by amphibolitized gabbro in age), subaqueous, calc-alkaline rocks of the dissected by numerous isolated diabasic dykes. Hijaz arc complex in the middle (c, 750-715 Ma), The overlying sheeted doleritic dyke complex is and subaerial volcanic rocks of an arc complex on 200-600 m thick and is overlain by a c. 300 m top (c. 680-615 Ma) (Stoeser & Camp 1985, and thick extrusive sequence made mainly of pillow references therein). Magmatism of the Hijaz arc basalt. The sedimentary cover includes pelagic complex was terminated by an orogenic event chert, dark shale and siltstone. Ophiolitic subunits (Samran orogeny) along the Bir Umq suture zone are intruded by granitic plutons of a volcanic arc around 700-680 Ma, leading to the development complex.

Ophiolites of the Arabian Shield

Table 1. Occurrence, age and classification of Proterozoic ophiolites in the Arabian Shield Remarks

Suture zone/Fault zone Ophiolite name and location

Age (Ma)

Lithological units

Classification

YANBU

Jabal Ess (26°22'N, 37°37'N)

706 ± 11 U-Pb zircon age from an isotropic gabbro

Mediterranean-type, SSZ ophiolite

Intruded by island arc plutons; basement of ensimatic arc terrane(s)

Al' Ays (25°5'N, 38°6'E); massifs in Jabal Al Wask, Wadi Osman, Jabal Al Ays, Jabal Abu Sweira, Jabal Ash Sharthah

770-740 U-Pb zircon ages from gabbro and trondhjemite in the Jabal Al Wask massif

Penrose-type pseudostratigraphy; shale and chert as sedimentary cover; melange unit beneath the ophiolite Peridotites, isotropic gabbro, rare layered gabbro, quartz diorite, splitic basalt, tuff, chert, isolated dolerite dykes; no sheeted dyke complex observed

Mediterranean-type, SSZ ophiolite

Intruded by island arc plutons; basement of ensimatic arc terrane(s)

Bir Umq (23°57'N, 40°58'E)

838 ± 10 U-Pb zircon age from a dioritic pluton; 782 ± 5 and 764 ± 3 U-Pb zircon ages from dykes in peridotites 870 ± 11 U-Pb zircon age from a gabbro pluton; inherited zircon (same rock) with an age >1250 Ma

Harzburgite and dunite, pyroxenite, isotropic gabbro, Mediterranean-type, SSZ ophiolite >2 km thick extrusive sequence overlain by chert and tuffaceous rocks; no sheeted dyke complex observed Locally well-preserved Penrose-type Mediterranean-type, SSZ ophiolite pseudostratigraphy; igneous contacts between lavas and sheeted dykes and between gabbros and sheeted dykes

Ophiolitic basement of ensimatic arc terrane(s)

843-821 U-Pb zircon ages from plagiogranite dykes in peridotites

Harzburgite, dunite, ultramafic cumulates, gabbro, plagiogranite, basaltic—andesitic lavas, tuff (rhyodacitic); no sheeted dyke complex or pelagic sedimentary cover 2 km thick mainly dunitic peridotites; gabbros and diabasic rocks as dykes in peridotites; basaltandesite-rhyolite flows, tuff, lahar deposits, turbiditic rocks; no sheeted dyke complex or pelagic sedimentary cover

BIR UMQ

Jabal Thurwah (22°36'N, 39°22'E)

HULAYFAH-RUWAH Bir Tuluha (25°50'N; 40°54'E)

Darb Zubaydah (24°30'N, 41°8'E) near Jabal Malhijah; (24°55'N, 41°29'E) near Bir Nifazi

c. 830 (Quick 1990); intruded by 720-640 Ma granitoid plutons

HALABAN

Halaban (23°30'N, 44°25'E)

694 ± 8 U-Pb zircon age from a gabbro; 680 40Ar/39 Ar hornblende date from the metamorphic sole

AL AMAR

Al Amar ophiolites in the Mogheirah 698 U-Pb zircon age from gabbros area (24°22'N, 45°E) and in the Jabal Bitran area (23°28'N, 45°10'E)

NABITAH-HAMDAH Nabitah massif (20°45'N, 44°13'E); FAULT ZONE Tathlith massif (19°32'N, 43°31 'E); Hamdah massif (19°2'N, 43°36'E)

627 ± 4 U—Pb zircon age from a gabbro in the Tathlith massif

Thrust over an Upper Proterozoic island arc sequence; ophiolitic basement of ensimatic arc terrane(s)

Mediterranean-type, SSZ ophiolite

Part of a volcanic archipelago fringing the Afif continental plate

Mediterranean-type, SSZ ophiolite

Part of a volcanic archipelago fringing the Afif continental plate; late-stage intra-arc rifting as evidenced by highTi pillow basalts (c. 500 m thick) in the Upper Volcanic Series

Serpentinized peridotites, isotropic gabbro, basaltic lavas, pelagic sedimentary rocks (stratigraphic top to the east); locally overlain by pyroclastic rocks, agglomerate and silicic lava flows; no sheeted dyke complex; locally associated with a serpentinite matrix melange

Mediterranean-type, SSZ ophiolite

Geochemical features typical of forearc ophiolites; incipient arcforearc tectonic setting above a westdipping subduction zone adjacent to the Afif continental plate; the AlBijadiyah and Jabal Tays massifs also in the same suture zone

Serpentinized peridotites, locally as fault-bounded blocks within volcanic rocks, layered and isotropic gabbros, diabasic dykes, and basaltic lavas; locally associated with a serpentinite matrix melange

Franciscan-type ophiolites incorporated into an accretionary complex (Abt Schist) via imbricate thrusting

Intruded by syn- to post-collisional granites (640-620 Ma); suture zone as a trench-arc (Ar Rayn terrane) collision site during terminal closure of the Mozambique Ocean

Different massifs in a 5-30 km wide transpressional Ligurian-type ophiolite with Hessshear zone within the Asir terrane; harzburgite, type oceanic crust dunite, wehrlite with minor gabbro and isolated dykes in them; basaltic pillow lavas with interbedded chert and limestone

Possible rift floor of an intracontinental pararift basin within the Asir terrane; MORB affinity of ophiolitic magmas; post-collisional, youngest ophiolites in the Arabian Shield

PROTEROZOIC ARABIAN SHIELD OPHIOLITES

689

Fig. 2. Simplified tectonic map of the Arabian Shield showing the occurrence of the Neoproterozoic ophiolites, suture zones, and major lithotectonic assemblages mapped as terranes by previous researchers. Crustal types of these terranes and their age brackets are also shown. The prominent NW-trending faults collectively make up the Najd fault system. Data sources: Vail (1985); Pallister et al. (1988); Berhe (1990); Johnson & Kattan (2001); Stern (2002). Ophiolites: 1, Jabal Ess; 2, Bir Fuqayr; 3, Al'Ays; 4, Jabal Thurwah; 5, Bir Tuluha; 6, Arja; 7, Darb Zubaydah; 8, Bir Umq; 9, Jabal Ghurrab; 10, Nabitah; 11, Tathlith; 12, Hamdah; 13, Al Bijadiyah; 14, Halaban; 15, Jabal Mogheira; 16, Jabal Bitran.

690

Y. DILEK & Z. AHMED

The Jabal Ess ophiolite has a Penrose-type complete pseudostratigraphy, although the original contacts between the subunits are strongly modified by successive episodes of deformation. An isotropic gabbro unit in the ophiolite yielded a UPb zircon age of 706 ± 11 Ma (Pallister et al. 1988). Al'Ays ophiolite. This ophiolite occurs in the south-central part of the Yanbu suture zone and consists of several isolated massifs (number 3 in Fig. 2; Ahmed 2003). The Jabal Al-Wask massif displays an ophiolitic sequence with serpentinized peridotites consisting of dunite, harzburgite, wehrlite and pyroxenite exposed in the core of a topographically well-defined domal structure. Lenses, pods and lenticular layers of chromite are dispersed in serpentinite, and show massive, banded and/or disseminated textures. Gabbroic rocks overlying the ultramafic cumulates are mainly isotropic and grade upward from pyroxene-rich to hornblende-rich gabbros and diorites. Layered gabbros are rare and relatively thin where present. Quartz gabbro is overlain by sodic granophyres and metavolcanic rocks. Sheeted dykes have not been observed in this massif. The sedimentary cover, which overlies spilitized metabasalts, includes clastic rocks, tuff and chert. Isolated doleritic dykes crosscut most of the ophiolitic subunits. Ophiolite bodies are technically emplaced in country rocks of the Farri Group composed of mafic volcanic and fine-grained clastic rocks with ages >780 Ma (Kemp et al 1980). Pallister et al. (1988) interpreted the Jabal Al Wask massif as a serpentinite-matrix melange. U-Pb zircon ages of gabbro (dyke in a peridotite) and trondhjemite samples from the Jabal Al Wask massif range from 770 to 740 Ma (Pallister et al 1988). Ophiolites of the Bir Umq suture zone The Bir Umq suture zone runs subparallel to the Yanbu suture and separates the Hijaz terrane to the north from the Jeddah terrane to the south (Fig. 2). Near the suture zone, the Jeddah terrane is composed of dioritic to tonalitic plutons of the Taif arc complex that were probably intruded into an older arc basement (Stoeser & Camp 1985). Magmatism in the Taif arc might have been waning by 715 Ma when the initial suturing between the Hijaz and Jeddah terranes was in progress (Nasseef & Gass 1977). The Bir Umq suture zone continues farther SW across the Red Sea into Sudan and connects with the Amur-Nakasib Suture, which separates the Gebeit terrane in the north from the Haya terrane in the south within the Nubian Shield (Nassief et al. 1984; Reischmann 2000). The

Ingessana ophiolite constitutes the main ophiolite complex in the Amur-Nakasib Suture in Sudan (Fig. 1). The Bir Umq and Jabal Thurwah massifs are the two main ophiolite occurrences in the Bir Umq suture zone (Table 1). Bir Umq ophiolite. The Bir Umq ophiolite occurs in an ENE-trending large thrust sheet at the eastern end of the suture zone (number 8 in Fig. 2) and consists of peridotite, gabbro, volcanic rocks, and overlying chert and tuff; no sheeted dyke complex has been observed in this ophiolite (Ahmed & Hariri 2001). The peridotites consist of altered harzburgite, dunite and pyroxenite. Crspinels in the harzburgite and dunite are TiC>2poor (<0.25%) and display a wide range of Cr number (Cr/(Cr + Al)) at relatively high Mg numbers (Mg/(Mg + Fe+2)), characteristic of forearc and/or immature island arc peridotites (Pallister et al. 1988, and references therein). Plutonic rocks are composed mainly of isotropic gabbros with thin horizons of layered gabbro. Basaltic rocks are over 2 km in thickness and locally display well-preserved pillow structures. They are intercalated with, and overlain by, chert and tuffaceous rocks. A dioritic rock from the western part of the ophiolite yielded a U-Pb zircon age of 838 ± 10 Ma (Pallister et al. 1988), whereas quartz keratophyre dykes and sills within serpentinized peridotites yielded U-Pb zircon ages between 782 ± 5 Ma and 764 ± 3 Ma, indicating continuous igneous activity on the ophiolitic basement at Bir Umq. Jabal Thurwah ophiolite. This ophiolite occurs in the western end of the Bir Umq suture zone (number 4 in Fig. 2), c. 20 km from the Red Sea, and includes a complete Penrose-type complete pseudostratigraphy. Ultramafic rocks are composed of harzburgite with minor dunite bodies, dunite and wehrlite cumulates, and pyroxenites. Crustal rocks include layered to isotropic gabbros, mafic dykes and basaltic rocks (Ramsay 1986). Small outcrops of the sheeted dyke complex and basaltic lavas in the SE part of the ophiolite are mostly metamorphosed to greenschist facies. Transitional igneous contacts between pillow lavas and sheeted dykes, and between gabbros and sheeted dykes, are locally well preserved. Trace-element chemistry of basaltic lavas from the Jabal Thurwah ophiolite suggests a suprasubduction zone affinity (Nassief et al. 1984). The ophiolite is covered in the east by Tertiary flood basalts of Harrat Rahat. The Jabal Thurwah ophiolite is thrust onto extensively folded and faulted, tholeiitic to calc-alkaline volcanic rocks of an Upper Proterozoic island arc sequence. Jackson & Ramsay (1980) correlated volcanic and

PROTEROZOIC ARABIAN SHIELD OPHIOLITES

691

Bir Tuluha ophiolite. This ophiolite constitutes the northernmost massif within the Hulayfah-Ruwah suture zone (number 5 in Fig. 2). Ophiolitic rocks mainly include serpentinized harzburgite and dunite, ultramafic cumulates, metagabbro, plagiogranite and spilitized basalt. Contacts between these Ophiolitic subunits are commonly faulted, although locally gabbro, diorite and plagiogranite intrusions occur as dykes within serpentinized peridotites. Chromites from the harzburgite and dunite tectonites show a wide range in Cr number characteristic of suprasubduction-zone-generated oceanic lithosphere (Pallister et al. 1988). Tectonized peridotites locally occur as fault-bounded slabs within a metavolcanic and metasedimentary sequence composed of basaltic pillow lavas, andesitic lavas, breccia, conglomerate, greywacke and rhyodacitic tuff (Le Metour et al. 1983). These relations indicate a strong spatial association of the ophiolite with a volcanic arc sequence. U-Pb zircon ages from plagiogranite dykes in serpentinized peridotites range from 843 to 821 Ma (Pallister et al. 1988).

(number 7 in Fig. 2) near Jabal Malhijah (global positioning system (GPS) 24°30'N, 41°8'E) and Bir Nifazi (GPS 24°55'N, 41°29'E). It was first mapped as an ophiolite by Petot (1976). It is intruded by granodioritic plutons in the west and by alkali granite and quartz diorite plutons in the east; it is also divided by a granitic pluton into two septa. The ophiolite comprises a crustal section > 11 km thick, underlain by a dunite-dominated ultramafic section of 2 km thickness. Ophiolitic exposures also occur as roof pendants in a younger granitic batholith and as slivers along the Najd system faults (Quick & Gregory 1995). At Darb Zubaydah, the ophiolite forms an east-dipping homocline. The orogenic granite-diorite-monzogranite intrusions are 720-640 Ma old, and a younger group of smaller granitic intrusions are 640-570 Ma in age (Quick 1990). Ultramafic rocks consist mainly of dunite cumulates, strongly altered to serpentine, talc, chlorite and magnetite. Chromian spinel grains are not foliated, as is commonly seen in tectonites, and possess very thick ferritchromite and magnetite rims with Cr numbers ranging from 0.75 to 0.85 (Quick 1990). Such high Cr numbers are typical of modern forearc peridotites, and are distinctly different from those of mid-ocean ridge or fracture zone peridotites (Dick & Bullen 1984). The existence of harzburgite is suggested by the presence of sparse poikilitic bastite pseudomorphs. The dominance of dunite in the ultramafic section, high Cr numbers of chromian spinel, and the depleted nature of the peridotites suggest high degrees of partial melting, and/or the percolation of large volumes of transient melt through a more stagnant upper mantle. Gabbroic and diabasic dyke intrusions are widespread in the ultramafic rocks. No sheeted dyke complex has been observed in the Darb Zubaydah ophiolite. The crustal section of the ophiolite contains interbedded basalt-andesite-rhyolite flows in addition to tuffs, lahar deposits, and turbidites formed in a submarine environment within, or on the flanks of, a volcanic arc. Pelagic sedimentary rocks are absent. Basalts and andesites have calcalkaline compositions with high-alumina (16.919.9%) and low-Ti (<8400ppm) contents (Quick 1990). High-Ti pillow basalts, c. 500 m thick, with interlayered turbidites, marble and chert form the Upper Volcanic Series and might have formed during intra-arc rifting (Quick 1990). The Darb Zubaydah ophiolite is c. 830 Ma old and is interpreted to represent the basal and upper-crustal units of an island arc developed between 850 and 800 Ma (Quick 1990).

Darb Zubaydah ophiolite. This north-south elongated ophiolite complex is exposed farther south

Jabal Ghurrab ophiolite. Other ophiolite massifs occur east of Jabal Ghurrab (GPS 22°26.6'N,

sedimentary rocks of the area with the Samran Series, which formed at 780-900 Ma. A metagabbro sample from the northern part of the ophiolite yielded a U-Pb zircon age of 870 ± 11 Ma; the same gabbro also contained inherited zircon with an age >1250 Ma (Pallister et al. 1988).

