Tectonic development of the Eastern Mediterranean region

Tectonic Development of the Eastern Mediterranean Region The Geological Society o f L o n d o n Books Editorial Comm...

Author: A. H. F. Robertson | Demosthenis Mountrakis

83 downloads 1766 Views 75MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Tectonic Development of the Eastern Mediterranean Region

The Geological Society o f L o n d o n

Books Editorial Committee Chief Editor BOB PANKHURST(UK)

Society Books Editors JOHN GREGORY (UK) JOHN HOWE (UK) NICK ROBINS (UK) JIM GRIFFITHS (UK) PHIL LEAT (UK) JONATHAN TURNER (UK)

Society Books Advisors MIKE BROWN (USA) JON GLUYAS (UK) RANDELL STEPHENSON(NETHERLANDS) RETO GIERI~(GERMANY) DOUG STEAD (CANADA) SIMON TURNER (AUSTRALIA)

Geological Society books refereeing 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 Book Editors 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. More information about submitting a proposal and producing a book for the Society can be found on its 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: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260. MUCEKU, B., MASCLE, G. H. & TASHKO, A. 2006. First results of fission-track thermochronology in the Albanides. In: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 539-556.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 260

Tectonic Development of the Eastern Mediterranean Region

EDITED BY A. H F. R O B E R T S O N University of Edinburgh, UK and D. M O U N T R A K I S Aristotle University, Thessaloniki, Greece

2006 Published by The GeologicalSociety 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 U K 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 0BG: 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, U K Tel. +44 (0)1225 445046 Fax +44 (0)1225 442836 Online bookshop: www.geolsoc.org.uk/bookshop Orders:

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. 9 The Geological Society of London 2006. 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 WIP 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/06/$15.00.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 10 1-86239-198-X ISBN 13 978-1-86239-198-7 Typeset by The Charlesworth Group Ltd, Wakefield, 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 +1 918 560-2652 E-mail [email protected] India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. +91 11 2327-9113/2326-4180 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 ROBERTSON,A. H. F. & MOUNTRAKIS,D. Tectonic development of the Eastern Mediterranean region: an introduction SMITH,A. G. Tethyan ophiolite emplacement, Africa to Europe motions, and Atlantic spreading HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. Late Proterozoic and Silurian basement units within the Serbo-Macedonian Massif, northern Greece: the significance of terrane accretion in the Hellenides YANEV, S., GONCIOOI3LU,M. C., GEDIK, I., LAKOVA,I., BONCHEVA,I., SACHANSKI, V., OKUYUCU, C., OZGf0L,N., TIMUR, E., MALIAKOV,Y. & SAYDAM,G. Stratigraphy, correlations and palaeogeography of Palaeozoic terranes of Bulgaria and NW Turkey: a review of recent data ROMANO, S. S., BRIX, M. R., DORR, W., FIALA,J., KRENN, E. & ZULAUF, G. The Carboniferous to Jurassic evolution of the pre-Alpine basement of Crete: constraints from U-Pb and U-(Th)-Pb dating of orthogneiss, fission-track dating of zircon, structural and petrological data ROBERTSON, A. H. F. Sedimentary evidence from the south Mediterranean region (Sicily, Crete, Peloponnese, Evia) used to test alternative models for the regional tectonic setting of Tethys during Late Palaeozoic-Early Mesozoic time KARAMATA,S. The geological development of the Balkan Peninsula related to the approach, collision and compression of Gondwanan and Eurasian units KAZMIN, V. G. & TIKHONOVA,N. F. Evolution of Early Mesozoic back-arc basins in the Black Sea-Caucasus segment of a Tethyan active margin GARDOSH, M. A. & DRUCKMAN,Y. Seismic stratigraphy, structure and tectonic evolution of the Levantine Basin, offshore Israel DANELIAN, T., ROBERTSON,A. H. F., COLLINS,A. S. & POISSON,A. Biochronology of Jurassic and Early Cretaceous radiolarites from the Lycian M61ange (SW Turkey) and implications for the evolution of the Northern Neotethyan ocean RASSIOS,A. H. E. & MOORES, E. M. Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab (northern Greece) KOLLER, F., HOECK, V., MEISEL,T., IONESCU,C., ONUZI, K. & GHEGA, D. Cumulates and gabbros in southern Albanian ophiolites: their bearing on regional tectonic setting GARFUNKEL, Z. Neotethyan ophiolites: formation and obduction within the life cycle of the host basins RIZAO~LU, T., PARLAK,O., HOECK, V. & lSLER, F. Nature and significance of Late Cretaceous ophiolitic rocks and their relation to the Baskil granitic intrusions of the Elam~ region, SE Turkey MORRIS, A., ANDERSON,M. W., INWOOD,J. & ROBERTSON,A. H. F. Palaeomagnetic insights into the evolution of Neotethyan oceanic crust in the eastern Mediterranean SHARV, I. R. & ROBERTSON,A. H. F. Tectonic-sedimentary evolution of the western margin of the Mesozoic Vardar Ocean: evidence from the Pelagonian and Almopias zones, northern Greece RICE, S., ROBERTSON,A. H. F. & USTAOMER,T. Late Cretaceous-Early Cenozoic tectonic evolution of the Eurasian active margin in the Central and Eastern Pontides, northern Turkey RIMMELI~, G., OBERHANSLI,R., CANDAN,O., GOFFI~,B. & JOLIVET,L. The wide distribution of HP-LT rocks in the Lycian Belt (Western Turkey): implications for accretionary wedge geometry DEGNAN, P. J. & ROBERTSON,A. H. F. Synthesis of the tectonic-sedimentary evolution of the Mesozoic-Early Cenozoic Pindos ocean: evidence from the NW Peloponnese, Greece

1 11 35

51

69

91

155 179 201 229

237 267 301 327

351 373

413

447

467

PIPER, D. J. W. Sedimentology and tectonic setting of the Pindos Flysch of the Peloponnese, Greece DOUTSOS, T., KOUKOUVELAS,I. K. & XYPOLIAS,P. A new orogenic model for the External Hellenides VAMVAKA,A., KILIAS,A., MOUNTRAKIS,D. & PAPAOIKONOMOU,J. Geometry and structural evolution of the Mesohellenic Trough (Greece): a new approach MUCEKU, B., MASCLE,G. H. & TASHKO,A. First results of fission-track thermochronology in the Albanides WZSTAWAu R. Late Cenozoic extension in SW Bulgaria: a synthesis AL~ICEIr M. C., WENVEEN, J. H. & OZKUL, M. Neotectonic development of the ~ameli Basin, southwestern Anatolia, Turkey BOULTON, S. J., ROBERTSON,A. H. F. & ~JNLOGENC,U. C. Tectonic and sedimentary evolution of the Cenozoic Hatay Graben, Southern Turkey: a two-phase model for graben formation PAVLIDES, S. B., CHATZIPETROS,A., TUTKUN, Z. S., ()ZAKSOY,V. & DOriAN, B. Evidence for late Holocene activity along the seismogenic fault of the 1999 Izmit earthquake, NW Turkey MOUNTRAKIS,D., TRANOS,M., PAPAZACHOS,C., THOMAIDOU,E., KARAG1ANNI,E. & VAMVAKARIS,D. Neotectonic and seismological data concerning major active faults, and the stress regimes of Northern Greece TRANOS, M. D., KARAKOSTAS,V. G., PAPADIMITRIOU,E. E., KACHEV,V. N., RANGUELOV, B. K. & GOSPOD~NOV,D. K. Major active faults of SW Bulgaria: implications of their geometry, kinematics and the regional active stress regime PAPAZACHOS,B. C., KARAKAISIS,G. F., PAPAZACHOS,C. B. & SCORDIL1S,E. M. Perspectives for earthquake prediction in the Mediterranean and contribution of geological observations

493

Index

709

507 521 539 557 591 613

635

649

671

689

Tectonic development of the Eastern Mediterranean region: an introduction ALASTAIR

H . F. R O B E R T S O N 1 & D E M O S T H E N I S

MOUNTRAKIS

2

l Grant Institute o f Earth Science, School o f GeoSciences, University o f Edinburgh, West Mains Road, Edinburgh EH9 3JW, UK (e-maik alastair, robertson@ed, ac. uk) 2Department o f Geology, Aristotle University, GR-54142, Thessaloniki, Greece Abstract: The Eastern Mediterranean is one of the key regions for the understanding of fundamental tectonic processes, including continental rifting, passive margins, ophiolites, subduction, accretion, collision and post-collisional exhumation. It is also ideal for understanding the interaction of tectonic, sedimentary, igneous and metamorphic processes through time that eventually lead to the development of an orogenic belt. Below, we will outline some milestones in the development of tectonic-related research in the Eastern Mediterranean region. We will mention how studies of the Eastern Mediterranean contribute to our understanding of fundamental tectonic processes and indicate how papers in this volume contribute to this aim. Current and emerging research themes will be highlighted. We will also outline the main alternative tectonic reconstructions of the region (see Fig. 1), and mention which of these the different contributors favour. Tethyan nomenclature remains controversial and we will suggest an appropriate informal terminology for the various oceanic basins that existed. An entr6e to some of the key literature sources is also provided. Citations here are mainly to edited volumes, which provide access to this large subject area. Many of the papers in this book integrate and synthesize large amounts of geological information for extended periods of geological time. The papers are ordered in a general time sequence with a view to linking those that consider comparable tectonic setting and processes. The locations of the areas are shown in Figures 2 and 3. Figure 2 also shows the main sutures, and Figure 3 illustrates the main neotectonic elements of the region.

Development of research

1970s to mid-1980s The Eastern Mediterranean region figured in preplate tectonic geosynclinal models (e.g. Aubouin et al. 1970). The plate tectonic framework for the modern tectonic setting was established in seminal papers by McKenzie (1972, 1978) and Le Pichon & Angelier (1979). Modern interpretations of this region in terms of plate tectonics effectively began with the pioneering work of Smith (1971) and of Dewey et al. (1973). During the 1970s field-based information was amassed by the French-led Tethys project, culminating in sets of palaeogeographical maps that evolved through several editions (Dercourt et aL 1986, 1993, 2000). The Tethys group initially envisaged the existence of a Mesozoic-Early Tertiary Tethyan ocean dating from Triassic time, bordered by the African and Eurasian continents. They interpreted the Mesozoic ophiolites as forming at mid-ocean ridges. The Tethyan ocean was subducted northwards beneath Eurasia in this interpretation. Others developed alternative tectonic models for parts of the region. By the early 1980s the

existence of an Early Mesozoic oceanic basin in the easternmost Mediterranean region had been proposed (Robertson & Woodcock 1979; Garfunkel & Derin 1984). In western Greece a belt of ophiolites was interpreted as evidence for the existence of a Mesozoic ocean basin, separate from a second belt of ophiolites further east (Smith et al. 1975; see also Smith 1993). Ophiolites and deep-sea sediments were distributed throughout m a n y areas of Turkey, suggesting that several Mesozoic oceanic basins, rather than one, might have existed there. In 1981 ~eng6r & Yllmaz published a seminal plate tectonic synthesis of Turkey, which depicted the interaction of microcontinents and small ocean basins. In addition, based initially on information from the Eastern Pontides (northern Turkey), Seng6r et al. (1980) introduced a tectonic model for Late Palaeozoic-Early Mesozoic time, later applied to Eurasia as a whole ($eng6r 1984). This envisaged southward subduction of a Late Palaeozoic-Early Mesozoic ocean (PalaeoTethys) and the related opening of several marginal basins along the northern margin of Gondwana. Closure of this ocean culminated in continental collision by the latest Triassic-Early

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 1-9. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

2

A.H.F. ROBERTSON & D. MOUNTRAKIS

~

~

.

.

9

~n o o o p e 2

" "

~Palaeotethys

"

Neotethys N. Neotethys

Drama

9

.

,

a, .Seng6r 1984 r" M e l i a t a " ~

.~ ",~, '

"

o-

,,b, Dercourt et al. 199a

--' /

~

~

z -.

",Sor

9 9 . Serbo-Macr

~ '~

Late Palaeozoic Tethys

" erbo.._.. K a r a k a y a ,Cal

l'~rcl'~r ~ ~

~

-

9; ~ 0

c, Stampfli & Borel 2002

e~,~

-

Triassic Tethys

Kirsehir

~tra-Tauride Teth s

d, Robertson et al. 2004

"~

MID - LATE TRIASSIC Fig. 1. Alternative tectonic models of Tethys in the Eastern Mediterranean region. (See text for explanation.) Jurassic time, and was followed by opening of a new, Jurassic ocean basin (Northern Neotethys). The nomenclature of Tethyan oceanic basins is, however, rather confusing and an attempt to clarify this aspect is made at the end of this introduction. At an international conference on the Eastern Mediterranean region, held in Edinburgh in October 1982, several different tectonic interpretations and much supporting evidence were presented and later published as Special Publication of the Geological Society of London No. 17, Tectonic Evolution of the Eastern Mediterranean (Dixon & Robertson 1984). A plate tectonic model of the Eastern Mediterranean by Robertson & Dixon (1984) in that book combined the concept of the area as a mosaic of microcontinents and ocean basins with the hypothesis that many of the Mesozoic ophiolites formed above intra-oceanic subduction zones. Two key elements: northward subduction and suprasubduction ophiolite genesis, were retained in many of the more recent reconstructions.

Mid-1980s on wards

A critical mass of information had by then become available for many areas and geological units including the Mesozoic-Early Cenozoic land geology of Greece, former Yugoslavia, Cyprus and parts of Turkey, whereas many other areas and units remained poorly understood. Chief amongst these were the regional metamorphic complexes, which by then could no longer be seen simply as 'basement units', but still remained poorly dated and little understood. Neotectonics (broadly post-Miocene) were known to affect many areas but also remained obscure. In addition, the marine geological setting remained largely unknown, despite the pioneering work of the Deep Sea Drilling Project (e.g. Hs/i et al. 1978). A major advance in recent years has been the testing and confirmation of the early plate tectonic model of McKenzie (1972, 1978) using a combination of field evidence, geophysical modelling (e.g. Jackson & McKenzie 1984) and,

INTRODUCTION

3

"

Former Yugoslavia Moesian platform Black Sea Turkey

%

Karakaya massif

j

solute

~$ s~

ETurkish

Mtns.

CypnJs-\~*r

Crete

.~]~

Syria

.34~ E. Mediterranean

Sea 8

~no

Egypt

.~,

ARABIA

~/

..... Main sutures .....

Fig. 2. Outline map of the main sutures showing the locations of studies mainly concerning the Palaeozoic and Mesozoic tectonic development of the Eastern Mediterranean region. 1, Smith; 2, Himmerkus et al.; 3, Yanev et al.; 4, Romano et al.; 5, Robertson; 6, Karamata; 7, Kazmin & Tikhonova; 8, Gardosh & Druckman; 9, Danelian et al.; 10, Rassios & Moores; 11, Koller et al.; 12, Garfunkel; 13, Rizao~lu et al.; 14, Morris et al.; 15, Sharp & Robertson; 16, Rice et al.; 17, Rirmnel~ et al. Studies 1 and 7 cover whole area. more recently, direct measurements by the global positioning system method (Reilinger et al. 1997). Large datasets have continued to be amassed that can now be used to test and develop tectonic hypotheses. Several international research programmes have aided this effort. Amongst these was the International Geological Correlation Project (IGCP) No. 276, 'Terrane Maps and Terrane Descriptions' (Papanikolaou 1996-1997); this analysed the region as tectonostratigraphic terranes supported by regional correlation maps and terrane descriptions. Another was IGCP No. 369, 'Comparative evolution of Peri-Tethyan Rift Basins', which focused on the rift basins of the Tethyan periphery (Ziegler et al. 2001). Recently, EUROPROBE GeoRift 3, 'Intraplate Tectonics and Basin Dynamics' (Stephenson 2004) has provided much information on the SE European craton and its margins. Palaeomagnetic studies have played an important role in regional interpretation (e.g. Morris & Tarling

1996). Ophiolites in the region have received particular attention (e.g. Hoeck et al. 2002). Several geological compilations have recently focused on parts of the Eastern Mediterranean region, notably Turkey (Bozkurt et al. 2000) and Greece (Pe-Piper & Piper 2002). There has also been an increasing focus on the neotectonic development of the region (Robertson & Comas 1998; Durand et al. 1999; Taymaz et al. 2004) Neotectonics refers to the strain resulting from a stress regime that essentially remains active at the present time, broadly from Miocene to Recent in the Eastern Mediterranean region. Conversely, palaeotectonics refers to stress regimes that are no longer active. One of the most important discoveries, especially from the study of the South Aegean region, is that many tectonic contacts that were traditionally interpreted as thrusts related to the emplacement of nappes are instead extensional faults (i.e. extensional detachments) related to Tethyan exhumation (e.g. Durand et al. 1999).

4

A.H.F. ROBERTSON & D. MOUNTRAKIS

Black Sea ~'~"

t

\ ~ 23 & 287 Bulgaria

Turkey

~

Aeg~ean"

e. ~

~

~ '~.4L.~

~ge accretionary

'

-

COmplex

Arabia

/

active

Majorcompression ~ zones ~.~ Majorstrike-slip 34~ ="='~zones ..~ Majorextensional ZOtl~S

,'-

E.Mediterranean

/~5l/~

100Km!

N. Africa I

"~

O ~.

/

7

active~

.

~

"'

///d -34~

~

1 I

...1.

30 ~ I

Fig. 3. Outline map of the main neotectonic features showing the locations of studies mainly concerning the Cenozoic-Recent tectonic development of the Eastern Mediterranean region. 18, Degnan & Robertson; 19, Piper; 20, Doutsos et al.; 21, Vamvaka et al.; 22, Muceku et al.; 23, Westaway; 24, Al~igek et al.; 25, Boulton et al.; 26, Pavlides et al.; 27, Mountrakis et al.; 28, Tranos et al.; 29, Papzachos et al. covers whole area.

Other advances include the results of ocean drilling in the Eastern Mediterranean Sea that allowed a closer integration of tectonic processes operating on land and under the sea (Robertson et al. 1998). Also, sedimentary basin evolution is increasingly seen as a response to kinematically linked tectonic processes, as shown by studies of the Mediterranean as a whole (e.g. Robertson & Grasso 1995), and most recently in southern Turkey (Kelling et al. 2005). Sets of papers resulting from the Third and Fourth International Meetings on Eastern Mediterranean Geology, held in Nicosia (Cyprus) and Isparta (Turkey) were edited by Panayides et al. (2000) and by Akmcl et al. (2003), respectively. The integrated set of papers on the classic Isparta Angle of SW Turkey (Akmcl et al. 2003) exemplifies the complex geology of interacting microplates. An outcome of these studies is that it is now possible to trace the well-known modern plate

tectonic setting backwards through time (i.e. neotectonics) and link this with the previous tectonic evolution of the region (i.e. palaeotectonics) to provide a more complete picture of the evolution of the orogen.

Tectonic processes The Eastern Mediterranean region is an ideal test-bed for the development of hypotheses for fundamental tectonic processes. Such work is also important to the well-being of those living in this highly populated region ( > 100 million people) as it can contribute to an evaluation of the resource potential, notably hydrocarbons, and is critical to the assessment of hazards, most obviously earthquakes. Process-oriented tectonic studies are outlined below. Many of these are developed in this book, as indicated.

INTRODUCTION Processes of rifting and continental break-up, especially of Triassic age, are well documented by volcanic and sedimentary rocks throughout the region. Processes of rifting are discussed here by Gardosh & Druckman for the easternmost Mediterranean area, and by Robertson and Degnan & Robertson for the Aegean area. The region includes some of the best examples of emplaced deep-water passive margin successions (e.g. Pindos zone, Greece), as summarized for the Peloponnese in southern Greece by Piper and Degnan & Robertson. The Mesozoic ophiolites include the classic Troodos ophiolite (Cyprus), the Pindos and Vourinos ophiolites (Greece) and the Mirdita ophiolite (Albania) that contribute much to our understanding of oceanic lithosphere genesis and emplacement. New results and interpretations are presented by Rassios & Moores and Koller et al. concerning Mesozoic ophiolites in Greece and Albania. Evidence for previously little known Late Cretaceous ophiolites in northern Turkey is given by Rice et al. The region includes some excellent examples of accretionary prisms (e.g. Cretaceous Ankara m61ange) related to subduction. For example, biostratigraphical studies of radiolarian cherts in m61ange blocks can shed light on Tethyan evolution, as discussed here for SW Turkey by Danelian et al. High-pressure-low-temperature (HP/LT) metamorphic rocks are widely distributed (e.g. NW Turkey) and are critical to our understanding of deep-seated subduction processes. Here, important new evidence of carpholite-bearing HP/LT rocks related to subduction and exhumation in SW Turkey is presented by Rimmel6 et al. The region provides good examples of arc volcanism and the formation of back-arc basins in active margin settings, as discussed by Rizao~lu et al. for eastern Turkey and by Rice et al. for northern Turkey. In general, evidence of subduction-related magmatism has increasingly been found in different areas, especially in eastern Turkey. As would be expected, the region is ideal for the study of collisional processes, allowing new models for continental collision to be developed, as presented here by Doutsos et al. The region is one of the key areas for studies of post-collisional processes, notably extensional detachment faulting and crustal exhumation (e.g. southern Aegean). Here, new fission-track thermochronology results are presented by Muceku et aL, which elucidate the exhumation of the Neotethyan suture in Albania. Syn- to post-collisional sedimentary basins are well exposed and yield important insights into

5

a range of tectonic processes, as exemplified by the Neogene Hatay Graben, southern Turkey (Boulton et aL), the Cameli Basin, SW Turkey (Al~i~ek et aL) and the Mesohellenic Trough, Greece and Albania (Vnmvaka et aL). The potential of the Eastern Mediterranean region increases still further when the wider regional setting is taken into account, including the Western Mediterranean and Central-North Atlantic regions, as discussed by Smith, and also Eurasia to the north and NE of the Eastern Mediterranean region, as summarized by Kazmin & Tikhonova. Traditional plate tectonic analysis is effective where large oceans and continental areas existed (e.g. related to subduction of Palaeotethys; see Kazmin & Tikhonova) but becomes difficult to apply in regions where numerous microplates existed. Alternative reconstructions, using terrane analysis, are presented here for parts of the Balkan Peninsula and adjacent areas (see Yanev et al. and Karamata). Particular difficulties are encountered with regions dominated by microplate interaction like the Eastern Mediterranean. For example, rift processes may be regionally variable and affect several microcontinental blocks simultaneously (see Robertson). Rifts may be constructed on several pre-existing orogens (Hercynian, or Pan-African) and this may affect the geometry of rifting or the geochemistry of rift-related igneous rocks (see Romano et al. and also Robertson). The nature and timing of sea-floor spreading may be difficult to determine where ophiolites are distributed through several adjacent sutures (see Garfunkel and Smith). Collision affecting several microcontinents is necessarily complex and diachronous (see Sharp & Robertson). Regional-scale crustal rotations may play an important role, and these, in favourable settings, may be restored and interpreted using palaeomagnetic techniques, as documented by Morris et al. for ophiolites and related units in Cyprus, southern Turkey and northern Syria. Strike-slip may also be important but is often difficult to recognize and restore in deformed regions (see Karamata). In addition, many parts of the Eastern Mediterranean region are seismically active and subject to earthquakes. Such hazards to the large populations living in this region can be investigated by studies of active faults and modern-day seismicity, as discussed by Mountrakis et al. and Tranos et al. The resulting data can be used to develop predictive models of earthquakes, as presented here by Papazachos et ai. Tectonic processes operate successively or interact through time to produce complex

6

A.H.F. ROBERTSON & D. MOUNTRAKIS

tectonic assemblages that are ultimately unique. However, recognizable patterns recur in different suture zones at different times. One important example is flexural foredeep development related to ophiolite emplacement, as for both the MidLate Jurassic of Greece and the Late Cretaceous of Turkey. Another is post-collisional extensional basin development related to exhumation, both for the Late Carboniferous-Permian of the Balkans and for the Neogene of the Eastern Mediterranean region as a whole. A number of papers in this book exemplify the successive activity and interaction of different tectonic processes through time, an excellent example being the tectonic development of the Vardar zone in northern Greece (see Sharp & Robertson).

Tectonic settings Most workers now accept that the Eastern Mediterranean region hosted a wide ocean separating Africa and Eurasia (Palaeotethys), at least by Late Palaeozoic time (see Kazmin & Tikhonova), but there is little agreement as to how, when and where this ocean formed and was ultimately consumed. There is an emerging consensus that some of the large allochthonous units ('terranes') of Late Palaeozoic, or earlier, age originated along the northern margin of Gondwana and were later accreted to Eurasia at different times (i.e. Carboniferous; Late Triassic; Late Cretaceous; see Yanev et ai. and Karamata). However, it is also suggested that some exotic terranes including igneous and metamorphic rocks formed and remained along the north margin of Gondwana during their entire history (see Romano et al. and Karamata). A further uncertainty is the timing, location and amount of lateral displacement related to strike-slip faulting (i.e. terrane displacement), especially during Late Carboniferous-Late Triassic time. The record of basement units of Pan-African age is sparse (e.g. western and N W Turkey) and thus their tectonic settings remain obscure. Hercynian-aged terranes are more widespread (e.g. Bulgaria; northern Turkey; southern Aegean) but were commonly deformed, metamorphosed and dispersed during later orogenesis, such that their tectonic settings have remained unclear. New evidence of Palaeozoic tectonic settings, supported by new radiometric dating, is given here for northern Greece by Himmerkus et aL, and for southern Greece by

Romano et aL Different tectonic interpretations also exist for the Mesozoic-Early Cenozoic tectonic evolution. Deep-water basins rifted along the North

Africa-Levant margin during Late PalaeozoicEarly Mesozoic time, but there is no agreement as to whether these represent intra-continental rifts (i.e. Red Sea-type rifts), or back-arc basins related to subduction. Here, Gardosh & Druekman argue in favour of an origin of the Levant Basin in the easternmost Mediterranean Sea as an early Mesozoic rift basin unrelated to subduction. The main Mesozoic ophiolites are, nowadays, widely viewed as forming above intra-subduction zones, as explained in papers by Smith, Rassios & Moores and Garfunkel. However, some geologists continue to believe that most ophiolites formed at mid-ocean ridges, unrelated to subduction. Assuming most of the ophiolites did indeed form in subduction-related settings, questions still exist concerning the timing of spreading, subduction initiation, and the number and location of subduction zones involved. In addition, because continental collision is progressive and diachronous it is difficult to determine when, where and how collision has taken place. For example, in northwestern Greece the initial closure of the Vardar ocean is seen by some as Late Jurassic in age but by others as latest CretaceousEarly Cenozoic. In central Turkey, within the Izmir-Ankara-Erzincan zone, collision is seen as either latest Cretaceous or Eocene (see Rice et ai.). In SE Turkey suturing of a southern Neotethyan ocean is variously thought to be latest Cretaceous, Late Eocene or Mid-Miocene (see Rizao~lu et al. and Boulton et aL ). Miocene-Recent tectonic settings are better understood as they can be directly related to the well-established modern plate tectonic setting of the region. There is a consensus that a Tethyan ocean in the south Aegean region (e.g. Ionian basin) was subducted northwards accompanied by back-arc extension, extending across the Aegean into western Turkey (e.g. Le Pichon & Angelier 1979). Evidence of related extension extending as far as northern Greece and Bulgaria is presented here by Mountrakis et aL, Tranos et al. and Westaway. However, questions remain, including when and where subduction-related extension began (e.g. in western Turkey) and the extent to which Mesozoic Tethyan oceanic crust remains in the easternmost Mediterranean Sea, e.g. within the Herodotus Basin SW of Cyprus. Was the driving force of neotectonics in the south Aegean region southward migration (i.e. roll-back) of the Aegean subduction zone or westward tectonic escape of Anatolia, or a combination of both? For SW Turkey this question is addressed by Al~;i~ek et ai. In the easternmost Mediterranean region, around Cyprus, did subduction continue until the present time along the

INTRODUCTION 'Cyprus arc', or end with collision during the Miocene or even earlier in this region (see Boulton et aL )?

Tectonic reconstructions The Eastern Mediterranean is a favourite region for tectonic reconstruction, especially as the bounding North African (Gondwana) and Eurasian (Laurasia) continents are clearly defined. Several of the papers in this book present reconstructions for certain regions or time intervals (e.g. Smith), and the alternatives are critically discussed (see Robertson). Palaeomagnetic studies show how the continental separation between Gondwana and Eurasia has evolved through time (e.g. see Morris & Tarling 1996). However, there are drastically different views of where the intervening pieces of the jigsaw puzzle should be placed and how they moved through time (see Robertson et al. 1996 for alternatives). There is still no consensus as to the most appropriate regional tectonic reconstruction. Controversial aspects include the timing and location of continental rifting and break-up to form one or more oceanic basins, the direction and timing of subduction, and the mode and timing of continental collision. The most currently discussed tectonic models are outlined in Figure 1. In one class of reconstruction ($eng6r 1984) a 'Palaeo-Tethyan' ocean was subducted southwards associated with the opening of 'Southern Neotethyan' marginal basins to the south. In most other reconstructions subduction was instead northwards (e.g. Garfunkel 1998, 2004). The reconstructions of Robertson et al. (1991, 1996, 2004), Dercourt et al. (1993, 1998, 2000) and Ricou (1996) envisage the rifting of several microcontinents from Gondwana. These fragments drifted northwards until they were accreted to Eurasia at various times. Even within this class of model (i.e. involving northward subduction), individual reconstructions vary considerably; for example, in the inferred location and age of oceanic crust in the Eastern Mediterranean region and whether the ophiolites mainly formed at mid-ocean ridges or above a subduction zone. In a third, different type of model, a 'Palaeotethyan ocean' was located in a more southerly position and completely closed by Early Jurassic time within the South Aegean region, whereas back-arc basins opened further north and did not then close until latest Cretaceous-Early Cenozoic time (Stampfli et al. 2001; Stampfli & Bore12002). The alternatives come sharply into focus for the Late Palaeozoic-Early Mesozoic evolution

7

of the South Aegean region, which is interpreted differently according to whether the Palaeotethyan suture is located within this area or much further north, close to the Eurasian margin (see Himmerkus et aL). One option is that subduction was dominantly northwards beneath the Eurasian margin (see Robertson), but that subduction also took place southwards beneath Gondwana at least during Late DevonianCarboniferous related to the Hercynian orogeny (see Romano et aL). Robertson presents evidence from the South Mediterranean region (Sicily, Crete, Peloponnese and Evia) that supports a model of rifting of microcontinents from Gondwana during the Triassic, followed by their northward drift during a time when northward subduction was active beneath Eurasia. However, southward subduction during the preceding Hercynian orogeny is not precluded.

Tethyan nomenclature At present, Tethyan nomenclature is confusing mainly because different researchers apply the same names (e.g. Neotethys) to entirely different oceanic basins in different areas. Here, we advocate a relatively loose, non-exclusive terminology for the various Tethyan ocean basins in the Eastern Mediterranean region. We take Palaeotethys to refer to oceanic crust of mainly Late Palaeozoic-Early Mesozoic age that was formed, subducted or emplaced regardless of its geographical location. We use the term Neotethys for oceanic basins that rifted and then opened during Early Mesozoic time, again regardless of their location or mode of formation. In principle, Neotethyan rift basins could have formed in several different settings, including cratonic areas or pre-existing orogens (either within their interiors or along their margins). Neotethys may also include oceanic lithosphere that was formed within a pre-existing (i.e. Palaeotethyan) ocean; for example, as a subduction-related marginal basin or a strike-slip controlled basin. Neotethys was clearly multi-stranded and in principle coexisted with Palaeotethys, in a manner similar to the multiple relatively young ocean basins that formed in the SW Pacific region while older oceans in the region coexisted. In Greece, Neotethys includes two belts of ophiolitic and related rocks (Pindos and Vardar), whereas Neotethys in Turkey includes the Southern Neotethys, south of the TaurideAnatolide platform and the Northern Neotethys to the north of this continental unit. Several other smaller Neotethyan oceanic strands have been proposed (e.g. Inner Tauride ocean; intraPontide ocean).

8

A . H . F . ROBERTSON & D. MOUNTRAKIS

This rather informal, non-prescriptive n o m e n clature that we advocate contrasts with some other approaches in which Tethyan oceans are n a m e d as specific basins in specific geographical regions that are indivisible f r o m particular tectonic reconstructions. This is unsatisfactory, as in different reconstructions the same names (e.g. Palaeotethys; Neotethys) are applied to entirely different oceanic basins in different areas by different workers. A genetic terminology needs to be avoided in principle, as it leads to circular reasoning and inhibits the testing o f alternatives. It seems increasingly likely there was, in any case, no sharp distinction between Palaeotethys and Neotethys, but rather one oceanic system existed and continued to develop t h r o u g h o u t P a l a e o z o i c - R e c e n t time, akin to the tectonic d e v e l o p m e n t of the SW Pacific region. We thank S. Pavlides and colleagues for convening the Fifth International Symposium on Eastern Mediterranean Geology in Thessaloniki, 14-20th April 2004. We also thank S. Pavlides for assistance with preparing this volume. J. Dixon is thanked for his review of the manuscript. J. Turner kindly advised on the structuring of this introductory chapter. References

AKINCI, O., ROBERTSON, A. H. F., POISSON, A. & BOZKURT, E. (eds) 2003. Special issue on the Isparta Angle, SW Turkey. Geological Journal, 38, 195-234. AUBOUIN,J., BONNEAU,M., CELET, P. et al. 1970. Contribution h la g6ologie des H611enides: le Gavrovo, le Pinde et la Zone Ophiolitique Subp61agonian. Annales de la Societk Gkologique du Nor& 90, 277-306. BOZKURT, E., WINCHESTER, J. A. & PIPER, J. D. (eds) 2000. Tectonics" and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173. DERCOURT, J., ZONENSHAIN,t . 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. DERCOURT, J. RICOU, L. E. & VRIELYNCK, B. (eds) 1993. Atlas Peri-Tethys Palaeogeographical Maps. CCGM/CGMW, Paris. DERCOURT, J., GAETANI,M., VRIELYNCK,B. et al. (eds) 2000. Per# Tethys Palaeogeographical Atlas. Gauthier-Villars, Paris. DEWEY, J. F., PITMAN, W. C., III, RYAN, W. B. F. & BONNIN, J. 1973. Plate tectonics and the evolution of the Alpine System. Geological Society of America Bulletin, 84, 3137-3180. DIXON, J. E., ROBERTSON, A. H. F. (eds) 1984. The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17.

DURAND, D., JOLIVET, L., HORVATH, F. & SI~RANNE, M. 1999. The Mediterranean Basins: Tertiary Extension within the Alpine Orogen. Geological Society, London, Special Publications, 156. GAREUNKEL, Z. 1998. Constraints on the origin and history of the Eastern Mediterranean basin. Tectonophysics, 298, 5-37. GARFUNKEL,Z. 2004. Origin of the Eastern Mediterranean basin: a re-evaluation. Tectonophysics, 391, 11-34. GARFUNKEL, Z. & DERIN, B. 1984. Permian-early Mesozoic tectonism and continental margin formation and its implications for the history of the Eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 177-186. HOECK, V., TOMEK, C., ROBERTSON, A. H. F. & KOLLER, F. (eds) 2002. Eastern Mediterranean ophiolites: magmatic processes and geodynamic implications. Lithos, Special Issue 65. Hs~3, K. J., MONTADERT, L., et al. (eds) 1978. Initial Reports of the Deep Sea Drilling Project, 32A. US Government Printing Office, Washington, DC. JACKSON, J. & MCkENZIE, D. P. 1984. Active tectonics of the Alpine-Himalayan belt between western Turkey and Pakistan. Geophysical Journal of the Royal Astronomical Society, 77, 185-264. KELLING, G., ROBERTSON, A. H. F. & VAN BUCHEM, F. H. P. (eds) 2005. Cenozoic Sedimentary Basins of South Central Turkey. Sedimentary Geology, Special Issue, 173. LE PICHON, X & ANGELIER, J. 1979. The Hellenic arc and trench system: a key to the neotectonic evolution of the Eastern Mediterranean area. Tectonophysics, 60, 1-42. MCKENZIE, D. P. 1972. Active tectonics of the Mediterranean region. Geophysical Journal of the Royal Astronomical Society, 30, 109-185. MCKENZlE, D. P. 1978. Active tectonics of the AlpineHimalayan belt: the Aegean Sea and surrounding regions. Geophysical Journal of the Royal Astronomical Society, 55, 217-354. MORRIS, A. & TARLING, D. H. (eds) 1996. Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105. PANAYIDES, I., XENOPHONTOS,C. & MALPAS, J. (eds) 2000. Proceedings of the Third International Conference on the Geology of The Eastern Mediterranean. Geological Survey Department, Nicosia. PAPANIKOLAOU, D. J. (ed.) 1996-1997. International Geological Correlation Project 276. Terrane Maps and Terrane Descriptions. Annales G6ologiques des Pays Hell6niques. PE-PIPER, G. & PIPER, D. W. J. 2002. The Igneous Rocks of Greece. The Anatomy of an Orogen. Beitrage zur regionalen Geologie der Erde, 30. REIL1NGER, R. E., MCCLUSKY, S. C., ORAL, M. B., KING, R. W. & TOKSOZ, M. N. 1997. Global Positioning System measurements in the ArabiaAfrica-Eurasia plate collision zone. Journal of Geophysical Research, 102, 9983-9999.

INTRODUCTION RICOU, L.-E. 1996. The plate tectonic history of the past Tethys ocean. In: NAIRN, A. E. M., RICOU, L.-E., VRIELYNCK, B. & DERCOURT, J. (eds) The Ocean Basins and Margins, 8, The Tethys Ocean. Plenum, New York, 3-62. ROBERTSON, A. H. F. & COMAS, M. C. (eds) 1998. Collision-related Processes in the Mediterranean Region. Teetonophysics, Special Issue, 298. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the Eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean, Geological Society, London, Special Publications, 17, 1-74. ROBERTSON, A. H. F. & GRASSO, M. (eds) 1995. Late Tertiary Mediterranean tectonics and palaeoenvironments. Terra Nova, 7, 254-264. ROBERTSON, A. H. F. & WOODCOCK, N. H. 1979. The Mamonia Complex, SW Cyprus: the evolution and emplacement of a Mesozoic continental margin. Geological Society of America Bulletin, 90, 651-665. 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, 289-344. ROBERTSON, A. H. F., DIXON, J. E., BROWN, S., et al. 1996. Alternative tectonic models for the Late Palaeozoic-Early Tertiary development of Tethys in the Eastern Mediterranean region. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBERTSON, A. H. F., EMEIS, K. C., RICHTER, K.-C. & CAMERLENGHI, A. (eds) 1998. Proceeding of the Ocean Drilling Program, Scientific Results, 160. Ocean Drilling Program, College Station, TX, 723-782. ROBERTSON, A. H. F., USTAOMER, T., PICKETT, E. A., COLLINS, A., ANDREW, T. & DIXON, J. E. 2004. Testing models of Late Palaeozoic-early Mesozoic orogeny: support for an evolving one-Tethys model. Journal of the Geological Society, London, 161, 501-511. SENGOR, A. M. C. 1984. The Cimmeride Orogenic System and the Tectonics of Eurasia. Geological Society of America, Special Papers, 195.

9

~ENGOR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 81-241. SENGOR, A. M. C., YILMAZ,Y. & KETIN, I. 1980. Remnants of a pre-Late Jurassic ocean in northern Turkey: fragments of Permian-Triassic PaleoTethys. Geological Society of America Bulletin, 91, 599-609. SMITH, A. G. 1971. Alpine deformation and the alpine areas of Tethys, Mediterranean and Atlantic. Geological Society of America Bulletin, 82, 2039-2070. SMITH, A. G. 1993. Tectonic significance of the Hellenic-Dinaric ophiolites. In: PRICHARD, H. M., ALABASTER, T., HARRIS, N. B. W. & NANCE, D. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 213-243. SMITH, A. G., HYNES, A. J., MENZIES, M., NISBET, E. G., PRICE, I., WELLAND,M. J. ~; FERRII~RE,J. 1975. The stratigraphy of the Othris Mountains, Eastern Central Greece: a deformed Mesozoic continental margin sequence. Eclogae Geologicae Helvetiae, 68, 463-481. STAMPELI, G. M. & BOREL, G. D. 2002. A plate tectonic model for the Palaeozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrones. Earth and Planetary Science Letters, 169, 17-33. STAMPFLI, G., MOSAR, J., FAURI~, P., PILLEVUIT,A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethys realm: the Neotethys East Mediterranean basin connection. In: ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. • CRASQUINSOLEAU, S. (eds) Peri-Tethys Memoir 5. Per# Tethyan Rift~Wrench Basins and Passive Margins. M6moires du Mus6um National D'Histoire Naturelle, 51-108. STEPHENSON, R. A. (ed.) 2004. EUROPROBE, GeoRift 3. Intraplate Tectonics and Basin Dynamics. Tectonophysics, Special Issue, 381. TAYMAZ, T., WESTAWAY, R. & REILINGER, R. (eds) 2004. Active Faulting and Crustal Deformation in the Eastern Mediterranean Region. Tectonophysics, Special Issue, 391. ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. 8r CRASQUIN-SOLEAU, S. (eds) 2001. Peri-Tethys Memoir 5. Peri-Tethyan RiftIWrench Basins and Passive Margins. M6moires du Mus6um National D'Histoire Naturelle.

Late Proterozoic and Silurian basement units within the Serbo-Macedonian Massif, northern Greece: the significance of terrane accretion in the Hellenides F. H I M M E R K U S

1'2, T. R E I S C H M A N N

1 & D. K O S T O P O U L O S

3

~Johannes Gutenberg Universitgit Mainz, Department of Geology, Graduiertenkolleg: Stoffbestand yon Kruste und Mantel, Max-Planck-Institut fiir Chemie, Abteilung Geochemie, Mainz, Germany (e-mail: himmerku@mail, uni-mainz, de) 2Present address: Am Steinhiiusle 11, 76228 Karlsruhe, Germany 3Faculty of Geology and GeoEnvironment, Department of Mineralogy and Petrology, National and Kapodistrian University of Athens, Panepistimioupoli Zographou, Athens 15784, Greece Abstract: The Serbo-Macedonian Massif (SMM) is a large, elongate basement complex in the Internal Hellenides, which stretches from Serbia to the Chalkidiki Peninsula in northern Greece. As a result of similarities in lithology and structural grain, the SMM has long been considered part of the adjacent Rhodope Massif. Recent work, however, based on precise geochronological and geochemical data, has revealed that the SMM is not a homogeneous crustal entity but made up of several crustal units, only one of which is related to the Rhodope Massif. One of these units, the Pirgadikia Unit, occurs as a tectonic sliver in a m61ange zone bordering the western margin of the SMM that separates it from the adjacent Vardar Zone. The Pirgadikia Unit consists of leucocratic mylonitic para- and orthogneisses. According to trace-element and Sr-isotope data, the orthogneisses originated in a magrnatic arc setting. Dating of this unit by the Pb-Pb single-zircon evaporation method yielded Pan-African ages of 555.8 + 2.6 Ma on a paragneiss collected near the village of Taxiarchis, and two ages of 570.0+7.0 Ma and 587.6+3.4 Ma on orthogneisses from the quayside at Pirgadikia village. The rocks enveloping this Late Precambrian basement complex are gneisses of the Vertiskos Unit. This unit, which is regarded as a distinct terrane, occupies the northwestern part of the Greek SMM and consists of Silurian orthogneisses with a magmatic arc signature and subordinate metasediments. Orthogneisses of the Vertiskos Unit adjacent to the mylonites of the Pirgadikia Unit gave Pb-Pb ages of between 428.2 + 1.2 Ma and 433.0 + 2.1 Ma. One of these samples was additionally dated by the conventional U-Pb method. This sample has three concordant zircon grains confirming a Silurian intrusion age and two inherited cores pointing to an older basement into which precursor rocks to the Silurian gneisses were intruded. The upper intercept of a Concordia plot yielded an age of c. 2.5 Ga, which is a common age in the cratons of Gondwana. The Pan-African age of the Pirgadikia Unit and the inherited ages of the Vertiskos Unit support the notion that these units are terranes derived from Gondwana. They were finally accreted to the Hellenic orogen during the closure of one of the branches of the Tethys Ocean. The presence of exotic terranes in the Internal Hellenides contradicts the hypothesis that this part of the Hellenides formed a stable hinterland during the Alpine phase and thus the Hellenides can be considered an accretionary orogen. The S e r b o - M a c e d o n i a n Massif ( S M M ) is a crystalline b a s e m e n t inlier in the central part of the Internal Hellenides of n o r t h e r n Greece. It is bord e r e d to the west by the V a r d a r Z o n e a n d to the east by the R h o d o p e Massif. The t e r m SerboM a c e d o n i a n Massif was coined by Dimitrijevi6 (e.g. Dimitrijevi6 1977, 1997), and was treated as a separate unit in the classical subdivision of the geology o f n o r t h e r n Greece (e.g. J a c o b s h a g e n 1986). T h e zones f r o m west to east, or f r o m external to internal are as follows (see Fig. 1, inset). The

I o n i a n Z o n e o f the External Hellenides is largely built up of Mesozoic c a r b o n a t e s a n d clastic sediments. This unit is followed to the east by the ophiolitic Pindos Z o n e (e.g. R o b e r t s o n 2002, a n d references therein), w h i c h yielded Late Jurassic e m p l a c e m e n t ages (Liati 2004). East o f the P i n d o s Z o n e the I n t e r n a l Hellenides, are defined as units c o n t a i n i n g p r e d o m i n a n t l y b a s e m e n t rocks (e.g. J a c o b s h a g e n 1986, 1994). T h e w e s t e r n m o s t unit of the I n t e r n a l Hellenides is the P e l a g o n i a n Zone, w h i c h merges to the south with the Attico-Cycladic M a s s i f (Diirr et al. 1978).

From:ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the EasternMediterranean Region. Geological Society, London, Special Publications, 260, 35-50. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

36

F. HIMMERKUS E T AL.

Fig. 1. Simplified geological map of the Serbo-Macedonian Massif from the Chalkidiki Peninsula in the south to the Kerkini Mountains in the north (modified after Kockel & Mollat 1977). The tectonic position in the Hellenic orogen is shown in the inset. The SMM and the Rhodope Massif form the metamorphic hinterland of the Hellenides. Together with the Pelagonian Zone they form the Internal Hellenides, which are mainly constructed of granites and gneisses. The Vardar and Pindos Zones are remnants of oceanic basins and the external Hellenides mainly consist of Mesozoic carbonates. The study area is shown as a box (see Fig. 2). Both basement units are mainly constructed of Permo-Carboniferous arc-related gneisses (Engel & Reischmann 1998; Anders et al. 2003) and are separated by the ophiolites of the Vardar Zone from the metamorphic hinterland of the orogen (e.g. Mercier et al. 1975; Kockel & Mollat 1977; Mussalam & Jung 1986). The ophiolitic Vardar Zone is interpreted as a major suture and the oceanic material is Late Jurassic in age (e.g. Stampfli et al. 2004). The Serbo-Macedonian Massif and the Rhodope Massif form the most internal zones of the orogen. The Rhodope Massif is composed of two major nappes (e.g. Burg et al. 1996; Barr et al. 1999). The basement rocks there show two pulses of magmatism, Permo-Carboniferous and Late Jurassic (Turpaud & Reischmann 2003).

The Serbo-Macedonian Massif is also subdivided into two units on the basis of lithological characteristics (Kockel et al. 1971). These two units are the Vertiskos Unit in the NW and the Kerdillion Unit in the east (see Fig. 1). The metamorphism of the SMM reaches upper amphibolite facies conditions (Kilias et al. 1999) and in its eastern part, the Kerdillion Unit, the rocks are generally migmatic-banded biotite gneisses (Kockel & Mollat 1977). Kinematic analysis and different generations of porphyroblasts indicate that the SMM not only experienced polyphase deformation but also at least two metamorphic events. Structural and lithological similarities with the Rhodope Massif led to a correlation of the SMM with the Rhodope Massif (Burg et al. 1995; Ricou et al. 1998). However, the age and

LATE PROTEROZOIC AND SILURIAN BASEMENT provenance of the SMM are still a matter of dispute, as available geochronological data are scarce and are characterized by a large scatter (e.g. Papadopoulos & Kilias 1985; De Wet et al. 1989; Lips et al. 2000). Most of these data are based on Ar-Ar, K - A r or Rb-Sr measurements on micas, which have a rather low closure temperature for the above isotope systems. Such being the case, these ages should be regarded as cooling ages after the last tectonothermal event, especially as the SMM is a polymetamorphic terrane. Preliminary geochronological data for the southern SMM were produced by Frei (1996; and unpub, data) who obtained a zircon U - P b age of 560 Ma for metarhyolitic rocks collected c. 2 km west of Pirgadikia (Frei, pers. comm.). In the SMM of Bulgaria similar U - P b ages are known from the Osogovo-Lisets dome (Graf et al. 1998; Graf2001). No other geochemical and geochronological data exist for the pre-alpine evolution of this part of the Hellenic orogen. For this reason this study focused on this pre-alpine period. Special emphasis was placed on the primary intrusion ages and the chemical and isotopic signature of basement rocks incorporated into the m61ange east of the Vardar Zone, to constrain the provenance of continental blocks involved in the accretion of the Hellenides. The SMM is not a homogeneous crustal entity, but also is made up of several basement units. Three major units can be distinguished on the basis of rock types. The largest of these is the Vertiskos Unit (Himmerkus et al. 2002, 2003), which occupies the northwestern part of the Greek SMM and consists predominantly of augengneisses (Fig. 1). The southeastern part of the Greek SMM is composed of banded biotite gneisses of the Kerdillion Unit (Kockel & Mollat 1977). The third unit, in the central-eastern part of the Chalkidiki Peninsula, is the Pirgadikia Unit, which is the topic of this study. There are two major shear zones with metasediments, amphibolites and ultramafic rocks, which, in our opinion, represent m61ange zones. One of these m61ange zones stretches along the border between the Vardar Zone and the SMM; it was formerly known as the Circum-Rhodope belt (Kaufmann et al. 1976; Jacobshagen 1986), and included the Hortiatis Unit and the Svoula Schist Formation (Kockel et al. 1971). The term Circum-Rhodope belt has already been rejected by Ricou et al. (1998). The second m61ange zone runs approximately N W - S E between the Vertiskos and the Kerdillion Units (see Fig. 1). This m61ange zone contains the mafic and ultramafic complexes of Thermes, Volvi and Gomati (Dixon & Dimitriadis 1984) and was interpreted by Burg

37

et al. (1995) as a thrust contact between different

tectonic units of the SMM. The two m61ange zones seem to converge north of the Sithonia Peninsula. The Pirgadikia Unit is a small exposure of basement within the m61ange zone between the Vardar Zone and the SMM, at a place where the two crustal-scale shear zones described above meet, north of the Sithonia Peninsula (Fig. 1). The rocks of the Pirgadikia Unit display lithological and structural features that differ strongly from those of all the other units known from the SMM and the adjacent basement complexes of the Pelagonian Zone and the Rhodope Massif. This suggests that the Pirgadikia Unit could be an exotic terrane, and for this reason detailed geochronological, geochemical, isotopic and structural investigations were carried out to define the provenance and structural position of this unit. To aid the reader, results from the adjacent basement rocks of the Vertiskos Unit are also reported in this study in an effort to highlight the distinct features of both units.

Geology of the study area The study area is located in the south-eastern part of the SMM in the Chalkidiki region close to the Sithonia Peninsula (see Fig. 1). In this part of the SMM, basement rocks of the Vertiskos and Pirgadikia Units are incorporated into the m61ange zone of the eastern Vardar Zone as tectonic slivers. The Pirgadikia Unit is a distinct lithological unit of small areal extent around the Pirgadikia village quay and along strike further north near the village of Taxiarchis (Fig. 2). The rocks of the Pirgadikia Unit are mylonitic orthogneisses with a strong shear fabric characterized by grain-size reduction and a prominent lineation (Himmerkus et al. 2004a). By contrast, the orthogneisses of the Vertiskos Unit are biotite augengneisses with feldspar porphyroblasts up to 10 cm in length. The gneisses of the Kerdillion Unit in the eastern SMM are strongly migmatized, fine-grained biotite gneisses, which are generally cut by a variety of leucocratic dykes and pegmatites. There are also important structural differences between the mylonitic orthogneisses of the Pirgadikia Unit and those of the adjacent basement units. Within the entire SMM and in the western Rhodope Massif there is a very uniform top-to-the-SW sense of shear, defined by stretching lineations (Burg et al. 1996). In the Pirgadikia Unit the sense of shear is top-to-the-east, which is not recorded elsewhere in the SMM or in the western Rhodope Massif. The lineation is subhorizontal and plunges very gently to the east.

38

F. HIMMERKUS E T AL.

Fig. 2. Geological map after Kockel & Mollat (1977) including fieldwork of this study and a profile from Metamorphosis to Ierissos across the Pirgadikia Unit that shows the structural position. The profile has the same scale as the map. The rocks of the Pirgadikia Unit are faulted into the m61ange zone and are associated with metasediments and orthogneisses of the Vertiskos Unit. According to our field observations, the Pirgadikia Unit is not a tectonic window as the rocks are more or less monoclinal in this area. In the outcrop of the Pirgadikia Unit around the small village of Taxiarchis the rocks are metaquartzites, characterized by a strong shear fabric. They are in tectonic contact with their country rocks, which are essentially low-grade schists and marbles of the Svoula Schist Formation. The metaquartzites are mylonitic and represent the only occurrence of pure quartzites in the whole area. This rock type also occurs as a tectonic sliver in the sheared metasediments of the Svoula Schist Formation, which are exposed in the m61ange zone at the south western contact of the Arnea granite. The tectonic position of the two exposures of the Pirgadikia Unit is very similar. Both occur within metasediments and marbles in the m~lange zone bordering the Vertiskos Unit to the SW. This basement unit and the metasediments of the Svoula Schist Formation are always in faulted contact with the Pirgadikia

mylonitic orthogneisses. This is also supported by the shear-sense criteria mentioned above. The sense of shear in the Pirgadikia orthogneisses is quite different from that in the metasediments and gneisses that envelop them. The two gneiss units of Pirgadikia and Vertiskos are never in direct contact with each other, but are always flanked by metasediments. The mylonitic deformation and the metamorphic grade of the Taxiarchis metaquartzites show that they cannot be part of the Svoula Schist Formation. Lithologically, the Svoula Schist Formation consists of fine-grained clastic sediments with little sand and very few pure sand layers. These rocks are essentially pelites with minor psammitic horizons. In most places the rocks are strongly deformed and sheared but the metamorphic grade is that of lower greenschist facies. Pyrite cubes are a typical feature and pressure

LATE PROTEROZOIC AND SILURIAN BASEMENT shadows around the cubes indicate formation prior to deformation. The sense of shear is topto-the-SW. Another typical feature is the high carbonate content of the pelites and psammites, which is mostly calcite. Furthermore, lenses of marbles, which can be up to several hundred metres in length, are intercalated with the pelites (Fig. 2). West of Plana village, fine-grained marbles are also interbedded with pelites of the Svoula Schist Formation. Orthogneisses of the Vertiskos Unit are also incorporated into the m61ange zone. They are lithologically very similar to the orthogneisses known from the northern part of the SMM and occur as small slivers within the metasedimentary rocks belonging to the sequence. Figure 2 shows a simplified geological map of the southeastern SMM and a profile across strike, traversing the Pirgadikia locality. In this cross-section the basement rocks of the SW and NE of the Pirgadikia Unit belong to the Vertiskos Unit. The granite is part of the Arnea suite (De Wet 1989), which is associated with the Vertiskos Unit (Himmerkus et al. 2003, 2004b). The profile was drawn to show the structural relations between the different units. The planar fabric is the main foliation; primary layering can be identified only in rare cases in the low-grade rocks of the Svoula Schist Formation. From Pirgadikia eastwards the succession is monoclinal and the rocks are strongly transposed. The m61ange zone contains a large variety of lithological units assembled during accretion of the units. From Pirgadikia to Metangitsi the structure is not as clear because the metasediments of the m61ange zone are intensely folded and sheared. The main feature here is that there is a large body of Vertiskos Unit orthogneisses incorporated into the succession. The orthogneisses are also internally folded indicating at least two phases of deformation. Southwest of Metangitsi the low-grade rocks of the Svoula Schist Formation again dip more or less constantly to the SW. This formation is also strongly sheared and internally folded, but is very poor in marker horizons that could help to identify large-scale structures.

Samples and petrography Because of the strong lithological and structural differences between the Pirgadikia Unit and the remainder of the SMM, the Pirgadikia Unit rocks were subjected to a detailed geochemical and geochronological investigation. For comparison purposes, three typical samples of the Vertiskos Unit, in the direct vicinity of the Pirgadikia Unit were also included in this study. For sample localities the reader is referred to Figure 2.

39

The best outcrops of basement orthogneisses at the Pirgadikia locality can be seen along the road-cut sloping down towards the village and at the shore south of the quay. The rocks are very fine-grained two-mica gneisses. Sample SM 13 is very leucocratic, whereas sample SM 12 contains relatively more biotite. The mineralogy is typically granitic, the rocks being made up of quartz, K-feldspar, plagioclase, biotite and white mica. Additionally, there is a minor amount of garnet in SM 12. SM 14 is a dyke, which crosscuts the foliation, but is itself boudinaged. This implies that this rock is younger than the mylonitic fabric but older than the main deformation in the southern SMM. The entire outcrop of the Pirgadikia Unit at the type locality is about 1.8 km long and strictly confined to the area around the village. In both the north and the south, the rocks are tectonically overlain by massive marbles. Further inland along strike, near the village of Taxiarchis, there is a second occurrence of Pirgadikia-type rocks. The entire length of this exposure is 1 km, spread both to the north and south of the village. The best outcrops are to be seen along the recently built road to Polygiros. One good sampling site is on the road just N W of Taxiarchis village, and a second is near the Tjunction to Arnea (east) and Palaeocastro (west). The rock here (SM 56) is a mylonitic quartzite, which is more coarse-grained than the rocks at Pirgadikia, but equally sheared. As a result of strain partitioning, the diameter of quartz aggregates varies from 4 cm to 2-5 mm. The rock is essentially bimineralic, composed of quartz and white mica, with feldspar occurring only in minor amounts. Prior to deformation it may have been a feldspar-phyric sandstone. The augengneisses of the Vertiskos Unit are biotite gneisses with only minor white mica, which may be secondary in origin as a result of deformation and metamorphism. Furthermore, they are composed of quartz and large feldspar porphyroblasts, mainly plagioclase. SM 70 is a coarse-grained augengneiss from along the main road going south from Megali Panagia. This biotite orthogneiss has a strong foliation and a strong shear fabric. The feldspar augen are up to 10 cm across and delta clasts show a well-defined top-to-the-SW sense of shear. SM 95 is a finegrained mylonitic orthogneiss from the beach west of Ierissos. It is not as coarse-grained as SM 70 and has only small feldspar augen, but it is equally sheared with a prominent C/S fabric. This rock is in faulted contact with garnetiferous metasediments and the mafic and ultramafic rocks of the Gomati body (Dixon & Dimitriadis 1984). SM 98 is an L-tectonite gneiss from the road going from Metangitsi to Pirgadikia. This

40

F. HIMMERKUS E T AL.

gneiss has a strong greenschist-facies overprint as a result of emplacement into the m61ange zone bordering the Vertiskos Unit to the SW. Sample SM 99 represents the same body of basement rocks as SM 98, but has a crenulated foliation with a prominent C/S fabric.

Geochemistry Major- and trace-element concentrations were determined by X-ray fluorescence (XRF) analysis at the University of Mainz, and the results are listed in Table 1. According to the classification scheme of De la Roche et al. (1980) and the distribution in the TAS diagram (Le Maitre 1989; not shown) the rocks are granodiorites and granites; only the sample from Taxiarchis village is a quartzite. The alkali content is relatively high (5-8 wt% Na20 + K20) and all rocks are peraluminous.The granitoids have an intermediate Ti content. The amount of incompatible elements such as Nb, Y and Rb in granitic gneisses from Pirgadikia indicates that the precursor rocks originated in a magmatic arc setting. In the Rb v. (Y + Nb) discrimination diagram of Pearce et al. (1984; Fig. 3) the rocks plot on the borderline between volcanic-arc and within-plate granite. Sample SM 14 plots slightly inside the within-plate granite field, but this may be an effect of fractionation. Quartzite sample SM 56 is rich in silica, has a high alumina saturation index and very low concentrations of trace elements. The basement gneisses of the Vertiskos Unit are chemically similar to their counterparts in the northern S M M (Himmerkus et al. 2003). They are generally very rich in SiO2 and plot in the fields of granodiorites, granites and alkali granites in the total alkalis-silica (TAS; Le Maitre 1989) and De la Roche et al. (1980) discriminant diagrams. They are also mildly peraluminous. Potassium generally predominates over sodium. The rocks have a rather high content of Fe, P and Ti and significant amounts of compatible trace elements such as Sc and V. In the Rb v. (Y + Nb) discriminant diagram of Pearce et al. (1984) the rocks fall in the field of volcanic-arc granitoids. This is in good agreement with their lithology, which identifies them as I-type granitoids (see also Sr-isotope geochemistry, below). To conclude, trace-element geochemistry strongly suggests that the precursor rocks to both basement units formed in a magmatic arc setting.

Geochronology Most of the geochronological dating was performed using the P b - P b single-zircon evaporation method (Kober 1986, 1987). The zircons were

Table 1. Major and trace element concentrations of the samples of the Pirgadikia and Vertiskos Units used for geochronology Sample SM12 SM13 SM14 SM 56 Lithology Mylonite Mylonite Dyke Mylonite Locality Pirgadikia Pirgadikia Pirgadikia Taxiarchis wt%

SiO2 TiO2 A1203 Fe203(t) MnO MgO CaO Na20

K20 P205 LOI Sum

69.55 0.79 14.08 4.75 0.06 0.38 1.08 2.71 4.56 0.09 1.35 99.39

75.93 0.53 11.32 3.05 0.04 0.46 1.23 1.19 4.32 0.08 1.24 99.39

73.73 0.05 15.33 0.61 0.07 0.06 0.30 5.52 3.88 0.03 0.56 100.14

88.03 0.34 5.15 3.20 0.00 0.07 0.02 0.24 1.31 0.02 0.88 99.26

ppm

Sc V Cr Ni Cu Zn Ga Rb Sr Y Zr Nb Ba Pb Th La Ce Nd

9 72 16 7 3 63 18 173 169 33 258 23 928 23 17 40 87 40

Quartz 34.38 Orthoclase 27.48 Albite 23.39 Anorthite 4.86 Corundum 3.00 Hypersthene 0.97 Hematite 4.84 Ilmenite 0.13 Rutile 0.74 Apatite 0.22 Sum 100.00

8 61 26 9 3 21 15 137 188 22 196 14 897 25 14 57 101 44

3 5 8 3 3 3 20 97 93 41 49 17 257 28 8 6 16 5

3 38 23 6 14 10 6 43 31 20 134 7 312 5 4 11 15 7

50.31 26.01 10.26 5.68 2.69 1.17 3.11 0.09 0.49 0.19 100.00

26.24 23.02 46.90 1.30 1.58 0.15 0.55 0.10 0.00 0.07 100.00

82.87 7.87 2.06 0.00 3.40 0.18 3.25 0.00 0.35 0.05 100.00

handpicked and a representative fraction of the sample was mounted into low luminescent epoxy resin for investigation under the electron microscope. The zircons of the two Pirgadikia samples are small (50-250 ~tm) and basically euhedral. Most of them are colourless and only a few are partly resorbed, probably as a result of intense deformation. The zircons from Taxiarchis are

LATE PROTEROZOIC AND SILURIAN BASEMENT

41

Table 1. Continued

Sample Lithology Locality

SM 70 Bi-Gneiss Plana

SM95 Mylonite Ierissos

SM96 Bi-Gneiss Metangitsi

SM98 Gneiss Metangitsi

SM 99 Gneiss Metangitsi

66.51 0.78 14.99 4.98 0.05 1.76 0.47 2.52 4.26 0.15 2.74 99.20

75.02 0.10 12.72 1.65 0.06 0.24 1.38 3.10 4.20 0.02 0.51 99.03

64.89 0.82 16.59 5.83 0.06 1.89 0.37 1.55 3.73 0.13 3.83 99.69

74.29 0.27 12.57 1.89 0.02 0.48 0.55 1.84 5.55 0.13

68.15 0.30 16.03 2.70 0.04 0.91 2.48 3.39 3.14 0.14

wt%

SiO2 TiO2 A1203 Fe203(t) MnO MgO CaO NazO K20 P205 LOI Sum

1.20

1.90

98.81

99.18

ppm

Sc V Cr Ni Cu Zn Ga Rb Sr Y Zr Nb Ba Pb Th La Ce Nd Quartz Orthoclase Albite Anorthite Corundum Hypersthene Hematite Ilmenite Rutile Apatite Sum

13 88 39 13 30 103 17 124 78 34 218 16 1269 17 20 51 110 47 33.52 26.09 22.10 1.40 5.95 4.54 5.16 0.11 0.75 0.37 100.00

8 8 6 3 3 15 14 78 218 31 133 6 983 21 15 23 46 21

14 108 88 31 32 103 21 133 97 27 195 15 806 18 13 45 73 36

5 26 16 7 9 32 14 150 67 33 114 8 606 15 11 26 52 23

3 39 13 4 4 67 19 109 413 12 111 12 901 24 8 24 48 23

38.24 25.20 26.63 6.82 0.62 0.61 1.68 0.13 0.03 0.05 100.00

40.01 22.99 13.68 1.03 10.06 4.91 6.08 0.13 0.78 0.32 100.00

41.83 33.60 15.95 1.93 2.92 1.22 1.94 0.04 0.25 0.32 100.00

30.98 19.07 29.48 11.71 2.96 2.33 2.78 0.09 0.26 0.34 100.00

LOI, loss on ignition. also small (100-250 gm) a n d euhedral. A large n u m b e r o f t h e m have a d a r k pink colour. Some display pitted surfaces, a feature typical o f sedim e n t a r y zircons, w h i c h is due to abrasion during transport. Others do not show this feature, w h i c h m a y point to a short transport distance between erosion a n d sedimentation. To acquire m o r e i n f o r m a t i o n o n the internal structure of the zircons a representative fraction was studied by c a t h o d o l u m i n e s c e n c e imaging. The result is that

all of the zircons analysed show a m a g m a t i c z o n a t i o n and only a few have small inherited cores (see Fig. 4). We, therefore, interpret the ages o b t a i n e d as p r i m a r y intrusion ages o f the granitic p r e c u r s o r rocks to the gneisses. In the case of the quartzite o f Taxiarchis there is a magmatic p r e c u r s o r rock, w h i c h was eroded. The fact that there are no m e t a m o r p h i c zircons indicates t h a t there was little t h e r m a l overprint, only strong d e f o r m a t i o n .

42

F. HIMMERKUS E T AL. 1000 -

,,""

WPG

1000-

Syn -COLG

100--

\ . / /

E VAG + Z

10---

"gJ~,

syn-COLGO

9 -

/

I

rr

100 Y (ppm)

10-

ORG

I

10

100-

J

1

1000 0 Vertiskos 9 Pirgadikia

VAG

/ ORG

I

I

10

100

1000

Y + Nb (ppm)

Fig. 3. Discrimination diagrams for granites after Pearce et al. (1984). The incompatible elements Y, Nb and Rb define different tectonic settings. The samples of both units have rather low concentrations of these elements and plot in the field of the magmatic arc granitoids. WPG, within-plate granite; VAG, volcanic-arc granite; ORG, orogenic granite; syn-COLG, syn-collisional granite.

The zircons from the Vertiskos Unit are large (250-450 ~tm), euhedral and mostly translucent. Most of them are colourless or slightly yellow. They are long and prismatic, capped by various small pyramid faces. In comparison with the zircons from the Pirgadikia Unit they are paler in colour and have perfect crystal faces, which are not resorbed or pitted. Furthermore, they show a conspicuous magmatic zonation, which leads to the interpretation that the Pb-Pb age is the primary intrusion age of the granitic precursors to the gneisses (see Fig. 4). The Pb-Pb method faces the problems of inherited components and opening of the system that may cause lead loss as a result of metamorphism or fluid infiltration. These problems can be overcome by statistics. In this study a considerable number of the zircons from the Pirgadikia Unit show the phenomenon of inherited components, pointing to an older basement source. These zircon grains were identified by the method of cumulative probability and were not included in the calculation of the mean age using the Isoplot package (Ludwig 2001). The whole dataset is shown in Table 2, and because of open-system or inherited components a number of grains listed in the table do not appear in the weighted-average diagrams in Figure 5. The ages obtained for the two orthogneisses from Pirgadikia village are 570.0_+7.0 Ma and 5 8 7 . 6 + 3 . 4 M a . The two orthogneiss samples have a large number of inherited components indicating intrusion of the precursor rock into an

older Neoproterozoic crust that supplied the inherited cores in the zircons. To identify the grains with a complex system the measured zircons of the two samples were plotted in a probability plot, after Ludwig (2001), together with a histogram (Fig. 6) and show a clear maximum at c. 580 Ma, with a significant spread towards older ages. The Taxiarchis sample is slightly younger and gives an age of 555.8 + 2.6 Ma. This metasediment gives a very uniform zircon age, which leads to the interpretation that it has only one source and probably formed in the proximity of the Pan-African basement. It is interesting that there are no Ordovician or Silurian zircons present in this metaquartzite. The ages of the Vertiskos gneisses are 433.0 _+ 2.1 Ma for SM 70, 428.2 _+ 1.2 Ma for SM 95 and 430.7 _+3.7 Ma for SM 98. These ages are, within error, identical to the mean age of the entire unit, which is 435.0 _+3.0 Ma (Himmerkus et al. 2003). Sample SM 70 was also dated by the U - P b method. The results of this method are shown in Table 3 and the conventional concordia plot is shown in Figure 7. There are three concordant zircons supporting a Silurian Pb-Pb age. There are, however, two strongly discordant grains with an upper-intercept age of 2.5 Ga, suggesting the existence of an older basement source into which the precursor granite to this gneiss was intruded. This particular age is also known from zircon grains in metasediments associated with the Vertiskos Unit. This age is very common in the cratons of Gondwana, and the inherited components

LATE PROTEROZOIC AND SILURIAN BASEMENT

43

Fig. 4. Cathodoluminescence photomicrographs of typical magmatic zircons of the Pirgadikia and Vertiskos Units. The zircons of the Pirgadikia Unit are euhedral and show magmatic growth. In comparison with the zircons of the Vertiskos Unit they are rather small and short-prismatic. The zircons from the Vertiskos Unit are generally large, long-prismatic and clear, and have large pyramids with numerous small crystal faces. Zircons of both units show a magmatic zonation.

F. H I M M E R K U S E T AL.

44

Table 2. Individual Pb/Pb ages of the dated samples Sample

Grain

Ratios

207/206measured

206/204 corr

207/206corrected

2cr-mean

Age

2~r-mean

Mean age 2or

1 2 3 4 5 6 7 8 1 2 3 4 5 1 2 3 4

80 148 120 100 200 198 192 176 200 200 160 200 200 198 58 100 190

0.056231 0.056664 0.057659 0.056515 0.058657 0.056595 0.057090 0.057948 0.056425 0.055938 0.057591 0.056926 0.056147 0.059089 0.137585 0.056409 0.057857

20798 12653 4386 12702 4644 12721 9378 6032 14288 22480 6574 9255 11716 7764 201 12145 7666

0.055560 0.055561 0.054328 0.055364 0.055525 0.055438 0.055549 0.055540 0.055403 0.055342 0.055434 0.055398 0.054889 0.057214 0.055678 0.055472 0.055937

0.000100 0.000110 0.000160 0.000140 0.000100 0.000088 0.000044 0.000133 0.000053 0.000089 0.000140 0.000042 0.000081 0.000120 0.000360 0.000170 0.000060

433.6 434.9 384.4 426.9 433.4 429.9 434.0 434.0 428.6 426.1 429.6 428.3 407.7 430.4 431.1 431.3 449.9

5.2 4.4 6.6* 5.6 4.0 3.2 1.8 5.2 2.1 3.6 5.6 1.7 3.3* 4.8 14.5 6.8 2.4*

433.0

2.1

428.2

1.2

430.7

3.7

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10

176 58 196 184 198 200 118 198 112 134 72 100 98 110 36 112 176 50 170 170 184 196 156 98 40 200 196 20 178 110 34 198 136 60

0.063325 0.080928 0.068185 0.068300 0.067989 0.060752 0.066778 0.064283 0.067352 0.062523 0.065493 0.065677 0.067763 0.067226 0.062210 0.062256 0.070380 0.062159 0.063937 0.062127 0.062023 0.062212 0.062492 0.065237 0.060520 0.060420 0.059242 0.059217 0.059017 0.061020 0.061059 0.061097 0.060443 0.060456

4151 677 2247 1615 9633 18424 3009 3582 2131 4232 2474 4492 2556 2373 7641 5263 1257 5571 4278 5660 6444 5334 7028 3349 15525 7932 14657 14761 14341 6080 6033 5920 8909 8943

0.059704 0.060659 0.061772 0.059327 0.066488 0.059934 0.061995 0.060235 0.060657 0.059082 0.059034 0.062547 0.062070 0.061278 0.060351 0.059437 0.058796 0.059579 0.059883 0.059586 0.059817 0.059544 0.060421 0.060910 0.059585 0.058231 0.058251 0.058282 0.058007 0.058712 0.058650 0.058642 0.058767 0.058831

0.000100 0.000360 0.000097 0.000045 0.000052 0.000072 0.000180 0.000100 0.000100 0.000073 0.000100 0.000140 0.000140 0.000100 0.000160 0.000090 0.000079 0.000095 0.000110 0.000045 0.000067 0.000094 0.000110 0.000240 0.000250 0.000100 0.000060 0.000420 0.000120 0.000130 0.000280 0.000110 0.000170 0.000280

592.9 627.2 666.2 579.1 821.8 601.2 673.9 612.0 627.1 570.1 568.4 692.9 676.2 649.0 616.2 583.2 559.6 588.3 599.4 588.6 597.0 587.1 618.7 636.1 588.6 538.4 539.2 540.4 529.8 556.4 554.1 553.8 558.5 560.9

3.9* 13.0" 3.4* 1.7 1.7" 2.6* 6.4* 3.8* 3.8* 2.7 3.9 4.8* 5.2* 3.8* 5.7* 3.3 3.0* 3.5 4.1" 1.6 2.4* 3.5 3.9* 8.5* 9.1" 3.8* 2.3* 15.7" 4.5* 4.9 10.5 4.1 6.4 10.3

570.0

7.0

587.6

3.4

555.8

2.6

Vertiskos SM 70

SM 95

SM 98

Pirgadikia SM 12

SM 13

SM 56

*207/206-measured is the measured 2~176 ratio; 206/204 corr is the 2~176 ratio used for common lead correction; 207/206-corrected is the corrected 2~176 ratio; discussion of the age groups is given in the text. The inherited components indicate an older Neoproterozoic crust, which provided the majority of the cores.

LATE PROTEROZOIC AND SILURIAN BASEMENT 600

590

592

~S5M71.2P+_ 7.%Md ikia~ l ....... ...~ _ w

580

45

572 ~..............................................................................................................................

.......................................................................

590 588

..........

586 584 582

570

580

560 442

. -Maji3.4"i~

437

448

......................................................................................................................

440

.........................................................................................................................

429

432436 .................................................................................................. ...... 428f,

427

422, .

~ ..................................................................................................................................

433 431

......

(SM 70 M. Panagia l l 433..0..+2.1M a J

{SM56Taxiarchis~+ ............... k 555.8_ 2.6 M. j ...........

444.

430,

:...............

............... ...........

435

434

426,

'44 '48f

54ot

578

438.

418

9

423

421

428.2+ 1.

Ma .

..........

424§ ..............~ 42~

.....................

..........................

Metangitsi? ............... ;',;t'-~... 430.7 + 3.7 Ma J........... 98

Fig. 5. Weighted-average plots after Ludwig (2001) of the ages observed in the Pan-African rocks of the

Pirgadikia Unit and the Silurian orthogneisses of the adjacent Vertiskos Unit. Samples SM 12, SM 13 and SM 95 are granitic mylonites; SM 56 is a quartzite; and SM 70 and SM 98 are augengneisses. The fact that only a small proportion of the analysed samples were used for the weighed average is due to problems with inherited components or lead loss in several grains (discussed in the text).

in the orthogneisses and metasediments therefore place constraints on the provenance of the basement of the Vertiskos Unit. Remnants of this old basement may also be represented by the Pan-African rocks of the Pirgadikia Unit despite the fact that the contact between the two units is entirely tectonic. To summarize, the zircon ages obtained in this study define two consistent units of gneisses formed in Late Precambrian and Silurian times in a volcanic-arc or active continental-margin setting. The inherited cores support the idea that both the Pirgadikia and the Vertiskos Units formed on Gondwana-derived basement. St-isotope characteristics

To test the tectonic environment for a magmatic arc suggested by trace-element geochemistry, R b - S r isotope geochemistry was employed to gain additional information about the precursor rocks. Sr-isotope ratios were measured in the static mode on the Faraday cups of the M A T 261 Finnigan mass spectrometer of the Max-PlanckInstitut ftir Chemie in Mainz. For the 87Rb/86Sr ratio we used the X R F data, as the elemental concentrations were well over the detection limit and accurate enough to calculate 87Sr]86Sr initial ratios using the ages obtained by the zircon dating. Individual ratios are shown in Table 4. In this

study, the Sr-isotopic signature was used merely as a tracer for crustal components in the source of the granitic precursor rocks to the gneisses. The 87Sr]86Sr initial ratios calculated for the two orthogneisses from Pirgadikia are 0.70644 for SM 12 and 0.70734 for SM 13. These values are typical for I-type granites and indicate a source with a large proportion of juvenile material and little input from pre-existing continental crust. The initial ratio of the metaquartzite is meaningless for an age of 555 Ma, which is the age of the source. However, if a Late Palaeozoic to early Mesozoic age of around 250 300 Ma is assumed, the calculated 87Sr/86Sr initial ratio is between 0.707 and 0.709, i.e. within the range of the global seawater curve for that time. This may be a hint of a Late Palaeozoic to Early Mesozoic age of deposition; sediments of such an age are indeed known from the Aegean region (e.g. Chios; Zanchi e t al. 2003). The fact that no Silurian ages are present in the metaquartzite indicates that the Silurian rocks of the Vertiskos Unit were not in the source region at the time of deposition. With regard to the Vertiskos Unit orthogneisses collected in the immediate vicinity of Pirgadikia village, two of the samples (SM 70 and SM 99) are isotopically disturbed, whereas the other two have calculated 87Sr/86Srinitial ratios of 0.70805 (SM 95) and 0.70876 (SM 98). Also, the

F. H I M M E R K U S ET AL.

46

Fig. 6. Probability plot and histogram for samples SM 12 and SM 13 of Pirgadikia after Ludwig (2001). The two samples together show a well-defined peak at around 580 Ma. The spread to older ages is attributed to inherited components of pre-existing Neoproterozoic crust. The younger ages may be due to open system and lead loss during a metamorphic event.

Table 3. Results of the U-Pb dating of sample SM 70 Zircon

U (ppm)

Pb (ppm)

1 2 3 4 5 6

1096.08 491.22 520.38 212.28 51.85 309.41

122.04 62.53 59.24 13.57 3.34 22.30

Zircon 1 2 3 4 5 6

2~ 0.1035 0.1185 0.1093 0.0654 0.0654 0.0705

2or 0.0010 0.0010 0.0013 0.0008 0.0012 0.0040

Pb-nonrad (pg)

2~

42.96 27.07 16.17 58.42 20.50 57.33 Age 634.86+6.00 722.08+5.61 668.70___7.75 408.20+5.00 408.14___7.03 439.10+24.17

2or

1.4561 1.7480 1.6044 0.4989 0.4984 0. 5435 r 0.6029 0.4481 0.7614 0.5150 0.5503 0.9912

2~176 0.1020 0.1070 0.1065 0.0554 0.0553 0.0560

0.0312 0.0404 0.0346 0.0152 0.0198 0.0343 2or 0.0013 0.0016 0.0011 0.0011 0.0014 0.0008

Age 912.38 1026.41 971.93 410.95 410.64 441.11

__ 19.97 __ 15.03 __ 13.59 __ 10.38 __ 13.48 __+25.19

Age 1661.500-22.96 1748.22+28.22 1739.64___18.47 426.42__+45.12 424.77___57.55 451.64+30.41

In the conventional concordia diagram in Figure 7 grain 2 is left out, because it plots over the discordia.

LATE PROTEROZOIC A N D S I L U R I A N BASEMENT

47

Fig. 7. Conventional concordia diagram of sample SM 70 of the Vertiskos Unit, drawn using the Isoplot package (Ludwig 2001). There are three zircon grains that are concordant and confirm the Pb-Pb age. Two other grains are strongly discordant and point to the age of the basement into which the precursor to the Vertiskos gneisses intruded. Table 4. Sr-isotopic ratios of the Vertiskos and Pirgadikia Unit

Vertiskos (age=430 Ma) Pirgadikia (age = 560 Ma)

Sample

875r/86Sr

2s

Sr (ppm)

Rb (ppm)

SM SM SM SM SM SM SM

0.722502 0.714399 0.748591 0.721722 0.730144 0.724201 0.726473

8 11 16 11 13 5 16

91 218 67 413 169 188 31

7 78 150 109 173 137 43

70 95 98 99 12 13 56

87Rb/86Sr 0.22 1.04 6.50 0.76 2.97 2.11 4.02

(875r/86Sr)initial 0.694292* 0.708055 0.708760 0.717039* 0.706447 0.707341 0.694374 t 0.707-0.709

*Disturbed isotopic system. 1-Assumed age 250-300 Ma. The 87Sr/S6Sr initialratio was calculated using the zircon age. t w o u n d i s t u r b e d s a m p l e s define an i s o c h r o n with an age o f 439_+ 16 M a , w h i c h is in g o o d agreem e n t with the zircon ages. T h e 87Sr/86Sr initial

ratio o f this i s o c h r o n is 0.70792 4- 0.00036, w h i c h is, w i t h i n error, identical to the m e a n o f the c a l c u l a t e d initials.

48

F. HIMMERKUS E T AL.

Such 87Sr/86Sr initial ratios support the proposed plate tectonics scenario of a magmatic arc or active continental margin. However, the Vertiskos Unit generally shows a slightly higher 87Sr/S6Sr initial ratio than the Pirgadikia Unit, and this may be interpreted as the result of a different amount of old continental material in the source.

Discussion and conclusions The results of the present geochemical and geochronological study point to volcanic-arc magmatism during Late Precambrian times, as recorded in the Pirgadikia Unit of the SMM. This is not, however, the only occurrence of Late Precambrian orthogneisses in the Internal Hellenides and the eastern Mediterranean region. Graf et al. (1998) have reported orthogneisses with U Pb intrusion ages of 545.1 + 6.4 to 568 + 7.5 Ma from the Osogovo-Lisets dome of the Bulgarian Struma Unit. This unit is situated north of the Bulgarian SMM (Ograzden Unit), about 100 km from the Greek-Bulgarian border (see Fig. 1). Other occurrences of orthogneisses with a similar age are in the Menderes Massif of southwestern Turkey (Hetzel & Reischmann 1996; Loos & Reischmann 1999 and in the Karadere basement (Istanbul Zone) of northwestern Turkey (Chen et al. 2002). In the former locality, typical intrusion ages range from 520 to 570 Ma with a mean age of 550 Ma, whereas in the latter they range from 560 to 590Ma. In both localities the orthogneisses display a magmatic arc signature (Dannat & Reischmann 1997; Chen et al. 2002). It may thus be surmised that the small crystalline basement outcrops of Pirgadikia and Taxiarchis villages may be correlated with similar crustal blocks or micro-terranes in other parts of the Balkans and Turkey. The mere presence of these Pan-African rocks underscores the allochthonous character of at least parts of the SMM and its relation to similar basement massifs in the north and along a suspected Palaeo-Tethyan suture ($eng6r et al. 1984; Stampfli et al. 2004). A similar line of thinking was also proposed by Neubauer (2002), who summarized Late Precambrian and Early Palaeozoic ages in the Alpine orogenic belt. The Silurian orthogneisses of the Vertiskos Unit represent a second distinct basement complex with a magmatic arc signature. Correlation of this unit with adjacent basement complexes is not possible, thus making it another exotic block within the Hellenic orogen. There are, however, Ordovician and Silurian rocks known from the internal parts of the Variscan and Alpine orogenic belts (Neubauer & Von Raumer 1993; Von Raumer et al. 2003).

In the SMM, all contacts between the Pan-African and Silurian rocks and their neighbouring crustal units are tectonic, so no mutual relationships can be identified. Our interpretation is that the SMM is an assemblage of distinct crustal terranes of different origin that were amalgamated during closure of the Vardar Ocean in Late Jurassic times (Mercier 1966; Kaufmann et al. 1976). This assemblage contains exotic blocks such as the Vertiskos and Pirgadikia Units, but also rocks that can be correlated to the adjacent basement complexes of the Rhodope Massif and the Pelagonian Zone. Until recently, the Internal Hellenides east of the Vardar Zone were regarded as a stable craton in the hinterland of the Hellenides, little influenced by the Alpine phase (Kockel & Mollat 1977; Jacobshagen 1986). However, the identification of nappe structures in the Rhodope Massif (Burg et al. 1995, 1996) nullified this scenario. The fact that there are exotic terranes of Gondwana origin present in the crystalline basement of the SMM makes the whole Hellenic orogen an accretionary orogen, comprising NW-SE-trending terranes that contain Mesozoic sediments in their external parts and crystalline basement units in their intrnal parts. Basement units with Pan-African or Cadomian ages, comparable with those of the Pirgadikia Unit, occur in almost all orogenic chains in Western Europe as well as in the Alpine orogenic belt of Asia. They can be identified as slivers of Gondwana that were incorporated into young collisional orogens. The same is true for Ordovician to Silurian gneisses that might be related to the Vertiskos Unit. These rocks may be remnants of a large active continental margin that originated from the northern margin of Gondwana. Similar rocks of this age are correlated with the 'Hun Terrane' (Stampfli & Borel 2002). The Tethyan oceans were created by rifting of variably sized terranes of different origin from the Gondwana supercontinent, and the Variscan and Alpine orogens are the products of accretion of these terranes to the European craton. We conclude that the Hellenic orogen is an accretionary orogen and that terrane accretion studies can give valuable information with regard to the build-up history of a large part of Europe and Asia. The authors wish to thank the other members of the working group at the University of Mainz and the staff of the Max-Planck-Insitut ffir Chemie, Abteilung Geochemie in Mainz, especially Wolfgang Todt, Ulrike Poller and Ingrid Raczek. Special thanks go to J. E. Dixon (University of Edinburgh), Sarantis Dimitriadis (Thessaloniki) and the group of Chris Ballhaus (MUnster) for helpful reviews of the original draft.

LATE PROTEROZOIC AND SILURIAN BASEMENT

References ANDERS, B., REISCHMANN, T., POLLER, U. & KOSTOPOULOS, D. 2003. The oldest rocks in Greece: geochronological evidence for remnants of a Precambrian basement within the central Hellenides. Goldschmidt Conference Abstracts, Geochimica et Cosmochimica Acta Supplement, 67, A18. BARR, S. R., TEMPERLEY, S. T. & TARNEY, J. 1999. Lateral growth of the continental crust through deep level subduction-accretion: a re-evaluation of central Greek Rhodope. Lithos, 46, 69-94. BURG, J.-P., GODERIAUX,I. & RICOU, L.-E. 1995. Extension of the Mesozoic Rhodope thrust units in the Vertiskos-Kerdyllion Massifs (Northern Greece). Comptes Rendus de l'Acadkmie des Sciences, S~rie a, 320, 889-896. BURG, J.-P., RIcou, L.-E., IVANOV,Z., GODFRIAUX,I., DIMOV, D. & KLAIN, L. 1996. Syn-metamorphic nappe complex in the Rhodope Massif. Structure and kinematics. Terra Nova, 8, 6-15. CHEN, F., SIEBEL,W., SATIR, M., TERZIOGLU,M. N. t~ SAKA, K. 2002. Geochronology of the Karadere basement (NW Turkey) and implications for the geological evolution of the Istanbul zone. International Journal of Earth Sciences, 91,469-481. DANNAT, C. & REISCHMANN,T. 1997. A geochemical, isotopic and geochronological study of granitoid gneisses of the Menderes Massif, SW Turkey. Terra Nova, 9, 404. DE LA ROCHE, H., LETERRIER, J., GRANDCLAUDE,P. & MARCHAL, M. 1980. A classification of volcanic and plutonic rocks using R1-R1 diagrams and major and trace element analysis--its relationship to modern nomenclature. Chemical Geology, 29, 183-210. DE WET, A. P. 1989. Geology of apart of the Chalkidiki peninsula, Northern Greece. PhD Thesis, Cambridge University. DE WET, A. P., MILLER, J. A., BICKLE, M. J. & CAPMAN, H. J. 1989. Geology and geochronology of the Arnea, Sithonia and Ouranoupolis intrusions, Chalkidiki peninsula, Northern Greece. Tectonophysics, 161, 65-79. DIM1TRIJEVI~, M. D. 1974. Sur l'fige du m&amorphisme et des plissements darts la masse SrrboMacrdonienne. Bulletin de l'Association Gkologique Carpatho-Balkanique, 21, 45~18. DIM1TRIJEVI~, M. D. 1995. Geology of Yugoslavia. Geoinstitut, Belgrade. DIXON, J. E. & DIMITRIADIS, S. 1984. Metamorphosed ophiolitic rocks from the Serbo-Macedonian Massif, near Lake Volvi, North-east Greece. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Specical Publications, 17, 603-618. DI3RR, S., ALTHERR, R., KELLER, J., OKRUSCH, M. & SEIDEL, E. 1978. The median Aegean crystalline belt: stratigraphy, structure, metamorphism, magmatism. In: CLOSS, H., ROEDER, U. H. & SCHMIDT, K. (eds) Alps, Apenines, Hellenides. Report 38, IUGS. Schweizerbart, Stuttgart, 455-477.

49

ENGEL, M. & REISCHMANN, T. 1998. Single-zircon geochronology of the orthogneisses from Paros, Greece. Bulletin of the Geological Society of Greece, 32(3), 91-99. FREI, R. 1996. The extent of inner mineral isotope equilibrium: a sytematic bulk U-Pb and Pb step leaching (PbSL) isotope study of individual minerals from a tertiary granite of Lerissos (northern Greece). European Journal of Mineralogy, 8, 1175-1189. GRAF, J. 2001. Alpine tectonics in western Bulgaria." Cretaceous compression of the Kraiste region and Cenozoic exhumation of the crystalline Osogovo-Lisec Complex. PhD thesis, ETH, Zfirich, Switzerland. GRAF, J., BERNOULLI, D., BURG, J.-P., IVANOV,Z. & VON QUADT, A. 1998. Geochemistry and geochronology of igneous rocks of the central SerboMacedonian Massif (Western Bulgaria). Abstracts Carpathian-Balkan Geological Association XVI Congress. Geological Survey of Austria, Vienna, 191. HETZEL, R. & REISCHMANN, T. 1996. Intrusion age of Pan-African augen gneisses in the southern Menderes massif and the age of cooling after Alpine ductile extensional deformation. Geological Magazine, 133, 565-572. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. 2002. First evidence for Silurian magmatism in the Serbo-Macedonian Massif, northern Greece. Geochimica et Cosmochimica Acta, Goldschmidt Conference Abstract Volume. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. 2003. The Serbo-Macedonian Massif, the oldest crustal segment of the internal Hellenides, identiffed by zircon ages. Geophysical Research Abstracts, 5, number 05671. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. 2004a. The Pirgadikia Unit, the oldest crustal segment in the Serbo-Macedonian terrane assemblage. 5th ISMEG Conference Proceedings, 84-85. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. 2004b. Triassic rifting recorded in Gondwana derived Tethyan terranes, Serbo-Macedonian Massif, northern Greece. Berichte der Deutschen Mineralogischen Gesellschaft, Beihefte zum. European Journal of Mineralogy, 16, 57. JACOBSHAGEN, V. (ed.) 1986. Geologie yon Griechenland. Geologisches Borntraeger, Stuttgart. JACOBSHAGEN, V. 1994. Orogenic evolution of the Hellenides: new aspects. Geologische Rundschau, 83, 249-256. KAUFMANN, G., KOCKEL, F. & MOLLAT, H. 1976. Notes on the stratigraphic and palaeogeographic position of the Svoula Formation in the Innermost Zone of the Hellenides (Northern Greece). BulIetin de la Soci~tk Gkologique de France, 18, 225-230. KILLAS, A., FALALAKIS,G. & MOUNTRAKIS, D. 1999. Cretaceous-Tertiary structures and kinematics of the Serbomacedonian metamorphic rocks and their relation to the exhumation of the Hellenic hinterland (Macedonia, Greece). International Journal of Earth Sciences, 88(3), 513-531.

50

F. HIMMERKUS ET AL.

KOBER, B. 1986. Whole grain evaporation for 2~ 2~ investigations on single zircons using a double filament thermal ion source. Contributions to Mineralogy and Petrology, 93, 482-490. KOBER, B. 1987. Single zircon evaporation combined with Pb+emitter-bedding for 2~176 investigations using thermal ion mass spectrometry, and implications to zirconology. Contributions to Mineralogy and Petrology, 96, 63-71. KOCKEL, F. & MOLLAT, n . 1977. Erliiuterungen zur geologischen Karte der Chalkidiki und angrenzender Gebiete 1:100000 (Nord-Griechenland). Bundesanstalt fiir Geowissenschaften und Rohstoffe, Hannover. LE MAITRE, R. W. (ed.) 1989. A Classification of Igneous Rocks and Glossary of Terms. Blackwell Scientific, Oxford. LIATI, A., GEBAUER, D. & FANNING, C. M. 2004. The age of ophiolitic rocks of the Hellenides (Vourinos, Pindos, Crete): first U-Pb ion microprobe (SHRIMP) zircon ages. Chemical Geology, 207, 171-188. LIPS, m. L. W., WHITE, S. H. & WIJBRANS, J. R. 2000. Middle-Late Alpine thermotectonic evolution of the southern Rhodope Massif, Greece. Geodinamica Acta, 13, 281-292. Loos, S. & REISCHMANN, T. 1999. The evolution of the southern Menderes Massif in SW Turkey as revealed by zircon dating. Journal of the Geological Society, London, 156, 1021-1030. LUDWIG, K. R. 2001. Isoplot/Ex. Berkeley Geochronology Centre Special Publication. MERCIER, J. 1966. Etudes g6ologique des zones internes des Hell6nides en Mac6doine centrale (Gr6ce), II--Contribution h l'6tude du m6tamorphisme et de l'6volution magmatiques des zones internes des Hell6nides. Annales Gkologique de Pays Hellkniques, 20, 792. MERCIER, J., VERGI~LEY, P. & Bt~BIEN, J. 1975. Les ophiolites h611eniques 'obduct6s' au Jurassique sup6rieur sont-elles les vestiges d'un oc6an tethysien ou d'une mer marginale peri-europ6enne? Bulletin de la Sociktk Gkologique de France, 17, 108-112. MUSSALAM, K. & JUNG, D. 1986. Geologie und Bau des Sithonia-Ophioliths (Chalkidiki, NEGriechenland: Anmerkungen zur Bildung ozeanischer Krusten). Geologische Rundschau, 75(2), 383-409. NEUBAUER, F. 2002. Evolution of late Neoproterozoic to early Paleozoic tectonic elements in central and Southeast European Alpine mountain belts: review and synthesis. Tectonophysics, 352, 87-103. NEUBAUER, F. & VON RAUMER, J. 1993. The Alpine basement: linkage between west European Variscides and Alpine-Mediterranean mountain belts.

In: VON RAUMER, J. & NEUBAUER, F. (eds) Pre-Mesozoic Geology in the Alps. Springer, Berlin, 640-663. PARADOPOULOS, C. • KILIAS, A. 1985. Altersbeziehungen zwischen Metamorphose und Deformation im zentralen Teil des Serbomazedonischen Massivs (Vertiskos Gebirge, Nordgriechenland). Geologische Rundschau, 74, 77-85. PEARCE, J. A., HARRIS, N. B. W. t~; Tindle, A. G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, 25, 956-983. RIcou, L.-E., BURG, J.-P., GODFRIAUX, I. & IVANOV, Z. 1998. Rhodope and Vardar: the metamorphic and the olistostromic paired belts related to the Cretaceous subduction under Europe. Geodinamica Acta, 11, 285-309. ROBERTSON, A. H. F. 2002. Overview of the genesis and emplacement of Mesozoic ophiolites in the Eastern Mediterranean Tethyan region. Lithos, 65, 1-67.

SENGOR, A. M. C., YILMAZ, Y. & SUNGURLU,O. 1984. Tectonics of the Mediterranean Cimmerides: nature and evolution of the western termination of Palaeo-Tethys. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 603-618. STAMPFL1, G. M. t~ BOREL, G. D. 2002. A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrones. Earth and Planetary Science Letters, 196, 17-33. STAMPFLI, G. M., MOSAR, J., DE BONO, A. & VAVASIS, I. 1998. Late Palaeozoic, Early Mesozoic plate tectonics of the western Tethys. Bulletin of the Geological Society of Greece, 23, 113-120. STAMPFLI, G. M., ROSSELET, F., & BAGHERI, S. 2004. Tethyan oceans and sutures. 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, April 14-20, 2004, Ref: T1-21. TURPAUD, P. & REISCHMANN, T. 2003. Zircon ages of granitic gneisses from the Rhodope (N. Greece), determination of basement age and evidences for a Cretaceous intrusive event. Geophysical Research Abstracts, 5, number 04435. VON RAUMER, J. F., STAMPFLI, G. M. & BussY, F. 2003. Gondwana-derived microcontinents--the constituents of the Variscan and Alpine collisional orogens. Tectonophysics, 365, 7-22. ZANCHI, A., GARZANTI,E., LARGH1, C., ANGIOLINI,L. & GAETANI, M. 2003. The Variscan orogeny in Chios (Greece): Carboniferous accretion along a Palaeotethyan active margin. Terra Nova, 15, 213-223.

Stratigraphy, correlations and palaeogeography of Palaeozoic terranes of Bulgaria and NW Turkey: a review of recent data S. Y A N E V 1, M . C. G O N C O O G L U V. S A C H A N S K P ,

C. O K U Y U C U

2, I. G E D I K 3, I. L A K O V A 1, I. B O N C H E V A l, 3, N . O Z G U L 4, E. T I M U R 3, Y. M A L I A K O V 1 & G. S A Y D A M 3

~Bulgarian Academy of Sciences, Geological Institute, Acad. G. Bonchev St., Bl. 24, 1113 Sofia, Bulgaria (e-mail." snyanev@geology, bas. bg) 2Middle East Technical University, Department of Geological Engineering, 06531 Ankara, Turkey 3General Directorate of Mineral Research and Exploration, Department of Geological Research, 06520 Ankara, Turkey 4GEOMAR, Cengizhan S. 18/3, 34730 Istanbul, Turkey Within the Alpine tectonic units SE of the European Variscan Orogenic Belt in Bulgaria and NW Turkey several crustal blocks are identified. Although their contact relations with surrounding units are obscured by Alpine events, the differences in the succession of events, stratigraphy, sedimentology and palaeobiogeographical distribution within them permits recognition of the Moesian, Balkan, Istanbul and Zonguldak Terranes. The Moesian terrane corresponds to the pre-Variscan Palaeozoic and Neoproterozoic rocks of the Moesian microplate in north Bulgaria and south Romania. The Balkan Terrane in Bulgaria incorporates Neoproterozoic and Palaeozoic sequences in the Western Balkanides (part of the Carpathian-Balkan orogen) and another three allochthonous units (Kraishte, Central Balkanides and Strandzhides). In NW Anatolia in Turkey, the Caledonian basement and Ordovician to Carboniferous sedimentary succession are divided into the Istanbul Terrane and the Zonguldak Terrane. With the exception of the Moesian Terrane in the Bulgarian area, they all comprise a Cadomian basement with relicts of oceanic lithosphere, volcanic arc and a continental crust of unknown affinity. Based on characteristic features within their Palaeozoic successions, these terranes are correlated with the main terrane assemblages in Central and Eastern Europe. It is suggested that they all are of periGondwanan origin but behaved independently while drifting towards Laurussia. During the Early Devonian the Zonguldak Terrane docked to Baltica, whereas the others were still at similar palaeolatitudes to the Central European terranes (e.g. Saxo-Thuringian). This was followed by the successive accretion of the Moesian Terrane to Laurussia along the Rhenohercynian suture at the end of Devonian-Early Carboniferous and of the Balkan and Istanbul Terranes between the Early and Late Carboniferous. Abstract:

The Variscan Orogenic Belt in Europe is characterized by a mosaic of Gondwana-derived crustal blocks or terranes, which were successively accreted to Laurussia during the Palaeozoic. The position of the Palaeozoic terranes in Bulgaria (Balkan and Moesia) and in Turkey (Taurus, Istanbul, Zongulda~) (Fig. 1) is shown in the palaeogeographical reconstruction of McKerrow & Scotese (1990), although McKerrow & Scotese's suggestion is of rather a Baltican origin of the Istanbul and Zonguldak Zerranes. The purpose of this paper is to review the stratigraphic, sedimentological and palaeogeographical data accumulated recently on the

Palaeozoic of the Moesian and Balkan Terranes in Bulgaria (Fig. 2) (as defined by Yanev 1990, 1993, 1997, 2000; Haydutov & Yanev 1996) and the Istanbul and Zonguldak Terranes (Fig. 3) (G6ncfio~lu 1997, 2001; G6ncfio~lu & Kozur 1998, 1999; Kozur & G6ncfio~lu 2000) in N W Turkey. In addition, the palaeogeographical position of the Moesian, Balkan, Istanbul and Zonguldak Terranes during Palaeozoic time is discussed here in the light of the evolution of the Variscan Orogenic Belt and the Trans-European Suture Zone (Berthelsen 1993), where it separates Avalonia-Baltica from the members of the Armorican Terrane Assemblage.

From:ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the EasternMediterranean Region. Geological Society, London, Special Publications, 260, 5147. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

52

S. YANEV E T AL.

Fig. 1. Palaeogeographical reconstruction of Gondwana, Baltica, Laurussia and peri-Gondwanan European terranes (after McKerrow & Scotese 1990). In recent years, detailed work has been carried out the geology, palaeogeography and geodynamics of these terranes in Western and Central Europe (e.g. Pharaoh 1999; Franke 2000; Winchester & PACE TMR Network Team 2002). In the west, Avalonia was one of the earliest recognized Gondwana-derived terranes that was already accreted to Baltica at the end of the Ordovician. Recently, it was suggested that this was not restricted to Southern Britain but may well continue towards Central and Eastern Europe (e.g. Moravo-Silesian terrane; Pharaoh 1999) to include some small crustal blocks. The next group of Gondwanan terranes that were accreted to Baltica-Avalonia later in the

Palaeozoic is the Armorican Terrane Assemblage (Franke 2000), which includes several crustal blocks within the Variscan Belt in Central and SE Europe (e.g. Bohemian Massif). The far eastern part of the Variscan Belt, however, remains relatively less-known to the international community. Being located on the eastern extension of the Variscan Belt and being involved in post-Variscan orogenic events, this region should theoretically include dismembered pieces of the Eastern European Craton and its cover (Baltica-derived terranes), the Avalonian Terrane, the Armorican Terrane Assemblage or other peri-Gondwanan terranes. Regional palaeogeographical reconstructions (e.g. G6riir

TERRANES OF BULGARIA AND NW T U R K E Y

Fig. 2. Geological sketch showing the Palaeozoic terranes and outcrops in Bulgaria.

Fig. 3. Geological sketch showing the Palaeozoic terranes and outcrops in NW Anatolia, Turkey.

53

54

S. YANEV E T AL.

et al. 1997; Stampfli 2000; Kalvoda 2001; Von Raumer et al. 2002) for this area, including the northern part of Balkan Peninsula and N W Anatolia, were mainly based on oversimplified previous work and did not take new data and comprehensive stratigraphical evidence into account, and thus are somewhat speculative.

Moesian Terrane Stratigraphy and sedimentology Western and central part o f the Moesian Terrane.

In the western and central part of the Moesian Terrane, in Bulgaria, the Palaeozoic sediments consist of Upper Silurian to Vis6an marine deposits and a Permian continental cover. The oldest marine sediments are Lower Silurian (Pridoli) and Lower Devonian black shales about 200 m thick with some bivalves and trilobites, and chitinozoans, acritarchs and spores. Palynological evidence supports a latest Silurian and Lochkovian age (Lakova 1993, 2001a,b; Steemans & Lakova 2004) with continuous sedimentation across the Silurian-Devonian boundary. There is no record of Pragian to Lower Eifelian sediments. The Middle Devonian sequence comprises 800 m of dolomitic limestones, calcareous dolomites and micritic limestones, with 60m of Emsian shales at the base. The Mid-Devonian age is recognized using Foraminifera, brachiopods and conodonts (Spassov et al. 1978; Vdovenko et al. 1981; Boncheva et al. 2002). A slight angular unconformity with clastics (calcirudites) at the base is found at the EmsianEifelian and Eifelian-Givetian boundaries in the central part. The Upper Devonian sequence is missing. The boundary between the Middle Devonian and the Lower Carboniferous is an erosional surface as proved by conodont and sediment data (Boncheva et al. 2002). In the west, the Vis6an limestones with algae, crinoids and ostracodes, black shales and dolomites overlie Tournaisian limestones. An Early Carboniferous age was proved using conodonts and Foraminifera (Spassov 1977; Vdovenko et al. 1981; Boncheva et al. 2002). The Lower Carboniferous sequence is about 730 m thick whereas Upper Carboniferous units are missing. In the central part, 580 m of Carboniferous continental shales, siltstones, sandstones and coal-bearing shales were shown to be Tournaisian to Early Namurian in age by macro- and microflora. These are the only coal-bearing Carboniferous sediments outside the Dobrudgea coal basin in east Moesia (Nikolov et al. 1990;

Dimitrova 1996). The Westphalian sequence is missing. With a contrasting lithology and clear discordance, Permian continental clastic rocks cover either Middle Devonian or Vis6an-Namurian rocks. The Permian sequence consists of reddish breccias-conglomerates, sandstones and siltstones, 50-800 m thick. These drastic variations in the thicknesses are controlled by the pre-Permian palaeotopography. Eastern part o f the Moesian Terrane. The Palaeo-

zoic section consists of a marine sequence from Ordovician to Vis6an (with numerous local discontinuities) covered unconformably by continental Carboniferous and Permian deposits. The oldest subsurface sediments in the eastern part of the Moesian Terrane in Bulgaria are Ordovician pelitic rocks about 100 m thick. In Romania, the Ordovician sequence, mainly encountered in boreholes, is 750 m thick and dated by palynomorphs (Parashiv & Beju 1974). The overlying Silurian and Lower Devonian units are mainly dark shales and siltstones with minor limestones and marls, up to 2000 m thick. Conodont and graptolite faunas prove the existence of Llandovery and Wenlock Series (Spassov & Yanev 1966). Chitinozoan, acritarchs and spores provide evidence of a Pridolian and Early Devonian age (Lakova 1993, 2001a,b; Steemans & Lakova 2004). Locally, thin quartzites and sandstones of possible Emsian-Eifelian age cover the Lower Devonian sequence with shales. In other areas, the Lower Emsian sequence is directly covered by Eifelian carbonate sequences (Spassov 1987; Boncheva 1995). The Middle-Upper Devonian to Vis6an carbonate sequence in the subsurface is subdivided into six informal lithostratigraphic series: carbonate-sulphate, dolomite, banded limestones, intraclastic limestone, organic limestones and clastic limestones (calcirudites). The total thickness of carbonate platform deposits penetrated is 1200-2000 m, thickening from NW to SE. The assumed stratigraphical thickness may be as great as 3000 m. Fossil data on conodonts (Spassov 1983; Boncheva et al. 1994, 2000; Boncheva 1995; Yanev & Boncheva 1997) prove Eifelian, Givetian, Frasnian, Famennian and Vis6an stages. Spassov (1987) provided macrofossil constraints on the Eifelian age on corals, brachiopods, ostracodes and trilobites. The Upper Vis6an, locally developed to the east, is up to 2300 m thick and consists of limestones at the base, followed by dark shales with coal layers and sandstones. The characteristic feature of the carbonate~lolomite sequence in the eastern part

TERRANES OF BULGARIA AND NW TURKEY

55

Fig. 4. Generalized stratigraphy of the Palaeozoic in the Balkan and Moesian Terranes in Bulgaria. of the Moesian Terrane is a dozen widespread unconformities within the Middle Devonian to Permian sequence, as established by sedimentological data and conodont biostratigraphy (Yanev & Boncheva I995). In the Dobrudgea Coal Basin, Middle to Upper Devonian carbonates are unconformably overlain by Upper Namurian-Westphalian coalbearing terrigenous strata (Fig. 4, in the composite column EM of eastern part of Moesia and Dobrudgea Coal Basin). The Tournaisian and Lower Vis6an units are missing. The Permian sequence consists of conglomerates, sandstones, shales and evaporites. The

great variations in the thickness of the Carboniferous (0-3000 m) and Permian sequences (03500 m) resulted from post-Palaeozoic erosion.

Palaeogeography Palaeogeographical interpretation and reconstruction of the Moesian Terrane is based on combined biogeographical, palaeoclimatic and palaeomagnetic analysis. The palaeobiogeographical interpretations are based on palynomorphs from the Upper Silurian and Lower Devonian sequences. The chitinozoan faunas of the Lochkovian and

56

S. YANEV ET AL.

Emsian in the Moesian Terrane show clear peri-Gondwanan affinities with North Africa, Spain and Brittany (Lakova 1995). Coeval acritarchs show palaeogeographical affinities (Lakova 2001b) with Brittany, Spain, North Africa and Southern England. Recently, palaeobiogeographical analysis of Lochkovian spores has revealed affinities with Belgium, Southern Britain and Poland (Steemans & Lakova 2004). The northern position of the Moesian Terrane in the Lochkovian, as indicated by palaeophytogeography, supports the hypothesis of northward drift of Moesia in Ordovician to Devonian times from Gondwana to Laurussia. The palaeoclimatic interpretations are based on Palaeozoic rocks and minerals indicating specific climatological conditions and zones, and thus palaeo-latitudes (Yanev 1990, 2000). In the Ordovician to Early Devonian the abundance of organic matter in the predominantly shaly sequence and the presence of Fe-oolitic minerals provide evidence of sedimentation in a temperate zone. Anhydrites in the Givetian of the eastern part suggest a transition to an arid zone. The Upper Carboniferous coal-bearing succession indicates deposition in the equatorial zone. These palaeoclimatic interpretations support a northward migration of the Moesian Terrane depositional environment from the southern temperate zone in the Silurian to the southern arid zone in the Late Carboniferous. In the Permian, the presence of reddish clastic deposits, anhydrites and evaporites in the eastern part suggests sedimentation in a northerly arid zone. The Gondwanan v. Baltican affinities of the Moesian Terrane are a matter of discussion, because of controversial data from Romania. The palaeogeographical distribution of Cambrian trilobites and shelly fauna is shown to be of mixed affinities with Avalonia, Bohemia and Baltica (Iordan 1992). Lower Devonian chitinozoans of East Moesia and possibly West Moesia show Northern Gondwana affinities (Vaida & Verniers 2005). On the other hand, Seghedi et al. (2004) interpreted the Eifelian of the Moesian Terrane as part of Laurussia. Obviously, further palaeobiogeographical studies on both benthic and planktonic fossils are necessary to confirm the origin of the Moesian Terrane.

Balkan Terrane S t r a t i g r a p h y and s e d i m e n t o l o g y

Within the Balkan Terrane, two distinct areas of specific stratigraphical and sedimentological development can be recognized: the West Balkan

Mountains and the Kraishte region. In addition, allochthonous low-grade metamorphic Palaeozoic rocks occur in the Shipka part of Central Balkanides and in the Strandzhides. Western Balkanides. In the Balkan Terrane, an island-arc association of cumulates, dykes and pillow lavas metamorphosed to greenschistfacies outcrops in the Western Balkan Mts. Recent isotope-geochronological dating of the ophiolites indicates an age of 563 Ma, confirming a Early Cambrian or Late Proterozoic age of the island arc (Von Quadt et al. 1998; Carrigan et al. 2003). These ages are similar to Pan-African ages and provide further evidence of a Gondwana origin of the Balkan Terrane. The island-arc complex is transgressively and unconformably overlain by an Arenig olistostromal sequence. Non-metamorphic Middle and Upper Ordovician shales and sandstones with brachiopods and trilobites, in total 2000 m thick, cover the olistostrome sequence. Upper Ordovician glaciomarine diamictites possibly relate to emergence as a result of glaciation (Gutierrez-Marco et al. 2003). Following a continuous transition from the Ordovician, the Silurian sequence represents a pelagic pelitic succession of 300 m lydites, black graptolitic shales and laminated shales-siltstones dated graptolites (Sachanski 1993; Sachanski & Tenchov 1993). The succession of established graptolite zones proves a complete Silurian section and transitional sedimentation across the Silurian-Devonian boundary (Sachanski 1998). An outcropping 1500 m Devonian succession of shales and siltstones with scarce tentaculites, graptolites and chitinozoans (Lower and Middle Devonian), silicites, siliciclastic 'pre-flysch' alternations of shales and lydites (Middle Devonian) is followed by thick flysch deposits with macroflora of Late Devonian to Vis6an age. Whereas Lochkovian, Pragian and Emsian rocks were proved by means of fossils, the assignment of the siliciclastic 'pre-flysch' alternation to a MidDevonian age is based only on its stratigraphical position. The development of flysch sedimentation in a progressively subsiding basin occurred between the Late Devonian and the Vis6an (Yanev 2000). Age determination is based of macroflora and on conodonts in single carbonate layers (Boncheva & Yanev 1993). The continental cover consists of Upper Carboniferous and Permian sediments and pyroclastic rocks, and overlies variegated sedimentary and metamorphic rocks of different ages. Namurian-Westphalian and Stephanian coal-bearing deposits rich in macroflora crop out

TERRANES OF BULGARIA AND NW TURKEY in isolated basins. Permian reddish siliciclastic rocks 0-3000 m thick accumulated over folded basement including the Upper Carboniferous sedimentary, volcanic and intrusive rocks. Kraishte region. The oldest Palaeozoic sedimentary rocks in the Kraishte region are Silurian black shales with lydites at the base. The age was proved, using graptolites (Spassov 1963, 1964), as Late Silurian and Early Devonian. The Upper Silurian and Lower Devonian sequence is developed in continuous shaly-carbonate sedimentation. In the central and southwestern parts of Kraishte biogenic limestones are dated, using conodonts and tentaculites (Boncheva 1991; Sachanski & Boncheva 1994), as Lochkovian to Eifelian and Frasnian-Famennian (Spassov 1973). The Lower Devonian sedimentation is a non-rhythmic succession of limestones and shales, the shales being predominant. Characteristic of the Devonian 'pre-flysch' sedimentation is the occurrence of thick folded lydite packets. Olistostromes of Lochkovian and Pragian limestones (Boncheva 1991) occur in the Middle Devonian-Lower Carboniferous and the Upper Jurassic-Lower Cretaceous flysch. The total thickness of the Silurian and Devonian units is hard to estimate because of tectonic displacement and lack of outcrops. The Middle Devonian to Visran mainly turbiditic succession about 1500 m thick is represented by clastic rocks with some carbonate and lydites in the upper part (Yanev 1985; Yanev & Spassov 1985). Upper Carboniferous and Lower Permian units are missing. The continental cover is of Upper Permian sandstones, siltstones and scarce breccias-conglomerates about 300-400 m thick. Palaeozoic succession of Shipka part of Balkanides and Strandzhides. In the Shipka part of the Balkan Mountains several Alpine tectonic slices consist of disturbed Riphean-Cambrian to Devonian low-grade metamorphic rocks that contain an incomplete stratigraphic column. The generalized Palaeozoic section consists of a Riphean-Cambrian metasedimentary formation, an Ordovician quartzite-shale formation, an Upper Silurian-Middle Devonian limestoneshale formation and an Upper Devonian rhythmic flysch sandstone-shale formation (Yanev et al. 1995). There are scarce fossil data only from the limestone-shale formation. Several crinoidbearing horizons in the limestones were proved to be Devonian using crinoids (Kalvacheva & Prokop 1988) and conodonts provide data on the Early Devonian (Yanev et al. 1995).

57

In the Strandzhides, metamorphic rocks up to greenschist facies of probable Palaeozoic age occur as allochthonous units in several Alpine nappe structures. The Palaeozoic succession is overturned and thrust over the Triassic and Jurassic sequences. Three metasedimentary series are recognized (Maliakov 2003). The lower series consists of metaconglomerates, metasandstones, marbles and phyllites, and is more than 600 m thick. Above, metasandstones, phyllites, marbles and metadiabase crop out. The total thickness is 550 m. The conodont fauna from this series indicate an Early Devonian age (Boncheva & Chatalov 1998). This series is covered by 100 m of recrystallized limestones, 350m of black phyllites and 450 m of grey-green calc-phyllites. Palaeogeography Middle Ordovician benthic faunas of the Balkan Terrane in west Bulgaria and eastern Serbia are of Bohemian and North African affinities (Guttierez-Marco et al. 2003). The Emsian chitinozoans are of clear Gondwanan affinities. The Carboniferous macroflora (Cyclostigma) is characteristic of the humid zone. Palaeoclimatic interpretations for the Ordovician are based on Fe-oolitic rocks and diamictites, which suggest a depositional environment in the higher latitude humid zone at about 40 ~ S. The Llandovery post-glacial graptolitic black shales were possibly deposited in the cool temperate zone (Yanev 1997). The abundance of diverse macroflora and coal deposition in the Late Carboniferous is characteristic of the equatorial humid zone. The presence of anhydrite matrix in the reddish Permian clastic rocks indicates deposition in an arid climatic zone. Thus, palaeoclimatic interpretations may support a northward migration from a temperate latitude in the Ordovician to the equator in the Permian. Palaeomagnetic data are available for the Balkan Terrane in Serbia (Milicevi6 1993, 1994). They indicate a position between 50 ~ and 29 ~ S during the Tremadoc, of 30~ ~ S in the MidOrdovician and 38~ in the Late Ordovician. In the Early Devonian, the Kucaj Terrane in Eastern Serbian (considered to have the same sedimentary development as the Balkan Terrane) was located at about 16~ S. In the Permian, palaeomagnetic data suggest that the position of the Balkan Terrane was at 8-14~ N (Nozharov et al. 1980; Milicevi6 1993). However, palaeomagnetic data are better interpreted when combined with palaeoclimatical and palaeofaunal evidence.

58

S. YANEV ET AL.

Istanbul Terrane S t r a t i g r a p h y and sedimentology

The crystalline basement of the Istanbul Terrane is represented by a structural complex including fragments of meta-ophiolites, island-arc volcanic rocks and arc-type granitoids, together with pieces of a continental crust of unknown affinity (G6ncfio~lu 1997; Usta6mer & Rogers 1999; Yi~itba~ et al. 2004). Recently, Usta6mer et al. (2005) determined a new U - P b zircon age of 571-579 Ma from the arc-type granitoids in the Bolu Massif (Fig. 3). The lowermost unit of the Palaeozoic succession in the Istanbul area comprises almost 1500 m of fine-laminated siliceous shales with sandy interlayers in its upper part. It is conformably overlain by 750 m of thin- to mediumbedded greenish sandstones alternating with thin-bedded, laminated shales (Gedik et al. 2002). None of these formations has yielded fossils, so that an Early Ordovician age assigned to them is arbitrary. A formation of almost 1000m thickness of variegated conglomerates, conglomeratic sandstones, arkosic sandstones and pink shales unconformably overlies the earlier formations. These continental clastic rocks are transgressively covered by 50-100 m of quartzarenites and quartzites with conglomeratic intervals. The quartzites do not include any fossils but contain undetermined traces (?Crusiana) and vertical vermes tubes (Onalan 1982). Upwards, the quartzites are transitional to a succession with greenish shales, siltstones and sandstones in the lower part and violet-grey and green mudstones with carbonate-rich lenses with brachiopods. Reddish-black bands with oolitic and nodular chamosite and hematite occur both in lower and upper parts of the succession. The thickness varies between 250 and 750 m. The chamositic bands in the lower part yielded early Late Ordovician brachiopods (Sayar 1984) followed by sandstones with Late Ordovician (Villas, pers. comm.) brachiopods. The carbonate-rich upper part includes Telychian brachiopods and conodonts (Haas 1968). No glacio-marine rocks have been observed at this interval as yet. The uppermost part of this formation includes a 70 m band with oolitic chamosites and limestones with conodonts characteristic of the Wenlock. Upwards, 100 m of sparry, compact and laminated limestones follow, known as 'Halysites Limestones'. These limestones include corals in addition to brachiopods, cephalopods and crinoids. Conodont findings indicate a Late Wenlock to Late Ludlow age. The following 300 m of the succession is characterized from

bottom to the top by neritic limestones. The lower part comprises grey to pink stromatolitic limestones, followed by dark grey to black limestones and dolomites. The upper part of the succession is represented by nodular limestones with marly interlayers. This carbonate succession is dated on the basis of brachiopods, corals and conodonts (e.g. Haas 1968), and includes without a significant break the whole Pridoli to Early Emsian succession. The carbonates are transitional to an almost 800 m alternation of clayey sandstones, limy greywackes and discontinuous bands of limestones, rich in brachiopods, corals, goniatites, bivalves and trilobites, that indicate continuous deposition between late Emsian and early Eifelian. The carbonate succession above consists of limestones of mid-late Eifelian age and nodular limestones with chert bands. Brachiopod and conodont findings from this 'lower nodular facies' indicate a Givetian to early Frasnian age. The following grey to brown silicified shales and cherts with violet nodular limestone and chert intervals, almost 100m thick, include late Frasnian conodonts and are transitional to 'upper nodular facies', a 75-80 m thick band with nodular limestones and lydite bands. The lower part of this unit includes Famennian conodonts (~apkino~lu 2000), whereas the uppermost layers are mid-Tournaisian in age (G6ncfio[glu et al. 2004). After an intervening unit of black lydite with phosphate nodules, late mid-Tournaisian in age (Gedik et al. 2003), the succession passes into proximal turbidites with plant remains and olistostromal sandstone-conglomerate bands, very rich in detrital white mica and clasts of felsic igneous rocks. This unit is traditionally known as the 'Variscan flysch' in the Istanbul area and is more than 2500 m thick. The flora obtained from the lower half of the formation is Vis6an in age (Baykal 1963). The youngest foraminifer age is from reefal limestones within the greywackes and is Late Vis6an. The Palaeozoic rocks of Istanbul are intensively deformed and intruded by Late Permian granitoids (e.g. G6riir et al. 1997). The lower part of the unconformably overlying red continental clastic rocks has not yet yielded any fossils. However, the middle part is Late Permian in age, so that the orogenic event responsible for the deformation should be of Carboniferous to Permian age. Palaeogeography

No palaeomagnetic data are available from the N W Anatolian Palaeozoic and the palaeogeographical interpretations are mainly based

TERRANES OF BULGARIA AND NW TURKEY on biogeographical and palaeoclimatic data. The Ordovician to Silurian benthic faunas of the Istanbul Terrane are of Avalonian and Podolian affinities, as mentioned by Haas (1968). Starting with the Devonian (Emsian and throughout Frasnian), however, brachiopods and trilobites are of clear Bohemian and North African (Morocco) affinities. This affiliation is further supported by Emsian ostracodes indicating faunal relations to Thuringia and Morocco (Dojen et al. 2004). The Carboniferous macroflora (Cyclostigma) is also found in Bulgaria and Central Europe. The Late Vis6an foraminiferal assemblage and the Early Carboniferous development, on the other hand, have been correlated with the Moravo-Silesian (Brunovistulian, Kalvoda et al. 2003) zone. As in the case of the Balkan Terrane, Ordovician siliciclastic rocks comprise Fe-oolitic or chamositic sequences, suggesting deposition in a temperate humid zone at about 40 ~ S. The dominance of reefal limestones during the Devonian as well as the presence of diverse macroflora in the Early Carboniferous is characteristic of the equatorial humid zone, so that, as for the Balkan Terrane, a migration of the Istanbul Terrane from temperate latitudes in the Ordovician to near the equator in Late Palaeozoic times can be assumed.

Zonguldak Terrane Stratigraphy and sedimentology

To the east of Istanbul, a number of isolated Palaeozoic successions crop out within the Alpine tectonic units (Fig. 3). G6nciio~lu & Kozur (1998, 1999), Kozur & G6ncfioglu (2000) and Von Raumer et al. (2003) suggested that they represent a distinct terrane (Zonguldak Terrane, G6nciio(glu & Kozur 1999), separate from the Istanbul Terrane. The rationale for this suggestion is that their stratigraphy, starting with lower Middle Ordovician, is completely different from that of the Istanbul Terrane (Fig. 5) and that these differences cannot be explained simply by lateral facies changes. Moreover, a late Early Devonian regional angular unconformity in the Zonguldak terrane together with an accompanying thermal event (Kozur & G6ncfio~lu 2000) contrasts with continuous platform-type deposition in the Istanbul Terrane during the same time interval. The basement of this terrane occurs in the Karadere area, where Chen et al. (2002) dated the tonalitic and granodioritic rocks to 570 and 590 Ma using the U - P b zircon method. Thus, the

59

basement rocks of the Zonguldak and Istanbul Terranes are both related to the Cadomian magrnatism, characteristic of Gondwanan or peri-Gondwanan terranes in Central and Southern Europe. This basement is unconformably overlain by siliciclastic rocks, commencing with Tremadoc shales, followed by a series of laminated shales and siltstones. A 700 m thick quartzite unit with conglomeratic interlayers is conformably covered by black shales with rare limestone layers. Graptolite, acritarch and conodont data from this succession indicate that the succession includes the time-span Early Arenig to MidLudlow (Dean et al. 1997, 2000). Upper Silurian (Pridolian) and Lower Devonian (up to Pragian) rocks are missing. The unconformably overlying succession is of quartzites and oolitic chamosites with a thick packet of carbonates. The lower part of this unit is very rich in neritic fossils and includes Pragian palynomorphs and conodonts. The onset of carbonate deposition here is late Emsian, and it terminated in the late Vis6an (Dil & Konyali 1978). The thickness of this shallow-marine limestone-dolomite succession reaches 1200 m. In contrast to the Istanbul Terrane, the carbonates display typical features of reef, lagoon and restricted shelf deposition, and are very rich in corals, brachiopods, bivalves and foraminifers especially in the upper part. This carbonate succession is conformably overlain by shallow-marine sandstones with brachiopods, corals and land plants. Carbonate lenses within them yielded early Serpukhovian conodonts (G6ncfio~lu et al. 2004). Upwards, the succession is characterized by a regressive series that grades into floodplain deposits with numerous coal seams of Westphalian age (Kerey 1984). The youngest age obtained from the plants within this 700-1200 m succession of these continental clastic rocks in the Zonguldak area is Stephanian. The Carboniferous strata here are only slightly deformed and unconformably overlain by Permo-Carboniferous continental clastic rocks. Palaeogeography

The Tremadoc acritarchs in the Karadere area are known from localities in Avalonia, Baltica and Gondwana, and hence are not indicative for palaeogeographical interpretations. Dean et al. (1997) suggested that the Late Ordovician and Silurian fauna were of mainly Avalonian affinities. The Devonian benthic fauna of this zone, on the other hand, is typical of the Rhenohercynian

60

S. YANEV E T AL.

Fig. 5. Generalized stratigraphy of the Palaeozoic in the Istanbul and Zonguldak Terranes in NW Anatolia, Turkey.

TERRANES OF BULGARIA AND NW TURKEY in Central Europe and SW England (Tokay 1955). The Late Namurian-Westphalian sediments, fauna and flora in the Zonguldak Coal Basin correlate very well with Moesia, Balkan and other Upper Carboniferous coal basins in Europe deposited under tropical conditions.

Discussion The brief review of the recent data given above may help to answer the following questions regarding the palaeogeographical setting of the Balkan and N W Anatolian terranes: How do the Palaeozoic terranes of Bulgaria and N W Anatolia correlate with each other? Were these terranes part of Baltica, Avalonia or the Armorican Terrane Assemblage? What was the location of the Bulgarian and N W Anatolian terranes with regard to the Variscan suture zones?

How do the Palaeozoic terranes o f Bulgaria and N W A n a t o l i a correlate with each other? Two of the terranes described, the Balkan and Istanbul Terranes, show striking similarities in their Ordovician to Carboniferous sedimentary development, which may imply their common terrane affinities and origin. That these two terranes shared the same depositional environments between the Ordovician and Eifelian is expressed in the development of very similar sedimentary successions: shallowwater siliciclastic deposits with brachiopods in the Ordovician, mainly deeper water black shales with graptolites in the Silurian, an alternation of shales and limestones across the SilurianDevonian boundary, and predominantly carbonates in the Lower Devonian shales, with carbonate or lydite in the Eifelian. However, some differences as a result ofbathymetric conditions and local palaeo-relief exist, such as reefal limestone bodies in the Middle Silurian rocks of the Istanbul Terrane, compared with the shaly sedimentation in the Balkan Terrane. After the Givetian, flysch accumulation started in the Balkan Terrane, in contrast to the shallowmarine, chiefly carbonate, sedimentation in the Istanbul Terrane. These contrasting depositional environments, caused by tectonic activity, existed laterally and persisted until the Visran. The Lower Carboniferous flysch in the Istanbul Terrane developed later than the flysch sedimentation in the Balkan Terrane where it started in the Givetian-Frasnian, whereas during the Late Carboniferous, continental deposits with coal formed in the Balkan Terrane; in the Istanbul

61

Terrane no Carboniferous deposits younger than Visran are preserved. An excellent stratigraphical correlation is possible only between the East Moesian and Zonguldak Terranes for the Mid-DevonianCarboniferous interval. Regarding the pre-MidDevonian, concerning striking features of the Zonguldak Terrane such as the deposition of graptolitic shales with pelagic carbonates in Mid-Ordovician to early Late Silurian and the early Mid-Devonian unconformity, these are not common features of all the continental microplates ascribed to Moesia. However, in East Moesia as well as in Dobrudgea similar occurrences were reported (e.g. Seghedi et al. 2004) within tectonic intercaletions of the Alpine belt. It is important to note that the Silurian and Devonian chitinozoans in East Moesia are of North Gondwanan affinity (Vaida & Verniers 2005).

Were these terranes part o f Baltica or peri-Gondwana (Avalonia and the Armorican Terrane Assemblage) ? The question refers to the classical approach that considers the Moesian, Istanbul and Zonguldak Terranes as part of the Eastern European Craton (or Baltica) throughout their geological history (e.g. Grrfir et al. 1997; Von Raumer et al. 2002; Kalvoda et al. 2003). Two lines of evidence are against such an interpretation: the Cadomian affinity of the oceanic lithosphere and the palaeobiogeographical provinciality based on benthic faunas. Both the Eastern Balkan and N W Anatolian terranes are characterized by the presence of Cadomian oceanic lithosphere and arc-type magmatism that lasted until the Early Cambrian. The oldest sedimentary cover of these crustal pieces is Early Ordovician, which would indicate that their amalgamation and hence deformation would have lasted during the Cambrian time. Thus, their Cadomian affinity would imply that they were originally part of Gondwana. The Ordovician trilobite fauna in the Zonguldak Terrane is more akin to that of south Wales (Avalonia) and Bohemia than Baltica (Dean et al. 1997, 2000). Consequently, it is unlikely that the Balkan and N W Anatolian Terranes were parts of Baltica. The Avalonian Terrane is characterized by Pan-African (Cadomian) events of Late Proterozoic age, deposition of siliciclastic rocks during the early Ordovician, and deformation, magmatism and metamorphism related to the 'Caledonian' orogeny as a result of either the

62

S. YANEV E T AL.

Fig. 6. Schematic map of the relationships between the studied terranes.

Late Ordovician collision with Baltica or the subsequent Late Silurian accretion with Laurentia. All these geological events can be used as important geological criteria to identify the Avalonian terranes. Additionally, distinct zones of faunal provinciality, mainly controlled by global palaeoclimate established for the Ordovician-Silurian period (e.g. Cocks 2001) and the deposition of glacier-related sediments in Gondwana or periGondwana during the end-Ordovician could also be used for palaeogeographical interpretations. The Ordovician in the Zonguldak Terrane contains trilobites of clear Avalonian (Wales) affinities (Dean et al. 2000). The Devonian benthic fauna are the same as in the Rhenohercynian zone (i.e. Avalonian or Armorican Terranes). In the Balkan Terrane, Mid-Ordovician benthic faunas (trilobites and brachiopods) were found (Gutierrez-Marco et al. 2003) that are of North African affinities. The planktonic fossils (chitinozoans and acritarchs) of the Early Devonian in the Moesian Terrane indicate the high latitude of the Armorican Terrane Assemblage and not the low latitude of Baltica (Lakova 1995, 2001b). From the studied terranes, only the Balkan Terrane includes diamictites within the uppermost Ordovician strata representing very important palaeogeographical evidence that it was not part of Avalonia but of Gondwana or Armorica. During the Ordovician and Silurian

the terranes studied either comprise abundant organic matter in predominantly shaly sequences or include Fe-oolitic minerals, evidence for deposition in a temperate humid zone. During the Mid-Devonian and Carboniferous, the fauna and flora of the Bulgarian and Turkish terranes suggest a depositional migration from the southern arid zone to the equator. The palaeomagnetic data for the Balkan Terrane in Serbia further suggest a movement of the terrane from a southern subpolar latitude in the Ordovician to the equator in the Permian. Even if there are some faunal links to Avalonia, the absence of Shelveian (late Ordovician) and/or Scandian (late Silurian) events in the West Moesia, Balkan and Istanbul Terranes opposes a link with Avalonia. The Zonguldak Terrane and some continental microplates in East Moesia, on the other hand, may have been located in the eastern continuation of Avalonian and Moravo-Silesian terranes. This is due to the fact that especially the Zonguldak Terrane displays a key unconformity of late Early Devonian age, which may correspond to the Acadian event also known in the southern periphery of Avalonia (Pharaoh 1999). Taking into account the generalized stratigraphical column, the occurrence of palaeoclimatological indicators in the sediments, and the palaeobiogeographical affinities of the benthic and planktonic fauna, it seems very probable that the West Moesian, Balkan and Istanbul

TERRANES OF BULGARIA AND NW TURKEY Terranes were more closely linked to the Armorican Terrane Assemblage that includes the Bohemian and Saxo-Thurungian terranes in Europe.