Ophiolites of the Hulayfah-Ruwah suture zone This suture zone marks the western boundary of the Afif terrane against the Hijaz, Jeddah and Asir terranes (Fig. 2). It is also known in the literature as the Nabitah suture zone (Pallister et al. 1988, and references therein). However, we distinguish the Nabitah fault zone as a separate tectonic discontinuity within the eastern half of the Asir terrane, different from the Hulayfah-Ruwah suture zone bounding the Afif terrane against the Hijaz-Jeddah-Asir composite terrane (Fig. 2). The adjacent Afif terrane to the east is composed of granite, gneiss and supracrustal clastic rocks of the Murdamah and Jibalah Groups. These rocks and their crystalline basement (of mid-Proterozoic age and older) collectively constitute a continental microplate rimmed by magmatic arcs. Post-orogenic granites and intermediate to felsic volcanic rocks in the Afif terrane are 640-580 Ma and c. 660-600 Ma old, respectively (Delfour 1981; Darbyshire et al. 1983; Stoeser et al. 1984).

692

Y. DILEK & Z. AHMED

42°22'E) and at Jabal Zalm (GPS 22°49'N, 42°13'E) farther south within the Hulayfah-Ruwah suture zone (number 9 in Fig. 2). At Jabal Buram, serpentinized harzburgite, crosscut by numerous diabasic dykes, is thrust over granite gneiss and metasedimentary rocks. Serpentinite is widespread south of the east-west-trending Wadi Subay at Jabal At-Tin (GPS 22°N, 42°35'E).

Ophiolites of Nabitah-Hamdah fault zone Ophiolitic units crop out discontinuously along the north-south-trending Nabitah-Hamdah fault zone, a transpressional shear zone of 5-30 km width within the Asir terrane (Fig. 2). Ophiolite sheets are c. 1.5km thick and are commonly dissected by high-angle reverse faults and NWtrending strike-slip faults of the Najd system. These ophiolitic exposures include strongly sheared serpentinites, rare gabbros, pillowed metabasalt and deep marine sedimentary rocks such as chert and pelagic limestone. Ophiolitic subunits have been metamorphosed to greenschist facies. Nabitah ophiolite. The Nabitah ophiolite lies in a north-south-trending narrow shear zone (number 10 in Fig. 2) and consists mainly of serpentinite, separating high-grade granite-gneiss and schist rocks to the east from dark-coloured volcanic and volcaniclastic rocks to the west. A melange unit containing blocks of basalt and gabbro in a schistose serpentinite matrix occurs to the north. Basalt is locally pillowed and contains interbedded chert and limestone. Chromite pods and lenses in the serpentinized peridotites contain massive, disseminated and nodular ores. Tathlith complex. Serpentinite bodies and associated gabbro intrusions occur along the Nabitah fault zone (number 11 in Fig. 2) and north of the town of Tathlith (GPS 19°32'N, 43°31'E; Kellogg et al. 1986). Serpentinized harzburgite, dunite and wehrlite are technically emplaced over amphibolites that are intercalated with graphite-schist, quartzite and calcschist; these basement rocks probably represent metamorphosed passive margin sequences. Serpentinite occurs mainly in two areas: (1) SW of Tathlith town, as a thin, elongate outcrop west of Wadi Tathlith; (2) NE of Tathlith town, east of Wadi Tathlith. Serpentinites are nonschistose and carry abundant veins and stringers of magnesite. Small disseminations (1-2%) of chromite and titanomagnetite are present. Crustal rocks include widespread gabbros and spilitized basalts intercalated with pelagic limestone and chert. A gabbro from the Tathlith area has a U-Pb zircon age of 627 ± 4 Ma (Pallister et al 1988) that is significantly younger than the ophiolites along the

Hulayfah-Ruwah suture zone farther north. Hamdah ophiolite. Serpentinite bodies at and near Jabal Hamdah (number 12 in Fig. 2), just east of the town of Hamdah (GPS 19°2'N, 43°36'E), rest technically on amphibolite (Hariri 1989). This southernmost ophiolite along the Nabitah fault zone may extend farther south beneath the Phanerozoic cover. Sheet-like to lensoidal serpentinite bodies are associated with minor gabbro and are crosscut by isolated dykes. The serpentinite contains asbestos, magnesite veins and sparse pockets of pyroxenite. Younger gabbro-diorite plutons intrude the serpentinite. Al-Rehaili & Warden (1980) interpreted the mafic-ultramafic rocks along the Nabitah-Hamdah fault zone as diapiric intrusions that were subsequently metamorphosed to greenschist-amphibolite facies and then remobilized into shear zones and cores of folds. Those workers differentiated these ophiolites from the suture-related ophiolites of the Arabian Shield by virtue of the intrusive nature of their ultramafic rocks, the lack of a Penrose-type ophiolite pseudostratigraphy, and their spatial association with melanges.

Ophiolites of the Halaban suture zone The NNW-trending Halaban suture (Urd suture of Pallister et al. 1988) marks the eastern boundary of the Afif terrane (Fig. 2). It contains the northern Al-Bijadiyah ophiolite, and the southern, more extensively developed, Halaban ophiolite. East of the suture zone, the Abt Schist forms roof pendants in a complex granite-granodiorite batholith in the Ad-Dawadimi terrane. The occurrence of detrital chromite in the Abt Schist indicates an oceanic-ophiolitic provenance for the depocentre(s) in which its protoliths were deposited. Stacey et al. (1984) reported a mean model age of 710 Ma for detrital zircons from the Abt Schist. The Abt Schist has been interpreted as a subduction-accretion complex developed at a convergent margin (Pallister et al 1988). To the west of the Abt Schist and the Halaban suture zone lies the Afif terrane, which contains the following lithostratigraphic units: (1) Halaban Formation, consisting of andesite, agglomerate, conglomerate, greywacke, acidic flows, diorite, gabbro and serpentinite; (2) Murdamah Group, including conglomerate, quartzite, slate, greywacke and acidic volcanic rocks; (3) granodiorite-granite batholithic rocks, unconformably underlying the Murdamah Group; (4) Jibalah Sandstone, which fills a NNW-trending faultbounded graben; (5) grey alkalic granites; (6) Farida marble; (7) pink alkalic and peralkalic granite, dated at 590-550 Ma. These lithostrati-

PROTEROZOIC ARABIAN SHIELD OPHIOLITES graphic units collectively represent a magmatic arc system that constitutes the main body of the Afif terrane. The Abt Schist was probably the accretionary complex to this magmatic arc system at c. 660-600 Ma (Stoeser & Camp 1985). The ophiolites along the 1-10 km wide Halaban suture zone contain serpentinized harzburgite and dunite, sparse pods of chromitite, anthophyllite schist, listwaenite, metagabbro and metabasalt. Volcanic rocks dominate toward the top (Delfour 1979b). The Al-Bijadiyah ophiolite (number 13 in Fig. 2) contains gabbro, serpentinite and basalt, and is associated with an ophiolitic melange including blocks of basalt, gabbro, pyroxenite and listwaenite within a strongly schistose serpentinite matrix. Southward in the ophiolite, gabbro predominates and ends sharply against amphibolitegrade basaltic rocks that are strongly sheared and interbedded locally with mafic pyroclastic rocks, calcareous metasedimentary rocks and chert. The Halaban ophiolite farther south in the suture zone (number 14 in Fig. 2; Urd Group of Delfour 1979b; Pallister et al. 1988) is sandwiched between gneiss and schist of the Al-Ajal Group (Delfour 1979b) in the Afif terrane to the west, and the granite-granodiorite basement intruded by remobilized younger calc-alkaline granites of the Ad-Dawadimi terrane to the east. It is composed mainly of deformed gabbro, serpentinized peridotites and basalt, all of which are metamorphosed to greenschist facies. Isotropic gabbros predominate although gabbronorite is locally present. Ultramafic rocks consist of serpentinized cumulates and websterite metamorphosed to anthophyllite, talc and tremolite schist. The distribution of basalt and pelagic sedimentary rocks indicates that the top of the ophiolite is to the east. Stacey et al. (1984) reported a U-Pb zircon age of 694 ± 8 Ma from a hypersthene gabbro near the village of Halaban. Al-Saleh et al. (1998) determined the emplacement age of the Halaban ophiolite as 680 Ma based on 40Ar/39Ar hornblende dates from the metamorphic sole. The Jabal Tays complex of Al-Shanti & El Mahdi (1989) farther SE in the suture zone is an extension of the Halaban ophiolite and is composed of sheared serpentinite, pyroxenite and gabbroic rocks intruded by granitoid plutons. AlShanti & Gass (1983) regarded the Jabal Tays serpentinite as part of an ophiolite generated in, or adjacent to, a transform fault zone and then placed in an oceanic trench where it was partially subducted.

Ophiolites of the Al-Amar suture zone This easternmost suture zone in the Arabian Shield separates the north-south-oriented Ar-Rayn

693

terrane to the east from the Ad Dawadimi terrane to the west (Al-Husseini 2000; Johnson & Kattan 2001). The Ar-Rayn terrane is composed mainly of a volcanic arc sequence (Halaban Volcanics; Al-Husseini 2000) with plutonic rocks younger than 650 Ma (Calvez et al. 1984; Stacey et al. 1984). However, the presence of 2 Ga inherited zircons in a trondhjemite intrusion in the western part of the Ar-Rayn terrane suggests that either the terrane has an older crustal component or that these are relict, continentally derived zircons incorporated into the melt via previous subduction events. The Ad-Dawadimi terrane adjacent to the suture zone consists of the Abt Schist with a protolith of basin-trench turbidites (Shanti & Mitchell 1976). In the Mogheirah area (Fig. 2) the Abt Schist is composed of albite-sericite-chlorite schist showing multiphase deformation and lying above the basal amphibole-calcschist, sericite marble and quartzo-feldspathic schist. The Al-Amar Group east of the ophiolites consists of rhyolitic to andesitic flows, and pyroclastic rocks cut by granite and/or trondhjemite stocks, all tectonized and metamorphosed to greenschist facies. The ophiolites in the Al-Amar suture zone (Mogheira and Jabal Bitran, numbers 15 and 16 in Fig. 2) contain strongly sheared serpentinites exposed extensively along shear zones and fault systems, and locally as fault-bounded blocks within volcanic rocks. Layered and isotropic gabbros, diabasic dykes and basaltic extrusive rocks make up the rest of the ophiolites. Locally, ophiolitic subunits, such as pillow basalt, plagiogranite, gabbro and diabase, appear as blocks in a serpentinite matrix melange. At Jabal Humayyan, schistose serpentinite is intermingled with chlorite schist and talc-carbonate schist, and locally includes fault-bounded blocks of gabbro, basalt and brown chert (Al-Shanti & El-Mahdy 1989). The Al-Amar ophiolites are dated at 698 Ma based on zircon ages from metagabbros (Pallister et al. 1988). The collision along the Al-Amar suture zone is dated at 640-620 Ma based on the ages of syn- to post-collisional granites intruding various lithological units within the suture zone.

Tectonics of the Arabian Shield ophiolites Existing structural, petrological, geochemical and geochronological data from the Arabian Shield ophiolites allow us to derive some limited conclusions and inferences about the mode and nature of oceanic crust generation and terrane accretion around Afro-Arabia during the Proterozoic. Reliable U-Pb zircon ages from major ophiolite occurrences in different suture and fault zones range from 870 ± 11 Ma for the oldest (Jabal Thurwah; Pallister et al. 1988) to 627 ± 4 Ma for

694

Y. DILEK & Z. AHMED

the youngest (Tathlith; Pallister et al, 1988; Table 1). The distribution of ophiolite ages within the Arabian Shield indicates that oceanic crust generation in this region was broadly contemporaneous during the Neoproterozoic, although an eastward progression through time is discernible. The majority of the older ophiolites located within the Yanbu and Bir Umq suture zones in the west (i.e. Jabal Ess, Jabal Thurwah) have a Penrose-type pseudostratigraphy, complete with sheeted dyke complexes, and display transitional igneous contact relations between their crustal units (plutonic sequence, sheeted dyke complex and extrusive rocks). The chemistry of uppermantle peridotites, gabbros, dykes and lavas in these ophiolites consistently shows melt source and magma compositions characteristic of suprasubduction zone settings (foreac, arc, backarc) (Pallister et al. 1988; Quick 1990, 1991; Stern 1994, 2002). Collectively, these structural and geochemical characteristics of the Yanbu and Bir Umq suture zone ophiolites in the western part of the Arabian Shield are typical of the Mediterraneantype ophiolites (Dilek 2003), which have evolved in small ocean basins separated by continental fragments within the Mesozoic Neo-Tethys (Dilek & Flower 2003), reminiscent of the modern SW Pacific Ocean (Dilek & Moores 1990). Unlike the Mediterranean ophiolites, however, the Yanbu and Bir Umq suture zone ophiolites are spatially associated with arc volcanic-volcaniclastic rocks and are intruded by arc plutons. These relations indicate that the ophiolites might have constituted an oceanic basement on which volcanic arc terranes evolved through time. The volcanic arc complexes in the bounding Midyan, Hijaz and Jeddah terranes (Camp 1984) probably developed on this inferred ophiolitic basement. In this sense, these Proterozoic volcanic arc complexes are analogous to Mesozoic ophiolites in the Sierra Nevada Foothills of California where fragments of a Jurassic island arc complex (i.e. Smartville, Slate Creek) formed on an older, deformed ophiolitic basement (i.e. Jarbo Gap ophiolite) displaying a polygenetic crustal evolution (Dilek 2003, and references therein). We therefore infer that the Yanbu and Bir Umq suture zone ophiolites show a record of sea-floor spreading, subduction initiation, island arc development and arc collision(s) in the Proterozoic evolutionary history of the western Arabian Shield. Steep contacts and the locally well-developed serpentinite matrix melange associated with these ophiolites are probably artefacts of multiple collisional events that led to the amalgamation of the Midyan, Hijaz and Jeddah terranes (Camp 1984; Stern 1994). The ophiolites of the Hulayfah-Ruwah suture zone between the Afif continental plate to the east

and the Hijaz-Jeddah-Asir composite volcanic arc terranes to the west are nearly coeval with or slightly younger than their counterparts in the Yanbu and Bir Umq suture zones. Their ages (c. 840-830 Ma) indicate that the igneous evolution of these eastern ophiolites preceded the collisional events that amalgamated the Midyan, Hijaz and Jeddah volcanic arc terranes. The Hulayfah-Ruwah suture zone ophiolites lack sheeted dykes but contain most other ophiolitic subunits in a typical Penrose pseudostratigraphy. Unlike the Yanbu and Bir Umq suture zone ophiolites, the ophiolites of the Hulayfah-Ruwah suture zone do not have pelagic sedimentary rocks (i.e. chert, limestone) in their cover, which is composed mainly of turbiditic clastic rocks, lahar deposits and basaltic to rhyolitic tuff (Quick 1990). The volcanic rocks in these ophiolites range from mid-ocean ridge basalt (MORB) tholeiite at the bottom to calc-alkaline andesite and rhyolite toward the top, similar to the Mediterranean-type ophiolites (i.e. Mirdita and Semail; Dilek & Flower 2003). This chemo- and lithostratigraphy of the extrusive sequence, combined with the high Cr numbers of spinels in tectonized peridotites, suggests that the Hulayfah-Ruwah suture zone ophiolites may represent a fossil oceanic lithosphere developed in an unevolved island arc or in a back-arc basin. Quick (1990) reported the existence of c. 500 m thick, high-Ti basaltic lavas in the upper part of the volcanic sequence of the Darb Zubaydah ophiolite that are associated with diabasic intrusions; he interpreted this late-stage magmatism as a product of incipient intra-arc rifting. However, this short-lived intra-arc rifting event was apparently arrested at an early stage, as it did not result in a fully developed back-arc basin. Similar ensimatic island arc evolution via intra-arc rifting is reported from the Jurassic Josephine ophiolite in CaliforniaOregon (Harper 2003). Limited structural observations from the Hulayfah-Ruwah suture zone make it difficult to discern the emplacement history of its ophiolites and the mode and nature of collisional events that led to the amalgamation of the ophiolites and the Afif terrane to the Midyan-Hijaz-Jeddah composite terranes. It is likely, however, that a volcanic archipelago consisting of the Hulayfah-Ruwah suture zone ophiolites was accreted to the western margin (in present coordinate system) of the Afif terrane first, followed by the collision of this arc-continent complex with the Midyan-Hijaz-Jeddah-Asir composite terranes to the west. This geodynamic scenario is reminiscent of the tectonic evolution of the Jurassic-Cretaceous Kohistan island arc complex, which was accreted to Eurasia through an arc-continent collision first, followed by the In-

PROTEROZOIC ARABIAN SHIELD OPHIOLITES dia-Eurasia (including the accreted Kohistan arc) continental collision (Coward et al. 1986). The Halaban suture zone ophiolites east of the Afif continental plate occur in discontinuous exposures within a belt of 1-10 km width and consist mainly of serpentinized harzburgite and dunite, gabbro, basalt and pelagic sedimentary rocks. The layer-cake stratigraphy commonly seen in Penrosetype ophiolites does not exist in the Halaban suture zone ophiolites, which locally display a serpentinite-matrix melange character. The sedimentary cover showing top-to-the-east stratigraphic indicators is overlain locally by pyroclastic rocks, agglomerate and silicic lava flows. The existing age data from a gabbro and the metamorphic sole of the Halaban ophiolite (694 ± 8 Ma and 680 Ma, respectively) suggest that the 'Halaban' oceanic crust was displaced from its igneous tectonic setting within 20 Ma of its formation, reminiscent of the Mediterranean-type ophiolites (i.e. Tauride ophiolites in Turkey, Dilek et al. 1999; Semail ophiolite Oman, Searle et al. 2003). The high field strength element depletion of ophiolitic rocks in the Halaban suture zone suggests high-degree hydrous melting of peridotites in a forearc tectonic setting (Al-Shanti & Gass 1983). Pallister et al. (1988) showed from Pb-isotope data that the mantle source for the Halaban suture zone ophiolites was depleted in U relative to Th and Pb long before the derivation of ophiolitic magmas; they inferred that these ophiolites had formed in a volcanic arc with a primitive oceanic mantle source. The existence of slightly younger volcanic, plutonic and clastic sedimentary rocks of the arcrelated Halaban Formation in the adjacent Afif terrane to the west supports the spatial and temporal association of the Halaban suture zone ophiolites with a volcanic arc. The inferred forearc-incipient arc setting of the Halaban suture zone ophiolites is analogous to the tectonomagmatic history of the Jurassic Coast Range ophiolite in California (Shervais 2001). We infer that the Halaban suture zone ophiolites developed in a forearc environment over a west-dipping subduction zone (in present coordinate system) in close proximity to the Afif continental plate. The tectonic setting of the Abt Schist as a subduction-accretion complex, forming much of the Ad Dawadimi terrane east of the Halaban suture zone, is well established (Shanti & Mitchell 1976; Stacey et al. 1984; Pallister et al. 1988). Detrital zircon ages (c. 710 Ma) from the Abt Schist indicate that this subduction-accretion complex was coeval with the 'Halaban' forearcincipient arc system to the west. Therefore, we interpret the Abt Schist as a Franciscan-type accretionary complex that evolved at an Andeantype active margin of the Afif terrane.

695

The Al-Amar suture zone ophiolites east of the Abt Schist in general display a melange character with blocks of gabbro, diabase and pillow lavas in a sheared serpentinite matrix. The igneous age (698 Ma; Pallister et al. 1988) of these ophiolites is similar to that of the Halaban suture zone ophiolites (c. 694 ± 8 Ma) to the west. Based on the regional tectonics and age relations we think that the ophiolitic occurrences in the Al-Amar suture zone may represent scraped-off fragments of an ocean floor, which were subducted beneath the Afif terrane to the west (in present coordinate system). We thus interpret the Al-Amar suture zone ophiolites as Franciscan-type ophiolites (Dilek 2003) that were incorporated into the accretionary prism (Abt Schist) via imbricate thrusting, synthetic to the palaeo-subduction zone. As is common in Franciscan-type ophiolites and accretionary complexes, considerable differences in lithological units, metamorphic grades, chemical affinities and structural relations are expected along the Al-Amar suture zone ophiolites. If this interpretation is correct, then the Al-Amar suture zone may actually mark the site of a trench-arc (Ar Rayn terrane to the east) collision in the latest Neoproterozoic. The Proterozoic ophiolites along the 5-30 km wide Nabitah-Hamdah fault zone in the eastern part of the Asir terrane (Fig. 2) are the youngest (627 ± 4 Ma) reported so far from the Arabian Shield. In addition, they are distinctly different from other suture zone ophiolites in the Arabian Shield in two aspects. First, they locally display primary intrusive relations with their metasedimentary host rocks, indicating that they are relatively parautochthonous. Overstreet (1978) reported, for example, a fresh gabbroic intrusion in a metagreywacke unit near the town of Tathlith. Al-Rehaili & Warden (1980) interpreted serpentinized peridotites in the Hamdah ophiolite as diapiric intrusions in metasedimentary rocks. These metasedimentary rocks in the Asir terrane likely represent passive margin sequences associated with intra-continental rifting. Second, leadisotope data from the Nabitah-Hamdah fault zone ophiolites indicate Pb with equivalent composition to that from modern mid-ocean ridge basalt (Pallister et al. 1988), suggesting a MORE origin of their magmas. Trace-element chemistry of their volcanic rocks suggests that ophiolitic magmas of the Nabitah-Hamdah fault zone were derived from a less refractory source that underwent a distinctly lesser degree of partial melting (Z. Ahmed, unpublished data). The internal structure and stratigraphy of the Nabitah-Hamdah ophiolites, their spatial relations with the metamorphosed passive margin sequences, and the MORE origin of their magmas

696

Y. DILEK & Z. AHMED

are reminiscent of the Ligurian ophiolites in the northern Apennines and other Jurassic ophiolites in the western Alps. These Tethyan ophiolites in the Alpine-Apennine orogenic belt are composed of Hess-type oceanic crust, made of abundant serpentinized ultramafic rocks with gabbroic intrusions and a basaltic cover intercalated with thin horizons of pelagic sedimentary rocks; this kind of oceanic crust is likely to have developed in intracontiental rift basins-pararifts and/or slowspreading centres (Dilek 2003). This inferred pararift origin of the Nabitah-Hamdah fault zone ophiolites and the limited age data suggest that the Asir terrane might have undergone an episode of intracontinental rifting after accretion of the Afif and Ar Rayn terranes into the rest of the Arabian Shield by 640-625 Ma (Camp 1984). This intracontinental rifting and restricted oceanization might have developed in response to collision-driven mantle flow, reminiscent of the western Mediterranean and the SW Pacific regions (see Dilek & Flower 2003; Flower & Dilek, 2003), and/or impact-generated lithospheric-scale stretching and rifting in the hinterland of a major collisional orogenic belt. In this case, the Nabitah-Hamdah fault zone ophiolites do not really represent classic 'suture zone' ophiolites sensu stricto, as in Mediterranean-type ophiolites, but rather 'alpine-type peridotites' (and associated crustal units as in Steinmann trinity) developed in the strongly attenuated rift floor(s) within the Asir terrane (Dilek 2003).

Implications for the geodynamic evolution of the Arabian Shield The Proterozoic ophiolites and associated melange units exposed in the Arabian Shield occur in several subparallel belts with distinctive orientations (Fig. 2). The NE-trending ophiolite belts in the west contain some of the oldest ophiolitic occurrences in Saudi Arabia (Pallister et al. 1988) and separate composite volcanic arc terranes with lithotectonic units that are in part coeval with and/ or younger than the ophiolites. The presence of relict, older zircon grains (> 1250 Ma in age) in some ophiolitic subunits in these belts (Pallister et al. 1988) suggests that the melt source in the mantle was highly heterogeneous, containing continentally derived detrital material. The ophiolitebearing Yanbu and Bir Umq suture zones appear to continue westward beneath the Coastal Plains and across the Red Sea into the Nubian Shield in east Africa (Fig. 1), and may represent the northeasterly tracks (in present coordinate system) of separate ocean basins that evolved between different volcanic archipelagoes during the early stages

of development of the Arabian-Nubian Shield. These ocean basins might have been small seaways within and fringing the much larger Mozambique Ocean (Stern 1994; Dalziel 1997). They probably collapsed as a result of arc-trench, arcarc and arc-continent collisions (Frisch & AlShanti 1977) prior to the terminal closure of the Mozambique Ocean that resulted from the collision of East and West Gondwana during the Neoproterozoic (Dalziel 1997). North- to NW-trending (in present coordinate system) suture zones in the eastern part of the Arabian Shield separate terranes with continental affinities displaying high-grade metamorphic rocks and a broad range of zircon ages. Two of these terranes, Afif and Ar Rayn, are considered as continental plates that underwent extensive cratonization through intrusion of granitoids before and after their amalgamation into the Arabian Shield (Greenwood et al. 1976; Camp 1984; Kroner 1985; Vail 1985; Pallister et al. 1988). Thus, the suture zones in the eastern Arabian Shield, namely the Hulayfah-Ruwah and Al Amar, appear to mark the sites of continental collision reminiscent of the modern Himalayas. Ophiolites in these suture zones range in age from 830 to 680 Ma (Pallister et al. 1988; Quick 1990, 1991; Al-Saleh et al. 1998) and are thus slightly younger than their counterparts in the western Arabian Shield. The westernmost of the NNW-trending suture zones, the Hulayfah-Ruwah, sharply truncates the NE-trending Yanbu and Bir Umq suture zones and juxtaposes the continental Afif terrane against the volcanic arc terranes of Hijaz, Jeddah and Asir to the west (Fig. 2). The other two suture zones to the east (Halaban and Al Amar) mark the progressive eastward growth of the Arabian Shield through successive docking of terranes during latest Neoproterozoic time. Thus, the timing of development of the Hulayfah-Ruwah suture zone appears to mark a significant shift from mainly north-south-directed collision-accretion events to dominantly east-west-directed collision-accretion events during the Proterozoic. The HulayfahRuwah suture zone, the Afif terrane, and all other suture zones and terranes farther west are crosscut by en echelon splays of the NW-trending Najd fault system (Fig. 2; Sultan et al. 1988), which might have developed in response to collisiondriven escape tectonics during the latest stages (c. 600 Ma) of the crustal evolution of the Arabian Shield (Stoeser & Camp 1985).

Discussion The origin and evolution of the Proterozoic ophiolites in the Arabian Shield have strong implications for late Precambrian tectonics and Earth

PROTEROZOIC ARABIAN SHIELD OPHIOLITES history. The majority of the ophiolites in the western Arabian Shield formed during 870770 Ma, a time period that coincides with the initial episodes of rifting of the supercontinent Rodinia (Lindsay et al. 1987; Stern 1994; Dalziel 1997). Dismantling of Rodinia via continental rifting generated the Mozambique Ocean with multiple seaways reminiscent of the modern SW Pacific Ocean or the Mesozoic Neo-Tethys (Dilek & Moores 1990). Fragments of the Mozambique ocean floor may have been preserved within the western Arabian Shield ophiolites, although much of the original oceanic lithosphere, generated by sea-floor spreading, in this large basin was consumed to produce island arc complexes that were subsequently accreted to form the juvenile crust of the Arabian-Nubian Shield. The forearc-infant arc origin of the ophiolites in the western Arabian Shield, as inferred mainly from their chemical signatures, suggests that they had formed during subduction initiation, similar to most Phanerozoic ophiolites. We think, therefore, that the mode of oceanic crust generation through plate tectonic processes during the Neoproterozoic was reminiscent of that operating throughout the Phanerozoic. The Neoproterozoic oceanic crust partially preserved in the western Arabian Shield ophiolites also formed an oceanic basement on which the volcanic arc complexes of the Midyan, Hijaz, Jeddah and Asir terranes were developed. Collisional processes that accreted these disparate volcanic arc terranes preserved the suprasubduction zone ophiolites in discrete suture zones. Thus, the ophiolites along the Yanbu and Bir Umq suture zones in the western Arabian Shield preserve a complete record of rift-drift, sea-floor spreading, subduction initiation and terrane accretion during the evolution of the Mozambique Ocean. Future studies in these ophiolites should focus on delineating the structural, geochemical and geochronological fingerprints of these different tectonic episodes in the rock record. A systematic correlation of these ophiolites with their counterparts in the Nubian Shield across the Red Sea should also be undertaken to better understand the geodynamic evolution of the East African Orogen. The Neoproterozoic ophiolites in the central and eastern Arabian Shield, particularly those in the Hulayfah-Ruwah, Halaban and Al Amar suture zones, also display geochemical and tectonic features that suggest their evolution in suprasubduction zone settings. However, these ophiolites appear to have developed in volcanic arc terranes adjacent to and/or fringing a continental plate (Afif terrane), analogous to the modern Andaman-Nicobar Islands and the Sunda Arc in Asia and to the Triassic-Jurassic Klamath-

697

Sierran arc systems in the Mesozoic western USA. The collapse and accretion of these fringing arc systems into the continental margins of the Afif terrane preceded its collision with the composite volcanic terranes to the west. It is possible that some oceanic rocks now exposed in these suture zones (particularly the Hulayfah-Ruwah and Al Amar suture zones) may represent fragments of seamounts and large igneous province-generated oceanic plateaux. Future investigations should help us better distinguish between these inferred origins based on careful structural field studies and isotopic analyses. The post-collisional Nabidah-Hamdah fault zone ophiolites (c. 627 Ma) are unique among the Arabian Shield ophiolites in that their origin was not associated with subduction-accretion processes. We infer that these ophiolites developed as a result of lithospheric-scale continental rifting within a recently assembled supercontinent. In this sense, they may signal the beginning of a new Wilson Cycle and the initial stages of the Greater Gondwanaland breakup (c. 640 Ma) at the site of major orogenic crustal thickening and uplift associated with the terminal closure of the Mozambique Ocean (Stern 1994). Fragments of these Ligurian-type (Dilek 2003) pararift zone ophiolites of the Arabian Shield may exist in Somalia, Ethiopia and the Mozambique Belt of Tanzania in eastern Africa. Future ophiolite studies in the Asir terrane should focus on documenting the structure, stratigraphy and petrology of the lithotectonic units surrounding the Nabitah, Tathlith and Hamdah ophiolites, as well as on the isotope geochemistry, geochronology and structure of these ophiolites. The Arabian Shield ophiolites were a major component of Neoproterozoic continental growth. They took part in the lateral growth of continental plates through the accretion of oceanic material and island arc terranes at trenches and by volcanism and plutonism at active and passive (rifted volcanic) continental margins. The history of the Arabian Shield ophiolites and suture zones indicates a rapid rate of continental growth in comparison with that estimated for the Phanerozoic (Stern 1994) during a time span of c. 250 Ma in the late Neoproterozoic. The Arabian Shield ophiolites provide an excellent opportunity to investigate oceanic and juvenile crust evolution in the latest Precambrian. Systematic studies of Neoproterozoic ophiolites and suture zones in the light of new developments in the evolving ophiolite concept and sophisticated isotopic and geochronological techniques can help us delineate ancient oceanic crust with distinct tectonic settings of origin, and disparate mantle isotope domains and mantle source regions; we

698

Y. DILEK & Z. AHMED

can also better understand igneous, metamorphic, hydrothermal and tectonic processes that operated during rift-drift, sea-floor spreading and plate collisions as the Arabian Shield was assembled. Uncertainties about the Proterozoic ophiolite record are not an artefact of the lack of preservation and/or the fundamental differences in plate tectonic processes that operated then, but are a result of the persistent search over the years for a Penrose-type oceanic crust with a layer-cake pseudostratigraphy, complete with a sheeted dyke complex. We now know that different ophiolites have different structural architecture, chemical fingerprints and geodynamic evolutionary paths, pointing to different geotectonic settings of their origin. Thus, we should no longer look for ophiolites in the Precambrian record in the strict sense of an ophiolite-ocean crust analogy, as defined by the Penrose Conference description (Anonymous 1972), a call that was also made by Helmstaedt & Scott (1992) more than a decade ago. Future Precambrian ophiolite investigations should focus on delineating the internal architecture and mantle sources and domains of ophiolites based on systematic structural field studies, isotopic analyses and regional tectonic syntheses. We gratefully acknowledge the co-operation and support of our colleagues in the King Fahd University of Petroleum and Minerals in Saudi Arabia (KFUPM) and the facilities provided to Z. Ahmed by M. M. Hariri, Chairman of the Earth Sciences Department at KFUPM, during the course of our research. Our synthesis of the Arabian Shield ophiolites heavily relies on the previous work by many colleagues, too numerous to list here; we bear the full responsibility, however, for the views and interpretations presented in this paper and for likely omissions of other pertinent references on the Arabian Shield ophiolites. Y. Dilek acknowledges partial funding from Miami University through OARS and the Committee on Faculty Research in support of this study. Careful and constructive reviews by J. Alten, A. Polat (University of Windsor, Canada), P. T. Robinson (Dalhousie University, Canada) and M. Sultan (University at Buffalo, USA) significantly improved the manuscript.