What was the location o f the Bulgarian and N W Anatolian Terranes with respect to the Variscan suture zones? The accretion of Gondwana-derived crustal blocks to Laurussia has resulted in the formation of a distinct orogenic belt: the Variscan Zone. Geodynamic reconstructions (e.g. Franke 2000; Neubauer 2003; Von Raumer 2003) suggest a very complex network with numerous crustal blocks within the Variscan Zone. Obviously, there were several oceanic seaways (e.g. Rheic Ocean, Rhenohercynian Ocean, SaxoThuringian Ocean, Palaeotethyan Ocean, etc.) that separated the terranes or terrane assemblages. Of the terranes studied here, only the Zonguldak Terrane includes evidence for a late Early Devonian deformation. This event is frequently observed in the Avalonia-related terranes in central Europe and attributed to the docking of Armorican terranes to Laurussia by the closure of the Rheic Ocean. If this interpretation is confirmed by additional data, the Zonguldak Terrane can be positioned at the eastern edge of the Moravo-Silesian terranes to the south of Laurussia during this period. The uplift and the closure of the Palaeozoic basin during the Late Stephanian was accompanied by weak deformation, but no distinct Variscan metamorphic event is recorded in the basement of the Zonguldak Terrane (Chen et al. 2002). The Moesian Terrane has not been affected by the closure of the Rheic Ocean and its docking to Baltica should be somewhat later, between the Late Devonian and Early Carboniferous Variscan convergence. The striking lithological, faunal and floral similarities in the Tournaisian to Stephanian successions in Zonguldak, Moesia, Donetz, Silesia, the Ruhr, Belgium and Wales can be attributed to their common palaeogeographical location to the north of the Rhenohercynian margin. Considering the general sedimentological development from the Ordovician to the Late Devonian-Early Carboniferous in the Balkan and the Istanbul Terranes and their correlation with the Saxo-Thuringian or Moldanubian zones of Central Europe, their most probable position was to the south of the Rhenohercynian suture. In the Bulgarian and N W Anatolian realm the terrane boundaries are covered by MesozoicTertiary successions and complicated by Cimmerian and Alpine deformations. Hence, there

63

are no surface or subsurface data to locate exactly the sutures between these terranes. Moreover, no ophiolite-bearing subduction-accretion prisms of Palaeozoic age have yet been identified along the terrane boundaries. The only ophiolitic material between the Moesian, Balkan and Thracian Terranes in Bulgaria has been proven (e.g. Haydutov & Yanev 1997) to be of PanAfrican age. The absence of ophiolitic material suggests that the terranes may have been juxtaposed by wrench-faulting (Kerey 1984) or oblique docking (G6nciio(glu 1997).

Conclusions Existing data on the Palaeozoic rocks in the eastern part of the Variscan Suture Zone need to be enhanced by further detailed palaeomagnetic and geophysical data, especially in the Turkish part. However, the available data provide a solid starting point for a preliminary geodynamic interpretation. This interpretation is mainly based on stratigraphical, sedimentological, palaeofaunal and biogeographical data for palaeogeography and basin development. The present data support Yanev's hypothesis of the peri-Gondwanan origin of the Moesian Terrane, its northward migration between Ordovician and Devonian time, the lack of a Scandian unconformity, drifting to the subequatorial arid zone in Late Devonian-Early Carboniferous time and accretion to Baltica in the Carboniferous. On the other hand, the Balkan Terrane, also of peri-Gondwanan origin, is very similar to the Saxo-Thuringian Zone and belongs to the late Palaeozoic accreted terranes south of the Rheic Suture. The accretion of the Balkan Terrane to Moesia-Baltica postdates the Early Carboniferous and continued during the Late Carboniferous and Permian. The collision between the terranes was not a coeval event but a polyphase process. For the Zonguldak and Istanbul Terranes in N W Anatolia, the Late Pan-African-Cadomian crystalline basement and the fossil provinciality for the Early Palaeozoic are considered as important evidence for their peri-Gondwanan origin. After drifting across the Rheic Ocean, the Zonguldak Terrane probably collided with Baltica during the Early Devonian and the Istanbul Terrane accreted to the northern palaeocontinent in the Serpukhovian. As no Palaeozoic oceanic lithologies have yet been identified, their accretion during the Variscan convergence may have involved strike-slip tectonics. This paper is a contribution to the BAS-T[)BITAK Joint Project Number 102Y157 and the Bulgarian

64

S. YANEV ET A L

National Fund Projects NZ 1001/01, 1401/04 and 1404/04, and the authors acknowledge the contributions of both organizations. The editors of this Special Publication, A. H. F. Robertson (Edinburgh) and D. Mountrakis (Thessaloniki), and the referees J. A. Winchester and T. Usta6mer, are gratefully acknowledged for their comments. This paper is a contribution to IGCP Projects 497 and 499.

References BAYKAL, M. F. 1963. Geological stu@ of the area to the west of Bosphorus. Mineral Research and Exploration Open File Report, 3267 (in Turkish). BERTHELSEN, A. 1993. Where different geological philosophies meet: the Trans-European Suture Zone. Publications of the Institute of Geophysics, Polish Academy of Sciences, 255(A20), 19-31. BONCHEVA, I. 1991. Conodont biostratigraphy of the Lower Devonian from Southwest Bulgaria. Geologica Balcanica, 21(4), 55-72. BONCHEVA, I. 1995. Conodont biostratigraphy of the Middle Devonian in North Bulgaria. Review of the Bulgarian Geological Society, 56(3), 35-46. BONCHEVA, I. & CHATALOV, G. 1998. Palaeozoic conodonts from the Dervent Heights and the Stradza Mountain. SE Bulgaria. Comptes Rendus de l'Acaddmie Bulgare des Sciences, 51(7-8), 45-48. BONCHEVA, I. & YANEV, S. 1993. New data on the Paleozoic flysch of the Sofijska Stara Planina Mountain. Geologica Balcanica, 23(5), 15-22. BONCHEVA, I., DIMITROVA, T. & LAKOVA, I. 1994. Devonian and Carboniferous conodonts and palynomorphs from the wells C-11 and P-120 Ograzden, Northeast Bulgaria. Review of the Bulgarian Geological Society, 55(3), 55-63. BONCHEVA, I., DIMITROVA, T. & YANEV, S. 2000. Stratigraphic, lithologic and palaeoecologic studies based on conodont fauna and microflora from the Middle and Upper Devonian series in the section of P-1 Vaklino, Northeastern Bulgaria. Review of the Bulgarian Geological Society, 61(1-3), 27-34. BONCHEVA, 1., SARMIENTO,G. N. & YANEV, S. 2002. Conodont colour alteration index and thermal maturation in Devonian and Carboniferous sediments of Northwestern Bulgaria. Revista Espanola de Micropaleontologia, 34(2), 117-128. ~APKINOdLU, S. 2000. Late Devonian (Famennian) conodonts from Denizlikoyu, Gebze, Kocaeli, northwestern Turkey. Turkish Journal of Earth Sciences, 9, 91-112. CARRIGAN, C. W., MUKASA, S. B., HAYDUTOV, I. & KOLCHEVA, K. 2003. Ion microprobe U-Pb zircon ages of the pre-Alpine rocks in the Balkan, Sredna Gora and Rhodope terranes of Bulgaria: constraints on Neoproterozoic and Variscan evolution. Journal of the Czech Geological Society, 48(1-2), 32-33. CHEN, F., SIEBEL,W., SATIR, M. & TERZIOGLU, M. N. 2002. Geochronology of the Karadere basement (NW Turkey) and implications for the geological evolution of the Istanbul zone. International Journal of Earth Sciences, 91,469-481.

COCKS, L. R. M. & TORSVIK,T. H. 2002. Earth geography from 500 to 400 million years ago: a faunal and palaeomagnetic review. Journal of the Geological Society, London, 159, 631-644. DEAN, W. T., MARTIN, F., MONOD, O., DEMIR, O., RICHARDS, R. B., BULTYNCK,P. & BOZDOGAN,N. 1997. Lower Palaeozoic stratigraphy, KaradereZirze area, Central Pontides, N Turkey. In: GONCOO(3LU, M. C. & DERMAN, A. S. (eds) Early Palaeozoic Evolution in N W Gondwana. Turkish Association of Petroleum Geologists, Special Publication, 3, 32-38. DEAN, W. T., MONOD, O., RICHARDS,R. B., DEMIR, O. & BULTYNCK,P. 2000. Lower Palaeozoic stratigraphy and palaeontology, Karadere-Zirze area, Pontus Mountains, northern Turkey. Geological Magazine, 137, 555-582. DIE, N. & KONFAL1, Y. 1978. Carboniferous of Zonguldak area. In: Guide Book: FieM Excursions on the Carboniferous Stratigraphy in Turkey. MTA Publication. D1MITROVA, T. 1996. Early Carboniferous miospores from Novachene borehole, central North Bulgaria. Geologica Balcanica, 26(4), 41-49. DOJEN, C., (~)ZGOL, N., GONCOOdLU, M. C. & GONCOOGLU, Y. 2004. Thuringian ecotype early Devonian ostracods from NW Anatolia (Turkey). Neues Jahrbuch fiir Geologic and Palaeontologie, Monatshefte, 2002(12), 733-748. FRANKE, W. 2000. The mid-European segment of the Variscides: tectonostratigraphic units, terrane boundaries and plate tectonic extension. In: FRANKE, W., ALTHERR,R., HAAK,V., ONCKEN,O. & TANNER, D. (eds) Orogenic Processes: Quantification and Modelling in the Variscan Belt. Geological Society, London, Special Publications, 179, 35-61. GEDIK, I., TIMUR, E. & DURU, M., et al. 2002. Kocat6ngel ve Bakacak formations in the Istanbul succession. 55th Geological Congress of Turkey, Abstracts, 97-99. GEDIK, I., TIMUR, E. & DURU, M., et al. 2003. Geology of lstanbul and surroundings. Mineral Research and Exploration Open File Report (in Turkish). GONCOOGLU, M. C. 1997. Distribution of Lower Palaeozoic rocks in the Alpine terranes of Turkey: palaeogeographic constraints. In: GONCOO~LU, M. C. & DERMAN, A. S. (eds) Early Palaeozoic Evolution in N W Gondwana. Turkish Association of Petroleum Geologists, Special Publication, 3, 13-23. GONC00(~LU, M. C. 2001. From where did the NW Anatolian Palaeozoic terranes derive: a comparative study of Palaeozoic successions. ESF Europrobe Meeting, 30 September-2 October 2001, Ankara, Abstracts, 22-23. GONCOOdLU, M. C. & KozuR, H. W. 1998. Facial development and thermal alteration of Silurian rocks in Turkey. In: GUTIERREZ-MARCO, J. C. & RABANO, I. (eds) Proceedings, 1998 Silurian Field-Meeting. Temas Geologico-Mineros ITGE, 23, 87-90. GONCOO6LU, M. C. & KOZUR, H. W. 1999. Remarks on the pre-Variscan development in Turkey. In: LINNEMANN,U., HEUSE,T., FATKA,O., KRAFT, P.,

TERRANES OF BULGARIA AND NW TURKEY BROCKE, R. & ERDTMANN, B. T. (eds) Prevariscan Terrane Analyses of ' Gondwanean Europa'. Schriften des Staatlichen Museums, Mineralogie, Geologie, Dresden, 9, 137-138. GONCOOGLU, M. C., BONCHEVA,I. & GONCOOGLU, Y. 2005. First discovery of Middle Tournaisian conodonts in the griotte-type nodular pelagic limestones, Istanbul area, NW Turkey. Rivista Italiana di Paleontologia e Stratigrafia, 110(2), 431-439. GOROR, N., MONOD, O. & OKAY, A. I., et al. 1997. Palaeogeographic and tectonic position of the Carboniferous rocks of the western Pontides (Turkey) in the frame of the Variscan belt. Bulletin de la Socidt~ Gkologique de France, 168(2), 197-205. GUTTIEREZ-MARCO,J. C., YANEV, S. & SACHANSKI,V., et al. 2003. New biostratigraphical data from the Ordovician of Bulgaria. Serie Correlacion Geologica, 17, 79-85. HAAS, W. 1968. Das Alt-Palaeozoikum yon Bithynian. Neues Jahrbuch fiir Geologie and Palaeontologie, Abhandlungen, 131, 178-242. HAYDUTOV, I. 8r YANEV, S. 1996. The Proto-Moesian continent of the Balkan Peninsula--a periGondwanaland piece. Tectonophysics, 272, 303313. IORDAN, M. 1992. Biostratigraphic age indicators in the Lower Palaeozoic successions of the Moesian Platform of Romania. Geologica Carpathica, 43(4), 231-233. KALVODA, J. 2001. Upper Devonian-Lower Carboniferous foraminiferal palaeobiogeography and Perigondwana terranes at the Baltica-Gondwana interface. Geologica Carpathica, 52, 205-215. KALVODA,J., LEICHMANN,J., BABEK,O. & MELICHAR, R. 2003. Brunovistulian Terrane (Central Europe) and Istanbul Zone (NW Turkey): Late Proterozoic and Paleozoic tectonostratigraphic development and paleogeography. Geologica Carpathica, 54(3), 139-152. KEREY, I. E. 1984. Facies and tectonic setting of the upper Carboniferous rocks of NW Turkey. In: ROBERTSON, A. H. F. & DIXON, J. E. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 123-128. KOZUR, H. W. & GONCOOGLU, M. C. 2000. Mean features of the pre-Variscan development in Turkey. Acta Universitatis Carolinae--Geologica, 42, 459-464. LAKOVA, I. 1993. Biostratigraphy of Lochkovian chitinozoans from north Bulgaria. Special Papers in Palaeontology, 48, 37-44. LAKOVA, I. 1995. Palaeobiogeographical affinities of Pridolian and Lochkovian chitinozoans from North Bulgaria. Geologica Balcanica, 26(5-6), 23-28. LAKOVA, I. 2001a. Dispersed tubular structures and filaments from Upper Silurian-Middle Devonian marine deposits in North Bulgaria and Macedonia. Geologica Balcanica, 31(3-4), 29-42. LAKOVA, I. 200lb. Biostratigraphy and provincialism of Late Silurian-Early Devonian acritarchs and

65

prasinophytes from North Bulgaria. In: JANSEN, U., et al. (eds) 15th International Senckenberg Conference, Joint Meeting IGCP 421/SDS, Frankfurt am Main, May 2001, Abstracts, 58-59. MALIAKOV, Y. 2003. The problem 'Strandza'. Mining and Geology, 5, 21-27. MCKERROW, W. S. & SCOTESE, C. R. 1990. Palaeozoic Palaeogeography and Biogeography. Geological Society, London, Memoirs, 12. MILICEVI~, V. 1992. Palaeomagnetic study of Phanerozoic sedimentary rocks in Serbia. Comptes Rendus de la SociOtO Serbe Gkologique, Livre Jubilaire, 243-250 (in Serbian with English abstract). MIL~CEVIC, V. 1994. Preliminary palaeomagnetic results for Ordovician of Zvonyachka Banya, Dgerchek and Zrna reka (Eastern Serbia). Proceedings Geoinstitute, 29, 13-22 (in Serbian with English abstract). NEUBAUER, F. 2002. Evolution of late Neoproterozoic to early Paleozoic tectonic elements in Central and Southeast European Alpine mountain belts: review and synthesis. Tectonophysics, 352, 87-103. NIKOLOV, Z., POPOVA, K. & POPOV, A. 1990. Coalbearing Upper Carboniferous sediments in R-1 Novachene (Central North Bulgaria). Review of the Bulgarian Geological Society, 51(1), 38-47. NOZHAROV, P., PETKOV, N., YANEV, S., KROPACHEK, V., KRS, P. & PRUNER, P. 1980. A palaeomagnetic data and petromagnetic study of Upper Carboniferous, Permian and Triassic sediments, NW Bulgaria. Studia Geophysica Geodynamica, 24, 252-284. 0NALAN, M. 1982. Depositional environment of the Ordovician and Silurian successions in Istanbul. Istanbul University Earth Sciences Bulletin, 2(3-4), 161-177. PHARAOH, T. C. 1999. Palaeozoic terranes and their lithospheric boundaries within the TransEuropean Suture Zone (TESZ): a review. Tectonophysics, 314, 7-29. VON RAUMER, J. V., STAMPFLI, G. M., BOREL, G. & BUSSY, F. 2002. Organization of pre-Variscan basement areas at the north-Gondwanan margin. International of Journal Earth Sciences, 91, 35-52. VON RAUMER, J. V., STAMPFLI, G. M. • BUSSY, F. 2003. Gondwana-derived microcontinents--the constituents of the Variscan and Alpine collisional orogens. Tectonophysics, 365, 7-22. SACHANSKI, V. V. 1993. Boundaries of the Silurian System in Bulgaria. Geologica Balcanica, 23(1), 25-33. SACHANSKI, V. V. 1998. Ordovician, Silurian and Devonian graptolites from Bulgaria. In: GUTIERREZMARCO, J. C. 8r RABANO, I. (eds) Proceedings, 1998 Silurian Field-Meeting. Temas Geologico-Mineros ITGE, 23, 255-257. SACHANSKI, V. & BONCHEVA, I. 1994. Tentaculites from the type section of Vrabcha Formation (Lower Devonian), south-west Bulgaria. Review of the Bulgarian Geological Society, 55(3), 139-142 (in Bulgarian with English abstract). SACHANSKI, V. & TENCHOV, Y. 1993. Lithostratigraphical subdivision of the Silurian deposits

66

S. YANEV ET AL.

in the Svoge anticline. Review of the Bulgarian Geological Society, 54, 71-81 (in Bulgarian with English abstract). SAYAR, C. 1984. Ordovician brachiopods in Istanbul. Geological Society of Turkey Bulletin, 27, 99-109. SEGHEDI, A., VAIDA, M. & VERNIERS, J. 2004. Palaeozoic evolution of the Moesian Platform: an overview. In: Avalonia, Moesia, Symposium and Workshop, 9-11 October 2004, Ghent/Ronse, Abstracts Volume, 29. ~ENGOR, A. M. C., YILMAZ, Y. & SUNGURLU,O. 1984. Tectonics of the Mediterranean Cimmerides: nature and evolution of the western termination of Palaeo-Tethys. In: DIXON, J. E & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 77-112. SPASSOV, CH. 1963. Das Oberludlow mit Monograptus hercynicus und dessen Grenze mit dem Devon bei Stanjovci, Bezirk Pernik. Review of the Bulgarian Geological Society, 24(2), 119-142. SPASSOV, CH. 1964. Beitrag zur stratigraphie des Silurs und Devons in Kraiste. Review of the Bulgarian Geological Society, 25(3), 267-283. SPASSOV, CH. 1973. Stratigraphie des Devons in Sudwest-Bulgarien. Bulletin of the Geological Institute, Series Stratigraphy and Lithology, 22, 5-39. SPASSOV, CH. 1983. Biostratigraphy of Devonian in North Bulgaria. I. Upper Devonian conodonts. Paleontology, Stratigraphy and Lithology, 18, 3-24. SPASSOV, CH. 1987. The Devonian System in Bulgaria. In: FLUGEL, H. W., SASSI, F. P. & GRECULA, P. (eds) Pre-Variscan and Variscan Events in the A lpine-Mediteranean Mountain Belts. Mineralia Slovaca--Monograph, 435-444. SPASSOV, CH. & YANEV, S. 1966. Stratigraphy of the Palaeozoic sediments in drilling from N.E. Bulgaria. Bulletin of the Geological Institute, 15, 25-77. SPASSOV, Cn., TENCOV, J. & JANEV, S. 1978. Die palaozoischen Ablagerungen in Bulgarien. In: Ergebnisse der Osterreichischen Projecte des Internationalen Geologishen Korrelationsprogramms (IGCP) bis 1976. Springer, Berlin, 279-294. STAMPFLI, G. M. 2000. Tethyan oceans. In: BOZKURT, E., WINCHESTER,J. A. & PIPER, J. D. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 1-23. STEEMANS, P. & LAKOVA, I. 2004. The Moesian Terrane during the Lochkovian--a new palaeogeographic and phytogeographic hypothesis based on miospore assemblages. Palaeogeography, Palaeoclimatology, Palaeoecology, 208, 225-233. TOKAY, M. 1955. Gtologi6 de la region de Bartin (Zonguldak). Mineral Research and Exploration Bulletin, 46(47), 46-63. USTAOMER, P. A. & ROGERS, G. 1999. The Bolu Massif: remnant of a pre-Early Ordovician active margin in the west Pontides, northern Turkey. Geological Magazine, 136(5), 579-592.

USTAOMER, P. A., MUNDIL, R. & RENNE, P. R. 2005. U/Pb and Pb/Pb zircon ages for arc-related intrusions in the Bolu Massif (W Pontides, NW Turkey): evidence for Late Precambrian (Cadomian) age. Terra Nova, 17, 215-223. VAIDA, M. & VERNIERS, J. 2005. Biostratigraphy and palaeogeography of Lower Devonian chitinozoans from East and West Moesia, Romania. Geologica Belgica (in press). VDOVENKO, M. B., REITLINGER,E. A., IOVCHEVA,P. & SPASSOV, CH. 1981. Foraminifers in the Lower Carboniferous Deposits from Bore-Hole R-3, Gomotarci (Northwest Bulgaria). Paleontology, Stratigraphy and Lithology, 15, 3-50. VON QUADT, A., PEYTCHEVA, I. & HAYDOUTOV, I. 1998. U-Pb zircon dating of Tcherny Vrach metagabbro, the West Balkan, Bulgaria. Comptes Rendus de l'Acaddmie Bulgare des Sciences, 51 (1-2), 81-84. WINCHESTER, J. A. & PACE TMR Network Team 2002. Palaeozoic amalgamation of Central Europe: new results from recent geological and geophysical investigations. Tectonophysics, 360, 5-21. YANEV, S. 1985. Dessarrollo litofacial del Carbonifero en Bulgaria. DixiOme Congrks International de Stratigraphie et de Gdologie du Carboni~re, Madrid, 12-17 septembre 1983, Comptes Rendus, 3, 77-84. YANEV, S. 1990. On the peri-Gondwana origin of the Eo-Palaeozoic sediments in Bulgaria. In: SAVA~qIN, M. Y. & ERONAT, A. H. (eds) Proceedings, l l th Earth Science Congress Aegean Regions, 2, 334-344. YANEV, S. 1993. Gondwana Palaeozoic terranes in the Alpine collage system of the Balkans. Himalayan Geology, 4(2), 257-270. YANEV, S. 1997. Palaeozoic migration of terranes from the basement of the eastern part of the Balkan peninsula from peri-Gondwana to Laurussia. In: GONCUOGLU, M. C. & DERMAN, A. S. (eds) Early Palaeozoic Evolution in N W Gondwana. Turkish Association of Petroleum Geologists, Special Publication, 3, 89-100. YANEV, S. 2000. Palaeozoic terranes of the Balkan Peninsula in the framework of Pangea assembly. Palaeogeography, Palaeoclimatology, Palaeoecology, 161, 151-177. YANEV, S. & BONCHEVA, I. 1995. Contribution to the Paleozoic evolution of the recent Moesian platform. Geologica Balcanica, 25(5-6), 3-23. YANEV, S. & BONCHEVA, I. 1997. New data on the collision between peri-Gondwana Moesian terrane and Dobrudja periphery of PalaeoEurope. In: GONCI)O(3LU, M. C. & DERMAN, A. S. (eds) Early Palaeozoic Evolution in N W Gondwana. Turkish Association of Petroleum Geologists Special Publication, 3, 118-132. YANEV, S. & SPASSOV,CH. 1985. Lithostratigraphy of the Devonian flysch between Tran and Temelkovo. Paleontology, Stratigraphy, Lithology, 21, 82-96. YANEV, S., TZANKOV, T. & BONCHEVA, I. 1995. Lithostratigraphy and Late Alpine structure of the

TERRANES OF BULGARIA AND NW TURKEY Palaeozoic Terrains in the Shipka Part of Stara Planina Mountains. Geologica Balcanica, 25(2), 3-26. YI(~ITBA~, E., KERRICH, R., YILMAZ, Y., ELMAS, m. & XIE, Q. 2004. Characteristics and geochemistry of

67

Precambrian ophiolites and related volcanics from the Istanbul-Zonguldak Unit, Northwestern Anatolia, Turkey: following the missing chain of the Precambrian South European suture zone to the east. Precambrian Research, 132, 179-206.

The Carboniferous to Jurassic evolution of the pre-Alpine basement of Crete: constraints from U-Pb and U-(Th)-Pb dating of orthogneiss, fission-track dating of zircon, structural and petrological data S. S. R O M A N O 1, M . R. B R I X 2, W. D ( ) R R 1, J. F I A L A 3, E. K R E N N 4 & G. Z U L A U F 1

llnstitut fiir Geowissenschaften, Johann Wolfgang Goethe Universiti~t, Senckenberganlage 32-34, 60054 Frankfurt, Germany (e-mail." [email protected]) 21nstitut ffir Geologie, Mineralogie und Geophysik, Ruhr-Universitiit Bochum, Universitgitsstrasse 150, 44801 Bochum, Germany 3Czech Academy of Science, 16500 Praha 6, Suchdol, Czech Republic 4Institut fiir Mineralogie, Universit?it Salzburg, Hellbrunnerstrasse 34, A-5020 Salzburg, Austria The pre-Alpine evolution of the external Hellenides is poorly constrained because of the Alpine impact which largely erased the older orogenic imprints. Only a few outcrops with pre-Alpine basement exist, one of which is located in eastern Crete. The preAlpine basement, part of the Phyllite-Quartzite Unit, is composed of four sub-complexes, which are different in protolith age, type and age of metamorphism, and postmetamorphic cooling history. The lowermost, Kalavros crystalline complex (KCC) underwent Permian amphibolite-facies metamorphism related to top-to-the-NE shearing. The KCC exhibits a four-stage garnet zonation and a late, high-temperature event associated with the growth of K-feldspar. The KCC is overlain by the Myrsini crystalline complex (MCC), which underwent Carboniferous amphibolite facies metamorphism associated with top-tothe-north shearing. Late cooling of the MCC is documented by Jurassic fission track ages of zircon. The Chamezi crystalline complex underwent upper greenschist-facies metamorphism related to top-to-the-north shearing. In addition, the Vai crystalline complex, in an uncertain structural position, is characterized by Triassic emplacement of granite, followed by amphibolite-facies top-to-the-NW shearing and cooling, as is indicated by Jurassic fission-track ages of zircon. A preliminary tectonic model is presented, which invokes south-directed subduction, collision and accretion of the crystalline complexes to the northern margin of Gondwana.

Abstract:

The pre-Alpine basement of eastern Crete forms part of the Phyllite-Quartzite Unit (PQU), which underwent Alpine subduction and related highpressure-low-temperature metamorphism (Seidel et al. 1982; Theye & Seidel 1991; Theye et al. 1992). Previous investigations of this basement suggested the existence of two sub-complexes (the Chamezi and Myrsini crystalline complexes), which differ in the grade of pre-Alpine Barrovian-type metamorphism (Franz 1992). The 'age' of the pre-Alpine metamorphism has been constrained as Carboniferous by K - A r dating of mica and hornblende (Seidel 1978; Seidel et al. 1982). The K - A r ages are not very reliable, however, because of an Alpine overprint and associated pervasive fluid flow. Therefore,

the more robust U - P b systems of monazite and zircon have been investigated (Finger et al. 2002; Romano et al. 2004). The high closure temperature of the U - P b system of zircon ( > 9 0 0 ~ Cherniak & Watson 2000) and monazite (>725 ~ Parrish 1990) should exclude the influence of Alpine metamorphism (T c. 300+ 50 ~ on the isotopic systems. U - P b dating of zircon has indicated a Cambrian protolith age of the orthogneisses (Romano et al. 2004). Chemical dating of metamorphic monazite, by electron microprobe, yielded either Carboniferous or Permian ages for the metamorphism (Finger et al. 2002). New data on the U - P b system of zircon and rutile and fission-track ages of zircon, as

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 69-90. 0305-8719106/$15.00 9 The Geological Society of London 2006.

70

S.S. ROMANO E T AL.

Fig. 1. Geological map of eastern Crete after Zulauf et al. (2002). well as structural and petrological studies, can help to further distinguish the individual subunits. Because of the very low-grade metamorphic Alpine overprint, which is below the zircon fission-track annealing zone ( < 350 ~ Brix et al. 2002), the dating should reflect a lower temperature level of the pre-Alpine overprint. To further constrain the metamorphic temperatures and the type of deformation, microfabrics of quartz have been investigated.

Geological setting Crete forms a horst within a fore-arc region above an active northward-directed subduction

of the African plate (e.g. Jolivet et al. 1996). There is a lower and an upper nappe pile, which have been stacked together during Oligocene to Miocene convergence (Seidel et al. 1982; Jacobshagen 1986; Theye et al. 1992; Thomson et al. 1998). Parts of the lower nappes (Plattenkalk, Phyllite-Quartzite (PQU) and ?Tripolitza units) as well as the uppermost variable unit underwent Alpine deformation and metamorphism (Fassoulas et al. 1994; Jolivet et al. 1994; Kilias et al. 1994; Fassoulas 1999). The other nappes are not affected by Alpine subduction and thus are not metamorphic (Fig. 1). The Phyllite-Quartzite Unit of Crete consists of several slices, which change eastwards from

R A D I O M E T R I C D A T I N G OF C R E T A N B A S E M E N T

siliciclastic dominated to carbonate dominated. In eastern Crete, Carboniferous to Triassic rocks occur, which differ in age and composition (Krahl et al. 1983; Kozur & Krahl 1987; Fig. 2). A pre-Alpine basement is sliced within Carboniferous to Triassic phyllite-marble intercalations (below) and a Triassic carbonate-dominated sequence (above). Also, the anchimetamorphic Triassic carbonate-dominated sequence shows a decreasing metamorphic grade from bottom to top (Bonneau 1984; Krahl et al. 1986; Zulauf et al. 2002). The peak conditions of the Alpine metamorphism in eastern Crete were: 4.5-6.0 kbar and 250-310 ~ (Seidel et al. 1982; Franz 1992; Zulauf et al. 2002). As this temperature is below the zircon fission-track annealing zone, zircon fission-track ages do not reflect the Alpine overprint, but the origin of the source rocks of the metasedimentary rocks instead (Brix et al. 2002). Structural investigations of the Carboniferous to Triassic rocks have documented six deformation stages, related to Alpine subduction and subsequent exhumation (Zulauf et al. 2002). The pre-Alpine basement complex has been divided into two sub-complexes, which differ in grade and age of pre-Alpine metamorphism (Seidel et al. 1982; Franz 1992). The upper unit (the Chamezi crystalline complex, CCC) consists of micaschist, paragneiss and orthogneiss, which underwent upper greenschist-facies Barroviantype metamorphism (T=500-550 ~ P = 5 . 5 6.5 kbar; Franz (1992)). The lower unit consists of micaschist, paragneiss, orthogneiss, amphibolite, quartzite and marble, and is referred to as the Myrsini crystalline complex (Franz 1992). The latter is characterized by Barrovian-type amphibolite-facies metamorphism (T= 580630 ~ P=6.5-8.0 kbar; Franz (1992)). U - T h Pb electron microprobe dating of metamorphic monazite of this basement yielded both Carboniferous (c. 330 Ma) and Permian (c. 260 Ma) ages (Finger et al. 2002). Because of this difference in metamorphic age, the Myrsini crystalline complex of Franz (1992) has been subdivided into the Kalavros crystalline complex (KCC, with Permian metamorphism) and the Myrsini crystalline complex sensu stricto (s. s.) (MCC; with Carboniferous metamorphism). Orthogneisses of the CCC and MCC yielded Cambrian protolith ages of 511 _+16 Ma and 514+ 14 Ma (U-Pb on zircon; Romano et al. 2004). Apart from the basement complexes mentioned above a further crystalline complex with unknown age and grade of pre-Alpine metamorphism is present in the Val area.

71

Analytical procedure U - P b dating o f zircon a n d rutile

The separation of the density fraction (rutile and zircon) of the Sfakfi paragneiss was carried out at the Czech Academy of Science, Prague, whereas the Vai and Paraspori orthogneiss were prepared at Giessen University, Germany. The solution of zircon followed the system of Krogh (1982), whereas the solution of rutile followed the system for whole-rock analyses of Todt (1988). Further details on the analytical procedure have been given by D6rr et al. (2002a, b) and Romano et al. (2004). The applicable initial Pb ratios were: 2~176 2~176 =15.584 and 2~176 (Stacey & Kramers 1975). Fission-track dating o f zircon

One hundred zircons of the Vai (80-160 gin) and Paraspori orthogneiss (125-160~m) were separated for fission-track dating. Zircons were processed according to the techniques outlined by Hurford et al. (1991). The crystals were mounted in FEP-Teflon, polished, and etched in a K O H - N a O H eutectic melt at 217 _+4 ~ in steps varying between 1 and 4 h using a platinum crucible, until the majority of the grains was fully etched. Total etch times were between 10 and 12 h. Thermal neutron irradiation was performed in the TRIGA reactor of Oregon State University in Corvallis, USA with a neutron fluence of 1 x 10~5n cm -2. The samples were analysed using the external detector method (Naeser 1976; Gleadow 1981). The neutron fluence was monitored using uranium-doped Corning glasses CN-1 and CN-2. The muscovite detector micas were etched for 50 min in 40% HF at room temperature. Spontaneous and induced fission-track densities were counted on a Zeiss Axioplan optical microscope at 1250 x magnification with a 100 x oil immersion objective. Central ages (Galbraith & Laslett 1993) were calculated according to the Zeta-Calibration approach of Hurford & Green (1983). U - ( T h ) - P b dating o f m o n a z i t e using the electron microprobe

To constrain magmatic and metamorphic ages of the basement rocks, thin sections of gneisses and micaschists were investigated. Chemical analyses of monazite were obtained using a Jeol JX 8600 microprobe (Salzburg University, Austria). Analyses were carried out at 15 kV, 250 nA, and a beam size of e. 5 gm. Analytical details have been given by Finger et al. (2002).

72

S.S. ROMANO E T A L .

Fig. 2. Simplified tectonostratigraphic scheme of the Phyllite-Quarbzite Unit of eastern Crete. Stratigraphic ages: *Krahl et al. (1986); **Kozur & Krahl (1987); ***Haude (1989). Data for Alpine quartz recrystallization after Zulauf et al. (2002).

RADIOMETRIC DATING OF CRETAN BASEMENT

73

Fig. 3. Microfabrics of the rocks investigated. (a) Vai orthogneiss indicating top-to-the-NW shearing is shown by a plagioclase clast; crossed nicols, xz-section (sample 00240901). Co) Margarite (Mrg)-bearing micaschist of the CCC. Pseudomorphic growth of green biotite (Bt), muscovite (Ms) and albite (Ab) after garnet. Parallel nicols, xz-section (sample 02230404). (c) Nematoblastic microstructure of actinolite schist of the CCC overprinted by multiphase folding. Crenulation cleavage related to top-to-the-east shearing, as indicated by asymmetric stretching of the epidote blasts. Parallel nicols, xz-section (sample 00110904). (d) Quartz from cataclasite (see (f)) deformed by fracturing that is oriented parallel to the cracks. (e) Albite-bearing gneiss showing an older isoclinally folded quartz vein, which was refolded by Dc4 folds (sample 02240401). (f) Shear zone between Myrsini and Chamezi gneiss. Arrow points to the hammer for scale. Mineral abbreviations after Kretz (1983).

Chemical composition of garnet To obtain semiquantitative constraints on the metamorphic evolution of the basement, the chemical composition of 22 garnets of micaschist, paragneiss and amphibolite was investigated.

Given that temperatures were below those values at which solid-state diffusion is possible, the chemical zonation of garnet might reflect the metamorphic conditions during garnet growth and thus reflect the P-T path. The F e - M g ratio and the Fe-number Fe/(Fe + M g ) are sensitive

74

S . S . R O M A N O E T AL.

to the P-T evolution. Continued growth under increasing temperature is visible in a Mn-rich, to a Fe-rich, to a Mg-rich garnet. This 'theoretical' evolution depends on the presence of staurolite and biotite. In the staurolite zone increasing Fe and Mg and decreasing Mn and Ca contents developed. A bell-shaped zonation of Mn generally exists, because Mn is preferably incorporated into garnet (Hollister 1966; Atherton 1968). This zonation is especially visible in line-profiles, element distribution maps and spessartinepyrope-grossularite triplots (Spear 1993). A semiquantitative expression of pressure (Xc,) and temperature (XMg) evolution gives the Xc~ (Ca/ (Ca + Mg + Mn + Fe)) v. XMg (Mg/(Ca + Mg + Mn+Fe)) ratio of garnet (Miyashiro & Shido 1973; Martignole & Nantel 1982; Spear 1993). Line-profiles with 25-50 analyses were carried out using a JEOL Superprobe JXA-8200 at Erlangen University, Germany, under conditions of 20 nA and 15 kV. Ca (Ko0, Na (Ko0, Fe (K~x), Ti (K~x), Si (Kot), Cr (Ko0, AI (Ko0, and Mg (Ko 0 were measured for 20 s, whereas Mn (Ko0 was determined for 40 s. The calculation of garnet composition was based on 12 oxygens and eight cations. Element distribution maps of the Ca, Mg and Mn contents were obtained under conditions of 20 kV and 20 nA. The counting time per dot was 0.4 s.

O

.9

_o

8 0) O

8

O

..O

5

Structural investigations and microfabrics The structural and kinematic data presented result from c. 2000 measurements at 564 locations. The microfabrics were investigated using c. 400 thin sections. Results

U-Pb dating of zircon and rutile U-Pb analyses of zircon were carried out for the Vai orthogneiss, which is part of the Vai crystalline complex of easternmost Crete 9 The light grey to white protomylonitic orthogneiss (sample 00240901, Fig. 3a) is exposed along a path-cut NE of Toplofl monastery. The mylonitic foliation displays a NW-SE-trending stretching and mineral lineation defined by the shapepreferred orientation of stretched plagioclase and quartz. Porphyroclasts of magmatic K-feldspar show asymmetric pressure shadows of white mica, which have resulted from top-to-the-NW shearing. The orthogneiss contained about 5000, light pink milky, brown translucent to colourless long prismatic zircons. Many of the zircons show

8 O

o

+l

.~

eq

~o

~

r~

.N

RADIOMETRIC DATING OF CRETAN BASEMENT

Z) .Q n o oJ

Vai orthogneiss

~o 03

0.11

0.09

75

70~ 60 __/"J ~ / 5 8 5OO

767+_.23Ma

0.07

300

0.05

~

~

.....

1Ma

20zpb/2asu

0.4

0.6

0.8

1.0

Fig. 4. Concordia diagram for Vai orthogneiss. Numbering and error of ellipses of data points are given in Table 1.

21.8

;

21.6

/"

Sfakf paragneiss

.

21.4

21.2

~

21.0

20.8

/

90

Fig. 5. 238U/z~ v. z~176 points are given in Table 2.

94

'~.J

95

238U/2~

i

100

i

110

i

120

i

130

ratio of rutile of the Sfakd paragneiss. Numbering and error of ellipses of data

inclusions or microcracks. Several small zircons exhibit 'diamond-shining' and absorption seams. D-type zircons are common, whereas P~, Gt and S10 types were only rarely observed (Pupinclassification; Pupin 1980). The shape, colour and weight of the zircons are listed in Table 1. The analysed single grains and fractions of 2-10 zircons do not show any relation between the discordance and grain-size or colour (Fig. 4). The single grain analysis 58 shows the lowest

discordance. The distribution of the data in the concordia diagram suggests the existence of at least three zircon generations. A discordia with a lower intercept age at 223 _+ 11 Ma and an upper intercept age at 767 4-23 Ma is defined by analyses 55-58 (MSWD=0.52). The zircons consist of Neoproterozoic cores, which were overgrown by Triassic rims. The zircons 59 and 60 display apparent higher 2~176 ages of 678 M a and 898 Ma in spite of different 2~176 ratios of

76

S.S. ROMANO E T AL.

Table 2. U-Pb analytical results from rutiles of the

Sfakd paragneiss

o

V

radiogenic ratios Weight ~3~U] ~~ Sample (mg) ~~ _+2o(%) ~~ 94 95 96

4 98.02 0.47 11 117.89 1.38 7 124.52 0.46

m. ,-2.

az~

.,..~

_+2o(%)Cor.

20.979 0.193 21.239 0.207 21.580 0.222

+l

0.44 0.24 0.48

§ § tt-)

Cor., correlation coefficient. r ee~ tt3 tt~ tg3 tt~

405 (59) and 1643 (60) but comparable discordance. A line through these analyses points to a third zircon generation at around 1550 Ma. U-Pb dating of rutiles has been carried out to obtain age information on the Sfak/~ paragneiss of the MCC (road-cut 500 m east of Sfak/0. The paragneiss consists of plagioclase, quartz, white mica and chlorite 9 The pre-Alpine mylonitic foliation was almost destroyed by a low-grade cataclastic overprint 9 Rutiles are largely broken and red to red brown in colour. The analysis consists of 1-7 mg of rutile ( > 500 gm, 500-140 gm and 140-80 gm). The analyses 94-96 contain mainly common lead with a small spread in isotopic ratios. In the ~38U/z~ v. 2~176 diagram they define an isochron age of 146_+13Ma with a MSWD of 0.86 and a 2~176 ratio of 18.73_+0.22 (Fig. 5, Table 2).

r~ 0

tg'3 t~

§

,,-.,

g

k el

,~,

+l + r162 ~ ) M:)t~

eq,~-

2o

Zircon fission-track ages The investigation of zircon fission tracks of the protomylonitic Paraspori orthogneiss of the MCC yielded an age of 150_+ 14 Ma (Table 3; sample 00170902 reported by Romano et al. 2004). The Vai orthogneiss of the VCC yielded 184_+ 11 Ma (Table 3). Details on the samples and zircons are listed above.

U - ( T h ) - P b age o f monazite Five Carboniferous and three Permian ages have been determined from rocks exposed between the villages of Mochlos and Chamezi (Fig. 6 and Table 4). Only a few samples show an overlap with other generations within the analytical uncertainty. Monazite has survived in garnet micaschists and gneisses. The monazite ages of the MCC range from 380 to 260 Ma, whereas those of the KCC are significantly younger, ranging from 308 to 214 Ma.

tt~ tg-)

r/1

el) ~"~ ~

~ ~,~ ~o +'2 N

o

RADIOMETRIC DATING OF CRETAN BASEMENT

77

Table 4. U-(Th)-Pb ages, obtained by electron microprobe dating of monazite

A g e 600 [Ma] 500 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

100 ........................................................................................................................................................................................

[ ] Finger et al. 2002

9

This study

Fig. 6. Data for electron microprobe analyses of monazite with 2(y errors. Two different age spectra are obvious; Carboniferous ages are restricted to the MCC, whereas Permian ages prevail in the KCC.

S t r u c t u r a l evolution a n d microfabrics

The Myrsini crystalline complex s.s. (MCC) is characterized by a sequence of pre-Alpine ductile deformations (DM1--DM6). DMI and DM2 is preserved only as relics. Pervasive DM3 top-tothe-north shearing was active under amphibolitefacies conditions and is indicated by S-C fabrics, mica-fish and o-clasts of feldspar (Fig. 7e). The axes of DM3 folds trend parallel to the northsouth-oriented stretching lineation. Plagioclase is largely recrystallized. Quartz shows evidence for high-temperature grain boundary migration recrystallization and sub-grain rotation recrystallization. The quartz c-axes are distributed as asymmetric type I and type II cross girdles, secondary single girdles and small circles (Romano 2005). The growth of garnet, staurolite and biotite of micaschists postdates DM3 (Fig. 7c). Top-to-the-east shear sense during DM4 is indicated by asymmetric pressure shadows of mica behind garnet. DM5 is documented by chevron folds with east-west-trending axes in gneisses, micaschists and quartzites. DM6 is related to top-to-the-NE shear zones, which were active under retrograde metamorphic conditions within the brittle-viscous deformation quartz regime (Fig. 3d and f). DM6 shear zones are rich in chlorite, which resulted from replacement of garnet, staurolite and biotite. In the KCC relics of a first deformation (D~:I) are preserved within muscovite, biotite, garnet and staurolite in the form of opaque phases, which show a shape-preferred orientation. Dm isoclinal folding under amphibolite-facies conditions is associated with a pervasive foliation (Sin).

Sample

Rock type

Myrsini CC 00170902a 00170903 01010501 01020504 01040503

orthogneiss ab-paragneiss st-grt-micaschist micaschist pl-paragneiss

Kalavros CC 00110916 ms-micaschist 00300901 a bt-micaschist 00101003 quartzite

Number Age (Ma) of grains _+2~ 3 1 3 1 4

341 ___39 379 _+140 298 _+38 325 _+50 327 _+21

1 1 3

273 _ 140 258 _ 44 268 _ 40

Quartz was deformed by high-temperature grainboundary migration recrystallization effects (GBM, Fig. 8b). Chessboard patterns in some of the quartz grains indicate subgrain boundaries, which are aligned parallel to both the prism and the basal planes (Fig. 8c). The distribution of quartz c-axes shows asymmetric type I cross girdles (Romano 2005). Plagioclase underwent dynamic recrystallization. Saussuritization of plagioclase and pseudomorphic growth of muscovite after plagioclase, as well as of muscovite after staurolite and garnet are documented (Fig. 7f). Subsequent amphibolite-facies deformation resulted in isoclinal folds with east-westtrending axes and SI~3. Elevated temperature caused growth of K-feldspar (Fig. 7d). Rotation of plagioclase and biotite as well as mica-fish, ~-clasts and asymmetric pressure shadows of mica behind garnet indicate top-to-the-east shearing during DK4(Fig. 7b). Within the CCC relicts of Dcl and Dc2 are preserved as aligned micas and opaque phases, which are folded within albite and tourmaline, both of which grew synkinematically during Dc3. Open folds and parasitic minor folds with northsouth-trending axes and a stretching lineation (elongated feldspar) with the same north-southorientation developed in orthogneiss. The rotation of albite and clinozoisite blasts indicates top-to-the-north shearing. The quartz fabric resulted from subgrain rotation recrystallization (SGR; Fig. 8a). The distribution of quartz c-axes shows the development of type II cross girdles (Romano 2005). Pseudomorphic growth of muscovite, biotite and albite after garnet as well as of muscovite after biotite is obvious in micaschists. Dc4 led to chevron folds with east-west trending axis, parallel to a stretching lineation that results

78

S.S. ROMANO E T AL.

Fig. 7. Microphotographs showing microfabrics of the rocks investigated. (a) Large biotite grows across pervasive SMt cleavage, which is affected by a younger crenulation cleavage (DMs); parallel nicols, xz-section (MCC, sample 01010501). (b) Main foliation (SK4)was overprinted by a crenulation cleavage (SKs). The folded quartz vein shows boudinage; parallel nicols, xz-section (KCC, sample 01150504). (e) Contact between staurolite (St), biotite (Bt) and garnet (Grt) of the MCC micaschist. Staurolite postdates a folded slaty cleavage. Chlorite is present only along the contact between staurolite and garnet; parallel nicols, xz-section (MCC, sample 01010501), (d) Pressure shadow of muscovite on garnet is replaced by K-feldspar (Kfs); crossed nicols; xz-section (KCC, sample 3111). (e) Growth of albite across an older folded foliation; parallel nicols, xz-section (MCC, sample 260404). (f) Staurolite with internal foliation was replaced by white mica (Ms1). Staurolite forms a 5-clast, which resulted from DK4 top-to-the-east shearing. Arrow points to the internal foliation in staurolite, parallel nicols; xz-section (KCC, sample 01150504). (Mineral abbreviations after Kretz 1983).

from noncoaxial deformation (Fig. 3e). A top-tothe-east transport is indicated by o-clasts of epidote and plagioclase in actinolite schist (Fig. 3c). Dvl and Dv2 of the VCC are present only as relics. Top-to-the-NW shearing during Dv3 is

indicated by o-clasts of plagioclase. Both plagioclase and quartz were recrystallized during Dv3. Dv4 was active under retrograde metamorphic conditions and led to a crenulation cleavage in micaschists.

RADIOMETRIC DATING OF CRETAN BASEMENT

79

Fig. 8. Microphotographs showing microfabrics of the rocks investigated. (a) CCC quartz vein of a micaschist showing evidence for subgrain rotation recrystallization (SGR); crossed nicols (sample 260402). (b) Quartz lens of KCC shows grain boundary migration recrystallization (GBM) associated with top-to-the-east shearing, crossed nicols (sample 00180901). (c) Chessboard pattern in quartz of the KCC; crossed nicols (sample 00180901). (d) Quartz vein of the lower violet slates showing undulose extinction and low-temperature bulging recrystallization as a result of Alpine subduction; crossed nicols (sample 151001).

Chemical zonation o f garnet Garnets of the MCC, KCC and CCC show a multistage optical zonation. Franz (1992) identified a triphase optical and chemical zonation in garnets of the Myrsini crystalline complex sensu lato (s. l.) and the CCC. Usually the garnet cores (Z1) are without inclusions; the second zone (Z2) shows a shape-preferred orientation of opaque phases. The third zone (Z3) is free from inclusions. Garnets of the KCC show a fourphase zonation. The central part of the KCC garnets is similar to that of the garnets of the MCC and CCC. In contrast to the latter, the marginal zone of garnets is characterized by inclusions that display a shape-preferred orientation. The marginal zone is often replaced by white mica or chlorite. The weakest degree in zonation of garnets of the CCC is displayed by the grossularite component, which decreases from 18% (core) to 14% (rim). The Fe-number also decreases slightly from core to rim (Fig. 9a). Also typical

of a greenschist-facies overprint is a cloudy arrangement in the pyrope-spessartine-grossularite-triplot (Fig. 10). The scatter-plots-indicate isobaric heating, which is also suggested by the Xca v. Xug ratio and the translation into P-T paths (see Martignole & Nantel (1982) ratio; Fig. 11a). The Ca content of plagioclase changed during the younger Alpine metamorphism, as indicated by albitization of plagioclase and growth of zoisite. Moreover, chloritization of biotite and staurolite have caused an Fe-Mg exchange. Thus, no thermobarometer can be applied. Therefore, only a relative, semiquantitative P-T path can be derived (Fig. 11). The garnets of the MCC show a correlation of the optical and chemical zonation. Garnet growth was related to prograde metamorphism, as indicated by the highest Fe-number in cores (about 0.925) and by the bell-shaped zonation of the spessartine component (Fig. 9b). The temperature peak was reached in zone 3, as suggested by increasing pyrope component at the expense of the grossularite component (see also pyrope-spessartine-grossularite triplot, Fig. 12).

80

S.S. ROMANO E T AL.

Fig. 9. Representative line scans of element distribution in garnet based on electon microprobe data. (a) Micaschist of CCC: diameter 2 cm (sample 980523/1-1, grt A). (b) Micaschist of the MCC: diameter 1.1 cm (sample 01010501, grt A). (c) Amphibolite: diameter 1.2 cm (sample 00180902, grt B). (d) Micaschist of KCC: profile length 0.2 cm (rim) and 0.8 cm (diameter) (sample 02080502, grt D).

The translation into the P - T path based on the XMg v. )(Ca indicates a complex P - T history with isobaric heating in the marginal zone (Fig. 11 b).

The optically unzoned garnets of the garnet amphibolites (00180902, Table 5) of the M C C show a multiphase chemical zonation, as is indicated by the increasing pyrope component and

RADIOMETRIC DATING OF CRETAN BASEMENT

Fig. 10. Ternary diagram of spessartine, pyrope and grossularite component of the CCC garnets. decreasing grossularite component (Figs 9c and 12). The optical zonation of garnet of the KCC is also related to chemical zonation. The distribution of the Mg content, in particular, suggests Z 1 and Z2 to be anhedral, whereas Z3 and Z4 are euhedral and concentric (01150504; Table 5). Zone 4 is characterized by a grossularite increase at the expense of pyrope as well as an XCa increase, whereas Z1 to Z3 indicate a similar chemical trend to that of the garnets of the MCC (Fig. 13). The translation of the data into a P - T path indicates a multiple sequence of isobaric heating and isothermic loading that is succeeded by a phase of isobaric cooling (Fig. 1 lc).

Discussion M a j o r constituents o f the pre-Alpine basement o f eastern Crete The new data presented above suggest that the pre-Alpine basement of Crete includes at least four different complexes, which are intercalated between Carboniferous to Triassic metasedimentary and metavolcanic rocks of the PhylliteQuartzite Unit. Utilizing the names of adjacent villages, these crystalline complexes are referred to as the Chamezl crystalline complex (CCC), the Kalavros crystalline complex (KCC), the Myrsini crystalline complex (MCC), and the Vai crystalline complex (VCC). The most important criteria that have been used to distinguish these complexes are listed in Table 6. The new monazite ages are compatible with those obtained by Finger et al. (2002). As the

81

closure temperature for the U - P b system of monazite (725+25 ~ Parrish 1990) was not reached in all of the complexes investigated, the monazite ages are interpreted to reflect the growth of the monazites during amphibolitefacies Barrovian-type metamorphism. The monazite ages unequivocally confirm that the Myrsini crystalline complex, after Franz (1992), consists of two complexes, the KCC with Permian metamorphism, and the Myrsini crystalline complex s.s. (MCC) with Carboniferous metamorphism (Finger et al. 2002). There is a clear correlation of the spatial distribution of monazite ages and garnet types (Fig. 14). The four-phase garnets prevail in the area where Permian monazite ages are observed (i.e. in the KCC). The three-phase garnets, on the other hand, are restricted to domains where Carboniferous monazite ages have been found (i.e. the MCC). Also, the K - A r ages of white mica (Seidel et al. 1982) are compatible with the distribution of monazite ages and garnet types (Fig. 14); that is, Late Carboniferous K - A r in the MCC, but Permo-Triassic K - A r in the KCC (Fig. 15). The MCC forms the core of the ENE-WSWtrending Myrsini syncline, which was formed during the Alpine cycle. In the course of Carboniferous convergence and related top-to-the-north movements, both the Cambrian granitoids and the paragneisses and micaschists of the MCC were pervasively sheared under amphibolitefacies conditions. As the dominant fabric is a D1 fabric in orthogneiss, but a D 3 fabric in micaschist and paragneiss, it cannot be excluded that the granitic protolith of the orthogneisses was intruded into a pre-existing metamorphic basement. The Cambrian age of the granitoids suggests this basement to be of Neoproterozoic (Cadomian-Pan-African) origin. The Carboniferous top-to-the-north movements could be related to north-south convergence. Given that subduction was active at this time, the dip of the subducting slab should have been towards the south (see also Xypolias et al. 2006). Although the age of metamorphism of the MCC and KCC is different (within uncertainties), the kinematic relations of both complexes are similar. Exposures of the KCC are largely restricted to the northern and southern limb of the kilometre-scale Myrsini syncline, whereas the MCC forms the core (Fig. 14). Thus, the map-scale structure suggests that the MCC rests tectonically above the KCC. The contact should be a thrust, as MCC rocks with higher metamorphic ages are resting above KCC rocks with lower metamorphic ages. The chemical zonation of the KCC garnets suggests prograde metamorphism; this implies multiple sub-stages with

82

S.S. ROMANO E T AL.

Fig. 11. XCaV. XMgratio of garnet of(a) CCC (01030501 grt B), (b) MCC (01010501 grt B) and (c) KCC (02080502 grt D). increase in pressure and temperature, succeeded by a stage of isobaric cooling. It is important that, in contrast to the MCC, the temperature increase in the KCC led to (1) growth of K-feldspar and breakdown of garnet and white mica, (2) high-temperature grain boundary migration in quartz and incipient slip along the prism planes, the latter being indicated by the chessboard pattern in quartz (Mainprice et al.

1986). The amphibolite-facies top-to-the-north shear zones of the KCC were apparently active in the Permian related to north-south convergence. Moreover, north-south kinematics also prevailed during the exhumation of the crystalline complexes, as indicated by late east-west-trending fold axes and by discrete phyllonitic shear zones, the latter showing top-to-the-north sense of shear under retrograde metamorphic conditions.

RADIOMETRIC DATING OF CRETAN BASEMENT

83

Fig. 12. Ternary diagrams of spessartine, pyrope and grossularite components of garnets of the MCC. Table 5. Sample description of the documented garnet-bearing rocks Sample

Location Rock type Minerals Microstructure

01150504 (KCC)

00290901 (MCC)

00180902 (MCC)

Path cut 1 km NE of Paraspori grt-st-micaschist st, ms, grt,bt, qtz, mr Amp foliation: older grt cores & slaty cleavage; st-) ms; SK4 foliation: ms, fsp and qtz; ms & tur mineral ha; top-to-the-E; asymmetric pressure sliadows of ms on bt & st; CC, undulose extinction of ms, kinking bt & boudinage

0.5 km S of Mochlos; E of the road to the quarry grt-micaschist grt, ms, bt Grt cores were ambient by aligned ms, opaque phases and bt; foliation is Si in grt & bt; N-S lineation; second foliation of aligned ms & tur; asymmetric pressure shadows of pl & bt on grt; E-W folds & crenulation cleavage

0.5 km N of the road crossing Messfi Moulianfi-Kalavros grt-amphibolite grt, amp, bt, ms, pl, ep, tur, zo Metamorphic layering: euhedral grt, green amp, bt & white ms in matrix of equigranular pl, ep, tur & zo; E-W amp lineation, grt with Si of pl; alteration of amp, bt ---~chl& sericitization of pl.

Mineral abbreviations after Kretz (1983). The C C C displays the lowest m e t a m o r p h i c grade o f all o f the crystalline complexes (upperm o s t greenschist facies). It rests above the M C C along a brittle-ductile shear zone, the age of which has yet n o t been determined. As the C C C shows striking similarities to the M C C , such as

the presence o f C a m b r i a n granitoids and top-tothe-north kinematics, we suggest that b o t h complexes belong to a single basement unit. There is only a difference in Carboniferous burial and thus a difference in the grade o f m e t a m o r p h i s m . The C C C f o r m e d the lower part o f the upper

84

S.S. ROMANO E T AL.

Fig. 13. Ternary diagrams of spessartine, pyrope and grossularite components of garnets of the KCC.

Table 6. Diagnostic criteria o f the different crystalline complexes o f eastern Crete Crystalline complex Criterion

CCC

KCC

MCC

VCC

Recrystallization of quartz ~ Garnet zonation ~ Age of metamorphism Protolith age of granitoids K-Ar white mica 3

BL + SGR Z~-Z3 ? Cambrian 2'4 ? ?

SGR + GBM Z~-Z3 Carboniferous 1'2'3 Cambrian 2'4 Carboniferous Permianlate Jurassic

SGR + GBM ? Triassic-Jurassic ~ Triassic l

Zircon fission-track ages ~

GBM ZI-Z4 Permian 1'2'3 '~ PermianTriassic ?

Early Jurassic

Based on Jnew data presented in this paper as well as 2Finger et al. (2002), 3Seidel et al. (1982) and 4Romano et al. (2004).

crust, whereas the M C C f o r m e d the lower crust. T h e r e t r o g r a d e m e t a m o r p h i c c h a r a c t e r of the intervening shear zone suggests the latter to result f r o m extensional m o v e m e n t s during e x h u m a t i o n . As the U - P b age of rutile of the Sfakfi paragneiss (146_+ 13 M a ) is similar to the fission-track age

o f zircon of the P a r a s p o r i orthogneiss 14 Ma), the M C C rocks should have below c. 350 ~ in Late Jurassic times. It be n o t e d that the closure t e m p e r a t u r e for Ar system of white mica was already by the C a r b o n i f e r o u s - P e r m i a n b o u n d a r y

(150_+ cooled should the K passed (Seidel

RADIOMETRIC DATING OF CRETAN BASEMENT

85

Fig. 14. Distribution of K-Ar ages of muscovite (Seidel et al. 1982), U-(Th)-Pb model ages of monazite (Finger et al. 2002, and this study) and type of the garnet zonation (this study).

et al. 1982; Fig. 15); that is, the MCC rocks were situated within the temperature interval of c. 350 ~ to c. 400 ~ for a relatively long period. The Vai crystalline complex (VCC) differs from the other complexes discussed above. The new U-Pb and fission-track data for zircon indicate that the tectonometamorphic imprints of the VCC are much younger than those of the other complexes. Based on the zircon typology (Pupin & Turco 1972; Vavra 1990; Benisek & Finger 1993; Hanchar & Miller 1993; Finger & Helmy 1998), the investigated zircons should be of magmatic origin. As the temperature peak of the pre-Alpine and Alpine metamorphism was below the closure temperature of the U-Pb system of zircon (>900 ~ Cherniak & Watson 2000), recrystallization of the zircons can be ruled out. The upper intercepts at 767+23 Ma and c. 1500 Ma reflect two older generations. The lower intercepts at 223 + 11 Ma and 223 _+ 13 Ma are interpreted to reflect the time of magma emplacement. Assuming this assumption is correct, the amphibolite-facies pervasive shearing of the Vai granite must have been active in Late Triassic or subsequent times. A lower boundary for the

ductile shearing of the Vai orthogneiss is given by the fission-track data for zircon (184___11 Ma). Thus, the period during which the Vai granite was converted to orthogneiss is bracketed as between 236 and 173 Ma (Mid-Triassic to MidJurassic; see also the temperature-time path in Fig. 15). As the dominant structural elements and the kinematics of the orthogneiss are similar to those of the adjacent micaschists (top-to-the N W movements under amphibolite-facies metamorphism), and evidence for syn-emplacement shearing of the gneiss protolith is lacking, the ductile deformation of the entire VCC should be restricted to this time interval. Tectonic m o d e l f o r the Carboniferous to Triassic period

Before giving a preliminary tectonic model for the Carboniferous to Triassic period, we will briefly discuss the Alpine imprints. The latter significantly affected the present pile of basement rocks. During Alpine subduction and collision, the basement complexes behaved mechanically

86

S.S. ROMANO E T AL.

Fig. 15. T - t paths based on the new data and data from Seidel et al. (1982), Finger et al. (2002) and Romano et al. (2004).

more or less as rigid bodies, which were surrounded by the weaker Carboniferous to Triassic metavolcanic and metasedimentary rocks (Zulauf et al. 2002). There are two lines of evidence suggesting that the present tectonometamorphic sequence of the Phyllite-Quartzite Unit results from dramatic displacements along Alpine d6collements. (1) The volcanosedimentary sequence of the Phyllite-Quartzite Unit shows Carboniferous to Late Triassic ages (Krahl et al. 1986; Kozur & Krahl 1987; Haude 1989); similar ages have been obtained for the metamorphism of the basement complexes (Finger et al. 2002, and this study). (2) Most of the detrital zircons of the volcanosedimentary sequence of the Phyllite-Quartzite Unit yielded Devonian to Carboniferous fission-track ages (Brix et al. 2002). As these ages are largely greater than those determined for the metamorphism of the now exposed basement, the latter cannot be regarded as a source rock for the deposition of the Carboniferous to Triassic volcanosedimentary sequence of the Phyllite-Quartzite Unit. Based on the data presented above it is concluded that in pre-Alpine times (earlier than c. 330 Ma) the individual nappes of the present tectonometamorphic sequence of the Phyllite-Quartzite Unit

were situated at significantly different places, forming separate microplates (Fig. 16). This holds for both the basement slices and the younger Carboniferous to Triassic rocks, which are free of pre-Alpine metamorphism. The only candidate that might have acted as source for the Carboniferous to Triassic sediments of the Phyllite-Quartzite Unit is the Arabian-Nubian Shield, where sphene and zircon fission-track data give ages ranging from 339 to 410 and from 315 to 366 Ma (Bojar et al. 2002). As a result of north-south shortening in Carboniferous to Triassic times, the Palaeo-Tethys lithosphere was consumed and the microplates were deformed and metamorphosed during collision with Gondwana. The prevailing top-to-the-north kinematics of large shear zones suggest that possible subduction zones were dipping towards the south beneath the Cimmerian arc ($eng6r et al. 1984). A south-dipping pre-Alpine subduction zone is further indicated by the spatial-temporal distribution of Carboniferous granitoids (Xypolias et al. 2006, and references therein), but is not consistent with plate tectonic models presented by Stampfli & Borel (2002). We suggest that south-directed subduction of PalaeoTethyan lithosphere beneath the northern margin

RADIOMETRIC D A T I N G OF C R E T A N BASEMENT

87

8~ ".~o

~o

O @

d@

"'a "~ IS~ O

r~ 9

.= ~rd o-~

"~

~S

a.Z ~ :...q a~

$= ~

0

@ o

r~'~ ~,

88

S.S. ROMANO E T AL.

of Gondwana began during Early Carboniferous times. This subduction was followed by collision of the M C C - C C C microplate, the latter being accreted to the northern margin of Gondwana at c. 330 Ma. The crust behind the active margin extended, forming a back-arc basin (Fig. 16). In Permian times the microplate of the KCC collided and was pushed underneath the M C C - C C C (Eocimmerian event). The still extending back-arc basin was subdivided into a southern and a northern part, where the sediments of the Plattenkalk and PhylliteQuartzite Unit, respectively, were deposited. As volcanism started to be active along the arc, the volcanic rocks were eroded and redeposited within the adjacent basins. Consequently, parts of the Permo-Triassic sedimentary rocks of the Phyllite-Quartzite Unit are intercalated with andesitic lava and pyroclastic rocks. PermoTriassic orogenic activity is also indicated by the Scythian conglomerates of the PhylliteQuartzite Unit of eastern Crete (Krahl et al. 1986). Moreover, evidence for Permo-Triassic metamorphic events has been found in surrounding units such as the Cyclades, the Dodecanese and the Menderes Massif of western Anatolia (see references given by Finger et al. 2002). The terminal orogenic event occurred in Late Triassic time when the VCC collided under topto-the-NW kinematics. Simultaneously, the sedimentation in the northern part of the back-arc basin (Phyllite-Quartzite Unit) ceased. A switch from deep to shallow water facies (Gypsum Rauhwacke formation) in ?Mid-Triassic times predates the collisional event (Krahl et al. 1983). The latter in particular is attributed to the Cimmerian orogeny ($eng6r et al. 1984). The accretion of at least three microplates to the northern margin of Gondwana led to a northward shift of the south-dipping Cimmerian subduction zone (roll-back). The margin became inactive and subduction ceased in Early Jurassic times, when the whole area was affected by extension ($eng6r et al. 1984). Towards the north of the Cimmerian arc, the Tripolitza basin and the Pindos basin in particular subsided during this time. It is proposed that a change from an overall convergent to an overall extensional setting controlled the rapid exhumation of the VCC, the latter reaching upper crustal levels (with T < c . 350~ already during Early Jurassic times.

Conclusions From the new data presented here we conclude that the pre-Alpine basement of eastern Crete consists of at least four crystalline complexes.

Based on the new U - P b and fission-track ages of zircon it is demonstrated that the contact between the pre-Alpine basement and the metasedimentary rocks of the Phyllite-Quartzite Unit (Chamezi beds) involved significant Alpine shearing and nappe transport. Further information, such as geochemical data for (meta)granitoids and age constraints on the metamorphism of the CCC and VCC, are necessary to test the preliminary tectonic model persented here, which invokes south-directed Carboniferous to Triassic subduction, collision and accretion of the individual basement complexes to the northern margin of Gondwana. This work was supported by a grant from Deutsche Forschungsgemeinschaft Zu 73-8. We thank J. Schastok and B. Herrmann for help in zircon preparation. We also thank J. Krahl for his introduction to the area of Vai, as well as S. Barthelmes, N. Beau, R. Bolte, B. Borsanyi, C. Josenhans and S. Schwanz for providing their geological maps. The manuscript benefited from comments by U. Ring, C. Fassoulas, and A. and T. Usta6mer.

References ATHERTON, M. P. 1968. The variation in garnet, biotite, and chlorite composition in medium grade pelitic rocks from the Dalradian, Scotland, with particular reference to the zonation in garnet. Contributions to Mineralogy and Petrology, 18, 347-371. BENISEK, A. & FINGER, F. 1993. Factors controlling the development of prism faces in granite zircons: a microprobe study. Contributions to Mineralogy and Petrology, 144, 441-451. BOJAR, A.-V., FRITZ, H., KARGL, S. & UNZOG, W. 2002. Phanerozoic tectonotherrnal history of the Arabian-Nubian shield in the Eastern Desert of Egypt: evidence from fission track and paleostress data. Journal of African Earth Sciences, 34, 191-202. BONNEAU, M. 1984. Correlation of the Hellenide nappes in the south-east Aegean and their tectonic reconstruction. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 517-527. BRIX, M. R., STOCKHERT, B., SEIDEL, E., THEYE, T., THOMSON, S. N. & KI~STER, M. 2002. Thermobarometric data from a fossil zircon partial annealing zone in high pressure-low temperature rocks of eastern and central Crete, Greece. Tectonophysics, 349, 309-326. CHERNIAK, D. J. & WATSON, E. B. 2000. Pb diffusion in zircon. Chemical Geology, 172, 5-24. DORR, W., BELKA, Z., MARHEINE, D., SCHASTOK,J., VALVERDE-VAQUERO,P. & WISZNIEWSKA,J. 2002a. U-Pb and Ar-Ar geochronology of anorogenic granite magmatism of the Mazury complex, NE Poland. Precambrian Research, 119, 101-120.

RADIOMETRIC DATING OF CRETAN BASEMENT DORR, W., ZULAUF, G., FIALA, J., FRANKE, W. & VEJNAR, Z. 2002b. Neoproterozoic to Early Cambrian history of an active plate margin in the Teplg-Barrandian unit--a correlation of U-Pb isotopic-dilution-TIMS ages (Bohemia, Czech Republic). Tectonophysics, 352, 65-85. FASSOULAS, C. 1999. The structural evolution of central Crete: insight into the tectonic evolution of the south Aegean (Greece). Geodynamics, 27, 23-45. FASSOULAS, C., KILIAS, A. & MOUNTRAK~S, D. 1994. Postnappe stacking extension and exhumation of high-pressure/low-temperature rocks in the island of Crete, Greece. Tectonics, 13, 127-138. FINGER, F. & HELMY, H. M. 1998. Composition and total-Pb model ages of monazite from high-grade paragneisses in the Abu Swayel area, southern Eastern Desert, Egypt. Mineralogy and Petrology, 62, 269-289. FINGER, F., KRENN, E., RIEGLER, G., ROMANO, S. & ZULAUF, G. 2002. Resolving Cambrian, Carboniferous, Permian and Alpine monazite generations in the polymetamorphic basement of eastern Crete (Greece) by means of the electron microprobe. Terra Nova, 14, 233-240. FRANZ, L. 1992. Die polymetamorphe Entwicklung des Altkristallins auf Kreta und im Dodekanes ( Griechenland): eine geologische, geochemische und petrologische Bestandsaufnahme. Enke, Stuttgart. GALBRAITH, R. F. & LASLETT, G. M. 1993. Statistical models for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21, 459-470. GLEADOW, A. J. W. 1981. Fission track dating methods: what are the real alternatives? Nuclear Tracks and Radiation Measurements, 5, 3-14. HANCHAR, J. M. & MILLER, C. F. 1993. Zircon zonation patterns as revealed by cathodoluminescence and backscattered electron images: implications for interpretation of complex crustal history. Chemical Geology, ll0, 1-13. HAUDE, G. 1989. Geologie der Phyllit-Einheit im Gebiet urn Palekastro (Nordost-Kreta, Griechenland). PhD thesis, Technische Universit~it M/inchen. HOLLISTER, L. S. 1966. Garnet zoning: an interpretation based on the Rayleigh fractionation model. Science, 154, 1647-1651. HURFORD, A. J. 1990. Standardization of fission track dating calibration: recommendation by the Fission Track Working Group of the lUGS Subcommission on Geochronology. Chemical Geology (Isotope Geoscienees), 80, 171-178. HURFORD, A. J. ~r GREEN, P. F. 1983. The zeta age calibration of fission track dating. Isotope Geoscience, 1, 285-317. HURFORD, A. J., HUNZIKER, J. C. • STOCKHERT, B. 1991. Constraints on the late thermotectonic evolution of the Western Alps: evidence for episodic rapid uplift. Tectonics, 10, 758-769. JACOBSHAGEN, V. 1986. Geologie yon Griechenland. Borntr/iger, Berlin. JOLIVET, L., DANIEL, J. M., TRUFFERT-LUXEY, C. & GOFEr, B. 1994. Exhumation of deep crustal metamorphic rocks and crustal extension in back-arc regions. Lithos, 33, 3-30.

89

JOLIVET, L., GOFFI~, B., MONII~,P., TRUFFERT-LUXEY, C., PATRIAT, M. & BONNEAU, M. 1996. Miocene detachment in Crete and exhumation P-T-t paths of high-pressure metamorphic rocks. Tectonics, 15, 1129-1153. KILIAS, A., FASSOULAS, C. & MOUNTRAKIS, D. 1994. Tertiary extension of continental crust and exhumation of Psiloritis 'metamorphic core complex' in the central part of the Hellenic arc (Crete, Greece). Geologische Rundschau, 83, 417-430. KOZUR, H. & KRAHL, J. 1987. Erster Nachweis von Radiolarien im tethyalen Perm Europas. Neues Jahrbuch fiir Geologie and Paliiontologie, 174, 357-372. KRAHL, J., KAUFFMANN,G., KOZUR, H., RICHTER, D., FORSTER, O. & HEINRITZI,F. 1983. Neue Daten zur Biostratigraphie und zur tektonischen Lagerung der Phyllit-Gruppe und der Trypali-Gruppe auf der Insel Kreta (Griechenland). Geologische Rundschau, 72, 1147-1166. KRAHL, J., KAUFFMANN, G., RICHTER, D., et al. 1986. Neue Fossilfunde in der Phyllit-Gruppe Ostkretas (Griechenland). Zeitschrift der Deutschen Geologischen Gesellschaft, 137, 523-536. KRETZ, R. 1983. Symbols for rock-forming minerals. American Mineralogist, 68, 277-279. KROGH, T. E. 1982. Improved accuracy of U-Pb zircon ages by the creation of more concordant systems using an air abrasion technique. Geochimica et Cosmochimica Acta, 46, 637-649. MAINPRICE, D., BOUCHEZ, J.-L., BLUMENFELD, P. & TuBIA, J. M. 1986. Dominant c slip in naturally deformed quartz: implications for dramatic plastic softening at high temperature. Geology, 14, 819822. MARTIGNOLE, J. & NANTEL, S. 1982. Geothermobarometry of cordierite-bearing metapelites near the Morin anorthosite complex, Grenville province, Quebec. Canadian Mineralogist, 20, 307-318. MIYASHIRO, A. & SHIDO, F. 1973. Progressive compositional change of garnet in metapelite. Lithos, 6, 13-20. NAESER, C. W. 1976. Fission track dating. US Geological Survey Open-File Report, 76-190. PARRISH, R. R. 1990. U-Pb dating of monazite and its application to geological problems. Canadian Journal of Earth Sciences. 27, 1431-1450. PUPIN, J. P. 1980. Zircon and granite petrology. Contributions to Mineralogy and Petrology, 73, 207-220. PuPrN, J. P. & TURCO, G. 1972. Une typologie originale du zircon accessoire. Bulletin de la Sociktk Franfaise de Mindralogie et de Cristallographie, 95, 348-359. ROMANO, S. S. 2005. Ursprung und Entwicklung des Altkristallins Ostkretas, Griechenland." geochronologische und strukturelle Untersuchungen. PhD thesis, Universit/it Frankfurt. ROMANO, S. S., DORR, W. & ZULAUE, G. 2004. Cambrian granitoids in pre-Alpine basement of Crete (Greece): evidence from U-Pb dating of zircon. In: DORR, W., FINGER, F., LINNEMANN,U. & ZULAUF, G (eds) The A valonian-Cadomian Belt and Related Peri-Gondwana Terranes. International Journal of" Earth Sciences, 93, 844-859. SEIDEL, E. 1978. Zur Petrologie der Phyllit-QuarzitSerie Kretas. Habilitation thesis, Universit/it Braunschweig.

90

S.S. ROMANO ET AL.

SEIDEL, E., KREUZER, H. & HARRE, W. 1982. A Late Oligocene/Early Miocene high pressure belt in the External Hellenides. Geologisches Jahrbuch, E23, 165-206. ~ENGOR, A. M. C., YILMAZ,Y. & S0qqGIS/RLIJ,O. 1984. Tectonics of the Mediterranean Cimmerides: nature and evolution of the western termination of Palaeo-Tethys. In: DIxoN, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 77-112. SPEAR, F. S. (ed.) 1993. Metamorphic Phase Equilibria and Pressure-Temperature-Time Paths. Mineralogical Society of America, Monograph Series, Washington, D. C. STACEY, J. S. & KRAMERS, J. D. 1975. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters, 26, 207-221. STAMPFLI, G. & BOREL, G. 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. THEYE, T. & SEIDEL, E. 1991. Petrology of low-grade high-pressure metapelites from the External Hellenides (Crete, Peloponnese). A case study with attention to sodic minerals. European Journal of Mineralogy, 1991, 343-366.

THEYE, T., SEIDEL, E. & VIDALO, O. 1992. Carpholite, sudoite and chloritoide in low-temperature highpressure metapelites from Crete and the Peloponnese, Greece. European Journal of Mineralogy, 4, 487-507. THOMSON, S. N., STOCKHERT,B. & BRIX, M. R. 1998. Thermochronology of the high-pressure metamorphic rocks of Crete, Greece: implications for the speed of tectonic processes. Geology, 26, 259-262. TODT, W. 1988. Isotope dilution measurements of Pb, U and Th concentrations in lorandite from Allchar. Nuclear Instruments and Methods in Physics Research, A271, 251-252. VAVRA, G. 1990. On the kinematics of zircon growth and its petrogenetic significance: a cathodoluminescence study. Contributions to Mineralogy and Petrology, 106, 90-99. XYPOLIAS, P., Dt)RR, W. & ZULAUF, G. 2006. Late Carboniferous plutonism within the pre-Alpine basement of the External Hellenides (Kithira, Greece): eidence from U-Pb zircon dating. Journal of the Geological Society. London, 163, 539-547. ZULAUF, G., KOWALCZYK,G., KRAHL, J., PETSCHICK, R. & SCHWANZ, S. 2002. The tectonometamorphic evolution of high-pressure low-temperature metamorphic rocks of eastern Crete, Greece: constraints from microfabrics, strain, illite crystallinity and paleostress. Journal of Structural Geology, 24, 1805-1828.

Sedimentary evidence from the south Mediterranean region (Sicily, Crete, Peloponnese, Evia) used to test alternative models for the regional tectonic setting of Tethys during Late Palaeozoic-Early Mesozoic time A. H. F. R O B E R T S O N

Grant Institute o f Earth Science, School o f GeoSciences, University o f Edinburgh, West Mains Road, Edinburgh, EH9 3JW, UK (e-mail: alastair, [email protected], uk) Abstract: The south Mediterranean region, including western Sicily, Crete and mainland Greece (southern Peloponnese and Evia), is critical to an interpretation of the Late Palaeozoic-Early Mesozoic tectonic evolution of Tethys. Several contrasting tectonic models compete to explain the regional evolution. In a divergence-related hypothesis (Model 1) the south Aegean region experienced pulsed rifting along the northern margin of Gondwana that culminated in break-up to form the Pindos ocean in the region of Greece. In an alternative convergence-related hypothesis (Model 2) the south Aegean experienced Late Palaeozoic Early Mesozoic northward subduction, accretion and arc magmatism, culminating in 'Cimmerian' suturing of a Palaeotethyan ocean in latest Triassic time. In a third model, southward subduction of a Palaeotethyan ocean took place beneath the North Gondwana margin during Late Palaeozoic-Triassic time, giving rise to back-arc magmatism in an extensional setting. In addition, a more complex setting involving two opposing subduction zones (Andean-type and intra-oceanic) has also been suggested (Model 4), mainly based on lava geochemistry. To test these tectonic alternatives, mainly sedimentary studies were carried out in western Sicily, western and eastern Crete, the Peloponnese and Evia (eastern central Greece). Western Sicily was studied as a proxy for the unexposed deep Mediterranean south of Crete. Most of the available evidence supports the divergence-related (pulsed rift) hypothesis (Model 1). There is no clear evidence of sea-floor spreading (e.g. ophiolites) to the south of what became the Pindos ocean, or of plate convergence (e.g. magmatic arcs, subduction complexes), or collisional deformation in the south Aegean region that could be related to subduction or collision during the Mid-Carboniferous to Triassic, as in Model 2. Model 3 is not supported by evidence from the wider region (northern Greece, Turkey). Model 4 is not supported by evidence independent of igneous geochemistry. In the proposed interpretation, the northern margin of Gondwana initially rifted during Mid-Carboniferous to Early Permian time to form a wide deep-water basin. This was followed by further rifting, associated with volcanism during the Early Triassic; final continental break-up and spreading to form the Pindos ocean to the north during Late Triassic to Early Jurassic time then followed. Mid-Triassic uplift of part of the rift basin is explained as a flexural response to rifting as a precursor to opening of the Pindos ocean. Passive margin subsidence during the Early Mesozoic relates to opening of the Pindos ocean to the north. A subduction geochemical signature within some Triassic volcanic rocks, in this interpretation, is explained by melting of heterogeneous sub-crustal mantle, following an earlier, possibly Hercynian, subduction event.

The quest for 'Palaeotethys' of Late Palaeozoic to Early Mesozoic age in the Mediterranean region continues (Fig. 1). Most palaeomagnetic reconstructions suggest that a large westwardnarrowing gulf of the super-ocean, Panthalassa ('Palaeotethys'), existed in the Eastern Mediterranean region by Late Permian time (e.g. Smith et al. 1981). What was the nature of this ocean? Where are its remnants? How does it relate to younger Mesozoic Neotethyan oceanic basins in the Eastern Mediterranean region? Deepmarine facies are known to have bordered the north margin of Gondwana, at least from

Mid-Carboniferous time (Krahl et al. 1982; Kozur & Krahl 1984; Catalano et al. 1991; Kozur 1993, 1995), but their tectonic setting is controversial. In a first, divergence-related Model 1 (Fig. 2), a Palaeotethyan ocean was subducted northwards beneath Eurasia, as indicated by evidence from the Pontides of northern Turkey and elsewhere along the southern margin of Eurasia. The Pelagonian Zone of Greece, eastern Crete and all of the units south of this continental fragment, known as the Pelagonian microcontinent, rifted from G o n d w a n a during Early Mesozoic time

From: ROBERTSON,A. H. F. & MOtrNTRAKIS, D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 91-154. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

92

A.H.F. ROBERTSON

"' . . . . . . . . . .

'0 ~ /I f ~ '

..-"m20~

'

-5

=,:o..,_

~:~, ~ Mediterranean SeCrete , ) _ _

I Levant __1

o.

=-

-"~_.~/ L N Africa .....~a" ~ , . _

~

+

.

,

,

===1Phanerozoic CoMer

'~

Sea

Lb i ya~

~) ~

--~-

Precambrian basement

E Evia

[ ~

'

+"-,_

~

I o'N"

'.

'~l MesozoicCenozoic cover

•

Alpine deformation

Palaeozoic

~

Pre-Alpine basement

basement

P Peloponnese

Fig. 1. Outline map of the Mediterranean region showing the major tectonic elements and the study area (within box). Modified after Papanikolaou & Ebner (1996-1997).

to create several Neotethyan oceanic basins. The principal oceanic realm between Eurasia and Gondwana in the Late Triassic lay to the north of the Pelagonian continent in this interpretation and thus no arc remnants or collisional suture existed further south. Model 1, in several variants, was favoured by workers such as Smith et al. (1975), Robertson & Dixon (1984), Dercourt et al. (1986, 1993, 2000), Robertson et al. (1991, 1996, 2004), Papanikolaou (19961997), Ricou 1996, Yxlmaz et al. (1996) and Dornsiepen et al. (2001). In an alternative convergence-related Model 2 (Fig. 2), a Palaeotethyan ocean was also subducted northwards beneath the southern margin of Eurasia during Late Palaeozoic-Early Mesozoic time; and again, the southern, Gondwana margin remained passive. However, a crucial difference is that the Palaeotethyan suture is inferred to be located much further south, within the south Aegean region, to the south of the Pelagonian continent, which is considered as part of Eurasia. In this interpretation an ocean opened along the northern margin of Gondwana during Late Ordovician-Early Silurian time and a

continental fragment, termed the Hun terrane, was detached and drifted northwards until it was accreted to Eurasia, with the Palaeotethys opening in its wake along the northern margin of Gondwana. The northern Palaeotethys was in turn subducted beneath Eurasia during the Late Palaeozoic until the Hun terrane collided and was accreted during the 'Hercynian' orogeny. During this subduction a new ocean basin, termed Neotethys in this model, rifted along the Gondwana margin during Late Permian time detaching a Cimmerian microcontinent. The remaining Palaeotethys continued to subduct, opening several Triassic marginal basins (Vardar and Pindos) until it too sutured in the latest Triassic 'Cimmerian' orogeny. The remaining Neotethys survived in this interpretation until Early Cenozoic subduction and eventual suturing of the African and Eurasian plates in the Balkan region. Variants of this interpretation were proposed by several workers (i.e. Pe-Piper 1982; Stampfli et al. 1991, 1998, 2001; Stampfli & Borel 2002). In a radically different Model 3 (Fig. 2), Seng6r (1984) proposed that 'Palaeo-Tethys' was rooted in the north, adjacent to the southern

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS margin of Eurasia (e.g. Pontides; Crimea), and that a Neo-Tethyan ocean rifted to the south of this, as one of several back-arc basins above a south-dipping subduction zone during the Triassic. This model, like the first, implies that south of the Pelagonian continent the Triassic setting was one of rifting, not subduction, collision or magmatism. This model has been tested and shown to be problematic based on studies in northern Turkey (e.g. Usta6mer & Robertson 1997), but has recently received renewed support from several researchers (e.g. Smith 1999, Karamata et al. 2006; Romano et al. 2006). Finally, Pe-Piper & Piper (2002) have recently proposed an additional tectonic interpretation (Model 4; Fig. 2), based mainly on the geochemistry of Triassic volcanic rocks in Greece, which invokes double subduction (Fig. 2d). This infers Triassic northward subduction from a southerly Palaeotethys in the south Aegean region as in Model 2, but also the presence of an additional Triassic, southward-dipping intra-oceanic subduction zone located in the eastern part of a Triassic Pindos ocean. Models 2 and 4 require the existence of a Late Palaeozoic-Early Mesozoic convergent margin and a collisional suture in the region of Crete and the Peloponnnese, whereas Models 1 and 3 locate the subduction zone of this age well to the north (albeit with opposite polarities) and imply a rift and passive margin evolution to have characterized the south Aegean region during the Triassic. The different models thus involve starkly contrasting inferences about the tectonic setting at this time, in this region. The primary aim of the paper is to present field-based sedimentary evidence from Sicily, Crete, the Peloponnese and Evia which will be used to test the above tectonic hypotheses in the light of the existing literature. The key requirement is to distinguish between generic models, which infer either divergence (Models 1 and 3), or convergence (Models 2 and 4) during pre-Jurassic time, rather than to test any one specific model, as variants of each of these models have been published and further alternatives may exist. The end-product will be a new tectonic model for the south Aegean region for Late Palaeozoic-Early Mesozoic time. An immediate problem is that the evidence for the existence of any former oceanic crust located along the northern margin of Africa, south of Crete has been obscured by Cenozoic subduction and the present deep-marine basin. The timing and setting of Neotethyan continental break-up cannot be determined from the on-land record of North Africa alone (Guiraud et al. 2001).

93

However, further west, in Sicily, Cenozoic northward subduction has already resulted in collision of a Tethyan accretionary prism with a promontory of Gondwana and, as a result, fragments of Late Palaeozoic-Early Mesozoic crust are exposed within a thrust belt in western Sicily (Catalano et al. 2000a, b). These units are critical to determine whether or not a 'Neotethyan' ocean existed in the South-Mediterranean during Late Palaeozoic-Early Mesozoic time. This area will be discussed first as a proxy for crust of this age south of Crete. The Upper Palaeozoic-Lower Mesozoic metasedimentary and metavolcanic units of Crete and the Peloponnese will then be considered. Evidence from the Pindos and Pelagonian zones further north in Greece is also important, particularly to determine if an early Mesozoic 'Cimmerian' collisional event affected these areas. One persistent problem is that Tethyan nomenclature tends to be model-specific. Thus, for Seng6r (1984) 'Palaeo-Tethys' is rooted in a relatively northerly location, whereas for Stampfli et al. (2001) their Palaeotethys is rooted further south, and, by definition, Neotethys even further south again (Fig. 2). When such a modelspecific nomenclature is adopted, one is at once locked into hypothesis confirmation rather than hypothesis testing (see Robertson & Mountrakis 2006). For this reason, a looser, non modeldependent approach is used here. The term Palaeotethys as used here refers generally to older (i.e. pre-Mid-Jurassic) oceanic crust, and the term Neotethys to generally younger oceanic crust (i.e. Late Triassic-Early Cenozoic). The writer was unable to discriminate between the alternative tectonic models from the literature alone, and so decided to embark on a field-based study of the critical areas that has taken several years (Fig. 1, inset). There is no simple shortcut to understanding the pre-Jurassic tectonic evolution of the south Aegean region other than in-depth studies of the lithological assemblages in each of these areas, followed by comparisons and synthesis, which also takes account of evidence from the wider region and modern tectonic settings. A substantial body of new information has became available during this work, mainly concerning the sedimentary facies and palaeotectonic setting of the Upper Palaeozoic-Lower Mesozoic units in the region. The main results of a 10-year study of comparable units in western Turkey were recently summarized elsewhere (Robertson et al. 2002) and will be drawn on in the discussion section. The criteria for discriminating between tectonic settings are first outlined. The alternative

94

A.H.F. ROBERTSON

O

o

~

Z~

O

~o.-=

~

~eq . ,...~

+o

~.O

r~ ~

o

o

~

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS possible interpretations of each area are reviewed, and an indication of which model is favoured is given before moving on to the next area. Salient aspects of the wider regional setting, outside the area studied (e.g. Eurasian margin; central European Hercynian orogen) will be considered in the discussion section. Many of the Upper Palaeozoic-Lower Mesozoic units of Crete and the SW Peloponnese, in contrast to western Sicily and Evia, have undergone HP-LT (blueschist-facies) metamorphism (Seidel 1978). For example, the extensive Phyllite-Quartzite unit in Crete was at least partially metamorphosed under high-grade conditions (8-19 kbar, 300M00~ during Late Oligocene-Early Miocene time (Seidel et al. 1982; Theye et al. 1992; Zulauf et al. 2002). However, primary sedimentary lithologies and sedimentary structures, including stratigraphical way-up evidence, are still commonly recognizable. For this reason, rock types will be generally referred to here in their pre-metamorphosed states, dropping the ubiquitous 'meta-'; i.e. psammitic schists were commonly sandstones, marbles were limestones or dolomites, and pelitic schists were mudstones, etc. An informal stratigraphical terminology is used, as different local names have often been used for similar units in different areas. The time scale is that of Gradstein et al. (2004). Coordinates given refer to the present day unless specified otherwise.

Criteria for recognizing tectonic settings A combination of biostratigraphical, sedimentary, igneous and structural evidence (termed tectonic facies; Robertson 1994) allows different tectonic settings to be distinguished. Some of the main, relevant tectonic settings are as follows. Divergence-related tectonic settings include rifts, failed rifts (aulacogens) and intra-platform basins. Sedimentary environments associated with passive margins and marginal platforms include siliciclastic shelves and carbonate platforms. Tectonic settings associated with spreading centres and oceanic basins include spreading ridges, abyssal plains, continental fragments, oceanic seamounts and oceanic plateaux. Conversely, tectonic settings associated with convergencerelated settings include supra-subduction zone spreading centres (i.e. many ophiolites), oceanic arcs, subduction-accretion complexes, fore-arc basins, intra-oceanic back-arc basins and intracontinental back-arc basins. Tectonic settings associated with collisional tectonic settings include intra-oceanic collision zones, foreland

95

basins and the sedimentary products of collision ('molasse'). Additional tectonic settings characterize strike-slip-related settings (e.g. pull-apart basin), which could also be relevant here. Recognition of such tectonic settings in the south Aegean region should allow the alternative tectonic models to be distinguished. The recognition of such tectonic settings in metamorphic terranes as in the south Aegean region is obviously difficult, but still possible where the protoliths of the sedimentary and igneous rocks can be recognized and where the sediments are reasonably well dated. As a cautionary note, however, it should be noted that metamorphic rocks that have undergone HP-LT metamorphism, like those of the south Aegean region, have been exhumed from a subduction zone setting so that parts of the original record may have been lost. Also, some subduction settings involve net loss of material from the overriding plate (i.e. subduction erosion) such that some critical tectonic units (e.g. accretionary prisms) may be lost. The main tectonic settings that would be expected to occur for each of the four main alternative tectonic settings of the south Aegean region are as follows. In a divergence (rift)-related model (Models 1 and 3) the tectonic facies would be those of rifts, passive margins and Atlantictype ocean basins. In a convergence (subduction)-related model (Models 2 and 4) the expected tectonic settings for the Triassic would identify both active margin (e.g. subduction complexes; magmatic arcs) and collisional settings (e.g. foreland basins). Also in these models, additional divergence-related tectonic settings would characterize the Late Palaeozoic, inferred breakup and spreading of 'Neotethys' adjacent to Gondwana. Time relations are therefore clearly critical to distinguish the tectonic alternatives. The southward subduction hypothesis (Model 3) should also be characterized by divergencerelated tectonic settings, but coupled with igneous geochemical evidence of subduction. Finally, the model invoking both southward and northward subduction (Model 4) would imply the existence of two belts characterized by convergence-related tectonic settings and two belts of subduction-related magrnatism, one intra-continental (Andean type) and the other intra-oceanic. In summary, in the extension-related models (Models 1 and 3) only a limited number of tectonic settings would be represented, whereas many more would need to have existed for the convergence (subduction)-related models (Models 2 and 4).

96

A.H.F. ROBERTSON

Tectonic units of the south Mediterranean region

Evidence from the Permo-Triassic of eastern Sicily

The entire south Aegean region and areas to the west, as exposed in the Italian region (e.g. Calabria and Sicily), comprise piles of thrust sheets that were mainly emplaced during Cenozoic time related to northward subduction and suturing of the Neotethyan ocean. In this paper, units will be discussed in turn, working structurally upwards on a regional basis, beginning with those at the structural base that restore closest to Gondwana and ending with those that restore furthest north. As noted above, the most southerly unit, representing Neotethyan crust that formerly separated North Africa from the Cretan units, has been subducted or is located deep beneath the Sea of Crete and is not exposed and has not been sampled by drilling. The main evidence for the existence of this southerly oceanic basin is the record of subduction obtained from the Mediterranean Ridge accretionary complex south of Crete (Camerlenghi et al. 1995; ChaumiUon & Mascle 1997), and the record of Cenozoic H P LT metamorphism within the Cretan nappes (Seidel 1978). A history of rifting is documented by wells and exposures in North Africa to the south (Guiraud et al. 2001) but it is not possible to determine from this when spreading of an adjacent southerly Neotethyan ocean began; possibilities include Late Permian, Mid-Late Triassic or Late Jurassic-Early Cretaceous. For this reason it was decided to study the Late Palaeozoic-Early Mesozoic of Sicily as a proxy for the oceanic basin between North Africa and Crete. This is reasonable, as the entire North African margin from the Nile to Morocco shows evidence of a comparable history of rifting during Late Palaeozoic-Early Mesozoic time and, indeed, some workers restore the Sicanian basin (Sicily) of this age to a location south of Crete or even further east (e.g. Garfunkel 2004). In Sicily, the Cenozoic thrust belt exposes units that formed in a southerly, Sicanian basin bordering Gondwana from the inception of this basin, during Late Palaeozoic time. There is thus an opportunity to determine the tectonic setting of the North African margin in this region during this period. In particular, does the lithology present record a rift setting as in Models 1 and 3 or a spreading-related setting as in Models 2 and 4? In the discussion below relevant aspects of the geology of western Sicily will be considered and it will be shown that the available evidence supports Model 1 and that there is no firm evidence in support of Models 2, 3 or 4.

Relevant outcrops occur in two main areas: Lercara-Roccapalumba and in the Sosio Valley (Fig. 3). These units are unmetamorphosed, in contrast to many of those of the south Aegean, which will be discussed later. The Permo-Triassic units of western Sicily were finally emplaced as a result of Late Neogene (Miocene-Pliocene) southward thrusting over the North African continental margin (Catalano et al. 2000a, b). This took place during closure of the Mesozoic Piedmont-Ligurian ocean to the north. Field evidence is augmented by results from local shallow drilling (e.g. in the Sosio Valley) and from hydrocarbon exploration drilling (e.g. Roccapalumba1 well) (Catalano et al. 1991). It will be argued that the sedimentary sequences are most consistent with a progressively deepening rift basin that nevertheless continued to be supplied periodically with shallow-water carbonate debris and terrigenous elastic deposits. L o w e r P e r m i a n clastic s e d i m e n t s a n d basic igneous r o c k s

Exposures are few and far between (e.g. in railway cuttings) in an area of rolling hills and farmland. The main lithologies are terrigenous and carbonate turbidites, siltstones and shales, with subordinate detached blocks and coarse carbonate debris flows of mainly shallow-waterderived material. The subsurface thickness exceeds 1000 m (e.g. in Roccapalumba-1 well), although stratal repetition is possible. Exposures are commonly highly deformed, steeply dipping, or inverted. Typically, thinner bedded units are strongly sheared, whereas thicker-bedded, more competent units have survived as undeformed beds, cemented by calcite spar. The mudstones within the turbiditic sequence are dated as Early Permian (Artinskian) by conodonts (Catalano et al. 1991). Occasional alkaline basic sills have been reported (Di Stefano & Gullo 1997). The Early Permian elastic sediments were examined in small exposures, SSW of a railway station, SE of Roccapalumba (c. 300 m west of the River Torto). Mudstones are greyish to reddish in colour and include deep-water tracefossils of the Nereites ichnofacies (Kozur et al. 1996). The mudstones are intercalated with thinbedded (5-10 cm) siliciclastic turbidites, exhibiting well-developed, partial Bouma sequences. Individual sandstone turbidites reach 1.6m in thickness in this area. Thin-section study shows that carbonate grains are typically more

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

97

Fig. 3. Upper Palaeozoic-Lower Mesozoic sedimentary and volcanic units exposed in central western Sicily. (a) Location; (b) sketch section of unit exposed in the Sosio Valley; (c) simplified sedimentary logs; (d) possible tectonic setting. (See text for explanation and data sources.). numerous than terrigenous ones. In general, the sandstones contain quartz and muscovite, with subordinate biotite and feldspar and rare zircon (Di Stefano & Gullo 1997).

Thin-section study shows that both the terrigenous and carbonate grains are mainly angular to sub-angular. The terrigenous grains are mainly monocrystalline quartz, polycrystalline quartz

98

A.H.F. ROBERTSON

(quartzite), mica-schist, plagioclase (mainly altered plagioclase and perthite), muscovite (including large unstrained laths), biotite and rare zircon. Occasional large grains containing plagioclase, quartz and biotite were probably derived from granitic rocks. There are also grains of weakly recrystallized quartzose sandstone, with individual grains set in a matrix of microcrystalline silica. Siltstone rip-up clasts, rich in quartz and muscovite, are also seen, together with rare grains of reworked pelagic micrite with calcitereplaced radiolarians and detrital microcrystalline quartz (?metachert). Some quartz grains are coated with calcareous algae, indicating a shallow-water origin. Carbonate grains are dominated by detrital grains of algal micrite and encrusting calcareous algae together with pisoliths, small oncolites, grapestone intraclasts, echinoids (some coated with calcareous algae), shell fragments, coral (replaced by coarse calcite spar), bryozoans, benthic Foraminifera (e.g. Miliolina), ostracodes, gastropods and recrystallized carbonate (marble). The matrix is micritic with scattered diagenetic pyrite. There are also occasional interbeds of carbonate debris flows (up to 1 m thick), containing clasts (up to 2cm in size) of mainly neritic carbonate, including coral, encrusting algae, gastropods and fusulinids set in a partially recrystallized calcite spar cement. Rare blocks of deep-water carbonate contain an Early Permian fauna, including ammonoids, trilobites, brachiopods, crinoids and Radiolaria. In general, the siliciclastic turbidites tend to be uniformly fine to medium grained, whereas the associated redeposited carbonates are considerably coarser grained, suggesting a more proximal origin. In the Leccara area there are several exposures of diabase and basalt, forming sheets up to 30 m thick (e.g. Contrada Rettino body, c. 3 km SW of Roccapalumba). These are interpreted as high-level intrusions into wet sediments. They contain phenocrysts of altered plagioclase, olivine, A1-Ti augite and exhibit an enriched mid-ocean ridge basalt (E-MORB) composition (Censi et al. 2000). Igneous rocks of similar age and composition are present in the Hyblean area (Bianchini et al. 1998). M i d d l e P e r m i a n succession

Key outcrops of deep-water sediments of MidLate Permian age occur in the classic Sosio Valley, SW of Palazzo Adriano (Fig. 3a). These units are highly disorganized and were mapped as tectonic m61ange, 'up to 500 m thick, related to Late Neogene thrusting (Di Stefano & Gullo 1996, 1997). The Permian-Triassic deep-sea sediments in this area are sheared and imbricated,

and may include the deformed limb of a large inverted isoclinal fold. The stratigraphically lowest unit is composed of turbidites that are compositionally similar to, but younger (i.e. earliest Mid-Permian) than those of the Lercara-Roccapalumba area, described above (Catalano et al. 1991). Local successions, up to several hundred metres thick, in the Sosio Valley are highly disorganized and have been interpreted as olistostromes (Catalano et al. 1991). Detached blocks are strewn through a shaly matrix, which locally contains a reworked Early Permian fauna. The individual blocks exhibit layer-parallel extension of thick turbiditic beds to produce classic sandstone 'phacoids'. Similar 'olistostromes' are exposed slightly further upstream to the SE, in tectonic contact with contrasting red mudstones of Late Permian age (Fig. 3b and ci). Several of the large detached sandstone blocks exhibit sheared and slickensided margins, and internal brecciation, showing that they were well lithified before being incorporated into the 'olistostrome'. This questions whether these are really sub-aqueous debrisflows of Mid-Permian age or instead the uppermost levels of a thick Early-Mid-Permian turbiditic sequence that was sheared strongly to create m61ange during the Cenozoic thrusting. Thin-section study of the Middle Permian turbiditic clastic sediments shows that they are more mature, both texturally and chemically, than the turbidites of the Lower Permian interval described above. The terrigenous grains range from sub-angular, to rounded and very well rounded. Grains are mainly unstrained monocrystalline quartz with, in addition, rare strained quartz, plagioclase, fine-grained polycrystalline quartz (quartzite) and rare grains of partially recrystallized siltstone. The carbonate bioclasts are, by contrast, relatively angular, and include all of the neritic grains (e.g. lithoclastic algal micrite) as in the underlying Lower Permian redeposited sediments, with the addition of numerous radiolarians. The matrix is partially recrystallized micrite, with scattered diagenetic pyrite. Upper P e r m i a n succession

The 'olistostrome' is faulted against a younger unit composed of bright red, weakly consolidated, glutinous claystones that contain a rich fauna of Late Permian radiolarians, conodonts and ostracodes (Catalano et al. 1991; Kozur 1993; Fig. 3cii). These red mudstones are interbedded with several thin, graded interbeds of redeposited neritic carbonate (< 15 cm thick), containing Radiolaria, Foraminifera and conodonts of Late Permian age. The packstonegrainstones also contain minor amounts of

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS quartzose silt and sand (Catalano et al. 1991). Thin sections studied reveal mainly redeposited grains of neritic carbonate, as in the underlying Permian coarse clastic facies, especially algal micrite, together with shell fragments, benthic Foraminifera and reworked grains of radiolarian micrite. M i d d l e - U p p e r Triassic succession

The Middle-Permian succession is faulted against a contrasting, mainly pelagic carbonate succession of Mid-Late Triassic (Late Anisian-MidCarnian) age (Fig. 3b). The Lower Triassic (Scythian) succession is typically absent in Sicily (Catalano et al. 1991). The presence of wellgraded beds shows that the Triassic succession is inverted. It begins with thin- to medium-bedded, greenish to dark grey radiolarian mudstones, together with tuffaceous and siliceous pelagic limestones (in beds <20cm thick), of Late Anisian-Early Ladinian age, and then passes into grey to pink nodular and cherty limestones with marly partings (Catalano et al. 1991). The lowest interbeds are dark and apparently organic rich. Lenticles and nodules of quartzitic chert of replacement origin are common near the base, but generally decrease in abundance stratigraphically upwards. The highest exposed beds, of Late Ladinian-Early Carnian age, are thin-bedded pinkish pelagic carbonates, with minimal chert. Thin-section study shows that the carbonates begin with dark grey terrigenous silt, compositionally similar to the material in the underlying Permian terrigenous sediments, dominated by quartz, muscovite and subordinate plagioclase. There are also abundant calcite-replaced radiolarians and recrystallized shell fragments. The directly overlying silty limestone is composed of radiolarians, mainly replaced by microcrystalline and chalcedonic quartz, together with well-aligned straight-shell fragments (Halobia), set in a micritic matrix. Pink and grey pelagic carbonates above this are dominated by pelagic micrite (with no terrigenous component), with calcified radiolarians, Halobia and Daonella pelagic bivalves, pelagic gastropods, ostracodes and rare benthic Foraminifera. Coeval P e r m i a n pelagic and neritic carbonates

Several small limestone blocks are located further NE (downstream) in the Sosio Valley, and may record contrasting slope or fault-block-type facies (Catalano et al. 1991; Kozur 1995). Notably, the 'Rupe del Passo di Burghio block' comprises Middle Permian (Wordian),

99

white to grey, ammonoid- and conodont-bearing pelagic limestones (Hallstatt-type facies), with interbeds of redeposited bioclastic calcarenites (Kozur 1995; Di Stefano & Gullo 1997; Fig. 3ciii). In addition, a limestone block (SE of Rupe del Passo di Burgio) includes yellow to green clays with conodonts of Mid-Permian age (Catalano et al. 1991). There are also several other blocks of condensed rosso-type facies. Kozur (1995) suggested that these blocks record remnants of Permo-Triassic condensed successions that were derived from widely differing areas of Palaeotethys, characterized by the presence or near absence of a Pseudofurnishius (conodont) fauna. Alternatively, the blocks originated within several different palaeoenvironments, but all within a relatively local rift basin setting (Di Stefano et al. 1996). The relative abundance of conodonts was instead controlled by facies variation and sedimentation rate variation (i.e. dilution effects) in the latter view. The deformed Permian-Triassic deep-water succession is structurally overlain by a large block of highly fossiliferous limestone, the wellknown Pietra di Salmone (Fig. 3civ). This unit is interpreted as the dismembered limb of a WNW-facing monoclinal fold (Flugel et al. 1991). This large block is dominated by a generally fining-upward sequence of redeposited neritic talus (c. 70 m thick). The carbonate talus ranges from disorganized, angular blocks and clasts (up to 3.5 m in size), to crudely stratified clast- to matrix-supported debris flows, to pebbly and to gravel-sized calciturbidites with marly tops. The constituents include a rich and diverse fauna of corals, algae, sponges, Foraminifera and conodonts. Some of the clasts, dated as EarlyMid-Permian, were redeposited into a matrix of yellowish, partly dolomitized marls containing conodonts of Mid-Permian age. Carnian a n d younger successions

Succession of Late Triassic and younger Mesozoic age are well exposed in a more intact, structurally associated succession (e.g. Pizzo Mondello; Di Stefano & Gullo 1997). These sediments are dominated by marls, claystones and thin-bedded calcilutites and contain Radiolaria, Halobia and conodonts ('Mufara Formation'). Comparable facies, up to c. 400 m thick, occur widely in central-western Sicily and include black organic-rich anoxic facies. There are also redeposited neritic carbonates, probably derived from large carbonate platforms (e.g. Hyblean Plateau; Di Stefano et al. 1996). Quartzose sediments, including polycrystalline quartz, muscovite and feldspars, are present in some sections and

100

A.H.F. ROBERTSON

appear to increase in abundance northwards, suggesting derivation from a metamorphic basement exposed in this direction (e.g. Hercynian Kabilo-Peloritani units). The succession continues upwards into uppermost Triassic pelagic Halobia-bearing limestones, commonly siliceous, with continued interbeds of redeposited neritic carbonate. Locally, the Upper Norian interval exhibits huge slump breccias associated with terrigenous clastic input and a low-angle unconformity (e.g. Imerese unit, northern Sicily; Di Stefano et al. 1996). A strike-slip controlled setting was suggested by Di Stefana & Gullo 1997), although an associated unconformity in adjacent areas (e.g. Tunisia) might also be explained by a pulse of extensionrelated flexural uplift. Lower Jurassic cherty pelagic limestones and marls above this are followed by a more or less continuous deep-water succession dominated by radiolarian marls and pelagic limestone, with locally variable intercalations of redeposited neritic carbonate (e.g. Imerse and Sicanian units). Widespread volcanism occurred in the Late Jurassic within some basinal (e.g. Sicanian) and platform (e.g. Hyblean) units. Some Late Triassic Mesozoic platforms were drowned and covered by Ammonitico Rosso (e.g. Hyblean platform), whereas neritic platform deposition persisted elsewhere (e.g. Sicilide units). Passive margin conditions persisted until the end of Mesozoic time, after which convergence began (Catalano et al. 2000b). Interpretation." a s u b s i d i n g d e e p - w a t e r rift

The thick, Lower Permian sandstones record accumulation in a deep-water basin (Fig. 3d), bordered by a relatively local carbonate platform and a metamorphic landmass, presumably the Hercynian orogen to the north (e.g. in the Peloritani Mountains, NE Sicily; Di Stefano & Gullo 1997). The source area included crystalline basement (e.g. mica schist), granitic rocks, low-grade siliciclastic sediments and metacarbonates. The inception of the basin must significantly predate the earliest known Early Permian deep-water sediments. The Middle Permian red claystones accumulated in a fertile relatively deep sea, with deep-water currents connecting to the eastern Tethys (e.g. Crete; Oman; Catalano et al. 1991; Kozur 1993). The thin calciturbidites with neritic detritus attest to the proximity of a Late Permian carbonate platform. This platform and related slope possibly included the Pietra di Salomone and several other local limestone blocks in the Sosio Valley. The Pietra di Salomone unit began to accumulate on a carbonate ramp bordering a well-established carbonate platform during

Mid-Permian time. The slope later steepened, shedding coarse talus, then stabilized by latest Permian time. The Petra di Salomone 'megablock' specifically represens the fill of a steepsided channel or canyon cut into the margin of a carbonate platform (Flugel et al. 1991). By contrast, the blocks of pink pelagic Hallstatt-type facies accumulated on isolated highs within the basin (e.g. Bernouilli & Jenkyns 1974), but the nature of their basement is not known. It is probably not necessary to invoke tectonic transport of units from quite different oceanic areas, as suggested by Kozur (1993, 1995), but instead different facies may represent deposition in different local palaeo-environments with different sedimentation rates and variable preservation of fossils. The Middle Triassic (Ladinian) interval then documents a quiet pelagic carbonatedepositing setting, in a still-fertile sea, rich in radiolarians. Although no unbroken succession is preserved, it is likely that, prior to Neogene thrusting, a remarkably long-lived, more or less continuous deep-water succession existed, extending from Early Permian to Early Cenozoic time. No base of the succession is exposed and the Lower Permian siliciclastic turbidites are relatively distal, in contrast to the coarser neritic carbonate material that was supplied from a more local carbonate platform. The overlying 'olistostrome' could indicate a period of tectonic instability during Mid-Permian time according to Catalano et al. (1991). By contrast, stable seafloor conditions are indicated for Mid-Permian time when well-oxidized, fertile hemipelagic sediments accumulated in a deep-water basin (500 m or more) (Catalano et al. 1991). The Lower-Middle Permian turbiditic clastic facies might, in principle, represent different tectonic settings. The first could be a flexural foreland basin related to the Hercynian orogeny (Catalano et al. 1991; Ziegler & Stampfli 2001). A foreland basin setting is, however, unlikely as there is no evidence of a foreland basin stratigraphy (i.e. a thickening and coarsening-upward succession), of any overthrust load in the area, or of flexural rebound after a collisional event; rather, deepening continued during the Triassic. Second, the deep-sea sediments might record an accretionary prism (i.e. turbidites) and an associated perched fore-arc basin (i.e. pelagic carbonates) related to subduction. However, there is no evidence within the kilometre-thick siliciclastic turbidites of exotic accretionary material (e.g. pelagic sediments or ophiolites), or thrust repetition of the succession as in an accretionary prism. Third, the succession might record a rift-basin (?transtensional) that post-dated the Early Carboniferous Hercynian orogeny (Di Stefano et al.

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS 1996). Of these alternatives, a rift setting best fits the sedimentary record and is consistent with the presence of the intruded 'enriched' basaltic igneous rocks (i.e. E - M O R n ; Censi et al. 2000) and sparse Middle-Triassic silicic tufts. N-type M O R n that could indicate the existence of strongly stretched continental crust or oceanic crust in the vicinity is not recorded. A rift-related (or transtensional) setting can be related to post-Hercynian extension during Mid?-Late Carboniferous time. By Early Permian time a broad palaeogeographically diversified deep-water basin existed (Fig. 3d), open to

101

Tethys to the east. Condensed pelagic sediments probably accumulated on isolated horsts that were bypassed by gravity flows. By the Early Anisian there was a general switch to hemipelagic carbonate deposition, possibly reflecting increased input of peripelagic carbonate from adjacent carbonate platforms. Continuing high siliceous productivity is suggested by the presence of replacement chert. Variable, deep- or shallowwater carbonate conditions persisted throughout Mesozoic time, punctuated by periodic tectonic instability that triggered mass flows on the margins of carbonate platforms (Di Stefano & Gullo

S

N

N. Africa

Mediterranean ridge . -

.

. .

Gavrovo - Tripolitza (upper units)

-,

ic Middle Miocene - Recent

Ionian Sea ;. s

,r

j

..

,/7 =

j

Exhumation of lower HP/LT units

..

- ,

,..,0.

b Miocene

Late Olig.- E. Mio underplating of HPILT units

Late E o c e n e / E a r l y Oligocene of upper (LP/LT) units . ~ "Sz 'thrusting ~ Pindos & overlying units oa

0o

9 ". ". : : - . ~ <"~" ,,, , . t,,"~,~";-.,.:-_~,~.~"..t ~-~,,:._.~,.:-"-:~. - .'" ~li~" ,". ~ ' ~ t / ~ / ~ . ~ . j /

a Late Eocene - Late Oligocene

~j

" ,'-"~

Phye~tit-l Quartzite &

underlying

~

Fig. 4. Interpretative sketch sections to illustrate the Cenozoic Alpine evolution of the South Aegean, which was dominated by collision and exhumation. The effects of the resulting deformation and metamorphism have to be removed to allow interpretation of the Upper Palaeozoic-Lower Mesozoic units discussed in this paper. (See text for discussion and literature.).

102

A.H.F. ROBERTSON

1996, 1997). Widespread volcanism and subsidence of carbonate platforms (e.g. Hyblean platform; SE Sicily) during Mid-Jurassic time was broadly coeval with spreading of the central North Atlantic. This rifting could also relate to break-up and the onset of sea-floor spreading to form the Ionian Sea (Catalano et al. 2001), with implications for the timing of spreading further east (e.g. Sea of Crete). There is no evidence of the existence of an oceanic basin in the Sicily area until this time. In summary, the sedimentary facies and igneous rocks exposed in western Sicily are indicative of a rift or failed rift (aulacogen) setting as in Model 1. A deep-water basin open to the Tethys to the east clearly existed but there is no evidence of oceanic crust (ophiolites or related hydrothermal deposits), or of a subsiding passive margin in which an overall thinning- and deepeningupward sedimentary succession would be expected (e.g. similar to the NW African or the eastern USA Mesozoic continental margins). Instead, a slowly subsiding rift persisted until this was reactivated related to opening of the central North Atlantic during Mid-Late Jurassic time. The main conclusion of fundamental importance from Sicily carried over to the south Aegean region, discussed below, is that there is no evidence for the existence of a Late Palaeozoic ocean basin bordering North Africa in this region.

Cenozoic setting of the South Aegean region Crete and the Peloponnese document palaeotectonic units that were located to the north of a now largely subducted southerly Neotethyan ocean basin that rifted along the North African continental margin. In contrast to Sicily, the Upper Palaeozoic-Lower Mesozoic units there have experienced metamorphism ranging from HP-LT in the structurally lowest units to LP-LT in the higher units. Any interpretation of the pre-Cenozoic tectonic setting depends critically on correctly restoring the effects of Early Cenozoic-Recent deformation and metamorphism. During the Early Cenozoic, Neotethyan basins progressively closed, affecting, in turn, the Vardar zone to the NE, the Pelagonian zone, then the Pindos zone and finally Apulia-North Africa to the SW (e.g. Dercourt et al. 1993; Robertson et al. 1999). The higher thrust sheets, the Pindos zone, and the Tripolitza platform and the units above were largely detached from their former Mesozoicearly Cenozoic stratigraphic cover associated with the later stages of closure of Neotethys. They were then tectonically assembled during Late Eocene-Early Oligocene time related to generally northward subduction. The lower,

more southerly derived nappes are represented by the Plattenkalk, Tripali and Phyllite-Quartzite units; these units were underplated to the hanging wall of a subduction trench during Late Oligocene-Early Miocene time, resulting in H P LT metamorphism, internal thrust imbrication, large-scale structural inversion and polyphase folding. Fission-track dating (Thompson et al. 1998) indicates that crustal thickening culminated in bi-vergent ductile extensional exhumation during Late Oligocene-Early Miocene time (24-15 Ma) (Fig. 4b). The lower nappes experienced extension-related folding, shearing and large-scale detachment near the contact with the overlying unmetamorphosed Pindos and Tripolitza nappes and also internally (Bonneau 1984; Papanikolaou 1988; Fassoulas et al. 1994; Kilias et al. 1994, 2002; Jolivet et al. 1996; Zulauf et al. 2002). The lower nappes in western Crete were rapidly exhumed, whereas exhumation was possibly slower in central Crete, associated with retrograde metamorphism (Fassoulas et al. 1994). Subsequently, Mid-Miocene-Recent time was dominated by pulsed extension (Fig. 4c) in an above subduction zone setting, with an important component of strike-slip-transtension in eastern Crete (e.g. ten Veen & Meijer 1999). In Crete (Fig. 5), the Tripali and Plattenkalk units are discussed first as they restore to a more southerly position than the Phyllite-Quartzite unit when the Cenozoic tectonic effects are removed.

Tectonostratigraphy of Crete In Crete, the lowest exposed unit in the tectonostratigraphy is the Mesozoic-Lower Cenozoic Plattenkalk unit (Creutzburg et al. 1977; Fig. 6). The Plattenkalk is dominated by grey platy pelagic limestones, which contain numerous chert nodules. The lower part of the Plattenkalk unit is represented by the Talea Ori unit, best exposed in central northern Crete (Krahl et al. 1988; Fig. 7). The Talea Ori is dominated by Upper Palaeozoic shallow-marine terrigenous and carbonate facies above a probable continental basement. The next unit above the Talea Ori-Plattenkalk unit is the Tripali unit. Based on recent studies in SW Crete (e.g. Lefka Off) Krahl & Kauffman (2004) suggested that the Tripali unit extends in age from the Early Liassic to Late Cretaceous (or even Early Cenozoic; see below). In the type area, in the Tripali Mountains, the Tripali unit forms a separate tectonostratigraphic entity (Kopp & Ott 1977). However, in western Crete, an important mainly metacarbonate and meta-evaporitic succession of Triassic age was initially referred to the Tripali unit (Gips-Rauhwacke Formation of Wurm 1950) but was later reinterpreted as an

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

103

"

~r

Evia

t N

0

"

',~

~

1

Phyllite-Quartzite unit

LE~ Plattenkalk & Tripali units 0

Zarouhla

r

~d. ~ ;

!

',3

0

D

"Tyros

20 KM I

U

~

0

Molai

,~

o

"Fripolitza zone

[~

~

d~

Cyclades Cb ~:~{~

J

Tagetosl t~ ' S Peloponnese

~) v

c~

I

Kythira

i~ Talea Ori N Crete unit ~.lkJ~ ~ .~. Chemezi . II/.:::.:..-~r,._.,_ I : . ~ A ~ " - ' x ~/~_~ ' l ~ _ . l k ,Iraklion ~j"] W .L,. re[e ~ ~-~ 'kaSiti - a z~u

AA

IL~i~l~

~

Fdpali unit Flora S f a ~ ' P , ~jloritis M.Preveli Mtns ,,

I

~v~i

E Crete ..........

Fig. 5. Outline map of Crete and the Southern Peloponnese showing the main locations of the Upper Palaeozoic-Lower Mesozoic Phyllite-Quartzite unit and related units. A cross-section of north central Crete is shown in Figure 7. More detailed maps of areas within boxes are shown in subsequent figures.

upward continuation of a regionally inverted succession of the Upper Palaeozoic-Lower Mesozoic Phyllite-Quartzite unit (Krahl et al. 1983b; see below). The Phyllite-Quartzite unit includes Middle Carboniferous-Lower Cenozoic, mainly siliciclastic successions that are most widely exposed in western Crete (Fig. 5). In addition, contrasting units of 'Hercynian' basement slices, Permian hemipelagic sediments, Triassic volcanicsedimentary units, and Triassic shallow-marine to non-marine sediments are well exposed in eastern Crete. The Phyllite-Quartzite unit is, in turn, overlain by the relatively unmetamorphosed Tripolitza and Pindos units (Fig. 6). In Crete, the

Tripolitza unit, in places begins with MiddleUpper Triassic dolomites, schists and shallowmarine carbonates ('Ravdoucha beds'; Kopp & Wernado 1983) and then passes into a thick shallow-water platform carbonate unit that culminates in Upper Eocene turbidites (Creutzburg et al. 1977; Fleury 1980). The Ravdoucha unit is well exposed in several areas in western Crete (e.g. the south coast at Sellia), where dark mudrocks and carbonates with occasional ammonites pass transitionally upwards into the Tripolitiza carbonate platform (Fassoulas 2001). A comparable but thicker unit in the Peloponnese, known as the Tyros unit, includes extensive Triassic volcanogenic rocks and terrigenous sediments (see below).

104

A.H.F. ROBERTSON

S

N

OVERTHRUSTING

~.

,=,,. ' ~ ' ~ . ~ . ~ l i t i c

EXHUMATION

units

~ -

.,,.

Fig. 6. Stacking order of tectonic units in Crete. (See text for data sources.) The lower thrust sheets experienced Cenozoic HP-LT metamorphism. The overlying Tripolitza platform was detached from its substratum related to Early Cenozoic northward subduction. The detached lower units were deformed and imbricated in the subduction zone, then exhumed. Talea Ori Mtns.

1 0.5

Km "" HP/LT units

SE

NW

Tripolitza & Pindos 0 '

1 Km '

~

Phyllite - Quartzite

I~

Plattenkalk& Tripali (Talea Ori Unit)

Fig. 7. Simplified cross-section of central northern Crete showing the main thrust sheets, or nappes. (See text for explanation.) Arrows indicate bivergent extensional exhumation. Modified from Fassoulas (2001). The overlying Pindos unit includes Triassic deep-water sediments (e.g. pelagic limestones, radiolarites) and extrusive rocks (lavas, volcanielastic rocks and tufts), Jurassic deep-water sediments and ophiolitic rocks, and culminates in deposition of terrigenous turbidites of Paleocene-Eocene age (Krahl et al. 1982; Bonneau 1984). The Pindos and Tripolitza units are locally downfaulted in direct juxtaposition with the Phyllite-Quartzite unit, related to midCenozoic tectonic exhumation, as seen in western Crete. The Pindos unit is overlain by Upper Cretaceous metamorphic rocks, which are well

exposed in central southern Crete (Asteroussia nappe), and finally by Upper Jurassic dismembered ophiolitic rocks (e.g. serpentinitized peridotite and gabbro) at the highest preserved levels of the thrust stack. The palaeotectonic settings of the lowermost units, the Talea-Ori and Plattenkalk units are next outlined.

Talea Ori and Plattenkalk units These units include Upper Palaeozoic-Lower Mesozoic continental margin-type sediments and deep-sea sediments that restore to the northern

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS margin of the now largely subducted southern Neotethyan oceanic basin that rifted along the North African margin. In both Models 1 and 2 these two units would be expected to record divergence-related tectonic facies but in a different tectonic context. In Model 1 these units would represent part of the northern passive margin of Gondwana that rifted in Late Palaeozoic time, whereas in Model 2 they represent the southern margin of a rifted 'Cimmerian' fragment, which evolved as a subsiding passive margin during the Mesozoic-Early Cenozoic. The inferred latest Triassic 'Cimmerian' collision to the north could have affected this margin (i.e. by creating a distal foreland basin setting). In Model 3 an entirely Triassic (back-arc) rift would be expected, whereas a Triassic convergent margin might have existed in Model 4. Below, it will be shown that the sedimentary record favours Late Palaeozoic rifting with a further strong pulse of rift-related subsidence in latest Triassic-Early Jurassic time. The older facies of the Plattenkalk, represented by the much-discussed Talea Ori unit (e.g. Hall & Audley-Charles 1983; Hall et al. 1984), is well exposed in the Talea Ori Mountains of central northern Crete, west of Iraklion (Fig. 8) and in the Psiloritis Mountains further south (Fig. 5). The succession exposed in the type area (Fig. 8a) is structurally inverted. The Talea Ori unit is reported to begin with Upper Palaeozoic clastic sediments (Krahl et al. 1988) that contain detrital zircons as young as 297-325+_5 Ma; i.e. of inferred Hercynian age (Zulauf et al. 2002; Brix et al. 2002; Romano et al. 2002, 2004). The section stratigraphically above comprises phyllites and sandstones with minor cherty horizons (Galinos beds), of inferred Early Permian age. These sediments are then overlain by neritic dolomites and limestones (i.e. Fodele beds), dated as Late Permian. The succession continues with mainly clastic dolomites, limestones and interbedded clastic sediments (Sisses beds), passing into Upper Scythian marbles (Epting et al. 1972). The Middle-Triassic interval exhibits a depositional hiatus (Epting et al. 1972) marked by subaerial exposure, karst development, diachronous erosion and local bauxite development. This was followed by marine transgression with deposition of Norian stromatolitic dolomites and massive dolomitic carbonates. Carbonate breccias then record a further discontinuity. Champod et al. (2004) reported that Upper Triassic facies are transgressive on various different underlying units, down to the level of Late Permian platy pelagic limestones with chert nodules. In most interpretations the Upper Triassic facies are inferred to pass depositionally upwards into the typical deeper water Plattenkalk carbonates. However, Krahl & Kauffman (2004) considered

105

that these pelagic carbonates are not part of the Plattenkalk unit s e n s u s t r i e t o but rather representative of a pelagic facies that occurs in several different units. They now consider the Talea Ori unit, of Late Carboniferous-late Early Cretaceous age, to be a separate tectonic unit from the Plattenkalk unit although this is not supported here. During this work detailed sedimentological observations were made, particularly on the Upper Palaeozoic Fodele and Sisses units and the overlying Triassic carbonate succession that shed light on the depositional and tectonic setting. The Permian Fodele beds (e.g. 1 km east of Sisses) are dominated by regularly dipping, little deformed alternations of darker and lighter coloured, locally pebbly limestones with numerous shelly and bioclastic layers (Fig. 8). Highly fossiliferous dark limestones include well-preserved large brachiopods, coral and gastropods. The bioclastic material is suggestive of a storm-influenced death assemblage. In the succession studied, north of Fodele village, the base of the succession is in tectonic contact with phyllites of the Phyllite-Quartzite unit. There, the lowest exposed beds of the Sisses unit are very coarsely crystalline marbles intercalated with buff phyllites. These phyllites are lithologically similar to the underlying Phyllite-Quartzite unit so that the exact location of the thrust contact is indeterminate. A similar alternating carbonate-clastic succession is well exposed further east (0.5 km), along the national road, where thick-bedded shales and limestones are interbedded, although the contacts are commonly sheared. A major carbonate conglomerate (c. 4.5 m thick) contains well-rounded clasts, up to 3 cm in size, set in a buff-coloured matrix of phyllite. Pebbly phyllitic interbeds contain well-rounded clasts (up to 5 cm in size). These matrix-supported conglomerates form laterally continuous units, up to 6 m thick. Close to the inferred tectonic contact with the PhylliteQuartzite unit small thrust slices of carbonate rocks and shale include small slices of dark bioclastic and nodular limestone, correlated with the Fodele unit. The matrix-supported conglomerates are suggestive of deposition in a slope setting. Stratigraphically above, the Sisses unit is dominated by neritic carbonates (Fig. 8). The higher levels of the Sisses unit are uniformly finegrained, thick-bedded to massive carbonates, ranging in colour from white, grey and buff to pink. Massive beds, commonly rich in vugs, appear to include boundstones. Locally, thin beds (c. 10 cm thick) exhibit sand-sized clastic interbeds (up to 10 cm thick), showing grading and fine parallel lamination. There are also several thin micro-breccias dominated by small platy

106

A . H . F . ROBERTSON [~

Karst & bauxite Evaporite L.Te.? I:: :.'"i.I

Volcanic rocks I~ [~

i ~ 'iTM

Replacement chert Redeposited carbonate

Io

i

I o I I i 9

I *

.

''

" ' i

I

-

o" l I I I

.'..'

U.Cr.?~=; ' " ' ~

L.Te.J" ." -;" "'.'-'-I U.Cr.' ' '

Pelagic carbonate F~

'"'""~

Nedtic carbonate/stromatolitic rocks

I, "

L.Cr.

Pelagic siliceous shale

I

I

U.Ju.,

Terrigenous shale

|

9

-

I

[~

Conglomerate/(carbonate) oreccia Conglomerate (siliciclastic)

[-~

Basement

(ii)

L.Ju.U.Tr. L ; - I

gxY I Cam-

!~1

O

~ - - ~ , - ~ ~.

Cam.AnJs.

"O

Nor

I I-,-I

i-i-L I-

libl

a

L

i

I

I

U'Tr'l I

-

i

-J

I II I II

I o oV,,,,v.,.oI

I

Iv v v I

I---I

"

I,-.' T Tl I, I , I ,.l

;;

LJuu

~

n

i

U.Tr.

oJ o

II

I

a4a

II I

o

I "-~ I I 7= I I I1

I I iI ] U.Tr. LI..:..J. ,

u""r'"l

TRIPALI

I

C

I I.Itl U.Tc ~I I n'r

I I

M.Scy. | . . . . . . . L.Tr. I ~," ~' ~,~1

L

i

V-" V m "-V --v--v'-

U.Pe. ,, o o o

PHYLLITEQUARTZITE

Pe I - - I -

-I

Pe.

E CRETE

a

I

I

L.Tr.| , ~ ,~

U.Perm. I / "~ ' "f d

,7 I

4"1

!

I---I I---1

M.Tr,

F

|.

(i)

II

iD

i

,I I

, - , 9 " ~ =

],'"

I

U'Tr- I o

I

,. 9

..... ~ I , " : ",J'

Terrigenous sandstone

"" L','.'.',

I L

e

. . . . . . .

M.Po.

|

:2_----2. U. Caro.

TALEA ORI/ PLATTENKALK

d

r--( -i

l:.;.fl PHYLLITEQUARTZITE W CRETE

b Fig. 8. Summary of the main restored Upper Palaeozoic-Lower Cenozoic successions exposed in the lowest thrust sheets on Crete that experienced H P - L T metamorphism and are discussed in detail in this paper. (See text for discussion and data sources.)