References ABDEL-GAWAD, M. 1970. Interpretation of satellite photographs of the Red Sea and Gulf of Aden. Philosophical Transactions of the Royal Society of London, Series A, 267, 23-40. AHMED, Z. 2001. Geological features of the ophiolites along the Yanbu and Bir Umq sutures in Saudi Arabia. GeoArabia, 6(2), 334. AHMED, Z. 2003. Geochemical and genetic implications of adalates associated with the suprasubduction zone-type Al-Ays ophiolite, northwestern Arabian Shield. International Geology Review, 45, 176-190. AHMED, Z. & HARIRI, M.M. 2001. Ophiolites of the Northwestern Arabian Shield, Kingdom of Saudi

Arabia. King Fahd University of Petroleum and Minerals, Department of Earth Sciences, SABIC Research Project (SABIC/99-2) Final Report, Dhahram, Saudi Arabia, 66p. AL-HUSSEINI, M.L 2000. Origin of the Arabian Plate structures: Amar collision and Najd rift. GeoArabia, 5(4), 527-542. AL-REHAILI, M.H. & WARDEN, AJ. 1980. Comparison of the Bir Umq and Hamdah ultrabasic complexes, Saudi Arabia. In: COORAY, P.G. & TAHOUN, S.A. (eds) Evolution and Mineralization of the ArabianNubian Shield. Proceedings of a Symposium. Institute of Applied Geology Bulletin (King Abdulaziz University, Jeddah, Saudi Arabia), 3(4), 143-156. AL-SALEH, A.M., BOYLE, A.P. & MUSSETT, A.E. 1998. Metamorphism and 40Ar/39 Ar dating of the Halaban ophiolite and associated units: evidence for twostage orogenesis in the eastern Arabian Shield. Journal of the Geological Society, London, 155, 165-175. AL-SHANTI, A.M. & EL-MAHDY, O.R. 1989. Geological Studies and Assessment of Chromite Occurrences in Saudi Arabia: Final Report. KACST Project, AT-6094. AL-SHANTI, A.M. & GASS, I.G. 1983. The late Proterozoic ophiolite melange zones of the easternmost Arabian Shield. Journal of the Geological Society, London, 140, 867-876. ANONYMOUS 1972. Penrose Field Conference on ophiolites. Geotimes, 17, 24-25. BAKOR, A.R., GASS, I.G. & NEARY, C.R. 1976. Jabal AlWasq, northwest Saudi Arabia: an Eocambrian back-arc ophiolite. Earth and Planetary Science Letters, 30, 1-9. BERHE, S.M. 1990. Ophiolites in Northeast and East Africa: implications for Proterozoic crustal growth. Journal of the Geological Society, London, 147, 41-57. BROWN, G.F., SCHMIDT, D.L. & HUFFMAN, A.C. JR 1989. Geology of the Arabian Peninsula, Shield Area of the Western Saudi Arabia. US Geological Survey, Professional Papers, 560-A. CALVEZ, J.Y., ALSAC, C, DEFOUR, J., KEMP, J. & PELLATON, C. 1984. Geologic evolution of western, central, and eastern parts of the northern Precambrian Shield, Kingdom of Saudi Arabia. In: BAKER, A.R ET AL.. (eds) Pan-African crustal evolution in the Arabian-Nubian Shield. Proceedings of a symposium, International Geological Correlation Programme (IGCP) 164. Faculty of Earth Sciences, King Abdulaziz University, Jeddah, 24-48. CAMP, V.E. 1984. Island arcs and their role in the evolution of the western Arabian Shield. Geological Society of America Bulletin, 95, 913-921. CHURCH, W.R. 1988. Ophiolites, sutures, and microplates of the Arabian-Nubian Shield: a critical comment. In: EL-GABY, S. & GREILING, R.O. (eds) The Pan-African Belt of Northeast Africa and Adjacent Areas. Vieweg, Braunschweig, 289-316. COLEMAN, R.G. 1984. Ophiolites and the tectonic evolution of the Arabian Peninsula. In: GASS, I.G., LIPPARD, SJ. & SHELTON, A.W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, Lon-

PROTEROZOIC ARABIAN SHIELD OPHIOLITES don, Special Publications, 13, 359-366. COWARD, M.P., WINDLEY, B.F. & BROUGHTON, R.D. ET AL. 1986. Collision tectonics in the NW Himalayas. In: COWARD, M.P. & RIES, A.C. (eds) Collision Tectonics. Geological Society, London, Special Publications, 19, 203-219. DALZIEL, I.W.D. 1997. Overview: NeoproterozoicPaleozoic geography and tectonics: review, hypothesis, environmental speculation. Geological Society of America Bulletin, 109, 16-42. DARBYSHIRE, D.P.F., JACKSON, N.J., RAMSAY, C.R. & ROOBOL, M.J. 1983. Rb-Sr isotope study of latest Proterozoic volcano-sedimentary belts in the central Arabian Shield. Journal of the Geological Society, London, 140, 203-214. DELFOUR, J. 1981. Geologic, tectonic and metallogenic evolution of the northern part of the Precambrian Arabian Shield (Kingdom of Saudi Arabia). Bulletin du Bureau de Recherches et Minieres (deuxieme serie), Nos. 1-2, Section II, 1980-1981, 1-19. DELFOUR, J. 1979a. Upper Proterozoic volcanic activity in the northern Arabian shield, Kingdom of Saudi Arabia. In: TAHOUN, S.A. (ed.) Evolution and mineralization of the Arabian-Nubian Shield, Proceedings of a symposium, Institute of Applied Geology Bulletin. 3, 59-75, King Abdulaziz University, Jeddah, Saudi Arabia. DELFOUR, J. 1979b. Geological Map of the Halaban Quadrangle, Sheet 23G, Kingdom of Saudi Arabia. Geological Map, GM46A. DICK, H.J.B. & BULLEN, T. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 5-76. DILEK, Y. 2003. Ophiolite concept and its evolution. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 1-16. DILEK, Y. & FLOWER, M.FJ. 2003. Arc-trench rollback and forearc accretion: 2. A model template for ophiolites in Albania, Cyprus and Oman. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 43-68. DILEK, Y. & MOORES, E.M. 1990. Regional tectonics of the Eastern Mediterranean ophiolites. In: MALPAS, J., MOORES, E.M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues, Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 295-309. DILEK, Y., THY, P., HACKER, B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and mafic dike intrusions (Turkey): implications for the Neo-Tethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. FLOWER, M.FJ. & DILEK, Y. 2003. Arc-trench rollback and forearc accretion: 1. A collision-induced mantle flow model for Tethyan ophiolites. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218,21-41. FRISCH, W. & AL-SHANTI, A.M. 1977. Ophiolite belt and the collision of island arcs in the Arabian shield. Tectonophysics, 43, 293-306.

699

GASS, I.G. 1981. Pan African (Upper Proterozoic) plate tectonics of the Arabian-Nubian shield: Upper Proterozoic to Lower Palaeozoic tectonics (PanAfrican event). In: KRONER, A. (ed.) Precambrian Plate Tectonics. Elsevier, Amsterdam, 387—405. GREENWOOD, W.R., HADLEY, D.G., ANDERSON, R.E., FLECK, R.J. & SCHMIDT, D.L. 1976. Late Proterozoic cratonization in southwestern Saudi Arabia. Philosophical Transactions of the Royal Society of London, Series A, 280, 517-527. HARIRI, M.M. 1989. Gold Mineralization at the Ancient Hajr Mine. Directorate General of Mineral Resources, Jiddah, Open-File Report, DGMR-OF-0636. HARPER, G.D. 2003. Tectonic implications of boninite, arc tholeiite, and MORB magma types in the Josephine Ophiolite, California-Oregon. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 207-230. HELMSTAEDT, H.H. & SCOTT, D.J. 1992. The Proterozoic Ophiolite problem. In: CONDIE, K.C. (ed.) Proterozoic Crustal Evolution, Developments in Precambrian Geology. Elsevier, Amsterdam, 10, 55-95. JACKSON, NJ. & RAMSAY, C.R. 1980. Time-space relationships of Upper Precambrian volcanic and sedimentary units in the Central Arabian Shield. Journal of the Geological Society, London, 137, 617-628. JOHNSON, P.R. & KATTAN, F. 2001. Oblique sinistral transpression in the Arabian Shield: the timing and Kinematics of a Neoproterozoic suture zone. Precambrian Research. 107, 117-138. KELLOGG, K.S., DOMONIQUE, J., MINOUX, L. & FOURNIGUET, J. 1986. Geological Map of the Wadi Tathlith Quadrangle, Sheet 20 G, Kingdom of Saudi Arabia. Scale 1:250000. Deputy Ministry for Mineral Resources, Geoscience Map, GM-103 C. KEMP, J., PELLATON, C. & CALVEZ, J.Y. 1980. Geochronological Investigations and Geological History in the Precambrian of Northwestern Saudi Arabia. Saudi Arabian Deputy Ministry for Mineral Resources Open-File Report, BRGM-OF-01-1. KRONER, A. 1985. Ophiolites, and the evolution of tectonic boundaries in the late Proterozoic Arabian-Nubian shield of Northeast Africa and Arabia. Precambrian Research, 27, 277-300. KRONER, A., TODT, W., HUSSEIN, I.M., MANSOUR, M. & RASHWAN, A.A. 1992. Dating of late Proterozoic ophiolites in Egypt and the Sudan using the single grain zircon evaporation technique. Precambrian Research, 59, 15-32. LE METOUR, J., JOHAN, V. & TEGYEY, M. 1983. Geology of the Ultramafic—Mafic Complexes in the Bi'r Tuluhah and Jabal Malhijah. Saudi Arabian Deputy Ministry for Mineral Resources Open-File Report, BRGM-OF-03-40. LINDSAY, J.F., KORSCH, R.J. & WILFORD, J.R. 1987. Timing the breakup of a Proterozoic supercontinent: evidence from Australian intracratonic basins. Geology, 15, 1061-1064. MOORES, E.M. 2002. Pre-1 Ga (pre-Rodinian) ophiolites: their tectonic and environmental implications. Geological Society of America Bulletin, 114, 80-95.

700

Y. DILEK & Z. AHMED

NASSEEF, A.O. & GASS, LG. 1977. Granites and metamorphic rocks of the Taif area, western Saudi Arabia. Geological Society of America Bulletin, 88, 1721-1730. NASSIEF, M.O., MACDONALD, R. & GASS, LG. 1984. The Jabal Thurwah Upper Proterozoic ophiolite complex, western Saudi Arabia. Journal of the Geological Society, London, 141, 537-546. OVERSTREET, W.C. 1978. A Geological and Geochemical Reconnaissance of the Tathlith Oe-degree Quadrangle, Sheet 19/43, Kingdom of Saudi Arabia. US Geological Survey, Saudi Arabian Mission Project Report, (IR) 230. PALLISTER, J.S., STAGEY, J.S., FISCHER, L.B. & PREMO, W.R. 1988. Precambrian ophiolites of Arabia: geologic settings, U-Pb geochronology, Pb-isotope characteristics, and implications for continental accretion. Precambrian Research, 38, 1-54. PETOT, J. 1976. Geology and Mineral Exploration of the Jabal Sanam Quadrangle [24/408J. Bureau de Recherches Geologiques et Minieres, Technical Record, 76-JED-12. QUICK, I.E. 1990. Geology and origin of the Late Proterozoic Darb Zubaydah ophiolite, Kingdom of Saudi Arabia. Geological Society of America Bulletin, 102, 1007-1020. QUICK, I.E. 1991. Late Proterozoic transpression on the Nabitah fault system—implications for the assembly of the Arabian shield. Precambrian Research, 53, 119-147. QUICK, LE. & GREGORY, R.T. 1995. Significance of melt-wall rock reaction: a comparative anatomy of three ophiolites. Journal of Geology, 103, 187-198. RAMSAY, C.R. 1986. Explanatory Notes to the Geologic Map of the Rabigh Quadrangle, Sheet 22D, Kingdom of Saudi Arabia. Deputy Ministry for Mineral Resources, Jeddah, Map, GM-84. REISCHMANN, T. 2000. Ophiolites and island arcs in the late Proterozoic Nubian shield. Ofioliti, 25(1), 1-14. SEARLE, M.P., WARREN, C.J., WATERS, DJ. & PARRISH, R.R. 2003. Subduction zone polarity in the Oman Mountains: implications for ophiolite emplacement. In: DILEK, Y. & ROBINSON, P.T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 467-480. SHACKLETON, R.M. 1986. Precambrian collision tectonics in Africa. In: COWARD, M.P. & RIES, A.C. (eds) Collision Tectonics. Geological Society, London, Special Publications, 19, 329-349. SHANTI, A.M.S. & MITCHELL, A.H.G. 1976. Late Precambrian subduction and collision in the AlAmar-Idsas region, Arabian Shield, Kingdom of

Saudi Arabia. Tectonophysics, 30, T41-T47. SHANTI, M. & ROOBOL, MJ. 1979. A late Proterozoic ophiolite complex at Jabal Ess in northern Saudi Arabia. Nature, 279, 488-491. SHERVAIS, J.W. 2001. Birth, death and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, 2, Paper Number 2000GC000080. STACEY, J.S., STOESER, D.B., GREENWOOD, W.R. & FISCHER, L.B. 1984. U-Pb zircon geochronology and geological evolution of the Halaban-Al Amar region of the Eastern Arabian Shield, Kingdom of Saudi Arabia. Journal of the Geological Society, London, 141, 1043-1055. STERN, R.J. 1981. Petrogenesis and tectonic setting of a late Precambrian ensimatic volcanic rocks, Central Eastern Desert of Egypt. Precambrian Research, 16, 195-230. STERN, R.J. 1994. Arc assembly and continental collision in the Neoproterozoic East African Orogen: implications for the consolidation of Gondwanaland. Annual Review of Earth and Planetary Sciences, 22, 319-351. STERN, R.J. 2002. Crustal evolution in the East African Orogen: a neodymium isotopic perspective. Journal of African Earth Sciences, 34, 109-117. STERN, R.J., NIELSEN, K.C., BEST, E., SULTAN, M., ARVIDSON, R.E. & KRONER, A. 1990. Orientation of late Precambrian sutures in the Arabian—Nubian shield. Geology, 18, 1103-1106. STOESER, D.B. & CAMP, V.E. 1985. Pan-African microplate accretion of the Arabian shield. Geological Society of America Bulletin, 96, 817-826. STOESER, D.B., STAGEY, J.C., GREENWOOD, W.R. & FISHER, L.B. 1984. U/Pb Zircon Geochronology of the Southern Part of the Nabitah Mobile Belt and Pan-African Continental Collision in the Saudi Arabian Shield. Saudi Arabian Deputy Ministry for Mineral Resources Technical Record, USGS-TR04-5. SULTAN, M., ARVIDSON, R.E., DUNCAN, J., STERN, R.J. & KALIOUBY, B.E. 1988. Extension of the Najd shear system of Saudi Arabia to the central Eastern Desert of Egypt based on integrated field and Landsat observations. Tectonics, 1, 1291-1306. SULTAN, M., BECKER, R., ARVIDSON, R.E., SHORE, P., STERN, R.J., EL ALFY, Z. & ATTIA, R.I. 1993. New constraints on Red Sea rifting from correlations of Arabian and Nubian Neoproterozoic outcrops. Tectonics, 12, 1303-1319. VAIL, J.R. 1985. Pan-African (late Precambrian) tectonic terrains and the reconstruction of the ArabianNubian Shield. Geology, 13, 839-842.

Index Page numbers in italic, e.g. 320, refer to figures. Page numbers in bold, e.g. 323, signify entries in tables. Aarja geological map 320 oxygen isotopic and petrographic data 323-324 plagiogranite bodies 328-334, 329, 330-333 Acaeniotyle diaphorongona 157 Acaeniotyle umbilicata 157 Ad Dawadimi terrane 689 Adamsfield 519 Adelaide Rift 518 Adelbert Range 511 Aden 686 Adriatic Sea 24 Adstock-Ham Massif 232, 234-235 geological map 236 Aegean Sea 24 Afif 686, 689 African ophiolites 10, 11, 14 Ahin 316 Ailaoshan 543 Ainyn terrane 637 Akesai 553 Akros, Mount 112 Aktyubinsk 569 Al Ajaiz 468 Al Amar suture zone 686, 693 Al Hammah 469 Al Khaburah 316 Al Kuwayt 686 Aladag ophiolites 45 AT Ays ophiolites 690 Albania Mirdita interpolated P-T-t histories 29 model template for ophiolites 43-46, 59-61 interpretation 50-53, 51 mantle-driven rollback 50 Neo-Tethyan subduction rollback 49-50 overview of Tethyan ophiolite geology 46-48 paired ophiolite belts 48-49 tectonic diagram 60 Albanide-Hellenide Alpine erogenic belt 111 Alboran Sea 24 Algero-Provencal Basin 24 Aliabad 132 All 543 Almopias Massif 111 Alpine-Apennine peridotites 69-70, 80-82, 82 chemical refertilization and thermal erosion of lithosphere 85 extraction of melt in dykes 83-84, 84 geochemical data 78-80, 79-80, 81 Lanzo and Corsica field observations 73-75, 74, 76 impregnation textures 75-78, 77 residual mantle mineral assemblages 75, 77 Piedmonte Ligurian ophiolites 70-71, 71 extrusive rocks 73

field relations and petrography of serpentines and peridotites 71-72, 72 plutonic rocks 72-73 porous melt flow during opening of Piedmont Ligurian ocean 82-83 Alpine-Himalayan orogenic system 9-10 Altun 543 Altyn Tagh Fault 24 Amadeus Basin 518 Amami Plateau 289 Amur-Nakasib suture zone 686 Anadyr 599 Andaman Sea 486-487 Andaman-Nicobar Islands 24 Ankara 44 Ankara Melange 45 Antalya Complex 45 Anumaqin 543 Aoyougou 553 Apennines 24, 71 Appalachian ophiolites 11, 12, 14 Ar Rayn terrane 689 Arabian Shield Proterozoic ophiolites 685-687, 696-698 Al Amar suture zone 693 Bir Umq suture zone 690-691 geodynamic evolution 696 geological map 686 Halaban suture zone 692-693 Hulayfah-Ruwah suture zone 691-692 Nabitah-Hamdah fault zone 692 occurrence, age and classification 688 tectonics 693-696 tectonic map 689 Yanbu suture zone 687-690 Arabian-Nubian Shield 10 Arafura Sea 508 Aranami Fault 301, 310 outcrop sketch 377 Riedel shears 373 schematic diagram 372 arc-trench rollback and forearc accretion collision-induced mantle flow model for Tethyan ophiolites 21-23,24,32-33 collision-induced mantle extrusion 31 endogenous vs. exogenous causes 29-31 forearc complexes 25-26 interpolated P-T-t histories 29 model development 25-29, 27 partial melting models 28 plate kinematic effects 29-30 proto-ophiolite to ophiolite transition 31-32 slab pull and extrusion tectonics 30, 30 subduction nucleation 26-29 Tethyan history 23-25 model template for ophiolites in Albania, Cyprus and Oman 43-46, 59-61 interpretation 50-59, 57, 54, 57

702

INDEX

mantle-driven rollback 50 Neo-Tethyan subduction rollback 49-50 overview of Tethyan ophiolite geology 46-48 paired ophiolite belts 48-49 tectonic diagrams 60 Argolis Massif 111 Argolis Peninsula (Greece), Triassic mid-ocean ridge basalts 109-110, 124-125 geochemical results 117-118, 120, 121, 122 bulk-rock major and trace elements 116-117 incompatible element abundance patterns 119 geology of central-northern region 110 petrogenesis 119-122 petrography 117 sampling and methods analytical methods 113-114 locations 113, 114, 115 structure and stratigraphy of the Middle Unit 110-113, 112 tectonomagnetic interpretation 118-119 Triassic magmatism 122-123 geodynamic implications 123, 124 modern oceanic analogue 123-124 Arthur metamorphic complex 519 Arunta Inlier 518 As Sifah 451, 452, 468, 469 As Sifah subwindow 452, 453, 455-456 Asia, North-East see NE Asia Asimah-Masafi 468 Asir terrane 686, 689 Asjudi 468 Assayab 316 geological map 379 oxygen isotopic and petrographic data 321-322 plagiogranite bodies 334, 335 Aushkul, Lake 577 Australia, Tethyan- and Codilleran-type ophiolites 10-11, 11, 14, 517-518, 518 Delamerian-Tyennan ophiolites 518 Delamerian Orogen occurrences 520 Tyennan (Tasmanian) Orogen occurrences 518-520 Lachlan Orogen ophiolites 520-521, 527 Darkly River Belt 528-529, 529 Heathcote Belt 521-523, 522, 524 Mount Wellington Belt 535-528 occurrences of unknown connections 529-531 main features 533 New England Orogen ophiolites 531 northern occurrences 531-532, 537 southern occurrences 532-533 tectonic setting and evolution of proto-Tasmanides 533-534 Austroalpine Nappes 77 Avekov terrane 525 Ayu Trough 485 Babu 554, 560 geological map and cross-section 555 Baghdad 44 Baigang 168 Bailadu 555 Baimaxueshan 560 Bainang 150, 168, 192 Baiquanmen ophiolite 553 Bakhtegan Fault 737 Balhashu 750

Banda 24 Banda Ridges 485 Banda Sea 24 Banggai 485, 490 Bangkok 554 Bangong-Nujiang Suture Zone 44 Bangong Lake 543 Bangong Lake-Nujiang ophiolite belt 557-558, 558 Bani Hamid 468 Banja Luca 92 Barkly River Belt 527, 528-529 petrological and geochemical data 529 Barrabool Hills 527 Baryulgil ultramafic rocks 537 basaltic glass, bioalteration recorded in ophiolitic pillow lavas 415-416, 424-425 biogenerated textures 417, 418, 419, 420 carbon isotopes 419-421, 423, 424 element mapping 411-419,421, 422 organic remains 417, 420 perspectives 421-424 Baskil arc 45 Bay of Islands ophiolite complex (Canada) 369-370, 397-398 analytical methods 374 conditions of metamorphism 390 fluid circulation model structural domain 3 395, 396 structural domain 4 395-397, 396 geological setting 370-371, 377 geothermometry 387-390 comparison between mineralogical and oxygen isotope geothermometers 389 lithological and structural units 372-374 mineral chemistry amphibole 381-382, 384, 385, 386 chlorite 382, 387 clinopyroxene 375-376, 381 epidote 386-387, 388 olivine 376 petrography, mineralogy and isotope composition data 377-380 plagioclase 376-381, 382, 383 prehnite 382-386, 388 pyroxene 376 oxygen isotope geochemistry calculated fluid/rock compositions 393 fluid/rock ratios 394-395 hydrothermal fluids 392-394 isotope compositions 390-391, 390, 391, 392 isotope equilibrium temperatures 391-392 petrography dynamically recrystallized gabbros 374-375, 375 static recrystallization 374, 375 sampling sites 373 Bayda 375 geological map 320 oxygen isotopic and petrographic data 323-324 plagiogranite bodies 328-334, 329, 330-333 Bayingou 560 Beaconsfield ultramafic complex 57P Beagle canal 666 Beenleigh block 537 Beijing 542, 599

INDEX Beimarang massif 150, 166, 167, 170-171 mineral chemistry 775, 777, 779, 180, 181, 182, 183, 184 Beishan 543 Beloresk 569 Benham Rise 289 Benten Island 301 fault analysis 308-309 stereographic projection 310 lithological map 370 pillow lava 377 schematic diagram of Aranami Fault 372 Beregovoi massif 626 Beregovoi terrane 625 Berezov massif 606 Bering Sea 527 Beysehir-Hoyran Nappes 45 Bianmagou 560 Bianmagou ophiolite 553 Bikin 599 bioalteration recorded in ophiolitic pillow lavas 415-416, 424-425 biogenerated textures 417, 418, 419, 420 carbon isotopes 419-421, 423, 424 element mapping 417-419, 421, 422 organic remains 417, 420 perspectives 421-424 Bir Tuluha ophiolites 691 Bir Umq ophiolites 690 Bir Umq suture zone 686, 689 Jabal Thurwah ophiolites 690-691 Bismarck Sea 492, 508, 511 Bitlis-Zagros Collision Zone 44 Blow Me Down Mountain 377 Blue Ridge ophiolites 253-256, 254, 255, 272-273 analysis methods 257 dunites and their spinel compositions 264-271 cation ratios 269 chemistries of coexisting olivine and Cr-spinels 266 Cr number-Mg number diagram 270 electron microprobe analyses of representative spinels 268 relatioship between mantle rock, dunite alteration and MORE gabbro 269 TiO2-Al2O3 diagram 2 70 electron microprobe results 257 field setting of ultramafic rocks 258 hydrous vs. anhydrous rocks 262-264 aspect ratios 264, 265 data 263 distribution 252 metamorphic associations in meta-ultramafic rocks 256 metamorphic character of ultramafic rocks 258-261, 259 petrogenetic grid 260 whole-rock chemistries 261 tectonic implications 271-272 Bonin (Ogasawara) arc 507-509 Bonin Islands incompatible element distributions 23 TiO2 vs. FeO plots 22 boninites boninite-HMA volcanism 26 presence in proto-arcs 26 tectonic implication in Josephine Ophiolite 207-208, 215-220

703

pillow lavas and breccias 218 tectonic setting for eruption of boninites 224-225 Th/Yb vs. Ta/Yb diagram 279 Ti vs. V discriminant diagram 216 TiO2 vs. La/Sm diagram 220 Y vs. Cr discriminant diagram 277 Boone ultramafic body 254 Borce 93 Borja massif 92 Borneo 24 Bosnian-Durmitor terrane 92 Boso triple junction 280 Bougainville Island 492 Bowen Basin 537 Boyes, River 519 Brazilian ophiolites 10, 77, 14 Brevard Fault Zone 254 Brezovica ultramafic massif (Serbia), petrology and evolution 91-92, 105-106 analytical procedures 94 field occurrence of ultramafic rocks 94 geological setting 92-93, 93 initial conditions 105 location 92 petrochemistry 98-100 bulk-rock compositions 98, 99, 100 primary mineralogy 94-97 olivine compositions 95 pyroxene compositions 96 spinel compositions 95, 97 secondary mineralogy 97-98 compositions 97 thermal history 100-102, 707 magmatic evolution 102-104 metamorphic evolution 104-105 Brisbane 537 Broken Hill Inlier 518 Buck Creek meta-ultramafic body 254 Budapest 44 Bulqiza Massif 777 Buqingshan 550 Buribai 569 Calabria 24 Calcutta 554 Caledonian ophiolites 77, 12, 14 Cangaldag arc 45 Cape Horn 666 Caribbean ophiolites 10 Caroline Sea 508 Carpathian-Balkan arc terrane 92 Carpathians 24 Caucasus 44 Celebes Sea 24, 490 Central Bosnian mountains terrane 92 Central (Irian) ophiolite belt 491-492 Chaibukha 607 Chalkidiki Massif 777 Changawuzi 560 Changning 543 Chelyabinsk 55P Chenaillet Ophiolites 77 Chengdu 554 Chersky Range 520

704

INDEX

Chiang Mai 554 China, ophiolite distribution, ages and tectonic settings 541-542,559,560,561-563 circum-Pacific ophiolites East Taiwan ophiolites 558-559, 559 Kaishantun ophiolites 559 generalized maps 542, 543 intracontinental oceanic basins 560-561 major oxides and trace elements 544-545 Neo-Tethyan ophiolites 554-555, 554 Bangong Lake-Nujiang ophiolite belt 557-558, 555 Yarlung Zangbo ophiolite belt 555-557 Paleo-Asian ophiolites 543-545 East Junggar ophiolite belt 548, 550 Inner Mongolian ophiolite belt 548-549, 552 Tianshan ophiolite belt 548 West Junggar ophiolite belt 545-547, 549 Paleo-Tethyan ophiolites 551-552 Jinshajiang-Ailaoshan ophiolite belt 552-554, 554 Precambrian ophiolites 542-543 Qinling-Qilian-Kunlun ophiolites North Qilian ophiolites 550-551, 553 West Kunlun ophiolites 549-550 rare earth elements 546-547 suprasubduction zone ophiolites 559-560 temporal-spatial evolution of ophiolites 561, 562 Chukchi Sea 621 Chukotka 620 Clarence Moreton Basin 531 Coast Range ophiolite (USA), accretionary emplacement mechanism 438-440 Coff's Harbour block 531 Connors Auburn Arc 537 Coral Plateau 508 Coral Sea 508 Cordilleran-type ophiolites 13, 14, 430 accretionary emplacement 436-437, 438-440 eastern Australia 517-518, 518 Lachlan Orogen 533, 534-536 Corsica 24, 71 peridotites field observations 73-75, 74, 76 geochemical data 79-80, 81 impregnation textures 75-78, 77 residual mantle mineral assemblages 75, 77 Corundum Hill meta-ultramafic body 254 Cotobato Trench 484 Crete 24 Cr-spinel Blue Ridge ophiolites 253-256, 272-273 analysis methods 257 backscatter image 277 cation ratios 269 chemistries of coexisting olivine and Cr-spinels 266 Cr number-Mg number diagram 270 dunites 264-271 electron microprobe analyses of representative spinels 268 electron microprobe results 257 relatioship between mantle rock, dunite alteration and MORE gabbro 269 TiO2-Al2O3 diagram 2 70 Josephine Ophiolite 219-220 composition 222

Crucella hispana 157 Cyclops ophiolites 493-494 Cyprus 24 model template for ophiolites 43-46, 59-61 interpretation 53-56, 54 mantle-driven rollback 50 Neo-Tethyan subduction rollback 49-50 overview of Tethyan ophiolite geology 46-48 paired ophiolite belts 48-49 tectonic diagram 60 Troodos ophiolites incompatible element distributions 23 TiO2 vs. FeO plots 22 Troodos-Mamonia: interpolated P-T-t histories 29 Cyprus Trench 45 Daban Col 553 Dabie-Sulu metamorphic belt 599 Dachadaban 560 Dachadaban ophiolite 553 Daito Ridge 289 Dakovica 92 Dalmatian-Hercegovinian composite terrane 92 Dangxiong 148 key characteristics of ophiolitic rocks 149 Darb Zubaydah ophiolites 691 Darbui 560 Day Brook meta-ultramafic body 254, 255 Dayuntang 555 Dazhuka 168, 560 Dazhuqu massif and terrane 150, 166, 168, 111, 192 mineral chemistry 776, 777, 779, 180, 181, 182, 183, 184 Ngamring Tso-Saga-Zongba 157-158 ophiolites 192-193 Renbung-Quxu 158 Xigaze ophiolites 154-156, 755, 757, 194-196 Dead Sea Fault 44, 45 Delamerian Orogen 518, 575, 527 Tethyan-type ophiolites 533, 534-536 D'Entrecasteaux Islands 492 depleted harzburgite-type ophiolites (DH) 588, 598 Devils Elbow Remnant of the Josephine Ophiolite 208, 210 Dhaka 554 Dhimaina 772 Dibba 468 Dinaride-Hellenide Alpine erogenic belt Triassic magmatism 122-123 geodynamic implications 123, 724 modern oceanic analogue 123-124 Dinarides 45, 92 Dingqing 554, 560 Djacovica-Mirdita massif 92 Djibouti 656 Dolodrook 527, 527-528 petrological and geochemical data 525 Dongqiao 554, 560 Dookie 527, 523-524, 525 U-Pb zircon data 526, 529 Dookie Fault 525 Dowlatabad 732 Drina-Ivanjica terrane 92 Duba 659 Dundas Trough 579 Dunhuang 553

INDEX dunite and metadunite petrology 253-256, 258, 272-273 cation ratios 269 chemistries of coexisting olivine and Cr-spinels 266 Cr number-Mg number diagram 270 electron microprobe analyses of representative spinels 268 relatioship between mantle rock, dunite alteration and MORE gabbro 269 spinel composition 264-271 TiO2-Al2O3 diagram 270 Dzhetygara 569 East Anatolian Fault 45 East Junggar ophiolite belt 543, 548, 550 East Siberian Sea 620 East Taiwan ophiolites 558-559, 559 Eastern Sunda Arc 485 Edmunds meta-ultramafic body 254 Elistratova 607 Erro Tobbio peridotites 71 Esk Trough 537 Euboea Massif 111 Evensk 607 Fathabad Fault 131 Fidalgo 208 Finisterre Range 511 Flores 485 Flores Basin 485 Flores Thrust 485 forearc complexes 25-26 forearc extension in Thetford Mines Ophiolite Complex 231-232, 246 deformation 239-241 lower crust 241-242 upper crust 240, 242 formation of boninitic crust 246, 247 geological setting 232, 233 model for the structural and magmatic evolution 244-245, 244, 245 sequence of crystallization and magmatic affinity 243 sheeted dyke complex and crustal polarity 243 slow-spreading environment 245-246 stratigraphy 235 hypabyssal facies 239 mantle section 235 plutonic crustal section 235-239, 236, 237, 238 stratigraphic columns 234 volcano-sedimentary facies 239 volcanic stratigraphy 243 forearc ophiolites in the western Pacific 507, 514 Bonin (Ogasawara) arc 507-509 gravity profiles 509 New Caledonia 509-510, 570 New Guinea ophiolites 510-514, 577, 572 Forth metamorphic complex 579 Franciscan Accretionary Complex 209 Franciscan-type ophiolites 10 Franklin metamorphic complex 579 Fries Block 254 Fries-Gossan Lead Fault 254 Futago 295 Futomi 307 Gangdese Batholith 793