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS carbonate clasts (< 1 cm long) of probable microbial origin (i.e. reworked microbial mats). The Sisses unit is overlain by very well exposed algal limestones, which are mainly medium to thick bedded, but include thin (several millimetres) lenticular siltstones, suggestive of current reworking. Above the hardground and bauxitic horizons (Epting et aL 1972; Fassoulas et al. 2004), an upper limestonemlolomite succession is dominated by dark foetid interlayers that increase in abundance upwards and culminate in very dark organic-rich carbonate. Dark stromatolitic carbonates and lighter vuggy carbonates are interstratified as alternating units c. 5-8 m thick. Massive vuggy white limestones follow with occasional pebbly horizons containing grey to white finely laminated carbonate clasts, up to 4 cm in size. Above this (near Aloides village), the succession is dominated by a spectacular highdensity carbonate turbidites, slumps, carbonate debris flows and coarse carbonate talus. Individual mega-breccia units reach 5 m thickness and include a range of very angular to wellrounded clasts, some > 1 m in size. Several depositional units (up to 2.5 m thick) comprise very well exposed carbonate debris flows grading into high-density type calciturbidites, showing evidence of dewatering, slumping and soft-sediment deformation. Most of the clasts are of lightcoloured carbonate facies but a few are dark, similar to the underlying intact carbonate succession. The highest levels of the redeposited facies comprise thick-bedded limestones with limestone conglomerates forming several metre-thick units with well-rounded limestone. Some beds are graded confirming stratigraphic inversion. The typical pelagic Plattenkalk facies then follows without a break (Fig. 8). The lowest of the Plattenkalk sediments are very thick-bedded (up to 0.8 m thick), pale grey crystalline limestones with numerous white chert lenticles and nodules. I n t e r p r e t a t i o n : a rift s e t t i n g

The lower part of the succession is consistent with a Permian transgression onto a Late Palaeozoic basement and input of clastic sediment in a tectonically unstable shallow-marine setting. This was followed by a marked regression, probably related to tectonic uplift during Early-MidTriassic time, which gave rise to the Mid-Triassic depositional hiatus. The thick organic-rich stromatolitic carbonates accumulated on a subsiding carbonate platform. This was followed, during the Late Triassic, by catastrophic collapse of the carbonate platform leading to the genesis of slumps, mega-breccias, carbonate debris flows and calciturbidites. The deep-water Plattenkalk

107

facies then began to accumulate as calciturbidites and relatively siliceous pelagic carbonates. In Model 1 the above succession records rifting of the North African margin, including a phase of flexural uplift in the Mid-Triassic, followed by a further major rift pulse in the Late Triassic. In Model 2, the succession (Galinos and Fodele units) records extension and subsidence related to opening of a Neotethyan ocean to the south. This was followed by flexural uplift, and crucially collision and post-collisional marine transgression during Mid-Late Triassic time related to and following closure of Palaeotethys (Champod et al. 2004). Here, the sedimentary record of the Talea Ori unit is considered to be consistent with Model 1 (or Model 3). The strong relative subsidence in the Late Triassic is attributed to a pulse of rifting to create a deep-water (Ionian) basin to the south. This pervasive rifting, fault scarp degradation, platform break-up and subsidence from a neritic to a carbonate-depositing setting during the Late Triassic is fully consistent with Model 1. However, these features cannot be explained as a flexural response to a Palaeotethyan 'Cimmerian' collision, as a contrasting flysch-type basin would be expected above the carbonate platform. Also, there is no evidence of the required foreland basin within the Plattenkalk, which restores to a more northerly position, as discussed below.

Tripali unit The regionally overlying Tripali unit (Fig. 8) is dominated by platform carbonates and is exposed in many mountainous areas (e.g. Levka Ori and Tripali Ori). The Tripali unit was traditionally seen as being restricted to very thick ( > 1000 m) successions of Upper Triassic-Lower Jurassic, partly dolomitized platform carbonates (e.g. Jacobshagen 1986). However, Krahl & Kaufmann (2004) reported an upward passage into siliceous metacarbonates (i.e. Plattenkalk-type facies) and metasiliciclastic sediments, extending in age from Early Jurassic (based on ammonites), to at least Albian (based on planktonic Foraminifera). Cenomanian planktonic Foraminifera and rudists occur in debris related to the emplacement of this unit over the Plattenkalk. This suggests that the original succession spanned the Late Triassic to Late Cretaceous-Paleocene. The higher, post-Triassic, part of the Tripali succession, for example, is well exposed along, and adjacent to, the main north-south road linking Vrisses with Hora Sfakion (Figs 5, 6 and 8). The succession structurally overlies the Plattenkalk in the north. The lowest part of the local succession exposed further south (near the turning to Asfendou) comprises massive to weakly

108

A . H . F . ROBERTSON

bedded, pale grey crystalline neritic carbonates, of inferred Late Jurassic age (J. Kxahl, pers. comm.). Upwards, there is an incoming of medium-grey to dark grey limestone with abundant chert nodules concentrated in finer-grained intervals. The limestones include thick-bedded matrix-supported bioclastic carbonate conglomerates, with clasts up to 5 cm in size, bivalve debris and belemnites. There are also several finely laminated, soft-weathering, shaly intervals. Upwards, the succession becomes increasingly thinly bedded and chert rich, and then passes into c. 20 m of brownish shale with chert-rich beds (up to 0.3 m thick), formed by almost complete replacement of thin carbonate beds. Above, there is a return to thicker bedded, purer limestone with abundant chert nodules. The highest levels of the succession, dipping southwards at a moderate angle towards the south coast, comprise medium- to thick-bedded (up to 2.9 m) dark, foetid, marble with abundant cannonball-shaped chert nodules. Individual beds (up to 0.3 m thick) are graded, with scoured bases and shaly tops. Higher parts of the succession, exposed further north (on the main road north of Kares near the col), are dated by a rare occurrence of Upper Cretaceous rudist limestone (J. Krahl, pers. com.). Thick-bedded coarse bioclastic limestones comprise subangular to subrounded carbonate clasts (up to 10 cm in size), including algal carbonate, and thin interbeds of pale grey calcilutite.

N TIP

~

. . . . .

Ir

Interpretation: a subsiding rift basin The Upper Triassic-Lower Jurassic limestones of the Tripali unit record a widespread neritic carbonate platform facies. By contrast, the deposition during Jurassic-Late Cretaceous time took place in deep water in a relatively proximal carbonate slope setting. Facies ranged from carbonate debris flows, to high- and low-density turbidity current deposits and siliceous hemipelagic carbonates. The distinctive shaly interval rich in replacement chert may correspond the regional Late Jurassic-Early Cretaceous interval of high siliceous productivity and chert formation (e.g. De Wever 1989). The Tripali unit can be restored as a distal equivalent of the Plattenkalk, hence its similarity in facies. The source of the redeposited carbonate material is inferred to be the Tripolitza carbonate platform, then located to the north. In Model 1, the Tripali unit records Late Triassic rifting to create a deep-water basin to the south in which deep-water slope and pelagic carbonates accumulated. The Tripali unit can be correlated with the margin of the deep-water Ionian rift basin in the Peloponnese, whereas the Plattenkalk can be correlated with the more distal deep-water pelagic carbonate facies of the Ionian basin which overlies stretched continental crust in the Peloponnese (e.g. British Petroleum Company Ltd 1971). This contrasts with Model 2, in which the Tripali and Plattenkalk units would

Recumbent nappe

-

S

sl.

- 3 Km I

!

I-/P Tripolitza & Pindos unmetamorphosed units Mana unit, Late Triassic

m

Gypsum, latest Triassic ? Dolomitic carbonates, shales, Late Triassic Thin- bedded limestones, shales, Early-Mid Triassic Quartzitic sandstones & phyllites, Late CarboniferousEarly Triassic & minor carbonates

Fig. 9. Simplified cross-section of western Crete showing the division of the Phyllite-Quartzite unit into an upper right-way up succession and a lower inverted succession. This distribution has been explained by the existence of a huge south-facing recumbent nappe (Krahl et al. 1983a, b, c). The structure was modified by high-angle faulting. (See text for discussion.)

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS be expected to overlie south-Neotethyan oceanic crust, for which there is no evidence in either Crete or the Peloponnese.

Phyllite-Quartzite unit of western Crete Lithofacies of the Phyllite-Quartzite unit as exposed in western and eastern Crete are substantially different and so will be discussed separately below. In general, the Phyllite-Quartzite unit of western Crete comprises a Middle Carboniferous to Lower Triassic deep-water siliciclastic sequence, with alkaline igneous rocks of mainly Early Triassic age. The overall succession shallowed upwards into an Upper Triassic, evaporitic sequence and an Upper Triassiclowemost Jurassic? conglomeratic sequence in different areas. In Model 1 (Fig. 2) the siliciclastic sequence represents a divergence-related tectonic setting, namely a Late Palaeozoic deep-water rift. A further pulse of rifting later gave rise to alkaline magmatism in this interpretation. The observed shallowing upwards is interpreted as the effect of flexural uplift related to break-up to form the Pindos ocean to the north. This inferred flexural uplift also affected the Plattenkalk (Talea Ori unit), as noted earlier. In Model 2 the Late Palaeozoic succession would record rifting of a Cimmerian fragment from Gondwana coupled with opening of Neotethys, whereas the Triassic shallowing-upward succession would record a foreland basin or collisional setting. In Model 3, the Triassic volcanic rocks would be expected to show a subduction influence, although the successions might be located too far south from the trench to show such an effect. In Model 4 there would be a greater chance that the volcanic rocks would be subduction related as they should directly overlie a northward-dipping subduction zone. The key discriminants are again the time and length scales of the inferred tectonic controls on the sedimentary sequences: how proximal or distal are the successions? Is there evidence of contractional deformation coeval with sedimentation, and what is the character and consequent explanation of coeval magmatism? It will be argued that the sequences are consistent with deposition in a rift setting, which then shallowed, rather than with a developing passive margin bordering an already extant oceanic basin or with any form of convergent setting. One lithological assemblage of the PhylliteQuartzite unit is widely exposed in western Crete and is now relatively well dated following careful biostratigraphical work, mainly using conodonts, benthic Foraminifera and ostracodes (Krahl et al. 1983a,b,c). The dating evidence suggests that both an inverted and a right-way-up

109

succession are present (Fig. 8), which have been interpreted as the lower and upper limbs, respectively, of a regional-scale south-facing recumbent nappe (Fig. 9; Krahl et al. 1983a,b,c). The Late Palaeozoic successions are similar in both of these sections, but the facies differed markedly during the Triassic, as outlined below; assuming the large-scale structure is correctly interpreted, this has important implications for sedimentary polarity, when restored.

C a r b o n i f e r o u s - L o w e r Triassic succession In general, the lower part of the succession (c. 600-800 m thick; Fig. 10b) is dominated by siliciclastic sandstones (quartzites), interbedded with shales (phytlites) and subordinate thinbedded limestones (marbles), ranging in age from Mid-Late Carboniferous to Early Triassic. The thickest-bedded and coarsest-grained siliciclastic sediments are inferred to be of Mid-Permian age (Krahl et al. 1983c). Several thin conglomeratic intervals, here interpreted as debris flows, occur especially in the Lower and Upper Permian intervals (J. Krahl, pers. com.). The Lower Triassic interval (starting at the Middle-Upper Scythian boundary) exhibits abundant evidence of gravity redeposition of variably consolidated sediment, derived from both deep-water and shallow-water settings, of both Permian and Early Triassic age. The oldest known part of the succession, of Mid-Late Carboniferous age (Krahl et al. 1983c), is, for example, well exposed in the N W (e.g. near Sfinari; Fig. 10a), where it dips northwards at a moderate angle. The succession between Sfinari and Kambos, estimated as c. 700 m thick, dips generally northwards, with an average dip of 10-15 ~, and, in places, is strongly sheared, isoclinally folded and faulted. Southwards, a continuous (inverted) succession exposes facies of Early, Mid-? and Late Permian, and then Early Triassic (Scythian) age (Krahl et al. 1983b; Fig. 10b).

S e d i m e n t a r y facies The Middle-Upper Carboniferous part of the succession (near Sfinari, Fig. 10a) comprises medium- to thick-bedded, grey to black, quartzitic meta-sandstones, with dark pelitic intercalations and thin-bedded dark metacarbonates. Individual packets of medium-bedded sandstones, commonly several metres thick, are commonly transitional upwards, over several metres, to grey mudstones, interbedded with subequal thicknesses of relatively fine-grained quartz- and muscovite-rich sandstones (in beds up to 0.25 m thick). These sandstone fade out over several metres and pass into silver-grey mudrocks

110

A . H . F . ROBERTSON

~

Conglomerate Meta-debrisflows Meta-gypsum ~ Meta-dolomite Meta-limestone Meta-sandstone/siltstone i ~ Meta-mudstone Meta-basicigneousrock Keyto logs

TriassiccarbonateevaporiteFaciesAssociation; TripaliGp.

9

"

.

.

.

.

"

"

"

~ A

"

D siliciclastic MCarbonFacies ifemt~-MT . riassic

i

S shalesamples ~V~o~:;~

~Kefa':.

Sk;ph,

i/larn.-~no s LefkaOr,

Key to map

.

N Mana

\/

SVoutas

M.-U. Tr. ! !

An.Cam.

!

_ "-_ -L i -j-- - .

~

LO

1 Kambos

C

2 Voutas

3 Kondokinigi

4 Argilas/ Rodovani

5 Ag.loannis

Keyto logs

L.-M.Scy.~ Bar,e-of-slopef

'

a

~

7_.

~ -

;

i M.Carb.-M.Perm.

[

: iii L,Trias? e.g.Kandanosarea

d

b Fig. 10. Late Palaeozoic-Early Mesozoic evolution of western Crete. (a) Outline map; (b) composite sedimentary log; (c) measured logs of matrix-supported conglomerates of late Early-Mid-Triassic age; (d) possible interpretation of the units of Late Palaeozoic-Early Mesozoic age. (See text for discussion.)

(phyllites), with only a few thin lenses of fine-grained sandstone (up to 15 cm thick). Occasional carbonate-cemented lenses and concretions are less compacted and retain traces of cross-lamination. A less well-exposed, argillaceous succession further south (disrupted by landslipping) includes several isolated sandstone lenses, individually 3-5 m thick. Further south

(near Ano Sfinari) there is a return to thickbedded sandstones (in beds up to l m thick), forming packets up to 2(L40 m thick, alternating with finer-grained facies. The beds within individual packets tend to thicken stratigraphically upwards, assuming the succession is inverted. The Upper Palaeozoic-Lower Triassic succession is cyclical, with alternations of thicker- and

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

111

r~ r

0

;>

>,-J

Q~

o o o o o,,

>-;>

[.T3 o,~ 0

Z

=S ~

=

eq

.

~

~

~

~

~= , _ f . .

. o o ~ID o

>>

,.o

>> 4-,

r.~ ,-o

~

~,..q

c~ o

4~

~.~

0

o

Oo 0

o 0

a

~ 2 0",

g g

r~ 0 .0

~c~c~oo o

N N

112

A.H.F. ROBERTSON

thinner-bedded sediment packets, each around 100 m thick. Despite isoclinal folding, outcrop patterns suggest that the alternating packets are laterally lenticular, on scales of several hundred metres. The thin-bedded packages are dominated by medium-bedded sandstones, up to 25-30 m thick, alternating with thinner-bedded shales, limestones and sandstones. By contrast, the thicker-bedded packets comprise massive sandstones, up to 5 m thick, with subordinate thinnerbedded sandstones, shales and rare limestones. The thickest and most homogeneous Upper Permian? metasandstones ( > 5 m thick) appear to be the most laterally persistent along strike. Most of the thick-bedded sandstone appears to be massive, possibly reflecting partial recrystallization during Alpine high-pressure metamorphism. However, the bases of some of the thinner-bedded sandstones, interbedded with dark shales, exhibit sharp bases and traces of grading. The thickest-bedded sandstones are locally pebbly, with well-rounded quartzitic pebbles (up to 1 cm in size), and are invariably poorly sorted. In thin section, medium-coarse grained sandstones (e.g. near Sfinari) exhibit moderately to well-rounded quartz grains set in a matrix of finer grained quartzose sandstones and siltstones. Other sandstones are dominated by subangular grains of quartz and include subordinate lithoclasts of muscovite schist and quartzite. Minor plagioclase is commonly recrystallized. The matrix includes muscovite, probably entirely recrystallized, and scattered heavy minerals (e.g epidote, amphibole and zircon). Higher in the succession (e.g. near Kambos), contrasting calcareous interbeds are composed of sandy limestone with scattered, mainly subrounded grains of quartz within a matrix of recrystallized coarsegrained carbonate. Thinner, dark coloured, sandstone interbeds are well graded with abundant pyrite and organic-rich material within a finegrained matrix. Overall, the sandstones are mainly quartzarenites and sublitharenites, showing ubiquitous evidence of textural inversion. Sediment chem&try

To help determine sediment provenance seven samples of dark metashales were collected from the stratigraphically higher part of the succession (of Late Permian?-Early Triassic age) located on a mountainous ridge (Ayias Zinas, NE of Kandanos; Fig. 10a) and were analysed for majorand trace elements by X-ray fluorescence (XRF; see Table 1 for representative analyses). When normalized against the composition of average North American shale (Gromet et al. 1984), the samples are compositionally similar to average shale, although relatively enriched in several

elements (e.g. La, Ce and Nd), but depleted in Sc, Zn and Ba (Fig. 11). The Ba depletion could relate to mobility of this element during Alpine metamorphism. The shales are compositionally similar to the average shale derived from the north margin of Gondwana in northern Syria (A1-Riyami & Robertson 2002). When plotted on several tectonic discrimination diagrams (Th v. Sc v. Zr/1000; K20/Na20 v. SiO2; TiO2 v. Fe203 + MgO) the samples plot in the oceanic island arc or active margin fields. However, the inferred igneous component is unlikely to relate to contemporaneous volcanogenic input, as sediments are interbedded with siliciclastic sandstones and lack volcanogenic material. In agreement with Romano et al. (2006; see also Bojar et al. 2002), the petrographic and geochemical evidence suggests that the terrigenous sediments were derived from the North African craton, or possibly from a rifted continental fragment of the same crustal composition. M a g m a t i c intercalations

Alkaline magmatic rocks, metamorphosed under HP-LT conditions, are locally present, mainly in the higher levels of the siliciclastic-shale succession (Seidel 1978). Small volumes of metaalkaline volcanic rocks, volcaniclastic sediments and tufts are interbedded with deep-sea sediments of Early Triassic (i.e. Early Scythian) age (Krahl et al. 1983c). Using a modern time scale (e.g. Gradstein et al. 2004), 'Late Permian' ages, radiometrically determined by Seidel (1978) are reassigned to the Early Triassic, although volcanism may have begun in the Late Permian. Prominent exposures of basic igneous rocks mainly occur in the higher levels of the siliciclastic-dominated succession. Metabasic igneous rocks were studied from well-exposed ridges to the NE of Kandanos (e.g. Palea Roumata; Spina; Ano Kefala; see Seidel 1978). These form lenticular, competent bodies, ranging from several metres to c. 15 m thick, usually traceable laterally up to several hundred metres. The most common lithologies show no obvious extrusive features and are likely to have originated as sills. For example, on a high ridge (near Ayios Ionnis Apopigadi), medium- to thick bedded quartzites and grey phyllites are interbedded with metabasic rocks, in layers up to 3-5 m thick. Further north, at Palea Romata, bluffs of very hard, very resistant metabasites, c. 10 m thick, are intercalated with isoclinally folded dark phyllites. Metaigneous lenses, 10-12 m thick, are traceable laterally for hundreds of metres along the hillside. Similar igneous bodies (5-8 m thick) were also studied further north (at Ano Kefala), within similar NE-east dipping sandstones and phyllites. Resistant bands, up to 10-20 m thick, are also

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

113

I}

",/

/

,1~

-

-

P

"-"1 ~-

1.0

c-

d o e~

E 0

~_0.1 E

I Sc

I Cr

I Zr

(Y)

I La

I Ce

I Nd

I Sr

I Rb

I Ba

I Th

Fig. 11. Sedimentary geochemistry of shales from the Upper Palaeozoic-Lower Triassic interval of the PhylliteQuartzite unit in western Crete. The results suggest derivation from a continental basement, probably North Africa. traceable across the hillside, 2 km south of Kandanos (above Anisaraki), within mainly fine-grained metasedimentary rocks. Meta-basic rocks (17 samples) from several localities in the N W of the area (e.g. Skaphi, Orthouni and Chosti) were analysed for major and trace elements by X R F (see Table 1). When plotted on MORB-normalized 'spider diagrams' (e.g. Pearce 1980), the amphibolites are typical of alkaline rift-related basalts and more fractionated alkaline igneous rocks (Fig. 12), in agreement with previous studies (Seidel 1978). In addition, a further 18 samples of metabasalts were analysed from a small exposure of the Phyllite-Quartzite unit on the hillside above Amoundi Beach on the south coast, SE of Kato Preveli monastery (Fig. 5) and these showed very similar 'enriched' patterns (unpublished data). Triassic i n v e r t e d succession

Depositional transitions from the typical siliciclastic facies to the Middle-Upper Triassic more

carbonate-rich facies are well exposed in the N W (Kambos area) and in an inlier in the SW (Kandanos area; Fig. 10a). In general, a succession of mainly shales and platy limestones of Early-Mid-Triassic age passes into mainly dolomitic carbonates and shales of Late Triassic age, culminating in shales and dolomitic carbonates with gypsum, of Carnian-Liassic? age. An inferred hiatus during the Late Scythian-Anisian time interval may correlate with the prominent hiatus in the Talea Ori unit (Epting et al. 1972; Krahl et al. 1983a). In the NW, near Kambos, alternating sandstones and shales of the Phyllite-Quartzite unit (of Late Permian-Scythian age) pass stratigraphically upwards into dominantly thickbedded marble, associated with an interval of carbonate conglomerates (r 10 m) comprising clasts of marble (up to 15 cm in size), some of which are well rounded (near the Mid-Late Scythian boundary; Fig. 10c). Individual depositional units, up to several metres thick, are dominated by marble clasts (up to 10 cm in size)

A. H. F. ROBERTSON

114

WEST CRETE: MORB Normalized 100 ! --4P-W Crete, Orthouni l ~W Crete, Skaphi --,t--W Crete, Chosti 10

0,1

Sr

K

Rb

Ba

Nb

La

Ce

Nd

P

Zr

Ti

Y

Sc

Cr

Fig. 12. MORB-normalized 'spider diagrams' of selected metabasic rocks from the Lower Triassic? interval of the Phyllite-Quartzite unit in western Crete. All the samples show 'enriched' trends and plot within a relatively narrow compositional range. (See text for explanation and Fig. 10a for locations.)

that are locally very well rounded. These conglomerates are interbedded with medium- to thick-bedded metalimestones (up to 3 m thick), purple-red shale (phyllite) and several thin interbeds of white sericitic-rich shale (up to 0.12 m thick), possibly tuff. The limestones include reworked ostracodes (Krahl et al. 1983c). Some beds contain numerous elongate sandstone or siltstone rip-up clasts (up to 0.4 m long), set in a poorly sorted quartz-rich matrix. Intraformational clasts of dark grey phyllite are also present within poorly sorted 'gritty' quartzose metasandstone. Long thin clasts were disrupted while still poorly lithified. Some reworked quartz grains are relatively large and well rounded. Thicker bedded, almost clast-supported conglomerates, occur higher in the succession (in beds up to 2.3 m thick) and contain subrounded limestone clasts (up to 0.6 m long), in which individual clasts range from pure to muddy carbonate. Several of these limestone conglomerates exhibit scoured bases and sharp tops, confirming that this succession is inverted. The succession then passes stratigraphically into well-bedded platy limestones and shales of inferred MidTriassic (Anisian-Ladinan) age. A comparable exposure in the Kandanos area extends into Late Triassic dolomitic carbonates and evaporitic facies. In addition, the inverted Triassic succession is well exposed in an arcuate belt further south, an area strongly affected by neotectonic

east-west faulting. However, Krahl (1983a,b,c) has reported a number of locally intact Triassic successions. The Triassic succession is, for example, locally exposed on the footwall (commonly landslipped) of a prominent east-west neotectonic fault zone. Just north of Voutas, typical thick-bedded quartzite (in beds up to 1.5m thick), of inferred Early Triassic age, is followed northwards (after a short gap in exposure) by thin- to medium-bedded platy granular carbonates with interbedded dark organic-rich phyllites, of inferred Mid-Triassic (Anisian) age. Higher parts of the succession, of inferred Late Triassic age are well exposed in a hilly area to the south (e.g. SW of Voutas). This succession, 100200 m thick, mainly dips eastwards at moderate angles (c. 38 ~ and is dominated by dark grey dolomitic carbonates, interbedded with thin- to medium- and thick-bedded carbonate-rich shales. The section is locally deformed into south-facing outcrop-scale folds and small duplexes. Individual dolomitic beds, up to 0.6 m thick, are composed of buff sugary carbonate, full of small vugs, apparently created by evaporite dissolution. Thinner-bedded, dark interbeds are finely crystalline, locally with small intraclasts of gypsiferous marl. Fissile intercalations are dark and organic rich. Elsewhere, Triassic evaporites have locally been mobilized to form lenticular masses of ductile-deformed gypsum, tens to several hundred metres thick.

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS Triassic right-way up succession: Mana unit A regional right-way up succession, best exposed in the NW (e.g. NE of Sfinari, near Mana and Sineniana; Fig. 10a), exhibits a relatively deep-water setting that persisted during MidTriassic time, with Radiolaria and pelagic conodonts, but then shallowed upwards during Late Triassic time (Norian), allowing the deposition of shallow-marine limestones and dolomites (without evaporites). The succession is capped by the Mana unit, which is composed of marbles (Mana marble) of inferred Late Triassic-Early Jurassic? age, followed by undated conglomerates (Mana conglomerate) (Krahl et al. 1983c). The Mana unit is critical to the interpretation of the later Triassic palaeoenvironments. For this reason three sections were studied, one in the NW, one further south and one in the far south, to provide a regional overview. In Model 1 this unit would be expected to relate to a rift setting, whereas in Model 2 it should record a foreland basin or post-collisional 'molasse'-type setting. In the NW, near Sineniana (Fig. 10a), a composite succession was pieced together in several adjacent fault blocks. The exposed succession above the valley floor (c. 1 km south of Sineniana; near Tsortsiana chapel; Fig. 13a, locality A) begins with thick-bedded, massive quartzitic sandstones (in beds up to 0.6 m thick) with subordinate mud-rock (dark phyllite). These sediments correlate with the stratigraphically higher parts of the Phyllite-Quartzite succession in the adjacent area. The quartzitic sandstones pass upwards into massive or thick-bedded marble (Mana marble). Similar marbles are downfaulted to the north forming two fault blocks (Fig. 13a). In the more northerly fault block the marbles are locally intercalated with, and then overlain by, coarse quartzitic facies (Mana conglomerate; Fig. 13c). The lowest exposed clastic facies depositionally overlie a white marble that exhibits a fissured upper surface. Fissures are infilled with grey calcareous mudstone, lenticular (0.3m) sandstone and then conglomerate (c. 0.6 m) with very well-rounded, dark grey quartzitic clasts (Fig. 13a, locality B). The clasts (up to 0.35 m in size) are poorly sorted and set in a sandy matrix. There is then further marble (c. 80 m thick) before the main succession of the Mana conglomerate comes in depositionally above. This comprises a gently dipping succession of alternating conglomerates, quartzitic sandstones and dark-coloured mud-rocks (phyllites) (Fig. 13c). The succession, although faulted, is locally well exposed along the road directly south of Sineniana (Fig. 13, locality C). There, several individual conglomeratic depositional units (up to 4.5 m thick) exhibit scoured

115

bases, overlain by conglomerate with mainly subrounded to well-rounded clasts (up to 0.5 m in size). The succession grades upwards through medium and fine sandstone to shale. There are also several amalgamated, ungraded intervals of conglomerate and sandstone, again with wellrounded clasts and common intraformational rip-up clasts. Similar graded, or ungraded, conglomerates are exposed more widely on the hillside to the SW (at Fig. 13, locality D) and can be traced laterally for hundreds of metres with little change in thickness. Interbedded sandstones are commonly graded, fining upwards from coarse sandstone, to fine sandstone, then dark shale. The mudrocks commonly include sandy partings (up to several millimetres thick) with very well-rounded quartz grains. The highest levels of the exposed succession include poorly exposed intercalations of shales, sandstone and conglomerate, and several laterally impersistent intercalations of marble (up to 10 m thick). In this area the Mana unit is structurally overlain by highly recrystallized carbonates and other facies (e.g. shales; radiolarian cherts) of the unmetamorphosed Pindos unit. At the second, more southerly locality, near Sarakina c. 14 km further south, the Mana unit is again well exposed as an isolated exposure near the top of a prominent hill (820m) above Sarakina village (Fig. 14). The succession above the village begins with pale grey, buff or yellow phyllite with occasional thicker quartzitic sandstone beds (up to 0.6 m), of inferred Early Triassic (Scythian) age. South of a prominent col (Fig. 14a) the succession passes upwards without a break into medium-bedded sugary marble with pale phyllite partings. A sheared and folded succession above this is dominated by medium- to thick-bedded grey dolomite and shale of inferred Mid-Late Triassic age. Above comes thick-bedded marble (Mana marble). The basal contact is marked locally by a c. 0.5 m thick zone of sheared and brecciated phyllite and recrystallized marble. The overlying, nearly massive Maria marble is overlain, apparently depositionally, by a veneer of conglomerate (Mana conglomerate) near the hilltop (Fig. 14a). The conglomerate comprises repeated weakly stratified depositional units, each up to several metres thick, made up of densely packed, clast-supported quartzitic conglomerates. The clasts range from well rounded to subrounded (average size 14 cm; maximum 60 cm), with occasional lenticular intercalations of dark sandstone. Despite the strong shearing and localized thrusting it is likely that an originally Lower TriassicUpper Triassic or Lower Jurassic? right-way up succession is represented at this locality (Fig. 14b and c).

116

A.H.F. ROBERTSON

Sineniana +oo

......

6"7 "~ e t, "o

,,

-~ i r-'qc '

,:. ,?o,:o;,

L_J "-100m

C

~ 1 7 6 111 76

.__,o_.2 o

o

o

o o

o o o o ~ o

o

o o o o o

o o o

.".-.T.Y. o o o o o

i

I

II I It i It

I I

J

II

I

I I1 I 1 I

Pindos thrust sheet

i,I I I I II I II I IJ I [1 I II

I I I I II J II ! II

I,I

,1I

II I II I

Bedded marble

9

J

Mud rock

lMan i

Meta- sandstone J unit o o o ~

o

["2m Meta- conglomerate Loc C Local log

II I H I 1 I

Late Triassic - lowermost Jurassic ?

r20m

Composite log Fig. 13. Tectonic-sedimentary relations exposed in the upper, right-way-up succession in the Phyllite-Quartzite unit in NW Crete (see Fig. 10a). The data presented highlight the importance of the conglomeratic Mana unit (of Late Triassic-Liassic? age). (See text for discussion.)

The third succession studied occurs further south again, in the hanging wall of an east-westtrending zone of major neotectonic downfaulting

to the south. Local successions are shown in Figure 10c, logs 2-4. For example, a partially landslipped succession, exposed near

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

S

117

820m .?

-

~

~

s

"g** ,-

Tr.? I

i

a

jj"

i

"'/

- 100m

~'~

L.Tr."

./ ~

......

..s ,...

~

~ .

"

[~ o=o o

r 20 [j I

I

t

i

U.T~.?~! M ,_L ! ,_

,__

r-

--

L.-M.

F~

Gavrovo- Tripolitza / Pindos units

rgg] Mana conglomerate [~

Mana limestone / marble

[~

Limestone/dolomite

r~

Shale/phyllite

Tr.?

Triassic- lowermost Jurassic ? C

--~

Fig. 14. Additional important tectonic-sedimentary relations exposed in the upper, right-way-up succession in the Phyllite-Quartzite unit in central southern Crete (see Fig. 10a). A Triassic succession is capped by the conglomeratic Mana unit (of Late Triassic-Liassic? age). (See text for discussion.)

Kondokinigi (by the turning to Voutas), passes from green phyllite (of Mid-Triassic? age, into thick-bedded marble (c. 16 thick; Mana marble?) and then into a coarsening-upward clastic succession of sandstone (10 m), coarse sandstone (8 m), then conglomerate ( > 2 5 m). In the same area (1 km south of Kondokinigi), a gently northward-dipping succession of dark phyllites and thick-bedded quartzitic sandstones passes depositionally upwards into typical Mana conglomerates, c. 60 m thick. This conglomerate is dominated by elongate, subrounded, pale yellow to grey, fine-grained sandstone and siltstone, in a matrix of yellowish grey phyllite. Occasional clasts of coarse sandstones are also present. Many of the clasts are elongate (up to 0.6 cm x 15 cm) and set in a sandy and gritty matrix. The conglomerates were in turn overridden by the unmetamorphosed Gavrovo-Tripolitza and Pindos units. Further west, on the westward extension of the fault zone, near Papadiana, local exposures of the Mana unit include conglomerate with subrounded sandstone clasts (up to 0.2 m in size) and quartzitic sandstones affected by soft-sediment deformation.

Interpretation: a developing rift The overall Middle-Upper Carboniferous to Lower Triassic succession accumulated on a relatively deep-marine slope, to base-of-slope, setting, judging from the redeposited nature of the sandstones and limestones, and the presence of Radiolaria and deep-water conodonts within the interbedded shales. The palaeowater depths are estimated as > 500 m (H. Kozur, pers. com.; Fig. 10d). The large-scale (tens to hundreds of metres thick) cycles record the interplay of local tectonics versus eustatic sea-level change, whereas the smaller scale (tens of metres or less) thickening- and coarsening-upward cycles may relate to autocyclic (steady-state) depositional processes (e.g. sand lobe progradation). Sand input peaked during the Mid?-Permian, then waned, whereas carbonate input relatively increased during Early Triassic time. The source of the sandstones was probably Pan-African basement, as exposed in Egypt and Libya, or possibly a rifted continental fragment of the same crustal type. The common well-rounded pebbles within thicker-bedded sandstones were derived from a high-energy shallow-marine or fluvial

118

A.H.F. ROBERTSON

setting. These sediments were redeposited into deep water by a range of low- to high-density turbidity current and mass-flow processes. Most of the thick-bedded, locally pebbly sandstones are seen as subaqueous sand flows (gravity flows). The prominent interval of matrix to clastsupported conglomerates of Early Triassic age towards the top of the siliciclastic succession records oversteepening of the pre-existing slope, resulting in widespread down-margin gravity transport. The mass movement was possibly triggered by uplift from a deep-sea setting to a shallow-marine carbonate-depositing setting during Mid-Triassic time, resulting in a hiatus in deposition. Material of mostly Permian and Triassic age was derived from settings ranging from shallow shelf (e.g. oolites) to deep water (e.g. micritic clasts with pelagic conodonts). Little evidence was observed to support the suggestion of sand deposition, within (or close to) an idealized deep-sea submarine canyon (Dornsiepen et al. 2001). There are few examples of coarse lenticular, channelized, debris-flow-type conglomerates, typical of such canyon-mouth settings. The sandstones are also dissimilar to typical 'classical' turbidites, as most are poorly sorted and only rarely well graded. These sediments were possibly deposited as sandy mass flows that were transported down a relatively steep, fault-controlled slope, possibly fed from line sources, rather than through regional-scale submarine canyons. The subordinate thinner bedded limestones are interpreted as relatively distal calciturbidites derived from a carbonate platform. The alternating thinner- and thickerbedded sand deposition persisted from MidCarboniferous to Early Triassic time. This is consistent with gentle subsidence of a rift rather than the long-term subsidence of a passive continental margin bordering an ocean basin, in which an overall thinning and fining-upward and deepening succession would be anticipated. As inferred from western Sicily, above, there is actually no evidence that a Late Palaeozoic ocean actually ever existed in the south Aegean region to the south, adjacent to Gondwana. Also, the Triassic alkaline volcanic rocks (extrusives and sills) that were erupted into deep water mainly during Late Permian?-Early Triassic time would not be expected in a mature passive margin succession. Given their characteristic 'enriched' composition, these volcanic rocks are better interpreted to represent low-degree melts that erupted in an extensional rift setting. The Phyllite-Quartzite unit of western Crete has also been suggested to represent the distal

facies of a south-Palaeotethyan passive margin that accumulated on oceanic crust (Ziegler & Stampfli 2001), but this is opposed by the relatively proximal setting of the sediments. This unit was also suggested to represent Palaeotethyan oceanic sediments preserved as a 'Cimmerian' accretionary prism of pre-Jurassic age (Ziegler & Stampfli 2001). However, this interpretation is opposed by the existence of long intact sedimentary successions and the absence of contemporaneous thrust-imbrication as in accretionary prisms. The observed deformation and metamorphism are instead believed to be entirely of Early Cenozoic age. The inferred rift basin generally experienced uplift of > 500 m during Mid- to Late Triassic time. This uplift took place especially during Late Early Triassic to Mid-Triassic time (Krahl et al. 1983c). This shallowing was linked to more calcareous, shallower water deposition, although with continuing siliciclastic input. Shallowing continued until deposition was restricted to semiisolated shallow-marine to lagoonal settings that were possibly fault-controlled. Relative sea-level fall culminated in gypsum deposition in evaporating marginal lagoons. Partial dissolution in response to freshwater leaching gave rise to local solution-collapse breccias (Pomoni-Papaioannou & Karakitsios 2002). By contrast, in the rightway-up succession, open-marine carbonate deposition persisted into the Mid-Triassic, followed by a relatively abrupt upward transition to a shallow-marine carbonate-depositing setting during latest Triassic-Early Jurassic time (Krahl et al. 1983c); this culminated in the deposition of the Mana marble and Mana conglomerate. The Mana conglomerates are interpreted as shallowmarine to non-marine facies that were deposited in high-energy deltaic settings, possibly including fan deltas. The source of the nearly monomict, remarkably pure quartzitic sandstones is problematic, as derivation from either Pan-African or Hercynian basement would be expected to have produced more polymict material, including schist and gneiss. The sandstones could instead have been derived from an uplifted part of the Phyllite-Quartzite succession, assuming this was already lithified, but again, a more heterogeneous composition, including quartzose, carbonate and basic igneous rocks, would be expected. One other possibility is that the Mana conglomerates were eroded from a succession of texturally mature sandstones, which accumulated on a marginal high (fault block or plateau) bordering a rift basin. Assuming the regional structure of a southfacing recumbent nappe is correct (Krahl et al.

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS 1983c), the inverted limb restores to a relatively southerly position, compared with the right-wayup limb. This would imply that the source area of the M a n a conglomerate was generally to the north. If correct, the source was a rifted margin or intra-basinal high to the north. In summary, the western Crete successions are consistent with Model 1 in which rifting took place in pulses, at least during MidCarboniferous (post-Hercynian orogeny) and Mid-Triassic time. In Model 2 the turbiditic sandstones of Permian-Early Triassic age would relate to subsidence of the southern margin of a Cimmerian passive margin. However, similar deep-water siliciclastic sedimentation started up to 150 Ma earlier. In Model 2 the Mid-Triassic hiatus and uplift could reflect the passage of a flexural bulge across the basin as Palaeotethys closed to the north. However, such flexural uplift, if significant, should have been followed by fiexural downwarping related to southward passage of a thrust load, itself related to the latest Triassic 'Cimmerian' collision with the Eurasian continent to the north. Instead, continued relative uplift is seen through Late Triassic time, culminating in evaporitic deposition. The Mana conglomerate cannot have been shed from the Eurasian margin, in view of its nearly homogeneous quartzitic composition. A more heterogeneous composition, including arc-related or ophiolitic rocks, would be expected if the conglomerates were related to a forearc, foreland basin or collisional setting. Also, if related to a convergent setting an upward-thickening and coarsening succession would be expected, which is not observed.

Phyllite-Quartzite unit of eastern and central Crete Distinctive lithofacies of the Phyllite-Quartzite unit are exposed in eastern Crete, particularly in the Chemezi and Vai areas (Fig. 5). These lithologies differ from those western Crete in terms of facies and structure. In particular, long intact successions are preserved in eastern Crete whereas successions in western Crete form parts of a tectonic slice complex. In eastern Crete there are two main exposure areas: the Chemezi area and the Vai area in the far east of the island. Here, the main focus will be on the Chemezi area, where most of the relevant units are exposed and the tectonostratigraphy is relatively simple. On the other hand, the local tectonostratigraphy of the Vai area is extremely complicated and to some extent controversial, mainly owing to Cenozoic deformation and metamorphism, such

119

that a fuller treatment is required elsewhere. Most of the units in the Vai area can be generally correlated with those of the Chemezi area, discussed below. However, two critical units in the Vai area are not known in the Chemezi area and these will be discussed below. Some thrust sheets (e.g. high-grade metamorphic rocks) are thin and laterally variable such that no simple tectonostratigraphy is applicable to the area as a whole, and so the various units must be studied and interpreted individually. In the Iraklion area of central Crete (Fig. 5) an additional outcrop area of Triassic sedimentary and volcanic rocks can be correlated with the units cropping out in eastern Crete. Importantly, these structurally overlie a unit with lithological affinities with the Phyllite-Quartzite unit of western Crete. This allows an assessment to be made of the relative tectonic settings of these two regionally contrasting units. Taking the Chemezi and Vai areas together, in Model 1 the Upper Permian-Lower Triassic succession represents a rift setting. Upper Permian-Lower Triassic deep-water sediments (i.e. radiolarian shales and hemipelagic limestones) record part of an early rift basin, a counterpart of the pelagic sediments exposed in western Sicily. Middle-Upper Triassic successions then represent a shallow-water to nonmarine rift-setting and also deeper marine riftrelated settings related to opening of the Pindos ocean to the north. Some Triassic lithologies are associated with volcanic rocks of basic to intermediate composition, which are interpreted as rift related in this model. Fragments of pre-rift continental basement are represented by thrust slices of 'Hercynian' high-grade metamorphic rocks. In Model 2 the successions in the different thrust sheets should record a wide range of tectonic settings, including an ocean basin with volcanic seamounts, a forearc basin represented by radiolarian sediments, a continental crust unit represented by 'Hercynian' basement and a backarc basin related to the opening of the Pindos ocean. In Model 3, only a Triassic rift would be expected, together with subduction-related magmatic rocks. In Model 4 a supra-subduction zone rift-related setting would be expected, with magmatism possibly affected by either, or both, of a northward-dipping and a southward-dipping subduction zone. For the central Crete Iraklion area, in Model 1 both structural units exposed would relate to Triassic rift successions. However, in Model 2 this occurrence would record the actual Palaeotethyan suture between a rifted Cimmerian passive margin (i.e. North Africa derived) and an Eurasian active continental margin.

120

A.H.F. ROBERTSON

Aegean N Sea

A Kalayros ~ - ~ r B

~

v G

SITIA <

9

Key to map Volcanicrocks Gypsum Plio - Quaternary Tripolitza nappe

I-El

~

variscan metamorphic rocks .Phyllite - Quartzite Unit Plattenkalk nappe

Chemezi Tripo .kefala unit

h, u'

'

l

0.s Km ~

0 01

JO

I.

unil

A A Exo/Meso

-I

|

|

[oooi [? o - ~ [-" o o l

,

I,-

-iI

Mouliana

& Ag.Theodoros

~

121o5 U. Pe.

--

--

--

--

-

Ill

--i

-'

-

o

o

o

r-~

~

V o

V

'~/

o

~-[

V

V

V

--.-i

,,t.,.

z Permo - Triassic

ci

"~

'

Key to section & logs Metamorphic clasts

o V

o

--

o

Tr. o ~ ,oq V

--

B

Triassic

Conglomerates t Volcanogenic rocks Triassic Limestone/ Limestone conglomerate Variscan metamorphic rocks

, e . . l l

Permo - Triassic pelagic limestone U. Permian radiolarian shale

cii

c

~:1

Plattenkalk

Fig. 15. Upper Palaeozoic-Lower Mesozoic units of eastern Crete from the Phyllite-Quartzite unit in the Chemezi area, west of Sitia. (a) Outline map; (b) sketch cross-sections; (c) measured logs. (See text for explanation and data sources.).

Evidence from the Chemezi area

Hemipelagic shale and pelagic limestone

In this area the Phyllite-Quartzite unit forms a number of thrust sheets that are folded into a broad east-west-trending syncline (Mersini syncline; Krahl et al. 1986; Fig. 15a). Two wellexposed traverses were examined during this study (Fig. 15b). Combining these, the following units are recognized from the structural base upwards.

This unit is well exposed in the NW, 0.3 km west of Kalavros (Fig. 15bii). The highest levels of the Plattenkalk unit there, the Lower Cenozoic Kalavros beds, are composed of muddy limestones that are overthrust by a distinctive unit of thin-bedded, relatively undeformed reddish brown platy shale (phyllite), up to 100 m thick, with Mn coatings on fractures. These shales

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS (Violetschiefer) contain Radiolaria and pelagic conodonts of Early Permian age (Kozur & Krahl 1984). In places, the shales are chemically reduced to a grey or green colour. In thin section, the typical shale is mainly ferruginous mudstone with silty laminae composed of quartz and feldspar. The reddish shale grades upwards into grey- green shale, then into thin-bedded hemipelagic limestone, of up to Early Triassic age (Krahl et al. 1986). A succession of relatively undeformed alternating well-bedded pure and muddier limestone with shale partings, c. 120 m thick (Agrilos Schichten), is then overthrust by a strongly contrasting mainly volcanogenic unit (Fig. 15ci). Similar, but faulted and thrust-imbricated, Permian shales and overlying Lower Triassic hemipelagic limestones are well exposed further south, especially along the main road from Ayios Nikolaos to Sitia. Local transitions from red shale to pelagic limestone are again seen. Notably, c. 200 m east of Exo Mouliana, a local hemipelagic limestone succession (c. 80 m thick) comprises alternations of thin- to medium- and thick-bedded facies, with muddy partings. Further east, near Ayios Theodoros, dark muddy partings are rich in fine plant material. In thin section, the limestones are finely laminated, with numerous small organic-rich grains. The Lower Permian reddish shales are interpreted as deep-sea hemipelagic sediments, with radiolarians, conodonts and fine-grained terrigenous sediment. The presence of reworked fossils suggests that the deep-water succession may date at least from Late Carboniferous time (Krahl et al. 1986), contemporaneous with the deep-water sediments in western Crete. The abundant manganese oxide in the pink shales could reflect continental runoff or even a hydrothermal source of manganese, but there is no known lithological association with igneous rocks. No basement to these deep-sea sediments is exposed. The siliceous sediments gave way to hemipelagic carbonates in the Early Triassic, with continued input of terrigenous silt and the addition of organic matter. The upward colour change possibly reflects a decrease in bottomwater oxygenation, or increase in organic matter, that was possibly climatically controlled. Volcanic-pelagic carbonate unit In the SW, near Mesa Mouliana (Fig. 15bii), the Plattenkalk unit dips at 40 ~ to the north and is structurally overlain by volcanic rocks (Seidel 1978), including strongly weathered aphyric, non-vesicular massive basalts and andesites. These volcanic rocks are locally interbedded with

121

grey to pink pelagic limestone, in depositional units up to 8 m thick. The volcanic rocks record eruption in a open-marine relatively deep-water setting, and may correlate with thicker and more intact Lower Triassic successions exposed on the northern flank of the Mersini syncline (see below). High-grade metamor phic basement The Permo-Triassic sediments and volcanic rocks are locally structurally overlain by several slices of high-grade basement rocks, mainly gneiss, mica schist, amphibolite and marble (Seidel 1978; Seidel et al. 1982). The protoliths include metasediments, basic igneous rocks and granitic rocks. For example, near Chemezi, the metamorphic rocks are dominated by dark grey, brown weathering, mica-schist intercalated with dark amphibolite. The metamorphic rocks are cut by numerous quartz veins and show evidence of extensive brittle deformation (e.g. small-scale duplexes), especially near the thrust contact with the underlying deep-sea sediments. Recent radiometric dating shows that the high-grade metamorphic rocks are divisible into several units, one with Cadomian (late Precambrian) and Late Carboniferous ages (Mirsini crystalline complex), and another with Late Permian ages (Kalavros crystalline complex; Finger et al. 2002; Romano et al. 2002, 2004, 2006). Some of the amphibolites exhibit MOR-type protoliths suggesting that they could have been accreted from a 'Hercynian' ocean prior to metamorphism. South-verging small-scale structures (e.g. C-S fabrics) have been reported by Romano et al. (2006), although these need to be treated with caution as they occur within relatively thin thrust sheets that have undergone Early to Mid-Cenozoic subduction and exhumation. V olcanogenic- limes tone- s haleconglomerate unit In the NW, on the northern flank of the Mersini syncline (Fig. 15bi), the Permian-Early Triassic deep-water radiolarian shale-pelagic carbonate succession or, locally, high-grade metamorphic rocks are structurally overlain by a contrasting succession that begins with metavolcanic rocks (mainly basaltic andesite), c. 100 m thick (Fig. 15cii). Occasional massive lavas near the exposed base of the succession are overlain by lava breccias, in units up to 8 m thick. These are mainly composed of pebbly volcaniclastic conglomerate or breccia (with stretched pebbles up to 10 cm long), volcaniclastic sandstones, volcanogenic shales. Repeated thin (< 1 m) lava flows are interbedded with limestone conglomerates. The lavas

122

A . H . F . ROBERTSON

EAST CRERE: MORB Normalized 100 --U--East Crete, Vai --e--East Crete, Chemezi 10

0.1

0.01 Sr

K

Rb

Ba

Nb

La

Ce

Nd

P

Zr

Ti

Y

Sc

Cr

Fig. 16. Geochemical plots of Triassic basalts from the Phyllite-Quartzite unit in the Vai and Chemezi areas of eastern Crete. (See text for explanation and Table 1 for locations.)

and associated relatively deep-water sediments are inferred to be of mainly Late Scythian (Late Olenekian) age (Krahl et al. 1986). Several samples of basalt of sub-alkaline composition were analysed by XRF and when plotted on MORB-normalized spider diagrams (Fig. 16) reveal a distinctive subduction influence, as shown by a relative depletion of immobile incompatible elements (e.g. Pearce 1980), and notably a pronounced negative niobium anomaly. The mainly volcanogenic lithologies are, in turn, overlain by thin- to medium-bedded limestones (c. 60 m thick). The thicker-bedded limestones are deformed into 'mega-boudins' (up to 10 m thick • 80 m long) by extensional shear, whereas the thinner beds are stretched to form detached' phacoids' within shale. Above this follows a strongly sheared interval, 10 m thick, probably also of Early Triassic age, made up of crudely stratified conglomerates and breccias, in beds up to e. 3 m thick, with clasts (up to 0.6 m in size) including schist and gneiss ('Chemezi beds'). Intercalated green volcanogenic shale (1 m thick) includes blocks of mica-schist (up to 0.4 m long). Overlying conglomerates are dominated by stretched pebbles composed of limestone, intercalated with sandstone lenses (up to 1.5 m thick). Higher levels of the succession, c. 150 m thick, comprise intercalations of siliciclastic sandstone, andesitic lava breccia (with clasts up to 0.35 m in size), volcanogenic debris flows, pebbly

sandstones (with angular limestone clasts), volcaniclastic sandstone (in <1 m thick interbeds) and volcanogenic shale. Occasional competent beds of limestone debris flows (in units <1 m thick) are strongly stretched to form phacoids in an incompetent volcanogenic matrix. Grading and scour structures indicate that the succession is mainly the right way up. The highest levels are isoclinally folded, suggesting the presence of a thrust, or at least strongly sheared contact, above which there are slightly metamorphosed brilliant reddish purple meta-mudrocks ('violet schists'), with thin (<10 cm), fine-grained sandstone interbeds (c. 80 m thick altogether). Fresh cuttings reveal the presence of dark organic-rich mudrocks prior to surface oxidation. The mixed volcanogenic-limestone-shaleconglomerate unit records andesitic volcanism in a highly unstable setting marked by redepositional processes, with volcanogenic and carbonate debris flows and turbiditic volcaniclastic sandstones. Volcanism later gave way to the accumulation of volcanogenic sediments and limestone conglomerates with clasts of locally derived metamorphic basement lithologies. This unit is inferred to be late Early to Early MidTriassic in age (Chemezi beds; Krahl et al. 1986). The overlying dark organic-rich muds and minor volcaniclastic turbidites possibly represent an original upward continuation of this setting in a quiescent relatively deep-water, reducing environment.

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS Coarse limestone conglomerate unit

The above unit is structurally overlain by a thick (<250 m) succession of very coarse limestone conglomerates and breccias (Tripokefala beds), well exposed in the vicinity of the OTE tower; Fig. 15a). To the north, the underlying volcanogenic unit is bounded by a major high-angle fault. According to Kozur (pers. com.), the limestone conglomerates are no younger than late Early to early Mid-Triassic in age, whereas Champod et al. (2004) suggested a Ladinian to Carnian age. Near the basal thrust contact coarse limestone conglomerates predominate, with stretched, but originally well-rounded clasts (Fig. 15cii). These clasts are locally fused, giving the impression of massive carbonate rock. The limestones locally include small slices (duplexes) of metavolcanic rocks (up to 4 m thick x 10 m long), as seen on a steep slope 0.4 km W N W of the OTE tower. Above this there is a generally thinning- and fining-upward succession of limestone rudites, sandstones and shales (Fig. 15cii). The limestone are mainly lenticular conglomerates and breccias, mostly composed of stretched limestone clasts. The interbedded sandstones are packed with elongate limestone clasts (up to 35 cm long) and include occasional large blocks of schist, gneiss (up to 0.6 m in size) and occasionally andesite. Several individual units grade from microconglomerate to sandstone with shale partings and occasional pebbly bands, suggesting that the succession is the right way up. The highest exposed levels (near the OTE tower) are dominated by brown-weathering shales and sandstones, with only occasional conglomerates (up to several metres thick). Most clasts are < 10 cm in size but a few intervals with larger clasts are also is present. Thin sections show that the typical sandstones are dominated by polycrystalline and monocrystalline quartz, with subordinate mica-schist, muscovite and plagioclase, set in a variably altered, chloritic fine-grained, or siliceous matrix. Some sandstones contain abundant relatively fresh basalt, mainly as angular elongate grains (shards) in a dark mesostasis of altered volcanic glass. Other sandstones contain coarse plagioclase (diabase or gabbro) and large clear tabular quartz grains (possibly reworked phenocrysts). The coarse limestone conglomerates (of Early-Mid-Triassic age) are interpreted as a proximal deltaic unit that accumulated in a shallow-marine to possibly non-marine setting. The clasts were derived from metamorphic basement, shallow-water carbonates and volcanogenic units. The structurally underlying units

123

(volcangenic units and high-grade metamorphic slices) could have supplied most of this material. However, a suggestion that the conglomerates represent a 'Verrucano-type facies' unconformably overlying metamorphic basement (Chainpod et al. 2004) could not be confirmed, as all the observed contacts are tectonic. Exotic gypsum

A notable feature of this area is the presence of large lenses of gypsum, up to several hundred metres thick, in which Triassic fossils are recorded (Krahl et al. 1986). A good example is exposed in, and around, a large working quarry in the west of the area (SSW of Mochlos; marked G in Fig. 15a), just above the Plattenkalk unit. This gypsum is entirely recrystallized, shows locally steep banding and is deformed by ductile folds. The sugary recrystallized gypsum includes numerous angular fragments of dark dolomitic carbonate (up to 15 cm long). The gypsum is exotic to the intact stratigraphic successions in the Chemezi area and is interpreted as having been extruded along the regional tectonic contact between the underlying Plattenkalk and the overlying Permo-Triassic units from a source of Triassic evaporite elsewhere in the basin. The possible regional tectonic significance of the Chemezi area will be considered after a summary of the comparable Vai area, below. Evidence from the Vai area The outcrops in the Vai Peninsula (Fig. 5) are located on the western and eastern parts of the peninsula with intervening Neogene-Recent sediments (Fig. 17). An initial non-trivial problem is how to correlate the tectonostratigraphy of these two outcrop areas. Recent workers agree that a stack of thrust sheets is present in the Vai Peninsula as a whole, as in the Chemezi area discussed above. The entire peninsula was mapped and described by Haude (1989). He interpreted the peninsula as a pile of thrust sheets, which were folded into a huge NNE-SSW-trending recumbent isoclinal fold, with the western limb mainly the right way up and the eastern limb inverted. This interpretation was largely followed by Krahl & Kauffman (2004). During this work it was found that by no means all of the successions within individual thrust sheets in the eastern Vai exposures are stratigraphically inverted, questioning the existence of a simple west-facing nappe structure. In the western peninsula area (Fig. 17d) most of the contacts between lithological units are sheared such that it is commonly difficult to

124

A.H.F. ROBERTSON A. rosso

WESTERN VAI AREA

#--10m

a

~-

A. rosso

b

u

100m

M. TOPLOU

Key to sections

Key to map

I ~ Limestone/marble/carbonate

Quaternary l ~ l Siliclastic/carbonateunit

r;~ Conglomerate Sandstone (psammite) [ ' ~ Shale(phyllite) I~ I~ I~

Uppervolcanic unit

I

|

0.5Km

Upper volcanic/sedimentaryunit ~.~ Schist/gneiss

Lowervolcanicunit

FI

Lowervolcanic/sedimentaryunit

Blackshale & limestone Schist/gneiss

I~

Plattenkalknappe

Fig. 17. Tectono-stratigraphy of the Phyllite-Quartzite unit exposed on the western side of the Vai peninsula, eastern Crete. (a-e) Representative cross-sections; (d) sketch map showing the locations. (See text for discussion and data sources.). distinguish between primary thrust contacts and sheared lithological contacts. However, several intact successions can be inferred in the western peninsula area (Fig. 17a-c). Details of these successions and measured logs will be given elsewhere. In general, moving structurally upwards

there is a lower unit dominated by relatively deep-marine, mainly andesitic volcanogenic rocks, including pelagic limestones (e.g. Ammonitico Rosso) and rare red chert, of inferred Early Triassic age, mainly based on dating of conodonts within the interbedded Ammonitico

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

I i

(

,.,,

kA

a 79~

]

I

125

2"

b -~%~-e ~ Vai beach 40

J [~

65.

'

Q

Palekastro

Tripolitza nappe Phyllite - Quartzite nappe

[T"I1 Piattenkalk nappe Neogene - Recent

50

~'~o

~._~

]'_.~ Megali --~.J K e p h a l a

Tripolitza nappe Siliciclastic deposits Permian pelagic limestone Melange High-grade metamorphic rocks Quartzitic & limestone conglomerates Nedtic carbonate Mixed siliciclastic/ carbonate Volcanogenic rocks

Kokino Kavo

9 ],m' k , v / . "

')

~

Fold axial trace

(~

~

Thrust

/

_ 0

M

500

- -

,,I

~ _a_

Normal fault Dip & strike of bedding

Fig. 18. Tectonostratigraphy of the PhyUite-Quartzite unit exposed on the eastern side of the Vai peninsula, eastern Crete. Inset: larger area including Toplou monastery and the west coast of the Vai peninsula (Fig. 17d). (See text for explanation and data sources.)

Rosso (J. Krahl, pers. comm). This volcanogenic unit is locally intersliced with schist and gneiss, and may include older (?Late Permian-Early Triassic) dark organic-rich shales and carbonates

near the base. The volcanogenic unit is overthrust by a mainly shallow-water succession of carbonates interbedded with siliciclastic sandstones and conglomerates of inferred deltaic origin. Clasts

126

A . H . F . ROBERTSON

O

O)

,

,

, .

.

,

> O

"~

E

g

"~~ ~ ~0

N [[][~ N E] ~] EqN) l @ NN qos/~g O^ml OUplO~l

o

X -<

n.

'"'" :" "'" "'" '"' I""11'"'II" fl"'I:I:I:II"'H:I'I:'H'I:H:I',I:I,I~'~IIH,H I ! ",' " ' " ' : " " ' ' " ' " ' " I ' " ' , ' " ' " " "" ,,~1!,:,,,..,,,~,,~,,o,O,,o o '~''-'r~ H.H ~ mI i' ! ; 'iiii'i~i ,,,,,, :,,,,,.,,,,,,,, ,.,,,.,,,,,,, ,,,,l:ll:lt,,,I,l:l,H,R,Ioil01,,H,l:l, , .i.

III:IIIII I~IIIII

r

,< A Oo.~.I,,.--~.-~ ."r,-~ T: -. :m, lo:O:O o.o o.o o.0 9 .o.o. O:o I# :o. oi/'A ,~ , ~ : ~ : ~ " ~ . i o : o'o. o:oo:o o:o o o:o: ~ :ooo:

~1~:

~ e~o

F ~ , ~ :" , :

.1"

" ,:"~'1",:.~.E!

,0:o 0"0 0"0 0"0

~

0

o : 0 o'.o o : o 0

"~'#

0 - 0 : O o o 0 o : ooo.o:o ,o

o,.o o ,o. o1,,,1:~ o,,,,,.. ,,oo~ ot!..oo :ooO.

I,:;;.EE',:I:,o ~ : _ ' _ "

;>

Id '~-

0[0"000000:~

O0000

~1',,1o~ ~ ~176ooooo

q

~,~

:

"E

g~ <

/I.o.l:o..~ o..~:o ~:o::~:o:~.o:,LILI~,~,,,9

,:o:,:o::,:o::, o:,:o:.,:o: '!~ C~q F~,~II

o,-~i

~oi,:o::,:o ::,:o:,:oh:o: ~:O:~l-I[ - ~ , . 9

O

o o ~ > -

Z O

I I~'-I.~H:,:,:'

>-

:= o

: , ~ i ,:o!:,i~176

,:o:I 14 I I o

I Ill bl: lli,:, :o: ,:o: ,: o :.,:o: :,: o: ,: oiHH I

~oo

O,,,~

r~

I "l:l"

I I" I " l > l

> 9

"

"~..

:> : >~ >.>...

: 9: 9

>::1o

Ho:I> >1~,~

:oo

;:'i~:il>.:l@l,,, I L ,

L~ ~W

s i i

~.1,,, >,, '>'

>o o

. 9

9

9 > ." " > :>>0 I1, I".,.,.I.> .. ~.>:~:il~on>~l> >l,,,g.l:o o~.1,> > 1 >,,,

9

I1i

I

I1"11:

~ o

'[~

l

>o " > )

'

.

>

:: H:,I,H, ,:, ':, 9 ,,.,.i,',l,,n > ,,,, I~ >o1.1,,I'"l~l '>. , . ,9 ~ ,.~ ,:, "~ e,,

[::;i

:= o

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS range from mainly carbonate to metamorphic basement-derived. The age of this succession is possibly Mid-Late? Triassic based mainly on lithological correlation with similar units exposed in the eastern side of the peninsula. The eastern peninsula area (Fig. 18) is more complex, lithologically and structurally. In general, four main eastward-dipping units were recognized during this work in ascending (present-day) structural order. Sedimentary logs were measured in each of these units (Fig. 19) and will be described in detail elsewhere. First, a mixed volcanogenic-siliciclastic unit ('arc unit' of Stampfli et al. 2003) is exposed in a small area (several square kilometres.) in the eastern part of the peninsula inland from Vai beach (Figs 18a, b & 19a, b). This outcrop correlates with a much larger volcanogenic succession of Early Triassic (Olenekian) age of the western area of the Vai Peninsula (Fig. 17a, b). Several samples of relatively fresh basaltic lava of sub-alkaline composition were chemically analysed from this unit (see Table 1). Samples from near the base of the succession (Fig. 16) show a subduction influence, as indicated by a pronounced negative Nb anomaly, similar to the lavas from the Chemezi area. Second, a mixed siliciclastic-carbonate unit is extensively exposed (Figs 18c, d & 19c, d), especially SE of Vai beach on a broad ridge running southwards c. 3 krn, as far as near Meridati (Fig. 18d inset), and is also well-exposed N W of Vai beach. The exposures, north and south of Vai beach, are likely to be offset by a down-to-thenorth normal fault (Haude 1989). The succession is equivalent to the 'Vai flysch, Upper forearc' unit of Stampfli et al. (2003). However, the succession reported here differs, as it necessary to take account of folding on north-south axes as indicated by changes in the younging direction (Figs 18 and 19). Third, a coarse quartzitic and carbonate conglomerate unit (Figs 18e, f & 19e, f) is exposed north of Vai beach, near the coast. This comprises siliciclastic conglomerates and breccias equivalent to the 'Vai flysch upper forearc unit' of Stampfli et al. 2003. Units 1-3 can be generally correlated with the Triassic coarse clastic units exposed in the Chemezi area. Fourth, there is a slice of m61ange, c. 300 m thick, equivalent to the 'olistostrome' of Stampfli et al. 2003 (Figs 18f and 19f). This is associated with a slice of high-grade metamorphic rocks (Fig. 18f). There is no known equivalent of this m61ange unit in the Chemezi area. However, it is critical to the interpretation of alternative

127

settings, as it has been interpreted as a Palaeotethyan accretionary complex in Model 2 (Stampfli et al. 2003) and so will be discussed in some detail below. The sedimentary matrix of the m61ange, which is well exposed along the coast, comprises intercalations of coarse clastic sediments and mud rocks (phyllites), ranging from reddish to pink and green, owing to local oxidation-reduction effects. Individual matrix-supported conglomerates, typically c. 1 m thick, and with scoured bases, grade from pebbly conglomerate into coarse quartzitic sandstone, then shale. Other interbeds include abundant limestone talus. Other coarse clastic sediments are interpreted as high-density turbidites and classical turbidites. Sedimentary way-up evidence is indicative of outcrop-scale isoclinal folding on north-south axes. However, the eastward-dipping m61ange fabric appears to be mainly inverted based on graded bedding and basal scour structures in most debris flows and sandstone turbidites. Individual m61ange blocks are best exposed several hundred metres inland. The most common m61ange block is limestone, which shows local transitions to sedimentary limestone talus. There are also rare small blocks of pelagic sediment. For example, a small block (0.6 m • 0.3 m) of red radiolarite was observed within quartzitic and carbonate-rich debrites. Small lenses and blocks of red shale are also present, together with several blocks of pink pelagic limestone (Ammonitico Rosso), up to c. 15 m in size, as seen in the east. Several blocks of andesitic lava and basaltic lava breccia are also present. The Ammonitico Rosso is inferred to be of Anisian age, whereas the red radiolarite was dated as Late Anisian (Mid-Triassic) (Stampfli et al. 2003). Blocks of both chemically andesitic and withinplate-type basalt were reported (Stampfli et al. 2003). One lava block analysed during this study shows a subduction-influenced signature (Fig. 16). In addition, in the highest structural position along the coast there is a well-exposed, remarkably intact, relatively undeformed succession of pelagic limestone, shale and sandstone (Figs 18g & 19g). No direct equivalent of this is known in the Chemezi area, although the pelagic limestones at the base of the succession in the eastern part of the Vai Peninsula could be equivalent to the pelagic limestones in the upper part of the Upper Permian-Lower Triassic succession at Chemezi. This eastward-dipping composite unit is well exposed between Megala Kephali and Kokino Kavo (Figs 18g and 19g), the latter name reflecting a brilliant red-brown-violet colour

128

A.H.F. ROBERTSON

of metashales and metasandstones exposed on headlands. The basal contact with the m61ange is associated with strong shearing, isoclinal folding, duplex formation, tension gashes and carbonate veining. Kinematic indicators indicate normal fault displacement (in present orientation). Above this contact, the unit begins with strongly deformed thin- to medium-bedded black limestones and black shales (c. 8 m). Grading and sharp bases in less deformed medium-bedded limestones near the contact are indicative of (local) stratigraphic inversion. Associated folded phyllite varies from green to purple, probably related to diagenetic alteration. Southwards, the succession passes into alternating thin-, medium- and locally thick-bedded grey limestone-phyllite alternations (c. 80 m thick). The limestones contain deep-water conodonts of Mid-Permian age (Stampfli et al. 2003; Krahl & Kauffman 2004). The individual pelagic limestone beds are laterally continuous and exhibit internal fine parallel lamination, suggestive of an origin as fine-grained calciturbidites, together with scattered nodules and lenticles of black chert of diagenetic origin. There is then a sharp, but apparently depositional contact between the fine-grained limestone and very coarse brecciaconglomerate above, composed mostly of quartzitic clasts. Traced laterally, this contact appears to cut stratigraphically downwards into the hemipelagic limestones and is therefore interpreted as a low-angle erosional unconformity. Above this contact, individual well-bedded (eastward-dipping) lenticular conglomerates and sandstones ('Vai flysch, lower forearc unit' of Stampfli et al. 2003) can be traced laterally (up to several hundred metres) before wedging out. Individual depositional units, up to 3 m thick, begin with clast-supported, quartzose conglomerates (clasts up to 4 cm in size), grading upwards into sandstones with muddy tops. Excellently developed grading in many beds confirms that this succession is the right way up. In all, nine graded packages of conglomerate-sandstone, each up to 2.5 m thick, were observed. The average thickness and grain size of each depositional package decreases upwards, until the exposed succession (in coastal cliffs) culminates (at Kokino Kavo) in up to 100 m of reddish metamud-rocks (phyllites), interbedded with thin- to medium-bedded, graded metasandstones (Fig. 19g). Although this succession generally thins and fines upwards, several thick, relatively coarse beds persist to the highest exposed levels along the coast. In thin section, the sandstones are composed of relatively uniform, well-sorted grains of mainly quartz and quartzite, with

minor muscovite and granular iron oxide. This spectacular succession remains undated but is probably of Triassic age based on regional comparisons with other facies. I n t e r p r e t a t i o n : rift-related settings

In the eastern Vai exposure area, the volcanogenic-siltstone-carbonate unit (Unit 1) of Early-Mid-Triassic (Late Scythian-Anisian?) age records the extrusion of mainly subductioninfluenced andesites, as massive flows, volcanic breccias and rare pillow lavas, interbedded with volcaniclastic sediments, tufts, pelagic carbonates, with deep-water conodonts (of Early Triassic age) and rare metacherts (jaspers). This volcanogenic succession (or a lateral equivalent) is a possible source for Mid-Triassic island-arc tholeiite (IAT)-type basalt blocks and also the Ammonitico Rosso and chert in the m61ange, assuming volcanism continued into the MidTriassic. A possible exception is one recorded instance of a block of within-plate basalt in the m61ange, which Stampfli et al. (2003) interpreted as remnant of an oceanic seamount. However, geochemically similar basalts also occur in rift settings, e.g. as in western Crete, discussed above. The mixed siliciclastic-carbonate unit (Unit 2), widely exposed in both the eastern and western outcrop, is interpreted as a shallow-marine to locally non-marine, channelized deltaic sequence of Early Triassic (Early Olekenian)-Late Triassic (Carnian-Norian?) to Early Jurassic (Rhaetian?) age (Stampfli et al. 2003). The conglomerates accumulated in a proximal deltaic setting where metamorphic basement was exposed. The interbedded limestone conglomerates probably record the erosion of a carbonate platform, in response to relative sea-level changes or local tectonics. A more intact carbonate platform, rich in coral, became established during the Norian-Rhaetian. The conodont assemblage within these limestones is reported to be similar to that elsewhere at the base of Tripolitza carbonate platform succession, which may suggest that the PhylliteQuartzite succession originally continued upwards into the Tripolitza platform (Stampfli et al. 2003). However, the actual contact is now a major structural and metamorphic discontinuity, probably a major Alpine thrust that was reactivated by exhumation. The structurally overlying coarse quartzitic and carbonate conglomerate unit (Unit 3), of inferred late Early-Mid-Triassic age, accumulated on relatively steep subaqueous slopes dominated by gravity-flow processes, probably as proximal fan deltas on a linear margin. These

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

1Km p t_.._a / ~ I

"

V" V"

\

-.::'::.

[

_-_-

(

v.

: ~"~--"'-~':~/P ~_ - X

Y

Sea of Crete

o0o,e 2 d L N-~'---..--~'-Vassilikon__~X'_"

i i

_-

129

I

I

I

I

I

I

~ lOOm

Wa,'aleoo

i

_

..i

[]

Clastic& volcanogenic rocks Vassilikon carbonate Mainly quartzose sandstone

Vassilikon [

-

-

l

O

O

~ m

~

X

Mainlyrnudrocks

l"771Talea Od unit

"~ ~" v' "V "v . sl. .

.

9

,~..Y Fig. 20. Occurrence and tectonostratigraphy of the Phyllite-Quartzite unit, west of Iraklion in central northern Crete. (See text for explanation and data sources.) sediments are unlikely to represent a channelized mid-fan of an idealized deep-sea fan, as suggested by Stampfli et al. (2003), especially as the conglomerates are tabular to broadly lenticular rather than markedly channelized; finer grained inter-channel fan muds and turbidites, as expected in mid-fan settings, appear to be absent. The abundance of quartzitic clasts may reflect preferential preservation of erosionally resistant lithologies derived from a nearby continental basement. The abundance of gneiss clasts confirms the existence of a metamorphic basement. This unit may be broadly coeval with, but more distal than the deltaic, to shallow-marine mixed

siliciclastic-carbonate unit (unit 2), and a moderate depth (e.g. outer shelf-type) rather than abyssal setting is preferred here. The m61ange unit includes pelagic carbonate, radiolarite and both within-plate basalt (WPB)type and subduction-influenced basalts within a sheared, siliciclastic and volcaniclastic matrix. The age of the m61ange is likely to be MidTriassic, of similar age to, or slightly younger than, the exotic blocks present. The structurally highest unit, exposed only on the east coast (Unit 4), is interpreted as turbidites and mass flows derived from a continental source area, controlled by relative sea-level changes. A

130

A.H.F. ROBERTSON

relatively proximal setting is particularly suggested by the presence of occasional relatively thickbedded debris flows and high-density turbidites towards the top of the exposed succession. This does not support the suggestion that these sediments accumulated on the distal outer part of a deep-sea fan (see Stampfli et al. 2003), where thinner bedded and more hemipelagic sediments would be expected. The small scale and thickness of the sedimentary units is inconsistent with regional-scale settings such as deep-sea fan in a fore-arc basin, as implied in Model 2. The bright red-purple colour ('violet schists') probably reflects secondary oxidation of iron, possibly controlled by a high amount of organic matter originally deposited in these sediments (as noted in the Chemezi area).

Evidence from the Iraklion area, central Crete The Phyllite-Quartzite unit is also extensively exposed in central Crete, west of Iraklion (Fig. 20), where two contrasting units are separated by an undated interval of coarsely crystalline marble (Vassilikon unit; Epting et al. 1972; Krahl et al. 1988). It is important to note that the lower of these units shows similarities to the Phyllite-Quartzite unit of western Crete, whereas the upper is more similar to the successions exposed in eastern Crete. This is one area in Crete where the two contrasting units of the Phyllite-Quartzite unit are known to occur together. The lower part of the exposed succession, structurally above the Talea Ori-Plattenkalk unit, is dominated by soft-weathering, pale micaceous phyllite with alternating packages of medium- to thick-bedded sandstones and dark shale. Laterally continuous medium beds of softweathering, fine-grained micaceous sandstones are interbedded with dark phyllite. Several cliffforming intervals of medium- to thick-bedded, laterally continuous quartzitic sandstones (in beds up to 3 m thick) include subordinate interbeds of black phyllite. The highest exposed levels beneath the Vassilikon marble comprise strongly cleaved grey phyllites with psammitic alternations. The gently north-dipping Vassilikon marble is virtually massive, with evidence of pervasive anastomosing shear zones in the lower part. A sharp contact with the underlying phyllites is indicative of a tectonic contact. The upper contact of the Vassilikon marble, exposed on the dip slope to the north, is also interpreted as tectonic. Above this, a contrasting

Phyllite-Quartzite succession includes thinbedded limestones, volcaniclastic sandstones and tuffaceous sediments, followed by massive, locally porphyritic andesitic lava and then further marble intercalations. Higher levels of the succession, exposed towards the coast, include massive andesitic lavas, lava breccias, volcaniclastic sandstones and tuffaceous sediments, >250 m thick.

Interpretation: additional rift-related settings The lower Phyllite-Quartzite unit exposed in the south is lithologically comparable with the Phyllite-Quartzite unit in western Crete, as discussed above, although inferred sandstone turbidites are thinner bedded and the succession is more shaly overall, suggestive of a relatively distal setting. The undated Vassilikon marble might represent a slice of Hercynian basement, Triassic neritic limestones, or a strongly recrystallized fragment of later Mesozoic carbonate platform rocks, the last-mentioned being plausible as the marble is homogeneous, comparable with the Mesozoic Tripolitza carbonate platform. The overlying mixed terrigenous-volcanogenic succession includes andesitic lavas that have geochemical affinities with the Triassic volcanic rocks exposed in eastern Crete (Seidel 1978) and is interpreted as part of a similar rift-related setting.

D&cussion o f tectonic settings in eastern and central Crete Most of the evidence from the Chemezi, Vai and Iraklion areas of the Phyllite-Quartzite unit is consistent with Model 1, in which all of the units, including the volcanic rocks, developed in a proximal to more distal rift setting, dating from the Permian, or earlier. In this interpretation the volcanic rocks relate to a pulse of Early-Mid?Triassic extension. The pre-existing sediments were uplifted related to flexural uplift of the rifted margin during Mid-Triassic time, ushering in shallow-water to non-marine deposition adjacent to Hercynian continental basement during Late Triassic-earliest Jurassic? time. This could be same regional flexural effect that also affected the Phyllite-Quartzite unit in western Crete and the Talea Ori (Plattenkalk) unit. The possible cause of this inferred flexural uplift is considered in the discussion section near the end of the paper. In Model 1, the contrasting exposures in the Iraklion area could be explained as different depocentres within a palaeogeographically varied rift setting. The Vassilikon marble might record

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS the remnant of a former intra-rifl high on which neritic carbonates accumulated, possibly during Late Triassic-later Mesozoic time. On the other hand, there is little evidence that the Iraklion outcrops record an actual Palaeotethyan suture, as in Model 2, as there is no evidence there of accreted oceanic material (deep-sea sediments or oceanic crust), or of any more intense deformation than seen elsewhere. Relatively intact successions are preserved in both the upper (eastern Crete-like) and lower (western Crete-like) thrust sheets there. In Model 2, the Vai, Chemezi and Iraklion areas restore as a series of volcanic arc, basement (backstop), fore-arc and accretionary tectonic settings. The andesitic volcanic rocks and related sediments record part of the arc. The high-grade metamorphic basement slices represent part of the backstop of a subduction zone. The deepwater clastic and carbonate sediments ('Vai flysch') accumulated in a proximal to distal forearc basin and finally the m61ange unit (eastern Vai area) represents a preserved fragment of an accretionary wedge created by northward subduction of Palaeotethys. The younger, coarser conglomerates and shallow-water carbonates accumulated in a post-collisional, transgressive setting in this model. Some problems with Model 2 include the following. (1) The inferred Permian-Lower Triassic 'forearc' sediments exposed in eastern Crete (i.e. Vai and Chemezi areas) contain terrigenous silt, but no arc-derived volcaniclastic sediment, as expected for a fore-arc basin. These sediments appear to have accumulated in a quiet, deep-water environment, unlike forearc basins that are typically unstable, and generally include turbidites, debris flows and slump deposits, rich in volcaniclastic debris. (2) There is no definite record of any related accretionary wedge (e.g. trench-type sediments, or slices of oceanic crust). The blocks of within plate-type basalt within the Middle Triassic m61ange unit in the eastern Vai outcrop could be related to Triassic rifting, rather than fragments of Palaeotethyan seamounts. 'Subduction-erosion' (e.g. von Huene & Scholle 1991) might account for the lack of an accretionary wedge. However, many comparable modern settings, including the eastern Mediterranean Sea south of Crete (Chaumillon et al. 1996) and the Gulf of Makran (Glennie et al. 1990) are associated with the development of accretionary prisms, which have a high potential for preservation in the stratigraphical record. (3) No major Triassic magmatic arc, e.g. involving large-scale central-type volcanism, is

131

known anywhere in the region. Triassic volcanic rocks that exhibit a subduction-related chemistry are interbedded with continentally derived subaqueous slope material, including coarse terrigenous debris flows and turbidites; there is no evidence of volcanic build-up, typical of continental margin arcs (e.g. Cascades, Andes). Also, air-fall tufts typical of Andean-type margins (e.g. Andes) are sparse. The absence of large central-type volcanoes is surprising, as modern back-arc rifts typically develop by the splitting of developed volcanic arcs (e.g. the Marianas and Tonga arcs; and the SW Pacific; see Robertson 1994, for literature). The fact that both the Pindos and Vardar Triassic basins are inferred to be back-arc basins in Model 2 would lead one to expect the existence of a substantially developed magrnatic arc, which in reality does not exist. On the other hand, the volcanic rocks are mainly andesitic and do show a subduction-related geochemical influence. This is consistent with Models 2, 3 and 4, but apparently not with Model 1. Possible reasons for this discrepancy are considered in the discussion and conclusion section. (4) The expected > 6 0 k m width of crust between the inferred trench, arc, continental backstop and back-arc rift appears to be absent. For example, in the Vai area, all four of these units, as inferred by Stampfli et al. (2003), appear locally as thin thrust sheets one above the other. Even taking into account Cenozoic deformation, too many plate-scale processes are inferred in too small an area. In the convergent Model 2, tiny slivers from a range of tectonic settings (oceanic, fore-arc continental) originally at least tens of kilometres apart, were preferentially incorporated in a 'Cimmerian' thrust belt during pre-Jurassic time (5) Collision of a 'Cimmerian' continent with the Eurasian margin during latest Triassic time would be expected to result in flexural collapse of the continental margin to form a regional, turbidite-filled foreland basin, yet neritic deposition prevailed, after a break in deposition during Mid-Triassic to Early Jurassic time. (6) Perhaps aware of some of the above difficulties, Ziegler & Stampfli (2001) suggested that only a 'soft collision' and 'docking' took place of the Cimmerian continent with the Eurasian (Pelagonian) margin, leaving no perceptible stratigraphic or structural record. This was, however, contradicted by Stampfli et al. (2003), who inferred more pervasive collision and metamorphism.

132

A.H.F. ROBERTSON However, there is, as yet, no firm evidence of a significant regional 'Eo-Cimmerian' compressional or metamorphic event dating from Late Triassic (Carnian-Norian) time in Crete or the Peloponnese (see below). For example, detailed structural studies in eastern Crete show that the main regional D2 event affects both the Phyllite-Quartzite unit and underlying Plattenkalk (Lower Cenozoic Kalavros beds) and thus must be Alpine. DI is represented by rare east-westtrending, isolated folds and a beddingsubparallel cleavage (Zulauf et al. 2002) of uncertain origin. There is no obvious petrographic evidence of an Eo-Cimmerian metamorphic event (e.g. relict textures) of upper greenschist facies, or higher grade, despite reported conodont colour indices suggesting temperatures of c. 500 ~ i.e.c. 100~ in excess of the temperature estimated for the reported regional HP-LT alpine metamorphism (Stampfli et al. 2003). However, the metamorphic grade of the lower thrust sheets (Plattenkalk, Tripali and PhylliteQuartzite) varies considerably throughout Crete (e.g. Zulauf et al. 2002) and so this may not be a problem.

Evidence from the Peloponnese The tectonostratigraphy of the Peloponnese is similar to that of Crete, with a similar pile of thrust sheets exposed in the same order. Most relevant outcrops are located in the southern and central Peloponnese, but there are also small isolated exposures in the NW Peloponnese (e.g. Zarouhla-Feneos area; De Wever 1975). Counterparts of the Plattenkalk and the PhylliteQuartzite units, and possibly the Tripali unit, are present, and above this there are counterparts of two different facies-associations representing the Phyllite-Quartzite unit. In general, these units are less well dated than in Crete. As in Crete, the HP-LT metamorphic units are structurally overlain by a low-grade metamorphosed to unmetamorphosed shelf to carbonate platform succession. There are, however, several differences between Crete and the Peloponnese, which make some discussion useful here in the attempt to discriminate amongst regional tectonic settings. First, in Crete the Mesozoic GavrovoTripolitza carbonate platform that overrides the entire underlying thrust stack is depositionally underlain only by a thin intact sedimentary succession (Ravdoucha beds). However, in the Peloponnese the equivalent Gavrovo-Tripolitza carbonate platform is stratigraphically underlain,

albeit with a sheared contact, by a much thicker Triassic unit, known as the Tyros beds, which include both volcanic and terrigenous sedimentary units. These are critical to an understanding of the Triassic rift history of the Pindos basin to the north. Second, fragments of unmetamorphosed Palaeozoic sediments, known locally in the Peloponnese, could record part of a preexisting continental basement. Third, there have been reports of a possible Palaeotethyan accretionary prism in the Peloponnese, which if correct would constitute important evidence for Model 2. In the discussion below evidence from the equivalents of the Plattenkalk unit and equivalents of the western Crete Phyllite-Quartzite unit will be summarized, highlighting features that are relevant to understanding the tectonic setting, although a wealth of new information available warrants a fuller discussion elsewhere. It will be concluded in this section that most of the evidence is again consistent with the rift-related Model 1, although as in eastern and central Crete some of the Triassic volcanic rocks appear to be anomalous, as they record a subduction-related geochemistry.

Metasediments of the structurally lower units in the Peloponnese Tectonostratigraphy

The Plattenkalk in the Peloponnese, as in Crete, is dominated by platy pelagic metalimestones with replacement chert, of inferred JurassicCretaceous age, passing into Eocene flysch (Lekkas & Papanikolaou 1980; Papanikolaou & Skarpelis 1986). The typical Plattenkalk facies is underlain by poorly dated 'Permo-Triassic' phyllites, quartzite and conglomerates (Psonias 1981). According to some workers (Dittmar et al. 1989; Dittmar & Kowalczyk 1991) these facies form the stratigraphic base of the Plattenkalk unit. However, in the SW Mani Peninsula, marbles (Mani marbles) and associated metaelastic facies, correlated with the Plattenkalk, are reported to be locally tectonically inverted and thrust over the Phyllite-Quartzite unit (Alexopoulos & Lekkas 1999). This raises the possibility that metaquartzose sediments underlying the Plattenkalk generally in the SW Peloponnese could represent thrust slices of the PhylliteQuartzite unit rather than a true stratigraphic basement, as some workers have previously assumed. The Phyllite-Quartzite unit in the Peloponnese is traditionally divided into three 'nappes', although these are only partially exposed in individual areas and are separated by Cenozoic

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS i

l

!

133

l

I I

~ 1 # 0 ~

100

t0 !

...-

'= 2

1 Kithera

thick ~ ]

i

,4, Vresthena -

'' 5 1 Kastania

(NW Peloponnese)

! !

3,

Neapoli

~l&O

I

Tagetos I I

I I

%>0

:

]37ON~

Jtl5..'1\

"~""~

"~

9 \

\

ee~8Mola

F5m ~'~

G Graniticgneiss H Serpentinised harzburgite

10

~j"

Finiki

~

Tyros unit

~

Ama unit (Phyllite- Quartzite)

[]-~ Plattenkalk I

Kastania (Tagetos)

'

/

36"

~.arouhl~30'

Gutfof

-'" lit IP } [t t|

e

1

I.~ o20'E . f 2- ~5

Nw-F~2 2

Kithera . ~

SW

i 1 : ~,,

.1

Key to map

"'"

Pelagia

f~k~,~,H Peloponnese

I Ill !

VM oo

V

! ....

oor

v

v

|

II/t

V 7~176 ,;,

v v v

v

"

vv

ool .,.., -" ".L:I F'-7

V

7 Krokee

v

v

v'

V

v

V w

>300 v ,m

OOQ

ooo vv

AA

v

till

IOOOi

v v v v v v v

8

Aridia/ Molai

V

~ ~

Silicictuff Shale

r~ r~

Sandstone ] Conglomerate/( Volcaniclastic breccia j rocks Basic- intermediate lava Intermediate- silicic lava

//tl

~.97,r A A

V V

v

Gypsum Terngenous shale

....

V g

~ ~

d9 ,d. o'r . IkA

v g

Neritic carbonate

aa

111111"

v

v

Volcanogenic rocks

ooC

.....

IIII v

V

II/I v

v

v

I

Temgenous rocks

r - ~ Limestoneconglomerate

-.Z

99 .--:.--:.

T ~

V ~.~,

!

~

L_3 ~ Km

1

Peleponnese

;

"%'-~ ~'~ 3 ~.

Lakonia

~ I I 1[,,

v

9 Marion

I

10

Tyros

! I

11

Zarouhl~

Key to logs

Feneos

Fig. 21. Measured logs of the Plattenkalk unit, including the Arna unit, and the Phyllite-Quartzite unit (Tyros unit) in the Peloponnese with locations (inset). (See text for explanation and data sources.)

134

A.H.F. ROBERTSON

sedimentary basins (Ktenas 1924, 1926; Thi6bault 1982; Triboulet & Bassias 1986). The 'lower nappe', dated as partially Triassic (Brauer et al. 1980; Doert et al. 1985), together with the Lower Cenozoic? meta-flysch (Papanikolaou & Skarpelis 1986), has a total structural thickness of c. 2000 m in the Parnon massif and c. 1300 m in the Taygetos massif (e.g. Doutsos et al. 2000; Xypolias & Doutsos 2000). This unit has experienced intense deformation and HP-LT metamorphism (c. 400~ and 10 kbar), although mineral assemblages indicative o f P - T conditions are commonly lacking; also P - T conditions may vary regionally as in Crete. The Phyllite-Quartzite unit is dominated by meta-mudrocks and metaquartzitic sandstones. However, exposures in the Tagetos Mountains, termed the Arna unit, include metashales (phyllites), metaquartzose conglomerates, MORB-type tholeiitic metabasalt and rare harzburgite (Skarpelis 1982). Occasional occurrences of harzburgite and metabasic igneous rock occur elsewhere in the Peloponnese (Thi6bault 1991; see below). In NE Kythira (Fig. 1), the Phyllite-Quartzite unit is associated with granitic gneiss of Hercynian age (Xypolias et al. 2006). Also on Kythira, according to Stampfli et al. (2003), Danamos (1992) has reported the presence of metabasic rocks and metacherts (lydites) that are said to exhibit primary contacts; if correct, this could be indicative of an origin as accreted Palaeotethyan oceanic crust, consistent with Model 2 (Stampfli et al. 2003). The ultramafic rocks and basic lavas within the Phyllite-Quartzite unit generally (including the Arna unit) have been interpreted to include exotic accretionary material (Skarpelis 1982), and could thus represent part of a Palaeotethyan accretionary complex (Stampfli et al. 2003). An alternative view is that these thrust-intercalated mafic-ultramafic rocks were derived from the Pindos ocean to the east, in the Cycladic region, during Early Cenozoic Alpine deformation and metamorphism (Papanikolaou 1996-1997; Pe-Piper & Piper 2002). In the literature a 'middle nappe' of the Phyllite-Quartzite unit is reported to comprise meta-clastic rocks, metacarbonates, metashales and sparse metavolcanic rocks, of possibly Carboniferous, Permian and Triassic age (Thi6bault 1982; Triboulet & Bassias 1986; Bassias & Triboulet 1994). Lithologies of no higher metamorphic grade than greenschist facies attributed to this unit are reported to be exposed in areas including the NW Peloponnese (Zarouhla and Feneos areas; De Wever 1975; Pe-Piper 1983), the SE Peloponnese (near Molai) and near Kalamata (Verga). However, these successions (Fig. 21) remain poorly dated and are reported to be overlain by carbonates similar to the

Gavrovo-Tripolitza unit (i.e. without any overlying 'upper nappe'), thus questioning the reality of a 'middle nappe' as a regionally significant tectonic unit. The mainly volcanogenic 'upper nappe' (Tyros beds), of greenschist-facies metamorphic grade, is widely exposed and less deformed. Dating is primarily based on an inferred depositional passage upwards into well-dated shallowmarine carbonates of the Gavrovo-Tripolitza zone of Late Triassic (locally Carnian) age (Thi6bault & Kozur 1979; Lekkas & Papanikolaou 1980; Skarpelis 1982). Successions in the SW Peloponnese (Stephania-Krokee area) are commonly assumed to be Early Triassic in age (Doert et al. 1985), but remain poorly dated. In the SE Peloponnese volcanic rocks are reported to be interbedded with carbonates and clastic sediments that are well dated as Carnian in age (Thi6bault & Kozur 1979; Brauer et al. 1980). According to Gerolynatos (1994) the Tyros unit as a whole comprises a Permian to Upper Triassic, mainly sedimentary succession with volcanic intercalations mainly of Scythian and Late Triassic (Carnian-Norian) age. However, it is questionable whether an intact stratigraphic succession is anywhere present. The volcanogenic rocks range from basalts and basaltic andesites to dacites in different areas, with commonly pyroclastic and tuffaceous sediments, previously believed to be mainly nonmarine, together with minor intrusive rocks (Pe-Piper 1983). Extensive geochemical studies have showed that the Triassic volcanic rocks are largely calc-alkaline, but locally tholeiitic, to alkaline in composition (Pe-Piper 1983; Pe-Piper & Piper 2002; see Degnan & Robertson 2006). The tectonic settings of eruption were seen as either intra-continental extension-related (Dornsiepen & Manutsoglu 1996), or subductionrelated (Pe-Piper & Piper 2002). Plattenkalk

( Mani unit)

The lowest unit, the Plattenkalk, locally termed the Mani unit, is structurally overlain, first by the HP-LT Phyllite-Quartzite unit (including the Arna unit) and then by the LP-LT Tyros unit, followed in turn by a deformed sedimentary transition to the Gavrovo-Tripolitza unit. These two units are separated by a regional low-angle tectonic contact, interpreted in different areas as a thrust fault (related to subduction), or a low angle-extensional detachment (related to exhumation). The base of the Plattenkalk unit, as exposed in the eastern Tagetos (e.g. near Kastania, south of Arna; Fig. 21, log 6), is a low-angle tectonic

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

135

contact, underlain by thick-bedded quartzose sandstones and phyllites that are lithologically very similar to the Phyllite-Quartzite unit of western Crete or the SE Peloponnese (see below). These rocks are thus unlikely to represent a true stratigraphic basement (Doert & Kowalczyk 1985), or to be equivalent to the shallow-water carbonates of the Talea Ori unit in Crete (i.e. Sisses and Fodele units).

meta-sills may also be present. Individual, metabasic units are mapped as lenses up to 4 km long (see Papanikolaou & Skarpelis 1986). In addition, a small exposure of glaucophane-bearing metabasic rocks (several metres thick) with WPB chemical affinities occurs in the N W Parnon massif (near Lakkomata), associated with small bodies of serpentinized ultramafic rock (Tribolet & Bassias 1986).

P h y l l i t e - Q u a r t z i t e unit

Metaserpentinite. The Phyllite-Quartzite unit includes small bodies of serpentinized ultramafic rocks at five well-documented localities, the first three of which mentioned below were studied in the field. All of these occur near the overlying tectonic contact with the Gavrovo-Tripolitza unit and appear to have undergone similar HP-LT metamorphism and deformation as the enclosing metasedimentary rocks. First, in the Tagetos massif, the Arna unit includes a lenticular, north-south-trending exposure of antigoritic harzburgite, several hundred metres long by several tens of metres wide (Skarpelis 1982; Fig. 21, log 1). Second, on Kythira, harzburgite is located (south of Ayia Pelagia; Fig. 21, log 2) within terrigenous metasediments, near the contact with the overriding Gavrovo-Tripolitza carbonate platform. Third, in the N W Parnon massif (at Vresthena; Fig. 21, log 4) serpentinized harzburgite forms a lens (c. 10 m thick) within terrigenous sediments. This is located several tens of metres beneath the overriding GavrovoTripolitza platform carbonates. Fourth, a smaller body (a few metres) elsewhere in the Parnon Massif (Agios Petra) is mapped as occurring directly along the tectonic contact of the Arna unit with the Gavrovo-Tripolitza carbonates (Skarpelis 1982). Finally, a small body of sheared serpentinized ultramafic rocks (possibly including dunite) is associated with phyllites and minor metabasic igneous rocks in the NE Parnon massif, at Lakkomata (Triboulet & Bassias 1986).

The Phyllite-Quartzite unit shows considerable lithological variation throughout the Peloponnese, as follows. Metasediments. In most areas the succession is dominated by quartzose sandstones, shales (locally carpholite-bearing) and subordinate quartzose conglomerates. The sandstones are most thickly bedded and coarse grained in the far SE (i.e. on Kythira) and in the Napoli area (Fig. 21, logs 2 and 3). However, sandstones are generally thinner, finer grained and more thinly bedded further north (e.g. Vresthena; Fig. 12, log 4) and in the NW Peloponnese (Fig. 21, log 5), although even in these areas occasional intercalations of relatively thick-bedded ( > 0 . 5 m ) and coarsegrained (conglomerate grade) sandstones are present. In all cases, the sandstones are mainly quartzitic and comprise well-rounded grains, as seen in western Crete. The Arna unit in the type area of the Tagetos Mountains (Fig. 21, log 1) is unusual, as the background grey or black (chloritoid-bearing) phyllites include numerous conglomerate intercalations composed almost entirely of quartzite clasts, ranging from clastsupported to locally matrix-supported conglomerates. Clasts (up to 80 cm in size) vary from angular to rounded, and locally to well rounded. Conglomeratic horizons, up to c. 20 m thick, can be traced laterally for hundreds of metres along strike, but no intact succession can be recognized. Metabasic igneous rocks within the Arna unit. In the type Arna unit in the Tagetos, the metasediments (mainly phyllites and quartzite conglomerates) are locally intercalated with metabasic rocks of MORB type (Skarpelis 1982). At one locality (Malevos, near the NeohoriGeorgitza road; see Papanikolaou & Skarpelis 1986) metabasalt, c. 100 m thick occurs within dark pelitic metasediments but the upper and lower contacts are poorly exposed. Further south (e.g. near Gorani, 14 km south of Arna village) metabasic rocks include definite volcanic breccias and appear to be interbedded with metasediments, including quartzitic conglomerates; some

Granitic gneiss. On the island of Kythira (Fig. 5) the Arna unit is associated with a small body of granitic gneiss (near Ayia Pelagia) that was recently shown to be of Hercynian age (Xypolias et al. 2006), similar to eastern Crete (see Romano et al. 2006). Metagranitic rocks, affected by neotectonic extensional faulting, occur near the coast, just north ofAgios Pelagia; these are rocks locally overlain by carbonates of the GavrovoTripolitza unit and are in faulted contact with unmetamorphosed, Triassic? sandstone, shale and limestone of the Pindos zone.

136

A.H.F. ROBERTSON

Interpretation: a rifted continental basement The quartzitic sandstones were derived from a continental setting, as in western Crete. Also, the Upper Palaeozoic gneissic rocks exposed on Kythira are suggestive of the existence of a Hercynian basement, as in eastern Crete. The serpentinized ultramafic rocks (mainly harzburgites) are likely to represent remnants of oceanic mantle (either oceanic or subcontinental). The MORB-type rocks in the Tagetos appear to be at least partially interbedded with metasedimentary rocks, rather than entirely exotic units. These associated sediments include quartzose conglomerates that are lithologically very similar to the Mani unit of western Crete, of inferred latest Triassic age. These volcanic rocks might conceivably relate to opening of the adjacent Pindos ocean during the Late Triassic. However, an origin as allochthonous thrust slices related to Alpine deformation cannot be ruled out in view of the strong deformation and high pressure metamorphism (Skarpelis 1982). In addition, the WPB-type metavolcanic rocks in NW Parnon (Lakkomata) might be of oceanic origin (seamounts), as they are associated with serpentinite, although Thirbault (1991) favoured an intracontinental origin. Indeed, the ultramafic rocks in all cases occur enclosed within terrigenous clastic sediments and lack evidence of significant amounts of other possible ophioliterelated rocks (e.g. cherts, gabbro, etc.). Also, the existence of metacherts and metabasic volcanic rocks on Kythira was not confirmed. In all cases the meta-ultramafic rocks occur in lenses or pods just beneath the major thrust fault or extensional detachment at the base of the Gavrovo-Tripolitza carbonate platform. They are thus entrained within or close to a profound tectonic and metamorphic discontinuity. It is, therefore, probable that these lithologies represent exotic units that were exhumed from a deep subduction environment related to exhumation of the HP-LT Mana (Phyllite-Quartzite) unit. It is possible that the serpentinites ultimately originated as fragments of subducted Pindos ocean (in the Cycladic region) that were later exhumed as diapiric pods in response to deep-seated out-ofsequence thrusting. There is, thus, no clear evidence that the ultramafic rocks record parts of a Palaeotethyan accretionary complex, which is otherwise not supported by field evidence within the Phyllite-Quartzite unit in the Peloponnese. Summarizing, the structurally lower HP-LT unit in the Peloponnese are consistent with a rift-related setting, as in Models 1 and 3. The claimed evidence of a Palaeotethyan accretionary complex, apparent evidence supporting Models

2 or 3 (e.g. on Kythira island), can now be discounted.

Evidence from the structurally higher units in the Peloponnese Triassic volcanic-sedimentary Tyros unit During this work it was found that an overall succession in this unit divides into a lower part, which is mainly volcanogenic, and an upper part, which is mainly terrigenous. The 'intermediate nappe' is here not considered to be a regionally significant unit (see below), but can instead be correlated with the traditional 'upper nappe'. The lower part of the volcanogenic succession is dominated by massive lavas, with, in addition, subordinate volcanic rudites (breccias and conglomerates; e.g. Aridia; Fig. 21, log 8). The upper part of the volcanogenic succession is commonly more varied and includes numerous thick (up to tens of metres) intercalations of poorly sorted volcanogenic rudites, volcaniclastic sandstones, occasional shales (typically pink coloured) and silicic tufts (e.g. Tyros, Fig. 21, log 10; Marion, log 9). The succession in the SW Peloponnese (Krokee area) is mainly volcaniclastic (Fig. 21, log 7), although there exposure is limited by low topography; potentially deeper levels of the succession are not exposed. In the NW Peloponnese (e.g. at Zarouhla, Feneos and Kastania; Fig. 21) tuffaceous sediments and volcaniclastic sediments are relatively more abundant, together with common intermediate to silicic composition lava flows. The upper part of the Tyros unit, where exposed, comprises strongly contrasting terrigenous sediments, mainly lithologically homogeneous quartzitic shales and mica-schists. The contact with the underlying volcanogenic unit is typically sheared, but is interpreted here as an deformed normal contact rather than a major tectonic break. The upper terrigenous unit is relatively thick ( > 100 m) in most areas (e.g. Tyros, Marion and NW Peloponnese; Fig. 21, logs 9-11). Elsewhere, the upper terrigenous unit is much thinner (tens of metres at Krokee; Fig. 21, log 7), down to only several metres (e.g. near Aridia and Floka). However, parts of the original succession may have been tectonically removed. Locally, evaporite (gypsum) has been reported from the upper terrigenous unit near the contact with the overlying Gavrovo-Tripolitza platform carbonates (e.g. Krokee and near Verga, Kalamata area; N. Skarpelis, pers. comm). The upper terrigenous unit, or locally the volcanogenic unit (e.g. near Aridia and Floka), is

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS overlain by neritic carbonates, commonly stromatolitic, forming the base of the Mesozoic Gavrovo-Tripolitza carbonate platform succession. In all areas the contact is moderately to strongly sheared, with much evidence of layerparallel extension and other features indicating at least partial detachment from the underlying Tyros unit. However, in some areas (e.g. south of Aridhia; Tyros) facies are transitional indicating that a normal contact was originally present (Lekkas & Papanikolaou 1980). The 'intermediate' nappe is problematic. In the NW Peloponnese a two-part volcanogenicterrigenous succession, as elsewhere, is structurally overlain by the Gavrovo-Tripolitza carbonate platform, providing no basis for the recognition of a separate, higher tectonostratigraphical unit. A large exposure in the SE Peloponnese (south of Aridia) includes folded phyllites, mica-schists and limestone conglomerates with subordinate volcanic intercalations. Local contacts are not well exposed. However, directly east of Molai (Fig. 21, log 8) comparable limestone conglomerates appear to pass depositionally into the Gavrovo-Tripolitza platform carbonates. On the other hand, well-dated metaclastic 'Permo-Carboniferous' sediments with Late Carboniferous sporomorphs structurally underlie the Gavrovo-Tripolitza carbonate platform in the SE Peloponnese (south of Monemvasia; Paraskevopoulou 1951; Fytrolakis 1971; N. Skarpelis, pers. comm), suggesting that the Triassic volcanic rocks in this area could have a continental basement. It, therefore, seems likely that the 'intermediate nappe' is a composite unit including lithologies underlying, laterally equivalent to, and overlying the overall volcanogenic-terrigenous succession.

Interpretation: rift settings The Triassic Tyros unit volcanogenic succession formed in a regionally extensive, subaqueous rift setting. Fragments of a sedimentary basement may be represented by the occurrences of Palaeozoic terrigenous and calcareous sediments (south of Monemvasia; Fytrolakis 1971). The rift basin was partially filled by flood basalt. The presence of abundant volcaniclastic sediments, largely subaqueous debris flows, with little terrigenous material, is suggestive of mass wasting on a subaqueous fault scarps. More fractionated (intermediate-silicic) volcanism predominated in the NW Peloponnese (Feneos-Zarouhla). After volcanism largely ended the inferred rift was covered by terrigenous muds and shallowed, culminating in the accumulation of varied carbonates, organic-rich muds and local evaporites.

137

The presence of volcanogenic horizons interbedded with transitional neritic carbonates (e.g. at Tyros) is suggestive of volcanism during the Carnian. This volcanism can be directly related to the opening of the Pindos ocean basin to the NE (in present coordinates). Where locally intact, the succession passes transitionally upwards into the Gavrovo-Tripolitza platform of Late Triassic-Early Jurassic age. Localized limestone conglomerates beneath the carbonate platform (i.e. west of Molai) are indicative of mass wasting of a nascent carbonate platform, as the pre-existing rift-related accommodation space was filled prior to regional covering by a thick Bahama-type carbonate platform. The Phyllite-Quartzite unit in the Peloponnese is generally similar to counterparts in western and central Crete. The intercalations of quartzitic conglomerates in the Arna unit (Tagetos massif) are lithologically similar to the Mana conglomerate in western Crete, of inferred latest Triassic-earliest Jurassic? age there. The clastic sediments and carbonates exposed in the transition between the Tyros unit and the Tripolitza platform are similar to the Ravdoucha Beds of western Crete, although Late Triassic volcanic rocks are not exposed in the latter unit. Despite differences in metamorphic grade (relatively high grade in eastern Crete, but lower grade in the Peloponnese) the Tyros unit shows some similarities to the Phyllite-Quartzite unit of eastern Crete (Vai-Chemezi areas) and central Crete (upper structural unit). In both the Peloponnese and eastern Crete, intact succession include a thick basaltic-andesitic volcanogenic sequence (lavas and volcaniclastic sediments) that passes upwards into shallow-water carbonates of Late Triassic-earliest Triassic age, correlated with the Tripolitza carbonate platform. In addition, slices of Hercynian granitic gneiss occur locally in both areas. However, conglomerates (with abundant basement-derived material) are much more extensive in eastern Crete than in the southern Peloponnese. Also, the Late Triassic volcanism in the Peloponnese is unknown in Crete (although dating remains limited). Similar volcanic rocks of the Tyros beds locally underlie the most proximal of the Pindos-Olonos nappes (Degnan & Robertson 1998, 2006), confirming that the Late Triassic Tyros volcanic rocks relate to opening of the Pindos ocean. Most of the evidence, outlined above, is consistent with Model 1, as for Crete. However, the presence of subduction-influenced Triassic volcanic rocks could also be consistent with Models 3 and 4, which invoke a southward-dipping subduction zone, although as noted below there is little or no evidence independent of geochemistry

138 that such a south-dipping existed in the Triassic.

A.H.F. ROBERTSON subduction zone

Evidence from the Pindos zone Additional relevant evidence comes from the regionally overlying Pindos zone, which is fragmentary in Crete but better exposed in the Peloponnese (Fig. 5) and in Greece further north. The Gavrovo-Tripolitza platform was overthrust by the relatively unmetamorphosed Pindos unit during Early Cenozoic time (e.g. Bonneau 1984; Jacobshagen 1986; Papanikolaou 1996-1997). Much evidence already exists in the literature, which can be used to test the alternative tectonic models. In Model 1 (divergence-related), the Pindos zone originated as a continental rift in the Triassic (Dercourt et al. 1986) but then developed into a subsiding passive margin as the Pindos ocean opened (Smith et al. 1975; Robertson & Dixon 1984; Robertson et al. 1991). The Pindos thrust sheets restore as an east-facing deep-water slope to abyssal plain (Degnan & Robertson 1998) that probably accumulated on 'transitional' crust within a continental-ocean transition zone (Degnan & Robertson 2006). In Model 2 (convergence-related) the Pindos ocean originated as a Late Triassic back-arc basin related to the later stages of northward subduction of Palaeotethys (Stampfli et al. 2003). This subduction culminated in the collision of a rifted 'Cimmerian' fragment with a Eurasiarelated unit represented by the Pelagonian zone during Late Triassic (Carnian-Norian) time. In principle, any such Cimmerian suturing related to northward subduction need not have affected a related marginal basin to the north, which could have remained isolated. However, Stampfli et al. (2003) specifically argued that a coUision-related compressional event is, indeed observed within the Pelagonian zone further north; this implies that stress was transmitted across the Pindos deep-sea basin from a suture zone to the south to a Pelagonian continent to the north. Is such a compression-related event actually recorded in the Late Triassic sedimentary fill of the Pindos basin? A regional 'Cimmerian' suturing to the south could have resulted in uplift and increased supply of clastic sediment to the basin during latest Triassic-earliest Jurassic time. Also, if the basin was internally deformed, sediment redeposition, slumping, or an intra-basin unconformity might be present: however, none of these features are apparent within the Triassic-Early Cenozoic Pindos succession (Degnan & Robertson 1998). The field evidence instead supports continuing passive margin subsidence of the

Pindos basin from Late Triassic to Early Cenozoic, punctuated by clastic influxes that can be mainly related to the effects of eustatic sea-level change. The Pindos thrust sheets are locally underlain and intercalated with a m61ange including blocks of volcanic rocks, some of which are dated as Triassic from associated sediments. Discrete thrust sheets including Triassic volcanic rocks are also locally present (see Pe-Piper & Piper 2002; Degnan & Robertson 2006). Extensive geochemical studies indicate that the Triassic igneous rocks commonly show a geochemical subduction influence that could be consistent with Models 2, 3, or 4. In addition, some 'enriched' basalts are present that could represent fragments of emplaced seamounts (Degnan & Robertson 2006). An origin related to a northward-dipping subduction zone (Models 2 and 4) is, however, unlikely as no ocean to the south has been identified, as discussed earlier. Pe-Piper & Piper (1998, 2002) have invoked an additional Triassic subduction zone (an intra-oceanic one that dipped southwards) to explain, in particular, localized occurrences of high-magnesian andesites (boninites) and plagiogranites. In this interpretation (Model 4) subduction would have culminated in collision of a trench with a Pelagonian passive margin to the north. This would have been expected to emplace Triassic (and younger) ocean crust (ophiolite) over a Triassic or younger accretionary prism. However, the overriding ophiolites are MidJurassic in age (Liati et al. 2004) and underlying accretionary units document a Triassic rifted margin (e.g. in Evia, Othris and Pindos; Robertson et al. 1991). There is thus no sedimentary or structural evidence for southward Triassic intra-oceanic subduction, as in Model 4. On the other hand, the presence of Mg-andesites locally that imply remelting of previously depleted mantle, clearly needs an explanation.

Evidence from the Pelagonian zone The Pelagonian zone, in turn, structurally overlies the Pindos zone; it is restricted to high-level fragments in Crete and the Peloponnese but is much more intact and widely present in central and northern Greece. Key areas include those NE of Athens, such as in Evia, where the Pelagonian zone has experienced only low-grade metamorphism, in contrast to northern Greece where the grade is higher (Mountrakis 1986). The Pelagonian zone comprises a pre-Triassic continental basement, which includes Upper Palaeozoic granites. Transgressive platform carbonate deposition was punctuated by ophiolite

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS emplacement and deformation during Late Jurassic-Early Cretaceous time (see Rassios & Moores 2006). This was followed by renewed platform deposition until emplacement as part of the Hellenide nappe pile during Early Cenozoic time (Mountrakis 1986). In Model 1 (divergence-related) the Pelagonian zone is interpreted as a microcontinent rifted from Gondwana in the Triassic (Dercourt et al. 1986) related to opening of a Pindos ocean (Robertson et al. 1991). In Model 2 the Pelagonian zone is interpreted as a microcontinent that was rifted from Eurasia related to opening of a Vardar back-arc oceanic basin to the north and a Pindos back-arc basin to the south (De Bono et al. 1998; Vavassis et al. 2000; Stampfli et al. 2001). In Models 3 and 4 the Pelagonian zone is seen as a microcontinent that was detached from Gondwana associated with opening of a backarc marginal basin, over either a south- or a north-dipping slab. Evidence to test the above alternatives mainly comes from the western (Pindos) and eastern (Vardar) margins of the Pelagonian zone. The evidence from the western Pelagonian margin is clearly consistent with Model 1, as there is evidence of Triassic rift-related sedimentation and alkaline volcanism (Mountrakis 1986), as is well exposed in the Othris area (Smith et al. 1975). Available evidence from the eastern margin of the Pelagonian zone (Vardar margin) also points to the existence of a Triassic rifted margin (see Sharp & Robertson 2006). Triassic basalts in the Vardar zone (i.e. within the Eastern Almopias zone) lack geochemical evidence of a subduction influence, opposing Models 2 and 3, in which the adjacent Vardar basin is seen as an abovesubduction zone back-arc rift or oceanic basin. In presenting evidence that, if valid, would support Model 2, Stampfli et al. (2003) argued that the Pelagonian zone experienced a pulse of regional 'Cimmerian' compression related to suturing of Palaeotethys to the south in latest Triassic time. Stress from this collision was transmitted across the Pindos, inferred back-arc marginal basin, triggering a stratigraphic inversion event within the Pelagonian zone during latest Triassic time. Stampfli et al. specifically argued that a Permian-Triassic rift succession exposed on the island of Evia (Fig. 5), termed the Liri unit, experienced compression-related uplift, associated with mass-wasting of 'olistostromes', and that this was then unconformably overlain by a Jurassic carbonate platform (Stropones Limestone). During the present work, this interpretation was tested in the field and it was found that the evidence is instead consistent with the extension-related Model 1.

139

The Liri unit is divisible into southerly and northerly exposures, separated by an inaccessible mountainous area (Fig. 22a). The Liri Unit has experienced greenschist-facies metamorphism and extreme layer-parallel extension, with the development of ubiquitous 'phacoidal' fabrics. Sedimentary structures (e.g. grading) are relatively well preserved, especially in the higher stratigraphic levels, and show that the sequence is mainly the right way up. The Jurassic shallowwater Stropones Limestone ('cover unit') is, in fact, located structurally beneath rather than above the Liri Unit; consequently, the Liri unit lacks any preserved overlying depositional cover in this area (Fig. 22b and c). The Liri unit is instead structurally overlain, above a major lowangle thrust contact, by a regionally extensive Pelagonian thrust sheet. The local Pelagonian sequence, of Late Permian? to Mid-Triassic age, includes metasiliciclastic sandstones, shale, ribbon chert, redeposited carbonates (including debris flows), andesitic-rhyolitic metavolcanic rocks and tuffaceous-volcaniclastic sediments, consistent with a rift-related origin. The succession passes upwards into a several-kilometrethick unit of platform carbonates of Late Triassic-Jurassic age, typical of the Pelagonian zone generally (Fig. 22d). This overall succession is stratigraphically underlain by schists and granitic rocks ('Hercynian basement') and coarse 'basal' clastic sediments derived from these lithologies (Fig. 22b). Petrographic study (19 samples) shows that the meta-sandstones of the lithic unit are mainly arkoses and lithic arkoses, mainly derived from granitic and metasedimentary lithologies, as widely exposed within the 'Hercynian' basement beneath the Jurassic platform carbonates throughout the Pelagonian zone (Mountrakis 1986). The Liri unit was mainly deposited by turbidity currents and mass-flow processes that were active during an inferred Permo-Triassic rift setting. However, there are few indications of water depths, which could have been relatively shallow (tens to several hundred metres). Radiolarian cherts or other evidence of pelagic deposition are absent. Some localized 'cherts' represent secondary alteration, of possibly hydrothermal origin. The uppermost part of the Liri unit, mostly < 10 m below the overriding Pelagonian thrust sheet, includes scattered small outcrops of highly fossiliferous shallow-water carbonate (Fig. 22b). This limestone is well dated as Late Carboniferous-Late Permian based on shallowwater calcareous fossils (e.g. benthic foraminifera) (Stampfli et al. 2003). These limestones apparently represent fragments of a long-lived

140

A . H . F . ROBERTSON

IA

Loc B

L ~ . ~ . k ~

r elagonian

lOnl 9 . . , 9 kiri : . "

/." ..........................

b

.~troponlask ' " " ""

9

Ay Georghios area. " "

:ZE5

W Pe~'~'g~

C

A

.,-,.Liri--.Unit..,..,

Stropones Limestone

0 ,

Km

~

Z

1 ,

A'

Pelagonianthrust U. Triassic~1 z U. Jurassic t- J t Pelagonian "~ T-Es carbonate platform 2 -~-:-4 M.- U. Trias.t- _ volcanic._~ v..v sedimentary unit er o " U. Permian" o o M. Triassic "~ i ~ ~ clastic rocks i ' l - " ~ i U. Carboniferous z. ~granite !i L Permian.

.

9 d

e

~

Upper Triassic Liri Unit

~'---~ '

"

P'-ermoCarboniferous sheet limestone blocks/ dismembered 9 thrust sheet

~

granitic blocks

i/

Zygos

.L..--.%;.~ 9 " """ ~

F

I l" 9 I 9 ~" 30 Olympos - - j . -- Mtns. - - / . I

// 1

siliceous black shale Arkosic & . lilhic " sandstone

[ "50 m

d

Ioo o , , ~

9

9

<__.

9 e.g.

~

12 lm L. Jurassic Stropones Limestone

I ~

~:]

f~

~18

25~. ~'3 " -/'30"

.~,

"~- 9 ~ " i~ _t

Seta

k Tharounla . . .o 9 ' ~) ~ e '~

25 (~

ni _---r-

"

/'40" 30 . ,~ Ay. Georghios B ~35 I- "~

/

-

Alluvium Pelagonian sequence (undiff.) Pelagonian metamorphic basement Mistros Unit

0~ 9 2 Km

/\ ""

N

Liri Unit Stropones Limestone

......

Fig. 22. Upper Palaeozoic-Lower Mesozoic sedimentary and volcanic units exposed on the Island of Evia. (a) Regional location; (b) Cross-section at location B (see d); (e) Cross-section at A-A; (d) Outline geological map; (e) Tectonic-stratigraphy in this area (left) and specific sedimentary log of the Liri Unit (right). ( See text for explanation and data sources.) carbonate platform that was possibly constructed on Hercynian basement within the Pelagonian zone. Individual blocks are typically less than a metre to several tens of metres in size. Small outcrops of distinctive dark grey shallow-water limestone can be traced for tens to hundreds of metres along strike, suggesting the existence of

dismembered sheets, in addition to detached blocks. Where smaller blocks (several metres across) are seen lower down the sides of valleys these are commonly landslipped. Measurements of bedding dip within the limestone blocks and sheets show that the bedding is everywhere moderately inclined

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS (Fig. 22d), subparallel to the tectonic contact with the overlying Pelagonian thrust sheet. Where the contact with the shale-sandstone matrix is rarely exposed this is seen to be a sharp tectonic contact. The margins of the limestone blocks are commonly brecciated and calcite veined. The adjacent matrix sediments do not contain sedimentary fragments of the same limestone (although some entrained phacoida! fragments are locally present). Petrographic study of the sandstone matrix enclosing the limestone blocks did not reveal sedimentary limestone clasts, but rather the composition remains unchanged from the underlying sandstones.

Interpretation: rift basin deformed in Early Cenozoic time In Model 1 the exotic blocks and sheets were shed into a rift basin related to rifting (i.e. rift shoulder uplift) that immediately preceded break-up to form the Pindos ocean in Late Triassic time (Fig. 23b). In Model 2 (convergent margin-collision) the upper part of the Liri unit is viewed as an 'olistostrome' containing blocks shed into a backarc basin, an event that was triggered by stratigraphic inversion related to 'Eo-Cimmerian' orogeny in latest Triassic time (Fig. 23a). However, as noted above there is no evidence of a critical Jurassic sealing unconformity, which is essential to validate this interpretation. Without

Pelagoni.an nappe

this, there is no evidence of pre-Jurassic deformation and instead the deformation is likely to be Cenozoic, associated with the SW-directed regional emplacement of Pelagonian thrust sheets. It is also possible that neither of the above alternatives is correct and that the 'olistostrome' instead relates to Cenozoic overthrusting. In this interpretation (Fig. 23c) the limestones could represent the dismembered remnants of a thrust sheet of neritic limestone of Pelagonian affinity that was entrained, together with the overriding Pelagonian thrust sheet, during regional Early Cenozoic deformation. Several observations are consistent with this interpretation: first, most of the blocks occur in highly sheared shales just beneath the overlying Pelagonian thrust sheet; second, many blocks join up as larger dismembered sheets at the same structural level; third, the dips are parallel to the overlying Pelagonian thrust sheet and are not variable as expected for a debris flow (olistostrome) origin; finally, there is an absence of associated limestone-derived debris flows and other gravity-flow deposits within the Liri unit, in contrast to typical large-scale debris-flow deposits (olistostromes). In summary, it is concluded that there is no evidence for latest Triassic 'Cimmerian' compression within the Pelagonian zone, as implied by Model 2.

b latest Triassic

a Late Triassic ..............

Detached blocks

Triassic siliciclastic sediments Mid- Late Carboniferous - Mid Permian neritic sediments Hercynian intrusive igneous rocks -Schists

r~

141

Hercynian metamorphic rocks Oceanic crust

Fig. 23. Alternative tectonic setting for the Triassic of central Evia (Pelagonian zone). (a) Compression (stratigraphic inversion interpretation) (Stampfli et al. 2003); (b) rift-related interpretation; (c) formation by layer-parallel extension (boudinage) of a thrust sheet beneath the Pelagonian nappe during the Cenozoic. A thrust-related interpretation is favoured, as discussed in the text.

142

A.H.F. ROBERTSON

Possible objections to rift-related settings Although the rift-related Model 1 explains most features of the geological evolution, there are several potential arguments against this interpretation, which are discussed below. However, each of these can be countered, as will be seen.

(1) The Mid-Carboniferous-Early Triassic basins, as documented in Sicily and western Crete, were sufficiently deep (c. 500 m or more) (Kozur 1993, 1995) to require the existence o f a contemporaneous ocean basin even if no oceanic crust is preserved There is, however, no requirement for an oceanic basin to have existed adjacent to Gondwana during Late Palaeozoic time, as in Models 2 and 4. Similar broad, deep basins existed widely around the margins of the Atlantic prior to spreading. These include the Jean d'Arc basin off Newfoundland (Reid & Keen 1990), the Hatton Bank and adjacent basins off Ireland (Fowler et al. 1989), the Lusitanian basin off Portugal (Wilson 1988), and comparable rift basins bordering the Central and South Atlantic and Indian oceans, and the Red Sea (Purser & Bosence 1998). There are also numerous examples of deep-water rifts marginal to now-sutured oceans that are exposed on land, notably around the Western Alps (Lemoine et al. 1986).

(2) No viable mechanism for the Triassic rifting o f continental fragments exists other than back-arc extension If true, this would favour Models 2, 3 and 4 over Model 1. However, similar rift settings are known from non-emplaced passive margins, notably the Exmouth Plateau off the NW Australia margin (Von Rad et al. 1992). Also, similarly rifted fragments appear be embedded in accretionary margins of Indonesia (Pigram & Pannabean 1984). A plausible mechanism might involve calving of weak marginal rift units, up to several hundred kilometres in size from a parent continent. The driving force could be slab-pull related to regional subduction, in this case northward subduction under the Eurasian margin during Late Palaeozoic-Early Mesozoic time. Although slab-pull by itself might be insufficient to initiate continental break-up (Smith 1999), it is possible that break-up could result from multiple rift events, especially once a rift was weakened by rift magmatism (Buck 1993). In the south Aegean there is indeed a history of pulsed rifting starting in Mid-Late Carboniferous time, with an

extensional pulse associated with magmatism in the Early Triassic, and final break-up to form the Pindos ocean in the Late Triassic.

(3) The change from deep-water to shallowwater deposition during the late Early-Mid Triassic, as documented within the PhylliteQuartzite unit o f western Crete and eastern Crete implies uplift o f > 500 m and so favours a convergence-related, foreland basin or collisional setting Undeformed rifts worldwide, including the Red Sea (Purser & Bosence 1998), the Gulf of Aden (Robertson & Bamakhalif 2001), the Avalon margin (e.g. Avalon platform; Tuckolke et al. 2004), the Indian ocean (e.g. off East AfricaMadagascar; Hankel 1994), and many other examples are known to have undergone hundreds of metres (to several kilometres) of marginal uplift related to extension, prior to the onset of sea-floor spreading. There are several possible mechanisms for such uplift. First, a model of inhomogeneous crustal stretching with depth predicts flank uplift of 1-2 km (e.g. Braun & Beaumont 1989; Steckler & Omar 1994), although this would be hard to test using field geological evidence. Second, a thermal pulse could cause regional uplift. A plume influence related to Triassic rifting has been suggested for the Balkan region (Dixon & Robertson 1999). The presence of ocean island basalt (OIB)-type basalts in many areas (e.g. western Crete) could reflect a plume influence but could alternatively be explained by low-degree melting of potentially inhomogeneous subcrustal mantle. As yet, there is no definite evidence of a plume-related setting in the south Aegean region. Third, a pre-existing rift basin could be flexurally uplifted related to a pulse of extension that was focused elsewhere in the rift zone. Such a change in the locus of rifting could cause a change in the dip of the related extensional faults such that the pre-existing rift footwall was transferred to the hanging wall of the subsequent rift. Such an effect alone would be capable of explaining the relatively rapid change from relatively deep-sea ( > 500 m) to neritic depositional conditions, as observed in the south Aegean region.

(4) The geochemical evidence o f Triassic igneous rocks requires coeval subduction in the south Aegean region The Triassic volcanic rocks of eastern Crete, the Peloponnese, many other parts of Greece

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS and also N W Turkey exhibit a chemical signature that is widely believed to require a subduction setting, at least locally (Pe-Piper & Piper 2002). Key points are the presence of Triassic arc-type granites (e.g. Cyclades, northern Menderes, eastern Crete; see Romano et al. 2006), the local occurrences of shoshonitic and high-K andesites (Lakmon Mtns.) and high-K andesitic intrusions (i.e. Kokkino, SW Peloponnese), the rare occurrence of boninitic-type lavas (Othris and Edipsos), and the presence of pyroclastic rocks (implying a high volatile content). Following an extended discussion, Pe-Piper & Piper (2002) concluded that the chemistry of some of the Triassic igneous rocks requires the involvement of subduction-derived fluid in the melt process and 'that subduction may be either Hercynian or of Triassic age' (p. 103). An inherited subduction influence, presumably related to Hercynian subduction in the south Aegean region, was previously proposed by various workers (Robertson & Dixon 1984; Dixon & Robertson 1993, 1999; Capedri et al. 1997; Pe-Piper & Piper 1998). Implicitly, Pe-Piper & Piper (2002) have acknowledged that these two alternatives, a coeval Triassic versus a Hercynian inherited subduction signature, cannot be resolved by geochemical evidence alone. Thus, the decisive factors can only be the geological evidence for subduction zones of the requisite age and location. No independent evidence for such Triassic subduction zones was found in the south Aegean region during this study and, therefore, the model of subduction zone inheritance is preferred. In keeping with this, volcanic rocks within units that restore further south (e.g. Phyllite-Quartzite unit, western Crete) are enriched in incompatible elements (with no subduction influence), similar to modern rift basalts (e.g. Fitton et al. 1998). By contrast, volcanic rocks extruded through Hercynian basement units, generally located further north, are relatively depleted in incompatible elements (e.g. Nb), possibly reflecting the extraction of a lithosphere-hosted subduction component of probable Hercynian age. It was similarly suggested that the presence of radiometrically dated Late Triassic calc-alkaline granitic rocks (orthogneiss) in the Vai area, eastern Crete, implies a convergent margin (subduction) setting during the Triassic, possibly related to southward subduction (Romano et al. 2006; Model 3). At least some of the granitic rocks in this area crystallized, then were exhumed and eroded throughout Mid-Late Triassic time, as similar granitic rocks are found as clasts within associated coarse clastic sediments of this age. The small Triassic granitic bodies might relate to melting in an extensional setting, followed by rapid exhumation, as inferred, for example,

143

for the Oligocene granites of northern Greece (Kolokotroni & Dixon 1991). (5) There is no evidence o f a Triassic subduction zone & northern Greece and thus the convergence o f Africa and Eurasia must be accommodated in the south Aegean However, several studies of units associated with the southern margin of Eurasia, including the Pontides (e.g. Usta6mer & Robertson 1997; Okay 2000), the Caucasus (e.g. Adamia et al. 1995) and the south margin of Eurasia generally (Nikishin et al. 2001) have concluded that a subduction zone dipped northwards beneath Eurasia and was active especially during Carboniferous to Mid-Jurassic time (Nikishin et al. 2001; Usta6mer et al. 2005; Kazmin & Tikhonova 2006). Palaeotethyan units have also been identified in former Yugoslavia (see Karamata 2006). Some workers have suggested that a Palaeotethyan suture is located in the Vardar zone of northern Greece (Robertson & Dixon 1984; Mountrakis 1986). However, the assumption that the Serbo-Macedonian and Rhodope zones formed part of the southern margin of Eurasia by Early Mesozoic time is now questioned by radiometric dating and structural studies that suggest that independent terranes existed until docking with Eurasia during Alpine (Jurassic) deformation (Himmerkus et al. 2006). A Palaeotethyan suture may thus be buried within northern Greece, removing the need for a more southerly located Palaeotethyan subduction zone. Such a subduction zone would have extended eastwards into the Pontides and westwards into former Yugoslavia. (6) Regional plate reconstructions favour the existence o f a Palaeotethyan ocean & the south Aegean region Garkunkel (2004) favoured a reconstruction akin to the convergence-related Model 2. He argued that between Late Permian and Late Triassic time, the regional palaeogeography evolved from a Pangaea A to a Pangaea A-2 type setting (see Smith et al. 1981; Smith 2006; Fig. 24). If correct, this would imply convergence along the southern margin of Eurasia of several hundred kilometres during this time. This reconstruction assumes c. 500 km of right-lateral motion between Africa and Eurasia, and that this was translated into clockwise tightening of a Palaeotethyan gulf in the east (comparable with the setting of the modern Gulf of Makran). Ziegler & Stampfli (2001) argued that up to 400 km of right-lateral displacement did indeed take place, but placed

144

A.H.F. ROBERTSON

9 "." .." ' .....~.Palaeotethys

'

'Pe Gr'~. E

9 .

Palaeotethys

Ta

Tauride

Po

Pontides

Pe

Pelagonian

L

Levant

I

Italy

Gr

Greece

Cy

Cyrenaica

Ap

Apulia

( 9

"

.

.

9

.

.

9 . Cy.X

.',xk. I. _i.

Cy

a

a

".

.\

L "".-..

9

f i,

9

~._..L.,.,,,.. ~7).

"

9

,

" ".//t'. " . "V

l/ (/..,' " ",\'.

,z cy/% 9 " (:. .

.I

I

~ ~ ~.

(,/,

c Pangea A-like

"

,

'

,~'

E

"-

AA Austro-Alpine

~

.

.

SI

.

t

PEL Peloponnese Palaeotethys E

Gr

..,.',.,

'

I

"

.

9

"

/

pQ PhylliteQuartzite

o~

"

' ~ ' ' ~ '

-

'~.

'~''~

,,~

d Pangea A2-1ike

Evia

CR Crete

. 1;I

":'~

'~..,,,.

Sicily

.

I

"

-

.

9 9

IIIAP~ 9

cR

"

)

~" ' { " i" f'~'l '" '

~ to?-~: ,-';..

\ "

I

9 , , /..~

'

/~ !

.

9

~

.

" ,.,~..

P . . [ alaeotethys

" I~ApG~.. ~,~.

9 f j

9 9

.~'-L"

~

I

AA

>.

."~.

'f 9 . . / 9

AA

""

,

" ~ "" ' ' "

Fig. 24. Alternative models for the Late Palaeozoic-Early Mesozoic tectonic evolution of the south Aegean region. (a) A single spreading axis propagated from the wider Tethys to the east (Ricou 1996); (b) northward subduction zone and continental rifting (Stampfli et al. 2001); (c) extensional setting, assuming a Pangaea A-like reconstruction (Garfunkel 2004); (d) convergent setting assuming a Pangaea A2-1ikereconstruction (Garfunkel 2004). AA, Austro-Alpine; Ap, Apulia; CR, Crete; Cy, Cyreanaica; E, Evia; Gr, Greece; I, Italy; L, Levant; Pe, Pelagonian; PEL, Peloponnese; Po, Pindos; PQ, Phyllite-Quartzite; SI, Sicily;Ta, Tauride. this within the latest Carboniferous-Early Permian time interval, associated with transtensional collapse of the Hercynian orogen and the development of related pull-apart basins (see, e.g. Scotese & Langford 1995). Regional magmatism (granitic intrusion and calc-alkaline extrusion) climaxed in Early Permian time. As a result, the preferred Pangaea A-2 type setting could have been established prior to Triassic time. A several hundred kilometre wrench offset might have been dissipated along several lineaments within the Hercynian orogen rather then being directly translated into regional clockwise rotation of the Eurasian margin in the Central and Western Mediterranean. Garfunkel (2004) considered an alternative, extension-related setting, which could be valid for the Triassic assuming Pangaea

reorganization largely occurred in pre-Triassic time, as favoured here (Fig. 24).