705

Gangrinboche 148 Gankuvayam sequence 644 compositions of rock-forming and accessory Cr-spinels 647 major and trace element compositions of ultramafic rocks 648, 649 microprobe analyses of Cr-spinels 646 ophiolite stratigraphy 644 Ganychalan terrane geological setting 636-639, 637, 638 major and trace element compositions of magmatic rocks 640 major and trace element compositions of plutonic ophiolitic rocks 641, 643 origin of gabbroic and volcanic rocks 642-643 petrography and geochemistry of plutonic and volcanic rocks 639-642 Ganzi 543 Gasquet 209 Gawler Craton 518 Gebeit terrane 686 Georgetown Inlier 518 Georgina Basin 518 Gerf terrane 686 Giabulin 750, 167, 192 giant dyke swarms 14, 15 Gizhiga Bay 625 global geochemical cycles and ophiolites 353-354, 365-366 boundary and initial conditions for hydrothermal activity 354 geometric constraints on accretion 354-355 interaction temperatures 354 global weathering rates 362 epicontinental seaways 363-364 short-term water cycle effects 362-363, 363 profile through oceanic crust 355-356 neodymium residence time 356 oxygen residence time 356-357 strontium residence time 356 rate constant 359-362, 360, 361 Samail ophiolites and continental layered gabbro complexes 358 Samail ophiolites and oceanic layer-3 oxygen isotope levels 357-358 seawater composition 354 significance of ophiolite studies 364-365 special significance of oxygen isotope profile for seawater 357 water balance 364 whole-rock <318O values for pillow lavas through time 358-359, 359 Gogango overfolded zone 537 Golmud 554 Gomsiqe Massif 777 Gondwana 10, 11, 14 breakup 23, 25 Gossan Lead Block 254 Gradiste 93 Grandfather Mountain Window 254 Great Artesian Basin 537 Greben massif 626 Greer Hollow meta-ultramafic body 254, 255 Guevgueli Massif 777 Guleman ophiolites 45 Gyangze 792 Gyantze 750 Gympie block 537

706 Hagab Thrust 468 Halaban suture zone 686, 689, 692-693 Halmahera 485, 490, 495 Hamdah ophiolites 692 Hanoi 554 harzburgite, depleted 597-599, 598 harzburgite-type ophiolites (HOT) 597-599, 598 Hashimoto 296 Hassanabad Fault 131 Hawasina Thrust 468, 469 Hawasina window 451, 456, 468 domal culmination 459 geological sections 473 Haya terrane 686 Haybi 468 Haybi Thrust 469 Haybi volcanic complex 451 Hayesville-Fries Fault 254 Heathcote Belt 521-523, 527, 522 petrological and geochemical data 524 Heathcote Fault 522 Heazlewood, River 519 Hegenshan 560 Hegurinaka 297 Heihe 553 Hellenic arc-trench 24 Hellenides 45 Hengchun 55P Hercynian ophiolites 11, 12, 14 Hidaka 599 'high tide marks' (HTMs) 25-26 high-magnesium andesites (HMAs) boninite-HMA volcanism 26 presence in proto-arcs 26 Hijaz terrane 686, 689 Hilti 316 Hiscocapsa grutterinki 157 Hokkaido 606 Hoots meta-ultramafic body 254 Hoste Islands 666 HOT ophiolites 597 Howqua 527, 526-527, 527 petrological and geochemical data 525 Howqua Fault 527 Hualien 55P Hulayfah-Ruwah suture zone 686, 689, 691 Bir Tuluha ophiolites 691 Darb Zubaydah ophiolites 691 Jabal Ghurrab ophiolites 691-692 Hulw 468 Huon peninsula 492 Huskisson, River 57P Huwl/Meeh subwindow 452, 453, 455-456 hydrothermal activity 354 geometric constraints on accretion 354-355 interaction temperatures 354 hydrothermal circulation and metamorphism, Bay of Islands ophiolite complex 369-370, 397-398 analytical methods 374 conditions of metamorphism 390 dynamically recrystallized gabbros 374-375, 375 geological setting 370-371, 377 geothermometry 387-390

INDEX comparison between mineralogical and oxygen isotope geothermometers 389 Hthological and structural units 372-374 mineral chemistry amphibole 381-382, 384, 385, 386 chlorite 382, 387 clinopyroxene 375-376, 381 epidote 386-387, 388 olivine 376 petrography, mineralogy and isotope composition data 377-380 plagioclase 376-381, 382, 383 prehnite 382-386, 388 pyroxene 376 oxygen isotope geochemistry calculated fluid/rock compositions 393 fluid/rock ratios 394-395 hydrothermal fluids 392-394 isotope compositions 390-391, 390, 391, 392 isotope equilibrium temperatures 391-392 structural domain 3 395, 396 structural domain 4 395-397, 396 sampling sites 373 static recrystallization 374, 375 Ibra Dome 469 Indochina 24 Indochina block 554 Indonesia and New Guinea (ING) region, geodynamic ophiolite patterns 480 convergence with Philippine Sea and Pacific plates 491 Central (Irian) ophiolite belt 491-492 Cyclops ophiolites 493-494 East Sulawesi ophiolites 495-496 Halmahera ophiolites 495 microplates 494 Molucca Sea region 494-495 New Guinea ophiolites 491 Papuan ophiolite belt 492-493 Talaud ophiolites 495 tectonic emplacement of New Guinea ophiolites 493 tectonic map 492 NE-directed subduction 486 Andaman Sea 486-487 Banda Sea basins 487-488 crustal cross-sections 490 geodynamics of Banda arc and ophiolite emplacement 489-491 Iherzolite bodies of Banda arc 489 Ocussi ophiolites of Timor 488—489 origin of ophiolites 496, 497, 499-500 changes in plate dynamics 498-499 mantle dynamics and marginal basin development 497-498 transtension 496-497 regional tectonic evolution 481-483, 484, 485 seismic tomography 486 Ingalls ophiolite complex 208 Inner Mongolia ophiolite belt 543, 548-549, 552 Ionian Sea 24 Irian (Central) ophiolite belt 491-492 Irian Jaya 485, 508 Ishigaki Island 599 Ispendere-Komurhan ophiolites 45

INDEX Istanbul 45 Iwai, River 298 Iwai Town 298 lyogatake, Mount 297 Izi-Ogasawara Trench 280 Izmir 45 Izmir-Ankara-Erzincan Suture Zone 45 Izu back-arc 289 Izu Islands 289 Izu-Bonin Islands 24 Izu-Bonin-Mariana (IBM) forearc system 27 incompatible element distributions 23 TiO2 vs. FeO plots 22 Jabal Ess ophiolites 687-690, 688 Jabal Ghurrab ophiolites 691-692 Jabal Thurwah ophiolites 690-691 Jadar block terrane 92 Japan, petrological diversity and origin of ophiolites 597-599, 598, 599, 608-610, 612-613 multiple nappe pile 608 NE Japan and Sakhalin 600, 601, 602-603 ophiolites with thick oceanic crust 610-612 origin of ophiolite-blueschist assemblage 612 SW Japan and Primorye 599-602, 600, 601, 602 Japan Trench 280 Java 24 Jebel Akhdar window 451, 468 Jebel Ja'Alan 468 Jeddah 689 Jeddah terrane 686 Jezerina 93 Jiangdong 556 Jiangsi 560 Jiaodong 543 Jiddah 689 Jiding 166, 167 mineral chemistry 176, 177, 179, 180, 181, 182, 183, 184 Jiding Donglha 150 Jiding massif 170 Jidong 192 Jinghong 554 Jingtai 553 Jinlu 166, 169 mineral chemistry 176, 177, 179, 180, 181, 182, 183, 184 Jinlu massif 171 Jinshajiang-Ailaoshan ophiolite belt 552-554 Jinzhuwan 556 Jiugequan 560 Jiugequan ophiolite 553 Jormua ophiolite complex carbon isotope distribution 423 general geology 416-417 Josephine Ophiolite (USA) 207-208, 208 geochemistry 212 element mobility during metamorphism 212-215 magmatic affinities and fractionation 215-220 pillow lavas and breccias 218 Th/Yb vs. Ta/Yb diagram 219 Ti vs. V discriminant diagram 216 TiO2 vs. La/Sm diagram 220 Y vs. Cr discriminant diagram 217 geological setting 208-210, 209 low-Ti dykes and lavas

major and trace element analyses 211, 213, 214, 221 MgO variation diagrams 275 occurrence 210 petrography 210-212 low-Ti magmas and mantle peridotite 222-223 petrogenesis 223-224 tectonic setting for eruption of boninites 224-225 Jungbwa, key characteristics of ophiolitic rocks 149 Jura 71 Kaishantun 560 Kaishantun ophiolites 559 Kalamaili 560 Kalinovka 599 Kamchatka Peninsula 607 Kamogawa City 295 Kamuikotan 599 Kanut 316 Karachi 44 Katmandu 554 Katsuyama Town 298 Kegiya 298 Kekesentao 560 Kempersay 569 Kengeveen 607 Kenting 560 Kenting melange 55P Khabarovsk 599 Khanka 599 Kharitonya terrane 637 Kimberley 518 Kinan Seamounts 289 Kizildag ophiolite 45 Klamath Mountains 208-209, 209 Kohistan island arc 44 Kojima 301 Koni-Pyagina Peninsula 620 Kopaonik block and ridge 92 Korab and Pelaginian terranes 92 Koryak Mountain ophiolite belts 599, 600, 603 Penzhina zone and Taigonos Peninsula 603-607 Koryak-Kamchatka fold belt 620 Koziakas Massif 111 Kraka 569 Krasnaya, Mount 609 Krivaja-Konjuh massif 92 KrstP3 Kuala Lumpur 554 Kuanping 560 Kudi 560 Kuerti 560 Kukes Massif 111 Kunming 542, 554 Kiire ophiolites 45 Kuyul terrane 637 geological setting 643-644 major and trace element compositions of ultramafic rocks 648, 649 microprobe analyses of Cr-spinels 646 mineral composition features 645 ophiolite stratigraphy 644 origin of the ophiolite 647-650 petrography and geochemistry 647 tectonostratigraphic units 644

707

708

INDEX

Lachlan Orogen 518, 521, 531 Cordilleran-type ophiolites 533, 534-536 Ladongri 558 Lagerny unit 626 Lajishan 553 Langeri 599 Lang0y Island 404 Lanzhou 553 Lanzo 71 peridotites field observations 73-75, 74, 76 geochemical data 79-80, 81 impregnation textures 75-78, 77 residual mantle mineral assemblages 75, 77 Laohushan 553, 560 large igneous provinces (LIPs) 10, 14, 15-16 Lasail 316 geological map 319 oxygen isotopic and petrographic data 321-322 plagiogranite bodies 334 Launceston 519 Laurasia 14 Lavrion Massif 111 Lepenac, River 93 Levantine Basin 24 Lewis Hills 377 Lhasa 44, 148, 166, 192, 554 Lhasa block 554 Lhasa terrane 148 Iherzolite-type ophiolites (LOT) 597-599, 598 Lichi 560 Lichi melange 559 Licola 521 age vs. rank of heating step 530 Pb isotope ratios 530 Ligurian ophiolites 10, 70 chemical refertilization and thermal erosion of lithosphere 85 extraction of melt in dykes 83-84, 84 general features 70-71, 71 extrusive rocks 73 field relations and petrography of serpentines and peridotites 71-72, 72 plutonic rocks 72-73 porous melt flow 82-83 Ligurian Sea 71 Ligurides 71 Litang 543, 560 lithological 'high tide marks' (HTMs) 25-26 Liuqu 148, 192 Lobastaya, Mount 609 Longlin 556 Longtang 555 LOT ophiolites 597 Lucipara Ridge 485 Luihe 556 Luobusa massif 148, 150, 166, 171, 792, 193, 554, 560 mineral chemistry 176, 177, 179, 180, 181, 182, 183, 184 ophiolitic rocks 158-159, 193-194 Luqu 167 Lusancay Rise 577 Lycian Nappes 45 Lyguorio 772

Macquarie Island ophiolites 431 emplacement mechanism 441 Madstone Cabin Thrust 209 Magadan 599 Magellan, Straits of 666 Maghrebides 24 Magnitogorsk 569 Mai Fault 527 Mainitis zone 599, 607-608 Makassar Strait 485, 490 Malay Peninsula 24 Malenco peridotite 77 mantle-driven rollback 50 Manus 492 Manus Trench 492 Maohashan 555 Maqin 554 Mariana arc 27 Mariana Islands 24, 289 incompatible element distributions 23 TiO2 vs. FeO plots 22 Mariana Trench 611 Mariana Trough 27, 289 incompatible element distributions 23 TiO2 vs. FeO plots 22 Markham Valley 577 Marlborough block 537 Mars Hill terrane 254 Marum Complex 508, 511 development phases 573 Mayile 560 Mclvor Hill 579 Mediterranean Sea, rollback cycles 25 Menglian 543, 560 Mentawai 484 Mergu Shelf 484 Mersin ophiolite 45 Mianlue 543, 560 Miass 569 mid-ocean ridge basalt (MORE) tectonic implication in Josephine Ophiolite 207-208, 215-220 pillow lavas and breccias 275 Th/Yb vs. Ta/Yb diagram 279 Ti vs. V discriminant diagram 276 TiO2 vs. La/Sm diagram 220 Y vs. Cr discriminant diagram 27 7 Midyan terrane 686, 689 Mikabu 599 Mindyak massif 569, 572-573, 572 bulk-rock and trace element data for mantle peridotites 578-579, 580, 581 influences on the geodynamic setting of peridotites 592-593 petrogenesis of peridotites origin of plagioclase 588-589 partial melting and melt flow processes 589-592 T and P estimates 5 88 whole-rock vs. pyroxene chemistry 589 petrographical and geochemical data for mantle peridotites 573, 575 trace element mineral chemistry clinopyroxene 582, 584-585, 587, 587 orthopyroxene 587-588 Mineoka 599

INDEX Mineoka ophiolites (Japan) 279-280, 286-287, 291 bathymetric map 280 eruption environment 288-291 faulting and deformation 299-300, 309-313 Benten Island 308-309 phases and stages of faulting 303 Shinyashiki outcrops 300-308, 302 formation site 287-288 plate tectonic reconstruction 288 geochronology Ar/Ar dating 286, 286, 287 K-Ar dating 284-286, 285 geology of Mineoka belt 280-283 geochemical compositions of bulk-rock samples 282-283 outcrop photographs 281 index map 300 lithological map 301 ocean basin formation ages 289 petrology and geochemistry 283-284, 284, 285 sampling locations 294, 295, 296, 297, 298 tectonic reconstruction models 290 Mineoka-Sengen 296 Mirdita (Albania) 45 carbon isotope distribution 423 general geology 416 interpolated P-T-t histories 29 interpretation of Tethyan ophiolites 50-53, 51 Miyamori 599 Moesian Platform 45 Molasse Basin 71 Molucca Sea 24, 494-495 crustal cross-section 490 Monte Maggiore 71 Moshkan Fault 131 Mount Dryden Belt 527 Mount Isa Inlier 518 Mount Wellington Belt 527, 535-528 Dolodrook 527-528 Dookie 523-524, 525, 526, 529 Howqua 526-527, 527 petrological and geochemical data 525 Tatong 524-526, 526 Mount William Fault 522 Munzur 45 Musandam Peninsula 468 Muscat 44, 451, 452, 468, 469 Musgrave Orogen 518 Nabitah-Hamdah fault zone 689, 692 Hamdah ophiolites 692 Nabitah ophiolites 692 Tathlith ophiolite complex 692 Nabitah-Hamdah suture zone 686 Nablyudeny 607 Najd fault system 689 Nambucca block 537 Namche Barwa 148, 192 Nankai Trough 280, 611 Naqqareh Khaneh 737 NE Asia, Early Cretaceous ophiolite accretionary complexes 619-621, 655-660 age, composition, structural location and geodynamic setting 657 analytical techniques 623-624