Regional tectonic development One of the remaining problems is the relationship between Hercynian deformation and metamorphism, as documented by the fragmentary high-grade units scattered around the south Aegean region, and the Triassic rift setting (see Romano e t al. 2006) 9How did the North African passive margin to the present south escape this deformation and metamorphism? By contrast, Hercynian compressional deformation affected the North African margin west of Tunisia (Guiraud e t al. 2001). Also, what was the nature of the contact between these metamorphosed

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS (e.g. Crete) and unmetamorphosed (e.g. North Africa) domains? In Model 1 (divergent setting) the Talea OriPlattenkalk unit in Crete and the Southern Peloponnese formed the distal edge of the North African margin during pre-Triassic time. Assuming that the Hercynian-age detritus within the Talea Ori-Plattenkalk unit (Brix et al. 2002) was locally derived, these sedimentary units are likely have been deposited on Hercynian basement, which was detached during Early Cenozoic subduction and is thus mainly not now exposed. In this model it is possible that the deformed and metamorphosed northern edge of the Hercynian orogeny, located along the North Gondwana margin, was later rifted to open the Neotethyan ocean basin to the north (Triassic or younger), stranding it entirely to the north. Models 2 and 4 (convergent settings) are problematic as no evidence of a northwarddipping Palaeotethyan subduction zone has been identified in the south Aegean region, ruling out any juxtaposition of metamorphosed and unmetamorphosed units as a result of subductioncollision (pre-Jurassic). In Model 2 the Talea Ori-Plattenkalk unit rifted from North Africa as the Cimmerian continent, well south of areas affected by Hercynian orogenesis, yet contains Hercynian-age detritus. To explain this, Champod et al. (2004) suggested that Hercynianaged detrital zircons were transported hundreds of kilometres southwestwards through a continental rift system from the well-established Hercynian orogen in the central-west Mediterranean region. However, a local provenance from exposures of high-grade 'Hercynian' basement is more consistent with the sedimentological evidence outlined earlier in the paper. In Model 3 (southward subduction), the Hercynian-age granitic rocks (e.g. Pelagonian and South Aegean) relate to southward subduction of Palaeotethys (e.g. ~eng6r 1984; Romano et al. 2006; Xypolias et al. 2006). Units affected by Hercynian metamorphism (e.g. Chemezi and Kythira) were close to the trench in the north relative to the North African continent further south. The deformation and metamorphism could then simply have tailed off southwards, with the original transition now being hidden beneath the Sea of Crete. In this interpretation the Carboniferous radiometric ages of the Mersini basement complex, eastern Crete, and associated structural evidence (Romano et al. 2006) are consistent with southward 'Hercynian' subduction. However, in this interpretation the Permian and Triassic ages from other crystalline units in the area (Romano et al. 2006) are surprising, as it is generally believed that Hercynian

145

orogeny had given way to extension-controlled exhumation by the Late Carboniferous (Ziegler 1988; Ziegler & Stampfl 2001). Also, the Triassic rift-related basaltic rocks of western Crete do not exhibit a subduction influence or contain arc-derived detritus, as would be expected if they represented back-arc marginal basins above a coeval southward-dipping subduction zone. There is evidence of Carboniferous subduction-related magmatism further east, in Turkey, but only in the north (e.g. in the N W Pontides; Usta6mer et al. 2005), which implies the existence of northward subduction beneath Eurasia. If southward subduction in the south Aegean region is also accepted, this would require the existence of two subduction zones, one dipping northwards beneath Eurasia and the other dipping southwards beneath Gondwana, both active during Late Palaeozoic time. However, in Turkey there is as yet no convincing evidence of Late Palaeozoic southward subduction; for eample, along the northern margin of the Tauride-Anatolide platform, where passive margin conditions persisted (Robertson et al. 2004). This suggests that any south-directed subduction would have mainly affected areas in the west, in the central and western Mediterranean regions and elsewhere in southern Europe but not Turkey further east. Devonian-Carboniferous southward subduction as well as northward subduction have indeed been inferred for the Hercynian basement in central and western Europe (e.g. Eastern Alps), giving rise to a doubly vergent orogen (Neubauer & Handler 1999). The apparent absence of Hercynian deformation and metamorphism within both North Africa and Gondwana-derived units (e.g. Taurides and Anatolides) raises the possibility that the Hercynian units of the south Aegean region might represent exotic terranes that were emplaced from the central Mediterranean region by right-lateral strike-slip (Dornsiepen et al. 2001). In this scenario, the fragmentary 'highgrade' metamorphic units of the south Aegean region (e.g. eastern Crete) formed in response to collisional suturing of Palaeotethys, some way westwards of their present position during Carboniferous time (Fig. 25). Open-ocean conditions (i.e. Palaeotethys) persisted further east, from western Turkey eastwards. Palaeotethyan exotic terranes were displaced eastwards in response to tectonic escape from the Hercynian suture zone to an open ocean to the east, during or soon after diachronous closure of Palaeotethys further west (SW and central Europe; Alps) during Late Devonian-Early Carboniferous time. This process would be comparable with the westward

146

A.H.F. ROBERTSON

ELi . 9 "

PT

" PO

0~

...

a Early Carboniferous

L a t e Triassic

b Late Carboniferous-Late Permian ECR EU GO H K KU L PE PI PT SMZ RH SI

East Crete

TA

Tauddes

(Phyllite-Quartzite unit)

Eurasia

Gondwana Harcynian orogen Kir~ehirMassif K~re backarc basin Levant

PelagonianZone Pindos ocean Palaeotethys

Sarbo-Macedonian Rhodope Sicily

~

N

VO Vardar ocean WCR West Crete ~PhylUte-Quartziteunit)

Fig. 25. Proposed tectonic evolution of the Upper Palaeozoic-Lower Mesozoic units of the south Aegean region. (a) Diachronous closure of Palaeotethys. (b) Syn-post-collisional right-lateral wrench faulting displaces exotic Hercynian terranes into the south Aegean region. Deep-water sediments accumulate in transtensional basins open to Palaeotethys to the east. (c) The south Aegean margin undergoes continental break-up to form the Pindos ocean and counterparts in the easternmost Mediterranean region.

tectonic escape of Anatolia after the Miocene. Such dextral displacement might have occurred at any time during Late Carboniferous-Early Triassic time, associated with reorganization to a Pangaea A-2 type assembly. Depending on the timing of any such displacement, the Late Palaeozoic deep-water basins of the south Aegean could have been strike-slip controlled. More evidence is needed to discriminate between the above alternatives, but the tectonic escape interpretation is promising. Following the Hercynian orogeny, the Permian-Triassic deep-sea sediments of western Sicily and eastern Crete accumulated in a broad, relatively deep, possibly transtensional rift basin. In Sicily, terrigenous turbidites sourced in the exhumed Hercynian orogen were deposited in deep water, followed by open-marine radiolarian sediments and pelagic carbonates. Condensed pelagic carbonates accumulated on intra-basin

highs and carbonate platforms developed around the periphery of the basin. Further east, in Crete, the Late Palaeozoic deep-sea siliciclastic sediments of the Phyllite-Quartzite unit and the shallow-marine siliciclastic sediments of the Talea Ori unit were deposited within, and along, the margins of a broad deep-water rift basin, fed from the North Africa craton and possibly from the detached northern margin of this basin. The coeval deep-sea sediments of eastern Crete formed in a relatively distal part of the rift, isolated from coarse terrigenous input. During this time the south Aegean region remained open to Palaeotethys further east. The Triassic Pindos rift basin in Greece extended northwestwards through the Budva zone of former Yugoslavia to connect with the Lagonegro zone in southern Italy. The rifted pelagic basin in the Lagonegro zone dates from the Mid-Triassic, but rifting apparently

TESTING ALTERNATIVE SOUTH MEDITERRANEAN TECTONIC MODELS

Carbonates

I'~

Volcanicrocks

~

Continentalcrust

Siliciclasticdeposits

~

SiLls

~

Oceaniccrust

Conglomerates

~

Evaporites

~ ~

Normal/reversefaults

S

/,','

~

,'T'

§

~ ~

T

~

147

'

+ / /+

I

I

,'

'§

,

. +

,','

" +

,','

,'~",-

*\ .+

+

+

. ~ \+

\ 9

§ x

+

Po

\ §

§ \"

~.+

? d Mid-Jurassic MC

~~i7

Y'

+

+

+

+

+l

C Late Triassic-earliest Jurassic TO -k

~

~

+

r + \~,+~,+:+,

+

-t-

+ ~,/

-4-

+ -k

+

-I-

+

-I-

-t-

.t.

+ -I-

-I-

b Early Triassic WP-Q -

-k

-

~

I

.-

-

~

~

EP-Q

~

_

_

,-,~ .

_

-I-

_

.

_

.

.-

_

-

~

-F

~

-I-

~

+ -I-

a Late C a r b o n i f e r o u s - P e r mi a n

9

EP-Q

E Crete P h y l l i t e - Q u a r t z i t e

PO

Pindos ocean

I

Ionian Basin

PR

Pindos rift

MC

Mana C o n g l o m e r a t e

T

Tripolitza p l a t f o r m

NA

North Africa

TO

Talea Ori

PZ

Pelagonian Z o n e

TR

Tripali

PK

Plattenkalk

WP-Q

4-

-k

H-

+

, ~ 10 km

West Crete P h y l l i t e - Q u a r t z i t e

Fig. 26. Inferred tectonic evolution of the Upper Palaeozoic-Lower Mesozoic units of Crete and the southern Peloponnese. Rifting along the north margin of Gondwana gave rise to a relatively wide and deep rift basin. During the Mid-Triassic, extensional faulting of this margin resulted in flexural uplift, whereas the basins to the north and south rapidly subsided culminating in opening of the Pindos ocean to the north. The former rift zone was capped by the Gavrovo-Tripolitza carbonate platform during Mesozoic passive subsidence.

148

A.H.F. ROBERTSON

commenced in the Late Permian, based partly on the evidence of reworked neritic fossils (see Ziegler & Stampfli 2001 for review). However, there is little evidence for the existence of a preexisting, Late Palaeozoic deep-water basin in the Lagonegro zone similar to the Sicanian basin. This, in turn, implies that the Pindos ocean did not simply widen the pre-existing Late Palaeozoic rift, but instead created a new basin, which reactivated an older rift in the east (e.g. Crete) but left it abandoned in the west (Sicanian basin). The Late Palaeozoic deep-water rift basin in the south Aegean region was reactivated in the Triassic as a precursor to opening of the Pindos ocean. A pulse of rifting, probably focused along the Pindos rift to the NE, resulted in flexural uplift of part of the pre-existing rift basin (i.e. riftshoulder uplift) in the south Aegean region (Fig. 26). By contrast, the Permo-Triassic rift basin further west, in Sicily (Sicanian basin), was abandoned and gently subsided until Mid-Jurassic time when it was reactivated related to opening of the central North Atlantic. Extension, however, reached as far west as this area and resulted in episodic destabilization of bordering carbonate platforms and localized Triassic volcanism. After spreading of the Pindos ocean began in Late Triassic time, passive margin subsidence was accommodated by the growth of large carbonate platforms bordering the Pindos ocean and the abandoned Permian rift basin in the Sicily area. The platforms were constructed right across the former rift basins represented by the PhylliteQuartzite unit after their Mid-Triassic flexural emergence and some erosion (e.g. to form the M a n a conglomerate; Fig. 26). During Late Triassic time, Crete, the Peloponnese and south Aegean as a whole experienced passive margin subsidence, building up the kilometres thick Gavrovo-Tripolitza carbonate platform and its passive margin units, including the Talea Ori and Tripali units. The platform was detached when the basement was subducted during the Early Cenozoic, followed by exhumation, whereas the platform cover and its local substratum (Tyros and Ravdoucha units) were accreted to the overriding plate. In Crete, the Triassic carbonate platform represented by the Talea Ori and Tripali units rifted and foundered, followed by deposition of the pelagic Plattenkalk, a counterpart of the deep-water Ionian zone in western Greece and Albania. In western Sicily, the Sicanian basin was reactivated during Mid-Jurassic time, related to opening of the Central Atlantic. A spreading centre possibly migrated eastwards to open the oceanic Ionian basin during Late Jurassic time (see Catalano et al. 2001), possibly even extending eastwards to open or widen the southernmost

Neotethyan oceanic basin between Crete and North Africa. During Cenozoic subduction in the south Aegean region the Tripolitza platform was detached from its pre-Jurassic rift-related substratum that was subducted, accreted to the overriding plate, and then was exhumed as the H P - L T units of the lower thrust sheets (Phyllite-Quartzite, Talea Ori-Plattenkalk and Tripali units).

Conclusion Of the alternative models for the Late Palaeozoic-Early Mesozoic setting of the south Aegean region, a pulsed rift model best fits the evidence, based on new field-based observations in western Sicily, Crete, the Peloponnnese and Evia combined with a review of the literature (Figs 2 and 26). A deep-water rift opened along the northern margin of Gondwana during the Mid-Late Carboniferous, followed by a further pulse of rifting in the Early Triassic, preparatory to opening of the Pindos ocean to the NE (present coordinates) during the Late Triassic. Mid-Triassic uplift and erosion in Crete is explained by upward flexure of the preceding Late Palaeozoic rift zone, related to renewed rifting to form the Pindos ocean in the south Aegean region. In the absence of evidence for contemporaneous Triassic subduction, it is inferred that the observed subduction signature in many the Triassic rift-related basalts (e.g. eastern Crete, Peloponnese) relates to melting of heterogeneous subcrustal mantle. The subduction fluids were probably introduced during Hercynian orogenesis. Our present understanding of the tectonic development of the south Aegean region owes much to the detailed biostratigraphical studies of J. Krahl in Crete. I would like to thank him for information on the literature and relevant outcrops in Crete. Thanks are also due to H. Kozur for written and verbal discussions during this work. I am grateful to R. Catalano for a helpful field introduction to the geology of western Sicily, and to N. Skarpelis for a similar introduction to the SW Peloponnese. J. Dixon is thanked for continuing helpful discussion. G. Karner provided useful insights into modern rifted margins. The manuscript benefited from comments by J. Dixon, P. Degnan and D. Mountrakis.

References ADAMIA,S., BAYRAKTUTAN,S. • LORDKIPANIDZE,M. 1995. Structural correlations and Phanerozoic evolution of the Caucasus-Eastern Pontides. In: EELER, A., ERCAN, T., BINGOL, E. & ORCEN, E. (eds) Geology of the Black Sea Region. Mineral Research and Exploration Institute of Turkey (MTA) Publications, 69-75.

TESTING ALTERNATIVE SOUTH M E D I T E R R A N E A N TECTONIC MODELS ALEXOPOULOS, A. & LEKKAS, S. 1999. The tectonic structure of Tainaro (Mani) Peninsula (Southern Peloponnese, Greece). Neues Jahrbuch fiir Geologie und Paldontologie Monatshefte, 11,698-704, AL-RIYAMI, K. & ROBERTSON,A. H. F. 2002. Mesozoic sedimentary and magmatic evolution of the Arabian continental margin, northern Syria: evidence from the Baer-Bassit M61ange. Geological Magazine, 139, 395-420. BASSIAS, Y. & TRIBOULET, C. 1994. Tectonometamorphic evolution of blueschist formations in the Peloponnese (Parnon and Taygetos Massifs, Greece): a model for nappe stacking during Cenozoic orogenesis. Journal of Geology, 102, 697-708. BERNOULLI, D. & JENKYNS,H. C. 1974. Alpine, Mediterranean and Central Atlantic Mesozoic facies in relation to the early evolution of Tethys. In: DOTT, R. H. & SHAVER, R. H. (eds) Modern and Ancient Geosynclinal Sedimentation. Society of Economic Mineralogists and Paleontologists, Special Publications, 19, 129-160. BIANCHINI,G., CLOCHIATTI,R., COLTORTI, M., JORON, J. L. & VACCARO, C. 1998. Petrogenesis of mafic lavas from the northernmost sector of the Iblean district (Sicily). European Journal of Mineralogy, 10, 301-315. BOJAR, A.-V., FRITZ, H., KARGL, S. & UNZOG, W. 2002. Phanerozoic tectonothermal history of the Arabian-Nubian shield in the Eastern Desert of Egypt: evidence from fission track and paleostress data. Journal of African Earth Sciences, 34, 191-202. BONNEAU, M. 1984. Correlation of the Hellenides nappes in the south-east Aegean and their tectonic reconstruction. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds). The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 515-527. BRAUER, R., ITTNER, R. & KOWALCZYK, G. 1980. Ergebnisse aus der 'Phyllit-Serie' SE-Lakoniens (Peloponnes, Griechenland). Neues Jahrbuch fiir Geologie und Paliiontologie Monatshefte, 1980, 129-132. BRAUN, J. & BEAUMONT, C. 1989. A physical explanation of the relation between flank uplift and the breakup of continental margins. Geology, 17, 760-764. BRITISH PETROLEUM COMPANY Ltd. 1971. The Geological Results of Petroleum Exploration in Western Greece. Institute for Geology and Subsurface Research, Athens, Special Report, 10. BRIX, M. R., STt)CKERT, B., SEIDEL, E., THEYE, T., THOMSON, S. N. & Kf0STER, M. 2002. Thermobarometric data from a fossil zircon partial annealing zone in high pressure-low temperature rocks of eastern and central Crete, Greece. Tectonophysics, 349, 309-326. BUCK, W. R. 1993. The effects of lithosphere thickness on the formation of high- and low-angle normal faults. Geology, 21,933-936. CAMERLENGHI, A., CITA, M. B., DELLA VEDOVA, B., FusI, N., MIABILE, L. & PELLIS, G. 1995. Geophysical evidence of mud diapirism on the

149

Mediterranean Ridge accretionary complex. Marine Geophysical Researches, 17, 115-141. CAPEDRI, S., TOSCANI, L., GRANDI, R., VENTURELLI, G., PAPANIKOLAOU,D. & SKARPELIS, N. S. 1997. Triassic volcanic rocks of some type-localities from the Hellenides. Chemie der Erde, 57, 257-276. CATALANO, R., DI STEFANO, P. & KOZUR, H. 1991. Permian circum-Pacific deep-water faunas from the western Tethys (Sicily, Italy)--new evidence for the position of the Permian Tethys. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 75-108. CATALANO, R., FRACHINO, A., MERLINI, S. & SULLI, A. 2000a. A crustal section from the eastern Algerian basin to the Ionian ocean (Central Mediterranean), Memorie della Societd Geologica d'Italia, 55, 71-85. CATALANO, R., FRACHINO, A., MERLINI, S. & SULLI, A. 2000b. Central Western Sicily structural setting interpreted from seismic reflection profiles. Memorie della Societd Geologica d'Italia, 55, 5-16. CATALANO, R., DOGLIONE, C. & MERLIN, S. 2001. On the Mesozoic Ionian Basin. Geophysical Journal International, 144, 49-64. CENSI, P., CHIAVETTA,S., FERLA, P., SPEZIALE,S. & DI STEFANO, P. 2000. Tholeiitic magrnatites in Lower Permian turbidites from Western Sicily. Memorie della Societd Geologica d'Italia, 55, 307-313. CHAMPOD, E. C., STAMPFLI, G. M. & KOCK, S. 2004. Permo-Triassic evolution of the Tethyan margins in the external Hellenides. 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 14-20 April 2004, Extended Abstracts, 53-56. CHAUMILLAN,E. & MASCLE,J. 1997. From foreland to forearc domains; new mutichannel seismic survey of the Mediterranean Ridge accretionary complex (Eastern Mediterranean). Marine Geology, 138, 237-259. CHAUMILLON, E., MASCLE, J. & HOFFMANN, J. 1996. Deformation of the western Mediterranean Ridge: importance of Messinian evaporite formation. Tectonophysics, 263, 163-190. CREUTZBURG, N., DROOGER, C. W. & MEULENKAMP, J. E. 1977. General Geological Map of Greece, Crete Island, Scale 1:200,000. Institute of Geology and Minerals Research, Athens. DANAMOS, G. (1992). Contribution to geology and hydrogeology of the island of Kithira. PhD thesis, University of Athens. DE BONO, A., VAVASSIS,I., STAMPFLI,G. M., MARTINI, R., VACHARD, D. & ZANINETTI, L. 1998. New stratigraphic data on the Pelagonian pre-Jurassic units of Evia island (Greece). Annales G~ologiques des Pays Hell~niques, 38A, 11-24. DEGNAN, P. J. & ROBERTSON,A. H. F. 1998. Mesozoic-early Cenozoic passive margin evolution of the Pindos ocean (NW Peloponnese, Greece). Sedimentary Geology, 117, 33-70. DEGNAN, P. J. & ROBERTSON, A. H. F. 2000b. Synthesis of the tectonic-sedimentary evolution of the Mesozoic-Early Cenozoic Pindos ocean evidence from the NW Peloponnese, Greece, In: ROBERTSON, A. H. F. & MOtmTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 467-491.

150

A . H . F . ROBERTSON

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. DERCOURT, J., RICOU, L. E. & VRIELYNCK, B. (eds), 1993. Atlas Tethys Palaeoenvironmental Maps. Gauthier-Villars, Paris. DERCOURT, J., GAETANI, M., VRIELYNCK, B., et al. (eds) 2000. Peri-Tethys Palaeogeographical Atlas (2000). CCGM/CGMW, Paris. DE WEVER, P. 1975. Etude gkologique des sbries apparaissant en fen~tre sous l'allochtone pindique (sgrie de Tripolitza et sgrie gpimetamorphique de Zarouchla). Peloponnkse septentrional, Grkce. Th~se 36me cycle, Universit6 de Lille. DE WEVER, P. 1975. Radiolarians, radiolarites, and Mesozoic paleogeography of the CircumMediterranean Alpine belts. In: HEIM, J.R. & OBRADOVI6, J. (eds) Siliceous Deposits of the Tethys and Pacific Regions. Springer, Berlin, 31-50. DI STEFANO,P. & GULLO, M. 1996. STOP 10. Valle del Sosio, Palazzo Adriano. I terreni permiani e triassici del bacino sicana nell'evoluzione dela Catena Siciliana Centro-Meridionale. Italian Geological Society, 79th National Congress, Excursion Guide, Western Sicily, 1, 95-119. DI STEFANO, P. d~ GULLO, M. 1997. Late PalaeozoicEarly Mesozoic stratigraphy and paleogeography of Sicily. In: CATALANO,R. (ed.). Field Workshop in Western Sicily, 11-13 June, Guidebook. 8th Workshop of the ILP Task Force 'Origin of Sedimentary Basins', Palermo, 7-13, June 1997. University of Palermo, Department of Geology and Geodesy, 89-101. DI STEFANO, P., ALESSI, A. & GULLO, M. 1996. Mesozoic and Palaeogene megabreccias in southern Sicily: new data on the Triassic paleomargin of the Sicilo-Tunisian Platform. Facies, 34, 101-122. DITTMAR, U. & KOWALCZYK, G. 1991. Die Metaklastite im liegenden der Plattenkalk-Karbonate der siidlichen Peloponnes. Zeitschrift der Deutschen Geologischen Gesellschaft, 142, 209-227. DITTMAR, V., Joos, C. & KOWALCZYK,G. 1989. DaN liegende der Plattenkalk-Karbonate im Taygetos (Sud-Peloponnese). Nachrichten, Deutsches Geologischen Gesellschaft, 41, 88-89. DtXON, J. E. & ROBERTSON, A. H. F. 1993. Arc signatures in Mediterranean Triassic rift basalts: a lithosphere-hosted inheritance from Hercynian subduction. Journal of Conference Abstracts, 10, 314. DIXON, J. E. & ROBERTSON, A. H. F. 1999. Are multiple plumes implicated in the Triassic break-up of the Gondwanan margin in the Eastern Mediterranean region? Journal of Conference Abstracts, 10, 314. DOERT, U., KOWALCZYK, G., KAUFFMANN, G. & KRAHL, J. 1985. Zur stratigraphischen Einstufung der 'Phyllit-Serie' von Krokee und der Halbinsel Xyli (Lakonien, Peloponnes). Erlanger Geologischen Abhandlungen, Erlangen, 112, 1-10. DORNSIEPEN, U. F. & MANUTSOGLU, E. 1996. Die vulkanite oder anorogene Trapp-Basalte? Zeitschrift der Deutschen Geologischen Gesellschaft, 147, 101-123.

DORNSIEPEN, U. F., MANUTSOGLU, E. & MERTMANN, D. 2001. Permo-Triassic palaeogeography of the External Hellenides. Palaeogeography, Palaeoclimatology, Palaeocology, 172, 327-338. DOUTSOS, T., KOUKOUVELAS, I., POULIMENOS, G., KOKKALAS, S., XYPOLIAAS, P. & SKOURLIS, K. 2000. An exhumation model of the south Peloponnesus, Greece. International Journal of Earth Science, 89, 350-365. EPTING, M., KUDRASS, H. R., LEPPIG, U. & SCHAFER, A. 1972. Geologie der Talea Ori, Kreta. Neues Jahrbuch fiir Geologie und Paldontologie, Monatshefte, 141, 259-285. FASSOULAS, C. G. 2001. Field Guide to the Geology of Crete. Natural History Museum of Crete, University of Crete, Heraklion. FASSOULAS, C., RAHL, J. M., AGUE, J. & HENDERSON, K. 2004. Patterns and conditions of deformation in the Plattenkalk Nappe, Crete: a preliminary study. Proceedings of the lOth International Congress, Thessaloniki, Greece, April 2004, Extended abstracts. FINGER, F., KRENN, E., RIEGLER, G., ROMANO, S. & ZULAUF, G. 2002. Resolving Cambrian, Carboniferous, Permian and Alpine monazite generations in the polymetamorphic basement of eastern Crete (Greece) by means of electron microprobe. Terra Nova, 14, 233-240. F/TTON, J. G., SAUNDERS, A. D., LARSON, L.M., HARDARSON, B. S. & NORRY, M. S. 1998. Volcanic rocks of the southeastern Greenland margin. In: SAUNDERS, A. D., LARSEN, H. C. & WISE, S. W, Jr (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 152. Ocean Dilling Program, College Station, TX, 331-350. FLEURY, J. J. 1980. Evolution d'une platforme et d'un bassin dans leur cadre alpin: les zones de GavrovoTripolitza et du Pinde-Olonos. Annales de la Societk Gkologique du Nord, Special Publication, 4, 651. FLUGEL, E., DI STEEANO,P. & SENOWARI-DARYAN,B. 1991. Microfacies and depositional structure of the allochthonous carbonate base-of-slope deposits: the Late Permian Pietra di Salomone Megablock, Sosio Valley (Western Sicily). FACIES, 25, 147-186. FOWLER, S. R., WHITE, R. S., SPENCE, G. D. WESTBROOK, G. K. 1989. The Hatton Bank continental margin--II. Deep structure from twoship expanding spread seismic profiles. Geophysical Journal, 96, 295-309. FYTROLAKIS, N. 1971. Die dis Heute unbekanten paleozoischen Sudostich von Kalamai. Bulletin of the Geological Society of Greece, 8, 70-81. GARFUNKEL, Z. 2004. Origin of the Eastern Mediterranean basin: a re-evaluation. Tectonophysics, 391, 11-34. GEROLYNATOS, I. K. 1994. Metamorphose und Tektonik der Phyllit-Quartzit-Serie und der Tyros-Schichten auf dem Peloponnes und Kithira Berliner Geowissenschaftliche Abhandlung, Reihe A,. 164. GLENNIE, K. W., HUGHES-CLARKE, M. W., BOEUF, M. G. A, PILAAR,W. F. H. & REINHARDT,B. 1990. Inter-relationship of the Makran-Oman Mountain

TESTING ALTERNATIVE SOUTH M E D I T E R R A N E A N TECTONIC MODELS belts of convergence. 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, 773-786. GRADSTEIN, J. G., OGG, J. G., SMITHA. G., et al. 2004. A Geologic Time Scale. Cambridge University Press, Cambridge. GROMET, L. P., DYMEK, R. F., HASKIN, L. A. & KOROTEV, R. L. 1984. The 'North American Shale Composite' its composition, major and trace element characteristics. Geochimica et Cosmochimica Acta, 48, 2469-2482. GUIRAUD, R., ISSAWl, B. & BOSWORTH, W. 2001. Permo-Mesozoic evolution of the western Tethys realm: the Neotethys East Mediterranean basin connection. In- ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUIN-SOLEAU, S. (eds). Peri-Tethys Memoir, 5. Peri-Tethyan Rift/ Wrench Basins and Passive Margins. M6moires du Mus6um National d'Histoire Naturelle, 469-510. HALL, R. & AUDLEY-CHARLES,M. G. 1983. The structure and regional significance of Talea Ori, Crete. Journal of Structural Geology, 5, 167-197. HALL, R., AUDLEY-CHARLES,M. G. & CARTER, D. J. 1984. The significance of Crete for the evolution of the eastern Mediterranean. In: DIXON, J.E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 499-516. HANKEL, O. 1994. Early Permian to Middle Jurassic rifting and sedimentation in East Africa and Madagascar. Geologische Rundschau, 83, 703-710. HAUDE, G. 1989. Geologie der Phyllite im Gebiet um Palekastro (Nordoest-Kreta). PhD thesis, Technical University Munich. HIMMERKUS, F., REICHMANN,T. & KOSTOPONLOS,D. 2006. Late Protozoic and Silurian basement units within the Serbo-Macedonian Massif, northern Greece: the significance of terrane accretion in the Hellenides. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic"Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 35-50. JACOBSHAGEN, V. 1986. Geologie yon Griechenland. Borntraeger, Berlin. JOLIVET, L., GOFFt~, B., MONIE, P., TRUFFERT-LUXEY, C., PATRIAT, M. & BONNEAU, M. 1996. Miocene detachment in Crete and exhumation of P-T-t paths of high-pressure metamorphic rocks. Tectonics, 15, 1129-1153. KARAMATA,S. 2006 The geological development of the Balkan Peninsula related to the approach, collision and compression of Gondwanan and Eurasian units. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediteranean Region. Geological Society, London, Special Publications, 260, 155-178. KAZMIN, V. G. & TIKHONOVA, N. F. 2006. Evolution of Early Mesozoic back-arc basins in the Black Sea-Caucasus segment of a Tethyan active margin. In: ROBERTSON, A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediteranean Region. Geological Society, London, Special Publications, 260, 179-200.

151

KILIAS, A., FASSOULAS, C. & MOUNTRAKIS, D. 1994. Cenozoic extension of continental crust and uplift of Psiloritis metamorphic core complex in the central part of the Hellenic arc. Geologisches Rundschau, 83, 417-430. KILIAS, A. A., TRANOS, M. n., OROXCO, M., ALONSOCHAVES, F. i . t~ SOTO, J. I. 2002. Extensional collapse of the Hellenides: a review. Revista de la Sociedad Geol6gical de Espa~a, 15, 129-139. KOLOKOTRONI,C. & DIXON, J. E. 1991. The origin and emplacement of the Vrondou granite, Serres, N.E. Greece. Bulletin of the Geological Society of Greece, 25, 469-483. KoPP, K. O. & OTT, E. 1977. Spezialkartierungen im umkreis meuer fossilfunde in Trypali und Tripolitza-Kalken Westkretas. Neues Jahrbuch fiir Geologie und Paldontologie, Monatshefte, 1977, 217-238. KoPP, K. O. & WERNADO, G. 1983. Uber eine intratriadische Deckenbwegung auf Kreta. Geologische Rundschau, 72, 895-910. KOZUR, H. 1993. Upper Permian radiolarians from the Sosio Valley area, western Sicily (Italy) and from the Uppermost Lamar Limestone of west Texas. Geologische Jahrbuch B, A, 136, 99-123. KOZUR, H. 1995. First evidence of Middle Permian Ammonitico Rosso and further new stratigraphical results in the Permian and Triassic of the Sosio Valley area, Western Sicily. Proceedings, First Croatian Geological Congress, 1, 307-310. KOZUR, H. & KRAHL, J. 1984. Erster Nachweis triassischer Radiolarien in der Phyllit-Gruppe auf der Insel Kreta. Neues Jahrbuch fiir Geologie und Paliiontologie, Monatshefte, 1984(7), 400-404. KOZUR, H. W., KRAINER, K. & MOSTER, H. 1996. Ichnology and sedimentology of the Early Permian deep-water deposits from the LercaraRoccapalumba area (Western Sicily, Italy). FACIES, 34, 34-41. KRAHL, J. t~ KAUFFMAN, G. 2004. New aspects for a palinspastic model of the External Hellenides on Crete. 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 14-20 April 2004, Extended Abstracts, 119-122. KRAHL, J., EBERLE, P., E1CHKOFF, J., FOSTER, O. & KOZUR, H. 1982. Biostratigraphical investigations in the Phyllite-Quartzite Group on Crete Island, Greece. International Symposium on the Hellenic Arc and Trench (H.E.A. T ), Proceedings Athens, 1, 306-323. KRAHL, J., FANDRICH,J., FORSTER,O. & HEINRITZI,F. 1983a. Neue Daten zur Biostratigraphie und zur tektonischen Lagerung der Phyllit-Gruppen auf der Insel Kreta/Griechenland). Zeitschrift der Deutschen Geologischen Gesellschaft, 137, 523-536. KRAHL, J., FORSTER, O., HEINRITZI, F., KAUFFMANN, G., KOZUR, H. & RICHTER, D. 1983b. A stratigraphical concept for the HP/LT-metamorphic Phyllite Group on Crete Island and its palaeogeographical implications for the External Hellenides. Terra Cognita, 3(2-3), 228. KRAHL, J., KAUFFMAN, G., KOZUR, H., RICHTER, D., FOSTER, O. & HEINRITZI,F. 1983c. Neue Daten zur Biostratigraphie und zur teklonischen Lagerung

152

A . H . F . ROBERTSON

der Phyllit-Gruppe und der Trypali-Gruppe auf der Insel Kreta (Griechenland). Geologische Rundschau, 72, 1147-1166. KRAHL, J., KAUFMANN, G., RICHTER, D., et al. 1986. Neue Fossilfunde in der Phyllit-Gruppe Ostkretas (Griechenland). Zeitschrift der Deutschen Geologischen Gesellschaft, 137, 523-536. KRAHL, J., RICHTER, D., FORSTER, O., KOZUR, H. & HALL, R. 1988. Zur Stellung der Talea Ori im Bau des kretischen Deckenstapels (Griechenland). Zeitschrift der D e u t s c h e n Geologischen Gessellschaft, 149, 191-227. KTENAS, C. A. 1924. Formations primaires s6mim6tamorphiques au P61oponn6se central. Comptes Rendus de la Societk GOologique de France, 24, 1-63. KTENAS, C. A. 1926. Sur le d6v61opment du primaire au P61oponn6se central. Praktikatis Academias, Athinon, 1, 53-59. LEKKAS, S. & PAPANIKOLAOU,D. 1980. On the phyllite problem in the Peloponnnese. Annales Gkologique des Pays Hellkniques, 29, 395-410. 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. LIATI, A., GEBAUER, D. & FANNING, M. 2004. The age of ophiolitic rocks of the Hellenides (Vourinos), Pindos, Crete); first ion microprobe (SHRIMP) zircon ages. Chemical Geology, 207, 171-188. MOUNTRAKIS, D. 1986. The Pelagonian Zone in Greece: a polyphase deformed fragment of the Cimmerian continent and its role in the geotectonic evolution of the eastern Mediterranean. Journal of Geology, 94, 335-347. NEUBAUER, F. & HANDLER,R. 1999. Variscan orogeny in the Eastern Alps and Bohemian Massif: How do these units correlate? In: NEUBAUER,F. & HOCK, V. (eds) Aspects of Geology in Austria. Mitteilungen der Osterreichischen Geologischen Gesellschaft, 92, 15-34. NIKISHIN, A. M., ZIEGLER, P. A., PANOV, D. I., et al. 2001. Mesozoic-Cenozoic evolution of the Scythian Platform-Black Sea-Caucasus domain. In: ZIEGLER, P., CAVAZZA,W., ROBERTSON,A. H. F. & CRASQUIN-SOLEAU,S. (eds) Peri-Tethys Memoir, 5. Peri-Tethyan Rift~Wrench Basins and Passive Margins. Mkmoires du Mus6um National d'Histoire Naturelle, 295-346. OKAY, A. I. 2000. Was the late Triassic orogeny in Turkey caused by the collision of an oceanic plateau? In: BOZKURT,E., WINCHESTER,J. A. & PIPER, J. D. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 25-42. PAPANIKOLAOU, D. J. 1988. Field Guide Book. Introduction to the Geology of Crete. IGCP Project, 276, Technical University of Crete, Chania. PAPANIKOLAOU, D. J. 1996-1997. Introduction to the terrane descriptions of the Alpine Tethyan belt. In: PAPANIKOLAOU,D. J. (ed.) Terrane Maps and Terrane Descriptions. IGCP Project, 276, 295-514. PAPANIKOLAOU, D. J. & EBNER, F. 1996-1997. Introduction to the terrane descriptions of the Alpine

Tethyan belt. In: PAPANIKOLAOU, D. J. (ed.) Terrane Maps and Terrane Descriptions. IGCP Project, 276, 195-197. PAPANIKOLAOU, D. J. & SKARPELIS, N. S. 1986. The blueschists in the external metamorphic belt of the Hellenides: composition, structure and geotectonic significance of the Arna unit. Annales G~ologique des Pays Hell~niques, 33, 47-68. PARASKEVOPOULOU, G. M. 1951. The coals of Monemvasia area. Annales Gkologique des Pays Hell~niques, 3, 32-41. PEARCE, J. A. 1980. Geochemical evidence for the genesis and eruptive setting of lavas from Tethyan ophiolites. In: PANAYIOTOU, A. (ed.) Proceedings of the International Symposium, 'Troodos' 1979. Geological Survey of Cyprus, Nicosia, 261-272. PE-PIPER, G. 1982. Geochemistry, tectonic setting and metamorphism of the mid-Triassic volcanic rocks of Greece. Tectonophysics, 85, 153-272. PE-PIPER, G. 1983. The Triassic volcanic rocks of Tyros, Zarouhla, Kalamae, and Epidavros, Peloponnese, Greece. Schweizerische Mineralogische and Petrographische Mitteilungen, 63, 249-266. PE-PIPER, G. & PIPER, D. W. J. 1998. The nature of Triassic extension-related magmatism in Greece: evidence for Nd and Pb isotope geochemistry. Geological Magazine, 13, 331-348. PE-PIPER, G. & PIPER, D. W. J. 2002. The Igneous Rocks of Greece. The Anatomy of an Orogen. Beitrage zur Regionalen Geologic der Erde, 30. PIGRAM, C. J. & PANNABEAN, H. 1984. Rifting of the northern margin of the Australian continent and the orgin of some microcontinents in eastern Indonesia. Tectonophysics, 107, 231-351. POMONI-PAPAIOANNOU, F. & KARAKITSIOS,V. 2002. Facies analysis of Trypali carbonate unit (Upper Triassic) in central-western Crete (Greece); an evaporite formation transformed into solutioncollapse breccias. Sedimentology, 49, 1113-1132. PSONIAS, K.T. 1981. Presence of Permo(?)-lower Triassic beds at the base of the Plattenkalk series in Taygetos. Description of a continuous section. Annales Gkologique des Pays Hklleniques, 30, 578-587. PURSER, B. H. & BOSENCE, D. W. J. (eds) 1998. Sedimentation and Tectonics of Rift Basins: Red Sea Gulf of Aden. Chapman & Hall, London. RASSlOS, A. H. E. & MOORES, E. M. 2006. Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab (northern Greece). In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications. 260, 237-266. REID, I. D. & KEEN, C. E. 1990. Deep crustal structure beneath a rifted basin: results from seismic refraction measurements across the Jeanne d'Arc Basin, offshore eastern Canada. Canadian Journal of Earth Science, 27, 1462-1471. RICOU, L.-E. 1996. The plate tectonic history of the past Tethys ocean. In: NAIRN, A. E. M., RICOU, L.-E., VRIELYNCK, B. & DERCOURT, J. (eds) The Ocean Basins and Margins, 8, The Tethys Ocean, Plenum, New York, 3-62.

TESTING ALTERNATIVE SOUTH M E D I T E R R A N E A N TECTONIC MODELS ROBERTSON, A. H. F. 1994. Role of the tectonic facies concept in orogenic analysis and its application to Tethys in the Eastern Mediterranean region. Earth-Science Reviews, 37, 139-213. ROBERTSON, A. H. F. & BAMAKHALIF,K. A. S. 2001. Late Oligocene-Early Miocene rifting of the northeast Gulf of Aden: basin evolution in Dhofar (South Oman). In: ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUIN-SOLEAU,S. (eds) Peri-Tethys Memoir, 5. Peri-Tethyan Rift~Wrench Basins and Passive Margins. Mtmoirs du Mustum National d'Histoire Naturelle, 641-671. ROBERTSON, A. H. F & DIXON, J. E. 1984. Introduction: aspects of the Geological evolution of the eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. ROBERTSON, A. H. F., CLIET, 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. & MOUNTRAKIS, D. 2006. Tectonic development of the Eastern Mediterranean Region: an introduction. In: Tectonic Development of the Eastern Mediterranean Region, Geological Society, London, Special Publications, 260, 1-9. ROBERTSON, A. H. F., DIXON, J. E., BROWN, S., et al. 1996, Alternative tectonic models for the Late Palaeozoic-Early Cenozoic development of Tethys in the Eastern Mediterranean region. In MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBERTSON, A. H. F., USTAtMER, T., PICKETT, E. A., COLLINS, A., ANDREW,T. & DIXON, J. E. 2004. Testing models of Late Palaeozoic-early Mesozooic orogeny: support for an evolving one-Tethys model. Journal of the Geological Society, London, 161, 501-511. ROMANO, S., DI3RR, W. & ZULAUF, G. 2002. U Pb-zircon datings and quartz textures from prealpine basement of Eastern Crete. In: Nurnberg L.f.g.U.E. (ed.) 9. Symposium TektonikStrukturgeologie-Kristallingeologie. Universitat Erlangen Nuernberg, 3, 81-82. ROMANO, S., DI3RR, W., FINGER, F. & ZULAUF, G. 2004. The complexity of the Cretan pre-Alpine basement: new age information and structural data. 5th International Symposium on Eastern Mediterranean Geology, ThessalonikL Greece, 14-20 April 2004, Extended Abstracts, 1, 179-181. ROMANO, S., BRIX, M. R., DORR, W., FIALA, J., KRENN, E. & ZULAUF, G. 2006. The Carboniferous to Jurassic evolution of the pre-Alpine basement of Crete: constialints from U Pb and U-(Th)-Pb dating of orthogneiss, fissiontrack dating of zircon and structural-petrological data. In: ROBERTSON,m. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 69-90.

153

SCOTESE, C. R. & LANGFORD, R. P. 1995. Pangea and the palaeogeography of the Permian. In: SCHOLLE, P. A., PERYT, T. M. & ULMER-SCHOLLE,D. S. (eds) The Permian of Northern Pangea, 1, Palaeogeography, Palaeoclimates, Stratigraphy. Springer, Berlin, 3-19. SEIDEL, E., 1978. Zur Petrologie der Phyllite-QuartzitSeries Kretas. Habilitationsschrift, Universitfit Braunschweig. SEIDEL, E., KREUZER, H. & HARRE, W. 1982. A Late Oligocene/Early Miocene high pressure belt in the External Hellenides. Geologisches Jahrbuch, 23, 165-206. ~ENGOR, A. M. C. 1984. The Cimmeride Orogenic System and the Tectonics of Eurasia. Geological Society of America, Special Papers, 195. SHARP, I. A. & ROBERTSON, A. H. F. 2006 Tectonicsedimentary evolution of the western margin of the Mesozoic Vardar Ocean: evidence from Pelagonian and Almopias Zones. northern Greece. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 373-412. SKARPELIS, N. 1982. Metallogeny of massive sulphides and petrology of the External Metamorphic Belt of the Hellenides (SE Peloponnesus). PhD thesis, University of Athens. SMITH, m. G. 1999. Gondwana: its shape, size and position from Cambrian to Triassic time. Journal of African Earth Science, 28, 71-97. SMITH, A. G. 2006 Tethyan ophiolite emplacement, Africa to Europe motions, and Atlantic spreading. In: ROBERTSON, A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 11-34. SMITH, A. G., HYNES, A. J., MENZIES, M., NISBET, E. G., PRICE, I., WELLAND,M. J. & FERRII~RE,J. 1975. The stratigraphy of the Othris Mountains, Eastern Central Greece: a deformed Mesozoic continental margin sequence. Eclogae Geologicae Helvetiae, 68, 463-481. SMITH, A. G., HURLEY, m. M. & BRIDEN, J. C. 1981. Phanerozoic Palaeocontinental Maps. Cambridge University Press, Cambridge. STAMPFLI, G. M. & BOREL, G. D. 2002. A plate tectonic model for the Palaeozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrones. Earth and Planetary Science Letters, 169, 17-33. STAMPFLI, G., MARCOUX,J. & BAUD, A. 1991. Tethyan margins in space and time. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 373-410. STAMPFLI, G., MOSAR, J., DE BONO, A. & VAVASSIS,I. 1998. Late Palaeozoic, early Mesozoic plate tectonics of the western Tethys. Bulletin of the Geological Society of Greece, 32, 113-120. STAMPFLI, G., MOSAR, J., FAURE, P., PILLEVUIT,A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethys realm: the Neotethys East Mediterranean basin connection. In: ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUINSOLEAU, S. (eds) Peri-Tethys Memoir, 5. Per# Tethyan Rift~Wrench Basins and Passive Margins.

154

A . H . F . ROBERTSON

M6moires du Mus6um National d'Histoire Naturelle, 51-108. STAMPFLI, G. M., VAVASSIS, I., DE BONO, A., ROSSELET, F., MATTI, B. & BELLINI,M. 2003. Remnants of the Paleotethys oceanic suture-zone in the western Tethys area. Bolletino della Societa Geologica Italiano, Special Volume, 2, 1-23. STECKLER, M. A. & OMAR, G. I. 1994. Controls of erosional retreat on the uplifted flanks of the Gulf of Suez and Northern Red Sea. Journal of Geophysical Research, 99, 12119-12173. TEN VEEN, J. H. & MEIJER, P. T. 1999. Late Miocene to Recent tectonic evolution of Crete (Greece): geological observations and model analysis. Tectonophysics, 298, 191-208. THEYE, T., SEIDEL, E. & VIDAL, O. 1992. Carpholite, sudoite and chloritoid in high-pressure metapelites from Crete and the Peloponnese, Greece. European Journal of Mineralogy, 4, 487-507. THII~BAULT, F. 1982. Evolution gkodynamique des HOllenides externes en PeloponnOse mOridionale (Grbce). Societ6 G6ologique du Nord, Special Publication, 6. THII~BAULT, C. 1991. Interpr6tation des donn6es g6ochimiques concernanant les metabasaltes associ6s a la Nappe Inferieur des Phyllades (P61oponn6se m6ridional, Gr+ce) Site g6odynamique de mise en place. Annales de la SocietO Gbologique du Nord, CIX, 193-205. THII~BAULT,F. & KOZUR, H. 1979. Pr6cisions sur l'age de la formation de Tyros (Pal6ozoique sup6rieurCarnien) et de la base de la s6rie de GavrovoTripolitza (Carnian), Peloponn6se m6ridional, Gr6ce. Compte Rendus de l'Acadkmie des Sciences, 288, 23-26. THOMPSON, S. N., STOECKHERT, B. & BRIX, M. R. 1988. Thermochronology of the high-pressure metamorphic rocks of Crete, Greece: Implications for the speed of tectonic processes. Geology, 26, 259-262. TRIBOULET, C. & BASS1AS, Y. 1986. Origine magmatique et g6odynamique des mgtavolcanites associ6es aux Phyllades (P61oponn6se, Gr6ce). Annales de la Societg GOologique du Nord, CV, 11-26. TUCKOLKE, B., SIBUET, J.-P., KLAUS, A., et al. (eds) 2004. Proceedings of the Ocean Drilling Program, Initial Reports, 210. National Science Foundation, Joint Oceanographic Institutions Inc. Texas A & M University, College Station, TX. USTAOMER, P. A., MUNDIL, R. & RENNE, P. R. 2005. U/Pb and Pb/Pb zircon ages for arc-related intrusions of the Bolu Massif (W Pontides, NW Turkey): evidence for Late Precambrian (Cadomian) age. Terra Nova, 17, 215-223. USTAOMER,T. & ROBERTSON,A. H. F. 1997. Tectonicsedimentary evolution of the north Tethyan margin in the Central Pontides of northern Turkey. In: ROBINSON, A. G. (ed.) Regional and Petroleum

Geology of the Black Sea and surrounding Region. American Association of Petroleum Geologists, Memoirs, 68, 255-290. VAVASSIS, I., DE BONO, A., VALLOTON,A., STAMPFLI, G. M. & AMELIN, Y. 2000. U-Pb and Ar-Ar geochronological data from Pelagonian basement in Evia (Greece): geodynamic implications for the evolution of Paleotethys. Schweizerische Mineralogische and Petrographische, Mitteilangen, 80, 21-43. VON HUENE, R. & SCHOLLE, D. 1991. Observations at convergent margins concerning sediment subduction, subduction erosion, and the growth of continental crust. Reviews of Geophysics, 29, 279-316. VON RAD, U., EXON, N., F., BOYD, R. & HAQ, B. U. 1992. Mesozoic palaeoenvironments of the rifted margin of NW Australia (ODP Leg 1221123). Geophysical Monographs, American Geophysical Union, 70, 157-184. W~LSON, R. C. L. 1988. Mesozoic development of the Lusitanian Basin, Portugal. Revista de la Sociedad Geologica de Espa~a, 1, 393-407. WURM, A. 1950. Zur Kenntnis des metamorphikums der Insel Kreta. Neues Jahrbuch fiir Geologie und Paliiontologie, Monatshefte, 1950, 206-239. XYPOLIAS,P. & DOUTSOS, T. 2000. Kinematics of rock flow in a crustal-scale shear zone: implications for the orogenic evolution of the southwestern Hellenides. Geological Magazine, 137(1), 81-96. XYPOLIAS, P., D()RR, W. 8r ZULAUF, G. 2006. Late Carboniferous plutonism within the pre-Alpine basement of the External Hellenides (Kithira, Greece): evidence from U-Pb zircon dating. Journal of the Geological Society, London, 163, 539-547. YILMAZ, P. O., NORTON, I. O., LEARLY, D. & CHUCHLA, R. A. 1996. Tectonic evolution and palaeogeography of Europe. In: ZIEGLER, P. A. HORVARTH, F. (eds) Peri-Tethys Memoir 2: Structure and Prospects of Alpine Basins and Forelands. M6moires de Mus6um National d'Histoire Naturelle, 48-60. ZIEGLER, P. A. 1988. Evolution of the Arctic-North Atlantic and the western Tethys. American Association of Petroleum Geologists, Memoirs, 43, 1-198. ZIEGLER, P. A. ~r STAMPFLI, G. 2001. Late PalaeozoicEarly Mesozoic plate boundary reorganisation: collapse of the Variscan orogen and opening of Neotethys. Natura Bresciana. Annali del Museo Civico Naturale, Brescai, Monografia, 25, 17-34. ZULAUF, G., KOWALCZYK,G., KRAHL, J. t~r SCHWANZ, S. 2002. The tectonometamorphic evolution of high-pressure low-temperature metamorphic rocks of eastern Crete, Greece: constraints from microfabrics, strain, illite crystallinity and paleodifferential stress. Journal of Structural Geology, 24, 1805-1828.

The geological development of the Balkan Peninsula related to the approach, collision and compression of Gondwanan and Eurasian units STEVAN KARAMATA

Serbian Academy o f Sciences and Arts, Knez Mihailova 35, 11000 Belgrade, SCG (e-mail." kristinas@mkpg, rgf.bg, ac. yu) Abstract: The Balkan Peninsula includes the margins of both Eurasia (the Moesian microplate) and Gondwana (the Adria microplate as a promontory); it also includes ophiolitic belts that represent remnants of Tethys and its marginal seas. Various terranes docked to larger crustal units and were incorporated to form new units. Most units within the Balkan Peninsula moved northwards to their present positions, jointly or independently, from positions around, or south of, the Equator from the end of the Palaeozoic to the present day. The assembly of these units was associated with generally northeastward subduction of Tethys. The first main period of docking was in the Carboniferous. Later, from the Permian to the Maastrichtian, marginal seas opened and later closed. Island arcs formed within the northwestern part of Tethys, and parts of continental margins were detached and relocated, or were transported along transcurrent faults. In the Maastrichtian the entire oceanic area was closed and the main units sutured. The resulting assemblage later underwent additional compression, rotation and transcurrent displacement of some units.

The Balkan Peninsula (BP) is situated in the northwestern part of the Eastern Mediterranean region and is of particular interest as it includes units of different provenance that are now sutured. This part of the Eastern Mediterranean region has been discussed by numerous workers (see e.g. Robertson et al. 1996) but because of its complexity requires further consideration. The BP includes the following main units: (1) the Moesian microplate, part of the southern margin of Eurasia; (2) Adria (often-termed Apulia), a microplate forming a promontory of Gondwana; (3) remnants of Tethys and related marginal seas. The first two units are made up of continental crust and the third by oceanic crust. During the Palaeozoic and Mesozoic different terranes were transported together with Tethyan oceanic crust and then docked to continental units, becoming larger entities. This process of terrane accretion lasted from the Early Palaeozoic until almost the end of the Cretaceous and was contemporaneous with subduction of oceanic lithosphere. Remnants of Tethys and its marginal seas are preserved as ophiolitic belts within the sutured continental units. Suturing was followed by compression, rotation of some units and large transcurrent movements. The northern boundary of the BP is here taken as a transcurrent fault zone that is located along the southern margin of the metamorphic basement of the Pannonian basin. This boundary, in Slavonia (west of the Danube), forms a c. 20 km wide, tectonically mixed zone resulting

from thrusting of metamorphic rocks of the Pannonian basement over younger deposits (Pami6 & Belak 1996-1997). East of the Danube, reflecting a relatively deep erosion level, this boundary is clearly exposed, trending eastwards from the Danube (Kemenci & Canovi6 1997). Further east again it cuts the Southern Carpathians, continues along the Danube and then deviates towards Constanca and the Black Sea. The southern boundary of the BP is not easy to define because of the combined presence of the Aegean Sea and a subduction zone to the south (south of Crete), and the effects of the westward movement of Anatolia, but is here taken as a line between Corfu and Olympos and the northern margin of the Aegean Sea. The eastern boundary of the BP is defined as the Black Sea coast, and the western boundary as the shores of the Adriatic and Ionian seas. Along the part of the collision zone between Gondwana-related units and Eurasia-related units that is exposed in Central Macedonia, strong post-collisional compression was followed by uplift, exposing the roots of some geological structures, displacing units and obscuring relationships. To the north of this zone, in Central Serbia, post-collisional compression was less strong and thus relationships are better preserved. Many models for the formation of the existing geological assemblage of the BP have been proposed. However, most units are mentioned only generally in regional overviews. In one such recent synthesis, by Stampfli & Borel (2004), only

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 155-178. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

156

S. KARAMATA

selected units of the BP were mentioned. Most other papers consider only some parts of the BP. The western part was discussed by Aubouin et al. (1970 a, b) and Aubouin (1973); its northern part (i.e. the Dinarides) by Dimitrijevid (1982, 2001), Herak (1986) and Dimitrijevid & Sikogek (1997); its central part (i.e. the Albanides) by Shallo (1992, 1994) and Shallo in Papanikolaou (1996-1997a; only the maps/sheet 1 and tectonostratigraphic diagrams of Albania), and its southern parts (i.e. the Hellenides) by Jacobshagen (1979). For the eastern parts (i.e. the Balkanides and the Carpathians) the most important studies are those by Bon6ev (1955) and Sandulescu (1984). For the central part of the BP (i.e. the Vardar Zone and its margins) the paper by Kossmat (1924) is still important, as well as those by Milovanovid (1950), Petkovi6 (1961) and Dimitrijevid (1997). In these contributions the units of the BP are considered, in line with the then current understanding, as parts of a double orogen with a median unit between, or characterized by large 'Alpine type' thrusts (i.e. nappe tectonics). The review of ophiolitic belts in central and southwestern parts of the BP by Robertson & K ar am a t a (1994), as well as papers by K a r a m a t a et al. (1999, 2000a), initiated a new approach to the interpretation of the Mesozoic evolution of the region.

During the 1980s and 1990s the units and terranes of the BP were carefully studied and correlated by Haydoutov et al. (1996-1997), K a r a m a t a et al. (1996-1997), Pamid & Belak (1996-1997) and Papanikolaou (1996-1997b). Papanikolaou (1996-1997a) summarized the results in I G C P Number 276. These results, together with new data obtained after 1996, now make it possible to formulate a new geological model for the formation of this region. New data obtained during the last few years from throughout the Balkan Peninsula indicate some problems with most of the existing models. The new model is based on information including inferred palaeolatitudes and palaeobotanical results, as well as other relevant geological, palaeontological, sedimentological and isotopic age data. Selected important data are set out in Tables 1-3, with key localities being shown in Figure 2. Combining all of these data allows the palaeolatitudes of units to be inferred for any given time, as discussed by Ka r a m a t a et al. (2003) for some of the western and central parts of the BP. In addition, pre-existing reconstructions, as summarized by Robertson et al. (1996), provide a regional framework for the new interpretation given here. One difficulty encountered relates to the terminology of geological units (e.g. terranes, blocks

Table 1. Palaeoinclinations, palaeofloristic and palaeoenvironmental properties o f selected geological units of the central part o f the Balkan Peninsula; Ordovician-Permian

Geological unit and age KT; Ordovician

Rock type and locality

Metasandstone metasiltstone KT; Silurian Low-grade schists SPPT; Early Devonian Siltstones KT; Early Devonian Siltstones RVT; Early Devonian Siltstones KT; Late Devonian Siltstones KT; Late Carboniferious Sandstone-siltstone; (Stephanian) Ranovac ESCB; Westphalian and Siltstone, coal-bearing; Stephanian SE and W part of ESCB DHCT; Stephanian Siltstone; Velebit JB(T); late Carboniferous Limestones JB(T); Stephanian Siltstone, limestone; Pecka, Ljubija ESCB; Permian Red beds; Ku6aj Belt JB(T); Late Permian Bituminous limestones JB(T), SUBKT, EBDT; Bellerophone limestones Late Permian

Palaeolatitude

Palaeontological & Reference; palaeoenvironmental affinity point in Fig. 2

25~

l

15~ 4~ 16~ 39~ 5~ 5~

1 2 2 2 1 1 EU flora GW flora

4~ 8~ 5-6~

3;14 3;15

4 GW flora

3;16

Warm and arid climate

1,5 4 6,7,8

NE margin of Adria

EU, Eurasian type; GW, Gondwana type; MED, Mediterranean type; other abbrevations are given in the text for Figure 1. References: 1, Krstid et al. (1996, 1996a), Milidevid (1996, 1998); 2, Krstid et al. (1996); 3, Pantid & Duli6 (1991); 4, Lazendid (2002); 5, Maslarevid & Krstid (2001); 5, Milidevi6 et al. (1995); 6, Protid et al., (2000); 7, Ramovg et al. (1984); 8, Haas et al. (1995).

TERRANE MODEL FOR THE BALKAN PENINSULA e,t

r

0

"2 2

2~ .5

"9

~

o~

"0

o~

..-.~ .,..~

9

,.s

ezO .,..~

k~

z oz

z

"toG" 0

=

.,-~

K =

. ~

o

?,

9~ . - ~

~

<m~=_ ~'N 9

6

etc.). In this paper existing names are used wherever possible but new terms have had to be introduced in some cases. The term 'terrane' is used in the sense of Keppie & Dallmeyer (1990) for a lithological assemblage that was transported by an oceanic plate until it docked with a block of continental crust and so lost its previous identity. The term 'unit' is used in a general way for any specified geological assemblage. In cases where the nature, boundaries, or the mode of transport of such an entity is not well documented the terms 'block', 'massif' or 'mass' are used in preference. Below, it will be noted at what time a 'terrane' becomes a 'unit', avoiding the name 'terrane' for some previously named crustal bodies (e.g. those termed 'Mass', 'Massif' or 'Belt'). Geological background

<

,-o

157

Figure 1 shows the geological framework of the BP, based on the work of Haydoutov et al. (1996-1997), Karamata et al. (1996-1997), Pami6 & Belak (1996-1997) and Papanikolaou (1996-1997b), with some additions by Karamata et al. (2000a, c, 2003), Kr~iutner & Krsti6 (2003), Karamata (2004) and some additional new data (see Tables 1 and 2). The following units were amalgamated to Eurasia as Tethys closed, mainly northeastwards: the Moesian plate (Moesia) (MP), to which were docked the following terranes: the Vr~ka (~ukaMiro6 terrane (VCMT/U) and the Stara PlaninaPore6 terrane (SPPT/U), both passing eastwards into the Forebalkan terrane (FOREB); the Ku6aj terrane (KT/U), passing eastwards into the Srednogora terrane (SRGT); the Homolje terrane (HT/U) continuing north of the Danube as the Sebes-Lotru or the Getic nappe; the Rhodope massif, made up of the Rila part (RHM) and the Pirin part (RHM'); the Ranovac-Vlasina terrane (or belt) (RVT/B); the Serbian-Macedonian mass (a composite terrane) (SMCT/U), amalgamated with the RVT by the Early Palaeozoic; the Circum Rhodope belt (CRHB); the Sakar terrane (SAK U). The above terranes exhibit different provenances. The VCMT, SPPT, FOREB, KT, SRGT and RVT originated from island arcs of different stages of development, or back-arc basins. The RHM, RHM', CRHB and SAK U, and probably the HT and the SMCT, represent detached fragments of continental crust. Each of these exhibits a different geological evolution (see Fig. 3; after Krsti6 & Karamata 1992; Karamata & Krsti61996; Karamata e l al. 1996-1997). After differing transport histories during the Carboniferous these terranes docked to the Moesian

158

S. K A R A M A T A

b

eo

"~ "F~

exl

o~ c~ o o

ii

0

0

0

0

z

Foz? ('4

F

o z

0"1

O'h

-~

. ,...~

o

o ~

0~

~

O0

~

~

z

z

z

~

~

~

e~

~

9.-~

8-q ~ . J

o

.~

o

oo

o,~

.o 8 .-

.~ 0~

o

o

o

....

~

~ ~

=

~ r

- ~ Z - ~

o o

.,..,

.~

,.~

>

N >

M >

0

o~

,....,

or.#a

! ~0 oo~

~

m M~Ua

m Ua

o .,..~ o o o

0

~oo

m Ua

m U~

N >

>

~

TERRANE MODEL FOR THE BALKAN PENINSULA plate to become parts of Eurasia, forming the Carpatho-Balkanides in the west and the Balkanides in the east; after the Cretaceous they then behaved as a single geological unit. The following units were incorporated into the Gondwana supercontinent as it moved gradually northwards: the Dalmatian-Herzegovina Composite terrane (DHCT) and further south the Eastern Hellenides Platform (EHP), representing the northeasternmost part of Adria and its post-Carboniferous cover; the Central Bosnian Mountain terrane (CBMT/U), which docked with the D H C T in the Carboniferous, including Mesozoic continental slope sediments deposited along its northern-northeastern margin; the East Bosnian-Durmitor terrane (EBDT/U), added to the DHCT in the Jurassic; the Dinaridic oceanic basin, with the Mirdita-Pindos ophiolitic basin as its continuation to the south, both representing a Mid-Triassic-Late Jurassic marginal sea, which has remnants in the Dinaridic (DOB) and Mirdita-Pindos (MPOB) ophiolite belts; the Drina-Ivanjica terrane (DIT/U); the K o r a b Western Macedonian terrane (K-WMT/U); the Pelagonian massif terrane, including the Olympos metamorphic rocks (PM). The CBMT, DIT, K-WMT, EBDT and the PM are all of different provenance and exhibit a different geological evolution (see Fig. 3; based on Papanikolaou 1996-1997b; Shallo in Papanikolaou 1996-1997a; Karamata & Krsti6 1996; Karamata et al. 1996-1997; Karamata & Vujnovi6 2000; Proti6 et al. 2000). These units represent different parts of a continental margin that were detached from its slope, shelf and parent continent and translated to their present positions by transcurrent faulting. The CBMT, DIT, K-WMT and the PM formed the northeastern border of Gondwana since the Carboniferous, except during the time of existence of the DOB-MPOB marginal sea. The EBDT was added to the DHCT in Mid-Late Jurassic time. After docking to Adria these units became part of Adria (Gondwana) and then behaved as a single geological entity. The Vardar Ocean was situated between Eurasian- and Gondwana-related units. It had a complex evolution leading eventually to the existing Vardar Zone (Karamata et al. 1999, 2000). The Vardar Zone is composed of the following components. A relic of the Main Vardar Ocean, the Main Vardar Belt (MVB) was inherited from an Early Palaeozoic ocean that existed between Gondwana in the south and continental masses that later became Eurasia to the north. From the Early Palaeozoic, island arcs and back-arc basins existed; these docked as terranes to the Moesian plate during the Carboniferous. The MVB closed

159

during the Late Jurassic and the Main Vardar ophiolite zone (MVZ) now represents its suture. The Veles series terrane (VS), part of a Carboniferous island arc, was transported with oceanic crust and docked to units to the east during the Late Jurassic. The Kopaonik block and ridge unit (KBRU) continuing to the south to the Paikon block, and as far north as Tissia ((~anovi6 & Kemenci 1997) represents a remnant of a section that detached from the DIT during the Late Triassic and formed a ridge separating the Main Vardar Ocean in the east from the newly formed Western Marginal basin of the Vardar Zone in the west. The Western oceanic basin of the Vardar Ocean, existing from the Late Triassic, became a wide oceanic basin during the JurassicEarly Cretaceous, and then closed by the latest Cretaceous; its suture is the Vardar Zone Western Belt (VZWB). The Sana-Una-BanijaKordun terrane (SUBKT/U) and the Jadar block terrane (JBT/U) were transported during the Early Cretaceous and incorporated into trench deposits of the Vardar Zone Western Belt. All of the terranes or units mentioned above became parts of the developing suture zone between Eurasia and Gondwana. Additional terranes and units were added after their initial amalgamation. Remnants of at least two oceanic basins, the MVZ and the VZWB are located within the suture zone, together with the VS island arc remnants. There are also the KBR unit and the JBT, and also the SUBKT terranes that were separated from a continental margin setting and later transported along transcurrent faults. All of these relics and terranes behaved as a single geological entity in the frame of the BP after the Maastrichtian. Figure 4 shows lithostratigraphic columns representing the evolution of the two oceanic realms of the Vardar Zone (the MVZ and the VZWB), as well as the DOB, for comparison (from Karamata et al. 1996-1997, 2000a, with some additions).

Cambrian to Devonian evolution Data for pre-Devonian time are very scarce and relate only to some widely separated units. Some units were located on the eastern side (i.e. present position) of Palaeotethys, or its precursor. These include several Precambrian, Cambrian and Silurian terranes now included within the Carpatho-Balkanides, Rhodopes and the SMCT. Units located along the western side (present position) of Tethys include Precambrian and Lower Palaeozoic units that now occur in the PM. In addition, Cambrian, Ordovician and Silurian formations are present in the DIT, CBMT, K-WMT, and EBDT (Fig. 3). There are many

160

S. K A R A M A T A

TERRANE MODEL FOR THE BALKAN PENINSULA differences between these units, particularly in lithology. The (meta)clastic and siliceous sediments differ, as Silurian black shales are present only in some terranes. Other differences include the timing of flysch deposition, hiatuses in sedimentation, the presence of basaltic or rhyolitic volcanism, and the time and grade of metamorphism (see Fig. 3; after Karamata & Krsti6 1996; Karamata et al. 1996-1997; Karamata & Vujnovi6 2000; Proti6 et al. 2000). Existing remnants are insufficient for a regional synthesis. For the Devonian, data are more abundant, but are still insufficient to determine the relative positions and the history of the units. However, the available results provide some regional considerations. At the beginning of the Devonian, units then forming parts of the BP can be divided into some that were oceanic (precursors of Tethys?) and later became parts of Eurasia or parts of Gondwana, and others that were parts of continental margins. The units located in the oceanic realm and at continental margins were later amalgamated to form various terranes and continental blocks. Their relative positions at this time are difficult to determine, as they were still widely separated. However, the latitudes at which they were at specific times can be determined using palaeomagnetic data. The longitude at any time is, however, unknown. During the Devonian the Moesian plate (MP) was the only part of Eurasia that later became part of the BP. The position of the Moesian microplate relative to Eurasia at that time remained ill defined. Stampfli & Borel (2004) placed Moesia, by Late Silurian time, at the southern margin of Avalonia-Baltica, one of the continental blocks that was later amalgamated with Eurasia. Robertson & Dixon (1984),

161

Stampfli et al. (1991), Neubauer & Von Raumer (1993), Robertson et al. (1996), Stampfli & Borel (2004), Muttoni et al. (2000), and some others have located the Moesian microplate with the Balkanides or the Serbo-Macedonian and Rhodope masses, effectively as part of the southern margin of the Eurasian plate during the Permian. Haydoutov & Yanev (1996), however, considered the MP as a block of Gondwanan continental crust, which was added to the East European platform in the Carboniferous, during the Variscan orogeny. The Ordovician to Early Carboniferous cover of the MP differs from the cover units of the same age now forming parts of the Carpatho-Balkanides to the west (Karamata et al. 2003), and also from the Balkanides to the south (Haydoutov & Yanev 1996). It is likely that the Moesian microplate already formed the southernmost part of the Eurasian plate before the Carboniferous. During the Mesozoic it was detached from the main continental mass along the North Dobrouga rift. During the Cenozoic the MP was separated from the main Eurasian plate and moved westwards related to the impact of the Pannonian basement from the west and the guiding effect of curvature of the Carpathian chain. The Adria microplate formed part of Gondwana that was incorporated later into the BP. However, pre-Carboniferous formations are not exposed. Most of Adria is covered by Mesozoic, mainly calcareous sedimentary rocks, including carbonate platforms of the D H C T and the EHP in the SW of the BP. The CBMT, K-WMT, DIT, JBT and SUBKT probably represent former marginal parts of Adria, or Gondwana that become separated. These include Cambrian or Devonian to Lower Carboniferous,

Fig. 1. Geological units of the Balkan Peninsula (after Haydoutov et al. 1996-1997, Karamata et al. 1996-1997 and Papanikolaou 1996-1997b, with modifications by S. Karamata). Inset shows the position of the Balkan Peninsula in the NE Mediterranean. CBMT/U, Central Bosnian Mountains terrane with continental slope deposits, from the Cretaceous a discrete unit; CRHB, Circum Rhodope belt; DHCT/U, DalmatianHercegovinan Composite terrane with its post-Carboniferous cover; DIT/U, Drina-Ivanjica terrane, from the Cretaceous a discrete unit; DOB, Dinaridic oceanic basin, after the Jurassic an ophiolitic belt; EBDT/U, East Bosnian-Durmitor terrane, from the Cretaceous a discrete unit; EHP, Eastern Hellenide Platform; FOREB, the Forebalkan unit; HT/U, Homolje terrane, from the Cretaceous a discrete unit; JBT/U, Jadar block terrane, from the Cretaceous a discrete unit; KBRU, Kopaonik block and ridge unit; KT/U, Kuraj terrane, from the Cretaceous a discrete unit; K-WMT/U, Korab-Western Macedonian terrane, from the Cretaceous a discrete unit; MD, Metohija Depression; MP, Moesian plate; MPOB, Mirdita-Pindos oceanic basin, after the Jurassic an ophiolitic belt; MVB/Z, main basin of the Vardar Ocean, from the Jurassic the main belt of the Vardar Zone; OPVD, Ov~e Polje-Vardar depression; PM, Pelagonian massif; RHM, Rhodope massif, Rila part; RHM', Rhodope massif, Pirin part; RVT/B, Ranovac-Vlasina terrane, from the Cretaceous a specific unit/belt; SAK U, Sakar unit; SMCT/U, Serbian-Macedonian Composite terrane, from the Cretaceous a specific unit; SPPT/U, Stara Planina-Pore6 terrane, from the Cretaceous a discrete unit; SRGT/U, Srednogora terrane, from the Cretaceous a discrete unit; SUBKT/U, Sana-Una-Banija-Kordun terrane, from the Cretaceous a discrete unit; VCMT/U, Vr~ka Cuka-Miro6 terrane, from the Cretaceous one unit; VS, Veles series terrane; VZWB, Vardar Zone western oceanic basin, from the Maastrichtian an ophiolitic belt.

162

S. K A R A M A T A

BLACK SEA

C'q

I

! oq |

I--"

0

TERRANE MODEL FOR THE BALKAN PENINSULA mainly terrigenous deposits, and granodiorite bodies, rhyolite and basalts (Fig. 3). Within the oceanic realm (MVZ in Fig. 4) several terranes were located between the Moesian and the Adria microplates as oceanic island arcs and back-arc basins. These included the SPPT! FOREBU, KT/SRGT and the RVT. There were also fragments of continental crust, i.e. the R H M , RHM', CRHB, SAK U, SMCT, and probably the HT and the PM. These docked to Eurasia during the Late Palaeozoic except for the PM. An impression of the wide dimensions of the oceanic realm between the MP and Adria, as represented by the basement of the D H C T and the EHP, can be achieved by reconstructing the positions of several terranes during the Early Palaeozoic, at a time when they remained widely separated within the oceanic realm (Fig. 5). During the Ordovician the KT, now part of the Carpatho-Balkanides of Eastern Serbia, was at a latitude of 29-25~ from there is moved northwards to 20-15~ in the Silurian, then to 16~ by the Early Devonian, 5~ by the Late Devonian,

163

5~ in the Carboniferous and then to 8~ in the Permian (Fig. 5; data from Krstid et al. 1996a, b; Milidevi6 1996, 1998). These results imply that steady northward motion took place, until after the amalgamation of units and the formation of the East Serbian Carpatho-Balkanides during the Mid-Carboniferous. During this motion only a moderate amount of rotation, mainly clockwise (up to 20-30~ took place. Other terranes now adjacent to the Ku6aj terrane experienced a similar history (Krsti6 et al. 1996a). During the Early Devonian, when the Ku6aj terrane was at 16~ the Stara Planina-Pore6 terrane (SPPT), now further east, was situated at 4~ whereas the Ranovac-Vlasina terrane (RVT), now to the west, was situated at 39~ (Fig. 5; data from Krstid et al. 1996a). The distances between the RVT, the KT and the SPPT corresponded to at least 35 ~ of latitude in the Early Devonian; that is, c. 4000 km of north-south oceanic separation between Eurasia and Gondwana at that time. The terranes discussed above were transported mainly northwards or northeastwards during the Early Palaeozoic and later docked to

Fig. 2. Positions of the characteristic localities mentioned in the text within the geological framework of the Balkan Peninsula. 1, granite of Vlajna (S); 2, ultramafic rocks of Plav6evo village and the Vitovni6ka river, south of Ku6evo (S); 3, talcized serpentinite in the Majdanpek mine open pit (S); 4, serpentinites, Juti (Ro); 5, gabbro, Donji Milanovac (S); 6, gabbro massif of Dell Jovan (S); 7, gabbro massif of Zaglavak (S); 8, gabbro to basalts of Cerni Vra6-Vrach (Bg); 9, serpentinites close to Stubik (S); 10, ultramafic rocks south of Blagoevgrad (Bg); 11 and 12, meta-ophiolite rocks (ultramafic rocks, metagabbros and metabasalts) in the belt Avren-Brusevci (Bg); 13, granite of Ziman, Neresnica (S); 14, Westphalian and Stephanian fora of Eurasian affinity, Ranovac (S); 15, Stephanian flora of Gondwanan-type, Southern Velebit Mt. (Cr); 16, Stephanian flora of Gondwanan-type, Pecka River, Jadar block (S); 17, Anisian palaeolatitudes, Belogradchik (Bg); 18, Late Scythian to Carnian rhyolitic to andesitic and basaltic lavas and volcaniclastic rocks of calc-alkaline affinity, northern Montenegro; 19, Krivaja-Konjuh amphibolites, Vijaka (Bill); 20, amphibolites of Bistrica, Priboj (S); 21, Zlatibor (S); 22, metamorphic sole of the Brezovica ultramafic rocks (S); 23, metamorphic sole of the Mirdita ultramafic rocks (Alb); 24, metamorphic sole of the Bulqize ultramafic rocks (Alb); 25, metamorphic sole of the Shpati and Shebeniku ultramafic rocks (Alb); 26, metamorphic sole of the Boboshtica ultramafic rocks (Alb); 27, metamorphic sole of the Pindos ultramafic rocks (Gr); 28, metamorphic sole of the Vourinos ultramafic rocks; 29, palaeolafitudes at the South-SE margin of the Cukali-Krasta zone (Alb); 30, Carnian(?)-Norian continental slope formation, western flank of the Kopaonik block (S); 31, the same formations at the eastern flank of the Studenica slice (S); 32, ophiolific basaltic lavas are interlayered at the highest levels and covered by cherts of Late Carnian to Mid-Norian age; Ov6ar-Kablar gorge (S); 33, amphibolites of the metamorphic sole in the Main Vardar Zone, Razbojna (S); 34, ophiolific pillow lavas covered by Tithonian reef limestones, Demir Kapija (Mac); 35 and 36, ophiolitic members and the m61ange with Tithonian and Early Cretaceous paraflysch cover, Brus-Kur~umlija (S); 37, amphibolites of the metamorphic sole at Banjska (S); 38, amphibolites of the metamorphic sole at Troglav (S); 39, amphibolites of the metamorphic sole at Teji6i (S); 40, crossite schist, Fru~ka Gora (S); 41, 42 and 43, calc-alkaline volcanic rocks of the Southern Carpathians-Eastern Serbia-Western to Eastern Bulgaria; 44, basalt with included Campanian limestone blocks, Krupanj (S); 45, basaltic pillow lavas interlayered with Late Campanian-Early Maastrichtian limestones, Gornji Podgradci (Bill); 46, Albian-Cenomanian sedimentary rocks with Eurasian type palynomorphs, Gledi6i Mts. (S); 47, Albian-Cenomanian sedimentary rocks with palynomorphs of Gondwanan affinity, Zlatibor (S); 48 and 49, palaeolatitudes of deposition of Albian-Cenomanian sedimentary rocks covering the ophiolitic rocks of the Main Vardar Zone, Rudnik to Kragujevac and Brus to Kur~umlija (S); 50, palaeolatitudes of Barremian-Cenomanian sedimentary rocks, Dugi Otok (Cr), 51, the Timok magmatic complex (S); 52, palaeolatitudes of Campanian-Maastrichtian sedimentary rocks, Eastern Serbia; 53, Campanian-Maastrichtian flysch, Toplica (S); 54, Campanian-Maastrichtian flysch around the Kopaonik Mt. (S); 55 and 56, Campanian-Maastrichtian flysch, Ca6ak and Jelica Mt. Countries: Alb, Albania; Bill, Bosnia and Herzegovina; Bg, Bulgaria; Cr, Croatia; Gr, Greece; Mac, Macedonia (former Yugoslavia); Ro, Romania; S, Serbia,

164

S. KARAMATA

Fig. 3. Correlation of selected Palaeozoic terranes in the Balkan Peninsula. Simplified presentation of the evolution of selected terranes that originated at the continental margin of Adria (four columns to the left) and in the oceanic basin (two columns to the right). After Krsti6 & Karamata (1992) Karamata & Krsti6 (1996), Karamata et al. (1996-1997) and Karamata & Vujnovi6 (2000), simplified with some additions. CMBT, Central Bosnian Mountains terrane with continental slope deposits; EBDT, East Bosnian-Durmitor terrane; DIT, Drina-Ivanjica terrane; JBT, Jadar Block terrane; KT, Ku6aj terrane; SPPT, Stara Planina-Pore6 terrane. 1, Continental clastic deposits; 2, red beds; 3, lagoonal facies; 4, psammites; 5, pelites; 6, cherts; 7, pelagic carbonates; 8, shallow-water carbonates; 9, rhyolites; 10, basaltic rocks; 11, intermediate-composition volcanic rocks; 12, granites to quartz diorites; 13, ophiolitic rocks; BS, black shales; F, turbiditic deposits; D, dolomite; MET, periods of metamorphism.

TERRANE MODEL FOR THE BALKAN PENINSULA A G VZWB

DOB-MPOB Pg

E

DOB-MPOB

MVZ

TF "NCH ASSEMBI..~ GES VZWB MVZ

FLYSCH Rudist limestone "~ ----~ 9 C('--- _ J BasaltsS=0Ma o>

Cr~

OLISTOSTROME Greywacke Radiolarite T2-J~ Basalt of MORB and IA type Gabbro Albite granite Ultramafic lenses Limestones T3-Cr2

Crossite schist Crl

123 Ma

I

POGARI SERIES = C ~z-~-

Metamorphic sole 157-146 Ma

L i~

PARAFLYSCH Reeflimestone OLISTOSTROME Greywacke Radiolarite T2-J~ Basalt at N: MORB, at S: MORB and IA type Gabbro Albite granite Palaeozoic granites Ultramafic lenses Limestones T & J

'9Metamorphic' sole 179-157 Ma

-

T3

T~

R

4~--- R .---~,

2. f

Veles seriesisland arc

D

Transport i of terranes to N-NE |

S O

MATRIX Siltstone

MATRIX Argillaceous-silty

t u

Cm

Time scale is not linear

VELES SERIES Low to medium grade metamorphism

ULTRAMAFIC SLABS In DOB Iherzolite, in MPOB Iherzolites and harzburgites obducted with metamorphic sole or intruded with metamorphic aureole

P C

OLISTOSTROME !ULTRAMAFICSLABS Greywacke Lherzolites obducted Radio rite , with metamorphic Basalt of IA type sole Gabbro Ultramafic lenses MATRIX Limestones (dark, Argillaceous-silty white)

OLISTOSPLAKAE Limestone T to J?

r

165

4---- R . ~

Opening

~

C~

Closing

~

Subduction

Fig. 4. Correlation of oceanic realms and ophiolitic belts within the Balkan Peninsula. After Karamata et al. (1990-1997, 2000a), with additions. DOB-MPOB, Dinaridic and Mirdita-Pindos oceanic basin, later becoming the Jurassic ophiolitic belt; VZWB, Vardar Zone Western oceanic basin, from the Maastrichtian an ophiolitic belt; MVZ, Main oceanic basin of Tethys, later the Vardar Ocean and after the Jurassic the main Vardar Zone ophiolitic belt. Pg, Palaeogene; Cr2, Late Cretaceous; Cr], Early Cretaceous, J3, Late Jurassic; J2, Mid-Jurassic; J~, Early Jurassic; T3, Late Triassic; Tz, Mid-Triassic; T1, Early Triassic; P, Permian; C, Carboniferous; D, Devonian; S, Silurian; O, Ordovician; Cm, Cambrian.

166

S. KARAMATA

Fig. 5. Positions of the Stara Planina-Pore6 terrane (above), the Ku6aj terrane (centre) and the Ranovac-Vlasina terrane (below) in the Early Devonian and the positions of the Ku6aj terrane during the Ordovician (1), Silurian (2), Devonian (3), Carboniferous (4) and Permian (5) related to the Equator (after Krsti6 et al. 1966a).

the Moesian plate. The SMCT was transported and amalgamated to the RVT during the Ordovician, as shown by the 500 Ma stitching granites of Vlajna (Locality 1 in Fig. 2). These terranes continued northward into the Southern Carpathians and southeastwards and eastwards into the Balkanides. The VCMT, SPPT, KT, HT, RVT and the SMCT continued into their respective units in Romania. The VCMT and SPPT continued eastwards to the FOREB; the KT continued to the SRGU, and the RVT continued to the Kraishtides. However, the SMCT, as well as the Rhodope units (RHM and RHM'), the Circum Rhodope belt (CRHB) and probably the Sakar unit (SAK U) did not continue further east, although similar fragments of microcontinental units exist there.

Carboniferous-Permian setting The docking of the terranes of the East Serbian Carpatho-Balkanides, the Balkanides in Bulgaria, and the blocks and belts of metamorphic rocks (SMCT, RHM, RHM', CRH B, SAK U) to the Moesian microplate took place during the Carboniferous, but when exactly this took place remains unclear. From the end of the

Early Carboniferous this amalgamated unit formed the margin of Eurasia, or the East Serbian Carpatho-Balkanides (ESCB) and the Balkanides in Bulgaria (BB). Its northwestern parts were situated at about 5~ (Milidevi6, 1996). Relics of oceanic crust are preserved as small bodies in the following settings: (1) as ultramafic rocks between Plav6evo village and the Vitovni6ka river, south of Ku6evo between the RVT and the KT (Locality 2 in Fig. 2); (2) as minute lenses of talcized serpentinite in the Majdanpek open pit along the boundary zone between the KT and the SPPT (Locality 3 in Fig. 2); (3) along the contact between the SPPT and VCMT as large bodies within the ophiolite belt known as Juti-Donji Milanovac-Deli Jovan-Zaglavakt~erni Vra6/Vrach (Localities 4-8 in Fig. 2); (4) between the VCMT and the Moesian plate close to Stubik (Locality 9 in Fig. 2) as small lenses of serpentinite; (5) as minute relics of ultramafic rocks along the boundary of the RVT and the SMCT; (6) as small bodies of ultramafic rocks (Locality 10 in Fig. 2) along the contact zone between the Rhodope units and the Ogra~den unit (i.e. southern part of the SMM); (7) as ophiolitic rocks (ultramafic rocks, gabbros and basalts) occurring along the boundary of the CRHB and the SAK U (localities 11 and 12 in Fig. 2), which were metamorphosed to a high grade together with associated sedimentary rocks (Haydoutov et al. 2004). The S-type granitic rocks of Ziman (Locality 13; Fig. 2) are assigned to a Late Carboniferous age. This is based on a 295 Ma Rb/Sr age on muscovite (Deleon 1969), although U/Pb ages on zircons as old as 434+ 35 Ma have been determined (Grunenfelder, pers. Comm.). The S-type granitic rocks were succeeded by large masses of I-type granites of Early Permian age, based on 275-256 Ma Rb/Sr data on biotite (Deleon 1969); there are also andesites, trachytes and rhyolites of Permian age. This magmatic activity relates to collisional processes and is found within and along the boundaries of all the terranes amalgamated to the Moesian plate. Permian red beds and Mesozoic sedimentary deposits cover the older formations of Adria and related units (Fig 3); for this reason it is not possible to determine their position during the Carboniferous. During the Early(?) Carboniferous the CBMT docked to the margin of Adria, followed later by the DIT and K-WMT. The PB was probably amalgamated at that time too, but was later thrust over neighbouring units (the K-WMT); for this reason the time of first amalgamation with Gondwana or Gondwana-related units is difficult to determine.

TERRANE MODEL FOR THE BALKAN PENINSULA The CBMT was transported from a shelf setting characterized by shallow-water Devonian deposits (Fig. 3) and was then located at the margin of the Adria microplate. The DIT is believed to have moved along a transcurent fault to its present position from a previous position associated with the Palaeozoic complexes of the southern part of the Massif Central, France (Dimitrijevid & Djokovid 1981). The K-WMT includes continental margin deposits and was probably transported from the SE. A wide oceanic realm, (Palaeo-)Tethys, or the Main basin of the Vardar Ocean (see Fig. 4) existed between Eurasia (i.e. the Moesian plate and attached terranes) and Gondwana (i.e. Adria and the attached units). The existence of this Palaeo-Tethyan realm, already mooted by Seyfert & Sirkin (1973), was later confirmed by numerous workers; its presence is also indicated by the existence of the Veles Series block (an island arc relic) of Carboniferous age (Grubid & Ercegovac 2002), which now forms a lens-like body within the ophiolite belt of the sutured Main Vardar Ocean. The contrasting Late Carboniferous flora within deposits of units related to both the Gondwanan and Eurasian units confirms the existence of a wide ocean. The ESCB Stephanian coal-beating lacustrine sediments of Ranovac (Locality 14; Fig. 2) that were deposited around the Equator (up to 5~ Milidevid 1998) contain Late Carboniferous (Westphalian and Stephanian) flora, which 'existed in swamps at the southern margins of Europe' (Pantid & Dulid 1991, translated by S.K.), whereas the deposits at Velebit in the DHCT (Locality 15; Fig. 2) and in the JBT (Locality 16; Fig. 2) display a Gondwana-type Stephanian flora (Pantid & Dulid 1991). During the Permian, units related to Eurasia and those related to Gondwana were still far apart; these deposits originated under different climatic conditions and contain different fauna. The Permian red beds were formed in the ESCB under equatorial semi-arid to arid conditions (Maslarevid & Krstid 2001). The northern parts of the NNW-SSE-oriented Ku6aj unit were located at a latitude around 8~ i.e. within the equatorial zone (Milidevid 1998). Along the northeastern margin of Adria (i.e. in the DHCT and EBDT), as well as in the JBT and the SUBKT, bituminous limestones (Ramov~ et al. 1984; Protid et al. 2000) of 'Bellerophon-type', characteristic of the southeastern Alps (Haas et al. 1995), were deposited during the Late Permian. Considering the distribution of the 'Bellerophon-type' limestones, the SUBKT and the JBT originated together, in an undefined part

167

of the southern or western margins of Tethys. During the Late Carboniferous the JBT was at 4~ (determined from Moscovian-Kasimovian limestones) and then moved northwards to 56~ in the Permian (Lazendid 2002). The JBT and the SUBKT were together detached from the margin of Gondwana but later separated during their transport and were incorporated into Cretaceous oceanic trench deposits of the VZWB. The inferred positions and relationships of units during the Late Carboniferous and the Permian are shown in Figure 6.

Triassic-Jurassic setting During the Triassic and Jurassic, oceanic lithosphere of the Vardar Ocean was first subducted south(west)-wards, then eastwards, resulting in significant geological effects in Adria, as well as in the Vardar Ocean. The geological evolution of the BP during the Mid-Late Triassic and the Mid-Late Jurassic is shown in Figures 7 and 8. The Triassic was a period of quiescence of the ESCB and also in the southern realm of the MP, as these units then formed parts of the passive eastern margin of the Vardar Ocean. During the Early Triassic continental red bed sedimentation persisted in this area, grading into shallowmarine deposits. The Mid- and Late Triassic are mainly represented by limestones that are similar to those of the Northern Alps and Central Europe. Some uplift and subsidence of blocks occurred but only in the units amalgamated to the MP. The western parts of the FOREB unit, close to the SPPU and to the south of the MP (Locality 17 in Fig. 2) were situated at 21-24~ during the Anisian (Muttoni et al. 2000). Even during the Jurassic, when the Vardar Ocean was subducting beneath this part of Eurasia, its effects on the inner eastern parts of the upper continental slab were negligible. During the Jurassic mainly shallow-water to pelagic calcareous sediments accumulated. Tectonic conditions were characterized by local subsidence and uplift of the basement (Tchoumachenko, pers. comm.). The Triassic and Jurassic cover of the SMCT, the Rhodope massif (RHM' and RHM) and the CRHB (i.e. frontal part of Eurasia) is absent, or represented by local remnants or by tectonic slices of undefined age of emplacement. These units were uplifted along the margins of the Eurasian continent during the Jurassic and thus lack sedimentation. Subduction beneath them during the Jurassic seemingly left little trace. During the Triassic and Jurassic significant geological events took place at the margin of Adria and within the Vardar Ocean. Already

168

S. KARAMATA

Fig. 6. The position of geological units during the Late Carboniferous-Permian, after the construction of the two main entities, but before the beginning of the separation of units within the Vardar Ocean. The abbreviations are as in Figure 1. Arrows indicate the direction of subduction of oceanic lithosphere.

Fig. 7. The position of geological units during the Mid-Late Triassic after the start of the separation of units within the Vardar Ocean. Only the latitudes determined are given. MVO indicates the main basin of the Vardar Ocean; other abbreviations are as in Figure 1; the arrows indicate the direction of subduction of oceanic lithosphere. by the Late Permian, and continuing into the Early Triassic, uplift was accompanied by arching along the northern border of Adria (DHCT and CBMU). This uplift was probably related to the southwestward subduction of the Vardar Ocean; this caused the deposition of shallowwater to continental sediments, dolomites,

lagoonal limestones, gypsum, etc. throughout this area. Subduction is also indicated by huge masses of Upper Scythian to Carnian rhyolitic to andesitic and basaltic lavas and volcaniclastic rocks of calc-alkaline affinity within the exotic EBDT (Locality 18; Fig. 2). These magmatic rocks were brought from the south in the Jurassic

TERRANE MODEL FOR THE BALKAN PENINSULA

169

Fig. 8. The position of geological units during the Mid-Late Jurassic, the time of the final closure of the Dinaridic oceanic basin and the main basin of the Vardar Ocean. MVO indicates the main basin of the Vardar ocean; other abbreviations are as in Figure 1. The arrows indicate the direction of subduction of oceanic lithosphere; the thin arrows show the transport direction of some terranes. Only the latitudes determined are given.

(in present coordinates). This subduction is also indicated by the geochemical evidence from the rift-related Middle Triassic volcanic rocks located along the northeastern margin of Adria (Kne2evid & Cvetkovid 2000). The Lower and Middle Triassic successions in the DHCT, DIT and the margins of the CBMU are similar to those of the Southern Alps. The characteristic feature is (Anisian)-Ladinian volcanism developed along subparallel rift-related faults. The volcanic members are of 'within-plate' character (Karamata et al. 2000b), but in some areas, including some fault zones, a subduction signature (Kne~evi6 & Cvetkovid 2000) is seen, especially within basalts closest to the eastern margin (Memovid et al. 2004). It is important to note that the Upper Scythian to Carnian calcalkaline volcanic rocks of the exotic EBDT brought in from the south during the Jurassic did not originate there. The Upper Triassic cover of Adria was characterized by the deposition of shallow-marine limestones, which continued into the Jurassic. Rifting began at the end of the Mid-Triassic (Figs 4 and 7) between the DHCU, the CBMU and the EHP on one side, and the DIU and the K-WMU on the other side. This was followed by formation of a marginal sea during the Late Triassic-Early Jurassic, which gave rise to the Dinaridic ophiolite and its continuation as the wide Mirdita-Pindos ophiolitic basin. Their

boundary is a transform fault now covered by the Metohija depression (MD in Fig. 1). During the Mid-Jurassic the DinaridicMirdita-Pindos oceanic basin (DMPOB) began to close (Figs 4 and 8), with tectonic inversion of the previous extensional regime within the oceanic realm, as indicated by the obduction of ultramafic or ophiolitic units over parts of the oceanic crust sited away from the oceanic ridge, or over the early parts of the oceanic trench assemblage. The age of metamorphism of the metamorphic sole beneath the obducted ultramafic slices in this belt was determined as 179-157 Ma; that is, 174-157 Ma by K/Ar and 174-162 Ma by the Ar/Ar method (e.g. Lanphere et al. 1975; Karamata & Lovrid 1978; Roddick et al. 1979; Spray & Roddick 1980; B6bien et al. 2000; Dimo-Lahitte et al. 2001). Both methods yielded very similar results and thus may be taken together. A younging in age from south to north was suggested by Dimo-Lahitte et al. (2001) for the amphibolites in Albania but such a tendency cannot be followed northwards into Serbia and Bosnia, or southwards into the western ophiolite belt of Greece. When interpreting the radiometric dating results it should be noted that different minerals from adjacent rocks can give differences in age of up to 10-12 Ma (Karamata & Lovrid 1978; Okrusch et al. 1978). Also, the obduction is likely to have been diachronous along the zone

170

S. KARAMATA

of emplacement and thus differing ages are to be expected. A K/Ar age of 157 Ma was obtained from pargasite within corundum-plagioclasepargasite amphibolites beneath the KrivajaKonjuh ultramafic massif in Bosnia (Locality 19; Fig. 2); by contrast, similar lithologies from Bistrica, Western Serbia (Locality 20; Fig. 2) yielded an age of 170 Ma (Lanphere et al. 1975). Amphiboles from the garnet-plagioclase amphibolite from Bistrica, exposed close to the dated sample mentioned above, yielded an age of 168-174Ma (Lanphere et al. 1975). Neighbouring amphibolites beneath the Zlatibor massif, western Serbia, gave a K/Ar age of 160Ma (Locality 21; Fig. 2; Karamata & Popevi6, unpubl, data). Further south, in the Brezovica area of southern Serbia (Locality 22; Fig. 2), hornblendes from amphibolites were dated at 171-176 Ma, but micas from associated metasediments yielded K/Ar ages of 159-168 Ma (Karamata & Lovrid 1978). Dimo-Lahitte et al. (2001) determined Ar/Ar ages for amphibolites from the metamorphic soles beneath ultramafic masses throughout Albania, ranging from the Mirdita area in the north (Locality 23; Fig. 2), and including both the eastern and the western belts in the central parts (i.e. Lura and Bulqiza massifs; Locality 24; Fig. 2), to the southern parts (i.e. Shpati and Shebeniku massifs; Locality 25; Fig. 2), and the southernmost parts (i.e. Boboshtica massif; Locality 26; Fig. 2) of the country. These results range from 162 to 168 Ma in the north and in the south, but from 170 to 174 Ma in the Bulqiza area. Further south the ages of hornblende from amphibolites of the metamorphic soles of the Pindos ultramafic massif (Locality 27; Fig. 2) gave a K/Ar age of 172 Ma (Thuizat et al. 1981) and Ar/Ar ages of 172-181 Ma (Roddick et al. 1979; Spray & Roddick 1980); amphibolites from the Vourinos massif (Locality 28; Fig. 2) yielded an Ar/Ar age of 179 Ma (Spray & Roddick 1980). It therefore appears that the metamorphic soles of the DMPOB originated from 181 to 157 Ma, i.e. during the Dogger, more precisely from the latest Lias to the end of the Callovian. The oceanic crust and the uppermost lithosphere were subducted northeastwards (present position) and huge masses were accreted at a subduction trench, as terrigenous-derived olistostromes, together with limestone gravity slides (from blocks moving over the subducting units) and obducted (or 'downslided') mainly ultramafic ophiolitic rocks (as lenses and masses). During the formation of this trench assemblage the EBDT was probably separated from the K-WMT and later transported from the SE into it.

During the Late Jurassic the Dinaridic and the Mirdita-Pindos ophiolite basins were closing (Fig. 4) and their relics were incorporated into the margins of Adria as the Dinaridic ophiolite belt, and further south as the Mirdita-Pindos ophiolite belt (i.e. the DOB and the MPOB, respectively). After closure, these relics were covered by transgressive shallow-water marine deposits (i.e. 'Pogari series') of TithonianValanginian age. Of interest is the Ladinian-Late Triassic to Tithonian position of the Cukali-Krasta zone, an aborted rift in the eastern part of Adria (Locality 29; Fig. 2; Mauritsch et al. 1996). This area was situated at appoximately the same latitude, i.e. about 13~ This could indicate that the opening of the Dinaridic-Mirdita-Pindos ophiolitic basin and its later closure did not affect the position of the inner parts of Adria. It may also indicate, however, that the widening of this basin hindered movement of parts of Adria towards the north, and that later, during closure of this basin, the spreading of the Vardar Zone Western Basin prevented such movements. At the end of Jurassic, after the suturing of the Dinaridic-MirditaPindos ophiolite basin general northward movement of these parts of Adria could resume. During the Triassic the Vardar Ocean was subducting southwestwards, causing opening of the Dinaridic-Mirdita-Pindos marginal sea during the late Mid-Triassic. A new oceanic realm originated (Fig. 4) during the Carnian(?)-Norian behind the Kopaonik Block and Ridge unit (KBR), which was detached from the eastern parts of the DIU. This basin probably opened during the Early Norian. Identical continental slope units mainly composed of fine-grained and rare coarser-grained terrigenous rocks, with some limestone intercalations and basaltic lava flows, occur on both sides of the basin (Localities 29 and 30; Fig. 2); these are now exposed along the western flank of the Kopaonik Block and the margin of the DIT, specifically, the eastern flank of the Studenica slice. In the north (OvcarKablar gorge; Locality 31; Fig. 2), the uppermost levels of ophiolitic basaltic lavas are interlayered with and covered by cherts of Late Carnian to Mid-Norian age (Obradovi6 & Gori6an 1988). This basin expanded into the western basin of the Vardar Ocean, and during the Jurassic and later became the main oceanic realm of the Vardar Ocean (i.e. Neotethys). During the Jurassic the direction of subduction of the Main Vardar oceanic basin was eastwards, beneath Eurasia. Simultaneously, the main basin of the Vardar Ocean began to close (Fig. 4). The first data indicating closure are the 182-187 Ma K/Ar ages of the amphiboles

TERRANE MODEL FOR THE BALKAN PENINSULA (Karamata & Popevid; unpubl, data) from the metamorphic sole of the ultramafic rocks near Razbojna (Central Serbia; Locality 33; Fig. 2). At the end of the Jurassic this basin closed (Fig. 4); its relics, the ophiolitic pillow lavas of the Gevgeli gabbro~tiabase-spilite massif were by then covered by transgressive conglomerates grading into Tithonian reef limestones (i.e. near Demir Kapija; Locality 34; Fig. 2; Hristov et al. 1965). Also, ophiolitic units and olistostrome m61ange were covered by Tithonian reef limestones and a flysch-like Lower Cretaceous formation (i.e. from Kragujevac to Kurgumlija (Localities 35 and 36; Fig. 2; Dimitrijevi6 1997). After the Jurassic, relics of the Main Vardar oceanic basin, the MVZ, together with the Kopaonik Block and Ridge unit (KBRU), formed the frontal part of Eurasia. During the Jurassic Palaeotethys coexisted with Neotethys, the former becoming narrower and the later wider with time. Within the Western basin of the Vardar Ocean, representing the precursor of the Vardar Zone western belt (VZWB), deep-water cherts and shales were deposited over basalts of the ophiolitic association from the Late Triassic to the Kimmeridgian (Obradovid & Gori6an 1988; Gori6an et al. 1999). Large masses of trench deposits, represented by olistostrome m61ange and gravity slides from the oceanic crust and the continental margin, accumulated within this basin from Mid-Jurassic time.

171

The above data indicate that during the Jurassic, Eurasia- and Gondwana-(Adria-)related units were widely separated and experienced different geographical and geotectonic histories. The Vardar Ocean represented a large oceanic realm between these units. The width of this ocean is demonstrated by clear differences between the Early and the Mid-Jurassic brachiopod fauna in the ESCB and the southern part of the DHCT, as observed by Radulovi6 (1995).

Cretaceous (to Maastrichtian) setting At the beginning of the Cretaceous the Main Vardar Ocean and the Dinaridic marginal sea were already closed. Three main units then existed until the Maastrichtian (Fig. 9): (1) Eurasia, composed of the Moesian microplate (MP) with, to the south and west, amalgamated terranes or units; i.e. the ESCB continuing to the FOREB and the SRGT; also the SAK U, RHM, RHM', CRHB and SMCT, and in the SW the added relics of the MVZ and the KBR; (2) Adria and its docked terranes; i.e. the CBMT, relics of the DOB, including the EBDT, relics of the MPOB; also the DIT, K-WMT and the PM; (3) the Western oceanic basin of the Vardar Zone between the two above assemblages. During the Early Cretaceous sedimentary formations that were deposited on the Eurasian unit were variable, ranging from terrestrial, to

Fig. 9. The position of geological units during mid-Cretaceous time during the late phase of existence of the western oceanic basin of the Vardar Ocean. The abbreviations are as in Figures 1, 6 and 8. Only the latitudes determined are given.

172

S. KARAMATA

shallow marine to pelagic, locally (Haydoutov et al. 1996-1997; Karamata et al. 1996-1997;

Dimitrijevid 1997). During the Cenomanian, psammites and pelites with basaltic to andesitic or trachytic volcanism formed in a narrow trough in the east, located close to the boundary of the Moesian plate and the ESCB (Karamata et al. 1996-1997). During the Late Cretaceous intense magmatic activity occurred in this region related to subduction of the western belt of the Vardar Ocean (Fig. 4). This lasted from the Turonian to the Palaeocene or even to the Eocene. This magmatic-volcanic arc-type activity forms a belt extending from the Southern Carpathians through Eastern Serbia and Western Bulgaria to Eastern Bulgaria; it may then continue in the Pontides beyond the Black Sea. The igneous rocks are mainly andesites, with subordinate dacites, grading into trachytes and related intrusive rocks, all of calc-alkaline type, plus local occurrences of magmatic rocks with an alkaline affinity (Karamata et al. 1997; Berza et al. 1998; Popov et al. 2000; Banjegevid et al. 2002; Ciobanu et al. 2002; Karamata et al. 2002). Adria and its attached units were situated further SW and represent the passive margin of the Western oceanic basin of the Vardar Zone. During the Cretaceous, areas in the SW, furthest from the edge of the continental units (i.e. the cover of the Adria), were characterized by shallow-water carbonate deposition, subject to hiatuses. Closer to the margin of the continental area (i.e. towards the Vardar Zone Western basin) relative uplift and subsidence of the basement gave rise to turbiditic basins, emergent areas and areas of quiet sedimentation (Fig. 3). Between these two continental entities was the Western oceanic basin of the Vardar Ocean (Neotethys), which was at that time the main oceanic area of Tethys in this area, as well as in areas further east. During the Cretaceous its development was characterized by long-lasting subduction (Fig. 4). The beginning of compression within this oceanic realm is reflected in K/Ar ages of 157(?) to 146 Ma (Karamata & Popevid, unpubl, data) of amphiboles from the amphibolites of metamorphic soles beneath the ultramafic rocks, at Banjska (154 Ma; Locality 37; Fig. 2), Troglav (146 Ma; Locality 38; Fig. 2) and Tejidi (157 Ma, Locality 39; Fig. 2) in Central and Western Serbia. A continuation of subduction is suggested by the 123 Ma K/Ar age of subduction-related crossite schists at Frugka Gora Mr., Northern Serbia (Locality 40; Fig. 2; Milovanovid et al. 1995), and also by the supra-subduction zone (SSZ) affinity of some basaltic pillow lavas (Robertson & Karamata

1994), the formation of intra-oceanic island arcs, and by Turonian to Eocene subduction-related calc-alkaline magmatism in the Southern Carpathians, Eastern Serbia and in the Srednogorje of Bulgaria (Localities 41-43; Fig. 2). A Sm-Nd isochron age of 136 + 15 Ma was obtained for ultramafic massifs south of the DOB-VZWB boundary (Lugovi et al. 1991). In the absence of precise information on sample locality, it is necessary to check whether the samples analysed come from only slightly displaced massifs of the DOB or from ultramafic rocks overthrust as much as some tens of kilometres from the VZWB. The youngest basaltic rocks, members of the ophiolite complexes, of this unit (Fig. 4) include Campanian limestone blocks from near Krupanj, Western Serbia (Filipovid, unpubl. data; Locality 44; Fig. 2); also, basaltic pillow lavas are interlayered with Upper CampanianLower Maastrichtian sandy limestones (dated by Sladid-Trifunovid; in Karamata et al. 2000) from near Gornji Podgradci in northwestern Bosnia (Locality 45; Fig. 2). The diabase of the sheeted dyke unit below these basalts is dated (K/Ar age) at 80 Ma, at Podgradci (Karamata et al. 2000). The first sequences that cover the trench deposits of the VZWB are rudist limestones grading into Upper Maastrichtian to Eocene flysch (Fig. 4). Accordingly, during the Cretaceous to the (Late) Maastrichtian, oceanic crust of the western branch of the Vardar Ocean continued to form at the same time as it was subducted. The character of palynomorphs in the sedimentary rocks of Albian-Cenomanian age deposited on both sides of the oceanic basin proves the width of the oceanic realm. The sedimentary rocks from the Gledidi Mts. (Central Serbia; Locality 46; Fig. 2), at that time a marginal part of Eurasia, contain Eurasian-type palynomorphs, but the palynomorphs in sedimentary rocks of the same age at Zlatibor (Western Serbia; Locality 47; Fig. 2), at that time a marginal part of Adria, are of Gondwanan affinity (Dulid 1999). Two types of pollen, which can be transported great distances, do not coexist; this can be explained only by the persistence of a large oceanic area. These continental units were at that time at different latitudes. The Albian-Cenomanian sedimentary rocks covering the ophiolitic rocks of the MVZ, situated at present between Rudnik and Kragujevac and between Brus and Kurgumlija (Central Serbia; Localities 48 and 49; Fig. 2), were deposited at 23-28~ (Veljovid & Milidevid 1987). The limestones of the Dugi Otok, Croatia (Locality 50; Fig. 2) were deposited at the same time, but on the other side of the oceanic basin at 33~

TERRANE MODEL FOR THE BALKAN PENINSULA (Marton & Mili6evi6 1994). These two entities were located at palaeolatitudes of c. 25~ and 33~ respectively, confirming the presence of an intervening ocean. The position of the Dugi Otok during the Barremian-Aptian, at 30~ compared with its position at 33~ during the Cenomanian (Marton & Milidevi6 1994) implies a steady northwards movement. The evolution of the BP during the midCretaceous (to the Maastrichtian) is shown in Figure 9.

Maastrichtian-Cenozoic setting The Western oceanic basin of the Vardar Ocean (Neotethys) was closed by the end of the Maastrichtian. Adria, with its already amalgamated units (the Dinarides sensn lato), collided with the Moesian plate and docked terranes and other units (the Carpathian-Balkan system and Rhodopes encircling Moesia). Between them, there remained only the VZWB and the MVZ opholitic belts, relics of former oceanic basins and the KBR unit. The positions of the units during the Maastrichtian-Paleocene are given in Figure 10. Since the Maastrichtian all of the geological entities of the present BP have behaved as one unit. The andesitic and intrusive rocks of the Timok magmatic complex on the K U of the ESCB (Locality 51; Fig. 2) originated at about 35~ and were oriented about 20 ~ counterclockwise with respect to the position of stable

173

Europe at that time (Stefanovi6 & Veljovi6 1981). In the ESCT, MVZ and the KBRU, CampanianMaastrichtian sedimentary rocks, mainly sandstones in the ESCB (Locality 52; Fig. 2) and flysch of the cover of the MVZ and around the KBR (Localities 53 and 54; Fig. 2) were deposited at latitudes of 33-36~ (Veljovi6 & Milidevid 1986, 1987). In the VZWB and the DIT (Localities 55 and 56; Fig. 2) Campanian-Maastrichtian flysch deposits originated at latitudes of 32-36~ (Veljovi6 & Mili6evi6 1986, 1987). All of these units were located about 10~ south of their present position (Stefanovi6 & Veljovi6 1981; Veljovi6 & Milidevi6 1986, 1987). The identical depositional latitudes of the (Campanian)-Maastrichtian sedimentary rocks and of the Timok magmatic rocks, which formed in different units, confirm that the various units and entities of the BP framework had been amalgamated by the beginning of the Maastrichtian. Oligocene sediments from the cover of the VZWB, MVZ, SMM and the ESCT were deposited at latitudes close to 37-38~ (Milidevid & Djuraginovid-Gavrilovi6 1990; Marovi6 et al. 2002). After the collision and suturing of units at the end of the Cretaceous, there was a clockwise rotation, increasing from west to east: in the MVZ, SMCU and in the western parts of the ESCB the rotation was c. 5-10 ~ whereas in the eastern parts of ESCB the rotation increased to c. 15 ~ or even to c. 20 ~ (Milidevi6 & Djura~inovidGavrilovi6 1990; Marovi6 et al. 2001). Dextral

Fig. 10. The position of geological units at the end of the Maastrichtian-beginning of the Paleocene (the time after the collision of the Eurasia and Adria and the closure of all oceanic realms). Abbreviations are as in Figure 1. Only the latitudes determined are given.

174

S. KARAMATA

strike-slip movements, of up to some hundreds of kilometres, were observed along the boundaries of some of the major structural units (mostly the boundaries of former terranes). There was also an eastward bending in the northeastern parts of the area related to the formation of the Pannonian basin and the Carpathian arc, as well as to the eastward movement of the Tisia block which, in turn, influenced units further south. Until the Cretaceous the orientation of the Dinarides, the Vardar Zone and the CarpathianBalkanides was probably about WNW-ESE. The pressure exerted by the Adria microplate caused the rotation of the whole system by about 20 ~ clockwise until the Oligocene; subsequently, units rotated differently because of indentation by the wedge-like Adria promontory. In the south, the compression was very strong, and the Albanides, and further south the Hellenides, additionally rotated c. 50~ clockwise (Kissel et al. 1995); this brought the units there to an almost north-south orientation. Because of post-collisional compression of the amalgamated continental units in the southern parts of the Vardar Zone, in Macedonia, the ophiolitic belts underwent reverse faulting and squeezing out of units after the Eocene (see Brown & Robertson 2003, 2004). Further north in the Dinarides, the western region (DHCT) underwent only negligible rotation (Kissel et al. 1995). However, units further NE and to the north underwent fan-like splitting in their northerly parts and in the NW rotated to almost an east-west orientation. The driving mechanism was northward escape of the units in the east and the existence of opposing continental masses to the north (i.e. Tisia). This indentation of Adria after the Oligocene, with the escape of units northwards, and the existence of the continental masses to the east and north caused the curvature of the DIU and the VZWB. The protrusion of Adria caused the units at its NE margin to change their direction in the south, in Western Greece to Central Serbia, to almost north-south, and further north, in Western Serbia and Northern Bosnia, to almost eastwest. This compression was also the cause of overthrusting and the formation of imbricate structures within Adria-related units and the Dinarides from the Eocene to the Miocene. Further east, within the Eurasian domain, after the Eocene, units additionally rotated clockwise by about 10~ Simultaneously, boundaries between the former terranes were reactivated. Large overthrusts occurred along the western flank of the Carpathian-Balkanides between the Cretaceous and Miocene. Further east, overthrusting took place along steeper surfaces (i.e. reverse faults) with smaller displacement. The movement

of the basement of the Pannonian basin along a dextral transcurrent fault at the northern boundary of the BP caused additional clockwise rotation in the northernmost parts of this region.

Conclusions The Balkan Peninsula is now an amalgamation of units related to Eurasia and to Gondwana, with additional material derived from oceanic realms. These units and terranes evolved separately until the Maastrichtian. During the Early Palaeozoic these units were widely separated; steady northward (and northeastward) movement then carried them closer until they collided in the Maastrichtian. The stages of formation are as follows. (1) The Early Palaeozoic to Carboniferous was a period of convergence of terranes and continental units and construction of initial large units. Terranes within Tethys (i.e. the Vardar Ocean) were transported from up to 30-40~ to a near-equatorial latitude; they then docked during the Carboniferous to the Moesian microplate. Simultaneously, terranes were transported along transcurrent faults along the margin of Adria, mainly towards the SE, and were added to the Adria microplate. (2) From the Permian to the JurassicCretaceous boundary the Vardar Ocean (i.e. northwestern part of Tethys) began to close. New marginal seas formed at its SW margin during the Triassic, first the DinaridicMirdita-Pindos basin and later the Western basin of the Vardar Ocean. The first marginal oceanic basin closed during the Late Jurassic. At the end of the Jurassic the Main Vardar Ocean basin was closed related to northeastward subduction. However, during the Jurassic the Western basin of the Vardar Ocean, a former marginal sea, became the main oceanic area in this region. (3) During the Cretaceous the convergence of Adria and Eurasia continued, and a long and complex closure of the Western basin of the Vardar Ocean took place until final collision of Adria and Eurasia during the Maastrichtian. (4) From the Permian to the Recent, steady northward and northeastward movement was common to all of the fragments (i.e. terranes and continental units) making up the Balkan Peninsula; these converged until their collision. (5) During the Cenozoic all of the units exhibited an interrelated development, which finally produced the present geological

TERRANE MODEL FOR THE BALKAN PENINSULA f r a m e w o r k of the Balkan Peninsula (Fig. 1). The i n d e n t a t i o n of Adria resulted in transcurrent m o v e m e n t s along some faults or the boundaries of geological units, the rotation of units, and overthrusting as a result of 'squeezing out' of some units or parts o f them. D. Stefanovid's help with interpreting palaeomagnetic data is gratefully acknowledged. The help and suggestions of M. Sudar, V. Cvetkovid, K. Sari6 and D. Milovanovid are also acknowledged, as is D. Milovanovid for his help with the preparation of the computer-drafted figures. To A. H. F. Robertson is expressed our gratitude for suggestions and critical reading. Gratitude is also due for critical comments by two anonymous reviewers. This work was supported by the Serbian Academy of Sciences and Arts, grant GEODYNAMICS.

References AUBOUIN, J. 1973. Des tectoniques superpos6es et de leur signification par rapport aux mod61es g6ophysiques: l'exemple des Dinarides: pal6otectonique, tectonique, tarditectonique, n6otectonique. Bulletin de la Socikt~ Gkologique de France, 7, XV5-6, 426-460. AUBOUIN, J., BONNEAU, M., CELET, P., et al. 1970a. Contribution /t la g6ologie des Hellenides: Le Gavrovo, le Pinde et la Zone ophiolitique subpOlagonien. Annales de la Societk G6ologique du Nord, 90, 277-306. AUBOU1N, J., BLANCHET, R., CADET, J. F. P., et al. 1970b. Essai sur la g6ologie des Dinarides. Bulletin de la Sociktk Gdologique de France, 7, XlI(6), 10601095. BANJESEVIC, M., CVETKOVIC, V., KOZELJ, D., PEYTCHEVA,I. & VON QUADT, A. 2002. The Timok Magmatic Complex and Ridan-Krepoljin Zone: geodynamical evolution. In: JELENKOVId, R. & KOZELJ, D. (eds) International Symposium, Geology and Metallogeny of Copper and Gold Deposits in the Bor Metallogenic Zone--Bor 100 Years, Special Issue, Bor Lake, Yugoslavia. Copper institute Bor, 199-202. BI~BIEN, J., DIMO-LAHITTE, A., VERGI~LY, P., INSERGUEIX-FILIPPI, D. & DUPEYRAT, L. 2000. Albanian ophiolites. I--Magrnatic and metamorphic processes associated with the initiation of a subduction. Ofioliti, 25(1), 39-45. BERZA, T., CONSTANTINESCU,E. & SERBAN-NICOLAE, V. 1998. Upper Cretaceous magmatic series and associated mineralisation in the CarpathianBalkan orogen. Resource Geology, 48(4), 291-306. BON~EV, E. 1955. [Geology of Bulgaria]. Tehnika, Sofia (in Bulgarian). BORTOLOTTI, V., CHIARI, M., MACUCCI, M., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2004. Comparison among the Albanian and Greek ophiolites: in search of constraints for the evolution of the Mesozoic Tethys ocean. Ofioliti, 29(1), 19-35.

175

BROWN, S. A. M. & ROBERTSON, A. H. F. 2003. Sedimentary geology as a key to understanding the tectonic evolution of the Mesozoic-Early Cainozoic Paikon Massif, Vardar suture zone, N Greece. Sedimentary Geology, 160, 179-212. BROWN, S. A. M. & ROBERTSON, A. H. F. 2004. Evidence for Neotethys rooted within the Vardar suture zone from the Voras Massif, northernmost Greece. Tectonophysics, 381, 143-173. CANOVI(}, M. & KEMENCI, R, 1997. Geologic setting of the Pre-Cainozoic basement in Vojvodina (Yugoslavia). Part II: The north part of the Vardar zone in the south of Vojvodina. Acta Geologica Hungarica, 42(4), 427-449. CIOBANU, C. L., COOK, N. J. & STERN, H. 2002. Regional setting and geochronology of the Late Cretaceous banatitic magmatic and metallogenic belt. Mineralium Deposita, 37, 541-567. DELEON, G. 1969. Pregled rezultata odredjivanja apsolutne geolo~ke starosti granitoidnih stena u Jugoslaviji. (A review of absolute age determination on granitic rocks from Yugoslavia.) Radovi Instituta za geolo~ka i rudarska istra~ivanja, 6, 165-182 (in Serbian with English. abstract). DIMITRIJEVI(~, M. D. 1982. Dinarides: an outline of the tectonics. Earth Evolution Sciences, 3, 4-23. DIMITRIJEVI(', M. D. 1997. Geology of Yugoslavia. BAREX, Belgrade. DIMITRIJEV1C, M. D. 2001. Dinarides and the Vardar Zone: a short review of the geology. Acta Volcanologica, 13(1-2), 1-8. DIMITRIJEVIC, M. D. & DJOKOVId, I. 1981. Pre-Triassic position of the Eastern Dinarides. Bilten-Bulletin (LMGK), 3, 37-49. DIMITRIJEVId, M. D. & SIKO~EK, B. 1997. Yugoslavia. In: MOORES, E. M. & FAIRBRIDGE, R. W. (eds) Encyclopedia of European and Asian Regional Geology. Chapman & Hall, London, 783-789. DIMO-LAHITTE, A., MONIE, P. & VERGELY, P. 2001. Metamorphic soles from the Albanian ophiolites: petrology, 4~ geochronology, and geodynamic evolution. Tectonics, 20(1), 78-96. DULI~, I. A. 1999. Middle Cretaceous palynomorphs of Serbia and paleophytogeography of Central Tethys. Bulletin de l'Acad~mie Serbe des Sciences et des Arts, CJ(IX, Arts, Classe des Sciences Mathkmatiques et Naturelles--Sciences Naturelles, 39, 151-161. GORICAN, S., KARAMATA, S. &; BATO(~ANINSRECI(OVIC, D. 1999. Upper Triassic (CarnianNorian) radiolarians in cherts of Sjenica and the time span of the oceanic realm ancestor of the Dinaridic ophiolite belt. Bulletin de l'Acadkmie Serbe des Sciences et des Arts, CXIX, Classe des Sciences Mathdmatiques et Naturelles--Sciences Naturelles, 39, 141-149. GRUBI~, A. & ERCEGOVAC, M. 2002. Age of the Veles 'Schistes Lustres' Formation from the Vardar Ocean. Proceedings of the XVIIth Congress of the Carpatho-Balkan Geological Association, Bratislava, 66-68. HAAS, J., KOVACS, S., KRYSTYN, L. & LEIN, R. 1995. Significance of Late Permian-Triassic facies zones in terrane reconstruction in the Alpine-North Pannonian domain. Tectonophysics, 242, 19-40.

176

S. KARAMATA

HAYDOUTOV, I. & YANEV, S. 1996. Palaeozoic terranes in the Protomoesian microcontinent--Bulgaria. In: KNE~EVI(_,V. & KRSTId, B. (eds) Terranes of Serbia. Faculty of Mining and Geology, Belgrade, 77-80. HAYDOUTOV, I., GOCHEV, P., KOZHOUKHAROV,D. & YANEV, S. 1996-1997. Terranes in the Balkan area. Annales Gkologiques des Pays Hellkniques, 37, 479--494. HAYDOUTOV, I., KOLCHEVA, K., DAIEVA, L.-A., SAVOV, I. & CARRIGAN, CH. 2004. Island arc origin of the variegated formations from the east Rhodope, Bulgaria--implications for the evolution of the Rhodope Massif. Ofioliti, 29(2), 145-157. HERAK, M. 1986. A new concept of the geotectoncs of Dinarides. Acta Geologica, 16, 1-42. HRISTOV, S., KARAJOVANOVIK,M. & STRACKOV,M. 1965. The Guidebookfor the Sheet Kavadarci of the Geological Map 1;100 000 of Yugoslavia. ( Tolkuvac za listot Kavadarci OGK SFRJ 1:100 000.) Federal Geological Institute of SFRY, Belgrade (in Macedonian with English abstract). JACOBSHAGEN, V. 1979. Structure and geotectonic evolution of the Hellenides. Proceedings of the 6th Colloquium of Geology of the Aegean Region, Athens, 3, 1355-1367. KARAMATA, S. 2004. Balkan Peninsula--a complex geological framework. Proceedings of the 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, 1, 389-391. KARAMATA, S. & KRSTIC, B. 1996. Terranes of Serbia and neighbouring areas. In: KNE2EVId, V. & KRSTId, B. (eds) Terranes of Serbia. Faculty of Mining and Geology, Belgrade, 25-40. KARAMATA, S. & LOVRIC, A. 1978. The age of metamorphic rocks of Brezovica and its importance for the explanation of ophiolite emplacement. Bulletin de l'Acaddmie Serbe des Sciences et des Arts, CXI, Classe des Sciences Math~matiques et Naturelles-Sciences Naturelles, 17, 1-9. KARAMATA, S. & VUJNOVI(~, L. 2000. Correlation of Palaeozoic units of the Dinarides and the northern part of the Vardar zone. In: KARAMATA, S. tfr JANKOVI(~, S. (eds) Proceedings of the International Symposium 'Geology and Metallogeny of the Dinarides and the Vardar Zone'. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 71-78. KARAMATA, S., KRSTI(~, B., DIMITRIJEVI(', D. M. DIMITRIJEVId, M. N., KNE~,EVld,V., STOJANOV,R. & FILWOVId, I. 1996-1997. Terranes between the Moesian plate and the Adriatic Sea. Annales Gkologiques des Pays Helldniques, 37, 429-477. KARAMATA, S., KNE2EVId, V., PECKAY, Z. 8r DJORDJEWd, P. 1997. Magmatism and metallogeny of the Ridanj-Krepoljin belt (Eastern Serbia) and their correlation with northern and eastern analogues. Mineralium Deposita, 32, 452-458. KARAMATA, S., DIMITRIJEVIC,M. N. & DIMITRIJEVIC, M. D. 1999. Oceanic realms in the central part of the Balkan Peninsula during the Mesozoic. Slovak Geological Magazine, 5(3), 173-177. KARAMATA, S., DIMITRIJEVIC_,M. D. & DIMITRIJEVIC, M.N. 2000a. A correlation of ophiolitic belts and oceanic realms of the Vardar Zone and the Dinarides. In: KARAMATA,S. & JANKOVId, S. (eds) Proceedings of the International Symposium.

'Geology and Metallogeny of the Dinarides and the Vardar zone'. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 191-194. KARAMATA, S., KNE:~EVI(~,V. & CVETKOVIC,V. 2000b. Petrology of the Triassic basaltoid rocks of Vare~ (Central Bosnia). Acta Geologica Hungarica, 43(1), 1-14. KARAMATA, S., OLUJI(~, J., PROTIC, LJ., et al. 2000c. The western belt of the Vardar zone--the remnant of a marginal sea. In: KARAMATA, S. • JANKOVIC, S. (eds) Proceedings International Symposium 'Geology and Metallogeny of the Dinarides and the Vardar Zone'. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 131-135. KARAMATA, S., KNE7.EVIC-DJORDJEVIC, V. & MmOVANOV~d, D. 2002. A review of the evolution of Upper Cretaceous-Paleogene magmatism in the Timok magmatic complex and the associated mineralization. In: JELENKOVI~, R. & KO~ELJ, D. (eds) Proceedings of the International Symposium, Geology and Metallogeny of Copper and Gold Deposits in the Bor Metallogenic Zone--Bor 100 Years, Special Issue, Bor Lake, Yugoslavia. Copper Institute, Bor, 15-28. KARAMATA, S., STEFANOVI(~, D. & KRSTI(~, B. 2003. Permian to Neogene accretion of the assemblage of geologic units presently occurring to the south of the Pannonian Basin~development of the Vardar Composite Terrane and adjacent units. Aeta Geologica Hungarica, 46(1), 63-76. KEMENCI, R, & CANOVI(~,M. 1997. Geologic setting of the Pre-Cainozoic basement in Vojvodina. Part I: The Tisza Mega-unit of North Vojvodina. Acta Geologica Hungarica, 40(1), 1-36. KEPPIE, J. D. & DALLMEYER,R. D. 1990. Introduction to the terrane analysis and the tectonic map of preMesozoic terranes in Circum-Atlantic Phanerozoic orogens. Abstracts, IGCP Meetings', 233, 24. KlSSEL, C., SPERANCA, F. F. & MILIdEVId, V. 1995. Paleomagnetism of external southern and central Dinarides and northern Albanides. Journal of Geophysical Research, 100, 1499-1508. KNEZEVI(~, V. & CVETKOVI(~,V. 2000. Triassic rifting magmatism of the Dinarides. In: KARAMATA,S. & JANKOVI~, S. (eds) Proceedings of the International Symposium 'Geology and Metallogeny of the Dinarides and the Vardar Zone'. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 149-160. KOSSMAT, F. 1 9 2 4 . Geologic der zentralen Balkanhalbinsel--Kriegsschauplaetze 1914-1918, geologisch dargestellt, 12, 1-198. KR~UTNER, H. G. & KRSTI~, B. 2003. Geologicalmap of the Carpatho-Balkanides between Mehadia, Oravita, Ni~ and Sofia. Geoinstitut, Belgrade. KRSTI~, B. & KARAMATA, S. 1992. Terrranes of the Eastern Serbian Carpatho-Balkanides. (Terani u Karpato-Balkanidima Istocne Srbije.) Comptes rendus des Seances de la Soci~tk Serbe de GOologie, Livre Jubilaire (1981-1991), 57-74. KRSTI(~, B., KARAMATA,S. & MILId-EVI~,V. 1996a. The Carpatho-Balkanide terranes--a correlation. In:

TERRANE MODEL FOR THE BALKAN PENINSULA KNEZEVId,V. & KRSTId, B. (eds) Terranes of Serbia. Faculty of Mining and Geology, Belgrade, 71-76. KRSTId, B., MASLAREVld,LJ. & MILId~vId, V. 1996b. Ordovician in the Ku6aj terrane of the East-Serbian Carpatho-Balkanides. In: KNE2EVld,V. & KRSTId, B. (eds) Terranes of Serbia. Faculty of Mining and Geology, Belgrade, 91-95. LANPHERE, M. A., COLEMAN,R. G., KARAMATA,S. & PAMId, J. 1975. Age of amphibolites associated with alpine peridotites in the Dinaride ophiolite zone, Yugoslavia. Earth and Planetary Science Letters, 26, 271-276. LAZENDI~, V. 2002. Paleomagnetic characteristics of the Jadar block, N W Serbia. (Paleomagnetske karakteristike mladjeg paleozoika Jadarskog bloka, SZ Srbija). MSc Thesis, Faculty of Mining and Geology, Belgrade (in Serbian). LUGOVIC,B., ALTHERR, R., RACZEK,I., HOFMANN,A. & MAJER, V. 1991. Geochemistry of peridotites and mafic igneous rocks from the Central Dinaridic Ophiolite Belt, Yugoslavia. Contributions to Mineralogy and Petrology, 106, 201-216. MAROVId,M., DJOKOVId,I., MILI~VI~, V., TOLJI(~,M. & GERZlNA, N. 2002. Paleomagnetism of the Late Paleogene and Neogene rocks of the Serbian Carpatho-Balkanides: tectonic implications. Annales Gdologiques de la Pdninsule Balkanique, 64, 1-12. MARTON, E. & MILld~vId, V. 1994. Tectonically oriented paleomagnetic investigation in the Dinarides, between Zadar and Split. Geophysical Transactions, 39(4), 227-232. MASLAREVId, LJ. & KRSTId, B. 2001. Continental Permian and Lower Triassic red beds of the Serbian Carpatho-Balkanides. In: CASSINIS, G. (ed.) Permian Continental Deposits of Europe and other Areas. Regional Reports and Correlations. Natura Bresciana, 25, 245-252. MAURITSCH, H. J., SCHOLGER, R., BUSHATI, S. L. & XHOMO, A. 1996. Paleomagnetic investigations in northern Albania and their significance for the geodynamic evolution of the Adriatic-Aegean realm. In: MORRIS, A. & TARLING, D. H. (eds) Paleomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 265-275. MEMOVId, E., CVETKOVId, V., KNE~EVI(~, V. & ZAKARIADZE, G. 2004. The Triassic metabasalts of Dudin Kr~, near Kosovska Mitrovica, Serbia. Annales G~ologiques de la Pdninsule Balkanique, 65, 85-91. MILIdEWd, V. 1985-1986. Paleomagnetism and paleogeographic position of the Turonian-Senonian and Senonian sediments from the Carpatho-Balkanic realm in Yugoslavia. (Paleomagnetizam i paleogeografska pozicija turon-senonskih i senonskih sedimenata sa prostora Karpato-balkanida u Jugoslaviji.) Bulletin/Vesnik, C--Geophysique, 261 27, 63-78 (in Serbian with English abstract). MILI~EVId, V. 1996. Ku6aj terrane in Palaeozoic time. In: KNE~EVld, V. & KRSTId, B. (eds) Terranes of Serbia. Faculty of Mining and Geology, Belgrade, 87-89.

177

MILI(~EVI(~, V. 1998. Palinspastics of the Hercynides in the Kudaj Zone of Eastern Serbia (Palinspastika hercinida u kudajskoj zoni Istodne Srbije.) Posebna izdanja. Geoinstituta (Special publication of the Geoinstitute), 26 (in Serbian). MILId-EVI(~,V. & DJURASINOVIC-GAVRILOVI(~,M. 1990. Paleomagnetism of the Paleogene (Oligocene) basins of Serbia. (Paleomagnetizam paleogenih (oligocenskih) basena Srbije.) XII Kongres Geologa. Jugoslavije, V, 276-286 (in Serbian with English abstract). MILI(ZEVI~, V., MILOVANOVI(}, D., FILIPOVIC, I. & JOVANOVI(~, D. 1995. Paleomagnetism of Palaeozoic formations in Jadar trough, NW Serbia. Proceedings of the X V Congess of the Carpatho-Balkan Geological Association. Geological Society of Greece, Special Publication, 4(3), 1125-1129. MILOVANOVI(~,B. 1950. Tectonic sketch of Yugoslavia. (Tektonska skica Jugoslavije.) In: M~LOVANOVId,B. (ed.) Geologija za rudare. Izdava~ko preduzede Saveta za energetiku i ekstraktivnu industriju Vlade FNRJ, Beograd, 411-452 (in Serbian). MILOVANOVIC, D., MARCHIG, V. & KARAMATA, S. 1995. Petrology of the crossite schists from Fru~ka Gora Mts (Yugoslavia), relic of a subducted slab of the Tethyan oceanic crust. Journal of Geodynamics, 20(3), 289-304. MUTTONI, G., GAETANI,M., BUDUROV,K., et al. 2000. Middle Triassic paleomagnetic data from northern Bulgaria: constraints on Tethyan magnetostratigraphy and paleogeography. Palaeogeography, Palaeclimatology, Palaeoecology, 160, 223-237. NEUBAUER,F. & VON RAUMER,J. F. 1993. The Alpine basement--linkage between Variscides and Mediterranean mountain belts. In: VON RAUMER,J. F. & NEUBAUER, F. (eds) Pre-Mesozoic Geology in the Alps. Springer, Berlin, 641-663. NEUBAUER, F., EBNER, F. & WALLBRECHER,E. 1995. Geological evolution of the internal Alps, Carpathians and of the Pannonian basin: an introduction. Tectonophysics, 242, 1-4. OBRADOVld,J. & GORI~AN, S. 1988. Siliceous deposits in Yugoslavia: occurrences, types and ages. In: HEIN, J.R. & OBRADOVId,J. (eds) Siliceous Deposits of the Tethys and Pacific Regions. Springer, New York, 51-64. OKRUSCH, M., SEIDEL,E., KREUZER,H. & HARRE, W. 1978. Jurassic age of metamorphism at the base of the Brezovica peridotite (Yugoslavia). Earth and Planetary Science Letters, 39, 291-297. PAMId, J. & BELAK, M. 1996-1997. The MoslavinaSlavonija terrane. Annales Gdologiques des Pays" Hellkniques, 37, 402-407. PAMIC, J., GugI(;, I. & JELASKA,V. 1998. Geodynamic evolution of the Central Dinarides. Tectonophysics, 297, 251-268. PANTI(~, N. & DULI(~, I. 1991. Jungkarbonische Floren der Balkanhalbinsel und ihre palaeobiogeographische Bedeutung. Proceedings of the PanEuropean Palaeobotanical Conference, Vienna, 371-375. PAPANIKOLAOU, D. 1996-1997a. IGCP Project No. 276--terrane maps and terrane descriptions. Annales Gdologiques des Pays Hell~niques, 37, 193-599.