709

Cape Povorotny geological setting 624 major and trace element compositions 628, 628, 629, 632 origin of igneous and metamorphic rocks 628-631 petrography and geochemistry of metamorphic and igneous rocks 624-628, 625 spinel composition 630-631 tectonic units 626 continental growth of NE Asia 627 Ganychalan terrane geological setting 636-639, 637, 638 major and trace element compositions of magmatic rocks 640 major and trace element compositions of plutonic ophiolitic rocks 641, 643 origin of gabbroic and volcanic rocks 642-643 petrography and geochemistry of plutonic and volcanic rocks 639-642 geological framework 621-623 Kuyul terrane geological setting 643-644 major and trace element compositions of ultramafic rocks 648, 649 microprobe analyses of Cr-spinels 646 mineral composition features 645 ophiolite stratigraphy 644 origin of the ophiolite 647-650 petrography and geochemistry 647 tectonostratigraphic units 644 tectonic map of NE Asia 620 temporal and spatial distribution of ophiolites 656 Ust-Belaya segment geological map 657 major element compositions of mafic and ultramafic rocks 652, 653 major element compositions of ophiolitic peridotites 654 origin of ophiolites 655 petrography and geochemistry 652-655 tectonic setting 650-652 Yelistratov Peninsula geological setting 631-632 geological setting 633 major and trace element content of dykes and lavas 636 olivine compositions 634 origin of mafic and ultramafic rocks 634-636 petrography and geochemistry of mafic and ultramafic rocks 632-634 rare earth element contents of peridotites 635, 635 spinel compositions 633 Nea Epidavros 772 Neo-Tethys 43 distribution of ophiolites 44 oceanization of the Pindos Basin 109-110, 124-125 geochemical results 116-117, 117-118, 779, 720, 727, 722 geology of central-northern Argolis Peninsula 110 petrogenesis 119-122 petrography 117 sampling and methods 113-114, 114, 115 structure and stratigraphy of the Argolis Peninsula Middle Unit 110-113, 772 tectonomagmatic interpretation 118-119 Triassic magmatism 122-124, 724

710

INDEX

stratigraphic and sedimentological constraints on age and evolution 147^148, 160 Dazhuqu terrane - Ngamring Tso-Saga-Zongba 157-158 Dazhuqu terrane - Renbung-Quxu 158 Dazhuqu terrane - Xigaze 154-156, 755, 757 emplacement 159-160 geological map of Xigaze-Renbung area 750 key characteristics of ophiolitic rocks 149 radiolarian-based ages 158 radiolarians 757 Zedong-Luobusa 158-159 subduction rollback 49-50 New Britain 508, 511 New Britain arc 492 New Britain Trench 492, 511 New Caledonia 508, 509-510, 570 New England 537 New England Orogen 518 New Guinea see Indonesia and New Guinea (ING) New Ireland 492 Neyriz 737 Neyriz (Iran) ophiolites, palaeo-transform fault zone 129-130, 142-143 fossil oceanic transform fault 142 intra-ophiolitic shear zone 134 hornblende c- and a-axis preferred orientations 138, 139-140 kinematic model of mylonite origin 141-142, 143 microstructures 141 petrography and chemistry of mylonitic rocks 134, 135 sense of shear 138-141 shear-band cleavages 138 structure 135-138, 136-137 Neyriz ophiolite geology 130-134, 732, 733 regional geology 130, 737 Ngamba block 537 Ngamring 148 Nishi 296 Nizhly Tagil 569 Norfolk Ridge 508 North Anatolian Fault 24, 45 North Arm Mountain 377 structural map of gabbroic unit 372 North Banda Basin 485, 487-488 North China block 543 North D'Aguillar block 537 North Qilian ophiolites 543, 550-551, 553 North Qinling 543 Novi Sad 92 Nujiang 543 Nurali massif 569, 570-572, 577, 572 bulk-rock and trace element data for mantle peridotites 577-578, 580, 581 influences on the geodynamic setting of peridotites 592-593 petrogenesis of peridotites origin of plagioclase 588-589 partial melting and melt flow processes 589-592 Tand P estimates 588 whole-rock vs. pyroxene chemistry 589 petrographical and geochemical data for mantle peridotites 573-575, 573, 574 trace element mineral chemistry clinopyroxene 581-587, 582, 583-584, 586 orthopyroxene 587-588

ocean basin formation 69-70, 80-82, 82 chemical refertilization and thermal erosion of lithosphere 85 extraction of melt in dykes 83-84, 84 porous melt flow 82-83 oceanic-type crust development in the Rocas Verdes region 665-668, 676 causes of basin formation 679-680 closure and deformation of Rocas Verdes Basin 675-676 geological map of southernmost Andes 666 implications for mid-ocean ridge spreading 678-679 lithotectonic units of southernmost South America 667 mixed mafic-felsic terrane flanking ophiolites 669-671 progressive stages of continental rifting 676-678, 677 regional geology 668-669 Sarmiento ophiolite complex 671-674 distribution of metabasalts and metagabbros 674 sequential lithotectonic sections 668 Tortuga ophiolite complex 675, 675 Ocussi ophiolites (Timor) 488-489 Oeyama 599 Officer Basin 518 Ogasawara Islands 289 Ohatsubatake 297 Okhotosk 620 Okhotosk, Sea of 606, 620, 621 Oki Daito Ridge 289 Oldra Island 404 oxygen isotope composition 407 Oman Mountains 450-453, 451, 452, 453 geological map 468 subduction zone polarity and implications for ophiolite emplacement 467-469, 477-478 model A - NE-dipping subduction system 469-475, 470 model B - nascent SW-dipping subduction zone 471, 475-477 Oman model template for ophiolites 43-46, 59-61 interpretation 56-59, 57 mantle-driven rollback 50 Neo-Tethyan subduction rollback 49-50 overview of Tethyan ophiolite geology 46-48 paired ophiolite belts 48-49 tectonic diagram 60 plagiogranite bodies 315-317, 348-349 Aarja-Bayda area 320, 328-334, 32P, 330-333 Aarja-Bayda area, oxygen isotopic and petrographic data 323-324 chemical characteristics 339-344, 341, 342, 343 geology and petrology of gabbro-sheeted dyke complex (SDC) 318-328 Lasail-Assayab area 37P, 334, 335 Lasail-Assayab area, oxygen isotopic and petrographic data 321-322 origin 344-346, 345 oxygen isotope studies 335-339, 336, 337, 339, 340 regional geological map 376 regional setting 317-318 relationship to ore deposits 346-348, 346, 348 sample geochemistry 326-327, 328 Suhaylah intrusion 325, 334-335 Suhaylah intrusion, oxygen isotopic and petrographic data 324

INDEX Semail ophiolites collisional emplacement mechanism 438 constraints on obduction process 454-460 incompatible element distributions 23 interpolated P-T-t histories 29 obduction and behaviour of underlying margin 449-450, 450, 453-454, 461-462, 462 revised obduction process 460-461 TiO2 vs. FeO plots 22 Omolon 620 Ondor Sum 560 Onib suture zone 686 Ontong Java Platform 508 ophiolites age distribution 11,14 Cambrian 10-11 Jurassic-Cretaceous 12-13 Palaeozoic 11-12 Phanerozoic 13-15 Precambrian 13 Proterozoic 10, 13-15 bioalteration recorded in pillow lavas 415-416, 424-425 biogenerated textures 417, 418, 419, 420 carbon isotopes 419-421, 423, 424 element mapping 417-419, 421, 422 organic remains 417, 420 perspectives 421-424 Cordilleran-type ophiolites 430 definition 9-10 Early Cretaceous accretionary complexes 619-621, 655-660 age, composition, structural location and geodynamic setting 657 analytical techniques 623-624 Cape Povorotny complex 624-631 continental growth of NE Asia 621 Ganychalan terrane 636-643 geological framework of NE Asia 621-623 Kuyul terrane 643-650 tectonic map of NE Asia 620 temporal and spatial distribution of ophiolites 656 Ust-Belaya segment 650-655 Yelistratov Peninsula 631-636 emplacement mechanisms 427-428, 441-442 accretionary emplacement of Cordilleran-type ophiolites 438-440 collisional emplacement of Tethyan-type ophiolites 435-438, 436-437 complex plate interactions 441 descriptive model 428, 429 emplacement of oceanic lithosphere over less dense rocks 435 exhumation of metamorphic soles 434 Macquarie Island ophiolites 441 ophiolite prototypes 428-431 ridge-trench intersection (RTI) ophiolites 440-441 structure and evolution of metamorphic soles 431-434, 433 subduction initiation models 434-435 formation 16 Franciscan-type 10 general geology of ophiolite complexes Jormua ophiolite complex (JOC) 416-417 Mirdita ophiolite complex (MOC) 416 Solund-Stavfjord ophiolite complex (SSOC) 416

711

Troodos ophiolite complex (TOC) 416 geochemical cycles 353-354, 365-366 boundary and initial conditions for hydrothermal activity 354-355 comparison with continental layered gabbro complexes 358 comparison with oceanic layer-3 oxygen isotope levels 357-358 global weathering rates 362-364 profile through Samail ophiolite 355-357 rate constant 359-362, 360, 361 seawater composition 354 significance of ophiolite studies 364-365 special significance of oxygen isotope profile for seawater 357 whole-rock 618O values for pillow lavas through time 358-359, 359 geographical distribution 11, 14 Africa 10 Appalachian 11-12 Australia 10-11 Brazil 10 Caledonian 12 Cordilleran 13 Hercynian 12 Tethyan-Caribbean 12-13 Uralian 12 Western Pacific 13 geology within Tethyan oceans association of mature arc complexes 47-48 initial rifting and precursory oceanic crust formation 46 metamorphic soles and ophiolitic melanges 46-47 Ligurian-type 10 Macquarie Island ophiolites 431 overview 1,7-8 emplacement mechanisms and processes 5 hydrothermal and biogenic alteration of oceanic crust 4-5 magmatic, metamorphic and tectonic processes in ophiolite genesis 3-4 ophiolites, mantle plumes and orogeny 1 regional occurrence and geodynamic implications 5-7 Tethyan ophiolites in the Alpine-Himalayan orogenic system 1-3 petrological diversity and origin in Japan and Russia 597-599, 598, 599, 608-610, 612-613 geological map 606 mineral analyses 604-605 multiple nappe pile 608 ophiolites in NW Pacific margin 599-608, 600 ophiolites with thick oceanic crust 610-612 origin of ophiolite-blueschist assemblage 612 plumes and orogeny 13-16 ridge-trench intersection (RTI) ophiolites 430-431 Sierran-type 10 Tethyan-type ophiolites 430 Orahovac 92 Orleans Thrust 209 Orsk 569 Ostrovica 93 Othrys Massif 777 Owen Stanley Ranges 577 Ozren of Doboj massif 92 Ozren of Sjenica massif 92

712

INDEX

Pacific Ocean, arc-trench rollback cycles 25-26 Pacific Plate 27 Pacific Rim ophiolites 10 Pacific, Western, ophiolites 13 Pacific, Western, ophiolites 14 Palau-Kyushu Ridge 24, 26, 27 Palea Epidavros 112 Paleo-Asian Ophiolite Zone 543 Paleo-Asian Orogenic Zone 542 Palu-Koro Fault 490 Pangea 14 Pannotia 14 Papua 492 Papuan ophiolite belt 492-493, 492 Parece Vela Basin 24, 27, 289 Paterson Orogen 518 Pec 92

Peel-Manning Fault 531 Penzhinskaya Guba Bay 625 peridotites melt migration 69-70, 80-82, 82 chemical refertilization and thermal erosion of lithosphere 85 extraction of melt in dykes 83-84, 84 geochemical data 78-80, 79-80, 81 Lanzo and Corsica 73-78, 74, 76 Piedmonte Ligurian ophiolites 70-73, 71 porous melt flow 82-83 Philippine Trench 484 Philippines 24 Philippines ophiolites 10 Phillip Island 521 Piedmont ophiolites 70 chemical refertilization and thermal erosion of lithosphere 85 extraction of melt in dykes 83-84, 84 general features 70-71, 71 extrusive rocks 73 field relations and petrography of serpentines and peridotites 71-72, 72 plutonic rocks 72-73 porous melt flow 82-83 Pilbara Craton 518 Pindos Basin 109-110, 124-125 geochemical results 117-118, 120, 121, 122 bulk-rock major and trace elements 116-117 incompatible element abundance patterns 119 geology of central-northern Argolis Peninsula 110 petrogenesis 119-122 petrography 117 sampling and methods analytical methods 113-114 locations 113, 114, 115 structure and stratigraphy of the Argolis Peninsula Middle Unit 110-113, 112 tectonomagmatic interpretation 118-119 Triassic magmatism 122-123 geodynamic implications 123, 124 modern oceanic analogue 123-124 Pindos Massif 111 Pindos ophiolites 45 Platta ophiolites 71 Pnom-Penh 554 Po Basin 71 Pocklington Rise 492

Pocklington Trough 492 Polio 150 Polyakovka 571 Pontides 44, 45 Post Davey metamorphic complex 519 Povorotny, Cape 607 ophiolite accretionary complexes geological setting 624 major and trace element compositions 628, 628, 629, 632 origin of igneous and metamorphic rocks 628-631 petrography and geochemistry of metamorphic and igneous rocks 624-628, 625 spinel composition 630-631 tectonic units 626 Pozanti-Karsanti ophiolites 45 Priboj 92 Pristina 92 proto-ophiolite to ophiolite transition 31-32 Pseudodictyomitra hornatissima 157 Pseudodictyomitra pentacolaensis 157 Pylgin suture 625 Qaidam block 554 Qiangtang block 554 Qinghai Lake 553 Qinling-Qilian-Kunlun Ophiolite Zone 543 North Qilian ophiolites 550-551, 553 West Kunlun ophiolites 549-550 Qumei 167 Qunrang massif 750, 168, 171 mineral chemistry 176, 177, 179, 180, 181, 182, 183, 184 Quriat 452, 469 Quxu 148 ophiolitic rocks 158 Radho 112 Ragmi 316 Rahab 320 Raishu 295 Rajmi 316 Ramu Valley 577 Raohe 560 Ras Al Hadd 468 Read, Mount 519 Red River Fault 24 Renbung 148, 150, 192 ophiolitic rocks 158 Rhodope Massif 45 ridge-trench intersection (RTI) ophiolites 430-431 emplacement mechanism 440-441 Rilong 558 Rocas Verdes ophiolites 665-668, 676 causes of basin formation 679-680 closure and deformation of Rocas Verdes Basin 675-676 geological map of southernmost Andes 666 implications for mid-ocean ridge spreading 678-679 lithotectonic units of southernmost South America 667 mixed mafic-felsic terrane flanking ophiolites 669-671 progressive stages of continental rifting 676-678, 677 regional geology 668-669 Sarmiento ophiolite complex 671-674 distribution of metabasalts and metagabbros 674 sequential lithotectonic sections 668 Tortuga ophiolite complex 675, 675

INDEX Rockhampton 537 Rocky Boat harbour 519 Rocky Cape Block 519 Rodinia 10, 14 Rostaqi Fault 131 Rusayl 469 Russia, petrological diversity and origin of ophiolites 597-599, 598, 599, 608-610, 612-613 geological map 606 Koryak Mountains 600, 603 Mainitis zone 607-608 mineral analyses 604-605 multiple nappe pile 608 ophiolites with thick oceanic crust 610-612 origin of ophiolite-blueschist assemblage 612 Penzhina zone and Taigonos Peninsula 600, 603-607, 607 Ryukyu Islands 24 Saas Zermatt ophiolites 71 Saga 148 Sagami Bay 298 Sagami Trough 280 Sagui 750, 192 Saham 316 Saih Hatat window 451, 452, 455, 468 domal culmination 456 fold-nappe 457, 458 structural sections 474 Sakhalin 606 Sakuma, River 298 Samarkand 44 San Cristobal Trench 492 Sana-Una terrane 92 Sangihe Arc 485, 490 Sangri 750, 792 Sangsang 7 48 Sarajevo 92 Sardinia 24 Sarmiento ophiolite complex 666, 671-674 distribution of metabasalts and metagabbros 674 geological map 670 mafic dykes 667 schematic sections 671, 672 sheeted dykes 667 sea-floor spreading in Thetford Mines Ophiolite Complex 231-232,246 deformation 239-241 lower crust 241-242 upper crust 240, 242 formation of boninitic crust 246, 247 geological setting 232, 233 model for the structural and magmatic evolution 244-245, 244, 245 sequence of crystallization and magmatic affinity 243 sheeted dyke complex and crustal polarity 243 slow-spreading environment 245-246 stratigraphy 235 hypabyssal facies 239 mantle section 235 plutonic crustal section 235-239, 236, 237, 238 stratigraphic columns 234 volcano-sedimentary facies 239 volcanic stratigraphy 243 seawater, ophiolites and oxygen isotopes 353-354, 365-366