178

S. KARAMATA

PAPANIKOLAOU, D. 1996-1997b. The tectonostratigraphic terranes of the Hellenides. Annales Gdologiques des Pays Helldniques, 37, 495-514. PETKOVId, K. 1961. La carte tectonique de la RPF de Yugoslavie. Glas SANU, 149, Odeljenje prirodnomatemati~kih nauka, Beograd, 22, 129-144. PoPov, P., BERZA, T. & GRUBIC, A. 2000. Upper Cretaceous Apuseni-Banat-Timok-Srednogorie (ABTS) Magmatic and Metallogenic Belt in the Carpathian-Balkan Orogen. ABCD-GEODE Workshop, Abstract Volume, Borovets, Bulgaria, 69-70. PROTId, LJ., FILIPOVIC, I., PELIKAN, P., et al. 2000. Correlation of the Carboniferous, Permian and Triassic of the Jadar Block, Sana-Una and 'Bukkium' terranes. In: KARAMATA, S. t~; JANKOVId, S. (eds) Proceedings of the International Symposium "Geology and Metallogeny of the Dinarides and the Vardar Zone'. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 61-69. RADULOVId, V. 1995. A review of the Lower and Middle Jurassic brachiopod distribution in the southern Carpatho-Balkan arc and the Yugoslav External Dinarides. Geologica Carpathica, 46(6), 371-377. RAMOVS, A., HINTERLECHNER-RAVNIK,A., KALENIC, M., et al. 1984. Stratigraphic Correlation Forms (SCF) of the Yugoslav Palaeozoic. In: SASSl, F. P. & JULIVERT, M. (eds) IGCP No. 5, Newsletter, 6, 81-109. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the Eastern Mediterranean. In: ROBERTSON, A. H. F. & DIXON, J. E. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. 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. H. F., DIXON, J. E., BROWN, S., et al. 1996. Alternative tectonic models of the Late Palaeozoic-Early Cainozoic development of Tethys in the Eastern Mediterranean region. In: MORRIS, A. & TARLING, D. H. (eds) Paleomagnetism and Tectonics of the Mediterranean. Geological Society, London, Special Publications, 105, 239-269. RODD1CK,J. F., CAMERON,W. E. 8s SMITH, A. G. 1979. Permo-Triassic and Jurassic Ar-Ar ages from Greek ophiolites and associated rocks. Nature, 279, 788-790. SANDULESCU, M. 1984. Geotectonics of Romania. Tehnica, Bucharest. SEYFERT, C. K. t~ SIRKINL. A. 1973. Earth History and Plate Tectonics. Harper & Row, New York. SHALLO, M. 1992. Geological evolution of the Albanian ophiolites and their platform periphery. Geologische Rundschau, 81(3), 681-694. SHALLO, M. 1994. Outline of the Albanian ophiolites. Ofioliti, 19/1, 57-75.

SPRAY, J. G. & RODDICK, J. C. 1980. Petrology and 4~ geochronology of some Hellenic subophiolitic metamorphic rocks. Contributions to Mineralogy and Petrology, 72, 4--5. S'rAMPFLI, 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, 163-185. STAMPELI, G. M. & BOREL, G. D. 2004. The TRANSMED transect in space and time-constraints on the paleotectonic evolution of the Mediterranean domain. In: CAVAZZA,W., ROURE, B., SPAKMAN, W., STAMPFLI, G. M. & ZIEGLER, P.A. (eds) The T R A N S M E D Atlas. The Mediterranean Region from Crust to Mantle. Springer, Berlin, 53-90. STAMPFLI, G. M., MARCOU, J. & BAUD, A. 1991. Tethyan margins in space and time. Palaeogeography, Palaeoclimatolgy, Palaeoecology, 87, 373-410. STEFANOVIC, D. • VELJOVIC, D. 1981. Paleomagnetic characteristics of some Upper Cretaceous volcanic rocks of the Timok eruptive complex. (Paleomagnetske karakteristike nekih gornjokrednih vulkanita Timo~ke eruptivne oblasti.) Glas 329, Acadkmie Serbe des Sciences et des Arts, Classe des Sciences Mathkmatiques et Naturelles--Sciences naturelles, 48, 53-62 (in Serbian with English abstract). THUIZAT, R., WHITECHURCH, H., MONTIGNY, R. & JUTEAU, T. 1981. K-Ar dating of some infraophiolitic metamorphic soles from the Eastern Mediterranean: new evidence for oceanic thrusting before obduction. Earth and Planetary Science Letters, 52, 302-310. VELJOVl(}, D. & MILIdEVI(, V. 1986. Report of the results of magnetic and paleomagnetic investigations of rock samples collected in Serbia in 1985 for the elaboration of the paleogeographic map. (Izve~taj o rezultatima magnetskih i paleomagnetskih ispitivanja uzoraka stena prikupljenih sa lokaliteta SR Srbije u toku 1985. godine u cilju izrade paleogeografske karte. ) Reports of the Geomagnetski lnstitut, Belgrade (in Serbian). VELJOVIC, D. t~ MILId'EVI(, V. 1987. Report of the results of magnetic and paleomagnetic investigations of rock samples collected in Serbia in 1986 for the elaboration of the paleogeographic map. (Izve~taj o rezultatima magnetskih i paleomagnetskih ispitivanja uzoraka stena prikupljenih sa lokaliteta SR Srbije u toku 1986. godine u cilju izrade paleogeografske karte. ) Reports of the Geomagnetski Institut, Belgrade (in Serbian). ZIEGLER, A. P. & STAMPFLi, M. G. 2001. Late Palaeozoic--Early Mesozoic Plate Boundary Reorganization: Collapse of the Variscan Orogen and Opening of Neotethys. In: CASSlNIS, G. (ed.): Permian Continental Deposits of Europe and other Areas. Regional reports and correlations. Natura Bresciana, 25, 17-34, Brescia.

Evolution of Early Mesozoic back-arc basins in the Black Sea-Caucasus segment of a Tethyan active margin V. G . K A Z M I N

& N . F. T I K H O N O V A

Institute o f Oceanology R A S , N a k h i m o v s k y Prospect 36, 117997, Moscow, Russia (e-mail: vkazmin@geo, sio. rssi. ru) Six new reconstructions illustrate the evolution of back-arc basins in the Black Sea-Caucasus region from the Mid-Triassic to the end of the Mid-Jurassic. The c. 2000 km long Tauric (Kiire) basin opened in the Late Permian-Early Triassic as the PontidesTranscaucasus and Rhodope microcontinents rifted from the Eurasian margin. The oceanic floor of the Tauric basin in the Mid-Triassic was at least 300 km wide. In the east the basin closed near the present-day Caspian Sea and to the west of the West Crimea transform it split into two branches to the south and north of the Moesian platform. The Tauric basin was partly inverted in the Carnian, when several Gondwanian terranes (Iran, South Armenia) collided with the Palaeotethyan subduction zone. Following the initiation of a new subduction zone, the back-arc extension resumed in the Norian-Early Jurassic. Opening of the Izmir-Ankara-Sevan back-arc basin commenced south of the Pontides-Transcaucasus. Simultaneously, rifting began in the Greater Caucasus and continued until the Early Pliensbachian. This was followed by the continental break-up in the Late PliensbachianToarcian. A narrow (100-150 km) strip of oceanic crust had formed by the beginning of the Aalenian. In the Late Aalenian a southward-migrating subduction zone at the southern margin of the Izmir-Ankara-Sevan basin had reached the central part of Neo-Tethys and presumably collided with a mid-oceanic ridge. Subduction was blocked and Africa-Eurasia convergence was compensated by inversion in the Tauric and Greater Caucasus basins. The basins were closed by the end of the Bathonian. Abstract:

During the last two decades several attempts were made to reconstruct the history of the early Mesozoic back-arc basins in the Black SeaCaucasus-South Caspian region (Dercourt et al. 1985, 1993, 2001; Adamia et al. 1990a; Kazmin 1990; U s t a r m e r & Robertson 1993; Stampfli 1996; Banks & Robinson 1997; Kazmin & Natapov 1998; Stampfli et al. 1998; Nikishin et al. 2001; Stampfli & Borel 2002). Although significant progress has been made, in most of the published works the reconstructions were schematic. The main problems concern the relationships between the Tauric (Kfire) back-arc basin and Greater Caucasus basin, the time and the mode of origin of the latter, and the configuration and the evolution of both basins. In most reconstructions, as listed above, the Greater Caucasus basin is interpreted as an eastward extension of the Tauric basin, although there is reliable evidence that the two basins were separated by the crustal block of the Shatsky rise. The opening of the Greater Caucasus basin in early Jurassic time and its subsequent evolution is usually related to a subduction zone along the southern margin of the Pontides. However, there is convincing evidence that in the Jurassic and Neocomian this margin was passive (Altiner & Ko~yi~it 1992;

Tiiysiiz et al. 1995; Okay & Sahintfirk 1997). Consequently, the interpretation of the early Mesozoic evolution of the Pontides-Caucasus region needs revision. Restoration of the early Mesozoic history is hampered by a lack of reliable palaeomagnetic data. For reasons still unknown, palaeomagnetic measurements of Jurassic rocks of the Pontides, Transcaucasus and Crimea yield very low inclinations, corresponding to remote southerly positions far from Eurasia (Asanidze & Pechersky 1979; Lauer 1984; Westphal et al. 1986; Saribudak 1988; Pechersky & Safronov 1993). The only attempt to reconcile the palaeomagnetic and geological data, by Kazmin & Natapov (1998), was unsuccessful. In the present paper, controversial palaeomagnetic data on terranes were not used. Movements of terranes relative to the Eurasian margin were instead deduced from geological data; i.e. the time of rifting and collision, the duration of rifting, spreading and subduction periods. Reasonable spreading and subduction rates were assumed. The position of the Eurasian margin was taken from recently published works (Kazmin & Natapov 1998; Daukeev et al. 2002), where it was calculated using oceanic magnetic anomalies and plate motion relative to hotspots.

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 179-200. 0305-8719106/$15.00 9 The Geological Society of London 2006.

180

V.G. KAZMIN & N. F. TIKHONOVA

Fig. 1. Main structures of the Alpine Belt in the Black Sea-Caucasus region. AL, Alborz; AR, Andrusov rise; A-T, Adjaro-Trialetia; CI, Central Iran; CRB, Circum Rhodope belt; CP, Central Pontides; DB, Donbass; DD, Dniepr-Donets aulacogen; EBB, Eastern Black Sea basin; EEP, East European platform; EI, East Iran; EP, Eastern Pontides; GC, Greater Caucasus; GCF, Greater Caucasus fold belt; IZ, Istanbul zone; KD, Kura depression; KDB, Kopetdag basin; KM, Kargl massif; KR, Kir~ehir massif; KS, Karpinsky swell; M, Mangyshlak; ME, Menderes massif; MG, Manych graben; MP, Moesian platform; ND, North Dobrogea; PB, Pre-Caspian basin; PC, Pechenega-Camena fault; PFB, Palaeozoic fold belt; PR, Paikon Ridge; RD, Rhodope massif; S, Strandja (Istranca) zone; SA, South Armenian terrane; SC, South Crimea; SCB, South Caspian basin; ScP, Scythian platform; SG, Sredna Gora zone; ShR, Shatsky rise; SK, Sakarya (Sakaria) block; SM, Serbomacedonian massif; SP, Stara Planina zone; SS, Sanandaj-Sinjar zone; SV, Svanetia; T, Turanian platform; TB, Tauric basin; TFB, Triassic fold belt; TM, Transcaucasus massif; TU, Tuarkyr; VC, Vardar suture; WBB, Western Black Sea basin; WC, West Crimea fault; WEP, West European Platform. There are two types of terranes involved in the evolution of the active Eurasian margin and the evolution of the related back-arc basins (see

Fig. 1). Terranes (microcontinents) of the first type have a Neoproterozoic basement strongly altered by Hercynian tectonics (Adamia et al.

BLACK SEA CAUCASUS BACK-ARC BASINS 1989; Okay & Sahinttirk 1997; Zakariadze et al. 1998). The wide development of pre- to syntectonic granitoids (330-280 Ma) and late Palaeozoic molasse with clear Eurasian affinity (Belov 1981) indicates that these blocks were rifted from the late Palaeozoic active margin of Europe. They formed a chain, including the Transcaucasian massif, the Pontides and also blocks of the Andrusov and Shatsky rises, which formed parts of the Pontides-Transcaucasus prior to opening of Mesozoic marginal basins. Less clear is the situation of the Rhodope massif. Traditionally its crust was described as Precambrian, strongly affected by Hercynian and Alpine tectonometamorphic events (Kronberg et al. 1970; Jones et al. 1992; Kozhoukharova 1996). According to others, the massif is an Alpine metamorphic complex formed by Cenozoic subduction-accretion processes (Barr et al. 1999; see also Himmerkus et al. 2006). Perhaps a compromise solution is acceptable: the Rhodope massif was perhaps a part of the Palaeozoic margin of Eurasia, to which magmatic material was added during the Alpine cycle. In the following reconstruction we envisage that a Triassic back-arc basin opened between the Rhodope massif and the Moesian platform and that the Rhodope massif was a part of a 'Rhodope-Pontide fragment' ($eng6r 1984). Terranes of the second type can be seen as fragments of Gondwana that collided with the Eurasian margin during the Mesozoic and Early Cenozoic. The largest of these fragments, Iran, belonged to the ribbon-like Cimmerian continent ($eng6r 1979) and had its western extension as a chain of blocks including the South Armenian terrane (Dercourt et al. 1986), probably the Kir~ehir massif and some smaller fragments. As there are few data on Alpine accretion, the present-day size of post-late Triassic terranes is assumed with some corrections (e.g. straightening of Alpine bends, approximate enlargement of partly underthrust terranes) in the following reconstructions.

Early-Mid-Triassic reconstruction (Fig. 2) Many workers suggested that a large basin existed in the Triassic and Jurassic between the Scythian platform and the Pontides (~eng6r & Ydmaz 1981; Seng6r 1984; Adamia et al. 1990a; Kazmin 1990; Usta6mer & Robertson 1993; Stampfli 1996; Stampfli et al. 1998). Seng6r viewed this basin as a relict of Palaeotethys. However, later studies demonstrated convincingly that this large basin was formed behind a north-dipping subduction zone in which the

181

Palaeotethyan crust was consumed. The Karakaya accretionary complex, including rocks that originated in abyssal, carbonate platform and trench settings, formed related to this subduction. In the Central Pontides a Triassic magmatic arc (the Qangaldag arc) and a back-arc basin were reconstructed (Pickett & Robertson 1996; Usta6mer & Robertson 1997; Robertson 2002). The basin has been given different names: the Ktire (Usta6mer & Robertson 1993, 1997; Nikishin et al. 2001; Stampfli & Borel 2002) or Tauric basin (Kazmin 1990). Fragments of its oceanic crust and sediments crop out in the fold belts of North Dobrogea and South Crimea, in the Strandja zone and in the Central Pontides. They were also penetrated by drill-holes in the northwestern shelf edge of the Black Sea. In the Tulchea zone of the North Dobrogea fold belt a continuous succession of sediments of early-mid-Triassic to mid-Jurassic age marks the northern passive margin of the basin (Gradinaru 1988, 1995). The facies become progressively deeper towards the axial zone of the belt, where late Triassic-early Jurassic flysch-type units are known. These sediments and a unit of mid-ocean ridge basalts (MORB) (Stampfli et al. 1998) intercalated with the deep-sea carbonates form north-vergent tectonic slices within the Niculitel nappe pile. The age of the basalts ranges from the late Early Triassic (Scythian) to Carnian (Sandulescu 1995). Very similar to the flysch-type units of the North Dobrogea is the Tauric Series of South Crimea. This comprises proximal and distal turbidites formed on the south-facing slope and rise (Mazarovich & Mileev 1989). The oldest sediments belong to the Ladinian, and the youngest to the Mid-Jurassic. In the Norian and Early Jurassic parts of the succession there are intercalated lavas and tufts ranging from basalts and andesite-basalts to a acidic varieties. Drilling shows that the sediments of the Tauric Series extend along the Black Sea shelf edge towards North Dobrogea (Ulanovskaya & Shevchenko 1992), thus marking the northern margin of the Triassic-Jurassic basin. The ophiolites and associated rocks of the Kiire area in the Central Pontides were first described as slices of the Palaeotethyan crust (Yllmaz & Seng6r 1985). Detailed structural and geochemical studies later demonstrated that two types of ophiolites are present (Usta6mer & Robertson 1997; Robertson 2002). The first type is represented by dismembered ophiolites in the Karakaya accretionary complex. Ophiolites of the second type are interpreted as tectonic slices of oceanic crust formed in a back-arc (Kiire) basin. They are covered by phyllites and

182

V.G. KAZMIN & N. F. TIKHONOVA

Fig. 2. Mid-Triassic reconstruction. Time of the maximum opening of the Tauric basin (Ladinian). Abbreviations as in Figure 1.

flysch-type sediments (the Akg61 Formation), of mid-Triassic to mid-Jurassic age accumulated. (Usta6mer & Robertson 1997; Robertson 2002). A close resemblance of the Akg61 Formation to

the Tauric Series of Crimea has been emphasized (Bocaletti & Manetti 1988). Data on the western extension of the Tauric basin come from the Strandja zone, on the

BLACK SEA CAUCASUS BACK-ARC BASINS southernmost periphery of the Balkanides. Here, low-grade metamorphic rocks of Early TriassicMid-Jurassic age unconformably overlie a metagranitic basement, intruded by 300 Ma granites (Okay et al. 2001). The cover and the basement form a series of north-vergent nappes. Usta6mer & Robertson (1993) were the first to suggest that early Mesozoic sediments were deposited on a south-facing passive margin of a back-arc basin that opened between the Moesian platform and the Rhodope massif. This interpretation was later confirmed by restoration of the predeformation structure, but the margin was referred to as 'Palaeotethyan' (Banks 1997; Okay et al. 2001). However, geological data show that in the Triassic the northern margin of Palaeotethys from Kunlun to the Pontides was active (Kazmin & Natapov et al. 1998). Most probably the Palaeotethyan subduction zone extended westward to south of the Rhodope massif (Golonka 2000; Dercourt et al. 2001). We support, therefore, the earlier suggestion that the western branch of the Tauric basin opened between Rhodope and Moesia. The above data confirm that in Triassic time a large Tauric basin with oceanic-type crust existed between the Rhodope-Pontide fragment and the Scythian platform. The eastern part of the Tauric basin opened between the Eastern PontidesTranscaucasus microcontinent and the Shatsky rise and its eastern extension, the Dzirula massif. To the east the Tauric basin narrowed and closed at the longitude of the present-day western coast of the Caspian Sea. Further east, in the south Turan, the Triassic active margin was of Andean type. Back-arc extension there (if any), resulted only in opening of small epicontinental basins (Boulin 1990). Accordingly, the PontidesTranscaucasus block occupied a diagonal position relative to the Eurasian margin; this places constraints on the width of the Tauric basin at the longitude of the Crimea-Central Pontides. It has been suggested that the Tauric basin extended directly eastwards into the Greater Caucasus (Stampfli et al. 1998; Nikishin et al. 2001), where Permo-Triassic sediments are usually included in the upper part of the Dizi Series of Svanetia (Somin & Belov 1967; Adamia 1968). Later studies have demonstrated that these sediments form an individual complex separated from the Palaeozoic sediments of the Dizi Series by a period of intensive folding and metamorphism, and that they accumulated in a backarc basin north of the Transcaucasian massif (Kazmin & Sborshchikov 1989). However, this basin is not seen as a direct extension of the Tauric basin. In Svanetia, Triassic sediments are shallow-water quartzitic and arkosic clastic

183

rocks, lacking volcanics rocks, and have nothing in common with the flysch and ophiolites of the Tauric basin. In the Late Triassic the Svanetia basin was inverted and unconformably overlapped by early Jurassic sediments, whereas the Tauric basin existed until the Mid-Jurassic. Finally, in the Mid-Jurassic, northward subduction of the Tauric basin was accompanied by formation of a volcanic arc on the Dzirula massif and Shatsky rise. This means that the Tauric basin was located south of the Shatsky rise, whereas the Permo-Triassic basin of the Greater Caucasus was to the north of it (Fig. 2). The western part of the Tauric Basin consisted of two branches. The North Dobrogea branch opened between the Scythian platform and a continental fragment that was rifted from it and located within the Pontides as the Istanbul zone. In the Istanbul zone Neoproterozoic basement is covered by platform-type Ordovician and younger Phanerozoic sediments, representing part of the south-facing Palaeozoic passive margin of eastern Europe ($eng6r 1984; Usta6mer & Robertson 1993; Okay et al. 1994; Yllmaz et al. 1997). The passive margin can be traced through the north Crimea to the Bechasyn zone of ForeCaucasus, which is geologically identical to the Istanbul zone. Data on the southern branch of the Tauric basin between the Moesian platform and the Rhodope massif are very limited. According to Banks (1997) and Okay et al. (2001), the northern passive margin of this branch originated in the 'earliest Triassic', whereas the final closure of the basin began in the late Mid-Jurassic. A brief period of inversion in the Carnian was followed by the accumulation of the Late Triassic Lipachka flysch. The significance of this event has been interpreted either as a transition to a compressional regime, or as a resumption of extension. The second interpretation is preferable in our opinion (see the next section for details). There is no evidence for the development of oceanic crust in this basin, but its width could be considerable, taking into account its long, Early Triassic-Early Jurassic, period of existence. A sharp change in the Tauric basin structure coincides with one of the major transverse features of the region, the West Crimea fault. Another major fault, the Pechenega-Camena strike-slip fault, was also active in the Early Mesozoic, constituting the southwestern transform boundary of the North Dobrogea basin (Gradinaru 1988). Both faults belong to a southeastern extension of the Tornquist lineament and were instrumental in the subsequent opening and evolution of the Western Black Sea basin,

184

V.G. KAZMIN & N. F. TIKHONOVA

acting as NW-trending transforms (Okay et aL 1994; Kazmin et al. 2000). As these faults were convex to the east, the Tauric basin narrowed and probably closed westwards, i.e. towards the pole of opening. The zone of the PechenegaCamena and West Crimea faults corresponds, accordingly, to the Euretian Equator. To estimate the probable width of the Tauric basin, the following considerations can be used. The period of spreading in the North Dobrogea branch lasted from the late Early Triassic (Scythian) to the Early Carnian, i.e. for about 15-16 Ma. At a spreading rate of 1 or 2 cm a -~, the newly formed crust was about 150-300 km wide. If the southern branch was opening at the same rate, the total width of the oceanic-type basement could have reached 300-600 km. Of the previously published reconstructions, the closest to that presented in Figure 3, although more schematic, is that by Usta6mer & Robertson (1993).

Late Triassic stage (Fig. 3) The Tauric basin was partly inverted in the Carnian. At that time a number of the

Gondwana-derived microcontinents collided with the Eurasian margin. This event is well dated. In northern Iran the Lower-Middle Triassic carbonates of 'Tethyan' type changed abruptly in the Late Carnian-Norian to continental coal-bearing clastic deposits of the Shemshak Formation, typical of the adjacent regions of Eurasia (Dercourt et al. 1986). North of the Alborz, in the area of the future south Caspian basin, the Cimmerian fold belt was formed. The frontal nappes of this belt, containing ophiolites, are known in western Alborz and in the Aladag-Binalud (easternmost extension of Alborz) (Alavi 1996). To the west the fold belt extended into the Greater Caucasus, where the Permian-Triassic (Svanetian) rift basin was inverted. West of Iran several smaller continental fragments docked with the Transcaucasus and Eastern Pontides. In one, the South Armenian microcontinent, the Tethyan Palaeozoic-Triassic succession and a Late Triassic transition to the Shemshak facies is well documented (Dercourt et al. 1986). South Armenia was either an extension of Iran or constituted an independent block. Less certain is the position of the Kir~ehir massif. According to Tiiysfiz et al. (1995), this block was

Fig. 3. Late Triassic reconstruction. Collision and partial inversion of the Tauric basin (Carnian). Abbreviations as in Figure 1; legend as in Figure 2.

BLACK SEA CAUCASUS BACK-ARC BASINS rifted from the Sakarya massif in the Early Jurassic to open the Izmir-Ankara Ocean. On the other hand, in most reconstructions (e.g. Dercourt et al. 1985, 1993, 2001; Golonka 2000) the Kir~ehir massif is regarded as a fragment of Gondwana, rifted from its margin in the Permian or Early Mesozoic. Because of very strong Alpine magmatic history and structural remobilization (Whitney et al. 2001), the history of Kir~ehir is still poorly understood, so its inclusion in Cimmeria, as suggested here, is hypothetical. In the same category as South Armenia possibly belongs the Kargi massif, a carbonate platform within the Triassic accretionary complex of the Central Pontides (Usta6mer & Robertson 1997). The effect of collision in the Pontides was mild: the eastern branch of the Tauric basin remained open but probably reduced. In the flysch sequences of the Crimea and the Central Pontides (Tauric Series; Akg61 Formation.) sedimentation continued from the Mid-Triassic (Ladinian?) to Carnian and Norian without a visible break or deformation that could be attributed to closure of the basin. As mentioned above, the western branch between the Moesian platform and the Rhodope massif was also not closed until the end of the Mid-Jurassic, although a short period of compression perhaps led to some shortening of the basin (Banks 1997; Okay et al. 2001). The compression resulted either in underthrusting (subduction?) at the northern margin of the basin (Fig. 3) and/or in the overthrusting of the Kirklareli nappe at the southern margin ($eng6r 1984). In the North Dobrogea branch of the Tauric basin the onset of accumulation of late Triassic flysch is usually regarded as marking a transition from extension to compression (E. Gradinaru, pers. comm.). The late Triassic compression was not restricted to the collision zone or back-arc basin but affected the adjacent portions of the Moesian and Scythian platforms. In the continental rift system extending from the Moesian platform to Mangyshlak and Tuarkyr (see above) the marine sediments were folded and faulted, and in some cases low-angle detachments developed (Tari et al. 1997; Volozh et al. 1999; Orel 2001). The inversion terminated in emergence and cessation of marine sedimentation. The Triassic rifts, and also the adjacent late Palaeozoic fold belts of the Karpinsky swell and Mangyshlak, were deformed and uplifted (Volozh et al. 1999).

Late Triassic-early Mid-Jurassic stage (Figs 4 & 5) As a result of the late Triassic collision, the Palaeotethyan subduction zone was blocked

185

and a new subduction zone developed south of the accreted microcontinents. This event was followed by extension, which led to rifting and then opening of the new back-arc basins in a wide back-arc region: in Iran, south of the PontidesTranscaucasus and in the Greater Caucasus. In Iran, the period of extension began in the Late Carnian. A system of east-west-trending continental rifts transected this territory, extending into the East Iran block. At present, the early Mesozoic rifts in this block strike NE. This implies rotation of east Iran by 90-130 ~ anticlockwise in post-Triassic time, as confirmed by palaeomagnetic data (Soffel & F6rster 1984). A spectacular discovery was made of a thick early to mid-Triassic marine clastic sequence with ammonites in central Iran, in the Anarek area (Aistov et al. 1984; Ruttner 1984) (Fig. 6). When rotated clockwise with the rest of east Iran, this area return to its initial position along the active Eurasian margin, where sediments of this type are known in the Aghdaraband area (Ruttner 1984). The back-arc rifting in Iran was a reaction to the onset of subduction at its southwestern margin. The evidence for the newly formed active margin comes from the northwestern part of the Sanandaj-Sinjar zone, where upper Triassiclower Jurassic turbidites and 'schistes lustres', intercalated with andesitic-basaltic pillow lavas, are known in the Mahabad and Esfahan area (National Iranian Oil Company 1975-1979; Cherven 1986) (Fig. 6). Following previous reconstructions (e.g. Dercourt et al. 1993), we believe that the Sanandaj-Sinjar block originally constituted the southwestern margin of Iran. In Norian-early Jurassic time the Alborz was a rapidly subsiding coastal plain, on which 30004000 m of the coal-bearing Shemshak clastic deposits accumulated in a paralic setting. The source of the terrigenous material was to the north, where the Cimmerian fold belt was eroded (Berberian & King 1981; Davoudzadeh & Schmidt 1981, 1984; Lensch et al. 1984). South of the Alborz a marine basin probably occupied, an east-west rift, separating the Alborz from the rest of Iran. West of Iran, at the southern margin of the Rhodope-Pontide fragment, extension behind a newly formed subduction zone led first to rifting (Fig. 4) and then to opening of the IzmirAnkara-Sevan basin (Fig. 5). The oldest continental sediments, related to the rift stage, are known at the margin of the Sakarya block, where they were dated as Hettangian (Altiner & Kogiy~it 1992; Kogiy~it 1998). Upwards, they pass into the marine sequence of the passive margin, which existed through the Jurassic and

186

V.G. KAZMIN & N. F. TIKHONOVA

Fig. 4. Early Jurassic reconstruction. The end of the first stage of extension (Norian-Early Pliensbachian). Abbreviations as in Figure l; legend as in Figure 2.

Early Cretaceous. According to Okay & Sahintiirk (1997), the marine transgression started in the Eastern Pontides in the Early Pliensbachian and spread from the south. A thick series of volcaniclastic sediments, intercalated with beds of 'ammonitico rosso' limestones and rare flows of the andesitic-basaltic lavas (the Kelkit Formation) was deposited on the subsiding passive margin. Geochemical parameters indicate an intraplate setting of volcanism. To the east, in the Transcaucasus, the earliest continental rift sediments are also dated as Hettangian (Panov 2000). The marine volcanic-sedimentary complex, extending from the Eastern Pontides, has been penetrated by drill-holes. Thin units of rhyolites and dacites are intercalated there with transgressive sediments of PliensbachianToarcian age (Lordkipanidze 1986). The Pliensbachian transgression probably coincided with the transition from rifting to spreading. It has been suggested that the passive margin was formed as a result of rifting of an unknown microcontinent from the Pontides (Okay & Sahintiirk 1997). In our reconstruction the rifted microcontinent included South Armenia, the Kir~ehir massif(?) and, perhaps, some other

blocks (Fig. 4). Behind the southward-migrating trench-arc system and continental fragments, the Izmir-Ankara-Sevan back-arc basin began to open. Evidence of an island arc formed on a rifted continental fragment comes from the northwestern part of the South Armenian block. According to Agamalyan (1987), on the western slope of the Tsachkunyak ridge in this area a thick (up to 6000 m) pile of lavas and volcaniclastic rocks, the Aparan Series, rests with a normal contact on the Precambrian basement. The basal unit, containing intercalations of shales and sandstones, was dated to the Toarcian-Aalenian. The overlying volcanic succession is only tentatively dated as Mid-Jurassic, although K/Ar determinations from lavas in the uppermost unit have yielded latest Jurassic to early Cretaceous ages. In the lower part of the succession basalts or andesitic-basalts are the main rock types; the upper part is built essentially of tufts, tuffites, lava-breccia and olistostromes. Pre-late Cretaceous intrusions of tonalites, quartz porphyry and granites cut the volcanic succession. Limited petrological and geochemical studies point to an island arc setting of volcanism. No data on Jurassic volcanic activity are known from the Kir~ehir massif.

BLACK SEA CAUCASUS BACK-ARC BASINS

187

Fig. 5. Mid-Jurassic reconstruction. The end of the second stage of extension (Late Pliensbachian-Early Aalenian). Abbreviations as in Figure 1; legend as in Figure 2.

The Izmir-Ankara-Sevan 'back-arc basin' extended to the southern periphery of the Rhodope-Serbomacedonian massif. A terminal western part of Neo-Tethys between the Serbomacedonian massif and the Pelagonian block (the Vardar or Axios basin) was studied recently in detail (Brown & Robertson 2004). It was demonstrated that in the Mid-Jurassic (or earlier) a continental fragment, the Paikon ridge, was rifted from the Serbomacedonian margin, following the onset of the eastward subduction of the Vadar oceanic crust. A volcanic arc formed on top of the Paikon ridge, while spreading and opening of the Guevgueli back-arc basin was in progress during the Mid-Late Jurassic. The time and style of evolution in this part of Greece are surprisingly similar to those deduced for the Izmir-Ankara-Sevan basin. The Guevgueli basin, or a branch of it, extended eastward to Thrace (NE Greece) to form the Jurassic-Early Cretaceous Circum-Rhodope belt (Magganas 2002). In the western branch of the Tauric basin shortening stopped and extension and opening(?)

was renewed. The renewed extension was marked by rifting on the northern (Moesian) margin of the basin in the Carnian-Norian (Dabovski & Georgiev 1996; Georgiev & Byrne 1995; Sinclair et al. 1997). At present, the Upper Triassic-Middle Jurassic sediments of this margin crop out in one of the nappes of Stara Planina, known as the Kotel zone. As demonstrated by geological and geophysical data, the sediments (clay-carbonate shales, flysch), accumulated at a south-facing rift margin, dominated by the Golitza master fault. The Golitza and associated normal faults dissected the Early-Mid-Triassic carbonate platform (Sinclair et al. 1997, p. 96, fig. 5); that is, the new continental slope was formed further north then the initial Early Triassic slope. Where the southern margin of the Tauric basin was located at the time is unknown. In any case, the Late Triassic (Lipachka) flysch spread as far south as the Strandja zone ($eng6r 1984; Banks 1997). There is no direct evidence of the situation in the eastern branch of the Tauric basin. However, termination of shortening and even reopening

188

V.G. KAZMIN & N. F. TIKHONOVA

/ L.Urmieh /'/

,@~\'~" , I , ~ ~

38

O. ~ ~ ~ " ~ k ' A '~

.Sanandajg /O' ' ~~B ~EJ

GT.Me

' /~ZZ27-1~ e n- O

/

,

,~,"~Agh 34

30

,

=0

48

~

26

o,u,77 54

60

KEY

i - ~ O u t c r o p of the Palaeotethyan complexes

[~-]

Nappe front

I ~

Mesozoic subduction-related complexes

~

Normal fault (a), strike-slipfault (b)

~

Rocks of the Shemshak Fm deformed and metamorphosed in the Middle Jurassic

Fig. 6. Index map of Iran (after Davoudzadeh & Schmidt 1984, with some additions). An, Anarek; Agh, Agharaband; Bi, Binalud; E, Esfahan; G, Golpaigan; GKF, Great Kevir Fault; Ha, Hamadan; Ma, Mahabad; MZF, Main Zagros Fault; To, Torbat; Yz, Yezd.

are likely, because at the same time rifting started in the Greater Caucasus to the north and at the Pontides-Transcaucasus margin to the south of the Tauric basin. Thus, the whole region, it seems, was affected by extension. In this geodynamic setting, late Triassic (Norian) volcanism on the Scythian platform is particularly

important. Following Carnian compressional deformation, sedimentation resumed there in the Norian in several widely dispersed subsiding basins, locally of irregular shape (e.g. the Nogaisk basin in the eastern Fore-Caucasus) and sometimes in reactivated early-mid-Triassic rifts (e.g. the East Manych grabens). Associated with

BLACK SEA CAUCASUS BACK-ARC BASINS shallow to moderately deep-water sediments, there are andesites, rhyolites, ignimbrites and volcaniclastic rocks. According to Nikishin et al. (2001), volcanic rocks have a calc-alkaline affinity, although Nikishin et al. noted that the data available are not sufficient for a reliable determination. Following V. E. Khain (1979), they viewed the Norian volcanic rocks as a subduction-related volcanic belt. As an alternative interpretation it can be suggested that volcanism and subsidence reflected regional extension and rifting in a wide area, far from the newly formed subduction zone, a situation also characteristic of the early Triassic extension. The Greater Caucasus basin is usually described as a long and narrow continental rift (Nikishin et al. 1998, 2001). However, several workers advocated limited spreading at the later stages of the basin's evolution (Adamia et al. 1987; Prutsky & Lavrishchev 1989; Dotduev 1989). The late Precambrian-Palaeozoic basement of the basin is exposed in the N W of the Greater Caucasus, forming a 'crystalline core' (Fig. 7a). On the southern flank this 'core' overthrusts a thick Mesozoic succession of the southern slope along the Main Caucasian thrust. To the SE and NE the basement complexes plunge below the overlying Jurassic sediments. Thus, the Main Caucasian thrust divides the basin into two sub-basins: southwestern and northeastern. The age of the basal Jurassic beds, transgressing the basement, is Sinemurian or younger, and this is usually accepted as the age of the Greater Caucasus basin (Nikishin et al. 2001; Panov 2000). However, in the deeper part of the northeastern sub-basin (e.g. along the northern tributaries of the Alazani river) Sinemurian microfossils occur within a monotonous shale sequences far upward stratigraphically from the unexposed basement. The lowermost Jurassic (Hettangian) succession is likely to be present there (Panov 2000). Hettangian and even Rhaetian sediments were described in the lower part of a continuous Jurassic succession in the southwestern sub-basin in Svanetia. A Triassic age for the lowermost part of the section was first established by the discovery of foraminifers (Saidova et al. 1988). Later a continuous succession, from the Rhaetian to Hettangian and Sinemurian, was proved by studies of palynomorphs (Adamia et al. 1990b). Shallow-water Upper Triassic sediments were also described in the westernmost part of the southwestern sub-basin (Krasnaya Polyana area) by Slavin (1958). Although contacts with the adjacent Jurassic rocks are tectonic, Triassic sediments may belong to the basal part of the Jurassic succession, as in Svanetia. There is enough

189

evidence, in our opinion, to date the onset of rifting in the Greater Caucasus basin as earliest Jurassic or even latest Triassic. The rift basin was bounded in the north by a master fault, which evolved along the Palaeozoic Tyrnyauz-Pshekish suture. Another major southdipping fault (the future Main Caucasus thrust) transected the basin obliquely, dividing it into two sub-basins (Fig. 7b and c). The associated monoclinal block had a maximum altitude in the NW, its surface gradually subsiding to the SE and NE. In general, the structure resembled that of the Baikal rift, where the diagonal monoclinal block of the Olkhon Island and Academician ridge separates the Northern and Central basins. In the southwestern sub-basin rifting propagated from the west, where the Greater Caucasus basin somehow connected with the Tauric basin (Fig. 5). The period of rifting lasted for about 2 2 M a (Rhaetian-Early Pliensbachian). The onset of spreading, in the Late PliensbachianToarcian, was marked by eruption of MORB in the axial zone of the southwestern sub-basin (Lordkipanidze 1980, 1986; Adamia et al. 1987), rapid subsidence and deposition of bathyal clays of the Tsiklauri horizon (Panov 2000), and transgression of the adjacent Scythian passive margin (Nikishin et al. 1998, 2001), which we interpret as a break-up unconformity. Spreading continued (perhaps sporadically) until the Early Aalenian, i.e. for about 14-15 Ma. The strip of newly formed crust was hardly wider than 100150 km, because the subsequent closure of the Greater Caucasus basin was not accompanied by supra-subduction volcanism. Accordingly, the spreading rate was about 1 cm a -1.

Mid-Jurassic stage (Figs 8 and 9) Major changes in the evolution of the marginal seas occurred in the Late Aalenian, when a period of compression began. As a result, almost all of the marginal basins were closed. The onset of compression in different basins was diachronous, from the Late Aalenian to Bathonian, perhaps as a result of the great complexity of the regional geological structure. Mid-Jurassic deformation was very important in Iran. In the Sanandaj-Sirjan zone (the Esfahan-Golpaygan-Hamadan area; see Fig. 6) the sediments of the Shemshak Formation were folded, slightly metamorphosed and intruded by diorite-granodiorite plutons with ages of 165175 Ma (Davoudzadeh & Schmidt 1984). Similar deformation, magmatism and metamorphism affected Late Triassic-Early Jurassic sediments in a wide belt between the East Alborz and Great Kevir fault and the Binalud ridge (National

190

V.G. KAZMIN & N. F. TIKHONOVA

Fig. 7. (a) Main geological features of the Caucasus. (b) Reconstruction of the Greater Caucasus basin for early Aalenian time (without scale). (c) Tentative cross-section. DS, Dizi Series; MT, Main Thrust; TPF, Tyrnyauz-Pshekish Fault. Iranian Oil Company 1975-1979; Lammerer et al. 1984; Lensch et al. 1984; Alavi 1996). It

appears that Central Iran was involved in MidJurassic deformation and magmatism, whereas

in the Central and Western Alborz this event resulted only in uplift and emergence (Delaloye et al. 1981; Alavi 1996). The whole of Iran was peneplaned in post-Mid-Jurassic time and then

BLACK SEA CAUCASUS BACK-ARC BASINS

191

Fig. 8. Mid-Jurassic reconstruction. The end of the first stage of compression (Late Aalenian-Bajocian). Abbreviations as in Figure 1; legend as in Figure 2.

covered by a diachronous transgression during Late Jurassic-Early Cretaceous time. The cause of the Mid-Jurassic deformation in Iran, as well as in the whole region, will be discussed later. Here, we wish to emphasize that deformation in Central-Eastern Iran may be of 'internal' origin and did not depend directly on events at its margins. It was noted that the Mid-Jurassic tectonomaganatic belt ran parallel to the northern margin of Iran, i.e. to the Alborz, and its origin was attributed to 'cratonization of the magmatic arc' (Davoudzadeh & Schmidt 1984). Sharp differences between the Alborz ('passive margin') and Central Iran ('arc') suggest that an important role in the 'cratonization' was played by the closure of the Intra-Iranian basin, as tentatively demonstrated in Figs 8 and 9. A good record of the closure of the eastern branch of the Tauric basin is preserved in the Central Pontides and the South Crimea (Yllmaz & Seng6r 1985; Bocaletti & Manetti 1988; Usta6mer & Robertson 1997; Nikishin et al. 1998,2001). The oceanic crust of the Tauric basin

was subducted below the Shatsky rise, on which the Bajocian volcanic arc was formed. Southward subduction of the Greater Caucasus basin should be excluded for two reasons: (1) in a narrow Greater Caucasus basin there was either no or very little oceanic crust present to generate arc magmatism lasting for about 4.0 Ma; (2) Jurassic deformation on the southern slope of the Greater Caucasus was strongly south-vergent, which is inconsistent with south-directed subduction below the Shatsky rise. In western Georgia, the Bajocian calc-alkaline arc volcanites and associated intrusions have been studied in detail on the Dzirula basement uplift (Lordkipanidze 1980, 1986; Adamia et al. 1990a) and also traced by onshore and offshore drilling to the adjacent part of the Shatsky rise. Further NW, Jurassic volcanic complexes are marked by characteristic magnetic anomalies on the Shatsky rise (Kazmin & Lobkovsky 2003). Finally, fragmemts of arcrelated complexes (lavas, tufts, volcaniclastic rocks and small dioritic plutons) crop out along the Black Sea Coast in the South Crimea, most probably in an allochthonous unit. Intrusive

192

V.G. KAZMIN & N. F. TIKHONOVA

Fig. 9. Mid-Jurassic reconstruction. The end of the second stage of compression (Bathonian). Abbreviations as in Figure 1; legend as in Figure 2.

rocks of diorite-granodiorite composition and large volcanic centres (seamounts?) developed synchronously in the back-arc region of the southern slope of the Greater Caucasus. The abrupt termination of volcanic activity at the end of the Bajocian marks the collision of the arc with the Pontides. In the Tauric Series of Crimea two stages of deformation are usually distinguished (Mazarorich & Mileev 1989; Nikishin et al. 1998, 2001). The early pre-Bajocian stage correlates with the onset of north-directed subduction in the Tauric basin. Following collision of the Pontides with the Shatsky rise volcanic arc and its western extension (volcanic complexes of the Crimea-Black Sea shore), a small remaining basin was compressed and finally deformed in the Bathonian. During this stage south-vergent thrusting of the Tauric Series took place. Usta6mer & Robertson (1997) demonstrated that in the Central Pontides north-vergent thrusts dominate the Kiire complex and can be attributed to accretion during closure of the Tauric basin. On the other hand, opposite-verging structures in the

Pontides and Crimea may have originated during a final stage of collision when convergence was directed to both sides of the relict basin towards the Shatsky rise and its western extention. As a result of collision, Crimea was welded to the Central Pontides and an orogenic belt was formed and then eroded. Products of erosion are known as the Demerji conglomerate in the South Crimea and the Muzun conglomerate in the Pontides. It was demonstrated long ago that the source of exotic blocks in the Demerji conglomerate was to the south (Chernov 1971), i.e. in the Pontides. In the Greater Caucasus basin the same two main stages of deformation are documented. During the pre-Bajocian stage the northeastern sub-basin was closed (Fig. 8). A system of southvergent thrusts formed within Jurassic sediments, resembling the structure of an accretionary prism (Panov 2000). The southwestern sub-basin (south of the MCT (Main Caucasus Thrust) remained undeformed and sedimentation there continued until the Bathonian. In pre-Callovian time the sedimentary pile was thrust southward (Panov &

BLACK SEA CAUCASUS BACK-ARC BASINS Prutsky 1983; Panov 2000) (Fig. 9). In front of the newly formed orogenic belt a narrow foredeep originated as a result of elastic bending of the lithosphere of the Shatsky rise. Late Jurassic-Early Cretaceous carbonates and siliclastic turbidites began to accumulate in an asymmetric trough. A very different evolutionary trend characterized the Transcaucasus massif and the adjacent (southeastern) portion of the Greater Caucasus basin. The Bajocian volcanic arc formed on the Transcaucasian massif; however, subducting lithosphere belonged there not to the Tauric but instead to the Izmir-Ankara-Sevan basin (Figs. 8 and 9). The arc was not affected by Mid-Jurassic deformation: magmatic activity continued uninterrupted through the Late Jurassic and part of the Neocomian (Lordkipanidze 1980, 1986; Kazmin et al. 1986). Sedimentation on the northern margin of the massif was also continuous, indicating that the southeastern part of the Greater Caucasus basin was not closed during Mid-Jurassic inversion. Further east, this part of the basin extended through the South Caspian to the Kopetdag basin, where no Mid-Jurassic deformation is reported and sedimentation continued uninterrupted from the Mid- to Late Jurassic (Lensch et al. 1984). Closure and deformation of the western branch of the Tauric basin are dated as postMid-Jurassic and pre-Cenomanian (Banks 1997; Okay et al. 2001). The youngest rocks in the north-vergent nappes of the Strandja zone have a mid-Jurassic age. A 155 Ma Rb-Sr age (biotite whole-rock) from the metagranitic basement of the Zwezdets nappe in Strandja dates regional metamorphism as Oxfordian-Kimmeridgian (Okay et al. 2001). Two events may provide additional information on the time of deformation: (1) At the northern margin of the basin (the Kotel zone) a transition from basinal to shallow-water facies took place in the Callovian (Georgiev & Byrne 1995); (2) in front of the Strandja nappes the Nish-Trojan foredeep evolved in the Late Jurassic and Neocomian (Okay et al. 2001). Its position and age are similar to the foredeep at the southern slope of the Greater Caucasus (see above). In both areas compressional deformation occurred penecontemporaneously at the end of the Mid-Jurassic to the beginning of the Late Jurassic. No precise data are available on mid-Jurassic deformation in the North Dobrogea. Indirect evidence comes from studies by Gradinaru (1988, 1995), who documented opening of the rift basin along the Pechenega-Camena fault in a transtensional setting and simultaneous transgression on the adjacent part of the

193

Moesian platform in the Late Bathonian. These events apparently postdate the closure of the North Dobrogea branch of the Tauric basin. Accordingly, the time of its closure is pre-Late Bathonian, i.e. probably simultaneous with the final deformation in South Crimea.

Discussion Four major epochs can thus be distinguished in the early Mesozoic history of the northwestern margin of Tethys. The first epoch lasted for about 20-22Ma, from the Scythian to the Early Carnian. This was a time of spreading and opening of the Tauric basin and associated basins of the North Dobrogea and the Greater Caucasus. Spreading in the Tauric basin was preceded by rifting, but evidence of this event is very limited. In the Istanbul zone of the Western Pontides there is the north-south-trending Kocaeli basin, which may represent a failed rift associated with opening of the K/ire (Tauric) basin (Usta6mer & Robertson 1993, 1997). The Kocaeli basin is filled by red clastic deposits with alkaline lavas at the base (Late Permian?) and the marine succession is dated from the Early Scythian to Carnian. Continental rifting of the Scythian and Turonian platforms also commenced in the Late Permian and evolved in the Early-Mid-Triassic (Orel 2001; Glumov et al. 2004). The Late Permian is provisionally accepted as the time of initial rifting of the Tauric basin. The Tauric basin opened behind the northdipping Palaeotethyan subduction zone (Pickett & Robertson 1996; Usta6mer & Robertson 1993, 1997; Robertson 2002). Subduction commenced at the southern margin of the PontidesTranscaucasus microcontinent after its collision with the Scythian margin in the Vis6an. The time of collision is constrained by the synchronous development of Serpukhovian-Bashkirian molasse on the Scythian margin and in the Transcaucasus (E. V. Khain 1979; Belov 1981). Intrusions of granodiorites and granites, together with subaerial volcanism in the Transcaucasus (c. 320-250 Ma) were related to late Palaeozoic northward subduction below this massif (Adamia et al. 1982, 1989). A question is why the back-arc extension only began in the Late Permian? The Late Palaeozoic evolution of the active margin was interrupted in the Early Permian by a strong compressional event. At that time north-vergent thrusting affected Palaeozoic sediments of the Fore-Caucasus and the Karpinsky swell (Volozh et al. 1999; Glumov et al. 2004). Early Permian compression is known in other parts of the active margin of southern

194

V . G . K A Z M I N & N. F. T I K H O N O V A

;'2

[...,

r.~

I r,~

O

"6 . ,,...,

,.o

. ,...~

BLACK SEA CAUCASUS BACK-ARC BASINS Eurasia. Kazmin (2002) noted that this event correlates with rifting of the Gondwana passive margin and formation of a new spreading centre behind a chain of separated microcontinents. It was speculated that compression at the active margin was caused by a trench-mid-ocean ridge collision and that rifting resulted from the slabpull transmitted to the passive margin at a time, when there was no spreading centre in Tethys. According to this idea, subduction resumed at the margin of the Pontides-Transcaucasus massif in the Late Permian and was immediately followed by back-arc extension, perhaps as a result of a strong intensive roll-back effect. The Tauric basin was partly inverted in the Carnian, when fragments of Cimmeria collided with the active margin. The main fragment of Cimmeria, Iran, had a western extension as a chain of blocks, including South Armenia, the Kir~ehir massif and the Kargi platform. Perhaps because of the small size of these blocks the effect of collision in the Tauric basin was relatively mild; the basin was shortened but not closed. Small-scale shortening explains the lack of Carnian arc magrnatism. In Figure 3, underthrusting or subduction is shown at the northern margin of the Tauric basin. However, this interpretation is arbitrary. South-directed underthrusting of ophiolites and sediments of the Kfire (Tauric) basin was described by Usta6mer & Robertson (1997) in their reconstruction of Central Pontides. More information is needed to determine if this structure formed in the Carnian or much later. Following accretion of Cimmerian fragments, a subduction zone originated south of the accreted microcontinents, and a new phase of extension in a back-arc area began. The main manifestations of this extension include rifting in Iran in Carnian-Norian time (Davoudzadeh & Schmidt 1984); rifting and formation of the Golitza passive margin (Kotel zone) in C a r n i a ~ Norian time on the southern periphery of the Moesian platform (Dabovski & Georgiev 1996; Sinclair et al. 1997), and opening of the Greater Caucasus basin in the Latest Triassic(?)-Early Jurassic. However, these events were of secondary importance compared with rifting and opening of the Izmir-Ankara-Sevan basin between the Pontides-Transcaucasus and the fragments of Cimmeria, which started in Hettangian time (Altiner & I~ogy~it 1992; Koqy~it 1998; Panov 2004). Continental rifting was followed by transgression and deposition of neritic then pelagic carbonates. In the Eastern Pontides and Transcaucasus subsidence and an extensive north-directed transgression started in the Early Pliensbachian (Lordkipanidze 1986; Okay &

195

Sahintfirk 1997), resulting, in our opinion, with the break-up of the continental lithosphere and the onset of spreading in the Izmir-AnkaraSevan basin. The width of the newly formed back-arc basin is unknown. However, if the opening continued until the mid-Cretaceous, i.e. to the onset of subduction at the Pontide margin, its width could be very considerable. It cannot be excluded that in the narrow western part of Neotethys (the Vardar, or Axios basin) migrating island arcs collided with the northwestern Neotethyan passive margin, as suggested by Dercourt et al. (1986). New data do not contradict this suggestion (Brown & Robertson 2004). In our reconstructon the Tauric basin evolved continuously from the Late Carnian to the Early Aalenian, i.e. for about 45 Ma. According to Nikishin et al. (1998, 2001), this uninterrupted evolution was punctured by an episode of compression and inversion in the RhaetianHettangian. No convincing evidence of this event can be found in the western or eastern branches of the Tauric basin. In the former, sedimentation was continuous, at least from the CarnianNorian to the Mid-Jurassic (Dabovsky & Georgiev 1996; Banks 1997; Okay et al. 2001). In the latter, no major deformation is known inside the Tauric series and its counterparts in the Central Pontides. A suspected stratigraphic lacuna in the Tauric series, corresponding to the Rhaetian-Hettangian interval (Nikishin et al. 1998, 2001); if present, this by no means proves the closure and inversion of the basin, but may reflect erosion or non-deposition on the continental slopes. As shown above, the Greater Caucasus basin originated in the latest Triassic-earliest Jurassic, i.e. at the time of the problematic inversion. We conclude that no Rhaetian-Hettangian inversion affected the Tauric basin. A period of compression and closure of back-arc basins began in the Late Aalenian and continued for about 8.5 Ma until the end of the Bathonian. As a result, the Tauric and Greater Caucasus basins closed and fold belts formed in their place. The process was accompanied by pre- to post-collisional magmatic activity. South-vergent structures dominated the eastern Tauric and Greater Caucasus basins, whereas in the western Tauric basin the structure was north-vergent. The change of polarity coincides with the West Crimea fault. The evolution of the part of the Tauric basin between the Sakarya and Istanbul blocks is still a matter of discussion. A controversy exists concerning the timing of suturing of these two blocks along the severely deformed ArmutluAlmacik zone. According to Okay et aL (1994),

196

V.G. KAZMIN & N. F. TIKHONOVA

the blocks collided in the Early Eocene, thus putting an end to the opening of the Western Black Sea basin behind the Istanbul zone. Ydmaz et al. (1997) associated the formation of the ArmutluAlmacik zone with closure of a branch of Neo-Tethys, the Intra-Pontide Ocean, and dated this event and the emplacement of ophiolites as 'post-Turonian and pre-Late Campanian'. Recent detailed studies (Robertson & Ustaomer 2004) confirmed the existence of a discrete Intra-Pontide oceanic basin that opened in the Triassic and closed in the Turonian. However, Elmas & Yi~itbas (2001) argued that the Sakarya and Istanbul blocks were welded together in prelate Jurassic time and that the ophiolites were emplaced along younger strike-slip faults. It is possible that the Intra-Pontide Ocean was part of the Tauric (Kfire) basin, situated between the Istanbul and Sakarya blocks (fig. 2; see also Usta6mer & Robertson 1993, p. 234, fig. 10). It was possibly closed in the Mid-Jurassic together with the whole Tauric basin. However, one cannot exclude that it reopened in the Late Jurassic in connection with dextral motion on the Pechenega-Camena fault in a transtensional setting (Gradinaru 1995). Comparison of the reconstructions in Figures 5 and 9 shows that the minimum Mid-Jurassic shortening along the Pontides-Greater Caucasus transect was about 300-400 km at a convergence rate of 3.5-4.5 cm a -1. The motion of AfricaArabia relative to Eurasia at this time was essentially left-lateral (Savostin et al. 1986), and the convergence between the two plates totals only a few hundred kilometres (Dercourt et al. 1985, 1993). Most, or all, of this convergence, was probably compensated by the closure of the back-arc basins. What caused the compression at the northern margin of Tethys in the Mid-Jurassic? As no collision with continental blocks occurred at that time, one must look for a tentative explanation at the remote plate boundaries. At the beginning of the Mid-Jurassic, spreading in the Izmir-Ankara-Sevan basin had already been active for about 16-17Ma (from the Pliensbachian to Early Aalenian). At a rate of 5-6cm a -1 the width of the basin reached 800-1000 km (Fig. 10; also see Figs. 8 and 9). According to global reconstructions, the width of Tethys in its westernmost part was about 20002200 km (Golonka et al. 1996; Golonka 2000; Dercourt et al. 2001). As a result, the southwardmigrating arc system at the southern front of the Izmir-Ankara-Sevan basin was able to collide with a Tethyan mid-ocean ridge. When subduction was temporarily blocked, convergence

between the main plates was compensated by shortening and closure of the back-arc basins. Compression at the northern Tethyan margin terminated at the end of the Bathonian, when subduction at the southern front of the Izmir-Ankara-Sevan basin was renewed.

Conclusions Evolution of the early Mesozoic back-arc basins in the Black Sea-Caucasus region was governed by several factors, as follows. (1) Two major periods of extension and opening of the Tauric and associated basins (Permian(?)-Early Triassic and Late Triassic-Early Jurassic) immediately followed formation of new subduction zones. This implies that the initiation of subduction was succeeded by the rapid sinking of a dense slab composed of the old oceanic lithosphere at the margin of the Palaeozoic or PermianTriassic ocean. Extension created by resulting roll-back affected a wide (up to 1000 km) area of the back-arc region. (2) Partial inversion of the Tauric and associated basins in the Carnian was related to closure of Palaeo-Tethys and collision of the Cimmerian fragments with the active margin of Eurasia. (3) The major compressional event in the MidJurassic resulted in deformation and closure of the Tauric and Greater Caucasus backarc basins. This event probably coincided with ridge-trench collision at the southern margin of the opening Izmir-Ankara-Sevan basin. For a period when the intra-oceanic subduction zone was blocked, AfricaEurasia convergence was compensated by shortening and closure of the back-arc basins. The authors are greatly indebted to A. H. F. Robertson for discussion of the manuscript and help with new information. A. Nikishin is also thanked for reviewing the paper. This work was financially supported by the Russian Fund for Fundamental Research (RFFI), Grant 04-05-64184.

References ADAMIA, S. A. 1968. Pre-Jurassic Complex of Caucasus. Metsnieraba, Tbilisi (in Russian). ADAMIA, S. A., ASANIDZE,B. Z. • PECHERSKY,D. M.

1982. Geodynamics of Caucasus (attempt of palinspastic reconstructions). In: MURATOV,M.V. & ADAMIA,S. A. (eds) Problems of Geodynamics of the Caucasus. Nauka, Moscow, 13-21 (in Russian).

BLACK SEA CAUCASUS BACK-ARC BASINS ADAMIA, S. A., BERIDZE, M. A., KIPIANI, YA. R., KULASHVILI, S. I. & LORDKIPANIDZE,M. B. 1987. The problem of the alpine geodynamics of the Greater Caucasus. In: MILANOVSKY, F~. E. & KORONOVSKY,N. V. (eds) Geology and the Mineral Resources of the Greater Caucasus. Nauka, Moscow, 55-61 (in Russian). ADAMIA, S. A., GABUNIA, G. L., KUTELIA, Z. A., KHUTSISHVILI, O. D. & TSIMAKURIDZE, G. M. 1989. The typical features of tectonics of the Caucasus. In: BELOV, A. A. & SATIAN, M. A. (eds) Geodynamics of the Caucasus. Nauka, Moscow, 3-15 (in Russian). ADAMIA, S. A., KUTELIA, Z. A., PLANDEROVA, E. & KHUTSISHVILI, O. D. 1990a. Rhaetian-Hettangian sediments of the Dizi Series of Svanetia (Greater Caucasus). Doklady Akademii Nauk SSSR, 313, 395-398 (in Russian). ADAMIA, S. A. LORDKIPANIDZE,M. B., BERIDZE, M. A., KOTETISCHVILI,E. & KUTELIA,Z. 1990b. Paleogeography of the Ukrainian Carpathians, the Crimea, and the Caucasus. In: RAKES, M. DERCOURT, J. & NAIRN, A. E. M. (eds) Evolution of the Northern Margin of Tethys, III, part I. M6moires de la Soci6t6 G6ologique de France, Nouvelle S6rie, 154, 123-146. AGAMALYAN,V. A. 1987. Mesozoic accretionary complex (the Aparan Series) in the Tsachkunyatski Ridge, Armenian SSR. Izvestiya Akademii Nauk Armyanskoi SSR, 40, 13-24. AISTOV, L., MELNIKOV,B., KRIVYAKIN,B., MOROZOV, L. & KIRISTAEV, V. 1984. Geology of the Khur Area (Central Iran).Vsesoyuznoe Obshchestvo Technoexport, Moscow. ALAVI, M. 1996. Tectonostratigraphic synthesis and structural style of the Alborz mountain system in northern Iran. Journal of Geodynamics, 21, 1-33. ALTINER, D. & KOGYIGIT, A. 1992. The Jurassic-early Cretaceous paleogeographic evolution of the south and northwest Anatolia. Turkish Journal of Earth Science, 1, 1-11. ASANIDZE, B. Z. & PECHERSKY, D. M. 1979. Results of paleomagnetic studies of Jurassic rocks in the Caucasus. Izvestiya Academii Nauk SSSR, Fizika Zemli, 10, 91-115 (in Russian). BANKS, C. 1997. Basins and thrust belts of the Balkan coast of the Black Sea. In: ROBINSON, A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG) Memoirs, 68, 115-128. BANKS, C. & ROBINSON, A. G. 1997. Mesozoic strikeslip back-arc basins of the Western Black Sea region. In: ROBINSON A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG) Memoirs, 68, 53-62. BARR, S. R., TEMPERLEY, S. & TARNEY, J. 1999. Lateral growth of the continental crust through deep level subduction-accretion: a re-evaluation of central Greek Rhodope. Lithos, 46, 69-94. BELOV, A. A. 1981. Tectonic Evolution of the Alpine Fold Belt in the Paleozoic. Nauka, Moscow (in Russian).

197

BERBERIAN, M. & KING, G. C. P. 1981. Towards a paleogeography and tectonic evolution of Iran. Canadian Journal of Earth Science, 18, 210-265. BOCALETTI, M. & MANETTI, P. 1988. The main unconformities and tectonic events in the Pontides. Bolletino di Geofisica Teorica ed Applicata, 30, 9-16. BOUL1N, J. 1990. Neocimmerian events in Central and Western Afghanistan. Tectonophysics, 175, 285315. BROWN, S. A. M. & ROBERTSON, A. H. F. 2004. Evidence for Neotethys rooted within the Vardar suture zone from the Voras Massif, northernmost Greece. Tectonophysics, 381, 143-173. CHERNOV, V. G. 1971. Composition of Upper Jurassic conglomerates of the Demerji Mt., Crimea. Vestnic Moskovskogo Gosudarstvennogo Universiteta (MGU), Seriya Geologiya, 2, 18-28 (in Russian). CHERVEN, V. B. 1986. Tethys marginal sedimentary basin in western Iran. Geological Society of America Bulletin, 97, 516-522. DABOVSKI, C. & GEORGIEV, G. 1996. Rifting in the south-eastern Moesian Platform margin during the Mesozoic-Paleogene evolution. In: Comparative Evolution of Peri-Tethyan Rift Basins IGCP Project 369, Third Annual Meeting, Abstracts Volume, Cairo, 16-17. DAUKEEV, S. ZH., YATSKEVICH, B. A., VAN FUTUN, et al. (eds) 2002. Atlas of Lithology Paleogeographic, Structural, Palinspastic and Geoecological Maps of Central Eurasia. Scientific Research Institute of Natural Resources, YUGGEO, Almaaty. DAVOUDZADEH,M. & SCHMIDT,K. 1981. Contribution to the paleogeography and stratigraphy of Upper Triassic to Middle Jurassic of Iran. Neues Jahrbuch fiir Geologie und Paldontologie Abhandlungen, 162(2), 137-163. DAVOUDZADEH, M. & SCHMIDT, K. 1984. A review of the Mesozoic paleogeography and paleotectonic evolution of Iran. Neues Jahrbuch fiir Geologie und Paliiontologie, Abhandlungen, 108(2-3), 183-208. DELALOYE, M., JENNY, J. & STAMPFLI, G. 1981. K-Ar dating in the eastern Elburz (Iran). Tectonophysics, 79, 2%36. DERCOURT, J., ZONENSHAIN, L. P., RICOU, L.-E., et al. 1985. Pr6sentation de 9 cartes pal6og6ographiques au 1:20000000 s'&endant de l'Atlantique au Pamir pour la p6riode du Lias h l'Actuel. Bulletin de la Socidte Gdologique de France, 1(5), 637-652. DERCOURT, J., ZONENSHAIN,L. P., RICOU, L.-E., et al. 1986. Geological evolution of the Tethys belt from Atlantic to Pamirs since Liassic. Tectonophysics, 123(1M), 241-315. DERCOURT, J., RICOU, L.-E., ADAMIA, S. A., et al. 1990. Anisian to Oligocene paleogeography of the European margin of Tethys (Geneva to Baku). In: RAKOS, M., DERCOURT,J. & NAIRN, A. E. M. (eds) Evolution of the Northern Margin of Tethys, III, part 1. M6moires de la Soci6t6 G6ologique de France, Nouvelle S6rie, 154, 159-190. DERCOURT, J., RICOU, L.-E. & VRIELVNCK, B. (eds) 1993. Atlas Tethys. Palaeoenvironmental Maps. Gauthier-Villars, Paris.

198

V.G. KAZMIN & N. F. TIKHONOVA

DERCOURT, J., GAETANI, M., VRIELINK,B., et al. (eds) 2001. Atlas Peri-Tethys. Paleogeographical Maps. CCGM/CGMW, Paris. DOTDUEV, S. I. 1989 Mesozoic-Cainozoic geodynamics of the Greater Caucasus. In: BELOV, A. A. & SATIAN, M. A. (eds) Geodynamics of the Caucasus. Nauka, Moscow, 82-92 (in Russian). ELMAS, A. & YI(~ITBA~, E. 2001. Ophiolite emplacement by strike-slip tectonics between the Pontide Zone and the Sakarya Zone in northwestern Anatolia, Turkey. International Journal of Earth Sciences, 90, 257-269. GEORGIEV, G. & BYRNE, P. 1995. South Moesian Triassi~Jurassic rift basin. In: Comparative Evolution of Peri-Tethyan Rift Basins. IGCP Project 369, Second Annual Meeting. Abstracts Volume, Mamaia, Romania, 14-15. GLUMOV, I. F., MALOVITSKY,YA. P., NOVIKOV,A. L. & SENIN, B. V. 2004. Regional Geology and Hydrocarbon Potentials of the Caspian Sea. Nauka, Moscow. GOLONKA, J. 2000. Cambrian-Neogene Plate Tectonic Maps. Wydawnictwo uniwersytety Jagiellofiskiego, Krakow. GOLONKA, J., EDRICH, M. E., FORD, D. W., PAUKEN, R. J., BOCHAROVA, N. Y. & SCOTESE, C. R. 1996. Jurassic paleogeographic maps of the world. Museum of Northern Arizona Bulletin, 60, 1-5. GRADINARU, E. 1988. Jurassic sedimentary rocks and bimodal volcanics of the Carjelari-Camena outcrop belt: evidence for a transtensile regime of the Pecenega-Camena fault. Studiisi Cercetary de Geologie, Geofizika, Geografie. Serie Geologie, 33, 97-121. GRADtNARU, E. 1995. Mesozoic rocks in North Dobrogea: an overview. In: Field Guidebook." Central and North Dobrogea. Geological Institute of Romania, Bucharest, 17-28. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOULOS,D. 2006. Late Proterozoic and Silurian basement units within the Serbo-Macedmian Massif, northern Greece: the significance of terrane accretion in the Hellenides. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 35-50. JONES, C. E., TARNEY, J., BAKER,J. H. & GEROUKI, F. 1992. Tertiary granitoids of Rhodope, Northern Greece: magmatism related to extensional collapse of the Hellenic orogen? Tectonophysics, 210, 295-314. KAZMIN, V. G. 1990. Early Mesozoic reconstruction of the Black Sea-Caucasus region. In: RAK0S, M., DERCOURT, J. & NAIRN, A. E. M. (eds) Evolution of the Northern Margin of Tethys, III, part 1. M6moires de la Soci6t6 G6ologique de France, Nouvelle S6rie, 154, 147-158. KAZMIN, V. G. 2002. The late Paleozoic to Cainozoic intraplate deformation in North Arabia: a response to plate boundary forces. In: CLOETINGH, S. & BEN-AVRAHAM, Z. (eds) European Geophysical Union (EGU) Stephan Mueller Special Publication Series, 2, 123-138.

KAZMIN, V. G., & LOBKOVSKY,L. I. 2003. Geological structure and evolution of the Shatsky Rise. In: Laverov, N. P. (ed.) Actual Problems of Oceanology. Nauka, Moscow, 221-243 (in Russian). KAZMIN, V. G. & NATAPOV, L. M. (eds) 1998. Paleogeographic Atlas of Northern Eurasia. Institute of Tectonics of Lithospheric Plates, Moscow (CD-ROM). KAZMIN, V. G. & SBORSHCHIKOV, I. M. 1989. Late Paleozoic-Early Mesozoic deformations in the Caucasus and their place in the evolution of Tethys. In: BELOV, A. A. & SATIAN, M. A. (eds) Geodynamics of the Caucasus. Nauka, Moscow, 46-54 (in Russian). KAZMIN, V. G., SBORSHCHIKOV,I. M., RICOU, L.-E., ZONENSHAIN, L. P., BOULIN, J. & KNIPPER, A. L. 1986. Volcanic belts as markers of the MesozoicCenozoic active margin of Eurasia. Tectonophysics, 123(1-4), 123-152. KAZMIN, V. G., SCHEIDER, A. A. & BULYCHEV,A. A. 2000. Early stages of evolution of the Black Sea. In: BOZKURT, E., WINCHESTER,J. A. & PIPER, J. A. D. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 499-513. KHAIN, V. E. 1979. North Caucasus-TurkmeniaNorth Afghanistan late Triassic volcanic-plutonic belt and opening of the northern branch of Tethys. Doklady Akademii Nauk SSSR, 249(5), 1190-1192 (in Russian). KHAIN, E. V. 1979. Ophiolites and nappe structure of the Peredovoi Range of the northern Caucasus. Geotektonika, 4, 63-80 (in Russian). KOCYt~IT, A. A. 1998. A geotraverse through the so-called 'Ankara m61ange' between Elmada~ and Bedesten, Ankara, Turkey. In: Guidebook of the Third International Turkish Geology Symposium. Middle East Technical University (METU), Ankara, 1-10. KOZHOUKHAROVA,E. 1996. New data for the geologic position of the Precambrian ophiolitic association in the Rhodope massif. Comptes Rendus de l'Acad~mie des Sciences de Bulgare, 49(1), 57-60. KRONBERG, P., MEGER, W. & PILGER, A. 1970. Geologie der Ril~Rhodope Masse zwischen Strimon und Nestos (Nordgriechenland). Beihefte Geologische Jahrbuch. 88, 133-180. LAMMERER, B., LANGHEINRICH,G. & MANUTCHEHRDANAI, M. 1984. Geological investigation in the Binalud Mountains (NE Iran). Neues Jahrbuchfiir Geologie und Palfiontologie, 168(2-3), 269-277. LAUER, J.-P. 1984. Geodynamic evolution of Turkey and Cyprus based on paleomagnetic data. In: DIXON, J. E. & ROBERTSON,A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 483491. LENSCH, G., SCHMIDT, K. & DAVOUDZADEH,M. 1984. Introduction to the geology of Iran. Neues Jahrbuch flit Geologie und Paldontologie, Abhanlungen. 168(2-3), 155-164. LORDKIPANIDZE, M. B. 1980. Alpine Volcanism and Geodynamics of the Central Segment of the Mediterranean Fold Belt. Metsnieraba, Tbilisi (in Russian).

BLACK SEA CAUCASUS BACK-ARC BASINS LORDKIPANIDZE,M. B. 1986. Mesozoic-Cainozoic volcanism and geodynamics of the central segment of the Alpine-HimalayanfoM belt. D. Sc. thesis, Geologicheskii Institut Akademii Nauk Gruzinskoy SSR, Tbilisi (in Russian). MAGGANAS, A. C. 2002. Constraints on the petrogenesis of Evros ophiolite extrusives, NE Greece. Lithos, 65, 165-182. MAZAROVICH, O. A. & MILEEV, V. S. (eds) 1989. Geological Structure of the Kacha Uplift, Mountainous Crimea. Moscowskii Gosudarstenny Universitet (MGU), Moscow (in Russian). NATIONAL IRANIAN Om COMPANY 1975-1979. Geological map of lran. Scale 1.'1 000 000. (Sheets 1 4 ) . National Iranian Oil Company, Tehran. NIKISHIN, A. M., CLOETINGrl, S., BRUNET, M.-F., STEPHENSON, R. A., BOLOTOV, N. & ERSHOV, A. 1998. Scythian Platform, Caucasus and Black Sea region: Mesozoic-Cenozoic tectonic history and dynamics. In: CRASQUIN-SOLEAU,S. & BARRIER,E. (eds) Peri-Tethys Memoir 3: Stratigraphy and Evolution of Peri-Tethyan Platforms. Mrmoires de la Musrum National de l'Histoire Naturelle, 177, 163-176. NIKISHIN, A. M., ZIEGLER, P. A., PANOV, D. I., et al. 2001. Mesozoic and Cainozoic evolution of the Scythian Platform-Black Sea-Caucasus domain. In: ZIEGLER, P. A., CAVAZZA,W., ROBERTSON, A. n . F. ~; CRASQUIN-SOLEAU, S. (eds) Peri-Tethys Memoir 6: Peri-Tethyan Rift~Wrench Basins and Passive Margins. Mrmoires de la Musrum National de l'Histoire Naturelle, 186, 295-346. OKAY, A. I. & ~AH1NTURK, O. 1997. Geology of the Eastern Pontides. In: ROBINSON, A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG), Memoirs, 68, 291-311. OKAY, A. I., ~ENGrR, A. M. C. & Gt3RUR, N. 1994. Kinematic history of the opening of the Black Sea and its effect on the surrounding regions. Geology, 22, 267-270. OKAY, A. L., SATIR, M., TOYSOZ, AKY~Z, S. & CHEN, F. 2001. The tectonics of the Strandja massif: late-Variscan and mid-Mesozoic deformation and metamorphism in the northern Aegean. International Journal of Earth Sciences, 90, 217-233. OREL, V. E. (ed.) 2001. Geology and Hydrocarbon Potentials of Fore-Caucasus. GEOS, Moscow (in Russian). PANOV, D. I. 2000. Tectonic structure of Jurassic terrigenous complex of the Greater Caucasus. In: General Problems of Tectonics, Tectonics of Russia, 33rd Conference on Tectonics. GEOS, Moscow, 387-389 (in Russian). PANOV, D. I. & PRUTSKY, N. I. 1983. Stratigraphy of Lower-Middle Jurassic sediments of the northwestern Caucasus. Byulleten Moskovskogo Obshchestva Ispytateley Prirody, ( MOIP), Otdelenie Geologicheskoe, 58(1), 94-112 (in Russian). PECHERSKY, D. M. & SAFRONOV, V. A. 1993. Palinspastic reconstruction of the mountainous Crimea position in the Mid-Jurassic-Early Cretaceous

199

according to paleomagnetic data. Geotektonika, 1, 96-105 (in Russian). PICKETT, E. A. & ROBERTSON, A. H. F. 1996. Formation of the late Palaeozoic-early Mesozoic Karakaya Complex and related ophiolites in NW Turkey by Palaeotethyan subduction-accretion. Journal of the Geological Society, London, 153(6), 995-1009. PRUTSKY, N. I. & LAVRISHCHEV,V. A. 1989. Northwestern Caucasus in the Mesozoic. In: BELOV, A. A. • SATIAN, M. A. (eds) Geodynamics of the Caucasus. Nauka, Moscow, 92-98 (in Russian). 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. & USTAOMER, T. 2004. Tectonic evolution of the Intra-Pontide suture zone in the Armutlu Peninsula, NW Turkey. Tectonophysics, 381, 175-209. RUTTNER, A. W. 1984. The Pre-Liassic basement of the Eastern Kopet Dag Range. Neues Jahrbuch for Geologie und Palgiontologie Abhandlungen, 168 (2-3), 256-268. SAIDOVA,K. M., KAZMIN,V. G., SBORSHCHIKOV,I. M., MATVEENKOV, V. V. & IVANOV, M. K. 1988. New data on stratigraphy of the Dizi Series of Svanetia. Doklady Akademii Nauk SSSR, 302(5), 407-409 (in Russian). SANDULESCU, M. 1995. Dobrogea within the Carpathian foreland. In: IGCP Project 369 Field Guidebook: Central and North Dobrogea. Geological Institute of Romania, Bucharest, 1-4. SARIBUDAK,M. 1989. New results and a paleomagnetic overview of the Pontides in northern Turkey. Geophysical Journal, 99, 521-531. SAVOSTIN, L. A., SIBUET, J. C., ZONENSHAIN,L. P., LE PICHON, X. & ROULET, M. J. 1986. Kinematic evolution of the Tethys belt from the Atlantic Ocean to the Pamirs since the Triassic. Tectonophysics, 123, 1-35. ~ENGOR, A. M. C. 1979. Mid-Mesozoic closure of Permo-Triassic Tethys and its implications. Nature, 29, 590-593. ~ENGrR, A. M. C. 1984. The Cimmeride Orogenic System and Tectonics of Eurasia. Geological Society of America Special Papers, 195. ~ENGrR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 131-241. SINCLAIR, H. D., JURANOV, S. G., GEORGIEV, G., BYRNE, P. & MOUNTNEY, N. P. 1997. The Balkan thrust wedge and foreland basin of Eastern Bulgaria: structural and stratigraphic development. In: ROBINSON, A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG), Memoirs, 68, 91-114. SLAVIN, V. I. 1958. New data on geological structure of Krasnaya Polyana and adjacent portion of the Main Ridge and southern slope of the Greater Caucasus. Izvestiya Vysshikh Uchebnykh Zavedeniy (VUZ), Geologiya i Razvedka, 6, 31-45 (in Russian).

200

V.G. KAZMIN & N. F. TIKHONOVA

SOFFEL, A. C. & FORSTER, A. G. 1984. Polar wander path of the Central-East Iran microplate including new results. Neues Jahrbuch fiir Geologie und Paldontologie, Abhandlungen, 168(2-3), 165-172. SOMIN, M. L. & BELOV, A. A. 1967. Stratigraphic subdivision of the Dizi Series of Svanetia (Central Caucasus). Byulleten Moskovskogo

active southern continental margin of Eurasia: evidence from the Central Pontides (Turkey) and adjacent regions. Geological Journal, 28, 219-238. USTA6MER, T., & ROBERTSON, A. 1997. Tectonicsedimentary evolution of the Northern Tethyan margin in the Central Pontides of Northern Turkey. In: ROBINSON, A. G. (ed.) Regional and

Obshchestva Ispytateley Prirody (MOIP), Seriya Geologicheskaya, 42(1), 4048 (in Russian).

Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum

STAMPFLI, G. M. 1996. The Intra-Alpine terrain: a Paleotethyan remnant in the Alpine Variscides. Eclogae Geologicae Helvetiae, 89, 13-42. 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., DE BONO, A. & VAVASSIS, [. 1998. Late Paleozoic, Early Mesozoic plate tectonics of the Western Tethys. Bulletin of the Geological Society of Greece, 32, 113-120. TARI, G., D~CEA, O., FAULKERSON,J., GEORGIEV, G., POPOV, S., STEFONESCU, M. & WEIR, G. 1997. Cimmerian and alpine stratigraphy and structural evolution of the Moesian Platform (Romania/ Bulgaria). In: ROBINSON, A. G. (ed.) Regional and

Geologists (AAPG) Memoirs, 68, 255-290. VOLOZH, YU. A., ANTIPOV, M. P., LEONOV, YU. G., MOROZOV, A. F. & YUROV, YU. A. 1999. Structure of the Karpinsky Ridge. Geotektonika, 1, 28-43 (in Russian). WESTPHAL, M., BAZENOV, M. I., LAUER, J. P., PECHERSKY,D. M. & SIBUET,J. C. 1986. Paleomagnetic implications on the evolution of the Tethys belt from the Atlantic Ocean to Pamir since Triassic. Tectonophysics, 123, 37-82. WHITNEY, D. L., TEYSSIER,C., DILEK, Y. & FAYON, A. K. 2001. Metamorphism of the Central Anatolian crystalline complex, Turkey: influence of orogennormal collision vs. wrench~tominated tectonics on P-T-t paths. Journal of Metamorphic Geology, 19, 41 IM33. YILMAZ, Y. & ~ENGOR,A. M. C. 1985. Palaeo-Tethyan ophiolites in northern Turkey: petrology and tectonic setting. Ofioliti, 10, 485-504. Y~LMAZ,Y., TOYS0Z, O., YI~ITBAS,E., CAN GEN(~,S. & SENG6R, A. M. C. 1997. Geology and tectonic evolution of the Pontides. In: ROnINSON,A. G. (ed.)

Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG) Memoirs, 68, 63-90. TI3YSOZ, O., DELLALO~JLU, A. A. & TERZIOGLU, N. 1995. A magmatic belt within the Neo-Tethyan suture zone and its implication for Turkey. Tectonophysics, 243(1-2), 173-191. ULANOVSKAYA,T. E. & SHEVCHENKO,T. V. 1992. The age of the volcanic-sedimentary unit penetrated by the drill-hole in the Desantnaya prospection area, Black Sea. Izvestiya Vysshikh Uchebnykh Zavedeniy (VUZ). Geologiya i Razvedka, 1, 50-57 (in Russian). USTAOMER, T. & ROBERTSON, A. 1993. A Late Paleozoic-Early Mesozoic marginal basin along the

Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists (AAPG) Memoirs, 68, 183-226. ZAKARIADZE,G. S., KARPENKO,S. F., BAZYLEV,B. A., ADAMIA, S. A., OBERHANSLY,P. E., SOLOV'EVA,N. V. & LYALIKOV,A. V. 1998. Petrology, geochemistry and Sm-Nd age of the pre-Hercynian palaeooceanic complex of the Dzirula salient, the Transcaucasus massif. Petrologiya, 6(4), 422.444.

Seismic stratigraphy, structure and tectonic evolution of the Levantine Basin, offshore Israel M I C H A E L A. G A R D O S H & Y E H E Z K E L D R U C K M A N

The Geophysical Institute o f Israel, P O B 182, L o d 71100, Israel (e-mail: miki@gii, co. il) Abstract: Multi-channel seismic reflection data and borehole information were used to study the structure and stratigraphy of the Levantine basin, offshore Israel. A new, 2D seismic survey that covers the southeastern Mediterranean Sea from the Israeli coast to the Eratosthenes Seamount shows the entire Phanerozoic sedimentary fill down to a depth of 14-16 km. The basin-fill is subdivided into six seismo-stratigraphic units interpreted as low-order, major depositional cycles (supersequences A-F). Correlation and mapping of these units allowed an investigation of the geological history of the basin and the analysis of two important tectonic phases: Neotethyan rifting, and Syrian Arc inversion and contraction. The Neotethyan rifting phase is recorded by the strata of supersequences A and B. Faulting took place during the Anisian (Mid-Triassic), continued through the Liassic and ceased during the Mid-Jurassic. The basin opened in a NW-SE direction, between the Eratosthenes Seamount and the Levant margin of the Arabian Massif, at an angle of about 30~ to the present-day shoreline. No indications for sea-floor spreading were found in the present study. Late Triassic to Liassic volcanic rocks of assumed intraplate origin accumulated in the northeastern part of the basin. It is hypothesized that the basin originated as an intracontinental rift associated with the nucleation of an oceanic spreading centre, but reached only an early magmatic phase. An inversion and contraction phase, associated with closing of the Neotethyan ocean system, is recorded by supersequences C and D. The contractional structures of the Syrian Arc extend in a wide and elevated fold belt along the eastern edge of the deep-marine basin. These structures were formed by the inversion of pre-existing normal faults. The folding occurred in several pulses starting in the Senonian and ending in the Miocene. The western limit of the main fold belt, located 50-70 km west of the coastline, is defined by a transition in crustal properties. Supersequences E and F record the Late Cenozoic history of the basin. A Messinian, evaporitic basin was limited to the east by the elevated and uplifted Syrian Arc fold belt composed of older, Oligocene to MidMiocene strata. During highstand episodes, the Messinian evaporites were deposited on the entire slope and within canyons incised into the shelf. High sedimentation rates of Nilotic and locally derived sediments during the Plio-Pleistocene resulted in the development of extensive submarine deltas and basinward progradation of the Levant shelf break.

The Levantine basin (Fig. 1a) occupies a considerable part of the eastern Mediterranean Sea. It is bounded to the east and south by the continental slopes of the African and Arabian plates along the Mediterranean coast of Sinai, Israel, Lebanon and Syria, and on the north by the Cyprian Arc plate boundary at the southern edge of Eurasia. It is a deep marine basin reaching water depth of up to 2200 m below mean sea level (MSL). The basin is a remnant of the Neotethys Ocean that opened following the break-up of Pangaea in Early Mesozoic times (Dewey et al. 1973; Robertson & Dixon 1984). During the Mid-Late Cretaceous the basin started to close. The northern margins were intensely deformed and subsequently subducted or accreted at the present-day areas of Cyprus and southern Turkey (Fig. 1) (Ben A b r a h a m 1989; Garfunkel

1998; Robertson 1998). The southeastern margins, however, remained stable at their original position near the Israel-Sinai coastline. The sedimentary sections on the slope and deep-marine basin of the southeastern Levant continental margin preserve important evidence of events associated with the opening and closing of the Neotethys. A vast volume of geological and geophysical data that were collected in the past 40 years have been used to study the structure and stratigraphy of the Levantine basin and to develop a conceptual model to explain its origin and tectonic evolution. The existence of a deep sedimentary basin in the southeastern Mediterranean Sea area was initially revealed during the late 1960s to 1980s (Ginzburg et al. 1975; Neev et al. 1976; Bein & Gvirtzman 1977; Druckman 1984; Garfunkel & Derin 1984; Cohen et al. 1988, 1990).

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic'Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 201-227. 0305-8719106l$15.00 9 The Geological Society of London 2006.

202

M . A . GARDOSH & Y. D R U C K M A N

o

0

o

~ 9

.~

~-~

o ~

on~

o~

~

o

~.~.~

r~9

.~go~

o ~

~ 0 . ~

0

~ , .._~

.~

~ - ~ ,~

"T

.i ~ -.~ ~~ ~ ~< ..., c~ 0

0

LEVANTINE BASIN, OFFSHORE ISRAEL Studies of onshore and offshore deep boreholes have shown marked lateral changes in the Jurassic and Cretaceous rock record, and a transition from shallow-marine shelf and slope facies in the south and east to deep marine facies in the west. Multi-channel seismic reflection surveys shot on the continental shelf have confirmed the presence of a thick Mesozoic to Neogene rock succession that extends throughout the southeastern Mediterranean area. The deep sedimentary basin and its bordering land were interpreted as a relic of the Neotethys Ocean and adjacent passive continental margin that developed on the northern edge of the Arabian plate in Early Mesozoic times (Fig. 1). The relation between the sedimentary rock record and the deep crustal structure was studied on a regional scale in a later phase of research. Long-range seismic refraction profiles and gravity and magnetic data were collected over the inland part of Israel and across the southeastern Mediterranean Sea, and revealed considerable variations in the density and velocity of the crust (Ginzburg et al. 1979; Ginzburg & Folkman 1980; Makris et al. 1983; Ginzburg & Ben-Avraham 1987; Makris & Wang 1994; Ben-Avraham et al. 2002). The geophysical dataset showed a 35 km thick, continental-type crust beneath southern Israel that thins to c. 10 km in the central part of the marine basin. The thickness of the sedimentary cover was found to change accordingly from 6 km near the Mediterranean coast to about 15 km in the offshore area. These findings further supported the initial interpretation of the basin as a major tectonic element in the area. The analysis of gravity and magnetic data, seismic refraction and single-channel seismic reflection data from the northeastern Mediterranean Sea has all revealed more information on the northern edge of the Levantine basin (Fig. 1) (Ben-Avraham et al. 1976; Woodside 1977; Makris et al. 1983; Makris & Wang 1994; BenAvraham et al. 1995). The elevated structures of Cyprus and the submarine Eratosthenes Seamount were found to be underlain by a 2535 km thick crust of assumed continental origin. An area of prominent deformation and wrench faulting found south and east of Cyprus was interpreted as an active boundary between the African-Arabian plate on the south and the Eurasian plate on the north. The uplift and deformation along the Cyprian Arc plate boundary was explained by the closure of the Neotethyan Ocean through plate collision, subduction and accretion processes in the northern part of the Mesozoic Levantine basin. The Eratosthenes Seamount was interpreted as a small continental

203

body detached from the African plate during the Neotethyan rifting and later moved northward to its present location just south of the present plate boundary. Although many details of the Levantine basin are now well recognized, there is disagreement among researchers regarding several important aspects of its origin and tectonic evolution. An 'oceanic' model assumes that the thin crust found in the central part of the basin is a relic of a Neotethyan oceanic lithosphere. According to this model the Levantine basin evolved since the break-up of Pangaea in Late Permian time as the southern arm of a large Neotethys ocean. Rifting and sea-floor spreading episodes in the Triassic to Early Jurassic were presumably followed by the emplacement of new oceanic crust in the central and northern part of the basin (Bein & Gvirtzman 1977; Makris et al. 1983; BenAvraham et al. 2002). An alternative 'continental' model suggests that the central part of the basin is composed of thinned continental crust. The Levantine basin is assumed to be underlain by a number of aborted or failed rifts that opened on the Mesozoic shallow shelf, south of the large Neotethyan Ocean (Hirsch et al. 1995). Another area of disagreement is associated with the geometry and nature of opening of the Levantine basin. Based on plate motion reconstruction several workers have suggested an east-west-oriented spreading centre within the basin that separated the Tauride microplate from Africa, and was associated with north-south, strike-slip motion on a transform fault along the eastern coast of the Mediterranean (Dewey et al. 1973; Seng6r et al. 1984; Dercourt et al. 1986; Stampfli et al. 2001). A NNE-trending discontinuity zone identified in geophysical data along the base of the Israeli continental shelf, termed the Pelusium Line, was interpreted as a major transcontinental shear associated with the postulated north-south strike-slip motion (Neev et al. 1976; Neev 1977). An alternative view proposes a north-southoriented spreading centre within the basin and a N W - S E extension and opening between the Mediterranean coastline and the Eratosthenes and Cyprus areas. This model postulates an east-west, strike-slip motion along the northern coast of Sinai (Garfunkel & Derin, 1984; Garfunkel, 1998). Hydrocarbon exploration activity has taken place in the Levantine basin, offshore Israel since the 1970s. In the last 10 years, following a renewed interest in its hydrocarbon potential and significant gas discoveries, a large amount of geophysical data were collected across the marine

204

M.A. GARDOSH & Y. DRUCKMAN

basin and several deep wells were drilled in the eastern margin. The new data provide a detailed and more accurate image of the subsurface that was not previously available. This paper presents an analysis of part of the new geophysical dataset. It focuses primarily on a regional 2D seismic reflection survey performed across the southeastern Mediterranean Sea during 2001. A seismic stratigraphic analysis was used to define and map the main rock units that make up the basin-fill. Several regional fault and fold systems were identified in the basin and on its margin. The results are used to evaluate and update the models previously suggested to explain the tectonic evolution and geological history of the Levantine basin.

Dataset The seismic dataset used in this research includes 52 multi-channel, 2D seismic reflection lines totalling 4000 km. The core of the study is a grid of 32 regional marine lines acquired during 2001 in the framework of hydrocarbon exploration activity offshore Israel (EM series in Fig. lb). These data, obtained by Spectrum Energy and Information Technology Ltd, extend from the Israeli shallow shelf to the submarine Eratosthenes Seamount, some 200 km to the NW. It covers most of the Levantine basin in a relatively dense seismic grid of about 10 km x 20 km (Fig. lb). The EM lines were acquired by the R.V. Geo Baltic vessel using a 7200 m long streamer with 576 recording channels (group interval 12.5 m), and an energy source of a four air gun array, each gun with a volume of 3410 cubic inches and a pressure of 3000 p.s.i. The optimal shooting parameters and data processing sequence resulted in a highly interpretable dataset in which seismic reflections are well resolved down to about 10s two-way travel time (TWT; Figs 2-6). The resolution and depth of penetration of the new EM series is superior to most other 2D reflection lines previously acquired for offshore Israel. Twenty additional 2D multi-channel reflection lines of older vintages were reprocessed and integrated into the seismic dataset (DS, AS, M, 83, 88 and 91 series in Fig. lb). These lines, which cover nearshore and onshore areas, allowed the correlation of the seismic data to 16 deep boreholes located on the eastern margin of the basin (Fig. 1, Table 1). An interpretation of the entire seismic dataset was performed on a workstation using timemigrated profiles. Synthetic seismograms and time-converted wireline logs were constructed

for the correlation of seismic horizons to stratigraphic units in the wells. Biostratigraphic and lithostratigraphic information, taken from published well reports, was used for age control and the geological interpretation of seismic events.

Seismic stratigraphy Seismic &terpretation and mapping A thick reflection series (up to 10 s TWT) comprises the entire sedimentary rock record that accumulated on the northern edge of the Arabian-African plate during the Phanerozoic (Figs 2-6). Six seismic packages were identified in the Phanerozoic sedimentary interval, between the crystalline basement and the water bottom surface (Fig. 2). The packages are distinct seismic units that are bounded by regional markers recognized through truncation and both onlapping and downlapping reflections. Correlation with deep boreholes shows that the seismic packages comprise major lithostratigraphic units (Fig. 2) and their boundaries are regional unconformity surfaces associated with relative sea-level changes. The seismo-stratigraphic units were interpreted as low-order depositional cycles that developed in response to the main tectonic events that shaped the Levantine basin and margin. The time span of most of these depositional cycles, labelled from bottom to top as units A to F (Fig. 2), ranges from 5 to about 80 Ma. According to sequence stratigraphic terminology they are defined as second-order depositional sequences or supersequences (Haq et al. 1988; Emery & Myers 1996). Supersequence E, comprising the Messinian evaporate, is a higher-order depositional cycle with a time span of <2 Ma. However, the Messinian strata represents a conspicuous seismic and lithostratigraphic unit that is well recognized throughout the basin and is therefore described here, for simplicity, within the supersequence scheme. Mapping of supersequences A - F involved several steps. First, time-structure and timethickness grids (isochrons) were prepared for each of the interpreted seismic horizons. The isochron grids were converted to isopach maps by multiplication of one-way travel time values by a single interval velocity (Fig. 7a-f). The isopach maps were then summed (in a layer-cake manner) to create the depth maps of the various unit tops (Fig. 8a and b). Seismic interval velocities used for depth conversion were calculated from check shot surveys of the 16 deep wells in the study area (Table 1,

LEVANTINE BASIN, OFFSHORE ISRAEL

205

Fig. 2. Interpreted, time-migrated seismic reflection line EM-24 (eastern part), showing the six seismo-stratigraphic units composing the Phanerozoic sedimentary fill of the Levantine basin (supersequences A-F). Each supersequence is dominated by a unique seismic facies: A is a series of parallel, high-amplitude reflections above the chaotic crystalline basement; B is characterized by discontinuous, low-amplitude reflections and reflection-free zones; C is a series of parallel, continuous reflections, in places onlapping the top of B; D is characterized by discontinuous, low-amplitude reflections; E is chaotic to reflection-free (Messinian evaporite); F is composed of a thinly layered, basinward-prograding reflection pattern. The supersequence boundaries are correlated with regional unconformity surfaces that are dated in onshore and offshore wells. An interpreted lithological column of the basin-fill is partly taken from cutting sample description of the Yam West-1 well. The composition of the Precambrian basement and the overlying supersequence A is assumed based on seismic character and the extrapolation from nearby onshore wells. CDP, common depth point. Location of the seismic profile and the wells is shown in Figure lb.

Fig. 1). The velocity data available for time to depth conversion are limited and come only from the shallow, eastern part of the basin (Fig. lb). Therefore, the following approximations are required: an interval velocity of 4200 m s-1 was used for depth conversion of the Messinian evaporates, as the velocity calculated for this interval from the Yam-2 well (3311 m s-1) is considered unrepresentative (Table 1); units D and C have similar compositions; both are dominated by argillaceous limestone and chalk, therefore, the same interval velocity value of 3000 m s-1 was used (Fig. 2, Table 1); for unit B we used the interval velocity of the deeper marine facies (3400 m s-l), which occupies most of the

study area. Correction for the effects of compaction was not applied. Although some of the interval velocities used here for depth conversion are approximate we consider them sufficiently adequate for the large scale and scope of the study.

The crystalline basement The deepest seismic horizon interpreted in this study is the near top of the crystalline basement. It is correlated throughout the study area with the transition from a chaotic and reflectionfree seismic character below to a continuous, parallel to divergent, high-amplitude reflections

206

M.A. GARDOSH & Y. DRUCKMAN

Fig. 3. Interpreted, time-migrated seismic reflection line EM-50 showing the deep structure in the central part of the Levantine basin. A series of normal faults offset the top of the basement, supersequence A and the lower part of supersequence B. Supersequences C-F are not affected by normal faulting. The graben and horst system in the lower part of the basin-fill developed during the Early Mesozoic, Neotethyan rifting phase. Thickness variations in supersequence A (Permian-Aalenian) indicate syntectonic deposition. Fault activity ceased during the deposition of the lower part of supersequence B (Bajocian-Bathonian). Some of the normal faults at the eastern margin of the basin (east of CDP 14000) were later reactivated in a reverse sense of motion during the Syrian Arc contractional phase. The shallow fault system at the top of supersequence E (Messinian evaporites) associated with Plio-Pleistocene halokinesis should also be noted. The location of the profile is shown in Figure lb. series above (Fig. 2). We interpret this transition to characterize a major acoustic boundary between non-layered, magmatic and metamorphic basement complexes to the overlying, layered Palaeozoic-Mesozoic sedimentary interval. The transition in seismic character is more evident in the southeastern corner of the study area, near the Yam West-1 well, where it is found at a relatively shallow depth of about 5-6 s T W T (Fig. 2). In this area the seismic boundary is correlated with the top of the Precambrian to Infra-Cambrian basement penetrated by the Heletz Deep-1 and the Gevim-2 wells (Fig. lb). The resolution of the seismic data generally decreases below 7-8 s as a result of the reduction of the seismic energy. However, the transition between chaotic seismic character and overlying high-amplitude, partly continuous reflection series was identified below 8 s in many seismic profiles located in the central, deep part of the basin (Figs 3 and 4). The basement dips gradually westwards from 7 km depth at the Levant shoreline to 14-16 km at the centre of the basin (Figs 5 and 8), and it

rises to 5-6 km beneath the Eratosthenes Seamount further N W (Fig. 3; Makris & Wang 1994; Garfunkel 1998; Ben-Avraham et al. 2002). The top of the basement marker is dissected by many faults. These are predominantly normal faults in the western and central parts of the basin and reverse faults at the eastern margin (Figs 3-5). Most of the normal faults in the central part of the basin trend to the NE, forming an extensive graben and horst structure extending across the entire basin, from the Eratosthenes Seamount area in the west to the base of the Israeli continental shelf in the east (Fig. 8a). An 80-100 km wide graben was identified at the centre of the basin where the top of the basement marker reaches its maximal depth (Fig. 8a). This central graben appears to terminate gradually to the NE, across the postulated seaward extensions of the Carmel fault and the Mount Carmel and western Galilee structures (Fig. 8a). No indication was found of a structural discontinuity along these features, as suggested by Ben-Avraham & Grasso (1991) and BenAvraham & Tibor (1994). A horst of 15-30 km

LEVANTINE BASIN, OFFSHORE ISRAEL

207

Fig. 4. Interpreted, time-migrated seismic reflection line EM-48 (central part), showing a deep-seated asymmetrical graben in the northern part of the Levantine basin. The graben is interpreted as a Neotethyan rift structure that was formed during the deposition of supersequence A (Permian-Aalenian). The chaotic strata of supersequence B on top of the graben (between CDP 2642 and 3642) are interpreted as post-rift clastic fill. The transition in seismic character from the chaotic crystalline basement to the continuous, high-amplitude parallel reflections in the lower part of supersequence A should be noted. The location of the profile is shown in Figure lb. width and 80 km length trending in a N E - S W direction is located in the centre of the graben, where the top of the basement is 2-3 km higher than in the adjacent flanks (Fig. 8a). Seismic refraction, gravity and magnetic data show that the crust underlying the area of Jordan and Israel thins from 35 km on land to about 10 km within the basin, and that it thickens to about 25 km further west beneath the Eratosthenes Seamount (Ginzburg et al. 1979; Makris et al. 1983; El-Isa et al. 1987; Ginzburg & Ben-Avraham 1987; Makris & Wang 1994; Ben-Avraham et al. 2002). The thick crust underneath Israel and the Eratosthenes Seamount is composed of a high-velocity layer at the bottom (6.7 km s-~), and a low-velocity layer on top (6.0 km s-~), interpreted as upper continental crust (Makris et al. 1983; Ben-Avraham et al.

2002). The absence of the low-velocity layer and the presence of high-velocity (6.7 km s-a), thinned lithosphere in the centre of the Levantine basin was interpreted as an indication of oceanic crust (Ben-Avraham et al. 2002). This type of crust is assumed to have been emplaced during Early Mesozoic sea-floor spreading as proposed by the 'oceanic' model. The nature of the upper crust below the near-top of the basement horizon cannot be directly determined from our seismic data; however, indirect criteria and a comparison with other passive continental margin were used to address the important issue of the crustal composition beneath the Levantine margin and basin. At the Atlantic continental margin of the USA, the transition between the continental

208

M.A. GARDOSH & Y. DRUCKMAN

Fig. 5. Interpreted, time-migrated seismic reflection line EM-22 showing the structure of the central and eastern parts of the Levantine basin. An elevated, highly deformed fold belt on the east side of the profile is the western edge of the regional, Syrian Arc fold system that developed in association with the closure of Neotethys. The Syrian Arc contractional deformation started during the deposition of supersequence C (Senonian-Early Oligocene) and ceased during the end of supersequence D (Late Miocene). The reverse, high-angle thrust faults within the fold belt are interpreted as reactivated normal faults of the Early Mesozoic rifting phase. The same type of faults remained inactive in the central part of the basin (west of CDP 10484). The onlapping reflections on the lower boundary of supersequence C (between CDP 9484 and 12484) and the increased thickness of the unit west of the elevated fold belt are associated with drowning of the Mid-Cretaceous shelf coupled with intense transport of clastic material into the basin. The location of the profile is shown in Figure I b.

crust in the west and the oceanic crust in the east is gradual and takes place through several intermediate zones with distinct physical properties and seismic character, ranging from continental, to rifted, to marginal oceanic and to oceanic (Klitgord & Hutchinson 1988). In our dataset, the basement layer shows similar seismic character across the entire Levantine basin and margin, and no indication for crustal zonation was observed (Figs 3 and 5). Additionally, in the centre of the basin the top of the basement layer does not show the hummocky or mounded character that often characterizes an oceanic basement in the Atlantic margin (Klitgord & Hutchinson 1988) or more locally, in the Ionian Sea at the centre of the Mediterranean Sea (Finetti 1985; Avedik et al. 1995). The deep-seated fault blocks, graben and horsts found throughout the Levantine basin are typical of a rifted or transitional-type crust

such as found in the western part of the Atlantic passive margin (Klitgord & Hutchinson 1988). Old rift-related structures would not have been present within new oceanic crust that presumably developed after the main rifting phase. It is therefore suggested, following Hirsch et al. (1995), that the basement of the Levantine basin is not oceanic in composition, but rather is a stretched, thinned and probably highly intruded continental crust. The velocity of 6.7 km s-~ found in seismic refraction studies is compatible with this type of crust (Hirsch et al. 1995). Other supporting evidence for a 'continental' model is the absence of linear magnetic anomalies and the extremely thick sedimentary cover (14-16 km) in the centre of the basin (Hirsch et al. 1995). It is assumed that the basement of the Levantine basin has not gone beyond the early magmatic phase of rifting and did not reach full-scale oceanic sea-floor spreading.

LEVANTINE BASIN, OFFSHORE ISRAEL

209

Fig. 6. Interpreted, time-migrated seismic reflection line EM-60 (eastern part) showing contractional deformation at the western edge of the Syrian Arc fold belt. Two folding episodes are observed: an early episode that affected only the strata of supersequence B resulted in high-amplitude, narrow folds and thrust faults (eastern part of the line below 3600 ms); a later episode that affected also the strata of supersequences C and D (eastern part of the line between 3600 and 2800 ms) produced lower-amplitude, wider folds and minor thrusting. Chaotic seismic packages filling the relief in the lower part of supersequence C (near CDP 1825) are interpreted as clastic-rich deposits that were transported into the basin during the early stage of drowning. Shallow, normal faults and folds within supersequences E and F are associated with Plio-Pleistocene halokinesis. The location of the line is shown in Figure lb.

The Permian to Aalenian supersequence (A) Supersequence A is a 3-8 km thick unit (1-3s TWT) extending across the entire width of the Levantine basin (Figs 3 and 7a). In the seismic data it is characterized by high-amplitude, relatively continuous, parallel to divergent reflections, locally changing to less reflective or reflection-free zones (Fig. 2). The disruption of seismic character commonly takes place in areas where the unit is highly deformed by normal and reverse faults (Figs 3 and 5). The lower boundary of supersequence A is the postulated near-top of the crystalline basement. The upper boundary is correlated with a distinct change in seismic character from high-amplitude, continuous reflection series below to discontinuous, low-amplitude, occasionally shingled,

reflections above (Fig. 2 at 4500 ms between the Yam-2 and Y a m West-1 wells). The change in seismic character is most evident in the eastern part of the basin; however, it is also observed in its central and western parts, where the top of supersequence A is a relatively continuous high-amplitude seismic event below less continuous and lower amplitude reflections (Figs 3 and 5). Well data indicate that the change in seismic character at the top of supersequence A corresponds to a significant transition in lithology and depositional environment. The continuous, parallel reflections in the upper part of the unit (Fig. 2) are interpreted as shallow-water deposits, probably similar to Triassic and Lower Jurassic platform carbonate found in various onshore wells on the eastern margin of the basin. The

210

M.A. GARDOSH & Y. DRUCKMAN

Table 1. Seismic interval velocities o f the basin-fill units (calculated from check shot surveys) in the wells o f the study area

Interval velocities (m s-1) Well names and abbreviations Asher Yam-1 (AYI) Ashkelon-2 (As2) Atlit-1 (Atl) Bravo-l(B1) Caesarea-3 (C3) Delta-l(D1) Foxtrot- 1 (F 1) Gaash-2 (Ga2) Gevim-1 (Gel) Heletz Deep-l(HD1) Hof Ashdod-1 (HA1) Item-1 (I1) Nisanit-1 (N1) Yam West-1 (YW1) Yam Yafo-1 (YY1) Yam-2 (Y2) Average values used for depth conversion

T.D.(m)

SequenceF

2020 4076 6531 4096 4600 4423 2153 5508 4620 6093 3152 3708 3750 5250 5787 5377

1921

SequenceE

Sequences SequenceB C+D

SequenceA

4865"t 27645

1750:~

5077"t 5630"? 2013

3700*

4660*w

1660 1794 1863 1800

2539* 2865 2961 3000

3348*w 2956*w 3197w 2765w 3400

3311 4200

5990t 5800t 5911t

4494t 5500

*Shelf facies. tIncomplete section. SEstimated from wireline logs and seismic data. w facies. T.D., total depth. The interval velocity values were used for depth conversion of interpreted, two-way travel time grids and the construction of isopach and depth maps.

overlying discontinuous, low-amplitude reflections of supersequence B correspond in the offshore, Yam-2, Yam West-1 and Yam Yafo-1 wells to a series of shale, marl, redeposited limestone and some sandstone of Bajocian to Bathonian age. These sediments accumulated in a slope and deep-water environment, basinward of the Mid-Jurassic shallow water carbonate platform (Derin et al. 1990; Druckman et al. 1994; Gardosh 2002). The seismic boundary at the top of supersequence A is interpreted as a regional unconformity surface between the Triassic-Lower Jurassic shallow-marine section and the overlying Middle Jurassic deeper marine section. This unconformity is recognized in deep wells at the eastern margin of the basin, where the entire Jurassic section was penetrated. In the onshore Helez Deep-l, Gaash-2, Nissanit-1, Gevim-1, Caesarea-3 and Atlit-1 wells (Fig. lb), a lithological and biostratigraphic break is identified at the top of limestone and dolomite series (Qeren Formation) of Aalenian age (Hirsch et al. 1998).

The shallow marine carbonate is overlain by a thin shale and siltstone bed (Rosh Pina Formation) interpreted as a regressive unit (Derin 1974; Hirsch et al. 1998). Hirsch et al. (1998) correlated this terrigenous unit with an Aalenian sea-level drop at the onset of the LZA-1 eustatic cycle (Haq et al. 1988). In some onshore seismic profiles the top of the Qeren carbonate appears as a continuous, high-amplitude reflection similar to the upper boundary of supersequence A in the offshore (Profile DS-3072 and Nisanit-1 well, Fig. lb). Supersequence A displays significant thickness variations. The presence of asymmetrical grabens indicates syntectonic deposition and continuous activity of deep-seated faults and basement blocks (Figs 3 and 4). The isopach map shows a large trough, some 90 km wide, trending NE-SW in the central part of the basin (Fig. 7a). The trough is up to 8 km thick in its centre and gradually thins to about 3 km towards the east and west. It is superimposed on a deeper graben mapped on the near-top of the basement horizon (see Figs 7a and 8a).

LEVANTINE BASIN, OFFSHORE ISRAEL

211

Figs. 7. Isopach maps of the six depositional supersequences composing the Levantine basin-fill. The maps were calculated from two-way travel time grids of the interpreted horizons by using constant, interval velocities (Table 1). (a) shows the marked thickness of supersequence A in the centre of the basin (4-8 km) that is associated with syntectonic deposition during the Early Mesozoic, Neotethyan rifting phase. (b) demonstrates the eastward shift of depocentre during the post-rift development of the Levantine continental margin.

212

M . A . GARDOSH & Y. D R U C K M A N

Figs. 7. (c) shows the increased thickness of supersequence C in the centre of the basin, probably caused by high influx of clastic material and creation of accommodation within the basin during the Syrian Arc deformation phase. The Jonah Seamount (JS) is interpreted as a Mid-Cretaceous volcanic structure that remained as an elevated submarine high during Senonian to Early Oligocene times. (d) shows a more or less equal distribution of supersequence D across the basin with minor thinning towards the eastern margin that reflects the last stage of the Syrian Arc deformation.

LEVANTINE BASIN, O F F S H O R E ISRAEL

213

Figs. 7. (e) shows the main accumulation of the Messinian evaporite in the centre of the basin and west of the elevated Syrian Arc fold belt. (f) demonstrates the thick accumulation of the Plio-Pleistocene prograding sequence along the eastern margin, probably as a result of high influx of both Nilotic and locally derived clastic material.

214

M . A . G A R D O S H & Y. D R U C K M A N

Fig. 8. Structural maps (below MSL) of (a) the near-top of the Precambrian basement horizon and (b) the top of the supersequence B horizon (Bajocian-Turonian). The near-top basement map shows the extensive SW-NE-oriented normal fault system formed during the Triassic-Jurassic Neotethyan rifting. The most intense rifting took place in the centre of the basin, where the depth to the top of the basement is in the range of 14-17 km. This Central Graben is dissected by an elongated and narrow horst that is possibly associated with intra-rift magmatic activity. The map of the top of supersequence B shows the elevated and highly deformed Syrian Arc fold belt along the eastern margin of the basin. Many of the folds contain high-angle thrust faults. The western edge of the Syrian Arc fold belt follows a zone of crustal discontinuity (dotted line) interpreted from deep refraction and gravity data (after Ben-Avraham et al. 2002). It is proposed that the westward termination of the Syrian Arc deformation is related to the deep crustal structure and reflects the transition from a continental crust in the east to a thinned, intruded crust in the west.

215

LEVANTINE BASIN, OFFSHORE ISRAEL Table 2. The displacement of various stratigraphic markers across the Heletz fault trace, represented by elevations in the hanging wall (Heletz Deep-l) and footwall (Gevim-1)

Stratigraphic tops in wells Aptian Callovian Bathonian Bajocian Triassic Near-top of Infra-Cambrian or Precambrian basement

Heletz Deep- 1, hanging wall (m below MSL)

Gevim- 1, footwall (m below MSL)

Measured vertical displacement (m)

Reconstructed normal throw (m)

1069 1783 (reconstructed) 1928 2724 4708 5714

1555 2022

+486 +239

247

2187 2494 3737 4460

+259 -230 -971 - 1254

227 716 1457 1740

The Heletz fault was active in normal motion during the Neotethyan rifting and in a reverse motion during the Syrian Arc contractional phase. The maximal reverse motion, estimated from the displacement of the Aptian level, amounts to 486 m. The normal throw (right column) is reconstructed by adding the reverse motion to the measured vertical displacement of the pre-Aptian units. The total amount of normal throw on the near-top of the Infra-Cambrian or Precambrian basement is 1740 m. A normal throw of 283 m is calculated for the Triassic phase; 741 m is calculated for a Liassic-Bajocian phase, and 489 m for an Early to Mid-Bathonian phase. (For further details see Fig. 9 and the text.)

Gevim-1

Heletz Deep-1

Metres I Aot. I " . . . . . . . below MS.~ -- /

7r

]

/ Cal, 1-- . . . . . . Bat. r . . . . . .

.-I -t

-2000J

,ooo3 / 9

~ Meai Sterranea e a. - . / . . .

-Ouuu

I

m~

Tr. -

h"

Triassic

i ....

m

!

Aptian Callovian Bathonian Bajocian

--"

Infracambrian/ Precambrian

mm

/.D-,,/ 4'n'r.L--J

1.7.

_.ooo-

-

I.F ~

rh

~if

Fig. 9. A schematic geological section across the trace of the Heletz fault, between the Gevim-1 and the Heletz Deep-1 wells. The fault was active in a normal, down-to-the-basin motion during the Neotethyan rifting phase and was reactivated in a reverse motion during the Late Cretaceous, Syrian Arc contractional deformation, Reconstruction of the normal throw on the Triassic and Jurassic stratigraphic levels provides an estimate of the time and amount of displacement during the Neotethyan rifting. (For further details see Table 2 and the text). The stratigraphy and age of supersequence A within the trough and throughout the Levantine basin are not well established because of lack of well control. Well data from the coastal plain show that Permian sediments overlie the Precambrian basement (Heletz Deep-l, Gevim-1). It is,

therefore, assumed that supersequence A is dominated by Permian, Triassic and Lower Jurassic strata. The thickness of this rock section onshore is in the range of 1.5-5 k m (Garfunke11998), considerably less then the 4-8 km of supersequence A in the offshore (Fig. 7a).

216

M.A. GARDOSH & Y. DRUCKMAN

The excessive thickness of supersequence A within the Levantine basin may be partly related to the presence of Palaeozoic or even InfraCambrian units that were regionally deposited and later eroded from elevated areas in the nearby continental margin. It is more likely, however, that the main cause of the additional thickness is the syntectonic accumulation of continental, shallow-marine and volcanic rock sections within Triassic and Early Jurassic rift structures. The occurrence of thick volcanic units in supersequence A is supported by well data from the Haifa area some 40 km eastwards of the main depocentre (Fig. lb). In the Atlit-1 well Derin et al. (1982), Dvorkin & Kohn (1989) and Korngreen (2004) described a Carnian-Norian (Upper Triassic) shallow-water carbonate section of 1141m thickness, interbedded with 320m of volcanic rocks. The Triassic sequence is overlain by a 2500 m section of olivine basalts of alkaline affinities known as the Asher Volcanics (Gvirtzman & Steinitz 1983). Argon dating has revealed a mid-Early Jurassic age (193-198 Ma) for the upper part of the volcanic section (Steinitz et al. 1983; Kohn et al. 1993). The Asher Volcanics do not show mid-ocean ridge basalt (MORB) characteristics and their geochemical signature resembles that of intraplate basalts (Dvorkin & Kohn 1989; Garfunkel 1989). The presence of interbedded shallow-water carbonate, palaeosols and lignites indicates shallowmarine to subaerial volcanism. Garfunkel (1989) proposed that the Asher Volcanics accumulated within a 2-3 km deep, fault-controlled depression in the Haifa area. Similar depressions may have filled with Triassic and Lower Jurassic basalts in the central part of the Levantine basin further west. The existence of thick volcanic units within supersequence A in the offshore is further supported by a strong positive magnetic anomaly over the northeastern part of the basin (Rybakov et al. 1997). In summary, based on the lithological composition found in onshore wells and the seismic character offshore, the strata of supersequence A are interpreted as continental to shallow-marine siliciclastic and carbonate deposits, in places containing thick volcanic sections. The unit was intensely faulted and its deposition took place contemporaneously with extensive rifting and extension throughout the Levant area. No clear evidence was found in the seismic and well data for a deep-marine basin during the Permian to Early Jurassic time span. A shelf break was formed in the Levantine basin only in Mid-Jurassic times, when an aggradational carbonate shelf was established on its eastern margin (Gardosh 2002).

T h e B a j o c i a n to T u r o n i a n supersequence ( B)

Supersequence B is an extensive unit attaining its maximal thickness of 3-3.5 km (2-2.5 s TWT) at the eastern margin of the basin (Figs 5 and 7b). Its thickness is reduced to about 1.5 km (1 s TWT) in the central and western parts of the basin (Figs 3, 5 and 7b). In the seismic data the eastern margin area is characterized by a relatively discontinuous, low-amplitude reflection series (Fig. 2, between Yam West-1 and Yam-2 wells), whereas more continuous, highamplitude reflections characterize the unit in the deep part of the basin, further offshore (Fig. 5). The lower sequence boundary is correlated with the Middle Jurassic unconformity surface. The upper boundary corresponds to a distinct transition of seismic character between several high-amplitude reflections at the top of the unit and overlying lower-amplitude reflections, marked by chaotic and reflection-free zones (Figs 2, 5 and 6). The upper boundary is evident throughout the deep part of the basin, where the high-amplitude package at the top of supersequence B is particularly continuous and is less disturbed by faulting (see left part of Fig. 5 at 6200 ms). In the eastern part of the basin the sequence is highly folded and its top is characterized by widespread onlapping reflections of supersequence C on the flanks of fold structures (Fig. 2, line EM-24, east of the Yam West-1 well; Fig. 5, the eastern part of line EM-22 between 3000 and 5000 ms). In the wells of the study area and in outcrops throughout the inland part of Israel the unconformity at the top of supersequence B is indicated by a depositional hiatus; there is a pronounced lithological transition between shelf and slope carbonates of Albian to Turonian age (Judea and Talme Yafe Formations) and pelagic marl and chalk of Senonian to Palaeogene Age (Mount Scopus and Hashefela Groups) (Gvirtzman & Reiss 1965; Flexer 1968; Flexer et al. 1986). The Middle Jurassic to Middle Cretaceous strata comprising supersequence B were penetrated by many onshore and offshore wells in the western part of Israel. An important feature identified in these wells is a pronounced facies change from coarse-grained, shelf-type units east of the present-day coastline to finer-grained, slope and basin-type deposits further west (Derin 1974; Bein & Gvirtzman 1977; Flexer et al. 1986). This facies variation indicates the development of a deep-marine basin bordered by a slope and a shallow-marine shelf on the southeastern Levant continental margin (Bein & Gvirtzman 1977).

LEVANTINE BASIN, OFFSHORE ISRAEL A detailed sequence stratigraphic analysis of the Middle Jurassic to Middle Cretaceous strata in the southern coastal plain of Israel (Gardosh 2002) shows that the continental margin is composed of six second-order and 22 third-order depositional sequences. These sequences include both lowstand and transgressive to highstand components. The lowstand-type deposits are fine-grained siliciclastic and carbonate debris that was transported into the basin by submarine gravity flows, whereas the transgressive to highstand deposits are coarse-grained bioclastic and siliciclastic deposits that accumulated on the eastern margin in shallow-marine rimmed platforms and ramps (Gardosh 2002). Rapid backstepping and aggradation associated with long-term sea-level rise characterize the carbonate platforms and ramps of the Bajocian-Oxfordian and Cenomanian-Turonian periods. Influx of both siliciclastic and detrital carbonate sediments into the basin was widespread during the TithonianHauterivian and Albian periods. The steep carbonate shelf edge of supersequence B found near the present shoreline is generally located east of the seismic coverage of the present study. The discontinuous, lowamplitude, occasionally shingled seismic facies of supersequence B in the offshore (Figs 2 and 5) was interpreted as lowstand-type, amalgamated, deep-water turbidite systems composed of siliciclastic and carbonate debris. These kinds of sediments are recognized within the BajocianTuronian strata in the Yam Yafo-1, Yam 2 and Yam West-1 wells (Derin et al. 1990; Druckman et al. 1994; Gardosh 2002). The turbidite flows bypassed the Mid-Jurassic and Mid-Cretaceous carbonate shelves through deeply incised channel and canyon systems. An example of such a channel is the Gevaram canyon, located in the Heletz area of the southern coastal plain (Fig. 1). The canyon is filled with up to 1000 m of Lower Cretaceous (deep marine) shale (Cohen 1976). Supersequence B reaches a maximum thickness of 2500-3500 m (Fig. 7b). Its depocentre extends from the present coastline to about 5 0 k m westwards, where large amounts of deep-water siliciclastic and carbonate strata accumulated on the upper and lower slopes of the Mid-Jurassic to Turonian margin. Towards the west and near the Eratosthenes Seamount the thickness of supersequence B is reduced to 1000-2000 m (Fig. 7b). It is estimated that in these distal parts of the basin the rate of accumulation of deep-water sediments was considerably smaller. No indication for a shallow-marine carbonate margin is recognized in the Eratosthenes area. Some small, thick accumulations of supersequence B found in the deep offshore are

217

interpreted as localized depocentres filled by clastic material that was eroded and transported from nearby submarine structural highs (note the depression in the eastern part of the line EM-48 between 5500 and 6500 ms in Fig. 4). The deep-seated normal faults at the centre of the basin affect only the lowermost part of supersequence B (Figs 3-5). At the eastern margin, most of the older, normal faults were later reactivated as thrust faults. Reconstruction of the reverse motion shows a normal displacement in the range of several hundred metres of the Middle-Upper Jurassic to Middle Cretaceous strata. The minor effects of faulting indicate that most of the deposition of supersequence B took place in a post-rift stage when the basin was dominated by cooling and thermal subsidence. The termination of this depositional cycle coincides with the onset of large-scale contractional deformation. This tectonic event, known as the Syrian Arc folding phase, is recognized throughout the southern and eastern margins of the Levantine basin, and is associated with the closure of the Neotethys Ocean during the Late Cretaceous to Early Cenozoic, at the time of deposition of supersequence C.

The Senonian to lower Oligocene supersequence ( C) Supersequence C attains its maximal thickness of about 2.5-3 km (2 s TWT) in the central part of the Levantine basin and displays marked thinning towards the eastern and western margins (Figs 3 and 7c). It is generally characterized by low- to medium-amplitude, continuous and parallel reflections, chaotic reflections and reflection-free zones (Fig. 2). Well data indicate that this seismic character corresponds to chalk and marl of pelagic to hemipelagic, deep-water origin throughout the basin (Almogi-Labin et al. 1993). Within the basin and on its eastern margin extensive onlapping is observed on the lower boundary of supersequence C (Figs 2 and 5). Similar onlapping of Senonian strata on Cretaceous slope and shelf carbonate (Talme Yafe and Judea Groups) is found in the southern coastal plain (Gardosh 2002). An erosional unconformity associated with subaerial exposure is recognized at the Turonian-Coniacian boundary (base of supersequence C) in outcrops throughout the inland part of Israel (Bentor & Vroman 1960; Flexer 1968). Both erosion and onlapping are interpreted as the result of a relative sea-level drop that was followed by eustatic rise and drowning of the Mid-Cretaceous carbonate platform (supersequence B). The drowning caused

218

M.A. GARDOSH & Y. DRUCKMAN

the cessation of the Albian-Turonian carbonate factory on the shallow shelf and the replacement of platform carbonate with Senonian pelagic chalk (Sass & Bein 1982; Almogi-Labin et al. 1993; Buchbinder et al. 2000). In the central part of the basin the lower several hundred milliseconds of the supersequence C are characterized by a reflection-free, transparent zone (Figs 4 and 5). In the eastern fold belt chaotic reflection packages are found within deep synclines, overlying supersequence B (the eastern part of line EM-60 between 3500 and 4000 ms in Fig. 6). The transparent to chaotic seismic character at the base of the unit is interpreted as coarser-grained, clastic-rich sediments probably associated with a frequent supply of hemipelagic gravity flows during the early stage of drowning. The top of the chaotic package constitutes, in places, a high-amplitude truncation surface (the eastern part of line EM-60 at 3500 ms in Fig. 6), interpreted as a higher-order sequence boundary within the Senonian to Lower Oligocene supersequence (Gardosh 2002). The upper boundary of supersequence C is correlated with the top of a relatively continuous higher-amplitude reflection series, in turn overlain by low-amplitude and occasionally mounded reflections (Figs 3 and 5). The change in seismic character corresponds to an unconformity surface in the offshore Yam Yafo-1, and Yam West1 wells (Fig. 2), as indicated by a pronounced lithological break and a biostratigraphic hiatus between Middle Eocene carbonates and an overlying section of Oligocene to Miocene siliciclastic deposits (the Saqiye Group; Druckman et al. 1994; Gill et al. 1995). On the isopach map supersequence C displays an almost symmetrical thinning pattern, from 3000 m in the central part of the basin to < 500 m on both its eastern and western margins (Fig. 7c). An exception to this pattern is a prominent elevated structure located in the centre of the basin where the unit is only 1500 m thick (Fig. 7c). This feature was termed the Jonah Seamount and interpreted by Folkman & Ben-Gai (2004) as an intrusive structure of OligoMiocene age. An Oligo-Miocene age for the Jonah Seamount is not supported by the present data, as the structure is not observed on the isopach map of supersequence D (Fig. 7d). An alternative interpretation suggests reduced deposition of Senonian to Early Oligocene strata over a preexisting submarine high, shown as a prominent structure in the centre of the basin on top of supersequence B (Fig. 8b). The Jonah Seamount is interpreted as a volcanic structure of MidCretaceous age. Relevant volcanic phases are known to have occurred in the area during

the Albian and Cenomanian (Garfunkel 1989). The reduced section overlying this high (supersequence C) may have been deposited as a carbonate build-up or an atoll. Calculated sedimentation rates in the central part of the basin are in the range of 33-50 m Ma -~ (uncorrected for compaction), but along the margin only about 8 m Ma -~. The low rate of deposition in the margin is similar to rates reported by Buchbinder et al. (1988) for undisturbed Eocene chalk in outcrops further onshore. The great thickness and high rate of sedimentation, as well as the widespread onlapping and the chaotic seismic character in the lower part of the unit are all indicative of significant gravity transport and influx of clastic sediments into the basin. These processes are associated with the creation of accommodation space probably through both tectonic subsidence and relative sea-level rise during the time of deposition of supersequence C. The O l i g o c e n e to upper M i o c e n e Supersequence (D)

Supersequence D is 0.5-1 km thick unit (0.3-0.7 s TWT), extending across the entire width of the basin (Fig. 7d). In the seismic profiles it is characterized by low-amplitude, partly discontinuous reflections, some mounded reflections and reflection-free zones (Fig. 2). The lower sequence boundary is the base Oligo-Miocene unconformity (base Saqiye siliciclastic rocks). The upper boundary is correlated in the central part of the basin with a marked transition in seismic character between the reflection series of supersequence D and the chaotic and reflection-free seismic package of the overlying Messinian evaporites (supersequence E) (Figs 3, 5 and 6). In the eastern part of the basin, where Messinian evaporites are either thin or missing, the upper sequence boundary corresponds to a conspicuous surface of erosional unconformity and a biostratigraphic hiatus, as recognized in both well and seismic data at the base of the Plio-Pleistocene section (Yafo Formation) (Gvirtzman & Buchbinder 1978; Druckman et al. 1995). In the offshore Yam West-1 and Yam Yafo-1 wells (Fig. l b), supersequence D is composed of pelagic marl and shale of Oligocene to Late Miocene age (Lower Saqiye Group) (Druckman et al. 1994; Gill et al. 1995). Thick intervals of coarse-grained sandstone and conglomerate were found onshore within the sequence in wells along the eastern margin (e.g. Hof Ashdod-1, Fig. l b). These sediments are interpreted as canyon-fill deposits associated with several

LEVANTINE BASIN, OFFSHORE ISRAEL cycles of subaerial and submarine erosion and incision on the Oligo-Miocene shelf (Gvirtzman & Buchbinder 1978; Druckman et al. 1995; Buchbinder & Zilberman 1997; Buchbinder & Siman Tov 2000). Mounded reflections identified within supersequence D in the offshore (the eastern part of profile EM-60 at 2800 ms in Fig. 6), are interpreted as siliciclastic, deep-water turbidite systems and basin-floor fans. These may be the distal equivalents of the proximal canyon-fill deposits found onshore. The presence of coarsergrained siliciclastic material within the finegrained strata of supersequence D may explain the overall low-amplitude and discontinuous seismic character of the unit in the offshore record (Fig. 5). The Messinian supersequence ( E )

The Messinian sequence is a distinct seismic package identified throughout the eastern Mediterranean region, composed of thick evaporitic series of Late Miocene age (Hsfi et al. 1973; Neev et al. 1976; Ryan 1978). In the Levantine basin this unit is about 1-2 km thick (0.5-1 s TWT), extending across most of the basin (Figs 3 and 7e). The lower and upper boundaries are the well-known seismic markers defined in previous works as the N and M horizons. Supersequence E generally displays a chaotic seismic character that is typical of massive rock salt (Figs 3 and 5). Continuous reflection packages found within the unit are interpreted as anhydrite and shale intercalations such as in the Yam-2 and Yam West-1 wells (Fig. 2). The area of evaporite accumulation extended throughout the deep part of the basin to about 20-40 km west of the present-day coastline (Fig. 7e). The evaporitic brine was probably limited to the east by a steep slope composed of the topographically higher strata of supersequence D (Fig. 5). However, during highstand episodes the brine covered this slope and penetrated through deep canyons further inland. Remnants of the Messinian brine in the form of halite and anhydrite beds a few tens of metres thick were encountered in various wells in coastal areas of Israel (Druckman et al. 1995). Halokinesis occurs throughout the basin, affecting the salt layer and the overlying PlioPleistocene strata up to the sea-bed surface (Figs 3, 5 and 6). In the eastern margin, salt-flow has resulted in the development of normal, occasionally listric faults, whereas in the central and northwestern part of the basin and near the Eratosthenes Seamount the salt layer is affected by many thrust faults and folds (Figs 3 and 5).