713

boundary and initial conditions for hydrothermal activity 354 geometric constraints on accretion 354-355 interaction temperatures 354 composition 354 global weathering rates 362 epicontinental seaways 363-364 short-term water cycle effects 362-363, 363 ophiolite record of ancient isotope ratios 401-402, 412 oxygen isotope profile of Ordovician seawater 409-412 quartz and epidote mineral pair analysis 407-409, 410, 477 whole-rock data 406-407, 407, 408, 408, 409 profile through oceanic crust 355-356 neodymium residence time 356 oxygen residence time 356-357 strontium residence time 356 rate constant 359-362, 360, 361 Samail ophiolites and continental layered gabbro complexes 358 Samail ophiolites and oceanic layer-3 oxygen isotope levels 357-358 sea-floor-seawater 18O exchange 402-403 significance of ophiolite studies 364-365 special significance of oxygen isotope profile for seawater 357 water balance 364 whole-rock <518O values for pillow lavas through time 358-359, 359 Seh-Qalatouin 737 Seh-Qalatouin Fault 737 Semail (Oman) ophiolites background structure of Oman Mountains 450-453, 451, 452, 453 collisional emplacement mechanism 438 constraints on obduction process age constraints 459-460 metamorphism of underlying rocks 456-458 Samail Nappe 454-455 structure of underlying rocks 455-456 incompatible element distributions 23 interpolated P-T-t histories 29 interpretation of Tethyan ophiolites 56-59, 57 isotopic profiles 355-356, 355 neodymium residence time 356 oxygen residence time 356-357 strontium residence time 356 obduction and behaviour of underlying margin 449-450,

450,461-462,462 previous ideas 453-454 plagiogranite bodies 315-317, 348-349 Aarja-Bayda area 320, 328-334, 329, 330-333 Aarja-Bayda area, oxygen isotopic and petrographic data 323-324 chemical characteristics 339-344, 341, 342, 343 geology and petrology of gabbro-sheeted dyke complex (SDC) 318-328 Lasail-Assayab area 379, 334, 335 Lasail-Assayab area, oxygen isotopic and petrographic data 321-322 origin 344-346, 345 oxygen isotope studies 335-339, 336, 337, 339, 340 regional geological map 376 regional setting 317-318 relationship to ore deposits 346-348, 346, 348

INDEX

714

sample geochemistry 326-327, 328 Suhaylah intrusion 325, 334-335 Suhaylah intrusion, oxygen isotopic and petrographic data 324 revised obduction process 460 age of eclogite metamorphism 460 P-T conditions of metamorphic sole 460-461 suprasubduction character 461 tectonic scenarios 461 vergence of underlying rock structure 461 subduction zone polarity in Oman Mountains 467-469, 477-478 model A - NE-dipping subduction system 469-475, 470 model B - nascent SW-dipping subduction zone 471, 415-411

TiO2 vs. FeO plots 22 Semail Thrust 468, 469 Seram 485 Serbo-Macedonian composite terrane 92 Sergeevka 599 Serikeyayilake 560 Serpentine Hill 519 Settlers schist 519 Shafan 316 Shanghai 542 Shan-Tai-Ma block 554 Sharipovo 577 Shchel Bay 626 Shebenik Massif 111 Shelting 599 Shelting Cape 606 Shelting massif 606 Shikoku Basin 24, 289 Shimosakuma 298 Shinganji 295 Shinyashiki 295, 301, 302 fault analysis 300-308 first-stage faults 302-303, 304 plan view of 'Peninsulet' 303 second-stage faults 303-305, 305, 306, 307, 309 third-stage faults 305-308, 307, 308 outcrop photograph 302 Shirataki 296 Shiraz 44 Shoalwater block 537 Shuanggou 554, 560 geological map and cross-section 555 Shuanghu 554 Siberian craton 620 Sibirskaya 542 Sicily 24 Sierra Navada ophiolites 10 Sierran-type ophiolites 10 Simao block 554 Sjenica 92 Skoplje 92 'slab pull' 30, 30 slab rollback 9-10 Smartville ophiolite 208 Smithton Basin 57P Sognefjorden 404 Sohar 316, 468 Solomon Islands 492, 508 Solomon Sea 508, 511

Solomon Trench 492 Solonshan 560 Solund-Stavfjord Ophiolite Complex (Norway) 401-402 carbon isotope distribution 423 general geology 416 isotopic data oxygen isotope profile of Ordovician seawater 409-412 quartz and epidote mineral pair analysis 407-409, 410, 411 whole-rock data 406-407, 407, 408, 408, 409 ophiolite record of ancient seawater isotope ratios 412 sea-floor-seawater 18O exchange 402-403 setting 403^405, 404 extrusive rocks 405 high-level gabbro 405 petrography 405-406 sheeted dyke complex 405 transition zone 405 Songpan block 554 Songshan 556 Songshugou 560 Sorell, Cape 57P Soro, River 296 Sorogawa, River 307 Sorogawa Fault 307 Sorong Fault 484 Sorong Fault Zone 490 South Banda Basin 485, 487-488 South Bismarck plate 492 South China Sea 24 South D'Aguillar block 537 Southern Alps 77 Southern Uralides, mineral chemistry of ultramafic massifs 567-568 analytical techniques 575-576 bulk-rock and trace element chemistry 576, 577-579, 580, 581 field relations Mindyak massif 572-573, 572 Nurali massif 570-572, 577, 572 geological background 568-570, 569 influences on the geodynamic setting of Nurali and Mindyak peridotites 592-593 major element mineral chemistry 576-581 petrogenesis of Nurali and Mindyak peridotites origin of plagioclase 588-589 partial melting and melt flow processes 589-592 Tand P estimates 588 whole-rock vs. pyroxene chemistry 589 petrography Mindyak peridotites 573, 575 Nurali peridotites 573-575, 573, 574 sample selection for analysis 573 trace element mineral chemistry clinopyroxene 581-587, 582, 583-584, 586 orthopyroxene 587-588 Spero Bay 57P Srbica 92 St Joseph Fault 233 St Lawrence, Gulf of 3 77 Stavely Belt 527 Stewart, Mount 57P Stichomitra communis 157 Strand Island 404

INDEX oxygen isotope composition 408 oxygen isotope vs. depth profile 409 subduction micleation 26-29, 32, 33 Suhaylah 316 geological map 325 oxygen isotopic and petrographic data 324 plagiogranite bodies 334-335 Sula 485, 490 Sulawesi 24, 485, 495-496 crustal cross-section 490 Sumatra 24 Sumatra Fault 484 Sumba 485 Sumeini 468 Sumisu Rift 24 incompatible element distributions 23 TiO2 vs. FeO plots 22 Sunan 553 Sunda 24 Sunda arc-trench system 484 Sunda craton 484 Suno 55P suprasubduction zone (SSZ) environments 9-10 Susunai 599 Sydney Basin 537 Table Mountain 377 Taigonos Peninsula 599 Taigonos, Cape 607, 623 tectonic map 625 Taipei 55P, 599 Taiwan 24 Takatsuru, Mount 296 Talaud Islands 490, 495 Talovka sequence 644 Tamworth belt 537 Tananao 560 Tananao melange 55P Tangbale 560 Tang-e-Hana 732 Tarim Basin 44 Tarim block 542 Tasman Line 518 Tasman Sea 508 Tasmania 57P Tasmanides, evolutionary implications of Australian ophiolites 10-11, 517-518,575 Delamerian-Tyennan ophiolites 518 Delamerian Orogen occurrences 520 Tyennan (Tasmanian) Orogen occurrences 518-520 Lachlan Orogen ophiolites 520-521, 527 Barkly River Belt 528-529, 529 Heathcote Belt 521-523, 522, 524 Mount Wellington Belt 535-528 occurrences of unknown connections 529-531 main features 533 New England Orogen ophiolites 531 northern occurrences 531-532, 537 southern occurrences 532-533 tectonic setting and evolution of proto-Tasmanides 533-534 Tatar Strait 606 Tathlith ophiolite complex 692 Tatong 527, 524-526, 526 Taurides 44

715

Tehran 44 Tenjingo Town 2P7 Tethyan-type ophiolites 430 collisional emplacement 435-438, 436-437 eastern Australia 517-518, 518 Delamerian Orogen 533, 534-536 Tyennan (Tasmanian) Orogen occurrences 518-520 Tethys collision-induced mantle flow model for ophiolite formation 21-23, 24, 32-33 collision-induced mantle extrusion 31 endogenous vs. exogenous causes 29-31 forearc complexes 25-26 interpolated P-T-t histories 2P model development 25-29, 27 partial melting models 28 plate kinematic effects 29-30 proto-ophiolite to ophiolite transition 31-32 slab pull and extrusion tectonics 30, 30 subduction nucleation 26-29 history 23-24 Palaeo-Tethys 24-25 Neo-Tethys 25 ophiolite geology association of mature arc complexes 47-48 initial rifting and precursory oceanic crust formation 46 metamorphic soles and ophiolitic melanges 46-47 Thanarla brouweri 157 Thetford Mines Massif (TMM) 232-234 Thetford Mines Ophiolite Complex (TMOC), forearc extension and sea-floor spreading 231-232 deformation 239-241 lower crust 241-242 upper crust 240, 242 description 232-235 formation of boninitic crust 246, 247 geological setting 232, 233 model for the structural and magmatic evolution 244-245, 244, 245 sequence of crystallization and magmatic affinity 243 sheeted dyke complex and crustal polarity 243 slow-spreading environment 245-246 stratigraphy 235 hypabyssal facies 239 mantle section 235 plutonic crustal section 235-239, 236, 237, 238 stratigraphic columns 234 volcano-sedimentary facies 239 volcanic stratigraphy 243 tholeiites tectonic implication in Josephine Ophiolite 207-208, 215-220 pillow lavas and breccias 218 Th/Yb vs. Ta/Yb diagram 27P Ti vs. V discriminant diagram 276 TiO2 vs. La/Sm diagram 220 Y vs. Cr discriminant diagram 277 Thomson Orogen 518, 531 Tianshan ophiolite belt 543, 548 Tien Shan 24 Tierra del Fuego 666 Timor 485, 490 Ocussi ophiolites 488-489 Toge 2P7

716

INDEX

Tokar terrane 686 Tokyo 599 Tolovka, River 637 Tongchang 556 Tongchangjie 554, 560 Topolovka 607 Tortuga ophiolite complex 666, 675, 675 geological map 670 schematic sections 671, 672 Totalp peridotites 71 Trial Harbour 519 Trobriand Trough 492 Tromeda 93 Troodos ophiolites 45 carbon isotope distribution 423 general geology 416 incompatible element distributions 23 TiO2 vs. FeO plots 22 Troodos-Mamonia interpolated P—T—t histories 29 interpretation of Tethyan ophiolites 53-56, 54 Tropoja Massif 111 Tukang Besi Platform 485 Turbocapsula costata 157 Tuzia 92 Tyennan ophiolites 518-520, 518 Tylpyntyhlavaam sequence 644 Tyrrhenian Sea 24, 71 Udachny sequence 644 Unnavayam sequence 644 Upupkin terrane 637 Uraga Path 298 Uralian Ocean, ophiolite petrogenesis 11-12, 11, 14, 567-568 analytical techniques 575-576 bulk-rock and trace element chemistry 576, 577-579,580, 581 field relations Mindyak massif 572-573, 572 Nurali massif 570-572, 577, 572 Nurali peridotites 573-575, 573, 574 geological background 568-570, 569 influences on the geodynamic setting of Nurali and Mindyak peridotites 592-593 major element mineral chemistry 576-581 petrogenesis of Nurali and Mindyak peridotites origin of plagioclase 588-589 partial melting and melt flow processes 589-592 Tand P estimates 588 whole-rock vs. pyroxene chemistry 589 petrography Mindyak peridotites 573, 575 sample selection for analysis 573 trace element mineral chemistry clinopyroxene 581-587, 582, 583-584, 586 orthopyroxene 587-588 Urumqi 542 Ust-Belaya segment geological map 651 major element compositions of mafic and ultramafic rocks 652, 653 major element compositions of ophiolitic peridotites 654 origin of ophiolites 655 petrography and geochemistry 652-655 tectonic setting 650-652

Vardar zone 92 Vanuatu 508 Vardar Zone ophiolites 45 Verkhoyansk-Chukotka fold belt 620 Verkn Ufaly 569 Vermion Massif 111 Veselaya sequence 644 Vitayetglia unit 626 Vladivostok 599 Vogel, Cape 508 Voshnezenka 577 Vothiki 772 Vourinos Massif 777 Vourinos ophiolites 45 Vstrechny sequence 644 Wadi Adai 452, 453 Wadi Ham 468 Wadi Hatta 468 Wadi Jizi 376, 468 Wadi Meeh gorge 452, 453 Wadi Tayin 469 Wagga-Omea metamorphic belt 527 Wandilla block 537 Waratah Bay 527 water balance, global 364 Weber Basin 485 Webster-Addie meta-ultramafic body 254 West Junggar ophiolite belt 543, 545-547, 549 West Kunlun ophiolites 543, 549-550 West Mariana arc 24, 27 West Philippine Sea Basin 24, 27, 289 Western Alps 77 Wetar Ridge 485, 490 Wetar Thrust 490 Wilson Cycles 31 mantle-driven sub-cycle 32, 33 Wilson, River 519 Wonominta Block 518 Woodlark Basin 492 Woody meta-ultramafic body 254 Xialu 750, 168, 192 Xigaze 148, 150, 166, 167, 192, 554, 560 key characteristics of ophiolitic rocks 149 ophiolitic rocks 154-156, 755, 757, 194-196 sheeted dyke swarm 557 Xining 553 Xitus clava 157 Yakuno 599 Yamada 297 Yanbu suture zone 686, 687-690, 689 AFAyes ophiolites 690 Jabal Ess ophiolites 687-690, 688 Yangtze block 554 Yarlung Tsangpo (Tibet) suture zone 44, 555-557 geochemical and geochronological constraints 191-192, 205 Dazhuqu terrane ophiolites 192-193 geochemical analyses 196, 197, 795, 799 geochronology 196-201, 200, 201, 202 geological map 793, 795 geological setting 192-196, 792 Luobusa massif 193-194

INDEX major terrane characteristics 193 ophiolite formation 201 regional geology 192 tectonic model 203, 204 tholeiitic and boninitic magmatism 201-204 geodynamic implications from mineral record 165-166, 187-188 amphibole 181-183 clinopyroxene 180-181 crustal unit overview 166-170 geodynamic model 185-187, 187 geothermobarometry 183-184 location map of ophiolite massifs 166 mineral chemistry 171-183, 172-175, 176, 177, 178, 179 olivine 179 orphopyroxene 179-180 palaeogeodynamic settings 185 petrogenesis 184-185, 186 plagioclase 181 spinel 177-179 ultramafic unit 170-171 stratigraphic and sedimentological constraints 147-148, 160 Dazhuqu terrane - Ngamring Tso-Saga-Zongba 157-158 Dazhuqu terrane - Renbung-Quxu 158 Dazhuqu terrane - Xigaze 154-156, 155, 157 emplacement 159-160 geological map of Xigaze-Renbung area 150 key characteristics of ophiolitic rocks 149 radiolarian-based ages 158 radiolarians 157 regional tectonic framework 151-154 sketch map 148 time-space plot for terranes 151

717

Zedong-Luobusa 158-159 Yarlung Tsangpo, River 166, 169, 193, 543 Yarrol block 531 Yaw Fault 527 Yekaterinburg 569 Yelistratov Peninsula, ophiolites geological setting 631-632, 633 major and trace element content of dykes and lavas 636 olivine compositions 634 origin of mafic and ultramafic rocks 634-636 petrography and geochemistry of mafic and ultramafic rocks 632-634 rare earth element contents of peridotites 635, 635 spinel compositions 633 Yenong 167 Yilgarn Craton 518 Yiti 469 Yooka Beach 301 Yueyashan 560 Yushigou 560 Yushigou ophiolite 553 Zagapu 150 Zagreb 44 Zedang 166 Zedong 148, 150, 192 key characteristics of ophiolitic rocks 149 ophiolitic rocks 158-159 Zhangye 553 Zhongzhan block 554 Zir Qalate 131 Zisong 167 Zlatibor massif 92 Zongba 148