219

Gradmann et al. (2005) recently suggested that both extension and compression of the Messinian evaporites are associated with thin-skinned salt tectonics related to the subsidence of the basin during post-Messinian time. Basinward creep of the salt layer resulted in normal faulting on the eastern slope and reverse faulting and buckling in the distal part of the basin. A considerable increase of salt thickness towards the western part of the basin may be associated with salt flow in this direction (Figs 3 and 7e). The Plio-Pleistocene supersequence ( F )

The Plio-Pleistocene supersequence attains its maximal thickness of about 1.6 km (1.6 s TWT) along the eastern margin and it is reduced to several hundred metres in the central part of the basin and towards the Eratosthenes Seamount (Figs 3 and 7f). The upper sequence boundary is the sea bed. The lower boundary is a composite unconformity surface. Within the basin this boundary corresponds to the top of the Messinian evaporites (M horizon). In the eastern margin, where the evaporites are missing, the M and N horizons merge and the composite surface coincides with the base Messinian unconformity (Fig. 5). In other parts of the Mediterranean region, this surface is marked by a short lived post-evaporitic fluvial and euryhaline episode that predated the establishment of Pliocene normal, deep-marine conditions, known as the Lago Mare event (Hsfi et al. 1978; Rouchy & Saint Martin 1992). Throughout the basin, the sequence is characterized by continuous, thin high- and lowamplitude reflections. In the central part, these are highly deformed as a result of underlying salt flow. At the eastern Levantine margin, where the unit attains its maximal thickness, it is dominated by sigmoidal progradational reflection patterns (Figs 2 and 5). Well data show that the Plio-Pleistocene section, termed the Yafo Formation, is a mud-dominated unit composed mostly of claystone and siltstone (Gvirtzman & Buchbinder 1978) (see Yam West-l, Fig. 2). Isolated, coarser-grained sand bodies containing large amounts of biogenic gas were recently found within submarine channels and basin-floor fans in the lower part of the supersequence, south of the Yam West-1 well (Fig. lb). The depositional mechanisms of the fine-grained clastic material composing the PlioPleistocene supersequence F are not yet adequately studied. Part of the material was probably transported through longshore currents from the Nile cone located about 200 km to the SW (Fig. 1) (Gvirtzman & Buchbinder 1978; Ben-Gai

220

M.A. GARDOSH & Y. DRUCKMAN

1996). Several lines of evidence suggest an additional local source to the east and SE. First, the conspicuous sigmoidal, progradational reflection pattern is oriented perpendicular to the presentday coastline, thus suggesting transport from exposed land to the east. Second, the unit attains its maximal thickness along the eastern coast, and only a minor increase in thickness towards the SW is observed (Fig. 7f). Third, a NW-SEoriented canyon and channel systems of Late Cenozoic age was identified in well outcrop and seismic data from the southern coastal plain of Israel (Afiq, Ashdod and Gaza canyons) (Gvirtzman & Buchbinder 1978; Druckman et al. 1995). These canyons were probably preferred conduits for submarine turbidite flows and basinward transport of clastic material supplied from a subaerial drainage system to the east and SE. It is suggested that during the time of deposition of supersequence F (Plio-Pleistocene) topset-clinoform systems composed of large shelf-edge deltas prograded along the Mediterranean coastline. Huge amounts of fine-grained siliciclastic sediments accumulated at the mouth of submarine canyons and channels. The great thickness of the unit west of the present coastline (Fig. 7f) is associated with particularly high sedimentation rates of about 300 m Ma -l. The rate of supply exceeded the rate of accommodation space, thus resulting in progradation of the shelf break to c. 20 km basinwards (at the top of line EM-22 near the Yam Yafo-I well; see Fig. 5). The intensity of the depositional processes is associated with the combined effects of global and local factors. These include a rapid rise of Pliocene sea level, subsidence within the basin as a result of sediment loading of the Nile cone, and the uplift of the Judea and Samaria mountain range located 20-50 km east of the basin.

Discussion

Tectonic evolution The mapping of the near-top Precambrian basement (base of supersequence A) (Fig. 8a) and top-Turonian surfaces (top of supersequence B) (Fig. 8b) reveals two important deformation phases of the Levantine basin and margin. An early, extensional phase caused normal faulting of the basement and the overlying Palaeozoic to Early Mesozoic strata. A later, contractional phase resulted in reverse faulting and folding and the formation of the Syrian Arc mountain range. The first phase is related to Early Mesozoic rifting that led to the separation of the AfricanArabian and Eurasian plates and the opening of

Neotethys. The second phase is associated with Late Mesozoic convergence and subduction that accompanied the closing of the Neotethys ocean system. Extensive evidence for the two deformation phases was previously found in outcrops and boreholes throughout the Levant region onshore (see Garfunkel 1998, 2004). Additional evidence for these events is presented here for the offshore Levantine basin.

Neotethyan rifting phase Within the study area, the eastern Mediterranean basin consists of a series of deep-seated, tilted fault blocks downfaulted from its margins (Israeli coastline and the Eratosthenes Seamount) towards its centre. The geometry of this fault system is recognized in the western and central parts of the basin where deep structures are well imaged. In this area the faults can be traced over long distances and extend from a few kilometres to several tens of kilometres (Fig. 8a). The spacing between the faults is in the range of 5-20 km and some of the blocks are asymmetrical half-grabens (Figs 2-4). The normal faults were active during the time span of supersequence A (Permian-Aalenian), dying out towards the top of the unit, or in some areas affecting the lower part of supersequence B (Bajocian-Turonian) (Figs 2-5). The vertical displacement on these faults is seen on the top of the basement horizon in the central and western parts of the basin, where it ranges from several hundred metres to 2-3 km (Fig. 8a). In the area east of the Eratosthenes Seamount normal faulting during the deposition of supersequence A accounts for almost the entire structural relief on top of the basement structural map (Figs 3 and 8a). On the eastern margin of the basin most of the deep-seated faults were inverted in the Senonian-Miocene and it is difficult to measure their original vertical displacement. However, thickness changes of supersequence A across some of the faults indicate downfaulting of at least a few hundred metres during the early, rifting phase. Unlike the western margin of the basin, normal faulting during supersequence A accounts only for half of the 8-10 km relief on top of the basement on the eastern margin (see line EM-22 in Fig. 5 and the isochore map in Fig. 7a). The amount and timing of vertical displacement in the eastern Levant margin are more accurately estimated from the subsurface of central and southern Israel. In this area, seismic and well data show evidence for considerable down-tothe-basin normal faulting in the Triassic and Jurassic section (Druckman 1977, 1984; Gelberman & Kemmis 1987; Garfunkel 1989, 1998; Cohen

LEVANTINE BASIN, OFFSHORE ISRAEL et al. 1990; Druckman et al. 1995). An example

of such a displacement is recorded on the Heletz Fault (Figs 8a and 9). This major, NE-SWtrending normal fault was mapped in the subsurface of the southern coastal plain of Israel by Gelberman (1995). A comparison of the depth of different stratigraphic markers found in the Gevim-1 well on the footwall, and the Heletz Deep-1 well on the hanging wall allows quantification of the history of motion (Table 2, Fig. 9). The Heletz fault was active in a down-to-thebasin, normal sense during the Neotethyan rifting phase and was reactivated in a reverse motion during the Late Cretaceous Syrian Arc contractional phase. The amount of reverse motion on the hanging wall of the fault can be estimated from the 486m displacement of the top-Aptian level (Table 2, Fig. 9). The normal throw of the deeper stratigraphic levels is calculated by adding the reverse motion to the measured vertical displacement (Table 2). The total, reconstructed normal throw of the near-top of the Precambrian basement is 1740 m. The vertical motion during the Triassic phase amounts to 283 m; this probably took place mainly during the Anisian and Carnian (Table 2) (Druckman 1984). A vertical motion of 741 m is calculated for a LiassicBajocian phase and 489 m for an Early to MidBathonian phase (Table 2). Increased thickness of the Triassic to Bathonian strata on the hanging wall of the Heletz Fault indicates syntectonic deposition (Fig. 9). The top-Bathonian and the reconstructed top-Callovian levels are similarly displaced (227 m and 247 m, respectively) suggesting some additional, post-Mid-Jurassic tectonic activity (Table 2). The Heletz Fault provides direct evidence for the activity of the Neotethyan rift system in the Levantine basin during the Mid-Triassic (Anisian), Early and Mid-Jurassic (LiassicBathonian). Rifting may have taken place already in the Permian, as can be inferred from other Neotethyan structures (Palmyra trough; Garfunkel, 1998). However, possible earlier activity in the Levantine basin is speculative because of lack of relevant well data. The overall thickness of supersequence A within the basin (central graben) is 4-8 km (Fig. 7a). Coeval units found within faultcontrolled depressions onshore, such as the interior basin of central Israel and the Palmyra trough of Syria, are only 3-5 km thick (Garfunkel 1998). It is, therefore, concluded that the centre of the Levantine basin was the most active part of the Early Mesozoic, Neotethyan rift system of the Levant area. The Levantine basin was bounded by the Arabian Massif to the SE and the Eratosthenes

221

continental block to the west. It was thus probably separated from the main body of the Neotethys Ocean that may have extended north of the Eratosthenes Platform and was later consumed underneath Cyprus and southern Turkey (Garfunkel 1998, 2004; Robertson 1998). An important question regarding the nature of the Neotethyan rifting processes is whether emplacement of new oceanic crust took place in the Levant area. Apart from faulting no pronounced disruption and conspicuous lateral variations in the seismic properties of either the upper part of the basement or supersequence A are observed in our dataset. The basement layer in the central part of the basin does not show any characteristics of oceanic crust such as described in other passive continental margins (Klitgord & Hutchinson 1988) or in the central part of the Mediterranean Sea (Finetti 1985; Avedik et al. 1995). The normal faults and basement structures associated with an early rifting stage are well preserved. Although a considerable amount of Triassic and Jurassic volcanic rocks may exist within the basin, their equivalent units in the onshore area (Asher volcanic series) do not show MORB characteristics associated with sea-floor spreading. Thus, the occurrence of spreading and introduction of new oceanic crust as implied by the 'oceanic' models are not supported by the present data. It is suggested that rifting during the Early Mesozoic was responsible for the significant thinning of the crust in the centre of the Levantine basin. However, rifting, thinning and oceanic physical properties, as interpreted from seismic velocities by Ben-Avraham et al. (2002), do not necessarily imply sea-floor spreading. The basin may not have reached the spreading phase as is the case in the northern parts of the Red Sea. There, the rift is continental, with only the nucleation of an oceanic spreading centre and an early magmatic phase. An oceanic spreading centre has come into existence only in the southern and central parts of the Red Sea (Martinez & Cochran 1988; Bohannon & Eittreim 1991; Cochran 2001). It should be emphasized that the Red Sea model compares well with the Levantine basin in term of physical dimensions and properties, although the duration of the rifting and early magmatic phases are entirely different: 10-15 Ma in the Red Sea and c. 75-80 Ma in the eastern Mediterranean. The general strike of the normal faulting mapped in our study in the Levantine basin is NE-SW. Accordingly, we interpret rifting and extension in a N W - S E direction perpendicular to the strike of the faulting. This interpretation

222

M.A. GARDOSH & Y. DRUCKMAN

cannot comply with rifting and spreading of the eastern Mediterranean in a north-south direction, accompanied by a transform fault along the eastern Mediterranean shoreline as proposed by Dewey et al. (1973), Bein & Gvirtzman (1977), Robertson & Dixon (1984) and Stampfli et al. (2001). The model of opening in a N W - S E direction and an east-west, strike-slip motion along the northern coast of Sinai, as suggested by Garfunkel & Derin (1984) and Garfunkel (1998) is in good accordance with our data. This model further implies that the eastern Mediterranean rift became detached from a Neotethys spreading centre that existed further north, beyond the northern palaeo-margin of the Arabian Massif. The passive nature of the eastern Levant margin with its rimmed carbonate shelf was established as early as late Mid- and Late Jurassic (Bathonian-Oxfordian), and persisted until the end of the Turonian (Gardosh 2002). Our data show marked differences between the eastern and western margins of the eastern Mediterranean basin during this time span: the density and dimensions of faulting are different on the two sides (Fig. 8a and b), thick sedimentary depocentres of supersequences A and B are located near the eastern margin but are missing from the western side (Fig. 7a and b), and no distinct passive margin geometry is identified near the Eratosthenes Seamount (Fig. 3). The only shallow-water sediments recorded from the Eratosthenes area are of Miocene age (Mart & Robertson 1998). Rybakov & Segev (2004) recently questioned the assumed shallow depth and continental nature of the basement underlying the Eratosthenes Seamount. Our dataset reaches only the eastern Eratosthenes and does not permit a comprehensive analysis of this area. A better understanding of the geology of the Eratosthenes Seamount is essential to fully reconstruct the evolution of the eastern Mediterranean basin during Early Mesozoic times. S y r i a n A r c inversion a n d c o n t r a c t i o n p h a s e

An elevated zone of contractional deformation extending offshore, some 70 km west of the present coastline (Fig. 8b), is interpreted as the subsurface continuation of the 'Syrian Arc' (Krenkel 1924) or 'Levantine' (Horowitz 1979) fold belt. This regional tectonic feature forms a S-shaped mountain belt extending from the Palmyride Mountains in Syria and Lebanon and the Anti-Lebanon Mountains in Lebanon, through the Judea Mountains and Negev anticlines in Israel and into the northern Sinai anticlines in Egypt (Fig. 1) (Picard 1943, 1959;

Ball & Ball 1953; Bentor & Vroman 1954, 1960; de Sitter 1962; Gvirtzman 1970; Bartov 1974; Neev et al. 1976; Garfunkel 1978; Horowitz 1979; Eyal & Reches 1983; Lovelock 1984; Beydoun 1988; McBride et aL 1990). Its extension further SW is masked by the thick MioPliocene sediments of the Nile delta (Aal et al. 2000). In the study area, supersequence B rises some 3500 m from the foot of the Syrian Arc mountain belt in the west to its top near the coastline (Figs 5 and 8b). The dominant deformational style within the fold belt is high-amplitude and shortwave length anticlines and monoclines, underlain by high-angle thrust faults (Figs 5 and 6). Individual structures range from 10 to 30 km in length and from 5 to 10 km in width. Their height ranges from several hundred metres to more than 1000 m, and their flanks dip at 10-30 ~ The thrust faults dip at 65-75 ~ and can be traced down into the basement and up to the Middle-Upper Cretaceous section. A second, less widespread style of deformation observed within the contractional belt is of wider folds (up to 15 km) with lower amplitudes of up to several hundred metres. In the eastern margin, these broad and shallow folds are occasionally found superimposed on high-amplitude, narrow folds (Fig. 6). Some low-amplitude folds of the second style of deformation are found also in the deeper basin, west of the Syrian Arc fold belt. The kinematics of these structures is currently not fully understood. The entire supersequences A, B and C and the lower part of supersequence D are affected by the contractional deformation. Thickness variations and onlapping reflections within supersequences C and D indicate that they were deposited contemporaneously with the folding. Assuming only minor mis-correlations and incorrect dating of our seismic units it is suggested that the contraction phase started in the Late Cretaceous and persisted until the Early-Mid-Miocene (Fig. 5). The folding took place in recurrent episodes, thus resulting in more intensive deformation of the older sections. A similar timing for the onset of folding was suggested from studies of the Syrian Arc (e.g. Bentor & Vroman 1960; Bartov 1974; Beydoun 1988; McBride et al. 1990; Garfunkel 1998). However, the time of cessation was previously not well established, mainly because of the lack of relevant field relations. Compelling evidence for inversion of Early Mesozoic extensional structures during the Late Cretaceous folding phase comes from various parts of the Syrian Arc fold belt in onshore Israel and Syria (Freund et al. 1975; Druckman 1981; Davis 1982; Gelbermann & Kemmis

LEVANTINE BASIN, OFFSHORE ISRAEL 1987; Bruner 1991; Best et al. 1993; Chaimov et al. 1993; Druckman et al. 1995; Gardosh et al. 2003). The interpretation of the present seismic data indicates that similar inversion took place on almost all faults underlying the fold structures of the Levantine basin and its margins (Figs 5 and 6). Various workers suggested that the folding of the Syrian Arc is associated with closing of the Neotethys Ocean and collision of the AfricanArabian and Eurasian plates. Our results further indicate an important association between the Syrian Arc deformation and the deep structure of the Levantine basin. The depth map of the top of supersequence B (Fig. 8b) and the isopach map of supersequence C (Fig. 7c) show that the uplifted and highly deformed Syrian Arc fold belt terminates sharply along a NE-SW-trending line located 50-70 km west of the present-day coastline. The western edge of the belt correlates well with a line of separation between the two main crustal units of the Levantine basin, as defined by deep-refraction, gravity and magnetic data (Fig. 8b; Ben-Avraham et al. 2002). Most of the Syrian Arc contractional deformation occurs within the continental-type crust on the eastern margin, whereas the area of thinned crust to the west is considerably less deformed. It is suggested that the distribution of the Syrian Arc deformation as well as the considerable uplift of the Syrian fold belt (Figs 5 and 8b) are controlled by variations in the physical properties of the crust underlying the Levantine basin and its margin. A more comprehensive analysis of this relation is beyond the scope of the present study. Finally, the western edge of the Syrian Arc fold belt correlates also with the Pelusium Line, as defined by Neev's et al. (1976). However, Neev's suggestion that this lineament is a megashear zone is not supported by our findings.

Summary and conclusions (1) The Levantine basin is a 15 km deep basin that came into existence during Early Mesozoic, Neotethyan rifting. Faulting took place during the Anisian (Mid-Triassic) and continued through the Liassic, Bajocian and Bathonian (Early-Mid-Jurassic). Normal faulting ceased during the Mid- to Late Jurassic. (2) The basin opened through break-up and extension in a N W - S E direction, accompanied by a strike-slip motion along its southern margin. Extension in this direction does not require the existence of compensation

(3)

(4)

(5)

(6)

(7)

(8)

(9)

223

by a north-south-oriented transform fault along the Levant shoreline. No indications of sea-floor spreading were found in the Levantine basin. Thick accumulations of Upper Triassic to Liassic volcanic rocks in the northeastern part of the basin show no MORB characteristics. The Eratosthenes Seamount may have been the basin's western margin. The basin is an intracontinental rift, which reached only an early magmatic phase, associated with the nucleation of an oceanic spreading centre. High-velocity basement in the centre of the basin is interpreted as thinned and intruded continental crust The contractional features of the Syrian Arc fold belt were formed by the inversion of movement on pre-existing normal faults at the eastern margin of the basin. The folding occurred in several pulses starting in the Senonian and ceasing in the Miocene. The distribution of the Syrian Arc deformation was controlled by basement properties. The deformation occurs predominantly within the area underlain by continental crust. The western limit of the main fold belt, located 50-70 km west of the coastline, coincides with a zone of transition in basement properties, previously described as the Pelusium Line. The Messinian evaporitic basin was limited in the east by the uplifted fold belt composed of older strata of Oligocene to Mid-Miocene age. During highstand episodes of the Messinian sea, evaporites were deposited on the higher slope and within canyons incised into the shelf. The Plio-Pleistocene basinward progradation of the shelf break by some 20 km across the drowned Mesozoic shelf break was caused by the high sedimentation rates of both Nile and locally derived sediments. These rates exceeded the accommodation space created by the rise of the Pliocene sea level and by local subsidence.

The authors would like to thank B. Buchbinder and A. Flexer for comprehensive reviews of the manuscript and their most helpful remarks. We would like to thank the Israeli Ministry of Infrastructure for allowing us the use of the well and seismic dataset.

References AAL, A. A., EL BARKOOKY,A., GERRITS,M., MEYER, H. SCHWANDER,M. & ZAKI, H. 2000. Tectonic evolution of the eastern Mediterranean Basin and

224

M . A . GARDOSH & Y. D R U C K M A N

its significance for hydrocarbon prospectivity in the ultra-deepwater of the Nile Delta. The Leading Edge, 19(10), 1086-1102. ALMOGI-LABIN, A., BE~N, A. & SASS, E. 1993. Late Cretaceous upwelling system along the southern Tethys margin (Israel): interrelationship between productivity, bottom water environments and organic matter preservation. Paleoceanography, 8, 671-690. AVEDIK, F., NICOLICH, R., HIRN, A., MALTEZOU, F., MCBRIDE, J., CERNOBORI, L. & the STREAMERS/ PROFILES Group 1995. Appraisal of a new, highenergy and low frequency seismic pulse generating method on a deep seismic reflection profile on the Central Mediterranean Sea. First Break, 13, 277-290. BALL, M. W. & BALL, D. 1953. Oil prospects of Israel. American Association of Petroleum Geologists" Bulletin, 37, 1-113. BARTOV, Y. 1974. A structural and paleogeographical study of the central Sinai faults and domes. PhD thesis, Hebrew University, Jerusalem (in Hebrew with English abstract). BEIN, A. & GVIRTZMAN, G. 1977. A Mesozoic fossil edge of the Arabian plate along the Levant coastline and its bearing on the evolution of the eastern Mediterranean. In: BIJU-DUVAL, B. & MONTADERT, L. (eds.) Structural History of the Mediterranean Basins. Technip, Paris, 95-110. BEN-AVRAHAM, Z. 1989. Multiple opening and closing of the eastern Mediterranean and south China Basin. Tectonics, 8, 351-362. BEN-AVRAHAM, Z. & GRASSO, M. 1991. Crustal structure variations and transcurrent faulting at the eastern and western margins of the eastern Mediterranean. Tectonophysics, 196, 269-277. BEN-AVRAHAM, Z. & TIBOR, G. 1994. Structure and tectonics of the Eastern Cyprean arc. Terra abstracts suppl. 1 to Terra Nova, 5, 254. BEN-AVRAHAM, Z., SHOHAM,Y. & GINZBURG, A. 1976. Magnetic anomalies in the eastern Mediterranean and the tectonic setting of the Eratosthenes Seamount. Geophysical Journal of the Royal Astronomical Society, 45, 105-123. BEN-AVRAHAM, Z., TIBOR, G., LIMONOV, A. F., LEYBOV, M. B., IVANOV, M. K., TOKAREV, M. YU. & WOODSIDE, J. M. 1995. Structure and tectonics of the eastern Cyprean Arc. Marine and Petroleum Geology, 12, 263-271. BEN-AVRAHAM, Z., GINZBURG, A., MAKRIS, J. & EPPELBAUM, L. 2002. Crustal structure of the Levant Basin, eastern Mediterranean. Tectonophysics, 346, 23-43. BEN-GAI, Y. 1996. Sequence stratigraphy of the PlioPleistocene in the continental margin of the southern Levant. PhD thesis, Tel Aviv University. BENTOR, Y. K. & VROMAN, A. 1954. A structural contour map of Israel (1;250 000), with remarks on its dynamic interpretation. Bulletin of the Research Council, Israel, 4, 125-235. BENTOR, Y. K. & VROMAN, A. 1960. The geological map of Israel (1:100 000). Sheet 16." Mount Sdom, explanatory text. Geological Survey of Israel. BEST, J. A., BARAZANGI,M., AL-SAAD, D., SAWAF, T. & GEBRAN, A. 1993. Continental margin evolution

of the Northern Arabian Platform in Syria. American Association of Petroleum Geologists Bulletin, 77, 173-193. BEYDOUN, Z. R. 1988. The Middle East: Regional Geology and Petroleum Resources. Scientific Press, Beaconsfield, UK. BOHANNON, R. G. & EITTREIM, S. L. 1991. Tectonic development of passive continental margins of the southern and central Red Sea with a comparison of Wilkes Land, Antarctica. Tectonophysics, 198, 129-154. BRUNER, I. 1991. Investigation of the subsurfaee in the northern Negev, Israel, using seismic reflection techniques. PhD thesis, Tel Aviv University. BUCHBINDER, B. & SIMAN-TOV, R. 2000. Oligocene clastics in the Ashdod area: mass flow and turbidite deposits in a submarine-fan setting. Israel Geological Society Annual Meeting (abstract), Ma'alot, 25. BUCHBINDER, B. & ZILBERMAN, E. 1997. Sequence stratigraphy of Miocene-Pliocene carbonatesiliciclastic shelf deposits in the eastern Mediterranean margin (Israel): effects of eustasy and tectonics. Sedimentary Geology, 112, 7-32. BUCHBINDER, B., BENJAMINI, C., MIMRAN, Y. & GVlRTZMAN, G. 1988. Mass transport in Eocene pelagic chalk on the northwestern edge of the Arabian platform, Shefela area, Israel. Sedimentology, 35, 257-274. BUCHBINDER, B., BENJAMINI, C. & LIPSON-BNITAH, S. 2000. Sequence development of Late CenomanianTuronian carbonate ramps, platforms and basins of Israel. Cretaceous Research, 21, 813-843. CHAIMOV, Y. A., BARAZANGI, M., AL-SAAD, D., SAWAF, T. & KHADDOURI, M. 1993. Seismic fabric and 3-D structure of the southwestern intracontinental Palmyride fold belt, Syria. American Association of Petroleum Geologists Bulletin, 77, 2032-2047. COCHRAN, J. R. 2001. Nucleation of an oceanic spreading center in a continental rift: the northern Red Sea. http://www.copernicus.org/EGS/egsga/ Nice01/programme/abstracts/aai1791.pdf. COHEN, Z. 1976. Early Cretaceous buried canyon: influence on accumulation of hydrocarbons in the Helez oil field, Israel. American Association of Petroleum Geologists Bulletin, 60, 108-114. COHEN, Z., FLEXER, A. & KAPTSAN, V. 1988. The Pleshet basin, a newly discovered link in the peripheral chain of basins of the Arabian craton. Journal of Petroleum Geology, 11(4), 403-414. COHEN, Z., FLEXER, A. & KAPTSAN, V. 1990. The tectonic mosaic of the southern Levant: Implication for hydrocarbon prospects. Journal of Petroleum Geology 13(4), 437-462. DAVIS, B. K. 1982. Structural inversion in Europe and Israel and its importance in oil exploration. Oil Exploration (Invest.) Ltd Report OEL/82104. DERCOURT, J., ZONENSHA1N, 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. DERIN, B. 1974. The Jurassic of central and northern Israel. PhD thesis, Hebrew University, Jerusalem.

LEVANTINE BASIN, OFFSHORE ISRAEL DERIN, B., GERRY, E. & LIPSON, S. 1982. Asher Atlit-1, northern coastal plain, stratigraphic log. Israel Institute of Petroleum and Energy Report 2/82. DERIN, B., LIPSON, S. & GERRY, E. 1990. Microbiostratigraphy and environments of deposition of the Yam 2 offshore well. Israel Institute of Petroleum and Energy Report 5190. DE SITTER, t. U. 1962. Structural development of the Arabian Shield in Palestine. Geologie en Mjnibouw, l l , 116-124. DEWEY, J. F., PITMAN, W. C., III, RYAN, W. B. F. & BONIN, J. 1973. Plate tectonics and the evolution of the Alpine system. Geological Society of America Bulletin, 84, 3137-3180. DRUCKMAN, Y. 1977. Differential subsidence during the deposition of the Lower Jurassic Ardon Formation in western Jordan, southern Israel and northern Sinai. Israel Jourmal of Earth Sciences, 26, 45-54. DRUCKMAN, Y. 1981. Comments on the structural reversal model as a factor of the geological evolution of Israel. Israel Journal of Earth Sciences, 30, 44-48. DRUCKMAN, Y. 1984. Evidence for Early-Middle Triassic faulting and possible rifting from the Helez Deep borehole and the coastal plain of Israel, In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 203-212. DRUCKMAN, Y., CONVCAY,B., ESHET, Y., et al. 1994. The stratigraphy of the Yam Yafo-1 borehole. Geological Survey of Israel Report GSI/28/94. DRUCKMAN, Y., GILL, D., FLEISCHER, L., GELBERMANN, E. & WOLFF, O. 1995. Subsurface geology and structural evolution of the northwestern Negev, southern Israel. Israel Journal of Earth Science, 44, 115-136. DVORKIN, A. & KOHN, B. P. 1989. The Asher Volcanics, northern Israel: petrography, mineralogy and alteration. Israel Journal of Earth Sciences', 38, 105-123. EL-ISA, Z. H., MECHIE, J., PRODEHL, C., MAKRIS, J. & RIHM, R. 1987. A crustal structure study in Jordan derived from seismic refraction data. Tectonophysics, 138, 235-253. EMERY, D. & MYERS, K. 1996. Sequence Stratigraphy. Blackwell Science, Oxford. EYAL, Y. & RECHES, Z. 1983. Tectonic analysis of the Dead Sea rift region since the Late Cretaceous based on mesostructures. Tectonics, 2, 167-185. FINETTI, I. 1985. Structure and evolution of the central Mediterranean (Pelagian and Ionian Seas). In: STANLEY, D. J. & WEZEL, F. C. (eds) Geological Evolution of the Mediterranean Basin. Springer, Berlin, 215-230. FLEXER, A. 1968. Stratigraphy and facies development of Mount Scopus Group (Senonian-Paleocene) in Israel and adjacent countries. Israel Journal of Earth Sciences, 17, 85-114. FLEXER, m., ROSENFELD, A., LIPSON-BEN1TAH, S. & HONIGSTEIN, A. 1986. Relative sea level changes during the Cretaceous in Israel. American Association of Petroleum Geologists Bulletin, 70, 16851699.

225

FOLKMAN, Y. & BEN-GAI, Y. 2004. The Jonah buried Seamount: intrusive structure in the southeastern Levant Basin, offshore Israel. Israel Geological Society Annual Meeting (abstract). FREUND, R., GOLDBERG, M., WEISSBROD, T., DRUCKMAN,Y. & DERIN, B. 1975. The Triassic-Jurassic structure of Israel and its relation to the origin of the eastern Mediterranean. Geological Survey of Israel Bulletin, 65. GARDOSH, M. 2002. Tile sequence stratigraphy and petroleum systems of the Mesozoic, southeastern Mediterranean continental margin. PhD thesis, Tel Aviv University. GARDOSH, M., BRUNER, I., DRUCKMAN, Y., REZNIKOV, M., BUCHBINDER,B. & SAGY, Y. 2003. Mesozoic tectonic phases and hydrocarbon potential of the southeastern Mediterranean basin. Results from new, offshore seismic surveys. Israel Geological Society Annual Meeting (abstract), Ein Boqeq, 39. GARFUNKEL, Z. 1978. The Negev--regional synthesis of sedimentary basins. In: lOth International Congress of Sedimentology, Guidebook, Part I, 34-110. GARFUNKEL, Z. 1989. Tectonic setting of Phanerozoic magmatism in Israel. Israel Journal of Earth Science, 38, 51-74. GARFtrNK_EL, Z. 1998. Constraints on the origin and history of the eastern Mediterranean basin. Tectonophysics, 298, 5-35. GARFUNKEL, Z. 2004. Origin of the Eastern Mediterranean basin: a reevaluation. Tectonophysics, 391, 11-34.

GARFUNKEL, Z. & DERIN, B. 1984. Permian-Early Mesozoic tectonism and continental margin formation in Israel and its implications to the history of the eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 18-201. GELBERMANN, E. 1995. Revised seismic structural and isopach maps, central and southern Israel. Institute for Petroleum Research and Geophysics Report 201/22/88-94. GELBERMANN, E. & KEMMIS, J. L. 1987. Triassic to Late Cretaceous tectonic alignments in the southern coastal plain and the northwestern Negev. Israel Geological Society Annual Meeting (abstract), Mizpe Ramon, 39. GILL, D., CONWAY, B.H., ESHET, Y., LIPSON, S., PERELIS-GROSSOWICZ,L., ROSENFELD,A. & SIMANToy, R. 1995. The stratigraphy of the Yam West 1 borehole. Geological Survey of Israel Report GSI/13195.

GINZBURG, A. & BEN-AVRAHAM, Z. 1987. The deep structure of the central and southern Levant continental margin. Annales Tectonicae, 1, 105-115. GINZBURG, A. & FOLKMAN, Y. 1980. The crustal structure between the Dead Sea rift and the Mediterranean Sea. Earth and Planetary Science Letters, 51, 181-188. GINZBURG, A., COHEN, S. S., HAY-RAY, H. & ROZENZWEIG, A. 1975. Geology of the Mediterranean shelf of Israel. American Association of Petroleum Geologists Bulletin, 59, 2142-2160.

226

M.A. GARDOSH & Y. D R U C K M A N

GINZBURG, A., MAKRIS, J., FUCHS, K., PRODEHL, C., KAMINSKI,W. & AMITA1, U. 1979. A seismic study of the crust and upper mantle of the Jordan-Dead Sea rift and their transition towards the Mediterranean Sea. Journal of Geophysical Research, 84, 1569-1582. GRADMANN, S., HUBSCHER, C., BEN-AVRAHAM, Z., GAJEWSKI, B. & NETZEBAND, G. 2005. Salt tectonics off northern Israel. Marine and Petroleum Geology, 22, 597-611. GVIRTZMAN, G. 1970. The Saqiye Group (Late Eocene to Early Pleistocene) in the Coastal Plain and Hashephela regions, Israel. Geological Survey of Israel Report ODIS/67. GVIRTZMAN, G. & BUCHBINDER,B. 1978. The Tertiary history of the coastal plain and continental shelf of Israel and its bearing on the history of the Eastern Mediterranean. In: Ross, D. A., NEPROCHNOV, Y. P., et al. (eds) Initial Reports of the Deep Sea Drilling Project, 42B. US Government Printing Office, Washington, DC, 1195-1222. GVIRTZMAN, G. & REISS, Z. 1965. Stratigraphic nomenclature in the Coastal Plain and Hashephela regions. Geological Survey of Israel Report ODIll 65. GVIRTZMAN, G. & STEINITZ, G. 1983. The Asher Volcanics--an Early Jurassic event in northern Israel. Israel Geological Survey Current Research, 28-33. HAQ, B. U., HARDENBOL,J. & VAIL, P. R. 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level changes. In: WILGUS, C. K., HASTINGS, B. S., KENDALL,C. G., POSAMENTIER,H. W., Ross, C. A. & WAGONER, J. C. (eds) Sea-level Changes." an Integrated Approach. Society of Sedimentary Geology, Special Publications, 42, 71-108. H1RSCH, E., FLEXER, A., ROSENFELD, A. & YELLINDROR, A. 1995. Palinspastic and crustal setting of the eastern Mediterranean. Journal of Petroleum Geology, 18(2), 149-170. HIRSCH, F., BASSOULLET,J. P., CARIOU, E., el al. 1998. The Jurassic of the southern Levant: biostratigraphy, palaeogeography and cyclic event. ln: CRASQUIN-SOLEAU, S. & BARRIER, E. (eds) Peri-Tethys Memoir 4: Epicratonic Basins of Peri-Tethyan Platforms. Mrmoires du Museum National d'Histoire Naturelle, 179, 213-235. HOROWITZ, A. 1979. The Quaternary of Israel. Academic Press, New York. HsO, K. J., RYAN, W. B. F. & CITA, M. B. 1973. Late Miocene desiccation of the Mediterranean. Nature, 242, 240-244. HsI~I, K. J., MONTADERT, L., BERNOULLI, D., et al. 1978. History of the Mediterranean salinity crisis. In. Hsu, K., MONTADERT, L., et al. Initial Reports of the Deep Sea Drilling Project, 42A. US Government Printing Office, Washington, DC, 1053-1078. KLITGORD, K. D. & HUTCHINSON, D. R. 1988. U.S. Atlantic continental margin; structural and tectonic framework. In: SHERIDAN, R. E. & GROW, J. A. (eds) The Atlantic Continental Margin, U.S. Geological Society of America, The Geology of North America, I-2, 19-55.

KOHN, B. P., LANG, B. & STEINITZ, G. 1993.4~ dating of the Atlit one volcanic sequence, northern Israel. Israel Journal of Earth Science, 42, 17-23. KORNGREEN, D. 2004. The Triassic in northern Israel. PhD thesis, Ben Gurion University, Beer-Shevu. KRENKEL, E. 1924. Der Syrische Bogen. Zentralblatt Mineralogie, 9, 274-281; and 10, 301-313. LOVELOCK, P. E. R. 1984. A review of the tectonics of the northern Middle East region. Geological Magazine, 121, 577-587. MAKRIS, J. & WANG, J. 1994. Bouguer gravity anomalies of the Eastern Mediterranean Sea. In: KRASHENINIKOV, V. A. & HALL, J. K. (eds) Geological structure of the north-eastern Mediterranean (Cruise 5 of the Research Vessel Akademik Nikolaj Strakhov). Historical Productions-Hall, Jerusalem, 87-98. MAKRIS, J., BEN-AVRAHAM,Z., BEHLE, A., et al. 1983. Seismic refraction profiles between Cyprus and Israel and their interpretation. Geophysical Journal of the Royal Astronomical Society, 75, 575-591. MART, Y. & ROBERTSON, A. H. F. 1998. Eratosthenes Seamount: an oceanic yardstick recording the Late Mesozoic-Tertiary geological history of the eastern Mediterranean. 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, 701-708 MARTINEZ, F. & COCHRAN, J. R. 1988. Structure and tectonics of the Red Sea: catching a continental margin between rifting and drifting. Tectonophysics, 150, 1-32. MCBRIDE, J. H., BARAZANGI, M., BEST, J., AL-SAAD, D., SAWAF, T., AL-OTRI, M. & GEBRAN, A. 1990. Seismic reflection structure of intracratonic Palmyrid fold-thrust belt and surrounding Arabian Platform, Syria. American Association of Petroleum Geologists Bulletin, 74, 238-259. NEEV, D. 1977. The Pelusium Line--a major transcontinental shear. Tectonophysics, 38, T1-T8. NEEV, D., ALMAGOR, G., ARAD, A., G1NZBURG,A. & HALL, J. K. 1976. The geology of the southeastern Mediterranean. Geological Survey of Israel Bulletin, 68. PICARD, L. 1943. Structure and evolution of Palestine with comparative notes on neighboring countries. Bulletin, Geological Department, Hebrew University, 4, 1-143. PICARD, L. 1959. Geology and oil exploration of Israel. Bulletin of the Research Council in Israel, GS, 1-30. 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. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds.) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74.

LEVANTINE BASIN, O F F S H O R E ISRAEL ROUCHY, J. M. & SAINT MARTIN, J. P. 1992. Late Miocene events in the Mediterranean as recorded by carbonate-evaporite relations. Geology, 20, 629-632. RYAN, W. B. F. 1978. Messinian badlands on the southeastern margin of the Mediterranean Sea. Marine Geology, 27, 349-363. RYBAKOV, M. & SEGEV,A. 2004. Top of the crystalline basement in the Levant. Geochemistry, Geophysics and Geodynamics, 5, 1-8. RYBAKOV, M., GOLDSHMIT, V. ~ ROTSTEIN, Y. 1997. New regional gravity and magnetic maps of the Levant. Geophysical Research Letters, 24, 33-36. SASS, E. & BEIN, A. 1982. The Cretaceous carbonate platform in Israel. Cretaceous Research, 3, 135-144. ~ENG~)R, A. M., Y1LMAZ, Y. & SUNGURLU, O. 1984. Tectonics of the Mediterranean Cimmerides: nature and evolution of the western termination of Paleo-Tethys. In: DIXON, J. E. & ROBERTSON,

227

A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 77-112. STAMPFLI,G. M., MOSAR, J., FAVRE, P. A., PILLEVUIT, A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethyan realm: the Neotethys East-Mediterranean Basin connection. In: ZIEGLER, P. A., CAVAZZA,W., ROBERTSON, A. U. F. & CRASQUIN-SOLEAU, S. (eds) Peri-Tethys

Memoir 6: Peri-Tethyan Rift~Wrench Basins and Passive Margins. M6moires du Museum National d'Histoire Naturelle, 186, 51-108. STEINITZ, G., GVIRTZMAN,G. & LANG, B. 1983. Evaluation of K-Ar ages of the Asher volcanics. Geological Survey of Israel, Current Research 1982, 34-38. WOODSIDE,J. M. 1977. Tectonic elements and crust of the eastern Mediterranean Sea. Marine Geophysical Researches, 3, 317-354.

Biochronology of Jurassic and Early Cretaceous radiolarites from the Lycian M61ange (SW Turkey) and implications for the evolution of the Northern Neotethyan ocean T. D A N E L I A N l, A. H. F. R O B E R T S O N % A. S. C O L L I N S 3 & A. P O I S S O N 4

1MicropalOontologie, UniversitO P. & M. Curie (Paris VI), C N R S - U M R 5143, C. 104, 4 place Jussieu, 75005 Paris, France (e-mail: [email protected]) 2Grant Institute of Earth Science, School of GeoSciences, University o f Edinburgh, West Mains Road, Edinburgh EH9 3JW, UK 3School of Earth and Environmental Sciences, University of Adelaide, Adelaide, SA 5005, Australia 4DOpartement des Sciences de la Terre, UniversitO de Paris-Sud, Orsay, France

Abstract: New radiolarian ages for blocks of radiolarian cherts associated with other blocks of distal pelagic facies and ophiolitic lithologies within the Lycian M61ange, SW Turkey, indicate deposition during Mid-Jurassic to Early Cretaceous time. Radiolarites overlying pink pelagic limestones of an allochthonous carbonate unit accumulated during the Mid- to Late Jurassic. On the basis of structural evidence the Lycian M61ange is inferred to have been rooted within the Northern Neotethys, to the north of the Tauride-Anatolide microcontinent. The Lycian radiolarites can be compared with other dated radiolarites from the Izmir-Ankara suture, the root zone of the Northern Neotethyan ocean. Based on all the available radiolarian data it is inferred that radiolarites accumulated within the Northern Neotethys in western Turkey from Late Triassic (Mid-Carnian to Late Norian) to MidCretaceous (Cenomanian) time. The radiolarites were later detached from their inferred oceanic basement and accreted within a subduction complex during the Late Cretaceous (Turonian-Maastrichtian) and emplaced over the northern margin of the TaurideAnatolide continent together with slices of continental margin and ophiolitic lithologies.

Following the development of Mesozoic radiolarian biostratigraphical schemes during the last 25 years, radiolarian biochronology has become invaluable for reconstructing the palaeogeography, palaeoceanography and tectonic evolution of Tethyan oceanic basins and margins (De Wever 1989; Robertson et al. 1991; De Wever et al. 1994; Danelian & Robertson 1997, 2001). Radiolarites in the Tethyan realm commonly occur within m61anges that are interpreted as parts of accretionary prisms related to subduction of ocean crust (Robertson 2002, 2004). Radiolarian ages from such m61anges provide valuable age constraints on related geological processes, including spreading, intra-plate volcanism and tectonic accretion (e.g. A1 Riyami et al. 2001; Beccaletto et al. 2005). Here, we will present the first radiolarian evidence from the Lycian M61ange, an eastwest-trending Late Mesozoic accretionary complex in SW Turkey (Collins & Robertson 1997; Fig. 1). This m61ange forms part of the wellknown Lycian Nappes in SW Turkey (De Graciancky 1972; Poisson 1977, 1984; Collins &

Robertson 1998). The root zone of the m61ange is considered to lie to the north of the regional 'basement' represented by the Tauride-Anatolide platform (Collins & Robertson 1998, 2003), and is inferred to lie within the Izmir-Ankara suture zone to the north of the Menderes Massif (Robertson & Pickett 2000; Robertson et al. 2004). We will also review the limited available data for radiolarites from comparable tectonic units within the Izmir-Ankara suture zone and the Bey Da~lan. We conclude by considering the implications of the available radiolarian evidence for the evolution of the Mesozoic Tethyan ocean.

Geological setting The Lycian Nappes occupy a large area to the NW of the Mesozoic Bey Da~lan carbonate platform and to the SE of the Menderes Massif (De Graciansky 1972; Bernoulli et al. 1974; Collins & Robertson 1997, 1998, 1999; Poisson 1977; Fig. 2). Five major tectonostratigraphic units are recognized in ascending order, as follows.

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds)2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 229-236. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

230

T. DANELIAN E T AL.

Fig. 1. Distribution of ophiolite and m61ange outcrops (in black) in central and western Turkey. A, B, and C are previously dated radiolarite localities (see text for explanation).

(1) The relatively autochthonous Menderes Massif to the NW and the Bey Da~larl platform to the SE, which represent the 'basement' onto which the Lycian Nappes were emplaced between the latest Cretaceous and Miocene times. (2) The Lycian Thrust Sheets comprise Mesozoic-Early Tertiary platform, slope and basinal successions that are dominated by sedimentary rocks with subordinate amounts of extrusive igneous rocks. The Lycian Thrust Sheets are reconstructed as a passive continental margin that faced northwards into the northern branch of the Mesozoic Tethys, known as the Northern Neotethyan ocean ($eng6r & Yllmaz 1981). However, the lowest thrust sheet, the Yavuz Thrust Sheet (Poisson 1977, 1984; Senel 1981) is restored as an Early Tertiary foreland basin that was not finally incorporated into the Lycian Nappes until the latest stage of southward emplacement during Oligo-Miocene time (Collins & Robertson 1999, 2003). (3) The Lycian M61ange tectonically overlies the Lycian Thrust Sheets and is interpreted as an accretionary prism related to the subduction

of Mesozoic Tethyan (i.e. Neotethyan) oceanic crust (Collins & Robertson 1997). This is composed of two intergradational, but tectonically juxtaposed, subunits. First, there is the structurally lower Layered Tectonic M61ange (up to several hundred metres thick) that is mainly composed of dismembered thrust sheets and detached blocks of pelagic limestone of up to 10 m (sized units), together with thin sequences (<10m) of highly disrupted ribbon radiolarites within a highly sheared shaly matrix. Second, above this, there is the Ophiolitic M61ange that comprises blocks and dismembered thrust sheets (up to 1 km in size) of ophiolitic lithologies including serpentinized harzburgite, gabbro, diabase and basalt. There are also generally smaller blocks of red-purple radiolarites and both neritic and pelagic limestone, all within a deformed matrix of terrigenous sediments, mainly turbidites and shales. In addition, there are large disrupted thrust sheets and detached blocks of Mesozoic shallow-water limestone, known as the Domuz Da~l unit (Poisson 1977, 1984). This typically forms dismembered unitsentrained within the Ophiolitic M61ange.

MESOZOIC RADIOLARITES, SW TURKEY

231

Fig. 2. Outcrops of the Menderes Massif, the Bey Da~larl carbonate platform, the Lycian Ophiolite and the Ophiolitic M61ange. The locations of the radiolarite outcrops dated during this study are also shown.

(4) The Lycian Peridotite Thrust Sheet, situated at the top of the Lycian thrust stack, represents the mainly ultramafic portion of oceanic mantle. This is interpreted as Cretaceous Neotethyan oceanic lithosphere that formed above a subduction zone (Collins & Robertson 1998). The serpentinized harzburgites are underlain by a dismembered metamorphic sole that has been radiometrically dated as Early Cretaceous (101-88 Ma, but mainly 95-90 Ma; Thuizat et al. 1981; Dilek et al. 1999). Geochemical studies of the metamorphic sole suggest that the protoliths were potentially derived from a range of island arc, mid-ocean ridge and seamount settings (~elik & Delaloye 2003).

Locality information and radiolarian ages Over 30 chert samples from different localities associated with the Ophiolitic M61ange yielded

Radiolaria; however, only four contained agediagnostic taxa and are discussed here. The location of dated samples is indicated in Figure 2. Age-diagnostic radiolarian species are illustrated in Figure 3. Radiolarians were extracted by repetitive leaching of samples with low-concentration hydrofluoric acid (4% HF). Sample 3s5b is from a roadside outcrop of radiolarite, 6 m thick, situated along a tarmac road c. 5 km south of the village of Sofular (29~ 36~ Red or purple chert is interbedded with siliceous red mudstones. Radiolarites are separated from adjacent neritic limestones by a fault. However, further SW, these radiolarites stratigraphically overlie similar limestones, and are overthrust by ophiolitic m61ange. Elsewhere, on a mountainous ridge between Cameli and Sofular, the limestone-radiolarite lithologies form kilometre-scale blocks within the ophiolitic m61ange. Sample 3s5b yielded a reasonably well-preserved radiolarian fauna

232

T. DANELIAN ET AL.

Fig. 3. Scanning electron micrographs of Mid-Jurassic to Early Cretaceous Radiolaria from radiolarite of the Lycian Ophiolitic M61ange. (a) Sethocapsa (?) zweilli Jud, sample 3s5b; (b) Transhsuum sp.cf. T. brevicostatum (Ozvoldova), sample A8; (c) Zartus praejonesi Pessagno & Blome, sample ALT 5; (d) Tethysetta dhimenaensis Baumgartner s.L, sample TOI-324; (e) Podocapsa amphitreptera Foreman TO1-324.

including Archaeodictyomitra apiarium (Riist), A. mitra Dumitrica, Sethocapsa (?) zweillii Jud, S. sp.cf. S. uterculus (Parona) s.l. and Xitus sp. The presence of species S. (?) zweillii indicates a Berriasian to Early Hauterivian (Early Cretaceous) age (Unitary Association Zones UAZ 14-19; Baumgartner et al. 1995). Sample A8 comes from a bedded red chert, of 2 m thickness, located c. 2 km north of the village of Seki (29~ 36~ Individual chert beds reach 50 cm in thickness. These cherts depositionally overlie pink fine-grained limestones with nodular chert. The limestones form kilometre-sized blocks and are completely surrounded by m61ange. Sample A8 yielded a poorly preserved fauna in which Transhsuum sp.cf. T. brevicostatum (Ozvoldova) and Triactoma sp.cf. T. jonesi (Pessagno) were identified, indicating a probable Mid- to Late Jurassic age (Baumgartner et al. 1995). Sample ALT 5 comes from 3 km SE of Altinyayla, along a road leading from Altinyayla to Seki (29~ 36~ The sample is from a metre-scale, tightly folded, red-brown block of radiolarite, which occurs within the Ophiolitic M61ange, together with blocks of serpentinite, basalt, gabbro and red cherty limestone. This m61ange is structurally overlain by a thick (hundreds of metres) thrust slice of serpentinized peridotite. A reasonably well-preserved radiolarian fauna was identified in Sample ALT 5, including Eucyrtidiellum sp., Laxtorum (?) hichisoense Isozaki & Matsuda, Palinandromeda sognoensis Baumgartner, Parahsuum sp., Paronaella sp., Stichocapsa convexa Yao and Zartus praejonesi Pessagno & Blome. The last species is known

only from Bajocian strata (Pessagno & Blome 1980). Sample TO1-324 was collected from radiolarite thrust sheets with lava intercalations just outside the town of Marmaris, along the road to Datqa (28~ 36~ Although the relationship between the radiolarites and lavas appears to be stratigraphic, a tectonic contact cannot be excluded. The lavas predominate towards the top of the hill overlooking Marmaris, where they are more massive and contain tectonic inclusions of marble and phyllite. The radiolarite sample TO1-324 yielded a moderately well-preserved radiolarian fauna including Archaeodictyomitra sp.cf. A. apiarium (Riist), Emiluvia sp.cf. E. orea Baumgartner s.l., Mirifusus dianae ssp.cf. M. d. dianae (Karrer), Obesacapsula sp.cf. O. morroensis Pessagno, Podocapsa amphitreptera Foreman, Protunuma sp.cf. P. japonicus Matsuoka & Yao, Tethysetta dhimenaensis (Baumgartner) s.l. and Tritrabs (?) sp. The sample is Late Jurassic (Mid- to Late Oxfordian to Late KimmeridgianEarly Tithonian) based on the co-occurrence of species P. amphitreptera and T. dhimenaensis (UAZ 9-11; Baumgartner et al. 1995).

Regional significance of the Lycian radiolarites The radiolarites within the Lycian Ophiolitic M61ange are associated with other fine-grained sediments (e.g. pelagic carbonates) that also appear to have accumulated in an oceanic setting, far from a source of coarse-grained sediment (e.g. sandstone turbidites, neritic calciturbidites). In

MESOZOIC RADIOLARITES, SW TURKEY

233

Maa n,

Izmir-Ankara M61ange

Lycian Nappes

Cam San

Da~kiJplfi

Tur O 100- UJ O < I-UJ

AIb

Lycian M~lange

O

Bar

3s5b

Hau Val

140-

Tokat

TO1-324

Ber

A8

~ Tit ~ . Kim

t50160-

~

c~ 170< 180-

Oxf

u-l, Cal o Bat _ Bai ~

Aal

n"

Toa

--3

ALT 5

F:==~

"

~ O

190-

Pli Sin Het

200-

Rha

C9 210- ~ ~ 220

Ankara

Cen

~

~

Nor -

Turun(;

.

Car 230 -

r ~ l Radiolarite

l--5-] Limestone

~

Lava

Fig. 4. Radiolarian ages determined from the Lycian Ophiolitic M61ange (this study) and the Turun9 succession in the Marmaris area (Tekin & G6ncfio~lu 2002), together with published ages from various other m~lange outcrops within the Izmir-Ankara suture zone (Bragin & Tekin 1996; Bozkurt et al. 1997; G6ncfio~lu et al. 2000, 2001; Tekin et al. 2002). Time scale after Gradstein et al. (2004).

contrast, more proximal successions, dominated by redeposited limestones, are associated with the structurally underlying Lycian Thrust Sheets, especially the K6yce(giz Thrust Sheet (Collins & Robertson 1999). The radiolarites are also associated with blocks of various ophiolitic rocks (e.g. serpentinite, gabbro, diabase, tholeiitic basalt). Geochemical study of the ophiolitic basalts is suggestive of eruption in a mid-ocean-ridge (MOR)-type setting (Collins & Robertson 1998). The radiolarites therefore provide clues to the age of the Neotethyan oceanic crust. Figure 4 synthesizes the ages determined during this study: i.e. from early Mid-Jurassic (ALT 5) to Early Cretaceous (3s5b).

The Lycian Ophiolitic M6lange, together with most of the Upper Palaeozoic-Mesozoic units of the Lycian Nappes, is widely considered to have been thrust southwards from the Izmir-Ankara suture zone (north of the Menderes Massif; Fig. 1; Robertson & Pickett 2000; Okay e t al. 2001; Collins & Robertson 2003; Robertson e t al. 2004), although emplacement from the south has also been suggested (Poisson 1984). The para-authochthonous Tavas-Boz Da~l Unit is the exception and is thought to represent a basin between the Menderes Massif and the Bey Da~lan (Poisson 1984; Poisson & Sarp 1985; Collins & Robertson 1999, 2003). Radiolarites have also been dated by other workers

234

T. DANELIAN E T AL.

from several units preserved within the Lycian Nappes and the Izmir-Ankara suture zone, as summarized below. Within the Lycian Nappes, a Mesozoic deepseawater succession exposed in the Marmaris area, near Turun~ (Fig. 2; Ersoy 1993, 1997), begins with basalts and volcaniclastic turbidites. Radiolaria reported from limestones intercalated between MOR-type (or transitional-type) basalts (Collins & Robertson 1998, 2003) have yielded a microfauna of Mid-Carnian (Late Triassic) age (Tekin & G6ncfio~lu 2002). The succession, as described by Ersoy (1993, 1997) and Collins & Robertson (1999), is overlain by Jurassic to Upper Cretaceous deep-water redeposited sedimentary rocks. Radiolarites are associated with basic lavas within the Da~ktiplfi M61ange, north of Eski~ehir within the Izmir-Ankara suture zone (locality A in Fig. 1); these were recently reported to be Late Carnian in age (Tekin et al. 2002). The m61ange crops out in the Central Sakarya Zone of the Northern Neotethyan suture, where it occurs beneath a regionally extensive ophiolite (Ta~tepe Ophiolite). The Da[gkiiplfi M61ange also includes several slices of mid-ocean ridge basalt (MORB)-type metabasalts that alternate with radiolarites dated as Mid-Late Jurassic (Bathonian-Tithonian), Early Cretaceous (Hauterivian-Aptian) and also as midCretaceous (Cenomanian) (G6ncfiofglu et al. 2000, 2001). Further east, within an exposure of the Ankara M61ange on the outskirts of Ankara city (locality B in Fig. 1), blocks of radiolarite were reported to occur in a volcaniclastic matrix together with blocks of gabbro, lava and pink micritic limestone. The interbedded radiolarites were dated as Late Triassic (Late Norian), Early Jurassic, Late Jurassic and mid-Cretaceous (Albian-Cenomanian; Bragin & Tekin 1996). A long history of radiolarite accumulation is thus inferred within a single short succession. Further north, a block of radiolarite associated with ophiolitic lithologies (e.g. serpentinite, gabbro, basalt) occurs within the Tokat Complex of the Central Pontides (locality C in Fig. 1) ; this was reported to be of Late Jurassic (Tithonian) age (Bozkurt et al. 1997). In addition, basic-intermediate composition lavas, green volcaniclastic siltstone and pelagic carbonates are interbedded with greenish siliceous tufts and tuffaceous radiolarian cherts, dated as Carnian in age (Tekin 1999), from within the Bey~ehir-Hoyran Nappes in the Bozkir area, central Taurides. This succession was attributed to rifting to form the Inner Tauride oceanic basin to the south of the main, northerly Neotethyan ocean and is not

considered further here (Andrew & Robertson 2002).

Discussion and conclusions: implication for Tethyan evolution The radiolarian data shed light on the evolution of the Northern Neotethys ocean, especially if it is accepted that the Lycian M61ange formed within this ocean basin, together with the other radiolarites preserved within the Izmir-Ankara suture zone. In some cases the radiolarites are preserved only as detached blocks, commonly associated with ophiolitic rocks, thus limiting their interpretative potential. However, at some localities the dated radiolarites are interbedded with basalts for which the tectonic affinity is known based on geochemical analysis (e.g. MORB); these radiolarian ages are particularly useful for determining the timing and mode of ocean crust formation. Accordingly, radiolarian ooze began to accumulate on the Neotethyan oceanic floor during the Late Triassic (Mid-Carnian) within units now forming parts of the Lycian Nappes (Tekin & G6nciio~lu 2002) and also during the Late Norian within the Ankara-Izmir suture (Bragin & Tekin 1996). Radiolarites of the former area are depositionally associated with MORB-type extrusive rocks, suggesting that the oldest oceanic lithosphere is Late Triassic (Collins & Robertson 1998, 1999). This is important as it supports initial genesis of oceanic crust within the Northern Neotethys by Late Triassic time (Robertson et al. 1996, 2004; Robertson & Pickett 2000; Tekin & G6nciiofglu 2002). This contrasts with an earlier view that Northern Neotethyan oceanic crust did not form until the Early Jurassic ($eng6r & Yllmaz 1981; G6riir et al. 1984). The youngest confirmed radiolarites are of Mid-Cretaceous (Cenomanian) age associated with m61ange units of the Izmir-Ankara suture zone (Bragin & Tekin 1996; G6nciio~lu et al. 2001). Taken together, these radiolarites appear to have accumulated within the Northern Neotethys during Late Triassic-Early Cretaceous time (c. 220-93.5 Ma). The radiometric ages of the metamorphic soles of the Lycian Peridotite, mainly dated at 95-90 Ma (Thuizat et al. 1981; Dilek et al. 1999), document the time of first tectonic disruption of young, still hot, oceanic lithosphere within the Northern Neotethys. This oceanic crust and mantle together with the distal deep-sea sediments of the Lycian M61ange, including the radiolarites dated here, were then accreted within a subduction complex during the Late Cretaceous (Turonian-Campanian; c. 90-75 Ma) (Collins & Robertson 1997; Okay et al. 2001).

MESOZOIC RADIOLARITES, SW TURKEY Field and laboratory work was partly funded by the UK Natural Environmental Research Council and partly by CNRS-FR32 (Cepage). We thank J. Aitchison and K. Tekin for helpful reviews of the original manuscript. C. Abrial and A. Lethiers helped with the drawings.

References AL-RIYAMI, K., DANELIAN,T. & ROBERTSON,A. H. F. 2001. Radiolarian biochronology of Mesozoic deep-water successions in NW Syria and Cyprus: implications for south-Tethyan evolution. Terra Nova, 14, 271-280. ANDREW, T. & ROBERTSON, A. H. F. 2002. The Bey~ehir-Hoyran-Hadim (B-H-H) nappes: genesis and emplacement of Mesozoic marginal and oceanic units of the Northerly Neotethys in Southern Turkey. Journal of the Geological Society, London, 159, 529-543. BAUMGARTNER, P. O., BARTOLINI,A., CARTER, E. S., et al. 1995. Middle Jurassic to Early Cretaceous radiolarian biochronology of Tethys based on Unitary Associations. In: BAUMGARTNER, P. O., O'DOGHERTY, L., GORICAN, S., URQUHART, E., PILLEVUIT, A. & DE WEVER, P. (eds) Middle

Jurassic to Lower Cretaceous Radiolaria of Tethys: Occurrences, Systematics, Biochronology. M6moires de G6ologie (Lausanne), 23, 1013-1048. BECCALETTO, L., BARTOLINI, A.-C., MARTINI, R., HOCHULI, P. A. & KOZUR, H. 2005. Biostratigraphic data from the Cetmi M61ange, northwest Turkey: palaeogeographic and tectonic implications. Palaeogeography, Palaeoclimatology, Palaeoecology, 221,215-224. BERNOULLI,D., DE GRACIANSKY,P. C. & MONOD, O. 1974. The extension of the Lycian nappes (SW Turkey) into the Southeastern Aegean islands. Eclogae Geologieae Helvetiae, 67, 4-90. BOZKURT, E., HOLDSWORTH, B. K. & KOCYIGIT, A. 1997. Implication of Jurassic chert identified in the Tokat Complex, northern Turkey. Geological Magafflne, 134, 91-97. BRAGIN, N. Y. & TEKIN, U. K. 1996. Age of radiolarian-chert blocks from the Senonian Ophiolitic M61ange (Ankara, Turkey). The Island Arc, 5, 114-122. ~ELIK, 6. F. & DELALOYE,M. 2003. Origin of metamorphic soles and their post-kinematic mafic dyke swarms in the Antalya and Lycian ophiolites. SW Turkey. Geological Journal, 38, 235-256. COLLINS, A. S. & ROBERTSON, A. H. F. 1997. Lycian m61ange, southwestern Turkey: an emplaced Late Cretaceous accretionary complex. Geology, 25, 255-258. COLLINS,A. S. & ROBERTSON,m. H. F. 1998. Processes of Late Cretaceous to Late Miocene episodic thrust-sheet translation in the Lycian Taurides, SW Turkey. Journal of the Geological Society, London, 155, 757-772. COLLINS, A. S. & ROBERTSON, A. H. F. 1999. Evolution of the Lycian Allochthon, western Turkey as a north-facing Late Palaeozoic to Mesozoic rift and passive continental margin. Geological Journal, 34, 107-138.

235

COLLINS, A. S. & ROBERTSON,A. H. F. 2003. Kinematic evidence for Late Mesozoic-Miocene emplacement of the Lycian Allochthon over the Western Anatolide Belt. Geological Journal, 38, 295-310. DANELIAN, T. & ROBERTSON, A. H. F. 1997. Radiolarian evidence for the stratigraphy and palaeooceanography of the deep-water northern passive margin of the Indian plate (Karamba Formation, Indus Suture Zone, Ladakh Himalaya). Marine Micropaleontology, 30, 171-195. DANELIAN, T. & ROBERTSON, A. H. F. 2001. Neotethyan evolution of eastern Greece (Pagondas M61ange, Evia island) inferred from radiolarian biostratigraphy and the geochemistry of associated extrusive rocks. Geological Magazine, 138, 245363. DE GRACIANSKY, P. 1972. Recherches gOologiques dans le Taurus lycien. Th6se Doctorat d'Etat, Universit6 Paris-Sud (Orsay). DE WEVER, P. 1989. Radiolarians, radiolarites and Mesozoic palaeogeography of the CircumMediterranean Alpine belts. In. HEIN, J. R. & OBRADOVI6,J. (eds) Siliceous Deposits of the Tethys and Pacific Regions. Springer, New York, 31-49. DE WEVER, P., AZI~MA, J. & FOURCADE, E. 1994. Radiolarites et radiolarites: production primaire, diagen6se et pal6og6ographie. Bulletin des Centres

de Recherches Exploration-Production Aquitaine, 18, 315-379.

Elf-

DILEK, Y., THY, P., HACKER,B. & GRUNDVIG,S. 1999. Structure and petrology of Tauride ophiolites and mafic dyke intrusions (Turkey): implications for the Neotethyan ocean. Geological Society of America Bulletin, 111, 1192-1216. ERSOY, S. 1993. The geological setting of the tectonic units situated in SW Anatolia (Turkey). Bulletin of the Geological Society of Greece, 28, 617-628. ERSOY, S. 1997. The syn-collisional deep-water sediments of the Marmaris Complex as a part of the Lycian Nappes, SW Turkey. International Earth

Sciences Colloquium of the Aegean Region, 9-14 October 1995, Izmir-Giilliik, Turkey, Proceedings Volume, 1, 95-111. GONCOO~LU, M. C., TURHAN, N., ~ENTORK, K., OZCAN, A., UYSAL, S. & YALINIZ, K. 2000. A geotraverse across northwestern Turkey: tectonic units of the Central Sakarya region and their tectonic evolution. 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, 139-161. GONCOOGLU, M. C., YALINIZ,M. K., TEKIN, U. K. & TURHAN, N. 2001. Petrology of Late BerriasianLate Hauterivian and Cenomanian oceanic basalts within the Central Sakarya ophiolitic complex, NW Turkey: constraints for the evolution of the Izmir-Ankara oceanic branch of Neotethys. Fourth

International Turkish Geology Symposium (ITGS IV), Cfikfirova University Adana (Turkey), 93. GORUR, N., OKTAY, F. Y., SEYMEN,I. & ,SENGOR,A. M. C. 1984. Paleotectonic evolution of Tuz G61fi Basin complex, central Turkey. In: DIXON, J. E. & ROSERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 81-96.

236

T. DANELIAN ET AL.

GRADSTEIN, F. M., OGG, J. G., SMITH,A. G., BLEEKER, W. &; LOURENS, L. J. 2004. A new Geologic Time Scale, with special reference to Precambrian and Neogene. Episodes, 27/2, 83-100. OKAY, A. I., TANSEL, I. & TI~YSOZ, O. 2001. Obduction, subduction and collision as reflected in the Upper Cretaceous-Lower Eocene sedimentary record of western Turkey. Geological Magazine, 138, 117-142. PESSAGNO, E. A., JR. & BLOME, C. D. 1980. Upper Triassic and Jurassic Pantanelliinae from California, Oregon and British Columbia. Micropaleontology, 26, 225-273. POISSON, A. 1977. Reeherches gkologiques duns les Taurides occidentales ( Turquie). Thesis Doctorat d'Etat, Universit6 Paris-Sud, Orsay. POISSON, A. 1984. The extension of the Ionian trough into southwestern Turkey. In: DIXON, J. E. & ROBERTSON, A. H. F (eds) The GeologicalEvolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 235-241. POISSON, A. & SARP, H. 1985. La zone de K1zdqat~orag61. Un example de sillon intra plate-forme ~t la marge externe du Massif du Menderes. In: IZDAR, E. & NAKOMAN,E. (eds) Proceedings of the VI Colloquium on the Geology of the Aegean Region, 19 September - - 4 October 1977, Izmir, Turkey. Dokuz Eyliil University, Bornova, Izmir, 555-564. 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. 2004. Development of concepts concerning the genesis and emplacement of Tethyan ophiolites in the Eastern Mediterranean and Oman regions. Earth-Science Reviews, 66, 331-387. ROBERTSON, A. H. F. & PICKETT, E. A. 2000. Palaeozoic-Early Tertiary Tethyan evolution of m61anges, rift and passive margin units in the Karaburun Peninsula (western Turkey) and Chios Island (Greece). In: BOZKURT, E., WINCHESTER, J. A. & PIPER, J. D. (eds) Geological Evolution of Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 2542. 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 Palaeozoic-Early Cenozoic development of Tethys in the Eastern Mediterranean region, In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBERTSON, m. H. F., USTAOMER, T., PICKETT, E. m., COLLINS, A. S., ANDREW, Z. • DIXON, J. E. 2004. Testing models of Late Palaeozoic-Early Mesozoic orogeny in Western Turkey: support for an openTethys model. Journal of the Geological Society, London, 161, 201-511. ~ENEL, Y. M. 1981. Palaeocene-Eocene sediments interbedded with volcanics within the Lycian Nappes: Faralya Formation. General Directorate of Mineral Research and Exploration (MTA) Bulletin, 113, 1-14. ~ENGOR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 81-241. TEKIN, U. K. 1999. Biostratigraphy and Systematics of Late Middle to Late Triassic Radiolarians from the Taurus Mountains and Ankara Region, Turkey. Geologisch-Pal/iontologische Mitteilungen Innsbruck, Sonderband, 5. TEKIN, U. K. & G6NCUOCJLU, M. C. 2002. Middle Carnian radiolarians from the intra-pillow limestones of the Turunq Unit within the Gfilmahar Nappe (Lycian Nappes, Marmaris, southern Turkey): geodynamic implications. First International Symposium of the Faculty of Mines (ITU) on Earth Science and Engineering, 16-18 May 2002, Technical University, Istanbu184. TEKIN, U. K., GONC00(~LU, M. C. & TURHAN, N. 2002. First evidence of Late Carnian radiolarians from the Izmir Ankara suture complex, central Sakarya, Turkey: implications for the opening age of the Izmir-Ankara branch of Neo-Tethys. Geobios, 35, 127-135. THU1ZAT, R., WHITECHURCH, n., MONTIGNY, R. & JUTEAU, T. 1981. K-Ar dating of some intraophiolite metamorphic soles from the East Mediterranean: new evidence for oceanic thrusting before obduction. Earth and Planetary Science Letters, 52, 302-310.

Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab (northern Greece) ANNE

H. E. R A S S I O S 1 & E L D R I D G E

M. M O O R E S 2

~Institute of Geology and Mineral Exploration, Lefkovrisi, 50100 Greece (e-mail." [email protected]) 2Department of Geology, University of California, One Shields Avenue, Davis, CA 95616 USA The Eocene-Miocene Mesohellenic Trough is an elongate sediment-filled tectonic basin trending NW across central Greece and into Albania. Neotethyan oceanic rocks, including Triassic-Jurassic rift-related volcanic rocks and deep-sea sediments, accretionary mrlange and ophiolitic complexes, crop out along its margins. These units were tectonically emplaced onto the Pelagonian microcontinent to the east and the Apulian-African continental margin to the west. In northern Greece, the mid-Jurassic Vourinos ophiolite on the eastern margin of the trough is geographically separated from the synchronous Pindos ophiolite along the western margin by a minimum c. 20 km distance. The sedimentary fill of the trough obscures their presumed subsurface continuation, although magnetic surveys identify thick 'ophiolitic' rocks beneath the basin. We interpret these ophiolites as parts of the same oceanic slab, two parts of a single larger oceanic complex we now term the Mesohellenic ophiolite. Comparable ophiolitic complexes to the south (the Koziakas and Othris) and the ophiolites of the Mirdita complex to the north in Albania are considered as members of this same complex. Geological and petrological data from the Vourinos and Pindos ophiolites define intra-slab heterogeneity. Vourinos essentially is a 'Penrose-style' ophiolite with 'supra-subduction' compositions; the less continuous Pindos ophiolite shows coexisting midocean ridge basalt and island arc characteristics. Ophiolitic rocks that seem to represent geographical overlap between these characteristic associations crop out along their northern (Dotsikos strip) and southern (Mesovouni) margins. Variations in mantle strain conditions across the ophiolitic slab have been mapped, and demonstrate a single orientation of deformation; this is explained by variable strain kinematics that persisted across the ductile-brittle boundary. A continuity from ductile to brittle emplacement structures spans the Mesohellenic Trough, independent of petrological association, and indicates the original relative positions of these ophiolites within the oceanic slab. These structures illustrate tectonic 'steps' of obduction from the ridge crest onto the Pelagonian margin to the east, and can be relatively timed by the overlap of magmatism with ductile deformation in different parts of the slab. Hence, rotations of original horizontality are dated to the period preceding cessation of ductile field deformation, while still in the oceanic environment. The morphology defined by these structures and the horizontal rotation of stratigraphy are analogous to a spoon- or scoop-shaped nappe originating in the ductile field at its base, and crossing into the brittle field rapidly at its leading edge (Vourinos), whereas the mylonitic deformation characterizes the 'trailing' end (Pindos). Age relations require that geochemical variation between the two complexes must be explained within a model of synchronous generation, possibly with apparent 'supra-subduction zone' rifting of originally heterogeneous mantle, or an overlapping series of diverse processes of magma generation in an initially homogeneous mantle. Indications of the original ridge crest directions suggest the operation of several simultaneous spreading centres, separated by transform faults or 'pseudofaults'. A palinspastic reconstruction of the slab constrains applicable oceanic models and provides the basis of future research.

Previous reconstructions of Neotethys in the region of the Mesohellenic T r o u g h o f Greece and its extension into Albania, including those o f the present study area o f the Vourinos a n d the Pindos M o u n t a i n s (Fig. 1), have centred a r o u n d establishing a stratotectonic framework o f the

region ( R o b e r t s o n 2002; Dilek et al. 2005). The significance o f paired ophiolitic belts, Jurassic remnants of Neotethys, figures p r o m i n e n t l y in these reconstructions a n d aids definition of a semi-independent seaway termed the Pindos Ocean. These ophiolites crop out along b o t h

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 237-266. 0305-8719/06l$15.00 9 The Geological Society of London 2006.

238

A.H.E. RASSIOS & E. M. MOORES

Fig. 1. Location of ophiolites proximal to the Mesohellenic Trough, Greece. Specific ophiolites mentioned in text: M, Mirdita; P, Pindos; V, Vourinos; K, Koziakas; O, Othris; Vr, Vrinena. margins of the Mesohellenic Trough, an Eoceneto Miocene-aged sedimentary-filled tectonic basin that straddles and obscures the precise suture of the Palaeozoic Pelagonian-Korabi microcontinent with Apulia. The eastern belt of Mesohellenic ophiolites is essentially considered

as a continuation of the supra-subduction zone (SSZ) Vourinos ophiolite, and the western belt as a continuation of the mid-oceanic ridge basalt ( M O R B ) + S S Z affinity Pindos. The Mirdita ophiolite of Albania, located at the northern termination of the Mesohellenic Trough, apparently

PINDOS-VOURINOS OPHIOLITIC SLAB transcends these zones with a west-east transition from a Pindos-style to Vourinos-style section (Robertson & Shallo 2000; Dilek et al. 2005). To the south of our study area in Greece, this pairing is less evident within the MORBdominant Othris ophiolite, although the ophiolitic section of Vrinena (Rassios 1990) in easternmost Othris resembles a highly sheared but petrologically similar section to Vourinos. At present, the petrogenetic differences between the eastern and western belts used to define the character of Neotethys relate primarily to their geochemical affinity. Vourinos is considered one of the 'type' SSZ ophiolites dominated by island-arc tholeiite (IAT) extrusive rocks in global-scale studies (Beccaluva et al. 1984; Pearce et al. 1984). The Pindos or western belt was initially interpreted as an 'SSZ' ophiolite because of the presence of boninites of unknown quantity (Caperdi et al. 1980; Dupuy et al. 1984), or as essentially an 'SSZ' ophiolite despite the presence of MORB lavas and lherzolite. Others (Jones & Robertson 1991; Jones et al. 1991; Robertson et al. 1991; Rassios & Smith 2000; Smith & Rassios 2003) interpreted the Pindos as a 'transitional' step between MORB and true SSZtype ridge activity. In short, the two ophiolites were considered as representing the evolution of supra-subduction oceanic lithosphere on a mature MORB substrata. Previously, Rassios & Smith (2000) concluded that genetic differences between the Mesohellenic ophiolites represent primary variations within a single oceanic (back-arc) slab. Primary differences between the mantle and cumulate sections of the two belts in Greece have not yet been applied as a means of constraining a palinspastic reconstruction of the Pindos ocean (as in Dilek & Moores 1990), nor have attempts been made to interpret the possible continuity of oceanic structures between ophiolites across the Mesohellenic Trough. In the present paper, we attempt to reconcile geomagnetic (Memou & Skianis 1993) as well as other evidence for the continuity of the Pindos-Vourinos ophiolite and new age data (Liati et al. 2004) with the geographical distribution of magmatic and structural heterogeneous features. This synthesis reveals a view of the evolutionary processes within the Neotethyan lithosphere and a 'step-by-step' progression from magmatism through emplacement. We note that the Vourinos and Pindos ophiolites are only two 'data' points available for describing the Pindos lithosphere and its riftdrift-emplacement legacy. Future syntheses including all of the 'Mesohellenic ophiolites', those presumed to be part of a single slab beneath the Mesohellenic-Mirdita trough extending from

239

Othris in Greece to the Mirdita complex of Albania, may document the evolution between constructional and emplacement processes of Neotethys.

Regional and ophiolitic background The Vourinos ophiolite complex (Fig. 2) was investigated by Moores (1969) during an era in which the basic tenets of plate tectonic theory and the importance of ophiolitic assemblages as ocean-floor analogues were first elucidated. Indeed, the 'idealized' consensus ophiolitic section (Anonymous 1972) is highly biased towards a 'Vourinos-style' pseudostratigraphy (Fig. 3). Key attributes of this complex and studies contributing to understanding its regional setting are as follows. (1) The Vourinos complex consists of a 12 km thick section with intact pseudostratigraphy from peridotite mantle rocks across the 'petrological Moho' and basal crustal cumulate section, into transitions of ultramafic to mafic magma chambers, dioritic cumulates and plagiogranites (in the Krapa and Vatolakkos areas), and an intact transition to a sheeted dyke swarm (Moores 1969; Jackson et al. 1975; Rassios 1981; Rassios et al. 1983b). Of the exposed area of the ophiolite 80% is depleted mantle lithosphere (harzburgite restite) with inclusions of dunite in deformed bodies that extend from several metres to several kilometres in size; some of these dunite bodies contain chromite concentrations. All chromite concentrations of economic scale, mined from the 1950s until 1992, crop out within the mantle suite in a belt termed the 'metalliferous zone' (Grivas et al. 1993). Many of the tectonic data we report in this paper were collected between 1984 and 1994 during research conducted at Vourinos that consisted of devising and testing a new strategy for chromite exploration (Rassios et al. 1986, 1991; Rassios 1994). This required structural mapping at 1:5000 to 1:500 scale and whole-rock analyses of representative suites, geochemical targeting of chrome deposits (Rassios & Kostopoulos 1990), and exploratory drilling of 11 ore districts. (2) Zimmerman's (1968, 1972) detailed description of the tectonic contact of the Vourinos ophiolite over Pelagonian basement rocks established the concept of metamorphism accompanying ophiolitic emplacement. The highly sheared sedimentary wedge he described between the Vourinos basal harzburgite

240

A.H.E. RASSIOS & E. M. MOORES

Fig. 2. Exposures and localities of the Pindos and Vourinos ophiolites and magnetically determined subsurface presence of ophiolitic and Pelagonian lithologies. Additional localities referred to in text: M, Monahiti; B, Vasilitsa; F, Flinga; S, Salatoura; A, Aspropotamos; L, Liagouna; K, Kisavos; V, Voidolakkos; PM, outcrop of petrological Moho in central area of Vourinos.

(3)

and Pelagonian carbonate rocks has since been reinterpreted as an accretionary m61ange essentially equivalent to that of the Avdella m61ange that tectonically underlies the Pindos ophiolite (Jones & Robertson 1991). At Vourinos, this m61ange has a metamorphic imprint varying from greenschist to low-amphibolite facies, ranges in thickness from metre scale to 500 m, and includes blocks of serpentinite, MORB lavas, tufts, and deep-sea Jurassic carbonates of centimetres to tens of metres in size in a matrix of phyllitic siltstones and tuffaceous shales. The largest m61ange block is a tectonically incorporated, pervasively mylonitic limestone of about 10 km length (the 'loferitic' limestone unit of Naylor & Harle 1976). The ophiolitic sole is delineated by discontinuous occurrences of amphibolite and garnet amphibolite of 0.2-2 m thickness. The Pelagonian rocks underlying the emplacement zone consist of passive margin carbonates (of early Mid-Triassic to Late

(4)

(5)

Jurassic age, Brunn 1956) that are crossed by mylonite zones within several hundreds of metres from, and parallel to, the ophiolitic base. Along the southern contact of Vourinos a metasedimentary rifting section of early-mid Triassic age (Mountrakis 1985) is exposed that includes conglomerates, metaclastics, tufts and deep-water carbonates, and radiolarian cherts with acid and basic intrusions. The upper portions of the Vourinos (Fig. 4) consist of dykes and flows beneath a wedge of Jurassic pelagic limestone (Moores 1969; Mavrides et al. 1979); these are eroded beneath and overlain by an angular unconformity at the base of a thick section of Upper Cretaceous (Cenomanian) limestone (Brunn 1956). The implications of the angular rotations denoted by this contact are examined later in this paper. Late and post-Alpine deformation of Vourinos consists of steep reverse faults at the eastward margins of the Krapa (Fig. 4) and Kissavos areas that probably relate to a

PINDOS-VOURINOS OPHIOLITIC SLAB

241

Fig. 3. Generalized 'Penrose' ophiolitic pseudostratigraphy (Anonymous 1972) compared with that of the Vourinos and Pindos complexes. IAT, island-arc tholeiite.

combination of mid-Tertiary Apulian collision (Doutsos et al. 1993), strike-slip faults (left-lateral) that cross the section with a major dislocation between the northern and southern blocks, and neotectonic normal faults that in some cases developed parasitically on older emplacement and strike-slip faults of the region (Pavlides & Caputo 1997). During the 1980s, Vourinos became accepted as one of the 'type' SSZ ophiolites based on developing models of geochemical affinity (Rassios et al. 1983a; Pearce et al. 1984; Beccaluva et al. 1984). The Pindos ophiolite (Fig. 2), unmapped since its original description by Brunn (1956), remained relatively inaccessible, and was correlated with the 'SSZ' model of Vourinos following several reconnaissance lava sampling projects (Capedri et al. 1980; Dupuy et al. 1984); none of these projects encompassed the entire immense

ophiolitic section of the > 2000 km 2 Pindos complex. The major structural and tectonostratigraphic relations of the Pindos had not been delineated at that time. Subsequently, Rassios with other Greek researchers (see end of article) systematically evaluated the chrome and copper ore potential of the Pindos ophiolitic rocks, a study that required the discrimination of mantle from crustal units and structures, imprinted tectonic structures, and petrological-geochemical analyses of peridotites. This study, influenced by the parallel studies of Jones, Robertson and Cann (Jones 1990; Jones & Robertson 1991; Jones et al. 1991) that delineated the tectonostratigraphy of the Pindos, and that of Kostopoulos (1989), which described the lava heterogeneity, resulted eventually in a new map that documents the distribution of ophiolitic rocks in the Pindos within the context of a plate tectonic-oceanic analogue framework (Rassios & Grivas 1999). Key features and studies

242

A.H.E. RASSIOS & E. M. MOORES contributing to this view of the Pindos are as follows. (1) The Pindos ophiolite is far more tectonically disrupted than Vourinos (Figs. 3 and 5a and b). As for the Othris ophiolite to the south (Smith 1993; Smith et al. 1975, 1979), the ophiolite is distributed among stacked nappe-imbricates; unlike Othris, these tectonic units include a great deal of petrological-pseudostratigraphic overlap, so that Rassios and Grivas (Rassios 1991; Rassios et al. 1994; Rassios & Grivas 1999, 2001) inferred minimal tectonic transport during dismemberment, and their map interprets the eastern margin of the complex as a ramped section verging to the NE. The largest and structurally highest unit is the Dramala complex, a nappe that extends from its type description over the Dramala massif (Rassios 1991) across the neighbouring mountains of Mavrovouni, Salatoura, Avgo, Flinga, Liagouna, Vassilitsa and Smolikas (Fig. 2). The Dramala complex contains the mantle suite of the Pindos ophiolite, and this includes harzburgite, harzburgite with a low content of clinopyroxene, lherzolite, and at least one occurrence of plagioclase lherzolite; thus, it resembles more closely the mantle section of a MORBtype ophiolite. A 100-300 m broad nonserpentinized zone occurs crossing from the Dramala massif southward to Mavrovouni (Rassios & Grivas 2001), but the remaining peridotites are partially to completely serpentinized. Dunite is included in the mantle peridotite as small lenses, rarely more than a few tens of metres in scale. The Dramala massif includes a preserved transition from mantle to crustal sequence rocks, a 'petrological Moho' consisting of a 0.5-1.5 km transitional section with increasing proportions up-section of magmatic cumulate rocks (dunite, plagioclase dunite and troctolite) in harzburgite. The petrological sequence of these cumulates also resembles more closely MORB crustal sections. However, coarse-grained pyroxene-bearing boninite intrusions occur within the lower crustal cumulates in the Dramala complex. (2) Jones and others (Jones et al. 1991; Robertson 2004) characterized the Aspropotamos unit of the Pindos ophiolite as a Fig. 4. Cross-section of the Krapa area of the Vourinos ophiolite. Transition from base of cumulates to sheeted dykes-pillow lavas demonstrates overturn of section towards the Mesohellenic Trough (west of section).

PINDOS-VOURINOS OPHIOLITIC SLAB

(3)

(4)

second complex, Aspropotamos, that consists of tectonically disrupted but overlapping units of lower cumulates, troctolite to diorite cumulates, plagiogranites, an in situ transition zone to a sheeted dyke complex, and a thick unit of pillow lavas that include several structural repetitions. The petrological overlap of the Dramala and Asprokambos crustal cumulate sections and their structural relations (Dramala overriding Asprokambos, Jones 1990; Rassios 1991; Rassios et al. 1994) imply these were once continuous. Nevertheless, the upper portions of the Aspropotamos section are overwhelmed by extrusive rocks that have a mainly SSZ geochemical affinity (Kostopoulos 1989; Jones 1990; Jones & Robertson 1991). The Pindos ophiolitic complexes of Dramala and Aspropotamos are underlain by an amphibolite sole, termed the Loumnitsa unit by Jones (1990). At Liagouna, the intact contact between serpentinized peridotite of the Dramala complex and its amphibolite sole rocks is exposed, and the sole formation grades downward stratigraphically with concurrent decrease of metamorphic grade for over 200 m into the underlying Avdella m61ange. At least one repetitive higher-grade unit is included in this transition. Metamorphic grade decreases from greenschist to zeolite facies within the underlying Avdella m61ange. The Avdella m61ange is a Jurassic accretionary wedge (Jones 1990; Jones & Robertson 1991) mapped over the extent of the Pindos ophiolite by Rassios & Grivas (1999). The m61ange is a 'block and matrix' complex in which blocks of Triassic to mid-Jurassic aged pelagic carbonates and cherts, lavas and assorted oceanic lithosphere fragments occur within a typically mudstone matrix; blocks range in size to kilometre-scale exposures of serpentinite and lavas that can be misinterpreted as parts of the ophiolitic section.

Rassios et al. (1994) demonstrated that the deformation that the ophiolite underwent while crossing the ductile-brittle boundary was NEverging (this evidence is reviewed below), and that the ophiolite and its m61ange base were originally emplaced to the NE. A later brittle field compressive 'back-thrust' verging to the SW placed the ophiolite-m61ange complex over the late Cretaceous-mid-Eocene Pindos flysch. In great part, the structural complexity of the Pindos is due to the relatively incompetent lithological nature of the Avdella m61ange. Compared with the ophiolitic and carbonate

243

formations of the Pindos, the m61ange unit preferentially accommodated both late emplacement (Jurassic-Cretaceous) imbrication, the younger (late Eocene) phase of back-thrusting, and the continuing deformation associated with the development of the Mesohellenic Trough. The most severe zones of tectonic complexity correlate with the 'corridors' (Brunn 1956) between the ophiolite-topped Pindos massifs in which the m61ange and tectonically incorporated Cretaceous limestones are severely deformed. Ductile structures in the peridotites along corridor margins (Rassios et al. 1994) suggest that these features initiated while still in oceanic conditions as 'tear' systems through the NE-emplacing slab; the corridors then went on to be reactivated as tear systems and imbricate fronts facilitating the mid-Eocene SW-verging thrust, and thence provided sites for extreme horizontal rotations of the ophiolitic and early Tertiary units during the formation of the Mesohellenic Trough in mid-Tertiary time. The 'type section' of sheeted dyke and pillow lava units of the Aspropotamos complex (Fig. 5a and b), which Rassios interpreted as the lowermost overridden member of the ophiolite, is severely imprinted by structures related to the formation and development of the Mesohellenic Trough. Horizontal rotation of the Tertiary sediments and lava-ophiolitic section by 70o-80 ~ accompanied an embayment of the Mesohellenic Trough into this area; dyke orientations from the Aspropotamos complex may not be representative of a 'relative' pseudostratigraphic position within the Pindos ophiolite. The entire study area extending to Albania has apparently been passively rotated about a vertical axis (see Robertson & Shallo 2000) in the mid-Tertiary from an originally more east-west trend.

Slab continuity Although subsurface continuity between the ophiolites of Vourinos and the Pindos has been postulated since initial studies of Greek ophiolites (summarized by Smith & Rassios 2003), a magnetic survey of the Mesohellenic area (Memou & Skianis 1993) verified the extensive presence of basement rock beneath the sedimentary deposits of the Mesohellenic Trough that display a 'serpentinized peridotite' signature. This confirms personal communications from several geologists that drilling programmes conducted in the 1960s in the Mesohellenic Trough for petroleum exploration recovered ophiolitic rocks at depth (Bornovas, pers. comm.; Mavrides, pers. comm.) although no cores or drilling records remain available.

244

A.H.E.

R A S S I O S & E. M. M O O R E S

o

~

o

o

~o~

~

0..1. ,..~

"t~ o

r

4.a o .0

0 '~

o,~.

o

o

~ ' - ~ . n ~ ~"n ~,, ~a. ~. = - , ~ - - ~

o

0a ~ ~ 8, ._ 0,.0

,~

;~,.cl

~

~ = ~ I=l

,...~. ~

~

Oa 1::I

0

;> 0

~.~=o .~

I

~

~

~

. ~

~x~

" ~ .nl

o~

~CI .0

~

-9 ~ *

O=-

~ ~ , . ~ . o.,..~ ,.~ ~ 0 ~

~o

"r~-~ ~l~O

..~ o _-

~ , .= o~O

-~

r

"~

~,~

"~o

PINDOS-VOURINOS OPHIOLITIC SLAB The magnetic survey (salient features summarized in Fig. 2) verifies field evidence that the Pindos is itself a thin thrust nappe underlain by magnetically 'neutral' material (presumably Pindos Flysch): no ophiolitic basement underlies the westward margin of the ophiolite belt. Magnetic signatures corresponding to the Pelagonian basement are evident beneath Vourinos and much of the region south and SW of Vourinos. The magnetic map is interpreted (Memou & Skianis 1993) to show that the most extensive ophiolitic rocks are located beneath the Mesohellenic Trough, and that Vourinos and the Pindos are merely thin edges preserved along the pre-Mesohellenic margins. Uncertainties within this magnetic-geological map are as follows. (1) The 'ophiolitic' magnetic signature is based on the presence of peridotite, which is volumetrically dominant in both the Vourinos and Pindos ophiolites as it is in oceanic lithosphere. Distinctions of other kinds of ophiolitic lithologies are not put forward in the present interpretation; thus, the subsurface expanse of ophiolitic rocks (lavas, gabbros, cherts, etc.) might be more extensive than the map indicates. (2) No distinction between the ophiolitic complex and the potential presence of large serpentinite blocks within the Avdella m61ange can be made. (3) Field checks of ophiolitic terrains correlates the position of magnetic lineaments to shear zones in outcrop, and these are assumed to represent the same in buried peridotite-rich areas.

245

on older Pindos oceanic crust (MORB to SSZ, essentially a fully developed continuation of the Aspropotamos complex), then it would be expected to be significantly younger than the Pindos. However, if the two ophiolites were emplaced together, the sole would be oldest where detachment originated, that is, towards the emplacement margin at Vourinos. Recently, Liati et al. (2004) published several zircon dates for Pindos and Vourinos plagiogranites, assigning ages of 171 _+3 Ma for the Pindos and 168.5_+2.4Ma and 172.9_+3.1 Ma for Vourinos. These crystallization ages show that the two ophiolites are essentially synchronous; oceanic spreading occurred in the midJurassic (Bajocian) and, consequently, sole formation followed immediately (Rassios & Smith 2000). If accepted, these zircon data preclude former evolutionary models and require us to consider how these two petrogenetically diverse ophiolites were produced simultaneously within Neotethys, then immediately emplaced as a contiguous oceanic slab. As previously shown through structural analyses (Rassios et al. 1994; Rassios & Smith 2000), there is overlap between magmatism and the emplacement period, and the effect of emplacement on the evolving lithospheric slab of Neotethys will be more closely examined herein.

Comparative slab processes

Although we cannot document tectonic disruptions to the slab beneath the Mesohellenic Trough, the fact that specific petrological and tectonic features can be recognized and traced between the ophiolites implies that their relative original positions have been preserved.

Recent evaluations (Moores et al. 2000) and descriptions of modern ridge crests (Robinson et al. 2000; Thy & Dilek 2000) spotlight the need for processes to explain observed ridge-crest complexities and the ophiolitic 'conundrum'. Given the constraint that the Vourinos and Pindos ophiolites formed concurrently and in apparent geographical proximity, the heterogeneous nature of the Mesohellenic slab requires that such processes were active in the Pindos ocean.

Age data

C o m p a r a t i v e m a n t l e section

Amphibolite sole dates show overlapping values that do not conclusively determine one complex to be older than the other, nor do they distinguish unique emplacement ages. Spray and others provided 4~ dates of the sole rocks of Vourinos as 171 _+ Ma and of the Pindos as between 173+3 and 1 7 2 + 3 M a , later corrected to 165_+3Ma (Roddick et al. 1979; Spray & Roddick 1980; Spray et al. 1984). K - A r dating (Thuziat et al. 1981) assigns an age of 167 Ma to the Vourinos sole, and 172 + 5 Ma to that of the Pindos. Were Vourinos, as an island arc, evolving

The mantle section of Vourinos is almost entirely harzburgite of nearly uniform composition: everywhere it contains 8-15% orthopyroxene and all geochemical studies show no more than minor compositional variation (Moores 1969; Rassios et al. 1994). Spinel chemistries in harzburgites and dunites, and the abundance of chromitites present, indicate derivation from the melting of a previously depleted mantle source (Pearce et al. 1984; Roberts 1988, 1992). The modal content of orthopyroxene in harzburgite along some contacts with dunite bodies (Grivas et al. 1993)

246

A.H.E. RASSIOS & E. M. MOORES

indicates volumetrically minor variation in melting levels. The mantle suite of the Dramala complex includes similar extremely depleted material but ranges to less-depleted and apparently nondepleted mantle rocks such as relatively fertile lherzolites and plagioclase lherzolite. The latter are more similar in chemistry to those of the Othris ophiolite (J. Pearce, in Rassios 1994). The scale of heterogeneity is difficult to generalize, as gradations in modal pyroxene are subtle. Most of Mavrovouni (Fig. 2), an area of 3 km 2, is lherzolite. Over the Dramala massif (Fig. 2), lherzolite occurs as bodies with outcrop areas of tens of metres. The Pindos mantle shows a mappable pseudostratigraphy (Fig. 5a) delineated by the presence of initially pyroxenite dykes, then pyroxenite and gabbroic dykes nearing the petrological Moho (Rassios 1991). Pyroxenite dykes are pervasive throughout the Vourinos mantle section, and gabbroic dykes occur only in several 'near-Moho' chrome districts. Dunite bodies in Vourinos vary in size from metre scale to the immense 3 km 3 Xerilivado dunite body, host of its largest chrome ore deposits (Rassios & Kostopoulos 1990). The morphology of these bodies is generally aligned with the regional mantle fabric and is infolded with harzburgite; this reflects the complex geometry generated during co-deformation of more competent peridotite with less competent dunite (Nicolas et al. 1980; Moat 1986; Roberts et al. 1988). Higher in the section, and lacking structural complications, an intrusive contact is apparent between harzburgite with basal cumulate dunite at the 'petrological Moho' (Harkins et al. 1980). High-temperature mantle deformation (diapiric fabric) terminates at this boundary. Lower-temperature ductile deformation crosses this boundary, as can be observed as a mineral fabric in crustal-section chromite deposits, and persists in the cumulate section as a strong, possibly synmagmatic, lineation (Rassios 1981; Rassios et al. 1983b). Mantle dunites of Vourinos belong to several generations. Rassios & Kostopoulos (1990) delineated the presence of several geochemically distinct dunites within a small area of the km2-scale dunite body at Aga Kouri, Vourinos. Within the same massive dunite body (Fig. 6a) at least two phases of dunite are mapped, the older with a strong overprint of ductile phase chrome spinel fabric, and the younger dunite, hosted by the older, lacking strong fabric and characterized by clots of euhedral, non-layered chrome spinel. The dunite bodies of the Dramala complex of the Pindos never exceed a scale of tens of

metres. The array of dunite generations in the Pindos lacks definitive geochemical distinction, but glaciated outcrops expose cross-cutting relations among generally smaller-scale dunite pods and lenses of metre to centimetre scale (Fig. 6b): older dunites are pervasively deformed by ductile deformation, and cross-cutting dunites are irregular in form and often include euhedral (undeformed) spinels. Chromitites are present in the Pindos mantle rocks, but nowhere in economically viable concentrations as present at Vourinos, although Rassios (1994) regarded this as possibly an artefact of pervasive ductile-stage tectonic thinning. Volatiles in the mantle section have been suggested as an explanation of a chromite-ore horizon in Vourinos (Rassios & Kostopoulos 1990). Similarly, the presence of a volatile 'front' parallel to mantle stratigraphy can be deduced in the Pindos based on the presence of euhedral chrome spinels growing into the fabric of highly deformed peridotite, the occurrence of sulphides in these same zones, and the common presence of amphiboles within zones of excessive deformation (mylonite and cataclastic zones). Amphiboles (clear, blocky-subhedral grains of about 0.1-1 mm size, not observed to make up more than 4% modal), have also been identified via petrographic and phase analyses in the mylonite zones of Vourinos (at the Voidolakkos locality, and within a deeper cataclastite zone parallel to the sole). Rassios (1994) and Rassios et al. (1994) pointed out that the addition of any volatile to an otherwise homogeneous peridotite undergoing ductile deformation reduces the 'plasticity' of the peridotite locally, and provides a mechanism for initiating ductile-brittle deformation. Pyroxenite pegmatites in the mantle section also suggest the presence of volatiles. The grain sizes of individual pyroxenes range from centimetre to metre scale. These sizes and subsequent amphibole alteration of the pyroxene preclude statistical petrographic analyses, but many dykes prove to be a clinopyroxene-orthopyroxene mix, i.e. websterite. In some localities, large grains show growth fabrics emanating perpendicularly from the dyke walls with smaller, apparently randomly oriented pyroxenes in the dyke interiors (Rassios 1981). These pegmatite dykes are pervasive in the mantle section of Vourinos (Moores 1969), and persist in the crustal section to the level of cumulate diorite. The pyroxenite dykes are tabular in shape, and crosscut nearly all ductile fabric. This implies that they post-date ridgecrest deformation, and perhaps represent local melt phenomena. The rare occurrences of ductile deformation of pyroxenite dykes at Vourinos

PINDOS-VOURINOS OPHIOLITIC SLAB

247

Fig. 6. Young dunite generations in the mantle suite: dunite apparently intrudes older dunite in Vourinos, whereas small dunite bodies intrude both harzburgite and lherzolite in the Pindos. (a) Morphological expression of a dunite body (Aga Kouri, Vourinos) deformed as a ductile ramp with inclusions of undeformed (younger?) dunite. (b) Undeformed 'young' dunite cross-cutting deformed lherzolite pods (circled and labelled 'HZ PX' in harzburgite (HZ N) network matrix (Mavrovouni).

248

A.H.E. RASSIOS & E. M. MOORES

are within distinct mylonite zones such as at Voidolakkos, and proximal to the ophiolitic sole (Grivas et al. 1993). Thus, the formation of these dykes occurred before the final cessation of ductile peridotite deformation. The stereonet diagram of Moores (1969) and subsequent fabric analyses (reported by Rassios et al. 1986, 1991) fail to describe any consistent orientation of these dykes at Vourinos; at present, they are best described as 'random'. By contrast, the pyroxenite pegmatite dykes of the Pindos are rarely tabular, commonly present as boudins within peridotite, and aid the delineation of a mantle 'pseudostratigraphy' across the Dramala massif of the Pindos ophiolite (Rassios 1991). The appearance of pyroxenite dykes in the mantle section of Dramala coincides with a stratigraphic zone containing abundant chromitite occurrences, again suggestive of a correlation between the presence of volatiles and chromite generation. As will be discussed in the next section on structural considerations, the orientation of pyroxenite dykes shows the effect of deformation on these dykes rather than their primary intrusional geometry. C o m p a r a t i v e crustal section

Cumulate petrology differs between the Vourinos and Pindos as follows. The 5 km thick section of Vourinos cumulates consists of basal dunite grading to ultramafic cumulates arrayed in cyclic units: gabbronoritic cumulates, diorites, and small non-cumulate plagiogranite bodies underplating the sheeted dyke zone. Orthopyroxene enters the cyclic units at various levels, generally as websterite beneath gabbronorite tops, but in places as basal cumulate harzburgite. Only rarely are cyclic units topped by troctolite (Rassios 1981; Rassios et al. 1983b). Conversely, the ultramafic to mafic cumulates of the Pindos, with an estimated thickness of not more than 1.5 km, lack orthopyroxene, consist of minor dunite lenses, plagioclase dunite and wehrlite, and are dominated by various gabbro and troctolite rocks underlying kilometre-scale dioritic and plagiogranitic bodies (Rassios 1991). Individual cumulate intrusions can be discerned by the presence of chilled contacts; in short, the cumulates resemble intruding sills (Rassios 2004). Vourinos has a well-developed sheeted dyke unit about 1.5 km thick, but scarce representation of pillow lavas, apparently as screens in upper parts of the dyke zone. The original stratigraphic continuity of the Vourinos extrusive section is preserved (Fig. 4) and coincides with oceanfloor zoned metamorphism (grading from

lower-amphibolite to greenschist and to zeolite facies; Rassios 1981). In their present position, dykes are within an overturned part of the Vourinos section and strike c. N20~ with shallow overturned western dips. Dips of cumulate fabric (vertical to overturned eastern dips) both in the gabbroic section and in the dyke-screen zone of Vourinos (Rassios 1981; Rassios et al. 1983b) are perpendicular to those of the dykes. Sulphides, epidosites, lava breccia pipes and autobrecciated lavas are recognizable as products of sea-floor hydrothermal systems (Rassios et al. 1983a). In its present orientation, the highest pseudostratigraphic section of Vourinos is vertical to overturned; possibly the pillow section is extant in the subsurface (Rassios 1981). Boninitic extrusive rocks among the sheeted dyke complex make up less than 15% of exposures, and dominate only a single locality (<0.25 km 2 in outcrop). This locality occurs in an east-west-trending zone bordering the northern extreme of the crustal block that also hosts exceptional occurrences of plagiogranites, lherzolite and flaser gabbros (the Vatolakkos area, described in a following section); the significance of this petrological-tectonic association within an otherwise island-arc extrusive series is unclear. Rassios (1981; Rassios et al. 1983b) speculated that proximity to an oceanic transform might account for this zone. Whereas the mantle to lower crustal section is continuous in the Dramala complex of the Pindos, the crustal-to-dyke and dyke-to-pillow sections, although extensive, are contained in underlying imbricate thrust slices. These latter are interpreted as a 'ramp pile' verging to the NE, following extensive mapping of this margin (Fig. 5a and b; Rassios 1991, 1994). Ophiolitic stratigraphy overlaps among these imbricates, and the general sense of 'up-section' is constant, although the degree of rotation in the lower ramps (sheeted dykes and pillows) cannot be precisely estimated. The sheeted dyke and pillow lava units are best exposed in the Aspropotamos area (Fig. 2), which Rassios interpreted as the lowest overridden member of the ophiolite (Fig. 5a). This part of the complex is severely imprinted by Alpine structures relating to the formation and development of the Mesohellenic Trough, and severe vertical and horizontal axis rotations of this unit seem to be indicated by the near-vertical positions of peripheral sections. In their present position, the sheeted dykes of the Aspropotamos section are vertical and strike NE. This orientation parallels the imprinting ductile deformation of the Pindos mantle rocks although this could be coincidental. Gabbro dykes cutting the Dramala massif generally parallel the younger phase of mantle

PINDOS-VOURINOS OPHIOLITIC SLAB ductile deformation, i.e. 040 ~ (see following section). Zoned lower-amphibolite to greenschist and to zeolite metamorphism, similar to that of the Vourinos uppermost crustal section, is represented within the Aspropotamos complex.Within extensive but horizontally rotated imbricate, blocks of pillow lavas, lengthy (kilometre-scale) hydrothermal breccia pipes cut the lavas within individual imbricates demonstrating tectonic repetition within the ramped section (Fig. 5b). Rassios, in recent mapping for copper exploration programmes (Rassios & Grivas 2001), has documented abundant hydrothermal metalliferous sediments within lava sequences both as intra-pillow deposits and 20-100 m continuous layers of < 1 m thickness within these imbricates, as also noted by Robertson & Varnavas (1993). These provide excellent indications of original horizontality, but vertical axis rotations cannot be estimated. Unfortunately for economic potential, the exhalative sulphide deposits once presumably associated with the hydrothermal conduits, as well as the base of the oceanic sediments, apparently were exploited as 'weak' zones facilitating out-of-sequence thrust faults.

Fast and slow spreading characteristics The crustal section of Vourinos (Rassios 1981; Rassios et al. 1983b) records the following magmatic stratigraphy characteristic of a fastspreading system (Thy & Dilek 2000): (1) the cumulate section, at 5 km, is very thick; (2) individual lower cumulate magma chambers range in map area from a few tens of square metres to 0.5 km2; (3) the upper (dioritic) cumulates show extensive 'along-strike' continuity (8 km in outcrop) suggestive of a large open magma chamber; (4) diabase dykes do not occur lower in the section than this highest dioritic unit, and form a sheeted dyke complex with a high statistical agreement of chilling direction. Rassios (1981) described a 7 k m horizon within the upper cumulates (between gabbros and diorites) t h a t shows evidence of having hosted a magma chamber or chambers 'feeding' the diabase dykes: this zone is demarcated by abundant penecontemporaneous deformation structures including roof pendants, extensional magmatic boudins along slumping layers, cumulate layers with entrapped diabase, and an appearance of volatile-rich features ('sand dykes') erupting through layers. Diabase dykes are rare below this horizon, but common above. The apparent lack of pillow lavas at Vourinos is interpreted as an artefact of erosion and structural overturn of the ophiolitic section, as will be discussed later.

249

The Pindos crustal section beneath the level of the dykes and lavas of the upper Aspropotamos complex seems to represent relatively slow spreading: cumulates are less continuous, having formed in elongate sills up to a few tens of metres long, with a total thickness of the cumulate section of about 1-1.5 km. Higher-temperature phases such as olivine continue into relative shallow crustal levels. Diabasic dykes cut the upper part of the Pindos mantle series, apparently merging into 'cumulate sills'. The 1-2 km thick dyke section, in addition to substantial pillow lavas of the upper imbricates of the Aspropotamos complex, imply a 'faster spreading' event on the older Dramala substrata. As a result of tectonic thickening, the original thickness of the pillow lava section cannot be estimated.

Geographical overlaps The Dotsikos strip ophiolite. The Dotsikos complex is an up-faulted outcrop of Neotethyan rocks along the western boundary fault of the Mesohellenic Trough and is geographically part of the Pindos Mountains (Fig. 2). Although highly sheared, this complex is sufficiently intact to include a recognizable ophiolitic assemblage, termed the Dotsikos strip ophiolite in this study. This c. 1 km x 40 km NW-trending fragment includes a tectonically attenuated dunite body (of metre scale to 0.5 km thickness and 5 km in length of outcrop) hosted by harzburgite tectonite, second in size in Greece only to the Xerolivado dunite of Vourinos. It includes chromite occurrences within a narrow but continuous 'metalliferous zone' along its length. These occurrences contain 'schlieren' (isoclinally folded) chromitite layers as well as concentrated clots of euhedral spinels; in these respects, the Dotsikos mantle suite resembles that of the Aga Kouri ore district in Vourinos. Also as at Vourinos, a pyroxenite dyke swarm is present. Although severely imprinted b y brittle shearing (rightlateral within the region of the dunite), several high-temperature kinematic indicators within the chromite occurrences appear to be left-lateral, i.e. showing high-temperature kinematic transport (facilitated by ductile shear) towards the east. The 'Moho' unit appears transitional rather than intrusive, and ultramafic cumulate sections fractionate towards troctolite. These features resemble similar pseudostratigraphic levels of the Pindos ophiolite. The Dotsikos strip ophiolite displays petrological similarities to both the Pindos and Vourinos, although incorporating a strato-tectonic section that includes the Avdella m61ange.

250

A.H.E. RASSIOS & E. M. MOORES

Mesovouni massif. Located as a geographical continuation of the Pindos mountains, the Mesovouni massif lies at the south and east extreme of the ophiolitic terrain (Fig. 2). Mesovouni contains a strip of mylonitic peridotite on its northeastern margin, but elsewhere consists exclusively of depleted harzburgite resembling that of Vourinos both physically and geochemically (Pearce, in Rassios 1994). This 'Vourinos-style' area lacks a strong late ductile phase fabric or overprinting ductile fabric, but includes a minor chromite 'metalliferous zone' with ores similar to those of Vourinos; notably, this is the only massif of Pindos-Vourinos with a high-temperature fabric dipping towards the NE. The Mesovouni massif is overlapped by Tertiary serpentine-matrix conglomeritic sediments; no 'boundary fault' with these Mesohellenic sediments is exposed as elsewhere along the Pindos-Mesohellenic boundary. The boundary between Mesovouni and a mylonitic-cataclastic strip of mantle peridotite and ophiolitic cumulate rocks to the north and west is also covered by molasse-type sediments. Vatolakkos section. The Vatolakkos area of Vourinos is located to the N W of the complex, apparently continuous to the east beneath a thin cover of recent sediments with exposures that include flaser gabbros. It is bounded to the north by a steep reverse fault against lower Mesohellenic sediments, and overlain by a series of Mesohellenic sediments containing angular unconformities reconciled by continued activity along this fault. Vatolakkos contains the largest body of plagiogranite of Vourinos (Rassios 1981); this body grades 'pseudostratigraphically' downwards into hornblende diorite and massive gabbro; the base of the section includes a mixture of mafic and ultramafic rocks ('magmatic m61ange' of Rassios), amongst them a body of hydrated lherzolite and oikocryst-bearing wehrlites. The orientation and bifurcation of diabasic dykes intruding the plagiogranite-dioritic section indicate that the section is overturned towards the west around a NW-trending horizontal axis in agreement with a similar overturn observed in the Krapa area to the south. The section is cut by an EW/vertical sinistral-slip fault that is tentatively attributed to the oceanic period as it does not extend into overlying Tertiary sediments and seems to be the source of clay-rich (hydrothermal) veinlets extending into the ophiolitic rocks (Rassios 1981). The Zygosti (Rodiani) ophiolite. Zygosti is a highly imbricated complex of ophiolitic lithologies located 2 km NE of the Vourinos ophiolite, and separated from it by a narrow strip of Pelagonian carbonates; the contact between these

rocks is a NW-trending dip-slip fault apparently raising the Pelagonian units against the serpentinite (Zimmerman 1968). The ophiolitic base is not well exposed, and outcrops of amphibolite and m61ange are not found in the area (Zimmerman 1968). The NW-trending Zygosti belt includes three ophiolitic units, as follows. (1) A highly-sheared 'mantle suite' of serpentinized harzburgite includes dunite bodies of tens of metres to 1 km 2 in area; dunites have been exploited for c. 250 000 tons of Al-rich chrome ore. The mantle section is cut by rodingite and pyroxenite dykes, suggesting hydration of the peridotites during intrusion. (2) A tectonically attenuated crustal section of ultramafic to dioritic cumulates is not more than 150 m in thickness. Pyroxene-olivine oikocrysts indicative of volatile presence, concurrent with magmatism and hydrated serpentinized peridotite, dominate ultramafic rocks; serpentinized dunite layers are found in the dioritic levels; the dioritic levels are intensively intruded by diabasic dykes and contain pervasive felsic-zeolitic veinlets. (3) Sheeted diabase dykes and pillow lavas appear always to be in tectonic contact with the ultramafic and crustal section rocks. Pervasive brittle shearing dominates the appearance of the mantle and crustal sequence in comparison with that of Vourinos or Pindos. These sheared formations were eroded and overlain by Cretaceous limestone (Brunn 1956) that lacks significant deformation, and thus dates the Zygosti shearing to the oceanic period. Tectonically incorporated metre-scale clasts of Pelagonian limestone in highly sheared serpentinites crop out along the north of the belt. Mineral foliation and mylonitic shears can be discerned in the least serpentinized harzburgites of the central area with c. NE/70~ strike/dip, i.e. parallel to orientations dominant in south Vourinos. Metrescale brittle imbricates in the ultramafic sections verge to the NE. Original horizontality within the ultramafic cumulates in the south of the belt is c. l l0/vertical, with 'up-section' defined by graded mineral layers and petrogenetic trend towards the north. Diabase dykes intrude a nearhorizontal lava section at c. 150/50~ i.e. parallel (by coincidence) to dykes in the Krapa Hills of Vourinos. Points of comparison with Pindos and Vourinos are as follows: (1) although the mantle suite resembles that of Vourinos, the chrome ores are significantly Al-rich compared with Vourinos (Economou et al. 1986); (2) pyroxenite and rodingite-altered gabbro pegmatite dykes are

PINDOS-VOURINOS OPHIOLITIC SLAB present in both Zygosti and Vourinos, but visually appear more abundant in Zygosti; (3) all levels of the Zygosti complex appear dominated by hydration, and thus more closely resemble portions of the Pindos (Aspropotamos) gabbrodioritic zone and Othris mantle section; (4) the presence ofdunite layers high in the cumulate section is reminiscent of the Aspropotamos sequence of Pindos.

Structural considerations Mantle structures

The following mantle structural features are considered as means of envisioning a single-slab framework for the Vourinos and Pindos ophiolites: (1) fabrics such as lithological layering, mineral foliation and lineation provide a geometric means of comparison of mantle terrane strain orientation analogous to deformed stratigraphic sections; (2) mantle kinematic textures demonstrate variation in the temperature-pressure regime effective at the cessation of ductile deformation and these trace ductile mass movement; (3) ophiolitic structures that appear continuous across the Mesohellenic Trough, including primary features such as chromite metalliferous zones and ductile shear zones; (4) structures forming at the ductile-brittle transition (Rassios et al. 1994) provide firm evidence for strain orientation during initial emplacement. These features are genetically overlapping: all formed in a short period coinciding with conditions of ductile strain of mantle rocks from rifting to emplacement. For the Pindos and Vourinos ophiolites, this appears to be < 6 Ma (Rassios & Smith 2000). At all stages of ductile deformation, brittle deformation was synchronous in higher and/or cooler parts of the slab. As the ophiolite left the spreading centre environment, the ductile-brittle margin deepened and brittle fabrics were imprinted on slightly older ductile fabrics. Kinematic zones and shear zone continuities

The temperature-pressure constraints for deforming peridotite and dunite have been studied comprehensively both in experimental and ophiolitic settings (e.g. Ave Lallemant & Carter 1970; Ave Lallemant et al. 1980; Ross et al. 1980; Nicolas 1989a). High-temperature plastic deformation initiates about c. 1250 ~ and coincides with magrnatic temperatures elsewhere in a rifting environment. Ductile deformation of peridotite ceases at about 700 ~ with some variation

251

depending on the presence or absence ofvolatiles. Between these ductile boundary conditions, peridotite deformation changes from intra-crystalline mechanisms to mylonitic, inter-granular processes. The appearance of peridotites during this transition (Nicolas 1989a; Fig. 7a) changes gradually from a blocky texture of orthopyroxene in a host of equigranular olivine, to an elongated appearance of orthopyroxenes, to long ribbons of orthopyroxene, and thence to total cataclasis at the cooler extreme. Rassios (1991) used the appearance of peridotite in the field, essentially orthopyroxene elongation ratios, and back-up petrography to map these ductile fabric regimes in the Dramala massif. Figure 8 extends the distribution of these fabric regimes over mantle rocks of the Pindos and Vourinos ophiolites. In this paper, we would like to introduce the concept that these mappable mantle domains be considered as kinematic zones: each gives information on the conditions of strain and ductile movement recorded in the mantle suite. Overall, the impression given in Figure 8 is one of heterogeneous movement within a slab; some areas (high-temperature ductile fabric zones) that ceased ductile deformation seem to have been transported passively amidst oceanic mylonite (lower-temperature ductile fabric) zones. Figure 8 also emphasizes that mantle fabrics that are blocky in appearance (high-temperature fabric) characterize Vourinos whereas mylonitic fabrics dominate the Pindos mantle rocks, as noted previously, utilizing a statistical approach (Rassios & Smith 2000). The orientation of these kinematic zones does not always correlate with that displayed by mantle rock fabric (Fig. 7b). For example, a mylonitic kinematic zone of the Dramala massif extends north-south, whereas peridotite fabrics strike NE (Rassios 1991). The ability of rocks to record strain apparently varies with heterogeneous physical parameters within the deforming unit: the orientation of the finite strain ellipse is constant, whereas the mode in which strain is recorded varies. In the case of Dramala, much of the heterogeneity corresponds to areas that suggest the presence of volatiles. The concurrence of mylonitic peridotite to zones directly above the emplacement sole has been observed in many ophiolites (Nicolas 1989a), and such zones are distinguished for both Vourinos and Pindos on the spatial map (Fig. 8). However, this map also delineates mylonitic kinematic zones at higher levels within the mantle slab. Some of these parallel the sole and are best interpreted as facilitating emplacement-related strain within higher levels in the slab. Elsewhere, mylonitic zones correlate in position with shear

252

A . H . E . RASSIOS & E. M. MOORES

Fig. 7. (a) Sketch adapted from Nicolas (1989) demonstrating elongation of mineral grains during constant ductile shear conditions and orientation of strain. These sketches correlate with decreasing temperature during deformation (c. 1200 ~ to left, and ductile-brittle boundary r 700~ towards the left). Areal mapping of such elongation leads to the 'kinematic zone' map of Dramala (Rassios 1991) and the Mesohellenic mantle suite of Figure 8. (b) Sketch demonstrating how strained zones in a mantle region of homogeneous temperaturepressure have been observed to have an orientation not parallel to regional mineral fabric. Such heterogeneous strain could be due to volatile presence inducing greater relative deformation in these zones. (c) Ductile shear zone crossing Mavrovouni, Pindos ophiolite. Sketch shows kinematic indicators of trailing spinel bands on the ductile surfaces. (d) Small but typical mantle fold exposed on glacially scoured surface atop Salatoura, Pindos ophiolite. Fold axial surface strikes 040~176 and axis plus regional spinel lineation trend 040~176 (e) Photograph with sketched geometric relations between folded chrome ore in dunite of the Kerasitsa mine district, Vourinos ophiolite (photograph by G. Konstanopoulou). zones (Fig. 7c) that are geometrically consistent with a sense of ductile tear within the slab or, in the case o f the pervasive cataclasis of the K o u k o u r e l o massif of the Pindos, delineate elongate N E - t r e n d i n g zones that m a y indicate the margins o f the slab itself. A similar zone is present across

the Mesohellenic at Vourinos a l t h o u g h with a less extreme mylonitic fabric. The m a p pattern combined with the depiction of magnetically defined shear zones b e n e a t h the Mesohellenic sediments, implies that these features span the two ophiolites.

PINDOS-VOURINOS OPHIOLITIC SLAB

253

Fig. 8. Map distribution of dominant fabric types interpreted as iso-kinematic zones within the mantle rocks of the Pindos and Vourinos ophiolites. Shear sense (arrows) in ductile conditions is deduced from ground observations as explained in the text, and inferred in some subsurface conditions based on magnetic lineament patterns and continuity with exposures. Metalliferous zones

Chrome ores at Vourinos are concentrated within a metalliferous zone within the mantle suite that includes all former economic workings and an estimated 80% of chromitite occurrences (Vrahatis & Grivas 1980). This zone, approximately half a kilometre in width, also contains a higher proportion of dunite than is present in regional harzburgite. Both dunite and chromitites appear to have preferentially recorded ductile deformation of ridge-crest structures such as mantle layering, dunite-harzburgite contacts and chrome ore layering. This zone generally parallels the sole contact with a N W trend in the north, is offset in a right-lateral sense by a Cenozoic fault between the northern and southern blocks of Vourinos, and within the southern block of Vourinos turns to a N E trend corresponding to regional fabrics. Frison (1987) and Nicolas (1989b) suggested that the metalliferous zone of Vourinos corresponds to the position of a palaeo-spreading ridge, although structural traverses exhibiting changes in dip of mantle layering and foliation across

this zone do not support this, nor can the zone be structurally continued into magmatic rocks at the petrologic Moho. Grivas et al. (1993) described the metalliferous zone as a strongly deformed, dunite-rich boundary between more coherent peridotite blocks. Dunite and chromites are concentrated within this margin. It remains unclear whether these zones are primary features that preferentially accommodate deformation because they are less competent than surrounding harzburgite blocks, or are an artefact of ductile-stage tectonic thinning. Could such zones comprise a marker 'horizon' within mantle pseudostratigraphy? In the Pindos, a metalliferous zone has been located in three areas, although without economic promise (Rassios & Grivas 2001). Two areas, in Dotsikos and in Mesovouni, are similar, both in general petrology and scale of dunite bodies, to the Vourinos metalliferous zone. Both seem to have been rotated by deformation at the limits of the ophiolitic complex, and Dotsikos, additionally, is severely attenuated by the boundary fault of the Mesohellenic Trough. A third zone is located in the Mavrovouni massif of the Dramala

254

A.H.E. RASSIOS & E. M. MOORES

complex. It contains ores essentially identical in composition (Pearce, in Rassios 1994) and orientation (NW trending) to those of Xerolivado in Vourinos: the host dunite is strongly attenuated by ductile-phase deformation. This may constitute a continuation of the Vourinos metalliferous zone across the Mesohellenic Trough into Pindos terrain. Fabric analyses

Mantle folding has been documented at Vourinos (on scales down to 1:500, Grivas et al. 1993) where tectonic analyses guided exploratory drilling for chrome ores in 11 districts. In all of these areas, high-temperature spinel lineations correlate with axial lineation of folds verging to the NE. Folds tend to be open in areas in the core of the northern block (e.g. Voidolakkos) in which the fabric is blocky, and tend to be sheath folds in more intensely deformed areas such as Xerolivado. Smaller-scale folding is always parasitic on the flanks of larger folds ad infinitum. The ore bodies themselves correspond in position to NEverging synformal axes as linear pods, parallel to regional spinel lineation (Roberts et al. 1988). The accuracy to which chrome was targeted using these principles (including 'blind' deposits at depths over 200 m in the subsurface) supports the 'compressional ductile nappe' model as the chief structural control on the position of chrome deposits (Rassios 1994). The relation between ductile and brittle fabrics mapped on a large scale (1:1000 to 1:500) at 11 chromite ore districts of Vourinos can be summarized as follows (Fig. 9). (1) Ductile and brittle geometries are parallel or overlapping at all localities, suggesting that there is a continuation between ductile and brittle strain (this has been documented for the Voidolakkos District by Grivas et al. (1993), and details of structures forming within the ductile-brittle boundary are by Rassios et al. (1994)). (2) Structures of both the ductile and early brittle episodes verge NE, in agreement with offsets observed in out-of-sequence displacements of ore bodies and dunite-harzburgite contacts. (3) Three areas in the 'core' of Mount Vourinos (NW part of the complex; including the Voidolakkos, Exarkos and Aga Kouri localities of Fig. 9) display high-temperature planar fabrics essentially parallel to the plane of the basal sole, i.e. striking N W with moderate western dip. (4) Three southern districts, including Xerolivado, Brovikas and Ano Konivos, are

dominated by NE-striking pervasive sheath folding (folding on scales of 500 m down to 5 m scale) verging to the NE (NE-trendingSE-plunging fold axes correlating with spinel lineations). This geometry grades into and parallels lateral ramps (tear shearing) in the brittle field (detailed in Fig. 9 for Ano Konivos and Brovikas). (5) Three districts in the central area of Vourinos include dunite bodies with outcrop patterns striking east-west. The Kersitsa and Rizo districts each demonstrate a dominance of z-folds that translate fabric from NW- or south-dipping into NE or steep north- or vertical-dipping orientations. Form-line maps (1:20 000 scale; general trends shown in Fig. 8) were generated over the entire extent of the Dramala complex of the Pindos. These show that the mantle fabric of the Pindos peridotite nappes is geometrically parallel to that of Vourinos. When stereonet diagrams (Fig. 10) are prepared for the mineral ductile fabric regimes and compared with primary structures such as mantle layering, dykes and early brittle shear systems, an evolution of structural style within a single geometric orientation is displayed, as follows: (1) mantle layering in the Pindos parallels the highest temperature mineral fabrics at N W strike/SW dip of Vourinos; (2) mineral fabrics rotate from this initial strike into a NE or vertical orientation with decreasing temperatures of ductile deformation; (3) gabbro dykes crossing the Pindos mantle section parallel the NE/vertical deformation; (4) ductile-brittle shear zones and the earliest brittle shear systems are NW- or SEdipping, i.e. parallel to and apparently in the same strain geometry as the highest temperature mantle fabrics. High-temperature ophiolitic peridotite fabrics originate in the mantle beneath an evolving spreading centre and represent extensional deformation (Nicolas 1989a). Ductile-brittle structures of Pindos and Vourinos are interpreted as representing an off-axis early emplacement geometry (Rassios et al. 1994; Rassios & Smith 2000). Their concurrence in orientation with high-temperature mantle features implies that the original So mantle structures have been rotated into this geometry as well. If so, the problem of determining the original orientation of ridgecrest mantle geometry based on these fabrics becomes equivalent to deducing the initial So orientation within any multiply deformed metamorphic complex as was done by Ramsey & Huber (1983). Lacking stratigraphic indicators within the mantle, the orientation remains inconclusive other than indicating a rotated original ridge-crest direction.

PINDOS-VOURINOS OPHIOLITIC SLAB

255

Fig. 9. A compilation of structural data in mantle rocks of Vourinos. Ductile structures summarized in stereonet diagrams and form lines are ductile-field mineral foliations. Brittle fabrics denote shears and imbricates of the mantle section. Stereonet diagrams demonstrate that ductile and brittle structures are essentially parallel. Lineations cited are mineral lineations, chiefly spinel lineations.

256

A . H . E . RASSIOS & E. M. MOORES

r

4 o

.~

2

s

.<

. ,...~

"-6 o

's

2 o

PINDOS-VOURINOS OPHIOLITIC SLAB The means of rotating the older mantle structures into the emplacement orientation is apparent from the deformation within the mantle complex itself. The observation of hightemperature folding of these mantle extensional fabrics was made as long ago as the studies by Ayrton (1968) and Moores (1969): if hightemperature folding was solely ridge crest in origin, it would require explanation of compressive structures in the extensional ridge crest. The resolution of this problem is that the folding of the older fabric to the younger generates a constant orientation of spinel and orthopyroxene lineations (NW, with c. 30 ~ plunge to the SE) to form the axial lineation of NE-verging fold systems. As previously mentioned, the chromite drilling programme served as a successful test of this deformation model. Are the primary mantle fabrics of all ophiolites deformed into an emplacement orientation? Without examination of these features elsewhere, we cannot answer this question; very few ophiolitic structural analyses include a synthesis of high-temperature deformation down to and crossing the ductile-brittle boundary. Ductile deformation of the Oman ophiolite was interpreted by Nicolas (1989a) as terminating the preceding establishment of detachment strain, such that oceanic ridge structures were considered separately from detachment and emplacement. As noted, the time sequence between rifting and emplacement of the Vourinos-Pindos slab is short, and this might be a facilitating factor. The obduction overprint onto the Pindos-Vourinos slab casts doubt on whether mantle layering can represent the original orientation of the Pindos ocean ridge crest. The present orientation is NW/S-dipping for both ophiolites, an orientation that can also have accommodated slab emplacement motions. However, by extending a sense of pseudostratigraphy into the mantle complex, as we attempt in the next section, an opposite sense of rifting is implied.

Orig&al horizontality and spreading centre geometry The most reliable indicator of original extensional geometry is the orientation of magmatic pseudostratigraphy and sheeted dykes. At Vourinos, the crustal block of Krapa (Fig. 4) contains an intact magmatic profile from ultramafic cumulate lithologies upwards to the oceanic sedimentary contact. The ultramafic base of Krapa overlaps petrologically with that of the lowermost cumulates above the petrological Moho. The cumulate layering and lamination in the ultramafic rocks overlying the 'Moho' as well

257

as those of Krapa are vertical, strike north to NW, and their 'up-direction' is towards the west. The Krapa block has been uplifted by a Cenozoic vertical reverse fault (with a probable right-lateral slip component) along its eastern margin. The magmatic section includes a preTertiary monocline, such that the section turns from vertical in the east to overturned towards the direction of the Mesohellenic Trough in the west. The upper contact of the Vourinos ophiolitic unit provides a significant indication of the horizontal-axis rotation of the ophiolitic unit within the oceanic environment. This contact includes Jurassic deep-water sediments generally interpreted as in situ (Moores 1969; Mavrides et al. 1979; Mavrides 1980), deposited above a section of eroded dykes and flows showing greenschist-facies metamorphism. A small exposure of pillow lavas is present as a screen within zeolite-facies extrusive rocks that include exhalative jaspers. This upper Jurassic surface is truncated by an angular and topographically irregular unconformity with Cretaceous carbonates. Rassios (1981) suggested an angular unconformity between the Jurassic sediments and the underlying extrusive section, with a < 12~ horizontal counter-clockwise rotation around a N W axis. The apparent lack of pillow lavas at Vourinos would then be explained by a combination of sea-floor erosion preceding deposition of the Jurassic sediments, with a remnant pillow section in the subsurface of the overturned section. The deposition of the Cretaceous limestone over this unconformity indicates a further rotation of 40-80 ~ (counter-clockwise around a NWtrending axis) in the period between deposition of the Jurassic sediments and Cretaceous limestone, i.e. during the same time period that included sole formation and ophiolitic emplacement. The sheeted dykes of Vourinos have a strong, systematic (overturned) orientation at northNW/40-70~ Rotating the orientation of the sheeted dykes backwards through these horizontal rotations indicates an original orientation, presumably ridge-crest parallel, of N20~ with dip c. 80~ The dykes also show a strong preferential chilling direction (a ratio of two to one favouring eastern margin chilling, based on >200 observations) as might be expected in a fast-spreading centre (Rassios 1981), the directionality of which would indicate a spreading ridge located to the west of their reconstructed position. An approximation of the present-day attitude of 'original horizontality' of the Pindos ophiolite based on the orientation of the petrological

258

A.H.E. RASSIOS & E. M. MOORES

Moho and preserved stratigraphic successions in the Dramala complex, and in general agreement with cumulate-bearing imbricates of the Aspropotamos complex, is c. NW/30-45~ with up-section towards the east (Fig. 5). The diabase dykes within the Pindos Aspropotamos section occur in a fault-bounded block. Within this, dykes strike 040 ~ and dip vertically, although their 'up-section' polarity is unknown. Severe rotations of this section associated with the formation of the Mesohellenic Trough bring into doubt whether the dyke orientations reliably constrain ridge-crest orientation. However, gabbro dykes within the Dramala mantle section in areas that have not been affected by the Mesohellenic structures also are oriented at about 040~ and this direction is the best estimate possible for the direction of the Pindos spreading axis. Structurally, the Vourinos and Pindos complexes describe a synform with an axis beneath the Mesohellenic Trough. A nearly vertical westfacing original horizontal surface of Vourinos on the eastern margin faces the east-topping, more gently sloping, formerly horizontal surface of the Pindos to the west (Fig. 11).

The emplacement geometry An inverted 'spoon-shaped' geometrical form (Fig. 12) is modelled by applying fabric orientations at distinct locations in the slab (essentially defined by the distribution of shapes shown by stereonet diagrams): the geometry of the strain ellipse is everywhere consistent. The form is that of a nappe. The synformal horizontality of Pindos and Vourinos is approximated by this shape, with further rotations in the Pindos being explicable by later deformation.

Structural summary The present orientation of the ophiolitic slab is one of a synform with an axis parallel to the Mesohellenic Trough; the Pindos represents the western limb, and Vourinos the eastern limb. All parts of this ophiolitic slab were overprinted by deformation spanning the ductile-brittle transition: this deformation was responsible for the folding of mantle features and the creation of a pervasive NE-southward trailing mineral lineation that conforms to fold axes. It also gave rise to other NE-verging structures, such as ductile mylonite zones that 'ramp' towards the NE or that accommodate differential NE-trending movement within the slab, and ductile-brittle to brittle structures that document a thrusting geometry verging to the NE.

Implications and conclusions The crystallization age and dating of sole formations of the Vourinos-Pindos ophiolites requires that they represent simultaneous generation of oceanic lithosphere in the Pindos ocean, and that they were obducted together within an originally coherent slab. Magnetic maps of the Mesohellenic Trough in Greece locate the presence of this slab in the subsurface. We consider VourinosPindos as a single segment within an even larger Mesohellenic slab that extends well beyond the study area. The Vourinos-Pindos segment is bordered to the south and north by ductile nappe boundaries, and thus was emplaced as a distinct structural unit. Possibly, the northern border overlaps an oceanic transform. All the Mesohellenic ophiolites appear to have undergone vertical-axis rotation within the Tertiary yet retain a coherent structural geometry (Robertson & Shallo 2000; Dilek et al. 2005) so that matching of these other segments should be possible.

Slab heterogeneity Increasingly detailed observations from modern ridge crests document a structurally complex environment (see Robinson et al. 2000; Thy & Dilek 2000). In ophiolitic settings, such complexities have in the past been relegated to 'artefacts' of obduction and preservation. Petrological and structural relations within the VourinosPindos slab touch on possible mechanisms for creating such complexity: in these ophiolites it is difficult to separate emplacement from spreading features. The tectonic processes displayed in the transition between ridge and emplacement could explain a great deal of oceanic and ridge-crest heterogeneity. The latest magmatic activity in both the Pindos and Vourinos mantle is the intrusion of pyroxenite and gabbro dykes. These dykes have undergone pervasive NE-verging ductile deformation in the Pindos. In Vourinos, pyroxenite dykes are deformed only in rare occurrences within distinct mylonite zones at the base of Vourinos, or in NE-verging mylonitic ramps as in Voidolakkos (Grivas et al. 1993). Gabbroic dykes are not deformed at Vourinos, but crosscut mantle section structures in mining districts near the petrological Moho (Aetoraches, Rizo). Ductile emplacement motions coincided with late ridge-crest magmatism in the Pindos, but crossed to the ductile-brittle boundary and the brittle field at Vourinos before cessation of magrnatism. Much of the Vourinos mantle suite retains high-temperature mantle ('diapiric') fabric lacking any significant intermediate ductile phase of deformation, whereas the reverse is true of

PINDOS-VOURINOS

O P H I O L I T I C SLAB

259

+..a . ,,,,,~ ,,,,~

O

O

O

r.~

,-=..

+..a

Oh"

,3,.-:.

"=~ O

,....~

260

A.H.E. RASSIOS & E. M. MOORES

Fig. 12. Reconstructed position of the chief ophiolites of the study area within the spoon-shaped nappe morphology defined by the geometry of their form lines. The generally higher-temperature mantle fabrics frozen into the Vourinos margin agree with a relatively rapid obduction across the ductile-brittle margin, whereas the trailing Pindos margin agrees with the observed prevalence of mylonitic~cluctile cataclastic fabrics. The mantle rocks of Mesovouni area resemble, in blocky appearance and petrology, those of the Vourinos complex, but form lines dip away from those of the Vourinos complex and all other mantle localities of the Pindos. the Pindos. Vourinos was apparently emplaced quickly, passing through the ductile-brittle boundary with scant intermediate deformation. The distribution of kinematic zones within the Vourinos-Pindos slab implies a variation in the mechanical rates of emplacement within the slab. These phenomena and internal NE-verging structures imply that some of the heterogeneity within the slab is explicable as geographical overlap along leading- and trailing-edge nappe boundaries. The ductile nappe model implies that Vourinos, as the leading edge, should not be continuous to the south to other ophiolites. Indeed, this is the case, as the Pelagonian Zone bounds the southern margin of Vourinos, and includes impressive dextral-extensional shear along this margin at the base of Mount Vounassa (Permo-Triassic section) continuing to the sole area of the ophiolite, the 'footprint' of the passage of the Mesohellenic slab. Across this lateral ramp margin, ophiolites in exposure and in the

subsurface (Figs 2 and 8) are removed far to the west. The Mesovouni massif is the nearest fully exposed ophiolitic block south of this margin. Thus, the reverse dip of mantle structures and the general resemblance of Mesovouni to Vourinos could be explained by it representing a continuation of Vourinos, itself trailing along and rotated by the lateral ramp. The ductile nappe model also implies that a northern lateral ramp margin should be dominated by sinistral shear motion, and that some 'Vourinos-style' phenomena might be present across the Mesohellenic Trough within the geographical area of the Pindos Mountains. So far, field studies are inconclusive, but the following supportive features can be pointed out. (1) Along the northern 'trailing edge' of Vourinos, Pelagonian carbonates demonstrate mylonitic zones parallel to the trend of the emplacement sole of Vourinos. Large blocks of marble tectonically included within the emplacement zone show left-lateral

PINDOS-VOURINOS OPHIOLITIC SLAB extensional shear boudinage. Form lines within the basal mylonitic peridotite turn slightly towards east-west. (2) The Vatolakkos section at the N W extreme of the Vourinos complex (Fig. 2) includes a left-lateral east-west-trending brittle fault tentatively dated to the oceanic period. The 'flaser' gabbros and hydrated appearance of this unit suggest transform motions concurrent with magmatism. The uplifted block of lherzolite may mark a tectonically upfaulted remnant of fertile mantle. (3) Across the Mesohellenic Trough to the west, the Dotsikos ophiolitic strip complex demonstrates a petrological overlap between the Pindos and Vourinos ophiolite. Ductile structures indicate sinistral shear at several localities, but penetrative shearing from the structural boundary of the Mesohellenic Trough precludes a statistical kinematical analysis (Rassios & Grivas 2001). The petrology of 'near-Moho' sections and the relation of this complex to the Avdella m61ange resemble those of the Pindos ophiolite. (4) The Zygosti complex displays a degree of shearing, magmatic hydration phenomena, and pre-Cretaceous faulting that may indicate the inland presence of a transform fault accentuated by nappe-marginal tectonism. Dunite occurs at high stratigraphic levels of the cumulate section in a crystallization order similar to that of the Pindos cumulates. The following observations place constraints on ridge-crest models for this segment of the Pindos ocean and result in several hypothetical palinspastic models: (1) the relative horizontal rotations of Vourinos and the Pindos within the slab as an oceanic deformation, as shown by the parallelism of the late-stage ductile imprint everywhere, and by the non-deformed, less severely rotated Cretaceous limestone; (2) the contemporaneous crystallization ages implied by the zircon dates; (3) overlapping but by no means identical petrological characteristics; (4) fast spreading observed in Vourinos, and slow spreading in the Pindos; (5) structural co-genetic features spanning the Mesohellenic Trough such as the metalliferous zone, late pyroxenite and gabbroic magmatism, and boninitic intrusion. The simplest model is the existence of a single ridge crest (Fig. 13a) with heterogeneous spreading behaviour along it. The ridge would extend from the Pindos to Vourinos (a NE direction in present coordinates). Limited areas along the ridge crest have faster or slower spreading rates, and varying degrees of depletion of mantle

261

sources. Primary ridge-crest structures are pervasively imprinted by early emplacement deformation and thus cannot be distinguished. The synformal mirrored stratigraphy is an artefact of oceanic-stage emplacement deformation. Although it is consistent with the VourinosPindos region of the Mesohellenic slab, it is doubtful that the model would hold if extended northward to include the Albanian Mirdita ophiolite or southward to Othris unless significant crystallization age differences could be established between these (presumably more remote from the spreading ridge and thus older) regions. A second ridge-crest model emerges when original horizontality as defined by crustal sections is reconstructed, along with an assumption that the original mantle orientation is an extension of these sections (Fig. 13b): in doing so, hightemperature mantle structures are now divergent (i.e. as if each is from a different side of the spreading centre), and a single ridge crest would be parallel to and now within (relative to present coordinates) the Mesohellenic Trough. The crystallization ages of Vourinos and Pindos represent similar ages on opposite sides of the ridge. The variation of petrology and spreading rate between the two ophiolites implies asymmetrical spreading. A more complex, but realistic, construction suggests that the ophiolites originated in a complex oceanic environment with NE (present coordinates) and N W spreading taking place simultaneously. The NW-trending spreading centre is represented by the Krapa Hills part of Vourinos, the Mesovouni massif and the Dotsikos strip ophiolite, whereas the Dramala complex and most of the Aspropotamos unit represent the NE-trending system. The Zygosti unit and Vatolakkos area suggest the presence of a spreading centre-transform fault sequence. Figure 13c is a schematic illustration of these relationships. Similar complex structures are present in such regions as the Lau-Havre basin, the complex microplate interactions at the Easter Island and Juan Fernandez triple junctions (Parson & Wright 1996; Bird et al. 1998; Neves et al. 2003). Many of the heterogeneities within the Vourinos-Pindos segment of the Mesohellenic slab can be explained by nappe-style emplacement tectonism. This is a feasible mechanism for explaining geographically overlapping phenomena such as those of Mesovouni (part of Vourinos in the Pindos?) and Zygosti (with Pindos-style upper cumulates east of Vourinos?). This concept needs to be extended to other ophiolites of the Mesohellenic slab, both towards Othris in the south and to the Albanian ophiolites

262

A . H . E . RASSIOS & E. M. MOORES

Fig. 13. Reconstructions of unified Mesohellenic ophiolite slab. (a) Model 1. 'Classic' structural analysis: Parallel primary and imprinting ductile and brittle structures imposed on slab undergoing horizontal-axis rotations during emplacement; all structures are controlled by emplacement. Synchronicity of magmatic sections requires a single ridge crest parallel to the section, with spreading at right angles. (b) Model 2. Assuming high-temperature structures (So) are imprinted but not rotated by subsequent ductile-brittle emplacement deformation, this reconstruction rotates ophiolitic orientation to original 'up-section'. This then requires that a ridge crest separates the Pindos from the Vourinos as their mantle structures dip away from each other, and that synchronous Zr dates reflect a symmetry about the spreading axis, located (present geography) beneath the Mesohellenic Trough. (c) Model 3. One possible palinspastic reconstruction places a propagating ridge crest between the major Pindos and Vourinos areas, explaining the variation in axis-parallel dyke geometry of the Krapa area of Vourinos with that of the Pindos, and by extension, the rest of Vourinos. The distribution of cumulates at Vourinos, lacking in the south, can be explained by oceanic period uplift of lithosphere as a result of movement along a (?)transform parallel to the Zygosti ophiolitic margin.

PINDOS-VOURINOS OPHIOLITIC SLAB

263

inferences from peridotite xenoliths. Tectonophysics, 70, 85-113. AVRTON, S. 1968. Structures isoclinals dans les pGridotites du Mont Vourinos (MacGdoine Grecque)---un exemple de dGformation de roches ultrabasics. Bulletin Suisse de Min&ologie et POtrologie, 48, 734-750. BECCALUVA, L., OHNENSTETTER, D., OHNENSTETTER, M. & PAUPY, A. 1984. Two magmatic series with island arc affinities within the Vourinos ophiolite. Contributions to Mineralogy and Petrology, 85, 253-271. BIRD, R., NAAR, D., LARSON, R., SEARLE, R. & SCOTESE,C. 1998. Plate tectonic reconstructions of the Juan Fernandez Microplate; transformation from internal shear to rigid rotation. Journal of Geophysical Research B, Solid Earth and Planets, 103(4), 7049-7067. BRUNN, J. 1956. Contribution h l'Gtude du Pinde septentrional et d'une partie de la MacGdoine occidentale. Annales G~ologiques du Pays Hellkniques, 7, 1-358. CAPEDRI, S., VENTURELLI,G., BOCCHI,G., 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. DILEK, Y. & MOORES, E. 1990. Regional tectonics of the Eastern Mediterranean ophiolites. In: MALPAS, J., MOORES, E., PANAYIOTOU,A. & XENOPHONTAS, C. (eds) Ophiolites, Oceanic Crustal Analogues," Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 296-309. Many of the data reported here were collected during DILEK, Y., SHALLO,M. & FURNES, H. 2005. Rift~irift, the compilation of technical reports on exploration seafloor spreading, and subduction tectonics of for chromite and copper ore in the western Hellenides Albanian ophiolites. International Geology Review, and have not been previously published in readily 47, 147-176. accessible literature (they are in Greek or in European ' DOUTSOS, R., PE-PIPER, G., BORONKAY, K. & Union reports), or these data are nonconfidential KOUKOUVELAS,I. 1993. Kinematics of the Central information contained in confidential mining reports. Hellenides. Tectonics, 12, 936-953. The data reported here include material collected DuPuv, C., DOSTAL,J., CAPEDRI,S. t~ VENTURELLI,G. during a 10 year EU-IGME (European Union1984. Geochemistry and petrogenesis of ophiolites Institute of Geological and Mineral Exploration of from northern Pindos (Greece). Bulletin Greece) development project to improve chromite Volcanologique, 47(1), 39-46. exploration methodology with contributions from the ECONOMOU, M., DIMOU, E., ECONOMOU, G., et al. following scientific team: (from Greece) A. Rassios, I. 1986. Chromite Deposits of Greece. Theophrastus, Athens, 129-159. Vacondios, E. Grivas, G. Konstanopoulou, G. Memou FRISON, J. 1987. Les pbridotites du massif ophiolitique and G. Skianis, and (from Britain and France) L. du Vourinos (Grkce). Etude petro-structurale mise Wright, S. Roberts, J. Pearce, D. Kostopoulos, T. en evidence d'une structure diapirique. PhD thesis, Moat and A. Nicolas. A. Robertson, S. Kokkalas and University of Paris. A. Magganas are thanked for thoughtful and challengGRIVAS, E., RASSIOS, A., KONSTANTOPOULOU, G., ing reviews. Y. Dilek's contagious enthusiasm added VACONDIOS, I. & VRAHATIS,G. 1993. Drilling for greatly needed impetus. 'blind' podiform chrome ore bodies at Voidolakkos in the Vourinos ophiolite complex, Greece. References Economic Geology, 88, 461-468. HARKINS, M., GREEN, H. & MOORES, E. 1980. MulANONYMOUS 1972. Penrose Field Conference on tiple intrusive events documented from the ophiolites. Geotimes, 17(12), 24-25. Vourinos ophiolite complex, northern Greece. AVE LALLEMANT,H. & CARTER, N. 1970. Syntectonic American Journal of Science, 280-A, 284-290. recrystallization of olivine and modes of flow in JACKSON, E., GREEN, H. & MOORES, E. 1975. The the upper mantle. Geological Society of America Vourinos ophiolite, Greece: cyclic units of lineated Bulletin, 81, 2203-2220. cumulates overlying harzburgite tectonite. AVE LALLEMANT,H., MERCIER, J.-C., CARTER, N. & Geological Society of America Bulletin, 86, 390-398. ROSS, J. 1980. Rheology of the upper mantle:

to the north. The misinterpretation of such geographically overlapping phenomena could create the modelling of redundant spreading axes that are difficult to correlate into a realistic palinspastic reconstruction of a unified Pindos ocean. Ideally, such a palinspastic model would show different views, from different ophiolitic data points, of the same phenomena of synchronous spreading systems with similar palaeogeographical relations. The Pindos and Vourinos ophiolites do not represent an idealized Penrose-style ophiolitic section. Their nature as simultaneously produced but heterogeneous petrological entities preserved within a single emplacing slab illustrates the complex reality of the oceanic crust and upper mantle, and aids in the explanation of similar phenomena observed in modern ridge-crest systems. It is not, perhaps, surprising that ophiolites may be emplaced with a geometrical orientation resembling that of ductile nappes, but it is remarkable that such observations remain largely undocumented for other ophiolitic slabs. Possibly, the short time interval between ridge crest and emplacement of the Pindos ocean served to promote an overlap between ductile field deformation and emplacement motions.

264

A . H . E . RASSIOS & E. M. MOORES

JONES, G. 1990. Tectono-stratigraphy and evolution of the Mesozoic Pindos ophiolite and associated units, northwest Greece. PhD thesis, University of Edinburgh. JONES, G. & ROBERTSON, A. H. F. 1991. Tectonostratigraphy and evolution of the Pindos ophiolite and associated units. Journal of the Geological Society, London, 148, 267-288. JONES, G., ROBERTSON, A. H. F. & CA~,q~, J. 1991. Genesis and emplacement of the supra-subduction zone Pindos ophiolite, northwestern Greece. In: PETENS, T., NICOLAS, A. & COLEMAN, G. (eds) Ophiolite Genesis and the Evolution of the Oceanic Lithosphere, Volume 5." Petrology and Structural Geology. Kluwer, Dordrecht, 771-800. KOSTOPOULOS, D. 1989. Geochemistry, petrogenesis and tectonic setting of the Pindos ophiolite, N W Greece. PhD thesis, University of Newcastle upon Tyne. LIATI, A., GEBAUER, D. & FANNING, M. 2004. The age of ophiolitic rocks of the Hellenides Vourinos, Pindos, Crete): first U-Pb ion microprobe (SHRIMP) zircon ages. Chemical Geology, 206, 21-24. MAVRIDES, A. 1980. Apropos de l'age de mise en place tectonique du cort6ge ophiolitique du Vourinos (Gr6ce). In: PANAYIOTOU,A. (ed.) Ophiolites. Proceedings of the International Ophiolite Symposium, Cyprus, 1979, 349-350. Ministry of Agriculture and Natural Resources, Geological Survey Department, Nicosia. MAVRIDES, A., SKOUTSIS-CORONEOU, V. & TSAL1AMONOPOLIS, S. 1979. Contribution to the geology of the subpelagonian zone (Vourinos area, West Macedonia). In: Sixth Colloquium on the Geology of the Aegean Region, Athens, 175-195. Institute of Geology and Mineral Exploration, Athens. MEMOU, G. & SKIANIS,G. 1993. Interpretation ofaeromagnetic data in the Pindos-Vourinos region --Part One." qualitative analysis. Internal Report, Institute of Geology and Mineral Exploration, Athens (in Greek). MOAT, T. 1986. Microfabric and rock deformation studies: competency contrast and its control on the structural behavior of mixed lithological sequences. In: RASSIOS,A., ROBERTS, S. & VACONDIOS,I. (eds) The Application of a Multidisciplinary Concept for Chromite Exploration in the Vourinos Complex (N. Greece). Institute of Geology and Mineral Exploration, Athens, 284-290. MOORES, E. 1969. Petrology and structure of" the Vourinos ophiolite complex, of Northern Greece. Geological Society of America, Special Papers, 118, 1-74.

MOORES, E., KELLOGG, L. & 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. MOUNTRAKIS, D. 1985. Geology of Greece. University Studio Press, Thessaloniki (in Greek).

NAYLOR, M. A. & HARLE, T. J. 1976. Palaeogeographic significance of rocks and structures beneath the Vourinos ophiolite, northern Greece. Journal of the Geological Society, London, 132, 667-675. MERES, M., SEANLE, R. & BOTT, M. 2003. Easter microplate dynamics. Journal of Geophysical Research, 108(4), 12. NICOLAS, A. 1989a. Structures of OphioBtes and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht. NICOLAS, A. 1989b. Analyse pr61iminaire des donn6es structurales du Vourinos. In: RASSIOS, A., KONSTANTOPOULOU,G. & VAKONDIOS,I. (eds) Tectonic Controls on Chrome Ore Localization in Ophiolites. Institute for Mining and Mineral Exploration, Athens, 72-78. NICOLAS, A., BOUDIER, F. & BOUCHEZ,J. 1980. Interpretation of peridotite structures from ophiolitic and oceanic environments. American Journal of Science, 280-A, 192-210. PARSON, L. & WRIGHT, I. 1996. The Lau-HavreTaupo back-arc basin; a southward-propagating, multi-stage evolution from rifting to spreading. Tectonophysics, 263(1~), 1-22. PEARCE, J., LIPPARD, S. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR, B. & HOWELS, M. (eds) Marginal Basin Geology. Blackwell Scientific, Oxford, 77-94. PAVLIDES, S. & CAPUTO, R. 1997. Field Trip to Western Macedonia and Thessaly: Recent and Active Tectonics in Western Macedonia and Thessaly--The Kozani Grevena Earthquake of 1995 (Guidebook). ZITI, Thessaloniki. RAMSEY, J. & HUBER, M. 1983. The Techniques of Modern Structural Geology. Academic Press, New York. RASSIOS, A. 1981. Geology and evolution of the Vourinos complex, northern Greece. PhD thesis, University of California, Davis. RASSlOS, A. 1990. Geology and copper mineralization of the Vrinena area, east Othris ophiolite, Greece. Ofioliti, 15(2), 287-304. RASSIOS, A. 1991. Internal structure and pseudostratigraphy of the Dramala peridotite massif, Pindos mountains, Greece. Bulletin of the Geological Society of Greece, 25(1), 293-305. RASSiOS, A. (ed.) 1994. Tectonic and Geochemical Methods for Chrome Exploration in Ophiolites. Institute for Mining and Mineral Exploration, Athens. RASSlOS, A. 2004. A Geologist's Guide to West Macedonia. The Development Agency of Grevena County, Grevena. RASSIOS, A. & GRIVAS, E. 1999. Geologic and Metallogenetic Map of the Pindos Imbricated Ophiolite and Associated Units (twelve 1:20 000 sheets," ~1000 sq km). Institute of Geology and Mineral Exploration, Athens. RASSIOS, A. & GRIVAS, E. 2001. The Metallogenetic Potential of the PindosIVourinos Ophiolite of West Macedonia. Institute of Geology and Mineral Exploration, Athens (in Greek).

PINDOS-VOURINOS OPHIOLITIC SLAB RASSIOS, A. & KOSTOPOULOS,D. 1990. The geochemistry of dunite and its relation to the position of chromitites in the Vourinos ophiolite complex, Greece. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues," Proceedings" of the Symposium 'Troodos' 87, Nicosia, Cyprus. Geological Survey Department, Nicosia, 593-604. RASSIOS, A. & SMITH, A. 2000. Constraints on the formation and emplacement age of western Greek ophiolites (Vourinos, Pindos 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 Ocean Drilling Program. Geological Society of America, Special Papers, 349, 473-483. RASSIOS, A., BECCALUVA, L., BORTOLOTTI, V., MAVRIDES, A. & MOORES, E. 1983a. The Vourinos ophiolitic complex: a field excursion guidebook. Ofioliti, 8, 275-292. RASSIOS, A., MOORES, E. & GREEN, H., 1983b. Magmatic structure and stratigraphy of the Vourinos ophiolite cumulate zone, northern Greece. Ofioliti, 8(3), 377-410. RASSIOS, A., ROBERTS, S. & VACOND1OS,I. (eds) 1986. The Application of a Multidisciplinary Concept for Chromite Exploration in the Vourinos Complex (N. Greece). Institute of Mining and Mineral Exploration, Athens. RASSIOS, A., KONSTANTOPOULOU,G. & VAKONDIOS,I. (eds) 1991a. Tectonic Controls on Chrome Ore Localization in Ophiolites, Greece. Institute for Mining and Mineral Exploration, Athens. RASSIOS,A., KONSTANTOPOULOU,G., VACONDIOS,I. & RIVAS, E. 1991b. Part 1, synthesis, In: RASSlOS, A., KONSTANTOPOULOU, G. ~:)VACONDIOS, I. (eds) Tectonic Controls on Chrome Ore Localization in Ophiolites, Greece. Institute of Geology and Mineral Exploration, Athens, 1-35. RASSIOS, A., GRIVAS, E., KONSTANTOPOULOU,G. & VACONDIOS, I. 1994. The geometry of structures forming around the ductile-brittle transition in the Vourinos-Pindos-Othris oceanic slab. Bulletin of the Geological Society of Greece, XXX(2), 109-121. ROBERTS, S. 1988. Ophiolite chromite formation: a marginal basin phenomenon. Economic Geology, 83, 1034-1036. ROBERTS, S. 1992. Influence of the partial melting regime on the formation of ophiolitic chromitite. In: PARSON, L. M., MURTON, B. J. & BROWNING,P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 203-217. ROBERTS, S., RASSIOS, A., WRIGHT, t., et al. 1988. Structural controls on the location and form of the Vourinos chromite deposits. In: BOISSONAS, J. & OMENETTO, P. (eds) Mineral Deposits within the European Community, Springer, Berlin, 249-266. 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. 2004. Development of concepts concerning the genesis and emplacement of

265

Tethyan ophiolites in the Eastern Mediterranean and Oman regions. Earth-Science Reviews, 66, 331-387. ROBERTSON, A. H. F. & SHALLO, M. 2000. MesozoicTertiary tectonic evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-254. ROBERTSON, A. H. F. & VARNAVAS,S. P. 1993. The origin of hydrothermal metalliferous sediments associated with the Early Mesozoic Othris and Pindos ophiolites, mainland Greece. Sedimentary Geology, 83, 87-113. ROBERTSON, A. H. F., CLIFT, P., DEGNAN,P. & JONES, G. 1991. Palaeogeographic and palaeotectonic evolution of the Eastern Mediterranean Neotethys. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 289-343. ROBINSON, P., DICK, H., NATLAND,J. & ODP Leg 176 Shipboard Party 2000. Lower oceanic crust formed at an ultra-slow-spreading ridge: Ocean Drilling Program Hole 735B, Southwest Indian Ridge. In: DILEK, Y., MOORES, E. M., ELTHON, D. & NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights fi'om Field Studies and Ocean Drilling Program. Geological Society of America, Special Papers, 349, 75-86. RODDICK, J., CAMERON, W. & SMITH, A. G. 1979. Permo-Triassic and Jurassic Ar-Ar ages from Greek ophiolites and associated rocks. Nature, 279, 788-790. Ross, R., MERCIER, J., AvI~ LALLEMANT,H., CARTER, N. • ZIMMERMAN,J. 1980. The Vourinos ophiolite complex, Greece: the tectonite suite. Tectonophysics, 70, 63-83. SMITH, A. G. 1993. The tectonic significance of the Hellenic-Dinaric ophiolites. In: PRICHARD, H. M., ALABASTER,T., HARRIS, N. B. W. & NEARY, C. R. (eds) Magmatie Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 213-243. SMITH, A. G. & RASSIOS, A. 2003. The evolution of ideas for the origin and emplacement of the western Hellenic ophiolites. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 337-350. SMITH, A. G., WOODCOCK, N. H. & NAYLOR, M. A. 1979. The structural evolution of a Mesozoic continental margin, Othris mountains, Greece. Journal of the Geological Society, London, 136, 589-603. SMITH, A. G., HYNES, A. J., MENZIES, M., NISBET, E. G., PRICE, I., WELLAND, M. J. P. & FERRIERE, J. 1975. The stratigraphy of the Othris mountains, eastern central Greece: a deformed Mesozoic continental margin sequence. Eclogae Geologicae Helvetiae, 68, 463-481. SPRAY, J. G., RODDICK, J. C. 1980. Petrology and 4~ 39mr geochronology of some Hellenic sub-ophiolitic metamorphic rocks. Contributions to Mineralogy and Petrology, 72, 43-55. SPRAY, J. G., BEBIEN,J., REX, D. C. & RODDICK, J. C. 1984. Age constraints on the igneous and metamorphic evolution of the Hellenic-Dinaric ophiolites. In: DIXON, J.E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean.

266

A . H . E . RASSIOS & E. M. MOORES

Geological Society, London, Special Publications, 17, 619-627. THUZIAT, R., WHITECHURCH, H., MONTIGNY, R. & JUTEAU, T. 1981. K-Ar dating of some infraophiolitic metamorphic soles from the Eastern Mediterranean: new evidence for oceanic thrusting before obduction. Earth and Planetary Science Letters, 52, 301-310. THY, P. & DILEK, Y. 2000. Magmatic and tectonic controls on the evolution of oceanic magma chambers at slow-spreading ridges: perspectives from ophiolitic and layered igneous intrusions. 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, 87-104. VRAHATIS, G. & GRIVAS, E. 1980. Geological mapping in Vourinos, 1:10 000 scale. Institute of Geology and Mineral Exploration, Athens. ZIMMERMAN, J. 1968. Structure and petrology of rocks underlying the Vourinos ophiolitic complex, northern Greece. PhD thesis, Princeton University. ZIMMERMAN, J. 1972. Emplacement of the Vourinos ophiolitic complex, northern Greece. Geological Society of America, Memoirs, 132, 225-239.

Cumulates and gabbros in southern Albanian ophiolites: their bearing on regional tectonic setting F. K O L L E R 1, V. H O E C K 2, T. M E I S E L 3, C. I O N E S C U 4, K. O N U Z P & D. G H E G A 5

1Department of Geological Sciences, University of Vienna, Geozentrum, Althanstr. 14, /1-1090 Vienna, Austria 2Department of Geography, Geology and Mineralogy, University of Salzburg, Hellbrunnerstr. 34, A-5020 Salzburg, Austria (e-mail: volker, hoeck@sbg, ac. at) 3Institute of General and Analytical Chemistry, University of Leoben, A-8700 Leoben, Austria 4Department of Geology, Babes-Bolyai University, 1 Kogalniceanu Str., R0-400084 Cluj-Napoca, Romania 5Institute of Geological Research, Blloku Vasil Shanto, Tirana, Albania The western belt of the southern Albanian ophiolites consists of six major ophiolite massifs (Voskopoja, Rehove, Morava, Devolli, Vallamara, Shpati) and two smaller ones (Luniku and Stravaj). Each massif has a distinct sequence of mantle tectonites, ultramafic cumulates (plagioclase-bearing peridotites and wehrlites), cumulate gabbros, troctolites and isotropic gabbros. Voskopoja, Rehove and Morava have predominantly lherzolites as mantle tectonites, Shpati lherzolites and harzburgites, and Devolli and Vallamara almost exclusively harzburgites. A volcanic section together with volcanogenic sediments occurs only in the Voskopoja and Rehove massifs as well as in the smaller Luniku and Stravaj massifs. Whole-rock geochemistry and mineral chemistry suggest a mid-ocean ridge setting for the origin of the cumulates and gabbros from the Voskopoja, Rehove and Morava massifs, with only a minor suprasubduction zone (SSZ) influence. The Shpati massif and the small Luniku massif show mid-ocean ridge (MOR) and SSZ signatures in their plutonic sequences. Cumulates and gabbros from Devolli and Vallamara formed in an SSZ setting. The predominance of MOR-generated crustal rocks and the relatively minor occurrence of SSZ-generated plutonic rocks together with the volcanogenic sediments in the Voskopoja and Rehove massifs are indicative of a back-arc basin origin of the western belt ophiolites above a westward-dipping subduction zone. Abstract:

The Albanian ophiolites are part of a large NNW-SSE-striking ophiolite zone, which comprises the Dinaric ophiolites as well as some Greek ophiolites such as Pindos, Vourinos and Othris. The total length of this ophiolite zone is c. 1000km, from Croatia in the N N W (e.g. Lugovic et al. 1991; Pamid et al. 2002) to Argolis (Greece) in the SSE (e.g. Robertson & Shallo 2000; Bortolotti et al. 2004). Their Jurassic age is constrained by palaeontological evidence from the sediments on top of the ophiolites, by the age of the metamorphic soles and by age determinations on the intrusive plagiogranite (Bortolotti et al. 2004; Dilek et al. 2005, and references therein). The overall setting of these ophiolites in the regional geological framework in Albania has been discussed by earlier workers (Shallo et al. 1990; Shallo 1992, 1994; Frasheri et al. 1996; Meco & Aliaj 2000; Robertson & Shallo 2000; Bortolotti et al. 2004) and is shown in Figure 1,

which follows the tectonic zones outlined by Meco & Aliaj (2000). The tectonic zones located west and N W of the Albanian (Mirdita) ophiolites comprise a westward-directed stack of thrust sheets related to the Apulian continental platform. The Sazani and the Kruja zones represent the carbonate platform, and the Ionian zone an intra-continental rift area (Robertson & Shallo 2000). The Krasta (Cukali) zone (Pindos zone in Greece) is the deep-water passive margin of the Apulian platform. It is, in turn, overthrust by the Albanian Alps, the Vermoshi and Gashi zones, platform-related units to the east of the Krasta zone (Meco & Aliaj 2000; Robertson & Shallo 2000). The Korabi zone (Pelagonian zone in Greece), east of the ophiolites, represents a continental fragment comprising a pre-Alpine basement and a Triassic-Jurassic sedimentary cover. The Albanian ophiolites form a link between the Greek and the Dinaric ophiolites. An

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 267-299. 0305-8719106l$15.00 9 The Geological Society of London 2006.

268

F. KOLLER E T AL.

Fig. 1. Overview of the geology of Albania, according to Meco & Aliaj (2000). The various zones are discussed in the text. understanding of their origin is essential to restore the continent-ocean distribution in the Jurassic in the Eastern Mediterranean realm and the possible sites of subduction zones and related thrusts. As outlined below, the setting of formation of these ophiolites and their mode of emplacement is still controversial. The Mirdita zone, as delineated in Figure 1, is commonly divided into two ophiolite belts, an eastern one and a western one (e.g. Shallo et al. 1990; Shallo 1992, 1994). The western belt is supposed to consist of predominantly mid-ocean ridge (MOR)-type ophiolites with lherzolites, troctolites, gabbros (Cortesogno et al. 1998) and MOR-type basalts (MORB). By contrast, the eastern belt is characterized by harzburgites,

wehrlites, gabbronorites, clinopyroxene gabbros, plagiogranites and volcanic rocks. These last show a wide range of suprasubduction zone (SSZ) compositions from basalts to andesites, dacites and rhyolites (Shallo 1992, 1994; Beccaluva et al. 1994a, b; Bortolotti et al. 1996, 2002, 2004; Robertson & Shallo 2000). Existing models explaining the formation of these contrasting ophiolites infer the presence of intra-oceanic (incipient) subduction zones, involving either east-dipping (Beccaluva et al. 1994b; Shallo 1994; Bortolotti et al. 2002) or west-dipping subduction (Insergueix-Fillippi et al. 2000; Robertson & Shallo 2000; Shallo & Dilek 2003; Dilek et al. 2005). Based on structural arguments, Nicolas et al. (1999) presented a

SOUTHERN ALBANIAN OPHIOLITES different concept for the northern Albanian ophiolites, in which both ophiolite belts formed in a slow-spreading MOR setting. The eastern belt would reflect magmatic episodes, whereas, the western belt reflects amagmatic episodes. The mantle tectonites in both belts were originally harzburgites, but in the western belt they were refertilized by basaltic melts to form lherzolites at shallow levels. Recently, Hoeck & Koller (1999), Hoeck et al. (2002) and Bortolotti et al. (2002) showed that basalts with an SSZ signature also occur in the western belt interlayered with MORB. In this paper we present evidence that in southern Albania, at least, the SSZ influence known in the western belt is not restricted to the volcanic suite but is also well recorded in the ultramafic tectonites, ultramafic, mafic cumulates, and gabbros. Based on the close spatial and temporal relationship of MOR and SSZ magmas and cumulates, we advocate a back-arc origin for the Albanian ophiolites.

Geological setting The ophiolites in southern Albania (see ISPGJF G J M - I G J N 1983: Geological Map of Albania) are confned to an elongate area ranging from SE of Tirana to the Greek border (Fig. 1). The detailed locations of the ophiolitic massifs are shown in Figure 2. They consist of several units belonging to either the western or the eastern belt. The two belts are separated by the Palaeogene and Neogene molasse sediments of the Neohellenic or Albanian-Thessalian trough (Shallo 1992; Robertson & Shallo 2000; Hoeck et al. 2002). The western belt units (Fig. 2) are, from south to north, the massifs of Voskopoja, Rehove and Morava, below referred to as VRM. The Voskopja and Morava massifs are separated by the north-south-striking Neogene to Quaternary basin of Korce. The Rehove massif is separated from the Morava and Voskopoja massifs by a serpentinite m61ange zone, which extends beneath the last two complexes. Towards the north, the Voskopoja massif is followed by the Devolli and Vallamara massifs (referred to as DV below) and furthermore by the Shpati massif with a spur towards the north. Two additional small massifs, Stravaj, east of Shpati, and, further north, the Luniku massif, crop out as windows beneath the Cenozoic molasse sediments between the two belts (Fig. 2). These are the only massifs containing volcanic rocks apart from Voskopoja and Rehove. A smaller ophiolite body called Erseka, occurring in

269

the south close to the Greek border, is excluded from this study. The eastern ophiolite belt is represented by the Shebenik massif, which extends from the southern tip of Lake Ohrid towards the north. It is followed further south by the small harzburgitic massifs of Bilisht, south of Lake Prespa (Fig. 2). Five columnar sections from the western belt including the massifs of Morava, Rehove, Voskopoja, Devolli + Vallamara and Shpati are shown in Figure 3a and b, and characterize the wide lithological variation of these ophiolite bodies. We will discuss first the southern three profiles: Voskopoja, Rehove and Morava (VRM) (Fig. 3a) and subsequently the northern ones: Devolli, Vallamara (DV) and Shpati (Fig. 3b). The total thickness of the ophiolites is indicated in the columnar sections. Voskopoja, R e h o v e a n d M o r a v a sections

Apart from the tectonically underlying m61ange, each of these ophiolite sections (Fig. 3a) starts with lherzolites, interlayered with minor harzburgites and rare dunites. Two of the three sections, i.e. Voskopoja and Morava, include a thrust unit within the ultramafic mantle tectonites (the half-arrows in Fig. 3a). A thin layer of amphibolites and metasediments separates a lower thrust unit, consisting only of mantle tectonites, from an upper thrust unit showing a continuous succession with mantle tectonites, cumulates, gabbros and extrusives (only in Voskopoja and Rehove). In Voskopoja, lherzolite of the structurally upper unit contains dykes of completely rodingitized troctolites and rare basalts. In the Rehove massif, as well in the upper thrust units of the Voskopoja and Morava sections, the ultramafic cumulates commonly include plagioclase wehrlites and minor plagioclase lherzolites followed by cumulate gabbros and troctolites. There are some differences between the three massifs. For example, intrusions of gabbronorites were found only in Morava. In the Rehove section, the ultramafic cumulates are overlain by cumulate gabbros and troctolites. Isotropic gabbros are relatively frequent in Rehove, but rare in Voskopoja and completely missing in Morava. Individual basaltic dykes in the gabbros are widespread in Rehove, rare in Voskopoja and absent in Morava. Sheeted dykes may have been originally present but are not now preserved, except for some remnants occurring together with pillow lavas, only in the Rehove breccias (see below). The ophiolite sequence in Morava ends with the plutonic sequence. Massive basalts are observed

270

F. KOLLER E T AL.

Fig. 2. Sketch map of the south Albanian ophiolite bodies with the sample locations from the individual ophiolite massifs.

SOUTHERN ALBANIAN OPHIOLITES

271

Fig. 3. (a) Columnar sections for the Morava, Rehove and Voskopoja massifs based on Hoeck et al. (2002). (b) Columnar sections for the Devolli + Vallamara and Shpati massifs. The half-arrows indicate thrust planes. in Rehove as well as in Voskopoja, directly overlying the gabbros, but the most conspicuous and widespread rock types of the extrusive sequence are volcaniclastic sediments. These start with breccias and often finish with volcaniclastic sandstones. (For further details of the extrusive section and the sediments above, see Hoeck et al. (2002).)

DevollL Vallamara and Shpati sections The Devolli and Vallamara ophioite sections are uniform in their lithology, thus only one

columnar section is shown for both (Fig. 3b). In contrast to the western ophiolite bodies they comprise harzburgites but no lherzolites. As in other places, mantle tectonites are underlain tectonically by a m61ange containing sediments, serpentinites, minor amphibolites and pillow basalts. Above the tectonic contact follows fresh harzburgite of c. 1500-1800mthickness, serpentinized only along small fracture zones. Orthopyroxenites, otherwise rare in the western belt, occur frequently as dykes or sills. Dunite lenses are restricted to the upper part of the harzburgites; these are overlain by

272

F. KOLLER E T AL.

plagioclase-bearing peridotites (dunite to lherzolites), representing ultramafic cumulates. In the upper part of the ultramafic cumulates thin gabbro layers (centimetres to decimetres thick) and a few gabbro dykes were observed. A thin cover of gabbros and troctolites, several tens of metres thick, ends the section (Fig. 3b). Extrusive and dyke sections are entirely missing. All of the volcanic rocks mapped in the vicinity of the DV ophiolite body are part of the m61ange. The Shpati massif further north (Fig. 3b) has a section that is intermediate between those of Voskopoja and Rehove on the one hand and Devolli and Vallamara on the other, as it contains a considerable amount oflherzolite. The profile again starts with a m61ange with sediments, gabbros and basalts, as well as amphibolites. A metamorphic sole is missing. Tectonically above the m61ange are harzburgites with some dunite lenses in the upper part, followed by lherzolites with occasionally websterite dykes and sills. The plagioclase-bearing ultramafic cumulates above the lherzolite contain some layers of troctolites and gabbros in the higher part. Similar to Devolli and Vallamara, the ophiolite section ends with gabbros and rare troctolites. Again, volcanic rocks are missing from the top of the section.

Analytical methods Geochemical data for the cumulates and gabbros were obtained for 128 samples collected from all of the massifs. These samples comprise a wide variety of lithological types and were taken randomly. After crushing in a metal jaw crusher all samples were ground in an agate mill. Major and trace elements were analysed by X-ray fluorescence (XRF) using a Philips PW 2400 at the Department of Geological Sciences, University of Vienna. For major elements a lithium borate melt pellet and for the trace elements a pressed powder pellet were used. The loss on ignition (LOI) was determined by heating in a furnace at 1000 ~ for 3 h. Rare earth elements (REE) and Th, Y, Zr and Nb were analysed by inductively coupled plasma-mass spectrometry (ICP-MS) after microwave assisted (Multiwave Perkin Elmer-Anton Paar) acid digestion with HNO3-HC1-HF. REE concentrations were determined with external calibration, using a sector field double focusing ICP-MS Element (ThermoFinnigan) system at the Institute of General and Analytical Chemistry, University of Leoben. Some samples were analysed after sodium peroxide sintering with a quadrupole ICP-MS system (at the Institute of General

and Analytical Chemistry, University of Leoben) following Meisel et al. (2002). Mineral analyses were carried out on polished thin sections with a JEOL 8600 electron microprobe including a Link control system (at the Department of Geography, Geology and Mineralogy, University of Salzburg). Measurement conditions were 15 kV acceleration voltages and 15 nA beam current. For the quantitative analyses synthetic and natural minerals were used as standards. The correction procedure included the background, dead time and a Z A F calculation built into the Link system.

Petrology of the plutonic section The main petrographic rock types are ultramafic cumulates that comprise a wide range of compositions including plagioclase dunites, plagioclase peridotites, wehrlites and pyroxenites. They are mostly serpentinized to a variable degree. Where the serpentinization is severe, it is often difficult to differentiate cumulates from the mantle tectonites, refertilized harzburgites or lherzolites. We distinguished the ultramafic cumulates mainly on the basis of visible centimetre- to decimetre-scale layering (Fig. 4) and magmatic textures. In general, olivine is the first liquidus phase to crystallize; it is often rounded (Fig. 5) and forms small layers in the range of millimetre to centimetre thickness. Very often olivine is partly or completely transformed to serpentine minerals. Cr-spinel forms small, brown, isotropic grains, often showing magmatic resorption. Both olivine and spinel are overgrown by clinopyroxene. When orthopyroxene is present, similar relations are found. Both pyroxenes are concentrated in layers alternating with olivine-rich

Fig. 4. Layering of an ultramafic-mafic cumulate sequence (Morava massif), with layer thickness ranging from 2 and 10 cm.

SOUTHERN ALBANIAN OPHIOLITES

273

Fig. 6. Thin section of a layered olivine-clinopyroxene gabbro (Voskopoja massif) with olivine (O1), plagioclase (Plag) and clinopyroxene (Cpx). Olivine is rimmed by clinopyroxene (Cpx rim); crossed Nicols. Fig. 5. Back-scattered electron image of a plagioclase-bearing wehrlite from the Rehove massif. O1, olivine; Cpx, clinopyroxene; Serp +Amp, serpentine and amphibole; AP, altered plagioclase.

layers. Occasionally, clinopyroxene is zoned and shows fine exsolution lamellae of orthopyroxene and, rarely, spinel. Plagioclase is the latest phase to appear, often being associated with a brown amphibole of magmatic(?), origin, with a pargasitic composition. Plagioclase is rarely preserved and was mostly transformed to low-temperature alteration products such as prehnite or hydrogrossular (hibschite to katoite). The dominant sequence of crystallization is olivine + spinel-clinopyroxene/orthopyroxeneplagioclase. Troctolites and cumulate gabbros differ from the ultramafic cumulates, as plagioclase becomes more dominant and orthopyroxene disappears except for occasional grains. As in the clinopyroxene gabbros, olivine is often rimmed by clinopyroxene, as shown in Figure 6. Olivine and plagioclase constitute troctolites, but with additional clinopyroxene they form cumulate gabbros. Layering is often visible (Fig. 4). There is a transition from ultramafic cumulates to cumulate gabbros and troctolites. As in the ultramafic cumulates, olivine and spinel are often found as inclusions in plagioclase. Inclusions of plagioclase in olivine and clinopyroxene are found as well. Olivine is strongly serpentinized; plagioclase is transformed into low-temperature alteration products, and clinopyroxene is changed to amphiboles. In general, olivine is the first mineral phase to appear, followed by plagioclase and finally clinopyroxene.

The isotropic gabbros and gabbronorites are mostly intrusive and form dykes and stocks in some mafic and ultramafic cumulates, as well as in lherzolites and harzburgites (metres to decametre in size). Particularly in the latter, chloritic intervals may develop at gabbroultramafic boundaries. Mineralogically, the isotropic gabbros consist mainly of clinopyroxene and plagioclase, with subordinate olivine. With the exception of gabbronorites in Morava (Fig. 7), orthopyroxene is commonly found, in minor quantities, in the DV and Shpati massifs. The sequence of crystallization is clinopyroxeneorthopyroxene-plagioclase. Most of clinopyroxene is changed to green amphibole with a composition varying from actinolite to magnesiohornblende. Brown magmatic(?) pargasitic amphibole is very rarely present. Plagioclase is also altered; where preserved it is weakly zoned

Fig. 7. Thin section of a gabbronorite (Morava massif) with clinopyroxene (Cpx), orthopyroxene (Opx) and plagioclase (Plag); crossed Nicols.

274

F. KOLLER ET AL.

with rims slightly lower in An content than cores. An-poor plagioclase is found only in the vicinity of the secondary amphiboles. Spinels are generally absent; instead, ilmenite and titanomagnetite frequently occur. Isotropic gabbros and gabbronorites reveal the presence of lowtemperature alteration products such as albite, prehnite, chlorite and zeolite minerals. Quite commonly, the gabbroic lenses and dykes in the ultramafic rocks and ultramafic cumulates show signs of rodingitization such as increasing Ca content. In these cases, plagioclase is normally replaced by hydrogrossular and prehnite. These oceanic, post-magmatic alteration processes affect mainly the large ion lithophile elements (LILE) and do not influence the high field strength elements (HFSE); this is important for classification.

is completely altered to amphibole (actinolite to magnesiohornblende). As for orthopyroxene, the XMg is usually very high, with little variation (0.86-0.93). Lower values are found only in the pigeonite-bearing gabbros and in an orthopyroxene-free cumulate gabbro from Rehove (0.765-0.824). All clinopyroxenes from the ultramafic cumulates classify as diopsides, whereas those from the cumulate and isotropic gabbros are diopsides or augites. They differ from clinopyroxene from the ultramafic cumulates by the enrichment in ferrosilite component and a slight depletion in the wollastonite end-member. There is also a significant change in the composition of clinopyroxene in the isotropic gabbros for minor elements such as A1203, TiO2, Na20 and Cr. A covariation of Na and Ti with XMgis displayed in Figures 9 and 10. Differences can be observed (1) between the ultramafic cumulates and the mafic rocks from the same massif Mineral chemistry and (2) between the cumulates and gabbros from different massifs. Olivine (Table 1) Several trends can be recognized in the All olivines of the ultramafic and mafic cumu- covariation of Ti, Na and Cr (not shown) with lates exhibit a forsterite content of > 0.80. In the XMg in clinopyroxene. Ti and Na are negatively samples from Voskopoja, Rehove and Morava correlated and Cr is positively correlated with they show an evolution trend with decreasing XMg. At a given concentration of these elements, XMg from ultramafic cumulates (0.895-0.881) to the XMg of clinopyroxene from the ultramafic cumulate gabbros and troctolites (0.860-0.845). cumulates is normally higher than for the Only one sample of an isotropic gabbro contains cumulate and isotropic gabbros. The high values olivine with XMg as low as 0.818 (Fig. 8). MnO of XMg in each group of clinopyroxene reflect the increases from 0.1 wt% in the ultramafic cumu- XMgof the whole rock. lates to 0.3 wt% in the gabbros; the NiO content The ultramafic rocks of DV, Shpati and VRM decreases from 0.31 wt% in the ultramafic define a steep trend with high variation of Na and cumulates to 0.09 wt% in the gabbros. Olivines Ti, but little decrease in XMg. The strong enrichfrom the ultramafic cumulates from Devolli ment trend of Ti and, in particular, Na in the and Shpati have higher XMg values, in the range ultramafic cumulates from all units is illustrated of 0.900-0.897. Gabbros from Shpati, with XMg by a series of samples without a large systematic of 0.892, show higher values than the troctolites variation within these samples. No dependence of the same unit (XMg0.875). between Na and Ti content in the whole rock and the content of these elements in the clinopyroxene Orthopyroxene (Table 2) is discernible. By contrast, the cumulate and This is represented mostly by enstatite. The XMg isotropic gabbro trends mostly show a larger varies from 0.845 in an orthopyroxene from a covariation of XMg with Ti and Na. Most of the gabbro from Devolli to 0.907 in an ultramafic samples form small patches with little internal cumulate also from Devolli and in a cumulate variation of Ti, Na and XMg. Only two samples, gabbro from Vallamara. The A1203 content is one gabbro from Shpati and one gabbro from more variable, ranging from 0.5 to 2.2 wt% in Luniku, exhibit perceptible trends with decreasa gabbro from Morava. CaO ranges from 0.6 to ing XMg and increasing Ti and Na from the cores 1.4 wt%. A gabbro from Shpati contains pigeo- to the rims. Another gabbro (as well as troctolites nite with XMg <0.80, CaO from 2.1 to 3.4 wt% from Shpati) shows trends of high Na and XMg and an A1203 content of 0.4 wt%, significantly clinopyroxenes, comparable with the ultramafic rocks. lower compared with enstatite.

Clinopyroxene (Table 3)

Spinel (Table 4)

This is the most widespread mineral in all of the cumulates and gabbros, except where it

Cr-spinel is mostly restricted to the ultramafic cumulates, troctolites and a very few cumulate

SOUTHERN ALBANIAN OPHIOLITES

275

.,o o o 0

~RR

.~E~RR~

RRR

.~RRRR

o

.,.,. |

r..)

o

k) ,..,._.

o

o

"~

~~ o

~

~.~'~ o

276

F. KOLLER ETAL.

100 _ ~ -

\

Turkish S S Z / ~ " ~ " ~ " -

ARC gabbros

~r/--

/

I

.,,.

80 O~ t<

70

UM cumulate Cumulate gabbro 9 -Troctol,te Gabbronorite Isotr~ gabbr~

9 .s*" uman j ~~ I:lrowrllng ~ 1 r 9 r~ n g / /.~ _l ( ) / ~ (

60

UM . . . . late -Troctolite

t ~

/

U,M,cu,rnu,late,

0.5

I,

0.6

,i

]

w ~

~

_ ~ O

/ ~

7 ~

~m~ Qa~tes^,

I , ~

0.7

AI=

~=

~

%"%. , / s "/ . . . . . . "~'" Oceanic

Gabbro A Oevolli, Vallamara

50

2a.bb._r .~ /;"I (1984)../~r

/ I /

Beard (1986) J .... I , , , L----'~,,,,, I

0.8

0.9

1.0

Fo m o l e (%) Fig. 8. Fo v. An relations in olivine and plagioclase, respectively. The arc gabbros and oceanic cumulates fields are after Beard (1986); the Turkish SSZ ophiolites field is after Parlak et al. (1996, 2000, 2002), and the Troodos ophiolite field is after H6bert & Laurent (1990).

gabbros. Only one isotropic gabbro from Shpati contains Cr-spinel. Spinels from wehrlites in Rehove (Fig. 11) are distinguished by their high Cr-number (0.55-0.62), low Mg-number (0.3-0.5) and high TiO2 contents (>0.8 wt%). The inset in Figure 11 shows the TiO2-A1203 relations (Kamenetsky et al. 2001), in which this group overlaps with ocean island basalts. All other spinels show a Cr-number of 0.32-0.55 and an Mg-number between 0.50 and 0.63. In the TiOz-A1203 diagram they plot in the MOR field. In particular, Cr-spinels from the Devolli ultramafic cumulates show the highest Mg-number (c. 0.63), and the lowest Cr-number (0.32) and TiO2 contents (<0.2 wt%). Many spinels from the DV and Shpati ultramafic cumulates and a few from VRM show similar Cr-number, Mg-number and TiO2 contents to the spinels from the mantle harzburgites and lherzolites (Fig. 11). Spinels from the mantle tectonites form a linear array with negatively correlated Cr-number and Mg-number. Harzburgites from Devolli, together with some from VRM, show the highest Cr-number; lower Cr-numbers are shown by the mantle tectonites from VRM and Shpati. Plagioclase (Table 5)

The An content in the preserved plagioclases ranges from Ans0 in an isotropic gabbro from Shpati to over Ang7 in an ultramafic cumulate from Devolli (Figs 8 and 12). In some samples the variability is much smaller. One sample

from the ultramafic cumulates in Morava shows An from 81 to 82, without any significant zoning. Plagioclase with the highest An content of 97 in the core also shows only minor zoning, with a value of 94 at the rim. Plagioclases from the cumulate gabbros, troctolites and isotropic gabbros are commonly zoned, with An-rich cores and An-poor rims. In a cumulate gabbro from Rehove, a plagioclase ranges from Any7in the core to An58 at the rim. The lowest An content comes from a gabbro at Shpati, with An55 in the core and Ans05 at the rim. The overall variability is shown in Figure 12. The variation along the An axis in each group illustrates the magmatic zoning between core (high An) and rim (low An).

Geochemistry Ultramafic cumulates

Analyses of the ultramafic cumulates (plagioclase-bearing dunites, lherzolites and wehrlites) are listed in Table 6. They have a variable composition, which mainly reflects the amount of olivine and plagioclase present. Orthopyroxene and clinopyroxene play a relatively minor role in the geochemical variability (Figs 13a, b and 14a, b). Spinels with a wide range of Crnumber and Mg-number are common accessory phases. The variability of MgO, A1203 and CaO is displayed in Figures 13a, b and 14a, b. The A120 3 content is generally < 4 wt%; only in Rehove

SOUTHERN

o~.~

ALBANIAN

OPHIOLITES

277

o m

~

o

-}

2

o

0

,.o

~ ~ ~ 1 7 6

e~

.~~176

~.

~ 1 7 6 1 7 6 1 7 6. ~ 1 7 . 6 1 7 6

278

F. K O L L E R

E T AL.

o ~

o

~

o

~

o

0

~

o

o

o

~ 0

o

0

0

o

0

o

o

0

o

0

~

~ o

o

o

o

~ o ~ o

0 o

0 o

0 0 0 0 ~ 0 0 ~ o o 0 o o o o o ~

~ o

o

o

o

~ o o 0

o

~

~

o

o

~

o

o

o

o

0

~

0

0

0

g

o

0

~ ~

o o

~ ~

o

o ~

o 0

~ 0

o o

~

~ 0

~

o

.o

~

~

o

0

~

o

~

o

o

o

o

~

o

o

o

o0 r

~0

o

0

O~

o

~o

0

d ~ M ~ d o ~ o d d N

~d

d d d d d d o o o o ~

o

0 0

0

0

~

0

0

~

0

0

0

d d d d d d d ~ d d ~

d ,.Q

m d d d d d d d d d d ~

-9 ~ ~

~ ~

~

~

~

.

d

o

SOUTHERN ALBANIAN OPHIOLITES

0.10

-''

'

'

'''

~_ C px .

''

I ' ' ' ' ' '

'''

9

I

'

''

'

''

279

''

'1

'

''

U M cumulate

9

Cumulate gabbro

--

Jl)

--

e~_

Troctolite Gabbllmodm Iootropic g a b b r o

0.08 ~

<> 0

:

Dew)ill, Vallamara UM oumulme

9

Z_

~

r-q-

=-

0.06

' ' ' ' '

VOskopoJa, Rohove, Morava

Sh~

i~

, m

I0.04

-

-

U M cumulallo Troctoilte Gabbro

/ ~

& /t A

"

Lunlku

-

0.02

0.00

1.0

0.9

0.8

0.7

0.6

XMg Fig. 9. XMgv. Ti relations for clinopyroxene from the south Albanian ophiolites.

0.05

i

i

i

I

i

l

i

i

I

i

I

'

~

'

i

i

i

'

I

i

I

R

'

i

'

C px

i

i

i

l

i

i

i

I

'

I

i

l

1

VoskopoJa, Rehove, Morav~ UM cumu~te 9 Cumulate g a b l x o Troclollte Gabbronod~ I ~ ~

0.04 :

4

A A

0.03

~_ 0 o

/x

G]

- -

I I -T-atN~

z

&zx /x

0.02

I)evolli, V a l l a m a r a U M cumulate ~btxo Shpati UM cumula~

+ il

0.01

0 J

0.00

i

1.0

i

i

i

i

!

i

9 Z [ ] -,k

:roc~.o Gebbm

A

Lunlku i

i

l

!

i

!

,

i

i

0.9

i

i

,

I

i

0.8

l

l

t

I

I

I

I

I

I

0.7

I

I

I

I

I

I

I

I

I

0.6

XMg

Fig. 10. XMgv. Na relations for clinopyroxene from the south Albanian ophiolites.

does it exceed 10 wt%. At low A1203 concentrations, the TiO2 content is positively correlated with A1203 in the range between 0.02 and 0.15 wt%. No correlation is observable at higher A1203 contents, where TiO2 ranges up to 0.4 wt%. The XMg expressed as MgO/(MgO + FeOtot) is high, at 0.86-0.92 (Fig. 15).

The composition of the ultramafic cumulates with low A1203 overlaps considerably with that of harzburgites and lherzolites in terms of both major elements and trace elements, such as Cr, Ni or Co. There seems to be only a slight difference in incompatible trace elements (e.g. Y or Zr), in that at least one of these (mostly both) is higher

280

F. KOLLER E T AL.

o

;D

o

~ ~

~ ~

-

-

~ . ~

~ ~ o o o ~ ~o~ .~oo~

~ ~

~

ch

+

%

2

o

0

g.~

o

SOUTHERN ALBANIAN OPHIOLITES 10

. . . .

I

281

. . . .

I

. . . .

I

. . . .

i

. . . .

0.8

0.6 . , . , . "

OO, o . . . .

/~zJ ~

<

L

I

L

....

~

,o . . . .

....

,'o . . . .

,o

Wt%AI203

0.4

v

0 0.2

- Gabbm t , ,

~ Troctdite i

I , ,

I,

I

,,

i,

I

I

I

,

J

I

,

,

,

,

0.0

0.3

0.4

0.5

0.6

0.7

0.8

Mg/(Mg+Fe) Fig. 11. J(Mgv. Cr-number and A1203 v. TiO2 (inset) relations for spinel from the south Albanian ophiolites.

in the cumulates. At present, REE data exist only for the VRM ultramafic cumulates and for one from Shpati. The cumulates exhibit strongly light REE (LREE)-depleted patterns with a chondrite-normalized (Ce/Sm)N ratio of 0.3-0.4 (Fig. 16a). In contrast, the mantle tectonites, as with the lherzolites and harzburgites from VRM and DV, are enriched in the LREE relative to the middle REE (MREE); thus, they show mainly U- or spoon-shaped patterns.

Cumulate gabbros and troctolites Analyses of cumulate gabbros are presented in Table 7. Together with the ultramafic cumulates, the cumulate gabbros and troctolites form linear arrays in the A1203 v. MgO and CaO v. MgO diagrams (Figs 13a, b and 14a, b). These arrays parallel a field between the Fo-rich olivine (Fo85_90) and An-rich plagioclase (An80_95). As for the ultramafic cumulates, clinopyroxene plays only a minor role in the chemical composition of the cumulate gabbros and troctolites. In general, the cumulate gabbros and troctolites are more A1203 rich (11-16 wt%) and CaO rich (8-15

wt%), but MgO poor (8-20 wt%) compared with the ultramafic cumulates. However, there is an overlap at high MgO and low A1203 and CaO contents. Na20 is considerably higher and varies from almost 0 to 2.5 wt%; high values for K20 (0.30 wt%) are also sometimes observed. The enrichment is probably due to sea-floor alteration. TiO2 is, in general, below 0.35 wt%, and the XMg is variable between 0.70 and 0.86 (Fig. 15). The compatible trace elements such as Cr or Ni show significant variations up to 2000 and 1400 ppm, respectively. The incompatible trace elements contents are very low (Zr < 10 ppm, Y < 6 ppm, V < 80 ppm). The chondritenormalized REE patterns are generally flat, with an overall enrichment factor of 1-4; a few are slightly depleted in LREE (Fig. 16b). The pronounced positive Eu anomaly is due to plagioclase accumulation.

Isotropic gabbros and gabbronorites Representative geochemical analyses are shown in Table 7. Gabbros and gabbronorites plot in the A1203 v. MgO diagram in a field that overlaps

282

F. K O L L E R E T AL.

~,.,,~ o

~

. 0 0 0 ~ ~

~

~

0 o

t~

0 0 ~

0

0

0

0

~

0

~

~

0

0

0

0

~

0

0

0

~

0

0

~

~

0

~

0

0

~

0

0

~

0 . .,...~

i

P:,

0

0

~

0

0

0

~

0

d ~ d o d ~ d d o

M~dddo

~ d d d ~ s

g~dddd

0

0

0

0 0 e~

M

0 ~ o o 0

~

0

0

0

~

0

~

~

0

0

0

0 oo

t~ ~

0

0

0

~

0

~

~

0

0

0

0

0 o

0

0

0

0

o

0

0

0

~ o o . ~ o ~ c~

[,,.

o

SOUTHERN ALBANIAN OPHIOLITES 1.0

-'''''''''1''~''''"1'''''''''

283

I'''''''''1'''"''''1

'''

''''''-

0.9 x O_ O c"=if= O~

0.8

0.7

/

~Ub~r~ J

~

../Leg .. 37 /I~

/

0.6

Voskopoja, Rehove, Morava

0.5

Shpati

UM cumulate 9 _ Cumulate gabbro ~ Troctolite ~ Gabbronorite Isotropic gabbro

0.4

/~ /~RabCbro~U ~....~/ ./cumulates/ ~'~'~ / / Turkish'SSZ Ophiolites

~

-

)

40

II

III

II

I

I

I

50

UM cumulate Troctolite Gabbro

II

Devolli,

+

UM cumulate Gabbro

Luniku

0 o I(

9 A z~

Isotropic gabbro I

III

i

I

I I I

60

II

I

In

i

I

I

I

I

III

70

I

P ]

I

80

In

l}

I

I

I

I

Vallamara

[

li~

90

9 [] II

III

I

100

An (%) Plag Fig. 12. An v. XMgfor clinopyroxene from the south Albanian ophiolites. The field of MOR gabbros is after Ross & Elthon (1993), the field of ARC gabbros is after Burns (1985), the fields of Turkish SSZ ophiolites are after Parlak et al. (1996, 2000, 2002) and the Leg 37 field is after H6bert et al. (1989).

only to a small extent the cumulate gabbros but deviates from the cumulate array towards lower contents of MgO and A1203. This is due to the dominant role of clinopyroxene, amphibole and (partly) orthopyroxene. A similar distribution is observed in the CaO v. MgO diagram (Figs 13a, b and 14a, b). Whereas in VRM, gabbros plot in a comparatively narrow field, the gabbros from Shpati, Devolli, Vallamara and Luniku show a much wider variation in their MgO, A1203 and CaO contents (Figs 13a, b and 14a, b). In particular, the Shpati gabbros show a high variability in CaO values; the high CaO content (up to 20 wt%) in some samples combined with low A1203 ( < 1 0 wt%) is probably due to rodingitization processes. The compatible elements such as Cr ( < 1000 ppm) or Ni ( < 600 ppm) are relatively low in the isotropic gabbros compared with the cumulate gabbros (up to 3000 ppm Cr and 2000 ppm Ni). The content of incompatible elements is generally low, with Zr < 30 ppm, Y <20 ppm and TiO2 < 0.4%. Nb is also low (1-3 ppm). In VRM and a few Shpati gabbros these elements are enriched, with Zr up to 110 ppm, Y up to 30 ppm and TiO2 > 2%. This is also valid for V, which may reach 300 ppm. The occasionally elevated values of K20 and Na20 are probably due to metasomatic

processes during oceanic metamorphism. The REE patterns of two gabbros, one from VRM (Fig. 16a) and one from Shpati (Fig. 16b) are 10 times chondrite-enriched and display a slight upward-convex pattern. The REE patterns are not distinguishable from those of the MORB from Voskopoja or Rehove, described by Hoeck et al. (2002). Assuming the isotropic gabbros represent melts rather than cumulates, the most evolved gabbros match the MORB melts. Gabbronorites are restricted to Morava, as small intrusive bodies in the cumulate gabbros. Their geochemistry falls within the range of the isotropic gabbros. They are similar to the gabbros, being high in CaO, MgO and A1203, but low in Na20 and K20, and very low in TiO2. Cr and Ni contents are in the normal range for the gabbros, but Nb, Zr and Y are extremely low.

Clinopyroxene geothermobarometry There are a number of methods and calculations for estimating the temperatures and pressures of crystallization of basalts based on phenocrystmelt equilibria (Beattie 1993; Danyushevsky et al. 1996; Putirka et al. 1996, 2003; Yang et al. 1996; M~trich & Rutherford 1998; Putirka 1999). All these thermobarometers require a knowledge of the composition of the phenocrysts

F. K O L L E R ET AL.

284

Table 6. Representative geochemical data for the ultramafic cumulate rocks of the southern Albanian ophiolites Sample: A99/020 A01/226 A991054 A l b l / 9 8 A03/343 A03/347 A03/372 A03/353 A02/319 A03/430 Massif: Voskopoja Rehove Morava Devolli Vallamara Shpati

Rock:

Plag-Per

PlagDunite

SiO2 TiO2 A1203 Fe203 MnO MgO CaO Na20 KzO P205 LOI Total

36.42 0.04 4.31 9.00 0.12 32.89 3.85 0.04 0.02 0.02 13.67 100.39

37.47 0.03 1.96 9.69 0.09 36.38 0.10 0.00 0.01 0.02 13.51 99.26

XMg Nb Zr Y Sr Rb Ga Zn Cu Ni Co Sc Cr V Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Th

0.879 b.d. 0.6 0.3 9.7 1.2 b.d. 48.3 44.1 1725.7 106.5 4.0 2081.1 14.2 23.5 0.022 0.004 0.027 0.028 0.027 0.007 0.001 0.010 0.004 0.008 0.001 0.007 0.004

0.002

0.881 0.1 b.d. b.d. 3.4 0.1 3.2 41.0 62.4 2277.2 119.3 3.1 2633.0 19.2 62.4

PlagPlagWehrlite Wehrlite Dunite 39.47 0.19 4.51 8.77 0.13 34.71 2.50 0.05 0.02 0.02 9.32 99.68 0.887 0.1 3.5 3.4 10.0 1.0 b.d. 48.3 24.3 1940.8 100.4 11.0 3462.0 57.1 15.0 0.113 0.337 0.070 0.515 0.265 0.129 0.464 0.084 0.615 0.138 0.437 0.063 0.423 0.065 0.177 0.058 0.001

38.45 0.13 7.73 8.41 0.12 31.45 4.02 0.16 0.11 0.02 9.69 99.60

41.23 0.15 3.35 8.28 0.12 35.41 2.67 0.10 0.01 0.02 8.33 99.66

0.881 b.d. b.d. 2.5 7.6 b.d. 2.9 42.5 18.5 1546.4 4.9 1825.8 31.9 17.8

0.894 b.d. 7.5 2.7 12.4 0.6 8.2 40.8 27.2 1950.5 92.4 11.8 2128.1 62.8 45.3

0.099 0.411 0.091 0.592 0.256 0.126 0.358 0.069 0.463 0.100 0.304 0.044 0.294 0.045

0.004

Plag-Per

PlagDunite

Plag-Per

42.98 0.11 3.03 8.96 0.13 38.64 2.73 0.10 0.01 0.01 3.35 100.04

38.26 0.07 1.57 10.15 0.14 43.25 1.04 0.08 0.01 0.01 4.10 98.68

41.87 0.06 1.73 8.28 0.12 38.79 2.05 0.00 0.00 0.01 6.70 99.61

0.895 b.d. 1.5 2.4 3.4 0.6 8.1 46.1 23.3 2022.4 95.0 13.6 2397.6 66.6 45.6

0.894 b.d. 1.4 b.d. 8.0 0.3 7.3 50.5 10.2 2370.0 116.9 7.1 2362.4 36.0 45.1

0.903 b.d. 0.2 b.d. 1.6 0.4 6.8 38.7 20.7 2150.3 98.9 11.3 2398.9 53.7 45.6

PlagDunite Plag-Per 41.38 0.10 2.78 9.13 0.13 36.97 2.31 0.16 0.02 0.01 6.94 99.93 0.889 0.1 5.2 2.5 10.9 0.6 8.2 44.2 9.7 2017.9 98.0 11.1 2326.1 59.2 41.2

46.25 0.38 5.70 5.98 0.12 23.11 14.34 0.86 0.01 0.01 3.11 99.86 0.884 b.d. 11.8 16.4 38.0 0.2 3.5 5.9 170.4 1049.9 52.8 33.3 2374.1 201.1 b.d.

0.133 0.480 0.088 0.509 0.206 0.090 0.317 0.061 0.437 0.100 0.312 0.046 0.313 0.050 0.847 0.054 0.024

REE, Hf, Ta and Th were determined by ICP-MS; all others besides LOI by XRF; Fe, o~as Fe203; XMgbased on Fetot as FeO; b.d., below detection limit; missing values not analysed; Plag-Per, plagioclase peridotite.

as well as the c o m p o s i t i o n o f the liquid in equilibr i u m w i t h the p h e n o c r y s t s . U n f o r t u n a t e l y , the c u m u l a t e g a b b r o s rarely m e e t this r e q u i r e m e n t , as the g a b b r o c o m p o s i t i o n does n o t reflect the

c o m p o s i t i o n o f the melt in e q u i l i b r i u m with the crystallized phases. T h e c l i n o p y r o x e n e b a r o m e t e r d e v e l o p e d by N i m i s (1995, 1999) a n d N i m i s & U l m e r (1998)

SOUTHERN ALBANIAN OPHIOLITES

285

(a) 25.0

......

~''' I .......

f I~ ~

20.0 O

I .........

~ I~'~[

I .........

I .........

Voskopoja, Rehove, Morava UM cumulate 9 Cumulate gabbro

Troctolite Gabbronorite

~Ang~

e <> o -_

15.0

-~

10.0

5.0 0,0

......... 0.0

I ......... 10.0

I. . . . . . . . . . 20.0

30.0

40.0

50.0

wt.% MgO (b)

30.0

. . . . . . . . .

I . . . . . . . . .

An95 0

I

. . . . . .

-~,.

= ~

~ ~

03

O <

20.0

_

~"~ o-

' ' ' 1

. . . . . . . . .

I

. . . . . . . . .

Voskopoja, Rehove, Morava

~ ~

~ _~

~ ~ ~ J ~ ~

UM cumulate Cumulate gabbro Troctolite Gabbronorite Isotropic gabbro

-

= ~ ~ o_ 9

m

10.0

0.0 0.0

10.0

20.0

30.0

40.0

50.0

wt.% MgO Fig. 13. (a) MgO v. CaO and (b) MgO v. m1203 for ultramafic cumulates from the Voskopoja, Rehove and Morava massifs (calculated on a dry base).

286

F. KOLLER E T AL. (a)

25.0 Devolli, Vallama

-~ 20.0

~,~o [

UM cumulate Gabbro Shpati UMcumulate Troctolite Gabbro Luniku o

~

~

a

_"'-,,,~ +-~,,,,,~ O 15.0 ~ t'~ (O A ~ i a A ~ 9 10.0

9 b

r

1 9 zk_ A-_ :: +-

5.0 ....

I ,A 'qllmlm~ ~ . . '',~ Fo9

0.0 0.0

10.0

20.0

30.0

40.0

50.0

wt.% Mg0 (b) 30.0

20.0 O ,:1:

. . . . . . .An95.8o ' ........

' .........

"'Dev0iii,'

qah'ama'ra'

v,,,,,v,,,,,, UM cumulate [] ~ ~ Gabbro [] ~ ~ Shpati A ~ ~ UM cumulate AA ~ ~ \\ Troctolite 4.. ~ ~ Gabbro A A [] ~ ~ Luniku D IE3D " ~ ~ c g a b b r o

9 [] 9 AA: +i

"~ 10.0 -s~ 0.0

0.0

.....

. ., . . . . . . . .

10.0

,

,

30.0

40.0

, . . . . . .

20.0 wt.%

-

,:

50.0

MgO

Fig. 14. (a) MgO v. CaO and (b) MgO v. AI203 for gabbros, troctolites and ultramafic cumulates from the Devolli, Vallamara, Shpati and Luniku massifs (calculated on a dry base).

approached the problem differently, utilizing structural parameters based on structural formula data for calculating the pressure. For

several calculations this method requires precise temperature values with a large dependence of pressure calculations on the temperature input.

SOUTHERN ALBANIAN OPHIOLITES

5.0

. . . .

I

Devolli,

. . . .

Gabbro 4.0

9 .===

. . . .

Troctolite Gabbro

I

. . . .

I

Voskopoja,

[]

. . . .

I

Rehove,

. . . .

Morava

Cumulate gabbro

Shpati

zx A

Troctolite

e

Gabbronorite Isotropic gabbro

<>

9 A

3.0

Luniku

I--~

I

Vallamara

287

Isotropic gabbro + 2.0 9

OoO a

1.0

o . . . .

0.0 1.0

.

0.9

0.8

,~I

.

.

0.7

.

[] .

t

0.6

,

,

, .I

i ,

0.5

,

,

.

0.4

XMO0 Fig. 15. XMg V. TiO2for ultramafic cumulates (UM-C), gabbros, troctolites and gabbronorites from the south Albanian ophiolites.

Of the four calibrations available at present for pressure calculations, only three could be applied to our samples. They cover the following compositional ranges: (1) anhydrous basic melts from basalts, through trachybasalts to low-alkali nephelinites (BA in Table 8); (2) hydrous melts of the same composition; (3) tholeiitic basalts to dacites (TH in Table 8). Calibration (1) is temperature independent and restricted to anhydrous systems; the others two can also be applied to hydrous systems but these are strongly dependent on temperature. In the absence of reliable temperature calculations for the clinopyroxene-melt equilibrium and the rare (questionable) occurrence of magmatic hydrous phases such as magrnatic amphiboles, we chose, as a first approximation, the first calibration for anhydrous quartz-normative to nepheline-normative melts. The standard error estimated by Nimis (1999) for calibration (1) is 1.7 kbar. The results are presented in Table 8 (calibration BA). It is worth mentioning that if a temperature of 1100-1150~ is assumed, a reasonable correspondence between the pressures calculated using calibrations (1), (2) and (3) is obtained. The agreement among the three calibrations is in general better than 0.5 kbar. However, some ultramafic, mafic cumulates and gabbros contain clinopyroxeneorthopyroxene pairs allowing temperature calculations (e.g. Brey & K6hler 1990; Taylor

1998). The application of the thermometers yielded a wide range of temperatures: 8451070 ~ (Brey & K6hler 1990) and 845-1080 ~ (Taylor 1998). These temperatures are the maximum values and clearly show that subsolidus equilibration took place to a wide extent. Texturally, this is mainly expressed by clinopyroxene exsolution lamellae in orthopyroxene. The variation of Fe, Mg and Ca in clinopyroxene is relatively small. As Nimis (1998) and later Tartarotti et al. (2002) pointed out, the calibration (1) by Nimis & Ulmer (1998) is independent of temperature, but underestimates the pressures for clinopyroxene recrystallized under subsolidus conditions. The pressures obtained by calibration (1) should thus be viewed as minimum values. Calibration (3) requires the temperature of crystallization. Because of the negative correlation between temperature and the calculated pressure, this calibration provides maximum values. Using temperatures calculated according to Brey & K6hler (1990) and Taylor (1998), respectively, calibration (3) results in pressures of 10-13 kbar for Morava, 9-11 kbar for Voskopoja, 9.5-11.5 kbar for Devolli and 3-9.5 kbar for the Shpati ultramafic and mafic cumulates (Table 8, calibration TH). Compared with the minimum pressure estimates (calibration BA in Table 8) for Morava, Voskopoja and Devolli the pressure brackets are large, varying from c. 2 kbar to a maximum of

288

F. K O L L E R E T AL.

(a) 10.0

I

I

I

I

I

I

I

I

I

I

I

I

I

I

D !

r .

-

m

"0 tO cO

1.0

-9

0.1

Q. E

0.01

0

Shpati 9 -

UM cumulate

Voskopoja,

00 I

0.001

I

La

I

I

I

Pr Ce

I

I

I Gd

Sm Eu

Nd

I

Rehove, I

UM cumulate I I I I

Dy Tb

Morava!

Er Ho

9 1

Yb Tm

Lu

(b)

50.0

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

10.0

,2, ~--D

1.0

F:

0.1

(13

O9

0.01

-- Devolll, Vallamara - Gabbro [] I I I I

La

Pr Ce

I

Shpati Troctolite Gabbro I I

Sm Nd

Li /~ I

I

Gd Eu

VoskopoJa, Rehove, Morava-: Cumulate gabbro ~Troctolils ~Isoln)picgabbro 0I I I I I I

Dy Tb

Er Ho

Yb Tm

Lu

Fig. 16. (a) C1 chondrite-normalized REE distribution patterns for the ultramafic cumulates from the Shpati, Voskopoja, Rehove and Morava massifs. (b) C1 chondrite-normalized REE distribution patterns for the mafic cumulates and gabbros from (a) the south Albanian ophiolites.

SOUTHERN ALBANIAN OPHIOLITES

289

Table 7. Representative geochemical data for the gabbroic rocks of the southern Albanian ophiolites Sample: Massif:

A1010/96 Voskopoja

A109/96

A00/139

A01/228 Rehove

A99/028

Alb2/98

A99/121 Morava

A00/125

Rock:

Gabbro

Gabbro

Gabbro

Gabbro

Gabbro

Gabbro

Troctolite

Gabbronorite

SiOz TiO2 A1203 Fe203 MnO MgO CaO Na20 K20 P205 LOI Total

39.92 0.05 12.88 7.48 0.10 25.16 6.27 1.33 0.02 0.02 6.55 99.38

43.96 0.09 19.09 4.57 0.07 13.94 11.28 2.35 0.02 0.02 4.55 99.92

50.65 1.62 15.14 10.84 0.18 7.46 10.65 3.12 0.28 0.16 0.76 100.84

46.69 0.33 18.06 6.42 0.10 12.55 11.05 2.18 0.11 0.04 3.56 101.07

48.95 1.21 15.01 8.80 0.11 8.78 10.00 3.14 0.56 0.11 2.84 99.41

42.16 0.15 11.48 6.21 0.09 20.32 12.05 1.06 0.04 0.02 6.47 100.05

50.44 0.14 17.60 4.13 0.09 12.09 14.37 1.39 0.10 0.02 0.82 101.18

XMg Nb Zr Y Sr Rb Ga Zn Cu Ni Co Sc Cr V Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Th

0.869 b.d. b.d. b.d. 25.0 b.d. 5.0 41.0 60.0 1333.0 105.0 3.0 1775.0 20.0 b.d.

0.858 b.d. 0.2 0.1 679.0 b.d. 10.0 24.0 30.0 622.0 59.0 11.0 793.0 33.0 b.d.

0.577 3.2 111.9 31.2 172.2 3.4 12.9 58.7 49.8 77.4 44.7 31.2 190.2 220.0 16.1

37.02 0.14 9.07 8.15 0.12 28.85 5.90 0.04 0.01 0.02 9.09 98.41 0.875 0.1 1.9 1.2 6.7 0.6 6.6 46.9 40.7 1699.5 89.7 6.5 8315.4 71.7 68.7

0.045 0.008 0.052 0.039 0.066 0.064 0.011 0.095 0.021 0.053 0.004 0.027 0.004

13 kbar. In Shpati they are smaller, at 2-9.5 kbar. F o r R e h o v e , no c l i n o p y r o x e n e - o r t h o p y r o x e n e pairs exist. In all o f the massifs the pressure brackets are wide, similar to those f r o m the R o m a n c h e F r a c t u r e Z o n e ( T a r t a r o t t i et al. 2002). U s i n g the

0.795 0.6 9.6 3.8 133.5 0.8 12.3 48.5 129.3 494.8 46.5 12.9 532.3 79.6 10.0 0.577 1.840 0.325 1.884 0.698 0.431 1.022 0.177 1.201 0.258 0.758 0.109 0.682 0.104 0.379 0.100 0.017

0.664 1.3 77.3 26.2 191.1 0.0 12.2 18.1 7.8 104.1 50.0 456.3 231.6 22.1 2.624 8.868 1.533 8.435 2.963 1.263 3.923 0.721 4.875 1.052 3.118 0.443 2.894 0.456

0.866 0.1 4.5 1.5 479.5 1.3 1.0 32.7 97.7 926.8 60.4 15.9 1327.0 59.5 9.6

0.853 0.8 b.d. 1.8 89.5 1.1 7.5 31.2 90.5 329.7 28.5 16.7 822.1 98.8 18.5

0.036 0.182 0.052 0.483 0.281 0.162 0.511 0.098 0.721 0.153 0.457 0.063 0.406 0.060 0.118 0.025

0.124

t e m p e r a t u r e s calculated a c c o r d i n g to T a y l o r (1998), the closest b r a c k e t s were f o u n d in a Shpati g a b b r o (02A281), b e t w e e n 1.8 a n d 3.3 kbar. This indicates a d e p t h o f m a g m a intrusion o f a r o u n d 5-10 km. T h e o t h e r brackets are n o t well c o n s t r a i n e d but are in a c c o r d a n c e with such

F. KOLLER E T AL.

290 Table 7. Continued

Sample: Massif:

A03/349 Devolli

A03/350

A03/363

A02/280 Shpati

A02/281

A02/285

A02/329 Luniku

A03/444

Rock:

Gabbro

Gabbro

Gabbro

Gabbro

Gabbro

Troctolite

Gabbro

Gabbro

SiO2 TiO2

44.64 0.16 16.30 4.69 0.09 13.42 14.57 3.01 0.02 0.01 3.07 99.97

45.17 0.12 13.70 5.34 0.09 19.17 10.82 2.43 0.02 0.01 3.13 99.99

42.24 0.13 10.91 8.12 0.13 23.62 8.61 2.09 0.05 0.01 3.14 99.04

48.34 2.03 15.05 13.14 0.21 5.42 8.45 4.33 0.30 0.15 2.19 99.62

47.76 1.09 15.74 9.47 0.16 7.76 11.70 2.81 0.13 0.07 2.16 98.83

44.61 0.05 19.52 4.36 0.06 15.20 10.90 2.66 0.01 0.01 2.59 99.97

47.83 0.35 11.83 8.06 0.14 14.76 13.80 0.74 0.02 0.02 2.64 100.18

47.90 0.27 18.42 5.71 0.11 8.59 15.26 1.65 0.13 0.01 2.17 100.23

A1203 Fe203 MnO MgO CaO Na20

K20 P205 LOI Total XMg Nb Zr Y Sr Rb Ga Zn Cu Ni Co Sc Cr V Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Th

0.850 b.d. 1.1 6.6 44.8 1.0 14.4 25.9 23.7 420.4 36.1 37.1 1006.6 170.8 43.0

0.877 b.d. 2.2 4.0 54.0 0.7 12.5 32.8 55.4 933.6 45.8 20.3 1753.7 90.8 43.9

0.852 0.1 2.9 2.6 71.3 1.0 12.2 43.7 64.8 1152.6 69.0 14.4 2721.7 69.9 45.3 0.346 0.795 0.141 0.604 0.268 0.162 0.421 0.100 0.629 0.144 0.420 0.388 0.066 0.032 0.290 0.282

0.449 b.d. 103.2 48.6 155.3 2.3 26.6 89.5 38.4 49.6 42.0 32.5 79.4 305.4 b.d.

0.619 b.d. 48.5 29.4 122.3 1.2 24.2 65.4 69.8 128.6 39.1 30.9 355.2 223.2 b.d.

1.369 5.013 0.985 5.689 2.214 0.811 3.231 0.564 3.807 0.800 2.373 0.342 2.256 0.325

0.065

0.873 b.d. 1.4 5.9 92.3 b.d. 14.2 27.6 49.7 661.7 36.2 ! 2.0 654.1 35.4 6.8 0.163 0.407 0.054 0.239 0.084 0.140 0.137 0.032 0.229 0.063 0.204 0.035 0.230 0.044 0.768 0.226 0.146

0.784 0.4 9.4 16.9 39.5 b.d. 15.9 46.8 133.6 347.8 49.5 34.1 1135.7 178.5 b.d.

0.749 2.4 13.8 9.3 48.8 4.9 13.5 22.6 93.3 147.2 19.6 17.1 538.6 141.8 30.5

0.429 1.168 1.486 0.868 0.417 1.493 2.221 1.541 1.477 0.242 0.654 0.303

REE, Hf, Ta and Th were determined by ICP-MS; all others besides LOI by XRF; Fe,,, as Fe203, XMgbased on

Fetot as FeO; b.d., below detection limit; missing values not analysed.

an intrusion depth. In Shpati a n d Devolli, there are n o volcanic or plutonic rocks superimp o s e d o n the thin g a b b r o sequence to a c c o u n t for the d e p t h o f the g a b b r o intrusions. T h e original

o v e r b u r d e n was p r e s u m a b l y tectonically eroded, possibly creating ' t u r t l e b a c k ' d o m e s by m a n t l e d e n u d a t i o n ( T u c h o l k e et al. 1998; Nicolas et al. 1999).

291

SOUTHERN ALBANIAN OPHIOLITES Table 8. Clinopyroxene geothermobarometry and clinopyroxene-orthopyroxene thermometry Rock Voskopoja Gabbro Plag-Peridotite Rehove Gabbro Gabbro Wehrlite Plag-Peridotite Morava Troctolite Plag-Dunite Plag-Peridotite Devolli Gabbro Plag-Peridotite Shpati Troctolite Gabbro Gabbro Plag-Dunite Plag-Dunite Luniku Gabbro

Sample

P (kbar) Cal. BA

02A323 99A020

1.8 2.4

98A05 99A028 99A028 99A054

1.8 4.1 2.1 2.8

99A 121 B 99A088 99A120

T (~ Taylor

P (kbar) Cal. TH

T (~ Brey & K6hler

P (kbar) Cal. TH

925

< 10.9

981

< 8.9

1.5 3.1 2.4

867 884

< 13.0 < 11.3

928 931

< 10.8 <9.7

03A350 03A362

2.2 1.6

940 843

< 9.5 < 11.5

919 846

< 10.3 < 11.4

2A285a 02A281 02A280 02A323 02A319

1.4 1.8 3.8 2.2 3.1

1033 1067 1032 922 1011

< 4.7 < 3.2 < 8.4 < 9.6 < 7.9

959 983 1050 956 1079

< 7.5 < 6.6 < 7.4 < 8.3 < 5.3

02A329

1.9

Clinopyroxene geothermobarometry according to Nimis (1995, 1999) and Nimis & Ulmer (1998). Clinopyroxene-orthopyroxene thermometry after Brey & K6hler (1990) and Taylor (1998). Pressure calibrations: Cal. BA, anhydrous basalt, Cal. TH, tholeiitic basalt. Discussion

M O R v. S S Z origin: geochemical and mineral-chemical evidence During the last 25 years a number of attempts have been made to solve a longstanding controversy concerning the interpretation of ophiolites, the so-called 'ophiolite conundrum' (Moores et al. 2000): did they form in a mid-oceanic ridge environment, which is documented by their structure, or in a marginal basin (SSZ), which is indicated by their geochemistry? Extensive reviews of this problem were recently published by Hawkins (2003) and Pearce (2003). Apart from the geological evidence, the whole-rock geochemistry, in particular the immobile trace elements, the mineral chemistry, and the petrological and isotopic evidence have been used to discriminate between M O R and SSZ environments (Pearce 2003). M a n y of the discrimination methods apply only to unaltered rocks and uncontaminated basalts. Whereas lavas are often severely altered, impeding geochemical discrimination, gabbros are often less altered but are affected by cumulus and assimilation processes or trapped melts.

Nevertheless, some gabbros can be discriminated on their whole-rock geochemistry (Serri 1981). The incompatible elements are low in some gabbros, but a number of isotropic gabbros from Morava, Rehove and Voskopoja, as well as some from Shpati, have higher values in the range of MORB. The former show a very low mafic index (M.I. <0.3), as defined by Serri (1981), where a discrimination of high- or low-Ti gabbros is not feasible; the latter classify as high-Ti gabbros (Serri 1981). Zr/Y ratio is 3-4 and the V/Ti x 1000 ratio is > 20 for most of the isotropic gabbros from V R M and for some from Shpati. Other isotropic gabbros such as those from Devolli, Luniku and some from Shpati show much lower values: Zr/Y < 2 and V/Ti < 20. Provided the former gabbros represent true melts with no cumulate components, they can be interpreted as being derived from MOR-type magmas. This is corroborated by a rock/MORB plot, which shows a flat pattern clustering around unity for the isotropic gabbros from V R M (Fig. 17a) and by the R E E pattern (Fig. 16b). Conversely, m a n y gabbros from Shpati (except one sample with M O R characteristics), and those from Luniku and Devolli show a marked depletion of H F S E (Nb to Y; Fig. 17b),

292

F. KOLLER E T AL.

(a) 10.0

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

to

r~ 0

E

1.0

0.1

co Voskopoja, Rehove, 0.01

I

I

[

Sr

I

I

Rb

I

I

Th

KzO

9

Morava I

I

Ta

I

I

I

I

I

1

I

Ce

TiO2 Yb Cr P205 Hf Sm Y Sc Zr

Nb

Ba

I

(b) 10.0

-

I

i

i

i

i

I

i

i

i

i

I

I

i

i

i

I

I

I

I

I

I

I

I

I

I

I

rn

rv' O

1.o

(1) Q.

E 0.1

O9 - Luniku - Shpati - Devolli 0.01

I

I

Sr

A [] I

I

Rb K20

Th Ba

Nb Ta

P205 Ce

Hf

Zr

TiO2 Sm

I

Yb Y

I

Cr Sc

Fig. 17. (a) Spider diagrams of MORB-normalized trace element abundances in the isotropic gabbros from the Voskopoja, Rehove and Morava massifs. (b) the Luniku, Shpati and Devolli massifs. consistent with their formation in an SSZ environment. The variability in LILE is due to lowtemperature alteration. Gabbros from Devolli, Luniku and most of them from Shpati represent SSZ-type magmas. A number of the compositional ranges of the most important minerals in cumulates, such as

clinopyroxene, orthopyroxene, spinel or plagioclase, have been used to distinguish between MOR-related and subduction-related cumulates and gabbros. In particular, the An content ofplagioclase is used commonly in connection with XMg in clinopyroxene (Fig. 12) or forsterite content in olivine (Fig. 8) to discriminate between M O R

SOUTHERN ALBANIAN OPHIOLITES and arc gabbros (Burns 1985; Beard & Borgia 1989; Parlak et al. 1996, 2000). In Figure 12, the MOR field was taken from Burns (1985), Ross & Elthon (1993) and Schmincke et al. (1998), the Leg 37 field from H6bert et al. 1989), and the arc field from Burns (1985), including the field of Turkish ophiolites from Parlak et al. (1996, 2000). The cumulates and gabbros show a wide range in An content (only An-rich cores were used) from 50 to 97 and a smaller spread in XMg clinopyroxene from 0.70 to 0.97. High XMghas been noted by Elthon et al. (1982, 1992) and by Komor et al. (1985) for gabbros from the Bay of Islands and from oceanic gabbros, and was interpreted as an indication of moderate- to high-pressure crystallization of the gabbros in an island arc environment. However, B6dard & H~bert (1996, 1998) showed that the high X'Mgin clinopyroxene is due to reaction-assimilation processes unrelated to high pressure. Most troctolites and isotropic gabbros from VRM, and some ultramafic cumulates and isotropic gabbros from Shpati, indicate a MOR composition. Ultramafic cumulates and gabbros from DV, isotropic gabbros from Luniku and a few cumulate gabbros from VRM plot in the area close to the arc gabbros field. Ultramafic cumulates from VRM, a few gabbros from VRM, and troctolites from Shpati with high An and XMgplot between the two fields. In particular, gabbros from Luniku and a few cumulate gabbros from VRM overlap with the Leg 37 field from the Central Atlantic. Gabbros with high-An plagioclase and XMg clinopyroxene described by Ross & Elthon (1993), from Site 334 west of the Mid-Atlantic Ridge, are also from Leg 37. Their Na20 and TiO2 contents in clinopyroxene are as low as in those from DV and Luniku. Ross & Elthon (1993) concluded that these depleted clinopyroxenes formed in a MOR environment from a strongly depleted liquid, which was then collected in a crustal magma chamber to form cumulates. Eventually, the cumulates interacted with more enriched liquids migrating through them by porous flow. In a similar way, the Cr-number in spinels is indicative of the environment of formation, as the MOR-type spinels have, in general, Cr-number below 0.50, and spinels from the subductionrelated gabbros and cumulates have Cr-number around 0.60 and higher (Dick & Bullen 1984; Allan et al. 1988; H6bert et al. 1989). In the Cr-number v. Mg-number diagram (Fig. 11), the VRM mantle tectonites plot in the field of oceanfloor peridotites (peridotites I of Dick & Bullen 1984), whereas the DV harzburgites in the

293

peridotite III field and the Shpati ultramafic rocks have an intermediate position. Spinels of ultramafic and mafic cumulates show a Crnumber of 0.3-0.6 and a Mg-number of 0.3-0.64. The variability in the Mg-number of VRM ultramafic cumulates at Cr-number around 0.6 is possibly due to subsolidus re-equilibration (Roeder 1994). Kamenetsky et al. (2001) presented a new discrimination diagram based on TiO2 and A1203 contents in spinel. They argued that other than Mg-number and Cr-number, TiO2 and Al203 reflect the melt composition and are thus potential discriminating elements for the tectonic setting of spinel-bearing magmas. The comparison of spinels from olivinebearing cumulates in terms of Ti and A1 are shown in the inset in Figure 11. There is a clear negative correlation of TiO2 with A1203. Two groups can be distinguished; the larger one is compatible with a MOR environment; the second is enriched in TiO2 and partially overlaps with the ocean island basalt (OIB) field (consisting in large part of the ultramafic cumulates from VRM). It should be noted that spinels of the ultramafic rocks from DV are high in A1203 and plot in the MOR field, suggesting that at least the spinels came from a magma with a MOR affinity, whereas other parameters such as high An, Fo and XMgin clinopyroxene suggest an SSZ affinity. Olivine is mainly confined to the ultramafic cumulates, troctolites and cumulate gabbros but is rare in the isotropic gabbros. The forsterite content is limited to between Fo82and Fo90.When combined with the An content of plagioclase (Fig. 8), almost all of the olivines plot in the field of oceanic cumulates, as defined by Beard (1986). Only few plagioclase-olivine pairs, with high An content and a lower Fo content, plot outside, in the field delineated for the Oman gabbros by Browning (1984) and closer to the arc gabbros and the field of the Turkish and Troodos gabbros, respectively (Elthon 1987; Parlak et al. 2000, 2002). Again, as with the XMgclinopyroxene v. An diagram, the anorthite component would be consistent with arc gabbros but the Mgnumber of clinopyroxene as well as the Fo content are too high to fit well into the delineated field of arc gabbros. The high XMg of clinopyroxene, orthopyroxene and probably olivine is interpreted as an indicator of magma crystallization at moderate pressures (Elthon et al. 1982; Elthon 1987; Parlak et al. 1996, 2000, 2002). The composition of clinopyroxene in terms of Ti, Mg, Mn, Si and A1 is used widely to discriminate pyroxenes from various environments. All of the diagrams, such as TiO2-NazO-MnO (Nisbet & Pearce 1977), TiOz-Na20-SiO2/100,

294

F. KOLLER E T AL.

N a v . Ti or A1w v. Ti (Beccaluva et al. 1989), A1rv v. Ti (Loucks 1990), or Ti + C r v . Ca (Leterrier et al. 1982) exhibit a considerable overlap between clinopyroxene compositions formed in a MOR environment and those formed in an intraplate or SSZ environment. The discrimination is mainly based on the Ti/Na and the Ti/A1w ratios. Clinopyroxenes from cumulates and gabbros show a wide variation of all ratios. Clinopyroxenes of most cumulate gabbros and some ultramafic cumulates show a Ti/Na ratio < 0.4 and Ti/A1TMratio < 0.15, as do those from the ultramafic cumulates, troctolites and some gabbros from Shpati. The low ratios are even more pronounced for the Devolli and Vallamara ultramafic cumulates and gabbros and are also observed in same isotropic gabbros from Luniku. The high Ti/Na (>0.4) and Ti/A1~v (>0.15) ratios are found in most ultramafic cumulates from VRM (high-Ti clinopyroxenes) in most of the Shpati gabbros, and in some Luniku isotropic gabbros. Most isotropic gabbros from VRM have an intermediate ratio between the two groups. The first group represents those cumulates and gabbros with high XMg in clinopyroxene and high An in plagioclase, which plot in Figure 12 outside the field of MOR gabbros and close to, or inside, the arc gabbros field. All of these features, the high XMg in clinopyroxene, the high An content in plagioclase, and the low Ti/Na and Ti/A1w ratio in clinopyroxenes, indicate that these cumulates and gabbros did not form in a MOR environment, but crystallized instead from primitive high-MgO and low-TiO2 melts in an environment related to a subduction zone. The second Ti-rich group has more in common with typical MOR magmas, i.e. high Ti and Na contents in clinopyroxenes, partly lower XMgin clinopyroxenes and a lower An content in plagioclase. It should be noticed, however, that both magma groups do not form distinct fields but rather show a gradation in composition. Summarizing, the ultramafic and mafic cumulates and isotropic gabbros show a wide variety of geochemical compositions of the whole rock as well as of the minerals such as olivine, plagioclase, clinopyroxene and spinel. In various discrimination diagrams they form continuous arrays, ranging in many cases from the MOR to SSZ fields. The most discriminating elements or element ratios are consistent as they show MOR and SSZ environments for the same groups of rocks. Peridotites are predominantly lherzolitic for VRM, harzburgitic for DV, and show lherzolitic and harzburgitic composition for

Shpati. This is also reflected in the spinel composition of the mantle tectonites in Figure 11. Based on XMg in clinopyroxene and the anorthite in plagioclase relationship (Fig. 12), most of the VRM cumulate gabbros and troctolites, as well as some isotropic gabbros from VRM and Shpati, and the Shpati ultramafic cumulate are consistent with a MOR origin. On the other hand, the DV ultramafic cumulate, gabbros, some isotropic gabbros from Luniku and troctolites from Shpati indicate an SSZ environment, whereas ultramafic cumulates and some gabbros of VRM are intermediate. Where olivines are preserved in the SSZ group, they plot on the An v. Fo diagram in the 'Oman' field (Browning 1984). The composition of clinopyroxene shows the same pattern. The low Ti content in some ultramafic to mafic cumulates and a few isotropic gabbros suggests an SSZ origin, similar to troctolites and some gabbros from Shpati, and some Luniku gabbros and ultramafic cumulates and gabbros from DV. Despite a relatively wide range of overlap, it is clear that a number of gabbros were most probably generated in an SSZ environment. Tectonic setting

Any discussion on the tectonic setting of the Albanian ophiolites has to take into account that the Albanian ophiolites are only a small fraction of a much larger belt extending between Croatia and Greece. Unfortunately, the Dinaric ophiolites are not very well investigated. More modern studies exist only for the northern part (Pami6 et al. 2002, and references herein). In Greece, the neighbouring ophiolites, Pindos and Vourinos, are much better investigated. (For recent reviews see Smith & Rassios (2003); Bortolotti et al. (2004) and Rassios & Moores (2006). Until a few years ago, the Albanian ophiolites were seen as a MORB-like western belt and an SSZ-like eastern belt. This view came from the structure of the ophiolites (Shallo 1992, 1994; Robertson & Shallo 2000; Shallo & Dilek 2003) and was supported by some petrological and geochemical studies (Beccaluva et al. 1994b; Bortolotti et al. 1996). All of the tectonic models focused on a duality involving a normal midocean spreading to create a MOR-type ocean crust, to which SSZ-type crust was added later above an intra-oceanic subduction zone. In the Pindos ophiolite, which can be considered as the continuation of the western MORtype ophiolite in Albania, geochemical analysis (Capedri et al. 1980; Pearce et al. 1984; Jones

SOUTHERN ALBANIAN OPHIOLITES et al. 1991; Smith & Rassios 2003; Saccani & Photiades 2004) revealed, besides a MORB character, an SSZ contribution in the lavas and dykes. Hoeck & Koller (1999) showed for the first time that in the western ophiolite belt in southern Albania, SSZ lavas also occur. Later Bortolotti et al. (2002) and Hoeck et al. (2002) reported a wider occurrence of SSZ lavas from the southern and northern parts of the western ophiolite belt, demonstrating that the western belt also shows at least some subduction influence. This is supported by the present study. An SSZ signature has been reported from cumulates and basalts from a mid-ocean environment (west of Mid-Atlantic Ridge, and the Chile Ridge adjacent to the Chile Trench) (Ross & Elthon 1993; Klein & Karsten 1995; Sturm et al. 2000). The close proximity of the western belt to the arc-related magmas (eastern belt, Pindos), combined with a relatively thick cover of volcaniclastic sediments and turbidites, makes it likely that it was generated in an SZZ environment; nevertheless, MOR-type lavas prevail. Whether the western belt formed in a fore-arc setting (Bortolotti et al. 2002) or a back-arc setting (Hoeck et al. 2002) is difficult to assess. The SSZ magma was erupted or intruded relatively late as the dykes and gabbros intrude earlier M O R magma-derived cumulates. In part, both lava types erupted simultaneously (Bortolotti et al. 2002). From this relationship and the late occurrence of (rare) boninites, Bortolotti et al. (2002) inferred a fore-arc setting above an eastward-directed subduction zone. On the other hand, the predominance of MOR magma is consistent with a back-arc basin (see Smith & Rassios 2003; Saccani et al. 2004), in which MOR lavas and island arc-related magmas may erupt simultaneously (e.g. Mariana back-arc basin, Hawkins 2003). In addition to a fore-arc spreading setting, Deschamps & Lallemand (2003) pointed out that boninites could occur within back-arc spreading centres if (1) a spreading centre propagates at a low angle to the associated volcanic arc, (2) a spreading centre intersects a transition between a subduction zone and a transform fault at a high angle or (3) a spreading centre intersects a transform fault at right angles, and this subsequently changes into an incipient subduction zone (see also Monzier et al. 1993). An example of scenario (1) is the Valu Fa ridge, which is at a very low angle to the Tofua arc (Kamenetsky et al. 1997), and an example of scenario (2) is the southern Andaman Sea (Moores et al. 1984; Harris 2003). Both basins

295

may serve as a model for the Albanian ophiolites. The strongly depleted DV massifs with their most pronounced arc signature are difficult to interpret. If they were originally situated within the western belt, they could be viewed as possible older rifted arc crust, which was later segmented into ridges and troughs, as reported by Parson & Hawkins (1994) for the western part of the Lau basin. Alternatively, they could be thrust westwards from an original position further east between the Shebeniku and Bilisht massifs, to their present position between Voskopoja and Shpati. However, this needs to be tested further by structural studies along the boundaries of both massifs. Our findings and even the occurrence of some boninites (Bortolotti et al. 2002) do not contradict an origin of the western ophiolites in a back-arc basin. However, the interpretation of the western ophiolite belt as a fore-arc or backarc basin has some implications for the direction of subduction. If the western belt represents a fore-arc and the eastern belt an incipient arc, an eastward-dipping subduction zone relative to the present coordinates is more probable. If the western belt formed in a back-arc, a model we favour, a west-dipping subduction zone is more probable (see also Saccani et al. 2004). Despite a wealth of petrological and geochemical data from the ophiolites of Albania and Greece, more systematic petrologicalgeochemical mapping along and across the ophiolite belts is still needed to elucidate their mutual relationship. Additionally, the sediments on top of the ophiolites and the denudation of the ultramafic massifs of Devolli, Vallamara, Morava and Shpati deserve detailed investigations to resolve the problems of genesis of the Albanide-Dinaride ophiolites.

Conclusions (1) The southern Albanian ophiolites consist of a number of individual bodies, each with a distinct geology and lithology. They comprise lherzolites, plutonic rocks and mainly basaltic volcanic rocks in Voskopoja and Rehove, units with lherzolites-harzburgites and plutonic rocks in Morava and Shpati, and massifs with only harzburgites and a thin plutonic cap in Devolli and Vallamara. (2) The plutonic sequence contains ultramafic cumulates (e.g. plagioclase-bearing dunites, wehrlites), cumulate gabbros, troctolites and isotropic clinopyroxene gabbros. Gabbronorites are rare. The whole-rock geochemistry of the isotropic gabbros and the

296

F. KOLLER E T AL.

mineralogy of all of the cumulate rocks such as ZMg of clinopyroxene, An content of plagioclase, forsterite content in olivine and spinel chemistry indicate a wide compositional field for gabbros, ranging from M O R - to SSZ-type rocks. (3) Cumulates and gabbros from Voskopoja, Rehove and Morava predominantly show a M O R composition with a minor SSZ fingerprint, whereas mantle tectonites and cumulates from Devolli and Vallamara almost exclusively exhibit an SSZ signature. In Shpati and the small Luniku massif SSZ plutonic rocks occur together with a considerable amount of MOR-type gabbros. Our findings show that in the western ophiolite belt of southern Albania a considerable amount of SSZ magmas occur, not only within the volcanic rocks (Bortolotti et al. 2002; Hoeck et al. 2002) but also in the plutonic rocks. (4) The predominance of MOR-type over SSZtype crustal rocks together with the occurrence of volcanogenic sediments on top of the ophiolites support an origin of the ophiolites in a back-arc basin. The harzburgitic bodies from Devolli and Vallamara, as well as the occasional occurrence of boninites, do not exclude a back-arc basin formation. This contrasts with an alternative view by Bortolotti et al. (2002), in which the western belt ophiolites formed in a fore-arc basin. The interpretation of the western ophiolite belt as back-arc or as fore-arc derived has some implications for the direction of the subduction. In the first case, a westwarddipping subduction is indicated; in the second, the subduction should dip to the east. This research was substantially supported by the Austrian Nationalbank (Jubil~iumsfond) Project No. 7602 and by the Albanian Geological Survey in Tirana. The branch in Korce was helpful with logistic support. We would like to thank also in particular F. Dafa, K. Gjata, H. Hallaci, E. Bedini, M. Besku and last but not least S. Bushati (all Tirana), as well as H. Pula and P. Kalina (Korce), for an introduction to the geology of Albania and the Albanian ophiolites. The microprobe measurements were carried out by D. Topa (University of Salzburg). The XRF analyst was P. Nagl (University of Vienna). Finally, we would also like to thank the Austrian Embassy in Tirana, and the OAD for the continual support of our research work. The manuscript benefited substantially from the careful reviews by P. Nimis and R. H6bert. Their comments improved our thinking on petrological problems in ophiolites. Special thanks go to A. H. F. Robertson, one of the editors of this volume, for his thoughtful review and his continuous support and help.

References ALLAN, J. F., SACK, R. & BATIZA, R. 1988. Cr-rich spinels as petrogenetic indicators: MORB-type lavas from the Lamont seamount chain, eastern Pacific. American Mineralogist, 73, 741-753. 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. BEARD, J. S. & BORGIA, A. 1989. Temporal variation of mineralogy and petrology in cognate gabbroic enclaves at Arenal volcano, Costa Rica. Contributions to Mineralogy and Petrology, 103, 110-122. BEATTIE, P. 1993. Olivine-melt and orthopyroxenemelt equilibria. Contributions to Mineralogy and Petrology, 115, 103-111. BECCALUVA, L., MACCIOTTA, G., PICCARDO,G. B. & ZEDA, O. 1989. Clinopyroxene composition of ophiolite basalts as petrogenetic indicator. Chemical Geology, 77, 165-182. BECCALUVA,L., COLTORTI,M., DEDA, T., et al. 1994a. A cross section through western and eastern ophiolitic belts of Albania (Working Group Meeting of IGCP Project 25~-Field Trip A). Ofioliti, 19(1), 3-26. BECCALUVA, L., COLTORTI, M., PREMTI, I., SACCANI, E., SIENA, F. • ZEDA, 0. 1994b. Mid-ocean ridge and suprasubduction affinities in the ophiolitic belts of Albania. Ofioliti, 19(1), 77-96. BEDARD, J. H. & HEBERT, R. 1996. The lower crust of the Bay of Island ophiolite, Canada: petrology, mineralogy, and the importance of syntexis in magmatic differentiation in ophiolites and at ocean ridges. Journal of Geophysical Research, 103(B3), 5165-5184. B~DARD, J. H. & HEBERT, R. 1998. Formation of chromitites by assimilation of crustal pyroxenites and gabbros into peridotidic intrusions: North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland, Canada. Journal of Geophysical Research, 101(B11), 25105-25124. BORTOLOTTI,V., KODRA,A., MARRONI,M., MUSTAFA, F., PANDOLF1,L., PRINC1P1,G. & SACCANI,E. 1996. Geology and petrology of ophiolitic sequences in the Mirdita region (northern Albania). Ofioliti, 21(1), 3-20. BORTOLOTTI, V., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2002. Interaction between mid-ocean ridge and subduction magmatism in Albanian ophiolites. Journal of Geology, 110, 561-576. BORTOLOTTI, W., CHIARI, M., MARCUCCI, M., MARRONI, M., PANDOLFI,L. & PRINCIPI,G. 2004. Comparison among the Albanian and Greek ophiolites: in search of constraints for the evolution of the Mesozoic Tethys Ocean. Ofioliti, 29(1), 19-35. BREY, G. P. & KOHLER, T. 1990. Geothermometry in four-phase lherzolites II. Journal of Petrology, 31, 1353-1378. BROWNING, P. 1984. Cryptic variation within the cumulate sequence of the Oman ophiolite: magma chamber depth and petrological implications.

SOUTHERN ALBANIAN OPHIOLITES

In: GASS, G. I., LIPPARD, S. J. & SHELTON, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 71-82. BURNS, L. E. 1985. The Border Ranges ultramafic and mafic complex, south-central Alaska: cumulate fractionates of island-arc volcanics. Canadian Journal of Earth Sciences, 22, 1020-1038. CAPEDRI, S., VENTURELLI,G., BOCCHI,G., DOSTAL, J., GARUTI, G. & ROSSI, A. 1980. The geochemistry and petrogenesis of an ophiolitic sequence from Pindos, Greece. Contributions to Mineralogy and Petrology, 74, 189-200. CORTESOGNO, L., GAGGERO, L., JAHO, E., MARRONI, M., PANDOLFI, L. & SHTJEEANAKU,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(2), 49-64. DANYUSHEVSKY,L. V., SOBOLEV,A. V. & DMITRIEV, L. V. 1996. Estimation of the pressure of crystallization and H20 content of MORB and BABB glasses: calibration of an empirical technique. Mineralogy and Petrology, 57, 185-204. DESCHAMPS, A. & LALLEMAND,S. 2003. Geodynamic setting of Izu-Bonin-Mariana boninites. In: LARTER, R. D. & LEAT, P. T. (eds) Intra-oceanic Subduction System. Tectonic and Magmatic Processes. Geological Society, London, Special Publications, 219, 163-185. DICK, H. J. B. & BULLEN, T. 1984. Chromium spinel as petrogenetic indicator in abyssal and alpinetype peridotites, and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DILEK, Y., SHALLO,M. & FURNES, H. 2005. Rift-drift, seafloor spreading, and subduction tectonics of Albanian ophiolites. International Geology Review, 47, 147-176. ELTHON, D. 1987. Petrology of gabbroic rocks from the Mid-Cayman Rise spreading center. Journal of Geophysical Research, 92(B1), 658-682. ELTHON, D., CASEY, J. F. & KOMOR, S. 1982. Mineral chemistry of ultramafic cumulates from the North Arm Mountain Massif of the Bay of Islands ophiolite: evidence for high-pressure crystal fractionation of oceanic basalts. Journal of Geophysical Research, 87(B10), 871%8734. ELTHON, D., STEWART, M. & Ross, K. 1992. Compositional trends of minerals in oceanic cumulates. Journal of Geophysical Research, 97(B11), 15189-15199. FRASHERI, A., NISHANI, P., BUSHATI,S. & HYSENI, A. 1996. Relationship between tectonic zones of the Albanides, based on results of geophysical studies. In: ZIEGLER, P. A. & HORWATH, F. (eds) Per# Tethys Memoir 2: Structure and Prospects of Alpine Basins and Forelands. M6moires du Mfiseum National d'Histoire Naturelle, 170, 485-511. 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.

297

HAWKINS, J. W. 2003. Geology of supra-subduction zones--implications for the origin of ophiolites. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 227-268. HI'BERT, 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 Crustal Analogues. Geological Survey Department, Nicosia, 149-163. HI'BERT, R., SERRI, G. & HI~KINIAN,R. 1989. Mineral chemistry of ultramafic tectonites and ultramafic to gabbroic cumulates fi'om the major oceanic basins and Northern Apennine ophiolites (Italy)--a comparison. Chemical Geology, 77, 183-207. HOEC~:, V. & KOLLER, F. 1999. The Albanian ophiolites and the Dinaride--Hellenide framework. EUG 10 Strasbourg. Journal of Conference Abstracts, 4, 406. HOECK, V., KOLLER, F., MEISEL, T., ONUZI, K. & KNERINGER, E. 2002. The Jurassic South Albanian ophiolites: MOR-vs. SSZ-type ophiolites. Lithos, 65, 143-164. INSERGUEIX-FILIPPI, D., DUPEYRAT, L., DIMOLAHITTE, A., VERGI~LY,P. & BI~BIEN,J. 2000. Albanian ophiolites. II--Model of subduction zone infancy at a mid-ocean ridge. Ofioliti, 25(1), 47-53. ISPGJ-FGJM-IGJN 1983. Harta gjeologfike e Shqiperise. Scale 1:200 000. Tirana. JONES, G., ROBERTSON,A. H. F. & CANN, J. R. 1991. Genesis and emplacement of the supra-subduction zone Pindos ophiolite, northwestern Greece. In: PETERS, T. NICOLAS, A. & COLEMAN, R.G. (eds) Ophiolite Genesis and Evolution of the Oceanic Lithosphere. Ministry of Petroleum and Minerals, Sultanate of Oman, 771-799. KAMENETSKY, V. S., CRAWFORD, A. J., EGGINS, S. & MOHE, R. 1997. Phenocryst 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. KAMENETSKY, V. S., CRAWFORD, A. J. & MEFFRE, S. 2001. Factors controlling chemistry of magmatic spinel: an empirical study of associated olivine, Cr-spinel and melt inclusions from primitive rocks. Journal of Petrology, 42(4), 655-671. KLEIN, E. M. & KARSTEN, J. L. 1995. Ocean-ridge basalts with convergent-margin geochemical affinities from the Chile Ridge. Nature, 374, 52-57. KOMOR, S. C., ELTHON, D. & CASEY, J. F. 1985. Mineralogic variation in a layered ultramafic cumulate sequence at the North Arm Mountain Massif, Bay of Islands ophiolite, Newfoundland. Journal of Geophysical Research, 90(B9), 7705-7736. LETERRIER, J., MAURY, R. C., THONON, P., GIRARD, D. & MARCHAL, M. 1982. Clinopyroxene composition as a method of identification of the magmatic affinities of paleD-volcanic series. Earth and Planetary Science Letters, 59, 139-154. LOUCKS, R. R. 1990. Discrimination of ophiolitic from nonophiolitic ultramafic-mafic allochthons

298

F. KOLLER ET AL.

in orogenic belts by the A1/Ti ratio in clinopyroxene. Geology, 18, 346-349. LUGOVI, 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. MECO, S. & ALIAJ, S. 2000. Geology of Albania. Beitrgge zur regionalen Geologie der Erde, 28. Borntraeger, Berlin. MEISEL, T., SCHONER,N., PALIULIONYTE,V. & KAHR, E. 2002. Determination of rare earth elements (REE), Y, Th, Zr, Hf, Nb and Ta in geological reference materials G-2, G-3, SCo-1 and WGB-1 by sodium peroxide sintering and ICP-MS. Geostandards Newsletter, 26(1), 53-61. MI~TRICH, N. & RUTHERFORD, M. J. 1998. Low pressure crystallization paths of H20-saturated basaltic-hawaiitic melts from Mt. Etna: implications for open-system degassing of basaltic volcanoes. Geochimica et Cosmochimica Acta, 62(7), 1195-1205. MONZIER, M., DANYUSHEVSKY, L. W., CRAWFORD, A. J., BELLON, H. & COTTON, J. 1993. High-Mg andesites from the southern termination of the New Hebrides island arc (SW Pacific). Journal of Volcanology and Geothermal Research, 57, 193-217. MOORES, E. M., ROBINSON, P. T., MALPAS, J. & XENOVHONOTOS, C. 1984. Model for the origin of the Troodos massif, Cyprus, and other mideast ophiolites. Geology, 12, 500-503. MOORES, E. M., 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. 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. NIMIS, P. 1995. A clinopyroxene geobarometer for basaltic systems based on crystal-structure modeling. Contributions to Mineralogy and Petrology, 121, 115-125. NIMIS, P. 1998. Clinopyroxene geobarometry of pyroxenitic xenoliths from Hyblean Plateau (SE Sicily, Italy). European Journal of Mineralogy, 10(3), 521-533. NIMIS, P. 1999. Clinopyroxene geobarometry of magmatic rocks. Part 2. Structural geobarometers for basic to acid, tholeiitic and mildly alkaline magmatic systems. Contributions to Mineralogy and Petrology, 135, 62-74. NIMIS, P. & ULMER, P. 1998. Clinopyroxene geobarometry of magmatic rocks. Part 1. An expanded structural geobarometer for anhydrous and hydrous, basic and ultrabasic systems. Contributions to Mineralogy and Petrology, 133, 122-135. NISBET, E. G. & PEARCE, J. A. 1977. Clinopyoroxene composition in mafic lavas from different tectonic

settings. Contributions to Mineralogy and Petrology, 63, 149-160. PAMIC, J., TOMJLENOVIC, B. & BALEN, D. 2002. Geodynamic and petrogenetic evolution of Alpine ophiolites from central and NW Dinarides: an overview. Lithos, 65, 113-142. PARLAK, O., DELALOYE, M. & BiNGOL, E. 1996. Mineral chemistry of ultramafic and mafic cumulates as an indicator of the arc-related origin of the Mersin ophiolite (southern Turkey). Geologische Rundschau, 85, 647-661. PARLAK, O., HOCK, V. & DELALOYE, M. 2000. Suprasubduction zone origin of the PozantlKarsantl ophiolite (southern Turkey) deduced from the whole-rock 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. PARLAK, O., HOCK, V. & DELALOYE, M. 2002. Suprasubduction zone Pozantl-Karsantl ophiolite, southern Turkey: evidence for high pressure crystal fractionation of ultramafic cumulates. Lithos, 65, 205-224. PARSON, L. M. & HAWKINS, J. W. 1994. Two-stage ridge propagation and the geological history of the Lau backarc basin. In: HAWKINS, J., PARSON, L., ALLAN, J. et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 135, Ocean Drilling Program, College Station, TX, 819-828. PEARCE, J. A. 2003. Supra-subduction zone ophiolites: the search for modern analogues. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 269-293. 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. PUTIRKA, K. 1999. Clinopyroxene + liquid equilibria to 100kbar and 240kbar. Contributions to Mineralogy and Petrology, 135, 151-163. PUTIRKA, K., JOHNSON,M., KINZLER, R., LONGHI,J. 8~; WALKER, D. 1996. Thermobarometry of mafic igneous rocks based on clinopyroxene-liquid equilibria, 0-30 kbar. Contributions to Mineralogy and Petrology, 123, 92-108. PUTIRKA, K. D., MIKHAELIAN, H., RYERSON, F. & SHAW, H. 2003. New clinopyroxene-liquid thermobarometers for mafic, evolved, and volatilebearing lava compositions, with applications to lavas from Tibet and the Snake River Plain, Idaho. American Mineralogist, 88, 1542-1554. ROBERTSON, A. H. F. & SHALLO, M. 2000. MesozoicTertiary tectonic evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-254. ROEDER, P. L. 1994. Chromite: from the fiery rain of chondrules to the Kilauea Iki lava lake. Canadian Mineralogist, 32, 729-746.

SOUTHERN ALBANIAN OPHIOLITES Ross, K. & ELTHON, D. 1993. Cumulates from strongly depleted mid-ocean-ridge basalt. Nature, 365, 826-829. SACCANI, E. & PHOTIADES, A. 2004. Mid-ocean ridge and supra-subduction affinities in the Pindos ophiolites (Greece): implications for magma genesis in a forearc setting. Lithos, 73, 229-253. SACCANI, E., BECCALUVA,L., COLTORTI, M. & SIENA, F. 2004. Petrogenesis and tectono-magmatic significance of the Albanide-Hellenide Subpelagonian ophiolites. Ofioliti, 29(1), 75-93. SCHMINCKE, H.-U., KLt0GEL, A., HANSTEEN, T. H., HOERNLE, K. & VAN DEN BOGAARD, P. 1998. Samples from the Jurassic ocean crust beneath Gran Canaria, La Palma and Lanzarote (Canary Islands). Earth and Planetary Science Letters, 163, 343-360. SERRI, G. 1981. The petrochemistry of ophiolite gabbroic complexes: a key for the classification of ophiolites into low-Ti and high-Ti types. Earth and Planetary Science Letters, 52, 203-212. SHALLO, M. 1992. Geological evolution of the Albanian ophiolites and their platform periphery. Geologische Rundschau, 81, 681-694. SHALLO, M. 1994. Outline of the Albanian ophiolites. Ofioliti, 19(1), 57-75. 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, 351-363. SHALLO, M., KODRA, A. & GJATA, K. 1990. Geotectonics of the Albanian ophiolites. In: MALPAS, J. MOORES, E. M., PANAYIOTOU,A. & XENOPHONTOS, C. (eds) Troodos 1987--Ophiolites, Oceanic Crustal Analogues, 265-269.

299

SMITH, A. G. & RASSlOS, A. 2003. The evolution of ideas for the origin and emplacement of the western Hellenic ophiolites. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 337-350. STURM, M. E., KLEIN, E. M., KARSTEN, J. L. & KARSON, J, A. 2000. Evidence for subductionrelated contamination of the mantle beneath the southern Chile Ridge: implication for ambiguous ophiolite compositions. In: DILEK, Y., MOORES, E. M., ELTHON, D. & NICOLAS,A. (eds) Ophiolites and Oceanic Crust." New Insights from FieM Studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 13-20. 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. TAYLOR, W. R. 1998. An experimental test of some geothennometer and geobarometer formulations for upper mantle peridotites with application to the thermobarometry of fertile lherzolites and garnet websterite. Neues Jahrbuch fiir Mineralogie, Abhandlungen, 172, 381408. TUCHOLKE, B. E., LIN, J. & KLEINROCK, M. C. 1998. Megamullions and mullion structure defining oceanic metamorphic core complexes on the MidAtlantic Ridge. Journal of Geophysical Research, 103, 9857-9866. YANG, H.-J., KINZLER, R. & GROVE, T. L. 1996. Experiments and models of anhydrous, basaltic olivine-plagioclase-augite saturated melts from 0.001 to l0 kbar. Contributions to Mineralogy and Petrology, 124, 1-18.

Neotethyan ophiolites: formation and obduction within the life cycle of the host basins ZVI GARFUNKEL

Institute o f Earth Sciences, Hebrew University, Jerusalem, Israel, 91904 (e-mail. zvi. garfunkel@huji, ac. il) To understand the Neotethyan ophiolites better, their place in the history of the host basins is explored, using the Jurassic Hellenic-Dinaric and some Cretaceous (mainly peri-Arabian) ophiolites as examples. These formed in mature (c. 60 Ma and 100 Ma old) seaways by spreading at rates that apparently were too high to persist for more than a fraction of the basin history. Each ophiolite group formed in a short time interval, about 10 Ma, soon after changes in the motions of plates in the host basins. Therefore, these ophiolites do not seem to have formed by normal long-term spreading along mid-ocean ridges, but their formation signifies special 'ophiolite events'. This fits well the widely accepted origin in a supra-subduction zone (SSZ) setting that was inferred from geochemical data. The ophiolites studied here are thus interpreted as having formed in new subduction zones that originated during changes in plate motions. The spreading during their accretion was driven by fast retreat (roll-back) of the subducting slabs. The western ophiolites of the Hellenic-Dinaric belt, dominated by mid-ocean ridge basalt (MORB)-like rocks but invaded by SSZ magmas, could have formed along ridges just before they failed (collapsed), but the age data fit better formation in a proto-back-arc setting alongside the more eastern ophiolites with the typical SSZ signature. The construction of the ophiolites examined here ended when they were detached from their substrate and pushed over the adjacent basins. At that stage they were underplated by metamorphic soles, but how the latter were emplaced still needs clarification. Continuing retreat of the subducting slabs consumed the host basins and pushed the ophiolites hundreds of kilometres until they were obducted over the nearby margins 15-20 Ma after formation. This framework seems to apply to many ophiolites and allows us to interpret them in terms of known processes, but also highlights problematic issues that still need to be resolved. Abstract:

Ophiolites are a conspicuous though minor component of Phanerozoic and Neoproterozoic orogens (Dilek 2003a). Following Hess (1965) and Gass (1968) they are recognized as fragments of oceanic crust and uppermost mantle. Being important indicators of the past operation of the Wilson cycle and the existence of former oceanic basins between components of orogens, ophiolites were extensively studied. Early workers favoured an origin along mid-ocean ridges (MORs) (e.g. Gass 1968; Moores & Vine 1971; Coleman 1977), but later geochemical studies pointed at a supra-subduction zone (SSZ) origin of many ophiolites, probably in a young pre-arc (nascent forearc) setting similar to what is found in the forearcs in the western Pacific (Casey & Dewey 1984; Hawkins et al. 1984; Leich 1984; Pearce et al. 1984; Stern & Bloomer, 1992). Moores (1982) recognized the diversity of ophiolites and distinguished between 'Tethyan (Mediterranean)' ophiolites that are obducted (thrust) over passive continental margins and 'Cordilleran (Pacific)' ophiolites that are incorporated into accretionary complexes, but Shervais (2001) stressed that the magmatic history of ophiolites of

both types followed similar evolutionary trends, although they differ in other respects (Beccaluva et al. 2004). Other ophiolite types were also recognized (Dilek 2003b). Studies of Mediterranean and Middle East ophiolites, especially the Troodos (Cyprus) and Semail (Oman) ophiolites, contributed much to these ideas (e.g. review by Robertson 2002). Although ophiolites were studied from various points of view, many questions regarding their origin and emplacement are still debated (e.g. Shervais 2001; Flower & Dilek 2003; Robertson 2004, and references therein). One way to advance the understanding of ophiolites is to examine how their history relates to the evolution of the host basins. The present work follows this approach, but in view of the ophiolite diversity the discussion is limited to a few groups of Tethyan ophiolites: Jurassic Hellenic-Dinaric ophiolites and some Cretaceous ophiolites (Fig. 1), whereas other ophiolites in that region that were emplaced along active margins (Robertson 2002, 2004) are not considered. The present work builds on previous discussions, but stresses plate kinematic considerations and features of subduction zones. Below, some

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 301-326. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

302

Z. GARFUNKEL

9

4

Black

S

9 N p'

\ ~ ~

%"

%

~. ?",,-~

,.9

30*

~ '-...~ i

/

Fig. 1. Distribution of ophiolites in the studied region. Alb., Albanian ophiolites; B, Baer-Bassit ophiolite; C, Cilo ophiolite; Ker., Kermanshah ophiolite; Klz]lda(gophiolite; Ne, Neyriz ophiolite; Pel., Pelagonian-Korab block; Pi, Pindos ophiolite; Ot, Othris ophiolite; San. -Sirj., Sanandaj-Sirjan block; Sem., Semail ophiolite; Tr., Troodos ophiolite; V, Vourinos ophiolite; Var., Vardar zone.

general considerations regarding ophiolite origin are first briefly presented, and then features of the ophiolites examined here that are relevant for the subsequent discussion are summarized. Based on these data the relations between these ophiolites and the histories of the host basins are discussed, paying attention to problems that need further study. Here stratigraphic ages deduced from fossils and from radiometric dating are compared using the time scale of Gradstein et al. (2004) 9

Neotethyan examples General setting and f e a t u r e s

The ophiolites considered here formed in Mesozoic seaways or basins that existed in the realm between Eurasia and Gondwanaland. The history of this realm, whose closure gave rise to the Alpine orogenic chain, has been interpreted in somewhat different ways (e.g. LePichon et al. 1988; Robertson et al. 1991, 1996; Stampfli et al. 2001), but all interpretations agree that early in the Mesozoic several blocks (micro-continents) rifted away from the major continents, especially from Gondwanaland, and that their subsequent drifting led to the growth of new (Neotethyan) seaways. The ophiolites considered here originated in these seaways and were obducted on their passive margins. Now these ophiolites occur

in belts that may be 1000-2000 km long that mark the sutures between blocks that bordered these basins. The ophiolites examined here are of two major types (Boudier & Nicolas 1985; Nicolas 1989). Most of them have a mantle part that consists largely of harzburgitic tectonites and an overlying crustal part that comprises the complete classic Penrose pseudo-stratigraphy (Anonymous 1972), i.e. (1) a plutonic unit of ultramafic and mafic rocks, often with some high-level plagiogranites, (2) a sheeted dyke complex, and (3) an extrusive unit, often overlain by (or interbedded with) abyssal sediments. The total thickness, where all units are preserved, may reach 12 km or more. Less common are ophiolites whose mantle part consists mainly of lherozlitic tectonites, and the crustal part comprises plutonic and volcanic units, whereas a sheeted dyke complex is absent or poorly developed; their total thickness does not exceed 3-4 km. The ophiolites are associated with allochthonous units that originated in the same basins and provide crucial information about these basins (e.g. Moores 1982; Robertson 2004, and references therein). They include: (1) strongly deformed metamorphic rocks, with inverted metamorphic gradients, that form soles up to several hundred metres thick at the base of the ophiolite

LIFE CYCLE OF NEOTETHYAN OPHIOLITES bodies; their protoliths consist of basinal sediments and volcanic rocks that differ from the volcanic rocks in the ophiolites; (2) accretionsubduction complexes (m61anges) that formed during ophiolite obduction; they consist of imbricated slices of pre-ophiolite rocks that were scraped off the floors of the host basins and the passive margins on which the ophiolites were obducted, but m61anges derived from the ophiolites themselves are often also present, and serpentinite bodies are sometimes also found. Ophiolites are generally accepted to represent new crust that formed by some type of sea-floor spreading, i.e. along zones of plate separation. Formation in an intra-oceanic setting, some distance from any continental margin, is also accepted, based on the nature of the overlying sediments (Robertson 2004). The metamorphic soles formed when their protoliths were overthrust by hot material, generally thought to be the ophiolites themselves (Spray 1984; Woodcock & Robertson 1977; but see below). On the other hand, it is still debated whether ophiolites formed along MORs or in a SSZ setting. The first interpretation is appealing because the structure of ophiolites is readily explained as formed by sea-floor spreading. Nicolas (1989) interpreted the harzburgitic and lherzolitic ophiolites as having formed by fast and slow spreading, respectively. However, the geochemical signature of the igneous rocks of the harzburgitic ophiolites differs from mid-ocean ridge basalt (MORB) but resembles rocks found in SSZ settings: low-Ti basalts, island arc tholeiites (IAT), or very low-Ti depleted magmas with a boninitic affinity. The crystallization of plagioclase after pyroxenes and the occurrence of orthopyroxene in cumulates also differs from MORB. Thus formation of the harzburgitic ophiolites in an SSZ setting was advocated (e.g. Cameron et al. 1980; Laurent et al. 1980; Pearce et al. 1984) and is widely accepted, although typical island arcs or sediments derived therefrom are most often lacking. In contrast, the volcanic rocks in lherzolitic ophiolites are high-Ti MORBlike rocks that could have formed along MORs. However, in some ophiolites rocks of both types occur together (see below). Ophiolites o f the H e l l e n i c - D i n a r i c orogen

An almost 1000 km long belt of Jurassic ophiolites extends through Greece, Albania and former Yugoslavia along the Hellenic-Dinaridic orogen (Fig. 1). They are thought to have formed in the Pindos Ocean, which existed between the Apulian and the Pelagonian (Korab) blocks (microcontinents) (Robertson et al. 1991; Doutsos et al.

303

1993; Smith 1993; Robertson & Shallo 2000; Robertson 2002). Another ophiolite belt with a different history, not treated here, extends east of the Pelagonian block and is considered to have formed in the distinct Vardar Ocean (Fig. 1; Robertson 2002). The underlying allochthonous units (Smith et al. 1979; Jones & Robertson 1991; Jones et al. 1991; Bortolotti et al. 1996, 2004; Robertson & Shallo 2000; Saccani et al. 2003) record rifting and formation of the Pindos oceanic basin on the western side of the Pelagonian block. Deepwater sediments accumulated in the basin until the Oxfordian (Danelian & Robertson 2001). Triassic rifting is also indicated by widespread magmatism of that age on the Pelagonian block (Robertson et al. 1991; Pe Piper 1998; Robertson & Shallo 2000). The ophiolites of the Hellenic-Dinaric belt were thrust eastward onto the Pelagonian (Korab) block (micro-continent) in Late Jurassic times (Smith 1993; Bortolotti et al. 1996; Rassios & Smith 2000; Robertson & Shallo 2000). However, the part of the Pindos basin west of them survived until the Eocene (Clift & Robertson 1989; Degnan & Robertson 1998) and it was only then that the area was deformed into a pile of SW-vergent thrust sheets. To represent this belt the better studied ophiolites in northern Greece and in Albania are considered in some detail. The ophiolites o f northern Greece. The Vourinos, Pindos, and Othrys ophiolites are well preserved (Fig. 1), whereas more southern ophiolites are disrupted. The Vourinos ophiolite comprises harzburgitic tectonites and displays the classic Penrose pseudo-stratigraphy, although the volcanic unit is reduced (eroded?) (Moores 1969; Smith 1993; Rassios & Smith 2000). The sheeted dykes and extrusive rocks consist mainly of low-Ti mafic rocks and minor andesites and dacites with an IAT affinity that are cut by dykes and small intrusions with a boninitic affinity (Beccaluva et al. 1984). The Pindos ophiolite comprises two thrust sheets (Capedri et al. 1980; Jones & Robertson 1991; Jones et al. 1991; Saccani & Photiades 2004; Saccani et al. 2004). The upper sheet consists only of cumulates and of harzburgitic tectonites. The latter generally occur in ophiolites with SSZ affinity, but the cumulates comprise also rocks that crystallized from MORB-like magmas. The lower sheet, although rather thin and disrupted, contains the entire pseudo-stratigraphy. The cumulates and some extrusive rocks have MOR affinities, whereas the volcanic rocks comprise alternating MORB, IAT and boninitic rocks, the last tending to be the younger ones. Dykes with boninitic affinity cross all units. Further south the Othris ophiolite

304

Z. GARFUNKEL

preserves slices of harzburgites as well as lherzolitic tectonites (Smith et al. 1975, 1979; Saccani et al. 2004). Thus, these ophiolites display a range of rock types, some comprising only rocks with SSZ affinities whereas others include such rocks as well as rocks with MORB affinities, the latter being the older ones. Dating of zircons by the U/Pb method (Liati et al. 2004) revealed that the Vourinos and Pindos ophiolites have the same ages within analytical uncertainty. A gabbro from the Pindos ophiolite yielded an age of 171 _+3 Ma, whereas a gabbro and a plagiogranite from the Vourinos ophiolite yielded ages of 168.5+2.4Ma and 172.9+ _ 3.1 Ma, respectively. Radiolarites overlying the Vourinos complex contain latest Bajocian-Early Bathonian (c. 168 Ma) and somewhat younger fossils (Chiari et al. 2003). Thus there is no clear age distinction between them. The metamorphic soles beneath these ophiolites yielded 4~ ages (using updated decay constants; Spray et al. 1984) of 171___4 Ma to 165_+ 3 Ma, close to the ages obtained by Liati et al. (2004). These ages and the occurrence of ophiolite-derived clasts in radiolarites of Early Bathonian to Early Callovian age (c. 166162 Ma) in a m61ange beneath the Pindos ophiolite (Jones et al. 1992) indicate tectonization shortly after ophiolite formation. Flexure in front of the approaching ophiolites caused the Pelagonian block to subside in OxfordianKimmeridgian times (Danelian & Robertson 2001). Final uplifting and obduction are recorded by Tithonian-Beriassian (c. 148-142 Ma) sediments overlying the Vourinos ophiolite (Smith 1993). Further south the ophiolites are disrupted but are similar, comprising rocks with both MOR and SSZ affinities, and have similar ages and similar emplacement histories (Robertson 2002; Liati et al. 2004; Saccani et al. 2004). The Albanian ophiolites. These ophiolites are well preserved. They mark the northward prolongation of the Greek ophiolite belt and had a similar history (Fig. 1; Beccaluva et al. 1994; Shallo 1994; Bortolotti et al. 1996; Robertson & Shallo 2000; Saccani et al. 2004). The ophiolites along the eastern side of the belt comprise harzburgitic tectonites, are thick ( > 8 km), and show a complete pseudo-stratigraphy, although in the south their upper parts were eroded. The sheeted dykes and volcanic rocks consist mainly of low-Ti basalts, basaltic andesites and dacites with an IAT character. Rocks with a boninitic affinity occur as dykes and as extrusive rocks in the higher parts of the volcanic units. The ophiolites

along the western side of the belt are only up to 3-4 km thick, and comprise lherzolitic tectonites and plutonic and volcanic units that have variable thicknesses because of faulting; sheeted dykes are generally absent. The volcanic rocks consist of MORB-like high-Ti tholeiites. However, in many places the western type ophiolites contain rocks of both MOR and SSZ affinity; either these alternate in time or the SSZ-type rocks are the younger ones (B~bien et al. 1998, 2000; Bortolotti et al. 2002; Hoeck et al. 2002). These occurrences show that originally the two ophiolite types were not widely separated and that at some times the two magma types were available in the same places. Dimo-Lahitte et al. (2001) obtained 4~ ages of 163.8 • 1.8 Ma (Callovian) for a plagiogranite in the NE and 172.6 _+ 1.7 Ma (close to the Aalenian-Bajocian transition) for a dyke in an ophiolite further south. Fossils in overlying cherts have Bathonian to Callovian or Oxfordian ages (Bortolotti et al. 1996). The metamorphic soles record P-T conditions that range from 800-860 ~ 0.9-1.2 GPa (equivalent 25-35 km depth) through 550-700 ~ 0.4-0.6 GPa (1218 km) to greenschist-facies conditions (Carosi et al. 1996; Dimo-Lahitte et al. 2001). They yielded 4~ ages of 173-169 Ma (AalenianBajocian) in the south and about 164Ma (Callovian) in the north, with few intermediate ages between these. There is no difference between the ages on the eastern and western sides of the ophiolite belt (Dimo-Lahitte et al. 2001). Ophiolites of all types are overlain by a Tithonian m61ange and Barremian-Lower Valanginian turbidites, both consisting of components derived from the two ophiolite types, from the adjacent basin floor, and from the Korab block (notably quartzitic sandstones) (Bortolotti et al. 1996). Thus, at c. 150-140 Ma the ophiolites were already close to the Korab micro-continent, but still under the sea. These sediments occur also as slices in the sub-ophiolitic allochthons, which confirms that they formed during ophiolite obduction over the Korab block. After uplift the ophiolites were covered by shallow-water sediments from the Barremian. As further south, remnants of the Pindos Ocean survived west of the ophiolites until they were tectonized in the early Tertiary. The Hellenic-Dinaric ophiolites extend northward into former Yugoslavia (Pami6 et al. 2002; Fig. 1), but these have been less studied. They are generally strongly deformed. Here the tectonites are predominantly lherzolitic, and volcanic rocks transitional between IAT and MORB are present. They had a similar history to the more southern ophiolites.

LIFE CYCLE OF NEOTETHYAN OPHIOLITES Summary. The foregoing account shows that the Hellenic-Dinaric ophiolites formed in a narrow time interval between e. 170 Ma and e. 165 Ma within an oceanic basin that originated in the Triassic, i.e. this basin was 60-70 Ma old when the ophiolites formed. These ophiolites comprise bodies of the harzburgitic type with a complete classic pseudo-stratigraphy and magmas with a SSZ character, and also bodies of the lherzolitic type with an incomplete pseudo-statigraphy and magmas with a MOR character, but many of the latter contain also magmas with a SSZ character. Thus the two ophiolite types appear to have formed in close proximity in time and space. The ages of the metamorphic soles (173-164 Ma) are close to or overlap the ages of ophiolite formation. Obduction over the PelagonianKorab block continued 15-20 Ma after ophiolite formation. The part of the Pindos basin west of the ophiolites survived some 100 Ma after ophiolite obduction, until its final closure in the Tertiary. O p h i o l i t e s o f the A n a t o l i a n d o m a i n a n d p e r i - A r a b i a n belt

In the Anatolian domain, east of the Aegean Sea, several belts of Cretaceous ophiolites are present (Fig. 1). They originated in different seaways that formed in the Early Mesozoic (probably Triassic) times and persisted into the Palaeocene ($eng6r et al. 1988; Yazgan & Chessex 1991; Yllmaz et al. 1993; Okay & Tiiysiiz 1999; Robertson 2002). Here the focus is on the southern belt that also extends along the periphery of Arabia, whereas the more northern belts are only mentioned briefly. A conspicuous belt of ophiolites extends along the Ankara-Erzincan suture (Fig. 1; Okay & Tfiysfiz 1999; Yaliniz et al. 2000; Robertson 2002; Onen 2003). They are incompletely preserved, but various components show a subductionrelated character. These ophiolites were obducted southward in Campanian-Maastrichtian times while a volcanic arc was active on the north side of the host basin. Metamorphic soles yielded 4~ ages of 101_+4Ma and 93_+4Ma (0hen 2003). A noteworthy feature is that in western Turkey ophiolites were emplaced over blueschists that formed at c. 20 kbar (55-60 km depth) in the Campanian (Okay et al. 1998). Further south ophiolites were obducted over the Tauride block (Fig. 1; Whitechurch et al. 1984; Dilek et al. 1999). They comprise harzburgitic tectonites and cumulates, whereas higher units are present in a few places only, probably because of syn-emplacement erosion (they may

305

be represented by clasts in the associated m61anges). In these ophiolites the cumulates and higher units, where present, show an SSZ fingerprint (Parlak et al. 1996, 2000). Metamorphic soles of the Taurus ophiolites yield K/Ar ages scattered mainly between c. 100 Ma and c. 85 Ma, with an age concentration around 95-90 Ma, and a few 4~ ages of 94-90 Ma (Dilek et al. 1999; Parlak & Delaloye 1999). A special feature is that these ophiolites and their metamorphic soles are crossed by numerous diabase dykes that formed from evolved IAT magmas with an SSZ signature that sometimes were less refractory than the magmas that from which the host cumulates crystallized (Whitechurch et al. 1984; Parlak & Delaloye 1996; Dilek et al. 1999; Celik & Delaloye 2003; Vergili & Parlak 2005). The dykes are not deformed, which shows that they were emplaced after the deformation of the metamorphic soles, but they do not extend into the underlying mhlanges. Two 4~ ages of 90-91 Ma and three ages of 90-87 Ma were obtained, and one sample gave an age of c. 63 Ma (Parlak & Delaloye 1996; Dilek et al. 1999), which shows that most dykes were emplaced shortly after formation of the metamorphic soles. Farther east, north of the Arabian platform, two ophiolite belts are present (Fig. 1): (1) the north Arabian belt, which is thrust over the Arabian platform, to be discussed below; (2) the SE Anatolian belt, preserved in higher tectonic slices, which also include sediments, m4langes and volcanic rocks, all being thrust over the neoautochthonous cover of the north Arabian ophiolites (Aktas & Robertson 1990; Yazgan & Chessex 1991; Ydmaz 1993; Ydmaz et al. 1993; Yi~itbas & Ydmaz 1996; Robertson 2002; Beyarslan & Bing61 2000; Parlak et al. 2004). The sediments in these tectonic units show that the host basin persisted until the Late Eocene, but was tectonized already at the end of the Cretaceous. The special features of the SE Anatolian ophiolites that are important in the present context are their SSZ affinity and the fact that in Coniacian-Maastrichtian times calc-alkaline volcanic-plutonic magmas were emplaced on some of them, whereas at the same time a magmatic arc along which granitoids were emplaced formed on the southern side of the Tauride block (Yazgan & Chessex 1991; Parlak & Rizao~lu 2004; Parlak et al. 2004). Thus the SE Anatolian ophiolites record a direct link between SSZ ophiolites and a calc-alkaline magmatic arc. The north Arabian ophiolite belt. This ophiolite belt, c. 1000 km long, originated in a basin next to the northern Arabian margin (Robertson 2002, 2004; Garfunkel 1998, 2004). It includes the

306

Z. GARFUNKEL

Troodos (Cyprus), Baer-Bassit (Syria), and Kizilda~ (Turkey) ophiolites, and smaller less well-studied bodies as far east as the Cilo ophiolite (Fig. 1). The continuity of this belt as far west as Cyprus is indicated by the similarity of the allochthonous units in Cyprus and in N W Syria (Robertson 2000) and by distinct magnetic anomalies over the intervening sea floor (Woodside 1977). This belt forms the northern part of the peri-Arabian ophiolite crescent defined by Ricou (1971) (see below). The allochthonous units associated with the Baer-Bassit ophiolite and similar units further east in Turkey as well as the Mamonia complex in Cyprus contain Triassic deep-water sediments, which indicates that the north Arabian margin and adjacent deep basin were shaped already in the Triassic (235-225 Ma), when MORB-like magmatism took place (Robertson & Woodcock 1979; Robertson 1990; Malpas et al. 1992; Yllmaz 1993; A1-Riyami et al. 2000; Garfunkel 2004). The margin and adjacent part of the deep basin persisted to mid-Cretaceous times at least. The little deformed Troodos ophiolite (Gass 1980; Robertson & Xenophontos 1993, and references therein), whose base is not exposed, consists of two parts. The main northern part shows a rather regular arrangement of all units of the classic pseudo-stratigraphy (actually it inspired the Penrose definition). The smaller southern part has a complex structure and is separated from the northern part by the eastwest-trending Arakapas fault zone, interpreted as a fossil transform (MacLeod & Murton 1993). The tectonites in the main part comprise a block of medium depleted lherzolites next to highly depleted harzburgites (Batanova & Sobolev 2000). The overlying units show an SSZ affinity: cumulates formed from wet magmas that crystallized plagioclase after pyroxenes (H6bert & Laurent 1990); sheeted dykes and extrusive units consisting of a basalt-andesitedacite-rhyodacite suite and a younger Mg-rich picrite-basalt-basaltic andesite suite with boninitic affinities, both derived from different depleted IAT (Baragar et al. 1990; Robinson & Malpas 1990; Bednarz & Schminke 1994; Portnyagin et al. 1997). Boninitic rocks occur in the Arakapas fault zone (Robinson & Malpas 1990; MacLeod & Murton 1993). In places the cumulates are cross-cut by later, much less deformed, ultramafic and mafic cumulate bodies (Malpas 1990) that record a relatively late-stage magmatic activity, perhaps related to the younger volcanic rocks. Some plagiogranites yielded U/Pb zircon ages of 90-93 Ma (Turonian), compatible with the ages of fossils in sediments overlying the volcanic rocks (Mukasa & Ludden

1987). Structural studies of the sheeted dykes and extrusive units revealed important normal faulting and block tilting indicative of extension perpendicular to the dykes (now approximately east-west), but the overall pseudo-stratigraphy was not disrupted (Varga & Moores 1985; Allerton & Vine 1991). Metamorphic soles are not known, as the base of the Troodos is not exposed, but amphibolite-grade slices resembling metamorphic soles of other ophiolites in the region are exposed next to the ophiolite (Malpas et al. 1992) and they yield similar 4~ ages of 90-83 Ma (Spray & Roddick 1981). The Kmlda~ (Hatay) ophiolite (Delaloye & Wagner 1984; Lytwyn & Casey, 1993; Dilek & Thy 1998; Dilek et al. 1999; Ba[gci et al. 2005) also shows all units of the classic pseudo-stratigraphy, but they are often in fault contact. The faulting was interpreted by Dilek & Thy (1998) and Dilek et al. (1999) as recording extension during accretion of the ophiolite. The tectonites consist largely of harzburgites. The sheeted dykes and volcanic rocks are similar to those in Troodos, ranging from IAT and basaltic andesites to rocks with boninitic affinity, indicating an SSZ affinity. The Baer-Bassit ophiolite, unlike the Troodos and Klzllda[g ophiolites, is broken into slices that are imbricated with the sub-ophiolitic allochthonous units, but all the units of the pseudostratigraphy are present (A1 Riyami et al. 2000, 2002). The tectonites consist predominantly of harzburgites. The dykes and extrusive rocks consist of depleted IAT and rocks with a boninitic affinity, resembling the higher lavas of the Troodos ophiolite. The metamorphic sole yielded K/Ar ages of 85-95 Ma (Delaloye & Wagner 1984; AI Riyami et al. 2002). The presence of lavas with both normal and reverse magnetization (Morris et al. 2002) records a somewhat younger age, but this magnetization may have been acquired during late hydrothermal alteration. Obduction of the ophiolites downflexed the north Arabian margin, which produced a foreland basin in the Campanian (Yxlmaz 1993). The Klzdda~ and Baer-Bassit ophiolites are overlain by ophiolite-derived clastic rocks and shallow-water sediments of Late Maastrichtian and younger ages (Delaloye & Wagner 1984; Dilek & Thy 1998; AI-Riyami et al. 2000), which indicates exposure after obduction over the continental north Arabian platform. Farther east similar relations are documented (Yllmaz 1993). The Troodos ophiolite had a different history (Robertson 2000, 2002; Lord et al. 2002). In the Campanian it was covered mainly by thin basinal sediments, except in western Cyprus where finegrained calc-alkaline volcaniclastic deposits (the

LIFE CYCLE OF NEOTETHYAN OPHIOLIIES Kannaviou Formation) accumulated. At the same time the Mamonia complex was tectonized and contributed to a m61ange (the Moni m61ange, now south of the ophiolites) that does not contain material derived from the ophiolite. The two units were brought together in the Maastrichtian and then contributed clasts to a common cover, which in turn is overlain by lower Tertiary pelagic sediments. This records an underwater position and emplacement on the thin crust of the Levant basin. Palaeomagnetic data (Morris et al. 1998) show that the assembly of these units in Cyprus involved considerable lateral motions. Palaeomagnetic data also indicate that the main part of the Troodos ophiolite rotated c. 90 ~ counter-clockwise on a vertical axis, mostly in Campanian-Maastrichtian times (Morris 1996), various parts of the Baer-Bassit ophiolite rotated from e. 100 ~ to 220 ~ counter-clockwise during obduction (Morris et al. 2002). While these ophiolites were obducted, a basin with a complex history persisted farther north until Late Eocene times and there the SE Anatolian ophiolites formed (see above). The east Arabian ( Z a g r o s ) ophiolite belt. The eastern part of the peri-Arabian ophiolite crescent originated in a basin between Arabia and the Sanandaj-Sirjan block (Fig. 1). Because the latter may have been distinct from the Tauride block, which controlled the seaway in which the north Arabian ophiolites originated, the two parts of the peri-Arabian belt are here considered separately. The east Arabian margin was shaped by late Permian and Triassic rifting (Lippard et al. 1986; Rabu 1993). Palinspastic reconstructions (e.g. Le Pichon et al. 1988; Stampfli et al. 2001) show that afterwards a wide oceanic area developed east of the east Arabian margin. The east Arabian belt is best represented by the well-studied Semail ophiolite (Fig. 1). The sub-ophiolite allochthonous units represent a complex 300-500km wide basinal area, the Hawasina basin, that existed off the Oman passive margin and received basinal sediments until Cenomanian-Turonian times. A similar feature has not been documented further north. The Semail ophiolite, c. 500 km long and 50-100 km wide, shows the classic pseudo-stratigraphy (Lippard et al. 1986; Rabu 1993). The tectonites, predominantly harzburgites, record hightemperature penetrative deformation that was interpreted as recording flow during and after melt extraction under a somewhat irregular spreading centre (Nicolas 1989). However, some chromites in harzburgites contain inclusions

307

of basalt with an SSZ character (Schiano et al. 1997). The cumulates mostly show a crystallization order different from that of MORB (Lippard et al. 1986). The older and most voluminous part of the volcanic rocks (Geotimes series) consists of basalts and basaltic andesites transitional between MORB and IAT with a weak SSZ signature; these are considered to be related to much of the sheeted dyke complex and the cumulates (Alabaster et al. 1982; Lippard et al. 1986; Ernewein et al. 1988). These are overlain by basalts, andesites and some dacites (Alley series) with a more pronounced SSZ signature (Alabaster et al. 1982; Lippard et al. 1986; Ernewein et al. 1988; Umino et al. 1990). Locally the latter series contains boninitic flows, and such magmas also form dykes in the underlying units (Ishikawa et al. 2002). The cumulates and tectonites are often crossed by late-stage discordant and little deformed minor intrusions of wehrlites and pyroxenites, as well as gabbros, diorites and trondhjemites that are related to the younger volcanic rocks (Lippard et al. 1986; Juteau et al. 1988; Umino et al. 1990). Some of the more acid rocks sometimes intrude the sheeted dykes. In places the late intrusions show a calc-alkaline differentiation trend (Lachize et al. 1996). The ultramafic tectonites are crossed by numerous dykes (Python & Ceuleneer 2003). These include a group of troctolites and olivine gabbros that crystallized from MORB-like magmas that precipitated plagioclase before pyroxene (unlike the cumulate unit) and intruded the harzburgites while they were still hot ( > 1000 ~ A second, more widespread group consists of gabbro-norites and pyroxenites that were precipitated from highly depleted magmas with SSZ affinity and were emplaced while the host rocks were cooling. High-level plagiogranites yielded Cenomanian U/Pb zircon ages of 93.5-97.9 Ma clustering around 95 Ma (Tilton et al. 1981), whereas 4~ 39Ar ages from plagiogranites, gabbros and veins range from c. 96 Ma to e. 93 Ma (mean 94.4_+ 0.5 Ma), and granitic and dioritic rocks of the wehrlitic series yielded a mean age of 93.8+_ 0.3 Ma (Hacker et al. 1996). These ages are compatible with the presence of late Cenomanian to early Turonian (c. 95-91 Ma) fossils in sediments within the volcanic rocks, but pelagic deposition continued into the Santonian (Tippit et al. 1981). The latest volcanic rocks (Salahi unit), associated with Coniacian (89-86Ma) fossiliferous sediments, consist of alkali basalts with a within-plate character (Alabaster et al. 1982; Lippard et al. 1986; Umino et al. 1990). Metamorphic soles underlie the entire ophiolite and record peak metamorphic conditions that

308

Z. GARFUNKEL

vary from c. 13 kbar, 800 ~ to c. 7 kbar, 700800 ~ corresponding to depths that considerably exceed the thickness of the ophiolite (Hacker & Gnos 1997; Searle & Cox 1999). They yielded 4~ hornblende ages that cluster in narrow ranges of 93.5 _0.1 Ma and 94.9 _+0.2 Ma in the north and south of the ophiolite (Hacker et al. 1996; K - A t ages show a much wider scatter) whereas micas from the soles yielded slightly younger ages, recording their cooling within a few million years (Hacker et al. 1996; Hacker & Gnos 1997). Some soles are crossed by undeformed dykes. A hornblende in one dyke gave a 4~ age of 93.7+_0.8 Ma (Hacker & Gnos 1997). In the north the tectonites and the lower cumulates are cut by dykes and lenses of peraluminous granites that gave 4~ ages of 8990 Ma and a K/Ar age of 85_+3 Ma (Hacker et al. 1996; Searle & Cox 1999). Palaeomagnetic data (Perrin et al. 2000; Weiler 2000) show that that the southern part of the ophiolite rotated c. 20 ~ counter-clockwise whereas its northern part rotated c. 120 ~ clockwise after the end of magmatism, but differential rotations between portions of the northern part had taken place already during the last stage of volcanic activity. The Hawasina basin and the nearby Oman margin were considerably deformed in Late Cenomanian-Turonian to Santonian times, which led to deposition of m61anges (the Muti Formation) that contain rock fragments derived from the margin and from within the basin, but not from the ophiolite (Lippard et al. 1986; Robertson 1987; Rabu 1993). This could arise if the ophiolite was still far from the deformed area, or if it was still low-lying and little deformed, and so did not contribute clastic deposits. The nearby platform edge was fractured and eroded at the beginning of this period, and locally minor volcanism took place, but in the Coniacian (87-86 Ma) the platform edge was downflexed and by the Campanian a foreland basin of > 100 km width developed and was filled by up to 4 km of sediments in front of the advancing nappes (Patton & O'Connor 1988; Boote et al. 1990; Warburton et al. 1990; Rabu 1993). Clastic deposits derived from the ophiolite and underlying allochthon appear in the Late Campanian (at c. 75 Ma), which shows that then the ophiolite advanced already over the continental margin and was uplifted enough to be eroded. The end of obduction is marked by erosion and subaerial weathering of the ophiolite and subsequent deposition of Late Maastrichtian marine sediments over the ophiolite. The oceanic basin NE of Oman persisted after emplacement of the Semail ophiolite, and a relict still survives, but in the Tertiary

additional deformation took place (Coleman 1981; Lippard et al. 1986; Rabu 1993). A special feature of the Semail ophiolite is that it overlies strongly deformed slices of rocks that were metamorphosed at high-pressure conditions, reaching eclogite-facies conditions that were variously estimated at 500-580 ~ and 2.0-2.4GPa (Searle & Cox 1999) and 550-580 ~ and 1.2-1.6 GPa (E1-Shazly 2001), which correspond respectively to depths of 70-80 km and 40-50 km. The protoliths consist of continental margin rocks, indicating subduction of the edge of the Arabian continent, but the history of these rocks is debated (Gray & Gregory 2003; Searle et al. 2003). In the present context the important point is that the Hawasina basin was still undeformed when the ophiolite and its metamorphic sole formed, so the latter were still far (hundreds of kilometres) from the continental margin, and so their evolution until then could not have been directly related to the subduction of the edge of Arabia. The more northern east Arabian ophiolites have been little studied. The Neyriz ophiolite (Fig. 1) is rather faulted but comprises all units of the pseudo-stratigraphy (Sarkarinejad 1994). The tectonites consist of harzburgites and subordinate lherzolites, whereas the dykes and volcanic rocks have a similar geochemical character to the main part of the lavas in the Semail ophiolite, and differ only little from MORB. The Kermanshah ophiolite comprises harzburgitic tectonics, and its dykes have SSZ chemistry (Desmonds & Beccaluva 1983). These ophiolites were obducted at the same time as the Semail ophiolite and then a basin still remained NE of them (Ricou 1971; Ghazi et al. 2004). S u m m a r y . The foregoing outline shows that the Cretaceous ophiolites of the Anatolian domain and the peri-Arabian belt formed within a basin that originated in Triassic or Permian times, and thus was c. 100 Ma old when the ophiolites formed. These ophiolites are constrained to have formed and evolved in a narrow time interval (often c. 10 Ma), although they originated in distinct seaways. Where metamorphic soles have been dated they mostly yield CenomanianTuronian ages, very close to, or even overlapping, the age of the ophiolite rocks. The sheeted dykes and extrusive rocks in these ophiolites have an SSZ geochemical signature, which is strongly expressed in the Anatolian and North Arabian ophiolites, but weakly in most of the lavas of the Semail and probably Neyriz ophiolites, although the younger lavas and dykes in the Semail ophiolite have a well-expressed SSZ signature. In the latter, some rocks that display a calc-alkaline

LIFE CYCLE OF NEOTETHYAN OPHIOLITES differentiation trend are present. In the east Anatolian ophiolites this trend is well displayed. Also noteworthy is the presence of undeformed dykes cutting the metamorphic soles of the Tauride ophiolite and the Semail ophiolite. Final obduction took place some 20 Ma after ophiolite formation, but the host basins persisted for variable periods (often 30-40 Ma, whereas the basin next to the Semail ophiolite has not closed yet).

Place of ophiolite development in the history of the basins To highlight the evolution of ophiolites in relation to the host basins it is convenient to divide their history into several stages. Shervais (2001) divided the history of SSZ ophiolites, mainly based on their petrological evolution, into the following stages: birth, formation of the bulk of the ophiolite; youth, formation of components derived from depleted sources; maturity, formation of arc-like components; death, end of magmatism and thrusting over the nearby basin; resurrection, emplacement of the ophiolite. Here the emphasis is on the ophiolite-host basin relations rather than on the petrological evolution, so a somewhat different division is used: (1) the pre-ophiolite stage, which produced the palaeogeographical setting in which the ophiolites originated; (2) the ophiolite formation stage; (3) the ophiolite tectonization and obduction stage, which may overlap the previous stage; (4) the final, post-obduction stage, ending with the closure of the basin. Stage (2) includes the first three stages of Shervais (2001), which he noted may overlap to some extent in time and space, and stage (3) overlaps his two subsequent stages. Pre-ophiolite stage The foregoing outline shows that the ophiolites considered here formed when the host basins were already well developed. Although these basins were destroyed, it is possible to constrain their widths and the spreading rates of ridges that existed in them. It is expected that after initial rifting the basins widened by nearly symmetrical sea-floor spreading along ridges, similar to the present ridges. As spreading progressed the ridges moved away from the passive margins and remained in the middle of the basins as long as there was no subduction (Fig. 2). If subduction began on one side of a basin, the ridge could move farther away from its other margin and could even be subducted. Thus if a ridge survives some time after basin initiation, its average long-term

309

spreading rate is constrained by the width of a basin at that time (Fig. 2). If faster spreading ridges existed, they could not persist, because that would imply production of an area wider than the basin. Such fast-spreading ridges would either be subducted or their spreading will have stopped or slowed down. Palinspastic plate reconstruction at c. 105 Ma, shortly before the ophiolites formed (Fig. 3), allows us to apply these considerations to the basins that hosted the Cretaceous ophiolites. It is seen that wide oceanic seaways existed then in the region, implying considerable sea-floor spreading since their inception. Along any transect the total width of the seaways is given approximately by the convergence between Eurasia and ArabiaAfrica (taking into account that a part of the convergence (100-200 km(?)) was taken up by shortening of continental areas). The convergence across the Anatolian domain was c. 2400 km since 105 Ma ago and c. 1800 km since 92 Ma ago, using the plate kinematics of Miiller & Roest (1992). Thus when the north Arabian ophiolites formed the host basin could have been 500800 km wide, leaving enough room for similar seaways farther north. If subduction had not occurred in this basin, then the average rate of its north-south opening was 0.5-0.8 cm a -1 (the actual opening could have been in a somewhat different direction). To accommodate this opening (and the opening of more northern seaways), together with the eastward motion of AfricaArabia relative to Europe, the subduction must have been oblique somewhere in this region before 105 Ma (Fig. 3). Farther east the motion of Africa-Arabia relative to Eurasia since Mid-Jurassic times reduced the oceanic area east of Arabia (Fig. 3; Le Pichon et al. 1988; Miiller & Roest 1992; Stampfli et al. 2001). This requires (oblique) subduction somewhere on the NE side of this oceanic area, perhaps along the Sanandaj-Sirjan block, where Late Jurassic and Cretaceous igneous activity took place (Berberian & Berberian 1981; Berberian & King 1981). The total convergence between Oman and Eurasia was c. 3200 km since 105 Ma ago and c. 2300 km since 92 Ma ago. Thus, taking into account the presence of other seaways in this area (McCall 1997), the basin hosting the Semail ophiolite could have been up to 2000-2200 km wide when this ophiolite formed. Thus, if a ridge survived in this basin for 120 Ma (since the Late Triassic) its average spreading rate must have been less than c. 3.5 cm a -1, or else it would have been subducted. The history of the Pindos basin is not well constrained, but it could not have been much wider than 500-800 km, given the space between the

310

Z. GARFUNKEL

(a)

(b)

~. "..

~

'.'~

w _++

w +

+

W

2~-

V< t

Fig. 2. Constraint on spreading rate of ridges. (a) Situation in ocean in which there is no subduction. (h) Situation in oceanic basin with a subduction zone on one side: ridge does not remain in middle of basin. Top, map view; bottom, cross-section, w, width of the basin at time t; v, half-spreading rate.

major continents in that area. Thus, if a ridge persisted in this basin for 60 Ma its average spreading rate could not have exceeded 1.62.7 cm a -1, if a subduction zone existed in this basin before ophiolite formation (otherwise longterm average spreading rate would have been half these rates). In summary, these considerations show that the ophiolites considered here formed in 60100 Ma old basins that were 500 km to a couple of thousand kilometres wide, which amplifies the conclusion of Shervais (2001) that ophiolites form in wide basins. The ridges that produced these basins must have been slow spreading if they persisted to the time of ophiolite formation. Otherwise they would have produced areas wider than the basins (alluded to by Robertson & Woodcock 1979), so either they would have been subducted or their activity would have been intermittent. The latter option, although not impossible, is difficult to evaluate in basins that no

longer exist, but there do not seem to be any firm arguments for the general applicability of such a scenario.

Ophiolite formation stage The setting in which the ophiolites discussed above formed is constrained by their regional framework, their structure, and their chemical signature.

The regionalframework. The regional framework of the basins hosting the Cretaceous ophiolites considered here is constrained, as discussed above, by the motions of Africa-Arabia relative to Eurasia. In the present context the important feature, accepted in all plate knematic models (e.g. LePichon et al. 1988; Miiller & Roest 1992), is that these motions changed considerably c. 105 Ma ago or somewhat later (Fig. 3), i.e. slightly before the formation of the Anatolian

LIFE CYCLE OF NEOTETHYAN OPHIOLITES

311

I

20 ~

~40 ~ Hell.-Din.

\

~o'~

~_.\

/o

~

'~,

67 Ma

\

AFRICA\ARABIA

Ma ,,~ ,,.oe

92 Ma

0

~05 Ma

1000 km

i

"~'~,"

i

Approx. Ma 0ol -

Fig. 3. Palaeogeographical situation about 105 Ma ago. Position of Africa-Arabia relative to Eurasia at various times is after Mfiller & Roest (1992). Coordinates relative to Eurasia are shown for reference. Positions of micro-continents are approximate, shown to outline the Neotethyan seaways (Iranian blocks are shown schematically based on Seng6r et al. (1988) and McCall (1997)). Also shown are possible positions of new subduction zones along which the ophiolite belts mentioned in the text formed. Older subduction zone are shown by thicker lines.

and peri-Arabian ophiolites. Figure 3 shows that in the Anatolian domain the motion of AfricaArabia changed from nearly parallel to its northern margin to convergence at an angle of c. 45 ~ to this margin at rate of c. 4 cm a -1. The convergence was probably partitioned between the seaways in this region, so new subduction zones should have formed in them. East of Arabia convergence at an overall rate of > 6 cm a -1 continued, but its direction changed by c. 30 ~ so that it became almost perpendicular to the A r a b i a n - O m a n continental margin. The plate motions across the Pindos Ocean are not known, but the initiation of opening of the Ligurian seaway on the western side of Apulia shortly before c. 170 Ma (Lombardo et al. 2002) would probably change the motion across the Pindos Ocean close to the time of formation of the Hellenic-Dinaric ophiolites. In a wider context this may be related to the opening of the central Atlantic at that time

(Smith 2004). Thus, the ophiolites considered here formed close to the time of changes in the motions of the plates or micro-continents bordering the host basins. Internal structure o f the ophiolites. The interpretation that ophiolites represent new crust that formed by some type of sea-floor spreading implies that their width perpendicular to the sheeted dykes measures the amount of spreading during their formation. Thus the largest bodies (Troodos, Semail, and Vourinos-Pindos) record spreading of _> 100 km, whereas smaller ophiolites record spreading of tens of kilometres. The spreading rates are difficult to determine, however, because dating is not precise enough. Therefore indirect indications were used. Nicolas (1989), who favoured ophiolite origin along MORs, proposed that differences in spreading rates controlled the structure of various ophiolite

312

Z. GARFUNKEL

types. As the factors that control generation of new crust in both MOR and SSZ settings are likely to be similar, it is instructive to compare the structure of ophiolites with that of oceanic crust, regardless of the setting envisaged for their production. Detailed studies of the structure of oceanic crust (Dilek et al. 1998; Karson 1998) show that the crust that formed along slow-spreading ridges (e.g. in the Atlantic Ocean, spreading rates 2-4 cm a -~) has an irregular structure, often as a result of faulting, so that its components have variable thicknesses. This is interpreted as resulting from limited and intermittent magma supply, which does not allow fast enough construction of new crust, so a part of the spreading is accommodated by slip on normal faults and shear zones (amagmatic spreading). The lherzolitic ophiolites that have such a structure (e.g. the west Albanian ophiolites) could therefore have formed by slow spreading. It is noteworthy that similar features were observed also in the Mariana back-arc basin (Ohara et al. 2002), supporting the idea that the same factors control spreading in SSZ and MOR settings. On the other hand, the layered structure of the harzburgitic ophiolites with a complete Penrose pseudo-stratigraphy resembles the structure of crust formed along the fast-spreading East Pacific Rise (spreading rates 8-12 cm a-l). Thus the Semail ophiolite is interpreted to have formed by fast spreading, perhaps 10cm a -~ (Nicolas 1989; Dilek et al. 1998). The Troodos and Kmlda(g ophiolites are faulted but still show the Penrose pseudo-stratigraphy better than lherzolitic ophiolites, suggesting formation by intermediate spreading, perhaps 4-6 cm a -~. These are, however, only approximate estimes. The important point in the present context is that such spreading rates could be maintained only during time intervals that are considerably shorter than the life span of the basins hosting these ophiolites, for otherwise areas wider than these basins would be generated. Thus a spreading rate of 10 cm a -~ could generate the entire width of the basin hosting the Semail ophiolite in c. 20-25 Ma, whereas a spreading rate of 2-3 cm a -~ could generate the entire width of the basin hosting the Troodos and Klzllda~ ophiolite in c. 20-30 Ma. Thus, ophiolite generation appears to mark short episodes of fast crustal accretion, considerably faster than the above estimates of the long-term average spreading rates in the host basins.

The sites of ophiolite generation. The sediments directly overlying the ophiolites record formation in an intra-oceanic setting, far from continental margins (Robertson 2004). The nature of the protoliths of the metamorphic soles (basinal

sediments and volcanic rocks, including MORBlike rocks) and the presence of such components in the sub-ophiolitic allochthons (Robertson 2002, 2004) also indicate that the ophiolites were thrust over intra-oceanic areas.

Geochemical signature. The foregoing review shows that the geochemical signature of igneous rocks in the harzburgitic ophiolites resembles that of rocks in SSZ rather than in MOR settings. The strength of the SSZ fingerprint is variable, however. At one extreme, in the Semail ophiolite, a considerable part of the lavas and cumulates formed from magmas transitional between MORB and IAT, whereas only the less voluminous younger volcanic rocks and intrusions have a well-expressed SSZ signature. In other ophiolites (e.g. in the Hellenic-Dinaric belt), rocks with both MOR-like and SSZ fingerprints are present, sometimes alternating in time. Generally, rocks with a boninitic affinity typical of an SSZ setting tend to appear late in the ophiolite history. It is the SSZ geochemical fingerprint of many ophiolites (which seems to characterize most ophiolites: Shervais 2001) that is the main argument for linking their formation with young subduction zones. If such ophiolites formed along MORs it is difficult to understand why the rocks normally found along the present-day ridge do not dominate, or at least are common, in ophiolites. Models o f Jormation The foregoing considerations show that the Semail, Troodos and other ophiolites considered here formed in special short-lived events (unless the spreading rates during their formation were much overestimated). Moreover, the fact that both the Jurassic and Cretaceous ophiolites considered here formed within short time intervals in belts that are often c. 1000 km long or even longer also points to special events, which, given the regional framework, took place close to times of changes in plate motions. This makes it difficult to consider ophiolite formation as a result of normal long-term spreading of ridges in the host basins. Thus, interpreting ophiolites as originating along MORs is not as simple a model as appears at first sight. These features fit, however, the alternative interpretation of the harzburgitic ophiolites as originating in an SSZ setting that was based primarily on their geochemical fingerprint. The common occurrences of IAT-like rocks, which are characteristic of SSZ settings but are uncommon in MORs, and especially the occurrence of rocks with a boninitic affinity, which are known

LIFE CYCLE OF NEOTETHYAN OPHIOLITES

313

Fig. 4. Development of ophiolites in an SSZ setting, from inception of a new subduction zone to obduction. (See text for discussion).

only in young intra-oceanic forearcs, point at an origin in an SSZ setting (Pearce et al. 1984), which is now widely accepted. Analogy with western Pacific forearcs (Casey & Dewey 1984; Hawkins et al. 1984; Leich 1984) inspired later interpretations (e.g. Stern & Bloomer 1992), and is generally followed here. Hence the formation of ophiolites in short time intervals (c. 10 Ma) along long belts is considered to express distinct 'ophiolite events' in which strips of crust were produced along new subduction zones that formed in response to changes in plate motions (which was suggested already by Moores (1982), from a different point of view). These considerations provide the framework for the following attempt at interpreting the features of the harzburgitic ophiolites considered here (the lherzolitic ophiolites will be discussed later). The polarity of the subduction is not directly constrained, but it can be assumed, following Coleman (1981) and Moores (1982), that during emplacement of Tethyan ophiolites the descending slabs dipped away from the passive margins on which the ophiolites were obducted. Intraoceanic subduction with such a polarity, but not the opposite one, will consume the basinal areas between these margins and the site of ophiolite formation. Assuming such a polarity during ophiolite formation allows the same subduction zone to serve for both ophiolite formation and obduction. This interpretation fits the ophiolites

considered here, as there is no record of subduction along the passive margins on which they were obducted (but does not necessarily apply to Cordilleran-type ophiolites). It should be noted that if subduction along an active margin brought an ophiolite to this margin, then another subduction zone should be invoked for the intra-oceanic formation of the ophiolite. The mechanism by which new subduction zones are initiated is not known, and perhaps they arise in several different ways (Casey & Dewey 1984; Stern 2004). Given the original intra-oceanic position of the ophiolites considered here, the subduction zones over which they formed must have originated within the host basins. Possibly an intact part of a basin failed under the influence of externally applied forces, as in the central Indian Ocean (Bull & Scrutton 1992; Krishna et al. 1998). At first, a system of thrusts develops, but later thrusting is localized to form a single subduction zone (Fig. 4a). Another possibility is that density differences along transforms may influence the process (Casey & Dewey 1984). However, as transforms are expected to be in isostatic equilibrium, as is the case today, there is no force to cause spontaneous foundering of the older and heavier sides. If such foundering occurred, then in the early stages the isostatic equilibrium would be greatly disturbed in a manner that will oppose the foundering (Fig. 5). Thus, failure induced by externally applied force, such as may arise during changes in plate

314

Z. GARFUNKEL

(a) Formation of new subduction zone

Ridge failure

New fracture

::::::::::::::::::::::::::::::::::: OR

+ + + ]

(b) Formation of ophiolite

Ophiolite

Sea level ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: -.. - _ _ _ _ ~

F:=-============:=:==_--'~=====_=_=IIIIIIIIIIIIIIIIIIIIIII~IIIIIIIIIIIIL=====-=. .

~-----------ST-__ (el Formation of metamorphic sole and initial thrusting 9

~:--:-:::---:-::

.

M T

sole ~,Late$

. . . .

z z z z z z 7. z

~L"-.~

..

0!

lO0,km

OR p:-z:::::-:S:':z;::::z:-_~~llllllllllllllllllllllllllllllllllllllll::z--z:=:l

--'~.-

(d) Obduction

km-{- -~- 4- -_-__~~ITIIlIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII-_---:-_--~=........ ......

+ + + IAccretionary-. . . . . . . . . . . . ' - ? ~ ' : ~ re#.e,~ prnsm

~-I-+-/-+++ff_

/. /. / . _ / . _ + + + + + . ~

~v-I-+ff"+++-/--/.../.

f./_++++

Fig. 5. Problems involved in the formation of new subduction zones along previous plate boundaries. (a) Formation along a ridge with small-offset transforms (map view): the subduction zone cannot follow only the weak zero-age crust along the ridge crest, but it must cut across transforms. (b) Formation along a ridge offset by long transforms (map view): the subduction zone cannot follow zero-age crust; it is also noteworthy that the lithosphere is older and heavier on alternating sides of each transform. (c) Formation of a new subduction zone at the expense of a long transform (cross-section): the older and colder side (+) may tend to sink (stage I), but in order to begin to subduct (stage III) it must pass through stage II, in which isostatic rebound will occur and prevent further sinking of the older lithosphere unless there is an externally driving force. MT, metamorphic sole.

motions, is more probable than spontaneous foundering. It should also be noted that the formation of new subduction zones of 1000 km length or more (the length of some ophiolite belts) requires the existence of much longer transforms, because the polarity of density reversals changes along the transforms. It is not clear how to place such long transforms in the basins hosting the ophiolites considered here, but this requires further study. A change in plate motions

may also cause failure along a ridge because this is the weakest place within the basin (e.g. Coleman 1981; called 'ridge collapse' by Robertson 2004). However, ridges are usually not simple zones of weakness but consist of segments separated by transforms, so their failure is more complicated than simply disrupting a continuous weak ridge crest, unless transform offsets are small (Fig. 5). It also involves to difficulty of subducting hot and buoyant ridge crests.

LIFE CYCLE OF NEOTETHYAN OPHIOLITES The Hellenic-Dinaric ophiolites may have formed by failure (collapse) of a ridge in the Pindos Ocean where there is no indication of former subduction. This allows the western part of the belt to be a remnant of a ridge, but it does not fit the absence of a clear east-west age difference across the belt, suggesting a different origin (see below). In the seaway between Arabia and the Tauride block, where the north Arabian and SE Anatolian ophiolite belts originated, two subduction zones may have formed: one off the Arabian margin along which T r o o d o s - K m l d a ~ ophiolites formed, and one farther basinwards, along which the island arc next to the Tauride block formed (Fig. 3), so perhaps only one of them initiated by ridge failure, but they could have formed differently. The palaeomagnetic data from Troodos indicate that the southern subduction zone was oriented close to east-west. Other subduction zones are assumed to have formed farther north of the Tauride block to account for the Tauride and the more northern Anatolian ophiolites. In the basin east of Arabia, which was shrinking since Jurassic time, a subduction zone probably existed already along the Sanandaj-Sirjan block (Fig. 4). However, the continuation of magmatic activity along this block into the Late Cretaceous (Berberian & Berberian 1981; Berberian & King 1981) suggests that this subduction zone remained active until that time, so it was not available as the site of formation of the east Arabian ophiolites. Therefore another subduction zone is assumed to have formed inside the basin (Fig. 3). In all these cases the new subduction zones could form by ridge failure, provided that ridges still survived in the basin, which is uncertain, and provided that this is a viable mechanism; otherwise, failure of the basin floor should be assumed. Hacker & Gnos (1997) suggested that the Semail ophiolite formed along a transform fault. This must have trended closer to east-west than to north-south, based on the palaeomagnetic data, and should have been comparable in length with the Semail ophiolite. It is not clear how such a fault fits into the history of the host basin. The production of ophiolites by some form of sea-floor spreading in an SSZ setting is interpreted as a consequence of retreat (roll-back) of the descending slabs (Fig. 4b; Elsasser 1971 (who called this behavior 'retrograde subduction'); Chase 1978). This is a common behaviour of subducted lithospheric slabs (Garfunkel et al. 1986), being a result of their negative buoyancy, which adds a downward component of motion to their dip-parallel descent, so that the slabs descend at a steeper angle than their dips (the dip need not change). This causes the hinges where

315

the lithosphere bends into the mantle to retreat, i.e. to move opposite to the slab dip, at rates of a few centimetres per year (Garfunkel et al. 1986; updated models of plate motion do not change this conclusion). Mantle flow may also promote slab retreat (Flower & Dilek 2003). In the present context it is important to note that slabs in young subduction zones (e.g. next to the Scotia Sea) also retreat at fast rates, and this also occurred along the western Pacific subduction zones immediately after their formation (Stern & Bloomer 1992). Thus if in a basin that shrinks at a rate of say 1-2 cm a-', slab retreat was faster, say 3 cm a -~, then sea-floor spreading at a rate of 2-1 cm a -~ will occur behind the trench, so there a 100 km wide strip of new crust will form in 5-10 Ma. In the case of the Semail ophiolite, considered to have formed by spreading at a rate of c. 10cm a -1, even faster slab retreat must be envisaged. Magma generation along the ophiolite-related subduction zones can be related to models of the dynamics and thermal state in the mantle wedge of subduction zones. These show that viscous drag of the descending slabs induces a corner flow in the wedges and that this flow is much enhanced by slab retreat, because material is sucked into the wedge to fill the space left behind the slabs (Fig. 6; Garfunkel et al. 1986; Davies & Stevenson 1992; Kincaid & Sacks 1997), and it is further modified when spreading behind the trench diverts some material towards the site of spreading (Ribe 1989). This flow replenishes the mantle wedge with hot material, which counteracts the cooling effect of the descending slab and thus allows continuing magma generation. However, the region close to the trench remains stagnant and cold, and there magmas are not generated (Davis & Stevenson 1992; Kincaid & Sacks 1997). Next to m o d e m trenches this cold amagmatic region is 150-200 km wide (Gill 1981) and igneous activity occurs only where the slab depth exceeds 65-130 km (England et al. 2004). In nascent subduction zones this cold region can be smaller, especially if subduction begins in young sea floor (e.g. near ridges) where the mantle is hot (Kincaid & Sacks 1997). In such cases igneous activity may occur closer (50 km or less?) to the trench above shallow (30 km?) parts of the slabs. Magmas with an SSZ fingerprint are considered to be derived from mantle sources that had already been depleted to varying degrees by previous partial melting events (e.g. M O R B production), and were remelted as a result of addition of slab-derived aqueous fluids that carry a 'subduction component' of various trace elements (Gill 1981; McCulloch & Gamble 1991;

316

Z. GARFUNKEL

Fig. 6. Model of flow and magma generation in the mantle wedge in subduction zones. (See text for discussion.)

Hawkesworth et al. 1993; Pearce & Peate 1995). Boninitic-like lavas are derived from sources that are more depleted than the IAT sources (Pearce et al. 1992). In all cases, the addition of water lowers the melting temperature and is essential for promoting magma generation from otherwise refractory sources. The flow system in the mantle wedge can entrain depleted peridotites from beneath the overriding plate (Fig. 6), where they were left behind after having contributed partial melts to generate the overlying crust. However, to be hot enough to flow and to melt again, such material must be derived from some depth beneath the Moho. More fertile mantle coming from greater depth can also be entrained by the flow and brought into the mantle wedges. Alternatively, spreading behind the subduction zone may also occur in a proto-back-arc basin (before a real arc develops) into which normal mantle material rises and melts to form magmas similar to MORB (similar to present-day back-arcs: Hawkins 1995, 2003), and then the still hot and ductile residue can be swept towards the narrower part of the wedge, where it is hydrated and remelted. This implies that the ophiolites represent only a part (that was obducted) of the zone of magma generation next to newly formed subduction zones. This picture is probably oversimplified and requires further elaboration. The occurrence side

by side of various magmas derived from sources that were depleted to different extents (e.g. Troodos, Semail, Pindos) strongly suggests that the mantle wedge is a heterogeneous region, probably on the scale of a few kilometres to several tens of kilometres, which results from the flow being more complex than suggested by simple models. The tendency of magmas with a strong SSZ character, and especially with a boninitic affinity, to appear in advanced stages of construction of the ophiolites considered here (and in many others as well: Shervais 2001) raises the possibility that diapiric upwellings that promote decompression melting of hydrated refractory mantle become more important as time passes (Pearce et al. 1992). These considerations allow us to assess the position of the lherzolitic ophiolites that occur along the Hellenic-Dinaric ophiolite belt. Their structure, which differs from that of the harzburgitic ophiolites, and the MORB-like character of their extrusive rocks raise the possibility that they formed along a slow-spreading MOR, perhaps just before it failed to produce a new subduction (Robertson & Shallo 2000), and when they became situated behind the newly generated trench they were invaded by SSZ magmas (Fig. 5b). Inserguieux-Filippi et al. (2000) suggested that a MORB source may continue to rise and melt some time after ridge failure and inception of a new subduction zone. Although this

LIFE CYCLE OF NEOTETHYAN OPHIOLITES model does not consider slab retreat and assumes a deeply rooted hot rising column beneath the ridge, which may not be realistic, the process envisaged deserves further evaluation. Still, the question arises of why such situations do not occur elsewhere. A more difficult problem is the absence of a resolvable age difference between the western and eastern ophiolites. These questions can be resolved if the western ophiolites formed in a proto-back-arc, i.e. by spreading some distance west of the subduction zone in a setting where now MORB-like magmas sometimes form (Hawkins 1995, 2003), whereas the eastern ones formed closer to the site of subduction. It remains to be seen whether the geochemical data allow such a model. If they do, then it is likely that such a distal part of the subduction system existed in other cases as well, but usually this not preserved in the obducted ophiolites. The Semail ophiolite reveals additional complexities. On the one hand, the chemical fingerprint of the sheeted dykes and extrusive rocks, especially the younger ones, points to formation in an SSZ setting, and this is supported by the presence ofboninitic volcanic rocks and by dykes with a pronounced SSZ character in the mantle tectonites. On the other hand some dykes that intruded the tectonites while they were still hot (1100-1200 ~ were derived from MORB-like magmas, and unlike the cumulates display the crystallization order of MORB (Python & Ceuleneer 2003). Moreover, the Semail ophiolite somewhat resembles the crust of the Hess Deep in the Pacific Ocean (Francheteau et al. 1990). These observations raise the question of the temporal and genetic relations between tectonites and the overlying parts of ophiolites, and whether all the parts of ophiolites formed simultaneously. Such a question was also raised regarding the Troodos ophiolite (Thy & Esbensen 1993). Alternatively, one may envisage that the position of the ophiolites relative to the subduction zones changed during their construction. In summary, an origin of ophiolites in an SSZ setting next to newly formed subduction zones fits their composition and petrography, and provides a framework for their interpretation. In particular, their formation can be related to the geometry and thermal-mechanical processes in the mantle wedges above the subducting slabs, although more needs to be known about the complexities of magma formation in such settings. T e c t o n i z a t i o n a n d o b d u c t i o n stage

The end of magmatic construction marks the transition to a new stage in ophiolite development in which tectonic rather than igneous

317

processes dominated their evolution as they approached the site of future obduction. This involves detachment from the original substrate, underplating by metamorphic soles, and internal deformation. Of these, the metamorphic sole formation is the best known, and probably among the earliest events. In most cases the metamorphic soles are not crossed by igneous rocks, indicating emplacement after the end of magma addition to the ophiolites. In some cases, such as the Semail and Tauride ophiolites, dykes intrude the sole and the overlying tectonites, but they account for limited and last SSZ magmatic additions to the ophiolites. Because the protoliths of the soles originated in the host basins, their emplacement beneath the ophiolites implies that the latter were detached from their substrate. This could allow displacement of the ophiolites away from the SSZ of magma generation and supply, and thus should follow closely the end of this supply. To extend under the entire width of the up to 1015 km thick ophiolites, the detachment surfaces should be rather flat (Fig. 4c), but cases such as the Semail ophiolite whose base cuts across the pseudo-stratigraphy (Boudier et al. 1988) show that they need not be horizontal. The detachments are thus distinct from the dipping tops of the descending slabs over which the ophiolites originated but are new fractures behind the trench (Fig. 4c), as recognized by Jones et al. (1991). To explain the timing of metamorphism close to the end of ophiolite construction and the inverted metamorphic gradients of the soles, they were interpreted to have formed when the still hot ophiolites were thrust over the adjacent basins where the sole protoliths originated. This is easily envisaged for ophiolites that originated along MORs, provided ridge failure and thrusting begin immediately after ophiolite production (Coleman 1981; Spray 1984). However, the finding that the igneous protoliths of the soles differ from the ophiolite rocks is incompatible with this model, because ridge spreading is expected to be symmetrical. This is an additional strong argument against a MOR origin of the ophiolites considered here (Searle & Cox 1999; Robertson 2002, 2004). Such a scenario can be envisaged also for ophiolites that originated in an SSZ setting, explaining their displacement away from the area of SSZ magma supply, but it encounters difficulties (e.g. Wakabayashi & Dilek 2000). To metamorphose the basin floor that they override the ophiolites must remain sufficiently hot, which requires very fast thrusting. Hacker et al. (1996) and Hacker & Gnos (1997) inferred 20 cm a -1 for

318

Z. GARFUNKEL

the Semail ophiolite, which considerably exceeds plate velocities of Neotethys (and most plate velocities in general) and thus is problematic. Another problem is that some soles record pressures corresponding to depths that considerably exceed the thickness of ophiolites. In Albania and Oman pressures reaching 0.9-1.2GPa and 0.7-1.3 GPa were found, corresponding to depths of 28-37 km and 2 2 4 0 km, respectively (Dimo-Lahitte et al. 2001; Searle & Cox, 1999). A sole of a Tauride ophiolite records T > 560 ~ P > 8 kbar, corresponding to a depth of c. 30 km or more (Dilek & Whitney 1997). An origin at such depths implies considerable thinning of the ophiolite after sole formation (Wakabayashi & Dilek 2000), but this has not been documented. Alternatively, the soles were metamorphosed in the subduction zone and then were exhumed and underplated the ophiolites (Fig. 4c), which explains the depth of metamorphism and also dyke intrusion after sole formation. Only material that was subducted during or shortly after ophiolite formation is expected to be exhumed, but not material that subducted earlier and descended to considerable depth, which explains why the metamorphism of the soles occurred later than ophiolite generation (see Wakabayashi & Dilek 2000). The exhumation mechanism remains obscure, however. Thus, both scenarios face dificulties, although the second is perhaps less problematic. Shervais (2001) proposed that the soles formed when the ophiolites overrode a ridge, but in the cases considered here this remains unproven. It is also unlikely that ridges are generally available to be subducted shortly after ophiolite formation, although perhaps this may happen in some places. In any case, after detachment and sole emplacement the basins between the ophiolites and the margins on which they were obducted were eliminated, which is generally ascribed to continuing subduction and slab retreat (Fig. 4). In the time intervals between the formation of the ophiolite considered here and their obduction (15-20 Ma) subduction at rates of several centimetres per year can consume basinal areas that are a few hundred kilometres wide. Probably, the ophiolites were internally deformed during this stage, as recorded in Oman (Boudier et al. 1988; Yanai et al. 1990), but insufficient studies in other cases do not allow us to outline a general picture. There is no clear record of the expected magmatism related to the subduction in this stage, though the late calc-alkaline rocks in the east Anatolia belt and the Kannaviou Formation of Cyprus may have originated in this setting. Apparently, the zone behind the ophiolites where such activity could occur is not obducted and its

record is lost. During this period the ophiolites should have formed the leading edges of the overriding plates, although no longer above sites of magma generation. As the ophiolites advanced they bulldozed the basin floor and the continental slope, and shed debris into the adjacent basin, all of which became stacked into the allochthonous subduction-accretion complexes (accretionary prisms) at their base (Fig. 4d). Palaeomagnetic data from the Semail, Troodos and Baer-Bassit ophiolites revealed large (90 ~ and more) rotations on vertical axes. A portion of the Semail ophiolite began to rotate already during the last stages of volcanism, but in the other cases it can only be established that at least in part the rotations took place during emplacement, indicating severe disruption. The causes and mechanism of the rotations may be related to oblique convergence, but further study is required to fully elucidate the process. Because of the lack of palaeomagnatic data from other ophiolites it is not clear whether this is a widespread feature of ophiolite obduction. As the ophiolites and the underlying allochthonous slices were obducted, their weight is expected to downflex the regions in front of them, producing foreland basins (Fig. 4d). Such foreland basins were documented in front of the Greek and Albanian ophiolites and in front of the north Arabian ophiolites, and over the Oman margin. In the first case only deepening was observed, indicating starved foreland basins, whereas in the latter cases the flexural lows were filled with sediments. As such basins extend some 100 km in front of the tectonic loads, depending on the strength of the flexed plate (Garfunkel & Greiling 2002), their formation over the continental margins records the final approach of the ophiolites. It is noteworthy that the top of a continuous flexed plate slopes down beneath the tectonic load and is not depressed in front of the latter. In such cases the ophiolites have to climb up this slope during obduction, which suggests pushing from behind. However, in reality the situation may be more complex, e.g. along the Oman margin where considerable faulting and deformation took place (Robertson 1987). In addition, there and in western Turkey exhumed high-P metamorphic rocks were emplaced not long before ophiolite obduction, although the timing of exhumation is not well constrained (Okay et al. 1998; Searle & Cox 1999; Gray & Gregory 2003; Searle et al. 2003). In these cases the development of the continental margins was much more complex than just being overrun by ophiolites and it cannot be described by simple flexure of continuous plates. The question then arises as to whether similar processes operated in

LIFE CYCLE OF NEOTETHYAN OPHIOLITES other places as well, but are hidden at depth. If this is the case, the obduction process may be considerably more complicated than envisaged in Figure 4d. In summary, although the scenario outlined here describes the end of ophiolite construction and its subsequent displacement toward the sites of obduction in terms of known processes, the above considerations also reveal several problems. This shows that in some cases at least the processes may be much more complicated than shown in Figure 4.

Post-obduction stage The ophiolites examined here were obducted over passive margins some time before the final closure of the host basins, which may persist for tens of millions of years. This stage of the host basin history is beyond the scope of the present work. It is only noted that the final continental collision that eliminated their remnants may deform the ophiolites and lead to their thrusting over the colliding blocks, which determines the final structural relations.

Discussion and conclusions The foregoing considerations highlight some general features of the ophiolites examined here and their relations with the histories of the host basins. (1) The ophiolites originated in mature basins that probably formed by slow spreading (0.5-3.0 cm a-l). If faster-spreading ridges existed in the basins, then they must have been subducted or were no longer active by the time the ophiolites formed, or else such ridges would have produced areas wider than the basins. Changes in spreading rate could also occur, but there are no direct constraints on this possibility. (2) The ophiolites formed in intra-oceanic positions by some form of sea-floor spreading. Spreading rates of the ophiolites with harzburgitic tectonites, which are inferred from comparison with oceanic crust, could produce the entire width of the host basins in 10-30 Ma, so they could be maintained for limited periods only. Production of these ophiolites thus signified changes in the evolution of the host basins and could not have been a long-lived process. (3) The ophiolites examined here occur in belts, and those in each belt formed in short time intervals (up to 5-10 Ma), ophiolite events, shortly after changes of the direction and/or rate of plate motions.

319

(4) The geochemical signature of the ophiolites with harzburgitic tectonites points at an origin in an SSZ setting. In view of the previous inferences the ophiolite events are therefore thought to signify the formation of new subduction zones following changes of plate motions. These ophiolites formed by spreading behind new trenches as a result of slab retreat, which also influences the dynamic and thermal structure of magma generation in the mantle wedges. (5) The less common ophiolites with lherzolitic tectonites and with a MORB-like geochemical signature formed near those with an SSZ signature and were invaded by magmas with an SSZ character. They may have formed along ridges just before the latter were disrupted. Alternatively they may have formed alongside the ophiolites with an SSZ signature in a setting analogous to back-arc basins (although there were still no true arcs). (6) Shortly after formation, the ophiolites were detached from their substrate and displaced from the zone of magma SSZ magma supply. Metamorphic soles were emplaced beneath the detachments, and in some cases were intruded by dykes. The way in which this took place remains problematic. Later the basins between the ophiolites and the margins on which the latter were eventually obducted were consumed by continuing subduction and slab retreat; the ophiolites became the leading edges of the overriding plates, and eventually were thrust over accretionary prisms built of materials scraped off the floor of the consumed basins and margins on which the ophiolites were obducted. (7) The weight of the approaching ophiolites and the underlying allochthonous units should downflex the margins on which they were emplaced. In simple flexure models the underlying plate slopes basinward, requiring up-slope pushing of the obducted ophiolites. The real situation may be much more complex, involving fracturing of the margin, and in some cases also emplacement of high-P metamorphic rocks, which limits the applicability of simple flexure models. The crucial point in the above considerations is that the harzburgitic ophiolites originated in an SSZ setting, soon after formation of the subduction zones. This is based primarily on petrological evidence from many such ophiolites worldwide (Shervais 2001) and is strongly supported by the analogy with western Pacific

320

Z. GARFUNKEL

forearcs. The above considerations regarding the place of ophiolite formation in the history of the host basins support and fit well into this interpretation, and also explain the remarkable similarity in age of large groups of ophiolites. Such distinct ophiolite events have been recognized also in other parts of the world (Ishiwatari 1994). This framework allows interpretation of the history and main features of the ophiolites of the type considered here in terms of known processes. However, the above discussion also reveals the need to further clarify many aspects of this history. The way in which new subduction zones form is still not clear. W h y is ophiolite construction a short-lived process, probably lasting < 10 Ma? W h y are they detached from their substratum and displaced away from the zone of SSZ magma generation, whereas other subduction zones (e.g. western Pacific) follow a different evolutionary trend that leads to construction of island arcs? If the distinctive ophiolite rock-associations normally form in new subduction zones, why are they not commonly found associated with old island arcs? Various aspects of metamorphic sole formation and of the obduction mechanism also need further clarification. In conclusion, the foregoing considerations show that examination of ophiolites from the standpoint of the histories of the host basins provides significant insights regarding their origin and history, supplementing the wealth of other data regarding ophiolites. This allows us to explain their formation and history in terms of known processes, but also reveals and highlights significant problems that must be clarified before ophiolite histories can be understood. This paper greatly benefited from thorough and very helpful reviews and from encouragement by A. H. F. Robertson, A. Rassios, J. Shervais and O. Parlak. I am very grateful to all of them.

References AKTAS,, G. & ROBERTSON, A. H. F. 1990. Tectonic evolution of the Tethys suture zone in SE Turkey: evolution evidence from the petrology and geochemistry of Late Cretaceous and Middle Eocene extrusives. In: MALPAS,J., MOORES,E. M., PANAYIOTOU,A. • XENOPHONTOS,C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 311-329. ALABASTER,T., PEARCE,J. A. 8r MALPAS,J. 1982. The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex. Contributions to Mineralogy and Petrology, 81, 168-183.

ALLERTON, S. 8r VINE, F. J. 1991. Spreading evolution of the Troodos ophiolite, Cyprus. Geology, 19, 637-640. AL-RIYAMI, K., ROBERTSON,A. H. F., XENOPHONTOS, C., DANELIAN, T. 8r DIXON, J. 2000. MesozoicTertiary tectonic and sedimentary evolution of the Arabian continental margin in Baer-Bassit (northwest Syria). In: MALPAS, J., XENOPHONTOS, C. 8r PANAYIDES,A. (eds) Proceedings of the Third International Conference on the Geology of the Eastern Mediterranean. Geological Survey Department, Nicosia, 61-82. AL-R1YAMI, K., ROBERTSON, A. H. F., DIXON, J. ~; XENOPHONTOS, C. 2002. Origin and emplacement of the Late Cretaceous Baer-Bassit ophiolite and its metamorphic sole in NW Syria. Lithos, 65, 225-260. ANONYMOUS 1972. Penrose field conference on ophiolites. Geotimes, 17(12), 24-25. BA(3CI, U., PARLAK, O. & HOCK, V. 2005. Whole rock and mineral chemistry of cumulates from the Klzllda~ (Hatay) ophiolite (Turkey): clues for multiple magma generation during crustal accretion in the southern Neotethyan ocean. Mineralogical Magazine, 69, 53-76. BARAGAR,W. R. A., LAMBERT,M. B., BAGLOW,N. & GIBSON, I. L. 1990. The sheeted dyke zone in the Troodos ophiolite. In: MALPAS, J., MOORES, E. M., PANAYIOTOU,A. & XENOPHONTOS,C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 37-51. BATANOVA, V. G. & SOBOLEV,A. V. 2000. Compositional heterogeneity in subduction-related mantle peridotites, Troodos massif, Cyprus. Geology, 28, 55-58. BI~BIEN,J., SHALLO,M., MANIKA, K. & GEGA, D. 1998. The Shebenik massif (Albania): a link between MOR- and SSZ-type ophiolites? Ofioliti, 23, 7-15. BI~BIEN, J., DIMO-LAHITTE, A., VERGI~LY, P., lnsergueix-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 ophiolite. Contributions to Mineralogy and Petrology, 85, 253-271. BECCALUVA, L., COLTORTI, M., PREMTI, I., SACCANI, E., SIENA, E. & SEDA, O. 1994. Mid-ocean and suprasubduction affinities in the ophiolitic belts of Albania. Ofioliti, 19, 77-96. BECCALUVA,L., COLTORTI, M., GUINTA, G. & SIENA, F. 2004. Tethyan vs. Cordilleran ophiolites: a reappraisal of distinctive tectono-magmatic features of supra-subduction complexes in relation to the subduction mode. Tectonophysics, 393, 163-174. BEDNARZ, U. & SCHMINKE, H. U. 1994. Petrological and chemical evolution of the northeastern Troodos extrusive series, Cyprus. Journal of Petrology, 35, 489-523. BERBERIAN, F. & BERBERIAN, M. 1981. Tectonoplutonic episodes in Iran. ln: GUPTA, H. K. & DELANY, F. M. (eds) Zagros, Hindu Kush,

LIFE CYCLE OF NEOTETHYAN OPHIOLITES

Himalaya. American Geophysical Union, Geodynamics Series, 3, 5-32. BERBERIAN, M. & KING, G. C. P. 1981. Towards a paleogeography and tectonic evolution of Iran. Canadian Journal of Earth Sciences, 18, 210-265. BEYARSLAN, M. & BINGOL, A. E. 2000. Petrology of supra-subduction zone ophiolite (Elazi~, Turkey). Canadian Journal of Earth Sciences, 37, 1411-1424. BOOTE, D. R. D., MOU, D. & WAITE, R. I. 1990. Structural evolution of the Suneinah foreland, 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, 397-418. 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 mid-ocean ridge and subduction magmatism in Albanian ophiolites. Journal of Geology, 110, 561-576. BORTOLOTTI, V., CHIARI, M., MARCUCCI, M., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2004. Comparison among the Albanian and Greek ophiolites: in search of constraints for the evolution of the Mesozoic Tethys Ocean. Ofioliti, 29, 19-35. BOUDIER, F. & NICOLAS, A. 1985. Harzburgite and lherzolite subtypes in ophiolitic and oceanic environments. Earth and Planetary Science Letters, 76, 84-92. BOUDIER, F., CEULENEER, G. & NICOLAS, A. 1988. Shear zones, thrusts and related magmatism in Oman ophiolite: initiation of thrusting on an oceanic ridge. Tectonophysics, 151,275-296. BULL, J. M. ~; SCRUTTON, R. A. 1992. Seismic reflection images of intraplate deformation, central Indian Ocean, and their tectonic significance. Journal of the Geological Society, London, 149, 955-966. CAMERON, W. E., NISBET, E. G. & DIETRICH, V. J. 1980. Petrographic dissimilarities between ophiolitic and ocean-floor basalts. In: PANAYIOTOU, A. (ed.) Ophiolites. Geological Survey Department, Nicosia, 182-192. CAPEDRI, S., VENTURELLI, G., BOCCHI, G., OSTAL, J., GARUTIO, G. & RossI, A. 1980. The geochemistry and petrogenesis of an ophiolitic 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 nappe, northern Albania. Ofioliti, 21, 21-40. CASEY, J. F. & DEWEY,J. F. 1984. Initiation of subduction zones along transform and accreting plate boundaries, triple-junction evolution, and forearc spreading centers--implications for ophiolitic geology and obduction. In: GASS, I. G., LIPPARD, S. J. & SHELTON, A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 269-290.

321

~ELIK, O. F. & DELALOYE, M. K, 2003. Origin of the metamorphic soles and their post-kinematic mafic dyke swarms in the Antalya and Lycian ophiolites, SW Turkey. Geological Journal, 38, 235-256. CHASE, C. G. 1978. Extension behind island arcs and motions relative to hot spots. Journal of Geophysical Research, 83, 5386-5387. CHIARI, M., BORTOLOTTI,V., MARCUCCI,M., PRINCIPI, G. • PHOTIADES,A. 2003. The Middle Jurassic siliceous sedimentary cover at the top of the Vourinos Ophiolite (Greece). Ofioliti, 28, 95-104. CLIFT, P. D. & ROBERTSON, A. H. F. 1989. Evidence of a late Mesozoic oceanic basin and subduction accretion in the southern Greek Neo-Tethys. Geology, 17, 559-563. COLEMAN, R. G. 1977. Ophiolites: Ancient Oceanic Lithosphere. Springer, Berlin. COLEMAN, R. G. 1981. Tectonic setting for ophiolite obduction in Oman. Journal of Geophysical Research, 86, 2497-2508. DAVIES, J. H. ~r STEVENSON,D. J. 1992. Physical model of source region of subduction zone volcanics. Journal of Geophysical Research, 97, 2037-2070. DANELIAN, T. & ROBERTSON, A. H. F. 2001. Neotethyan evolution of eastern Greece (Pagondas M61ange, Evia Island) inferred from radiolarian biostratigraphy and the geochemistry of associated extrusive rocks. Geological Magazine, 138, 345-363. DEGNAN, P. J. & ROBERTSON, A. H. F. 1998. Mesozoic-early Tertiary passive margin evolution of the Pindos ocean (NW Peloponnese, Greece). Sedimentary Geology, 117, 33-70. DELALOYE, M. t~ WAGNER, J. J. 1984. Ophiolites and volcanic activity near the western edge of the Arabian plate. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 225-249. DESMONDS, J. & BECCALUVA,L. 1983. Mid-ocean ridge and island arc affinities in ophiolites from Iran: palaeogeographic implications. Chemical Geology, 39, 39-63. DILEK, Y. 2003a. Ophiolite pulses, mantle plumes and orogeny. In: DILEK, Y. & ROBINSON, P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, %19. DILEIr Y. 2003b. Ophiolite concept and its evolution. In: DILEK, Y. & NEWCOMa, S. (eds) Ophiolite Concept and the Evolution of Geologic Thought. Geological Society of America, Special Papers, 373, 1-16. DILEK, Y. & THY, P. 1998. Structure, petrology and seafloor spreading tectonics of the Klzflda~ ophiolite, Turkey. In: MILLS, R. A. & HARRISON, K. (eds) Modern Ocean Floor Processes and the Geological Record. Geological Society, London, Special Publications, 148, 43-69. 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., MOORES, E. M. & FURNES, H. 1998. Structure of modern oceanic crust and ophiolites and

322

Z. G A R F U N K E L

implications for faulting and magmatism at oceanic spreading centers, ln: BUCK,W. R, DELANEY,P. T., KARSON, J. A. & LAGABRIELLE,Y. (eds) Faulting and Magmatism at Mid-Ocean Ridges. American Geophysical Union, Geophysical Monograph, 106, 219-266. DILEK, Y., THY, P., HACKER,B. & GRUNDVIG, S. 1999. Structure and petrology of Tauride ophiolites and mafic dyke intrusions (Turkey): implications for Neothethys ocean. Geological Society of America Bulletin, 111, 1192-1216. DIMO-LAHITTE, m., MONII~, P. & VERGI~LY, P. 2001. Metamorphic soles from the Albanian ophiolites: petrology, 4~ geochronology, and geodynamic implications. Tectonics, 20, 78-96. DOUTSOS, T., PE-PIPER, G., BORONKAY, K. & KOUKOUVELAS, I. 1993. Kinematics of the central Hellenides. Tectonics, 12, 936-953. ELSASSER, W. M. 1971. Seafloor spreading and thermal convection. Journal of Geophysical Research, 76, 1101-1112. EL-SHAZLI, A. E. K. 2001. Are pressures for blueschists and eclogites overestimated? The case from NE Oman. Lithos, 56, 231-264. ENGLAND, P., ENGDAHL, R. & THATCHER, W. 2004. Systematic variation in the depths of slabs beneath arc volcanics. Geophysical Journal International, 156, 377-408. 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. FLOWER, M. F. J. & DILEK, Y. 2003. Arc-trench rollback and forearc accretion: 1. A collisioninduced mantle flow model for Tethyan ophiolites. In: DILEK, Y. & ROBINSON, P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 21-42 FRANCHETEAU, J., ARMIJO, R., CHEMINI~E, J. L. HEKINIAN, R., LONSDALE, P. & BLUM, N. 1990. 1 Ma East Pacific rise oceanic crust and uppermost mantle exposed by rifting in Hess Deep (Equatorial Pacific). Earth and Planetary Science Letters, 101, 281-295. GARFUNKEL, Z. 1998. Constraints on the origin and history of the Eastern Mediterranean basin. Tectonophysics, 298, 5-35. GARFUNKEL, Z. 2004. Origin of the Eastern Mediterranean basin: a reevaluation. Tectonophysics, 391, 11-34.

GARFUNKEL, Z. & GREILING, R. O. 2002. The implications of foreland basins for the causative tectonic loads. In: BERTOT]?I, G., SCHULMANN, K. & CLOETINCH, S. A. P. L. (eds) Continental Collision and the Tectono-sedimentary Evolution of Forelands. European Geosciences Union, Stephan Mueller Special Publication Series, 1, 3-16. GARFUNKEL, Z., ANDERSON, C. A. & SCHUBERT, G. 1986. Mantle circulation and the lateral migration of subducted slabs. Journal of Geophysical Research, 91, 7205-7223. GASS, I. G. 1968. Is the Troodos massif of Cyprus a fragment of Mesozoic ocean floor? Nature, 220, 39-42.

GASS, I. G. 1980. The Troodos massif: its role in the unravelling of the ophiolite problem and its significance in the understanding of constructive margin process. In: PANAYIOTOU, A. (ed.) Ophiolites. Geological Survey Department, Nicosia, 23-35. GHAZI, A. M., HASSANIPAK,A. A., MAHONEY, J. J. & DUNCAN, R. A. 2004. Geochemical characteristics, 4~ ages and original tectonic setting of the Band-e-Zeyarat/Dar Anar ophiolite, Makran accretionary prisem, SE Iran. Tectonophysics, 393, 175-196. GILL, J. 1981. Orogenic Andesites and Plate Tectonics. Springer, Berlin. GRADSTEIN, M., OGG, J. G., SMITH, A. G., BLEEKER, W. & LOURENS, L. J. 2004. A new geologic time scale with special reference to Precambrian and Neogene. Episodes, 27, 83-100. (Also available at http:/lwww.stratigraphy.org.) GRAY, D. R. & GREGORY, R. T. 2003. Ophiolite obduction and Semail Ophiolite: the behaviour of the underlying margin. In: DILEK, Y. & ROBINSON, P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 449M65. HACKER, B. R. & GNOS, E. 1997. The conundrum of Samail: explaining the metamoprhic 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. HAWKESWORTH, C. J., GALLAGHER, E., HERGT, J. M. & MCDERMOTT, F. 1993. Mantle and slab contributions in arc magmas. Annual Review of Earth and Planetary Sciences, 21, 175-204. HAWKINS, J. W. 1995. Evolution of the Lau Basin-insights from ODP Leg 135. In: TAYLOR, B. & NATLAND, J. (eds) Active Margins and Marginal Basins of the western Pacific. American Geophysical Union, Geophysical Monograph, 88, 125-173. HAWKINS, J. W. 2003. Geology of supra-subduction zones--implications for the origin of ophiolites. In: DILEK, Y. & NEWCOMB, S. (eds) Ophiolite Concept and Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 227-268. HAWKINS, J. W., BLOOMER, S. H., EVANS, C. A. & MELCHIOR, J. T. 1984. Evolution of intra-oceanic arc-trench systems. Tectonophysics, 102, 175-205. 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 Crustal Analogues. Geological Survey Department, Nicosia, 149-164. HESS, H. H. 1965. Mid-ocean ridges and tectonics of the sea-floor. In: WHITTARD, W. F. & BRADSHAW, R. (eds) Submarine Geology and Geophysics. Colston Papers, 17, 317-333. HOECK, V., KOLLER, F., MEISEL, T., ONUZI, K. & KNERINGER, E. 2002. The Jurassic south Albanian ophiolites: MOR- vs. SSZ-type ophiolites. Lithos, 65, 143-164. INSERGIEUX-FILIPP1, D., DUPEYRANT, L., DIMOLAHITTE, A., VERGELY, P. & BI~BIEN, J. 2000.

LIFE CYCLE OF NEOTETHYAN OPHIOLITES Albanian ophiolites II.--Model of subduction zone infancy at a mid-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. ISHIWATARI, A. 1994. Circum-Pacific multiple ophiolite belts. Proceedings of 29th International Geological Congress, Kyoto, Japan, Part D. VSP, Utrecht, 7-28. JONES, G. & ROBERTSON, A. H. F. 1991. Tectonostratigraphy and evolution of the Pindos ophiolite and associated units. Journal of the Geological Society, London, 148, 267-288. JONES, G., ROBERTSON, A. H. F. & CAYN, J. R. 1991. Genesis and emplacement of the supra-subduction zone Pindos ophiolite, northwestern Greece. In: PETERS, T., NICOLAS, A. & COLBMAN, G. (eds) Ophiolite Genesis and the Evolution of Oceanic Lithosphere. Kluwer, Dordrecht, 771-800. JONES, G., DE WEVER, P. & ROBERTSON,A. H. F. 1992. Significance of radiolarian age data to the Mesozoic tectonics and sedimentary evolution of the northern Pindos mountains, Greece. Geological Magazine, 129, 385-400. JUTEAU, T., ERNEWEIN, M., REUBER, I., WHITECHURCH, H. & DAHAL, R. 1988. Duality of magmatism in the plutonic sequence of the Sumail Nappe, Oman. Tectonophysics, 15, 107-135. KARSON, J. 1998. Internal structure of oceanic lithosphere: a perspective from tectonic windows. In: BUCK, W. R, DELANEY, P. T., KARSON, J. A. & LAGABRIELLE,Y. (eds) Faulting and Magmatism at Mid-Ocean Ridges. American Geophysical Union, Geophysical Monograph, 106, 177-218. KINCAID, C. & SACKS,S. I. 1997. Thermal and dynamical evolution of the upper mantle in subduction zones. Journal of Geophysical Research, 102, 12295-12315. KRISHNA, K. S., RAMANA, M. V., GOPALA RAO, D., MURTHY, K. S. R., MALLESWARA RAO, M. M., SUBRAHMANYAN,V. & SARMA, K. V. 1998. Periodic deformation of oceanic crust in the central Indian Ocean. Journal of Geophysical Research, 103, 17859-17875. LACHIZE, M., LORAND, J. P. & JUTEAU, T. 1996. Calc-alkaline differentiation trend in the plutonic sequence of the Wadi Haymiliya section, Haylan Massif, Semail Ophiolite, Oman. Lithos, 38, 207-232. LAURENT, R., DELALOYE, M. & VAUGNAT, M., WAGNER, J. J., 1980. Composition of parental magma in ophiolites. In: PANAYIOTOU, A. (ed.) Ophiolites. Geological Survey Department, Nicosia, 172-181. LEICH, E. C. 1984. Island arc elements and arc-related ophiolites. Tectonophysics, 106, 177-203. LE PICHON, X., BERGERAT, F. & ROULET, M. J. 1988. Plate kinematics and tectonics leading to the Alpine belt formation. A new analysis. In: CLARK, S. P. JR, BURCHFIELD, B. C. & SUPPE, J. (eds) Processes in Continental Lithospheric Deformation. Geological Society of America, Special Papers, 218, 111-131. LIATI, A., GEBAUER, D. & FANNING, C. M. 2004. The age of ophiolitic rocks of the Hellenides (Vourinos, Pindos, Crete): first U-Pb ion microprobe

323

(SHRIMP) zircon ages. Chemical Geology, 207, 177-188. LIPPARD, S. J., SHELTON, A. W. & GASS, I. G. 1986. The Ophiolite of Northern Oman. Geological Society, London, Memoirs, 11. LOMBARDO, B., RUBATTO, D. & CASTELLI, D. 2002. Ion microprobe U-Pb dating of zircon from a Monviso plagiogranite: implications for the evolution of the Piedmont-Liguria Tethys in the Western Alps. Ofioliti, 27, 109-117. LORD, A. R., PANAYIDES, I., URQUHART, E. XENOPHONTOS, C. 2002. A biochronostratigraphical framework for the Late Cretaceous Recent Circum-Troodos sedimentary sequence, cyprus. In: PANAYIDES, I., XENOPHONTOS, C. & MALPAS, J. (eds) Proceedings of the Third International Conference on the Geology of the Eastern Mediterranean. Geological Survey Department, Nicosia, 289-297. LYTWYN, H. N. & CASEY,J. F. 1993. The geochemistry and petrogenesis of volcanics and sheeted dykes from the Hatay (Klzdda~) ophiolite, southern Turkey: possible formation with the Troodos ophiolite, Cyprus, along fore-arc spreading centers. Tectonophysics, 223, 237-272. 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. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 141-176. MALPAS, J. 1990. Crustal accretionary processes in the Troodos ophiolite, Cyprus: evidence from field mapping and drilling. In: MALPAS, J., MOORES, M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 65-74. MALPAS, J., XENOPHONTOS, C. & WILLIAMS, O. 1992. The Ayia Varvara Formation of SW Cyprus: a product of complex collision tectonics. Tectonophysics, 212, 193-211. MCCALL, G. J. H. 1997. The geotectonic history of the Makran and adjacent area of southern Iran. Journal of Asian Earth Sciences, 15, 517-531. MCCULLOCI-I,M. T. & GAMBLE,J. A. 1991. Geochemical and geodynamical constraints on subduction zone magmatism. Earth and Planetary Science Letters, 102, 358-374. MOORES, E. M. 1969. Petrology and Structure of the Vourinos Ophiolitic Complex, Northern Greece. Geological Society of America, Special Papers, 118. MOORES, E. M. 1982. Origin and emplacement of ophiolites. Reviews of Geophysics and Space Physics, 20, 735-760. MOORES, E. M. & VINE, F. J. 1971. The Troodos Massif, Cyprus and other ophiolites as oceanic crust: evaluations and implications. Philosophical Transactions of the Royal Society of London, Series A, 268, 433-466. MORRIS, A. 1996. A review of palaeomagnetic research in the Troodos ophiolite, Cyprus. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics in the Mediterranean Region. Geological Society, London, Special Publications, 105, 311-324.

324

Z. G A R F U N K E L

MORRIS, A., ANDERSON, M. W. & ROBERTSON, A. H. F. 1998. Multiple tectonic rotations and transform tectonism in an intraoceanic suture zone, SW Cyprus. Tectonophysics, 299, 229-253. MORRIS, A., ANDERSON, M. W., ROBERTSON, A. H. F. & AL-RIAMI, K. 2002. Extreme tectonic rotations within an eastern Mediterranean ophiolite (BaerBassit, Syria). Earth and Planetary Science Letters, 202, 247-261. MUKASA, S. & LUDDEN, J. N. 1987. Uranium-lead ages of plagiogranites from Troodos ophiolite, Cyprus, and their significance. Geology, 15, 825-826. MOLLER, R. D. & ROEST, W. R. 1992. Fracture zones in the North Atlantic from combined Geosat and Seasat data. Journal of Geophysical Research, 97, 3337-3350. NICOLAS, A. 1989. Structures of Ophiolites and Dynamics of Oceanic Lithosphere. Kluwer, Dordrecht. OHARA, Y., STERN, R. J., 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. OKAY, A. L. & TOYS0Z, O. 1999. Tethyan sutures of northern Turkey. In: DURAND, B., JOLIVET, L., HORVATH, F. & SERANNE, M. (eds) The Mediterranean Basins: Tertiary Extension within the Alpine Orogen. Geological Society, London, Special Publications, 156, 475-515. OKAY, A. I., HARRIS, N. B. W. & KELLEY, S. P. 1998. Exhumation of blueschists along a Tethyan suture in northwest Turkey. Tectonophysics, 285, 275299. ()NEN, A. P. 2003. Neotethyan ophiolitic rocks of the Anatolides of NW Turkey and comparison with Tauride ophiolites. Journal of the Geological Society, London, 160, 947-962. PAMI(~, J., TOMLJEVI~, B. & BALEN, D. 2002. Geodynamic and petrogenetic evolution of Alpine ophiolites from the central and NW Dinarides: an overview. Lithos, 65, 113-142. PARLAK, O. & DELALOYE,M. 1996. Geochemistry and timing of postmetamorphic dike emplacement in the Mersin ophiolite (southern Turkey): new age constraints from 4~ geochronology. Terra Nova, 8, 585-592. PARLAK, O. & DELALOYE, M. 1999. Precise 4~ ages from the metamorphic sole of the Mersin ophiolite (southern Turkey). Tectonophysics, 301, 145-158. PARLAK, O. & RIZAO(]LU, T. 2004. Geodynamic significance of granitoid megmatism in the southeast Anatolian orogeny (Turkey). In: 5th International Symposium on Eastern Mediterranean Geology. Thessaloniki, Greece, 157. PARLAK, O., DELALOYE,M. & BINGOL, E. 1996. Mineral chemistry of ultramafic and mafic cumulates as an indicator of the arc-related origin of the Mersin ophiolite (southern Turkey). Geologische Rundschau, 85, 647-661. PARLAK, O., HOCK, V. & DELALOYE, M. 2000. Suprasubduction zone origin of the PozannKersantl ophiolite (southern Turkey) deduced from whole-rock 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 Areas. Geological Society, London, Special Publications, 173, 219-234. PARLAK, O., HOECK, V., KOZLU, H. & DELALOYE,M. 2004. Oceanic crust generation in an island arc tectonic setting, SE Anatolian orogenic belt (Turkey). Geological Magazine, 141, 583-603. PATRON, T. L. & O'CONNOR, S. J. 1988. Cretaceous flexural history of northern Oman Mountains foredeep, United Arab Emirates. American Association of Petroleum Geologists Bulletin, 72, 797-809. PEARCE, J. A. & PEATE, O. W. 1995. Tectonic implications of the composition of volcanic arc magmas. Annual Review of Earth and Planetary Sciences, 23, 251-285. 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. PEARCE, J. A., VAN DER LAAN, S. R., ARCULUS, R. J., MURTON, B. J., ISHII, T., PEATE, D. W. & PARKINSON, I. J. 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., et al. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 125. Ocean Drilling Program, College Station TX, 623-659. Pc-PIPER, G. 1998. The nature of Triassic extensionrelated magmatism in Greece: evidence from Nd and Pb isotope geochemistry. Geological Magazine, 135, 331-348. PERRIN, M., PLEN1ER,G., DAUTRIA, J. M., COCUAUD, E. & PRI~VOT, M. 2000. Rotation of the Semail ophiolite (Oman): additional paleomagnetic data from the volcanic sequence. Marine Geophysical Researches, 21, 181-194. PORTNYAGIN, M. V., DANYUSHEVSKY, L. V. & KAMENETSKY, V. S. 1997. Coexistence of two distinct mantle sources during formation of ophiolites: a case study of primitive pillow-lavas from the lowermost part of the volcanic section of the Troodos ophiolite, Cyprus. Contributions to Mineralogy and Petrology, 128, 287-301. PYTHON, M. & CEULENEER, F. 2003. Nature and distribution of dykes and related melt migration structures in the mantle section of the Oman ophiolites. Geoochemistry, Geophysics, Geosystems, 4(7), 8612, doi:10.1029/2002GC000354. RABU, D., 1993. Stratigraphy and Structure of the Oman Mountains. Documents on BRGM Bureau de Recherches GOologiques et Minikres, 221. RASSIOS, A. & SMITH, A. G. 2000. Constraints on the formation and emplacement ages of wesrtern Greek ophiolites (Vourinos, Pindos, and Othris) inferred from deformed structures in peridotites. In: DILEK, Y., MOORES, E. M., ELTHON, D. A. NICOLAS, A. (eds) Ophiolites and Oceanic Crust: New Insights from Field Studies and Ocean

LIFE CYCLE OF NEOTETHYAN OPHIOLITES Drilling Program. Geological Society of America, Special Papers, 349, 473-483. RIBE, N. M. 1989. Mantle flow induced by back arc spreading. Geophysical Journal International, 98, 85-91. RICOU, L. E. 1971. Le croissant ophiolitique periarabe: une ceinture de nappes mise en place au Cretac6 sup6rieur. Revue de Gkographie Physique et Gkologie Dynamique, 13, 327-349. ROBERTSON, A. H. F. 1987. The transition from a passive margin to an Upper Cretaceous foreland basin related to ophiolite emplacement in the Oman Mountains. Geological Society of America Bulletin, 99, 633-653. ROBERTSON, A. H. F. 1990. Tectonic evolution of Cyprus. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. t~ XENOPHONTOS, C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 235-250. ROBERTSON, A. H. F. 2000. Tectonic evolution of Cyprus in its Easternmost Mediterranean setting. In'. PANAYIDES, I., XENOPHONTOS, C. 8~ MALPAS, J. (eds) Proceedings of the 3rd 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, 1-68. ROBERTSON, A. H. F. 2004. Development of concepts concerning the genesis and emplacement of Tethyan ophiolites in the Eastern Mediterranean and Oman regions. Earth-Science Reviews, 66, 331-387. ROBERTSON, A. H. F. & SHALLO,M. 2000. MesozoicTertiary tectonic evolution of Albania and its regional Eastern Mediterranean context. Tectonophysics, 316, 197-214. ROBERTSON, A. H. F. & WOODCOCK, N. H. 1979. Mamonia complex, southwest Cyprus: evolution and emplacement of a Mesozoic continental margin. Geological Society of America Bulletin, 90, 651-665. ROBERTSON, A. H. F. & XENOPHONTOS, C. 1993. Development of concepts concerning the Troodos ophiolite and adjacent units in Cyprus. In: PRICHARD, H. M., ALABASTER, T. 8r HARRIS, T. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 70, 85-120. ROBERTSON, A. H. F., CLIFT, P. D., DEGNAN, P. J. 8~ JONES, G. 1991. Palaeogeographic and palaeotectonics 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 Eastern Mediterranean. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics in the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBINSON, P. T. & MALPAS, J. 1990. The Troodos ophiolite of Cyprus: new perspectives on its origin and emplacement. In: MALPAS, J., MOORES, E. M.,

325

PANAYIOTOU,A. & XENOPHONTOS,C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 13-26. SACCANI, E. & PHOTIADES, A. 2004. Mid-ocean ridge and supra-subduction affinities in the Pindos ophiolites (Greece): implications for magma genesis in a forearc setting. Lithos, 73, 229-253. 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 and Earth History. Geological Society, London, Special Publications, 218, 109-127. SACCANI, E., BECCALUVA,L., COLTORTI, M. & SIENA, F. 2004. Petrogenesis and tectono-magmatic significance of the Albanide-Hellenide Subpelagonian ophiolites. Ofioliti, 29, 75-93. SARKARINEJAD, K. 1994. Petrology and tectonic settings of the Neyriz ophiolite, southwestern Iran. Proceedings of 29th International Geological Congress, Part D. VSP, Utrecht, 221-234. SCHIANO, P., CLOCCHIATTI, R., LORAND, J. P., MASSARE, D., DELOULE, E. & CHAUSSIDON, M. 1997. Primitive basaltic melt included in podiform chromites from the Oman Ophiolite. Earth and Planetary Science Letters, 146, 489-497. SEARLE, M. • Cox, J. 1999. Tectonic setting, origin, and obduction of the Oman ophiolite. Geological Society of America Bulletin, 111, 104-122. SEARLE, M. P., WARREN, C. J. & WATERS, D. J., PARRISH, R. R. 2003. In: DILEK, Y. & ROBINSON, P. T. (eds) Ophiolites in Earth History. Geological Society, London, Special Publications, 218, 467-480. ~ENGOR, A. M. C., ALTINER, D., CrN, A., USTAOMER, T. & Hst3, K. J. 1988. Origin and assembly of the Tethyside orogenic collage at the expense of Gondwanaland. In: AUDLEu M. G. & HALLAM, A. (eds) Gondwana and Tethys. Geological Society, London, Special Publications, 37, 119-181. SHALLO, M. 1994. Outline of the Albanian ophiolies. Ofioliti, 19, 57-75. SHERVAIS, J. W. 2001. Birth, death, and resurrection: the life cycle of suprasubduction zone ophiolites. Geochemistry, Geophysics, Geosystems, paper number 2000GC000080. SMITH, A. G. 1993. Tectonic significance of the Hellenic-Dinaric ophiolites. In: PRITCHARD,H. M., ALABASTER,T., HARRIS, N. B. W. & NEARY, C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 213-243. SMITH, A. G. 2004. Are N Atlantic hot-spots and N Atlantic continental breakup related to Tethyan ophiolite creation and emplacement? In: CHATZIPETROS, A. A. & PAVLIDES, S. B. (eds) 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 288-291. SMITH, A. G., HYNES, A. J., MENZIES, M., NISBET, E. G., PRICE, I., WELLAND, M. J. P. & FERRII~RE, J. 1975. The stratigraphy of the Othrys mountains, east central Greece: a deformed continental margin

326

Z. G A R F U N K E L

sequence. Eclogae Geologicae Heltetiae, 68, 463481. SMITH, A. G., WOODCOCK, N. H. & NAYLOR, M. A. 1979. The structural evolution of a Mesozoic continental margin, Othrys Mountains, Greece. Journal of the Geological Society, London, 136, 589-603. 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. 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,J., REX, D. C. & RODDICK, J. C. 1984. Age constraints on the igneous and metamorphic evolution of the Hellenic-Dinaric ophiolites. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 619-627. STAMPFLI, G. M., MOSAR, J., FAVRE, P., PILLEVUIT, A. & VANNAY, J. C., 2001. Permo-Meosozoic evolution of western Tethys realm: the Neo-Tethys East Mediterranean basin connection. In: ZIEGLER, P. A., CAVAZZA, W., ROBERTSON, A. H. F. & CARSQUIN-SOLEAU, S. (eds) Peri-Tethys Memoir 6. M6moris du Mus6um National d'Histoire Naturelle, 186, 51-108. STERN, R. J. 2004. Subduction initiation: spontaneous and induced. Earth and Planetary Science Letters, 226, 275-292. STERN, R. J. & BLOOMER, S. H. 1992. Subduction zone infancy: examples from the Eocene Izu-BoninMariana and Jurassic California. Geological Society of America Bulletin, 104, 1621-1636. THY, P. & ESBENSEN, K. H. 1993. Seafloor spreading and the ophiolitic sequences of the Troodos Complex: a principal component analysis of lava and dike composition. Journal of Geophysical Research, 98, 11799-11805. TILTON, G. R., HOPSON, C. A. & WRIGHT, J. E. 1981. Uranium-lead isotopic ages of the Semail ophiolite, Oman. Journal of Geophysical Research, 86, 2763-2776. TIPPIT, R. P., PESSAGO, E. A. & SMEWING,J. D. 1981. The biostratigraphy of sediments in the volcanic units of the Samail ophiolite. Journal of Geophysical Research, 86, 2756-2762. UMINO, S., YANAI, S., YAMAN, 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. Geological Survey Department, Nicosia, 375-374. VARGA, R. J. & MOORES, E. M. 1985. Spreading structure of the Troodos Ophiolite. Geology, 13, 846-850. VERGILI, O. & PARLAK, O. 2005. Geochemistry and tectonic setting of metamorphic sole rocks and mafic dikes from the Pinarba~1 ophiolite, Central Anatolia (Turkey). Ofioliti, 30, 37-52.

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. 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, 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, 419427. WEILER, P. D. 2000. Differential rotation in the Oman ophiolite: paleomagnetic evidence from southern massifs. Marine Geophysical Researches, 21, 195-210. WHITECHURCH,H., JUTEAU, T. & MONTIGNY, R. 1984. Role of the Eastern Mediterranean ophiolites (Turkey, Syria, Cyprus) in the history of the NeoTethys. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 301-317. WOODCOCK, N. H. & ROBERTSON, A. H. F. 1977. Origins of some ophiolite-related rocks of the 'Tethyan' belt. Geology, 5, 373-379. WOODSIDE,J. M. 1977. Tectonic elements and crust of the eastern Mediterranean Sea. Marine Geophysical Researches, 3, 317-354. YALINIZ, K. M., FLOYD, P. A. & GONCI3Ot3LU,C. 2000. Geochemistry of volcanic rocks from the Cicekda~ ophiolite, central Anatolia, Turkey, and their inferred tectonic setting within the northern branch of the Neotethyan Ocean. In: BOZKURT, E., WINCHESTER, J. A., PIPER, J. D. A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Areas. Geological Society, London, Special Publications, 173, 203-218. YANAi, S., UMINO, S., IBRAHIM,S. O., NAKAMURA,Y. & IIVAMA, J. T. 1990. Subduction- and collisionrelated emplacement of the Semail ophiolite, northern Oman Mountains. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. & XENOPHONTOS,C. (eds) Ophiolites, Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 385-396. YAZGAN, E. & CHESSEX,R. 1991. Geology and tectonic evolution of the southeastern Taurides in the region of Malatya. Turkish Association of Petroleum Geologists Bulletin 3, 1-42. Y1GITBA~, E., YILMAZ, Y. 1996. New evidence and solution to the Maden complex controversy of southeast Anatolian orogenic belt (Turkey). Geologische Rundschau, 85, 250-263. Y1LMAZ, Y. 1993. New evidence on the evolution of the southeast Anatolian orogen. Geological Society of America Bulletin, 105, 251-271. Y1LMAZ, Y., YIGITBA~, E. & GENff, 9" C. 1993. Ophiolitic and metamorphic assemblages of southeast Anatolia and their significance in the geological evolution of the orogenic belt. Tectonics, 12, 1280-1297.

Nature and significance of Late Cretaceous ophiolitic rocks and their relation to the Baskil granitic intrusions of the Elazl~ region, SE Turkey T A M E R R I Z A O t ~ L U 1, O S M A N P A R L A K 1, V O L K E R

HOECK 2& FIKRET ISLER 1

l(Sukurova Oniversitesi, Jeoloji Miihendisli~i B61iimii, 01330 Balcah, Adana, Turkey (e-mail: parlak@cukuro va. edu. tr ) 2University o f Salzburg, Department o f Geology, A-5020 Salzburg, Austria The Elaz~ region in SE Turkey comprises, in descending order, the PalaeozoicMesozoic Malatya-Keban platform, an ensimatic island arc unit (i.e. EIam~ magmatic rocks-YiJksekova complex), and ophiolitic rocks (i.e. K6mfirhan) of Late Cretaceous age. All of these were intruded by the Baskil granitic rocks. These tectonomagmatic-stratigraphic assemblages were emplaced over the Middle Eocene volcano-sedimentary Maden complex to the south during the evolution of the SE Anatolian orogen. The K6mfirhan ophiolite exhibits an intact ophiolite pseudostratigraphy. The base of this has been metamorphosed to amphibolite facies during intraoceanic subduction-thrusting. The amphibolitic rocks were intruded by synkinematic granitic rocks (Baskil magmatic rocks). The ensimatic island arc volcanic rocks are widely distributed in the region. The contact of the volcano-sedimentary unit with the underlying K6mfirhan ophiolite is a thrust dipping to the north. The rock assemblages of the volcano-sedimentary unit suggest formation of small volcanic edifices above a subduction zone, coupled with debris-flow deposits and volcaniclastic turbidites. The whole-rock and mineral chemistry of the K6mfirhan ophiolite and the ensimatic island arc volcanic rocks suggests that they represent a comagmatic tholeiitic suite, formed in the Late Cretaceous in a suprasubduction zone (SSZ) setting. The amphibolites beneath the K6mfirhan ophiolite indicate derivation from an island arc tholeiite (IAT) protolith. The geological and geochemical evidence from the Elaz~ region suggests the following evolutionary scenario. The K6mfirhan ophiolite was formed above a north-dipping subduction zone between the Arabian platform to the south and the Tauride platform to the north in Late Cretaceous (c. 90 Ma). An ensimatic island arc assemblage was then built on the SSZ-type crust. The metamorphic sole was formed by metamorphism of IAT-type basalts that were detached from the front of the overriding K6mfirhan ophiolite and then underplated. These units were then accreted to the base of the Tauride active margin to the north, where both units were cut by the Baskil granitic rocks around 85 Ma. Abstract:

Anatolia is situated in a critical segment of the Alpine-Himalayan orogenic system, where remnants of Neotethyan ocean basins crop out along east-west-trending tectonic zones located between metamorphic massifs or platform carbonates (~eng6r & Ydmaz 1981; Ydmaz et al. 1993; Robertson 2002). The remnants of Neotethys are characterized, in a structural descending order, by ophiolites, metamorphic soles and ophiolitic m61anges (Fig. 1). The ophiolites and related subduction-accretion units were generated during the closing stages of Neotethyan oceanic basins in the Late Cretaceous (Pearce et al. 1984; Yahmz et al. 1996, 2000; Robertson 2002, 2004; Parlak & Robertson 2004; Parlak et al. 2004; Robertson et al. 2006, 2007). The Late Cretaceous ophiolites in Turkey are located in five zones based mainly on their geographical distribution; namely, the Pontide ophiolite belt, the Central Anatolian ophiolite belt, the Tauride

ophiolite belt, the SE Anatolian ophiolite belt and the Peri-Arabian ophiolite belt (Fig. 1). The SE Anatolian orogenic belt is one of the best regions to study mountain-building processes resulting from the collision of the AfroArabian and Eurasian plates in Mid-Miocene time (Ydmaz 1993; Yllmaz et al. 1993). The ophiolites, ensimatic island arc units, ophioliterelated metamorphic rocks and granitic rocks within this orogenic belt are important elements of the Late Cretaceous tectonomagmatic evolution of the southern Neotethys. The Late Cretaceous ophiolites are the G6ksun (Kahramanmara~ or N Berit), [spendere (Malatya), K6miirhan and Guleman ophiolites (Elazl~). The ensimatic island arc volcanic unit is represented by either the Elazl~ magmatic rocks or the Yiiksekova complex. The ophiolite-related metamorphic units are the Befit metaophiolite (S Berit ophiolite) and the metamorphic sole of the

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 327-350. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

328

T. R I Z A O G L U

"._-'~ " "

\

,

.

.

.

.

/

~r ~ '

k~"

3'o ~

E T AL.

'

'

- I

'

~'

'

BLACK SEA

o

oV oo ,

,

,

,

,

,

,

,

solo

Fig. 1. Distribution of the Neotethyan ophiolites in the eastern Mediterranean region (from Robertson 2002).

K6m/irhan ophiolite. The granitic rocks are located in three different areas: the G6ksunAf~in (Kahramanmara~), Do~an~ehir (Malatya) and Baskil (Elazl~) regions. The petrology and geochronology of the G6ksun (N Berit) ophiolite and related granitic rocks to the north of Kahramanmara~ are well constrained (Parlak et al. 2004; Parlak 2006; Robertson et al. 2006, 2007). However, the relations of the Late Cretaceous tectonomagmatic units in the Elazl~ region are not well established because of very limited geochemical and geochronological data. The main uncertainty in the Late Cretaceous evolution of the region is the relationships between the Baskil granitic body, the Elazl~ magmatic rocks-Y/iksekova complex and the K6miirhan ophiolite. One interpretation is that the Elazl~ magmatic rock units are the extrusive equivalents of the Baskil arc plutonic rocks that represent an Andean-type active margin along the MalatyaKeban platform to the north; the K6m/irhan ophiolite formed away from the Tauride margin to the south (i.e more oceanic) in this model (Yazgan & Chessex 1991). A second interpretation is that the Elazl~ magmatic unit, comprising both intrusive and extrusive rocks, is an island arc assemblage. This island arc unit formed above the K6miirhan ophiolite during a mature stage of suprasubduction zone (SSZ) spreading (Beyarslan & Bing61 1996, 2000). A third interpretation is that the extrusive rocks in the Elazl~ region have nothing to do with the Baskil arc plutonic rocks and could be seen as the westward continuation of the Ytiksekova ensimatic island

arc unit formed above a subduction zone during the Late Cretaceous (Perinqek 1979; Aktas & Robertson 1984). More recently, geological mapping was carried out in the area between Baskil and Sivrice (Elazl~) regions (see Fig. 3) to investigate the field relations of the tectonomagmatic units. A detailed stratigraphic log was measured of the ensimatic island arc volcanic unit (Elazl~ magmatic rocks-Yiiksekova complex). This paper presents whole-rock and mineral chemical data for the K6mfirhan ophiolite and the ensimatic island arc unit from the Elazffg region; the results can be interpreted in terms of the spatial and temporal relations between the tectonomagmatic units and the Baskil granitic body during the Late Cretaceous.

Regional geology The Malatya-Elazl~ region comprises a number of tectonomagmatic-stratigraphic units that are important for the evolution of the southern Neotethyan ocean. These are the MalatyaKeban metamorphic unit, the P/it/irge metamorphic unit, Late Cretaceous ophiolites, the Baskil arc magmatic unit, the Elazl~ magmatic unit, the Maden unit and sedimentary cover units (Fig. 2). The Malatya-Keban metamorphic unit is a low-grade metamorphosed Late PalaeozoicMesozoic unit consisting of marble, schist, slate and black phyllite, with rare metaconglomerates (Asutay 1988; Turan & Bing61 1991; Yflmaz et al.

OPHIOLITES AND GRANITES, SE TURKEY

329

Fig. 2. Regional geological map of the Elazl~ region (MTA 2002).

1993). It has both tectonic and intrusive contact relationships with the Baskil arc magmatic unit and is, in turn, overlain by Tertiary unmetamorphosed sedimentary rocks in the Elaz~ region (Bing61 1984; Yazgan & Chessex 1991; Rlzao~lu et al. 2004). The Pfitfirge metamorphic unit comprises both core and cover units. The core rocks are dominated by augen gneiss, amphibole schist and biotite schist with an intruding granite (Yflmaz 1971, 1978), whereas the cover rocks consist of slates, phyllites, calc-schists and marbles (Yxlmaz et al. 1993; Erdem & Bing61 1995). The Pfitiirge metamorphic unit is unconformably overlain by the Middle Eocene Maden unit. Yflmaz et al. (1993) believed that, the metamorphism of the Malatya-Keban and Piitiirge units occurred during the Campanian to Early Maastrichtian interval because the uppermost units of the metamorphic sequences are Campanian in age (Yllmaz et al. 1987) and the massifs are unconformably overlain by an Upper Maastrichtian sedimentary cover. The Baskil arc magmatic unit is mainly exposed near Baskil town and to the north of Keban Lake (Fig. 2) where it cuts the MalatyaKeban platform, the Elazl~ magmatic unitYfiksekova complex and the K6miirhan ophiolite (Parlak 2006). This magmatic complex

was interpreted as I-type calc-alkaline intrusive rocks formed as a result of ensimatic island arc-continent collision during closure of the southern Neotethys. It is represented by basic to silicic plutonic rocks and swarms of dykes. K - A r ages from the Baskil intrusive rocks are reported as 76+2.45 and 7 8 _ 2 . 5 M a by Yazgan & Chessex (1991). The Maden group is a low-grade metamorphosed volcanic and sedimentary unit of MidEocene (Ypresian-Lutetian) age that crops out to the south of Elazl~ region (Fig. 2). It unconformably overlies the Pfitfirge metamorphic unit and is, in turn, tectonically overlain by ophiolites (Fig. 2). There is no consensus on the nature of the Maden Group. Various interpretations were proposed, such as an immature island arc association (Erdo~an 1977), a product of intracontinental subduction (Yazgan 1983, 1984; Michard et al. 1985), a back-arc basin ($eng6r & Yllmaz 1981; Hempton 1984, 1985), or an immature back-arc basin (Yi~itbas & Yflmaz 1996a, b). The Elazl~ magmatic unit is a basic to silicic volcanic and volcano-sedimentary rock assemblages of Late Cretaceous age. It is widely distributed around the Keban Lake and Elazl~ town

330

T. RIZAOGLU E T AL.

OPHIOLITES AND GRANITES, SE TURKEY (Fig. 2). This unit has been interpreted as the extrusive equivalents of the Baskil arc plutonic rocks (Yazgan & Chessex 1991; Beyarslan & Bing61 1996, 2000), or an ensimatic volcanic arc unit-Yiiksekova complex (Peringek 1979; Akta~ & Robertson 1984). Recently, based on field and geochemical data, Rlzao~lu et al. (2004) showed that this volcanic and sedimentary rock assemblage has a thickness of c. 750 m and has a tholeiitic nature. They interpreted this unit as the extrusive part of the K6miirhan ophiolite which was formed in an SSZ environment in the Late Cretaceous. The ophiolites in the region, from west to east, are represented by the ispendere, K6miirhan and Guleman ophiolites (Fig. 2). The K6miirhan is distinct as the lower part of the ophiolite pseudostratigraphy is metamorphosed (Yazgan & Chessex 1991). These ophiolites are interpreted as the emplaced remnants of southem Neotethys formed above a subduction zone (PerinCek 1979; Akta~ & Robertson 1984; Beyarslan 1996; Beyarslan & Bing612000).

Field relations and petrography The Baskil granitic rocks are represented by mafic to acidic plutonic rocks (diorite, granodiorite, granite, tonalite, quartz diorite, quartz monzonite) and swarms of dykes (aplite, diabase, microdiorite and granophyre). It has both tectonic and intrusive contact relationships with the Malatya-Keban platform in the central part of the study area near Ayranh (Fig. 3). The Baskil intrusive rocks are unconformably overlain by Palaeocene and younger sediments between Odaba~x and Hasanda~x (Fig. 3), whereas they intrude the K6mtirhan unit and the volcanosedimentary unit to the south (Fig. 3). The K6mtirhan ophiolite comprises a complete oceanic lithospheric remnant and is represented, from the bottom to the top, by mantle tectonites, ultramafic-mafic cumulates, isotropic gabbros, sheeted dykes, volcanic rocks and associated sedimentary rocks (Fig. 4). A thin metamorphic sole unit tectonically underlies the mantle tectonites to the south of Karakaya Tepe (Fig. 3). The volcano-sedimentary unit of the K6miirhan ophiolite crops out along an eastwest-trending belt in the central part of the study area (Fig. 3). It is represented by alternations of volcanic and sedimentary rock units and has a thickness of c. 750 m (see Fig. 5). At the base, the volcanic section has a sharp tectonic contact with the plutonic rocks (gabbro) of the K6mtirhan ophiolite, as seen along the Baskil-Ku~sarayl road, and is intruded by the Baskil granitic rocks at Sapanh and south of Kargada~l (Fig. 3).

331

Massive to stratified lithologies of the volcanic section are pillow lavas, lava breccias, massive lava flows, debris flow, alternations of volcanogenie sandstone and siltstone, siliceous tuff, mudstone-limestone alternations and columnarjointed lava flows (Fig. 5). The volcanic rocks are characterized by basalt, basaltic andesite, andesite, dacite and rhyodacite including secondary gypsum as veins and massive sulphfide deposits (B61ficek et al. 2004) (see Fig. 5). The basalts display amygdaloidal, intersertal, hyalomicrolitic porphyritic to microlitic porphyritic textures and are dominated by plagioclase and pyroxenephyric lavas. The andesites show amygdaloidal, hyalomicrolitic porphyritic to microlitic porphyritic textures and are plagioclase and amphibolephyric lavas. The rhyodacites display microlitic porphyritic to microgranular porphyritic textures, and are represented by plagioclase-phyric lavas. Plagioclase is seen either as phenocrysts or as microliths within the matrix. Euhedral to subhedral corroded quartz forms phenocrysts. Dacites show hyalo-porphyritic to amygdaloidal textures and are dominated by zoned plagioclase and corroded quartz set in a fine-grained matrix. Common secondary phases in the volcanic rocks are epidote, chlorite, calcite, albite, kaolinite and opaque minerals. The sheeted dyke complex of the K6mfirhan ophiolite is represented by diabase, microdiorite and quartz microdiorite, and is well preserved in the eastern part of the study area, east of Katm~hkda~ (Figs 3 and 4). Individual dykes exhibit variable thicknesses ranging from 15-20 cm to 100-150 cm, without obvious chilled margins. The dykes display intergranular, doleritic and microgranular textures. The main mineral phases are plagioclase, pyroxene, amphibole, quartz and magnetite. The sheeted dyke rocks are often associated with secondary calcite, amphibole, chlorite and epidote. The isotropic gabbros of the K6miirhan ophiolite crop out extensively in the southern part of the study area, between Eskik6y and Kamx~hk (Figs 3 and 4), and are represented by gabbro, diorite and quartz diorite. Gabbros display a non-cumulus granular to poikilitic texture and are characterized by primary plagioclase (An55 6o) (60-70 vol.%), clinopyroxene (15-20 vol.% ), orthopyroxene (< 5 vol.%) and opaque minerals (Fe-Ti oxide). Diorites exhibit granular to intergranular texture and are represented by plagioclase (Al135~40) (60-70 vol.%), amphibole (30 vol.% ), quartz (c. 1 vol.%) and opaque minerals (Fe-Ti oxide). Quartz diorites display granular textures and comprise slightly zoned plagioclase (50 vol.% ), amphibole (25 vol.% ), quartz (15-20 vol.% ) and opaque minerals (Fe-Ti oxide). The

332

T. RIZAOGLU E T AL.

Fig. 4. Tectonomagmatic-stratigraphic units in Baskil-Sivrice (Elazl~) region. rock isotropic gabbros include secondary calcite, chlorite, epidote and kaolinite. The ultramafic to mafic cumulate rocks of the K6mfirhan ophiolite crop out at Karada~l, Cortunlu and Kaml~hkda/g (Fig. 3). Ultramafic cumulates consist of wehrlite, whereas mafic cumulates are represented by olivine gabbro, gabbro-norite, gabbro and amphibole gabbro. The wehrlite displays a granular texture and is represented by olivine (60-70 vol.%), clinopyroxene (20-30 vol.%) and chromite (1-2 vol.%). The olivines and pyroxenes in the wehrlites are serpentinized to variable degrees. The olivine gabbro displays granular to poikilitic textures: it comprises olivine (Fo73_76; 20-30 vol.%) with a grain size of 1-6 mm, plagioclase (An92_94; 50-80vo1.%) with a grain size of 0.4-7 mm,

Fig. 5. Measured stratigraphic section from the volcano-sedimentary rocks of the K6mfirhan ophiolite.

clinopyroxene (En6970Wo22_27Fs4_8; 5-30vo1.%) with a grain size of 1~1 mm, orthopyroxene (En76 77W00.60.7Fs2223; < 5 vol.%) with a grain size of 1-5 mm, chromite and Fe-Ti oxide minerals. Serpentine, chlorite, talc, epidote and amphibole are secondary phases. The gabbronorite displays granular to poikilitic textures and is characterized by clinopyroxene (En40_51Wo21m4 Fs7 26; 20--30 vol.%) with a grain size of 0.53 mm, orthopyroxene (Ensv_61WOl.3_z.zFs37m0;

OPHIOLITES AND GRANITES, SE TURKEY 10-15 vol.%) with a grain size of 0.5-2.5 mm,

plagioclase (An53_77;c. 50 vol.%) with a grain size of 0.5-4 mm and opaque (Fe-Ti oxide) minerals. The gabbro displays granular to poikilitic textures and is characterized by plagioclase (60-80 vol.%) with a grain size of 0.5-7.5 mm, clinopyroxene (15-20%) with a grain size of 1-7 mm, orthopyroxene (1-2%) and amphibole (3-5%). Kaolinite, sericite, chlorite and magnetite are secondary phases. The amphibole gabbro has a granular to poikilitic texture and is represented by plagioclase (An43_57; 80-85%), amphibole (10-15%), biotite (2-3%) and opaque minerals (1-2%). Mantle tectonites within the K6mfirhan ophiolite are very limited, and are observed only in the SW of the study area (Fig. 3). The rock units are of serpentinized dunite, harzburgite, lherzolite and serpentinite. The metamorphic sole rocks crop out in the southwestern end of the study area especially south of Karakaya Tepe (Figs 3 and 4) where they have a tectonic contact with the mantle tectonites. They are cut by synkinematic granites near K6mfirhan bridge. The metamorphic sole is represented by amphibolite, plagioclase amphibolite, plagioclase-epidote-amphibole schist, quartz-plagioclase-amphibole schist and metasediments. The amphibolites exhibit granoblastic texture and comprise coarse-grained magnesio-hornblendes. The plagioclase amphibolites show granoblastic to grano-nematoblastic texture and are represented by plagioclase (20-25%), amphibole (70-75%) and accessory sphene and magnetite minerals. The plagioclaseepidote-amphibole schists display banded to nematoblastic textures and comprise amphibole (50-60%), epidote (15-20%), plagioclase (510%), secondary chlorite and magnetite. The quartz-plagioclase-amphibole schists exhibit banded to nematoblastic textures, and are characterized by amphibole (70-75%), plagioclase (e. 20%), quartz (c. 5%) and accessory sphene and magnetite.

Geochemistry Analytical methods A total of 75 samples from the metamorphic sole (11), cumulate (19), isotropic gabbro (six), sheeted dyke (15) and volcanic rocks (24) of the K6mfirhan ophiolite were analysed for major and trace elements by standard X-ray fluoresence (XRF) spectrometry. Major element contents were determined on glass beads fused from ignited powders to which LizB407 was added at a ratio of 1:5, in a gold-platinum crucible at 1150 ~ Trace element contents were measured

333

by XRF on pressed-powder pellets. A subset of 29 samples were also analysed for trace elements (including rare earth elements (REE)) by inductively coupled plasma-mass spectrometry (ICP-MS) at Acme Analytical Laboratories in Canada. The results of the analyses are presented in Tables 1 and 2. A total of eight representative polished sections were used for electron microprobe analysis on a JEOL JXA-8600 instrument in the Geology and Paleontology Department at Salzburg University (Austria). The analytical conditions for the elements were a counting interval of 13 s (10 s for peak and 3 s for background), a beam current of 20 nA and an acceleration voltage of 15 kV. The data reduction was done following the ZAF procedure. Fe 3+ and Fe 2+ were determined from stoichiometry of spinel using the equation of Droop (1987). The results of the analyses are presented in Tables 3-5.

Whole rock Major, trace and rare earth elements are given in Tables 1 and 2 for the volcanic, sheeted dyke, isotropic gabbro, cumulate and metamorphic sole rocks of the K6miirhan ophiolite. Loss on ignition (LOI) values reach 9.12% in the volcanic rocks, 2.41% in the sheeted dykes, 3.21% in the isotropic gabbros and 4.1% in the metamorphic sole rocks, reflecting variable secondary alteration, which is indicated by the presence of mineral phases such as epidote, calcite or chlorite. The mobility of many elements during low-grade submarine alteration has been well constrained by a number of studies (e.g. Hart et al. 1974; Humphries & Thompson 1978). For this reason, recourse is generally made to relatively immobile elements such as Ti, P, Zr, Y, Nb and REE, and to a lesser extent Cr, Ni, Sc and V, to designate lava groups, petrogenetic trends and tectonic environments (Pearce & Cann 1973; Floyd & Winchester 1975, 1978; Pearce & Norry 1979). The rock classification diagram for the rock units of the K6miirhan ophiolite is based on Zr/Ti v. Nb/Y (Pearce 1996). The lavas cover a wide compositional range from basalt to rhyodacite, but basaltic to andesitic compositions predominate (Fig. 6a). The sheeted dykes are characterized by diabase and microdiorite, whereas the isotropic gabbros are dominated by gabbroic rocks (Fig. 6b and c). The amphibolites of the metamorphic sole reflect their derivation from a basaltic protolith (Fig. 6d). Nb/Y ratios are in the range of 0.5-0.02 in volcanic rocks, 0.19-0.09 in sheeted dykes, 0.2-0.06 in isotropic gabbros and 0.33-0.07 in metamorphic sole rocks, indicating that all the analysed rocks from the K6miirhan ophiolite are tholeiitic in character (Pearce 1982)

334

T. RIZAOGLU E T AL. ' , ~ ~"-- O~ ~

oO o o r

t"q o o ',,~ er

',~- ',:~1" ~'.-

cq. "~. ~. o. ,--Z.~ eq.. ~. ~. o. o. eq.. o. o.

,-6 ~. o. ~. ~ --. ~. o. ~ ~. o. ,q. O. O. ~.

r

~D

~"~ oO t ' ~ "~" ,.~" tt'~ ~

['--.- ,-.-~ r

t"..- ' , ~ er

ee'~

.,~1- .,... [-.~ r ~ t-,q ,...~ tt-~ ,...~ ~ OO t.',q ~ " ~ - ~ t " ~ ' ~ , , ' ~ O O ~ C ' q ~ O

O.,~ ~,

~ o. ~. ~ ~ ~ o. o. o. o. ~. ~ ~ ,~.

r

~

',,D OO oO ,..~ r

r

t.'.q ~

,---.O'~oO

e-i I"--

t"-q ,.~- ,..~ ee'~ OO

~. e,-].~ eq.. --. t--: ,~. ~ --. O. ~. ~ o. o.

tt~ r

. ~ eq.. e-:. ~. ~.

o eq.. ~. ~ ~. O. o. ~. r

~5 r

~-,O

~

O",

t"q ,---.

tt~

r162 .~

o6 ,--; ,4 o ~5 t--: ~ ,,A o c5 c-,i o c5 o6

.~- r--- t--r t'q

" ~ 0 o ~".,.- ~ 1 o'~ o") 0~

',,,~ o o 0~.O 0 .,.-..~ ~

...~

tt'5 t"xl

tt'3 r

tt~

~,..,.q ~:3~ t..., t"~

oO

r

r

,,6 r oo

t"q

eZ

~..., , . ~ tt'~ .,.~ ~

m

..~' o o tt"~ ~

~

tt'~ ~

~

~

06

.~

OPHIOLITES A N D GRANITES, SE TURKEY

V

V

V

o

o

V

o

o

o

335

oc5c5

~ @ ~ v

Svvv ~

- -

~

~

s

o

~

~

~

-

~

~

2

~

4- ~

oe .i~ V O O

~

~

V

o

o

V

~m o

e,-~ em o~ c5,#o e-I oo

cq

~q

V

o

o

V

V ~

m

2

336

T. R I Z A O ( ~ L U

.

.

.

~

q

~ o ~ o ~

o

~

E T AL.

.

. ~

.

tr

',~oo

rr

~

r',l ~ rr I~

t"-I

~

t"-I

~

~ o ~

.

.

.q~

~

q

o

.

~

o

.~ er

k,~

er

eq

',~1-

~

I~

r

~

i

~ o ~

. o o

. ~ o ~

oR. oq. ~.

oo~o ,~r-,

00~

0

r.~

|

i

i

,.4, |

,~,- t"xl

o~. ~0

~o~ i

oo

i

i

2~ [.,

.o o

d

O P H I O L I T E S A N D G R A N I T E S , SE T U R K E Y

m m ~ ~ o m ~

.~

~

337

. ~ 1 7 6 1 7 6 1 7 6

~

0

0

0

0

0

0

~

0

0

0

0

~~ ~

~ o o o o ~ o ~ o o o o o

~

.

~

o

~

o

~

~

~ 0 ~ ~ 0 ~ 0 0 0 0 0 ~

0

~

0

0

0

0

~

0

0

0

.~

0

~

~

~ 0

~

. 0

~

~ ~ 1 7 6 1 7. ~ 6 1 7 6 1 7 6 1. 7~ 6~ 1 7 6

i

~

0

0

~

~

0

~

0

0

0

0

~

0

~

0

0

~

~

0

0

0

~ 0

0

0

O 0

0

0

~

~

0

0

0

~

0

0

~

0

~

O ~

O 0

o ~ ~ 1 7 6 1 7 6. ~ ~ 1 7 6 1 7 6 1 7 6 1 7 6

!

r

,~.

r

o,h

s

s

i

~ ~

~o

. ~ ~ ~ 1 7 6 1 7 6 .~

.

.

i

r

.

o

.

~

.

.

.

.

0

~

0

~

.

.

- -

.

.

~ o o o o o o ~ o o o o ~

~

0

~

0

0

~

~

0

~

0

~

0

0

~

0 0 0

0

0

0

~

0

0

~

~

0

~

0

0

~

0

O 0

~

O 0

~

i

eel

|

.

oh

~

.00

.

~

~ 0 0 ~ 0 ~ 0 ~ 0 0 0

~

.

|

~o r

o

,..t,

o

~6

9

i t',l

i r

o ~

0

~

0

~

0

~

0

0

~

~

0

0

0

0

0

0

0

0

0

0

0

~

0

~

i r o on r

N ooc

o

oZ~

Z~

338

T. RIZAOGLU E T AL. |

i

. ~ o ~ m ~

.~

~

o

.oqo

. ~ o

.

i

~ d d d d ~ d d

A

A o

o

~ d d d d d d d ~ d d ~

g i

i

i

i

er

t~

i

er 3 t ~

.-k

o

eel t'q

i

i

~

. o ~ m ~ o ~

~

~

o

.

o

~

~

o~

o

r

2~ [,,.

r~

z~

OPHIOLITES AND GRANITES, SE TURKEY

'"(a]Volcanic~ ~ iAlkali !''rocks t , (b)Sheeted~ i

~

.P,~

.

339

Alkali:l''gabbros ~ i t

Rhyolite&Dacite "

~

(d)Met's~

Alkali

~

Rhyolite~: Dacite "]~

Basalt A

Basalt

0.1 def

0.01

Basalt 0.001

[

........

0.01

n

Basalt

.....

h'l

0.1 Nb/Y

........

n

1

..............

u

0.1 Nb/Y

|

n

n

| n l l n |

0.1 NbfY

|

9

9

,,

0.1 NbfY

Fig. 6. Rock classification diagrams based on Nb/Y v. Zr/Ti (Pearce 1996) for the K6miirhan ophiolite rocks.

lOa'V~ OO l rocks/ 1

o

I

(b) Sheeteddykes

(c) Isotropicgabbros

/]

0

I-.!

zx

oo'

(d) Met. sole rocks

tl

o

%

"~ 100

10 0.01

.

.

Th~

.

.

.

.

.

"91

0.1 N b/ Y

Tholeiitic

~/

.

.

.

.

.

.

.

.

.

.

.

.

.

.

9

.

.

Tholeiitic

.

.

0.1 Nb/Y

.

.

.

.

un

1

.

.

.

.

.

.

.

.

Tholeiitic

9

0.1 Nb/Y

.

1

.

.

.

.

.

.

.

uu

0.1 Nb/Y

1

Fig. 7. Nb/Y v. Ti/Y diagrams showing tholeiitic nature of the K6miirhan ophiolite rocks (Pearce 1982). (Fig. 7). To exhibit the chemical relationships of the K6miirhan ophiolite rocks, several diagrams based on immobile elements are presented in Figure 8. In the TiO2 v. Zr diagram (Fig. 8a), the volcanic rocks define a decreasing trend with increasing Zr from basic (1.43%) to acidic (0.21%) rocks, suggesting magnetite or titanomagnetite crystallization in the more evolved rocks. By contrast, the sheeted dyke rocks have a high content TiO2 (1.20-2.69%) compared with the volcanic rocks and the metamorphic sole rocks (0.16-1.28%). The Y and FeO*/MgO ratios

of the rocks, plotted against Zr in Figure 8b and c show a positive correlation and coherent trends. By contrast, a decreasing Y content in some of the acidic volcanic rocks may be caused by amphibole fractionation. The chemical features displayed suggest that the volcanic rocks, sheeted dykes and the protolith of the metamorphic sole rocks represent a differentiated co-magmatic tholeiitic suite, exhibiting similar fractionation trends. Representative analyses of major and trace element contents of the cumulate rocks and the

340

T. R I Z A O ( ~ L U 3.0

Ca) +

2.5

++ 2.0

+ +

,~

1.5

O

L0 0.5

<>

0.0 0

D I

I

I

!

I

50

100

150

200

250

300

Zr (ppm) 80

(.~ Basic-intermediate volcanic 7 0 - [ ] Acidic volcanic --1- Sheeted dyke Metamorphic sole 60

(b)

+

§

++

50

..~ 40

[] D D

30

DD

20 10 0

I

0

50

I

I O0

I

I

I

150

200

250

300

Zr (ppm)

(c) 6 5

[] + +

2

+

O

1

0

I

0

50

I

100

J

I

I

150

200

250

300

Zr (ppm)

Fig. 8. Major and trace element variations against Zr for the K6mfirhan ophiolite rocks. isotropic gabbros are presented in Table 1. Loss on ignition (LOI) values are up to 3.05~ in the mafic cumulates and 6.97% in one ultramafic cumulate rock sample (Table 1), indicating variable amount of serpentinization or alteration. The A1203, CaO, Ni and Cr contents of the isotropic gabbro and the ultramafic to mafic cumulate rocks are plotted against Mg-number (100• as an indication of the degree of differentiation (Fig. 9). The CaO

E T AL.

content is 6.09 wt% in wehrlite and ranges from 20.9 to 9.44 wt% in gabbroic rocks, and from 15.59 to 8.72 wt% in the isotropic gabbros; it is negatively correlated with MgO (Fig. 9b). The A1203 content, which is also negatively correlated with increasing MgO, shows lower values in wehrlite (5.73 wt%), but higher values in cumulate gabbros (from 27.57 to 13.39 wt%) and in isotropic gabbros (from 18.39 to 14.81wt%) (Fig. 9a). The high CaO and A1203contents in the gabbroic rocks are an indication of the presence of plagioclase (An95~5). Ni and Cr contents decrease markedly from high values in wehrlite (633 ppm for Ni and 1081 ppm for Cr) to much lower values in plagioclase-rich gabbros (from 306 to 3 ppm for Ni and from 1483 to 4 ppm for Cr), consistent with the fractionation of olivine, spinel and clinopyroxene (Fig. 9c and d). The REE patterns of the volcanic, sheeted dyke and metamorphic sole rocks are presented in Figure 10. The basic to intermediate volcanic rocks exhibit (1) flat [(La/Yb)N=l.65-0.75] and (2) marked light rare earth element (LREE) enrichments with respect to heavy rare earth element (HREE) ((La/Yb)N=4.96-4.36). The acidic volcanic rocks also exhibit similar REE patterns; one group has a flat pattern ((La/Yb)N=l.12-0.87) and a second has an LREE-enriched pattern ((La/Yb)N=7.89-5.18) (Fig. 10). These samples exhibit a slight Eu negative anomaly, as a consequence of the removal of feldspar by fractional crystallization or the partial melting of a source material in which feldspar is retained in the source (Rollinson 1993). The sheeted dyke complex generally exhibits slightly LREE-depleted to flat ((La/Yb)N=0.96-0.57) REE patterns with an overall enrichment of 10-30times chondritic values (Fig. 10). The metamorphic sole rocks display slightly LREEenriched ((La/Yb)N = 2.99-1.40) to LREEdepleted ((La/Yb)N = 0.42) patterns (Fig. 10). The enrichment of the LREE is commonly interpreted as a consequence of mantle source enrichment by subduction-derived components (Floyd et al. 1991; Bortolotti et al. 2004). Flat to LREEenriched patterns are typically found in islandarc tholeiites (Pearce 1982; Peate et al. 1997) and suprasubduction-zone type ophiolites of the eastern Mediterranean (Desmons et al. 1980; Searle et al. 1980; Alabaster et al. 1982; Pearce et al. 1984; Parlak 1996; Yahmz et al. 1996, 2000; Parlak et al. 2000; A1-Riyami et al. 2002). Figure 11 presents normal mid-ocean ridge basalt (N-MORB)-normalized spider diagrams of the volcanic, sheeted dyke and metamorphic sole rocks of the K6mfirhan ophiolite. Some general features include (1) enrichment in large ion lithophile elements (LILE; Rb, Ba, Th, K)

O P H I O L I T E S A N D G R A N I T E S , SE T U R K E Y 30

(b)

. ~ Wehrlite i(a) ~ Olivine gabbro

\.9

Gabbro

25

g

341

C~

Diorite ~ [A Is~ gabbro L. 2

2O

g

e~

A AL: ....

0

.....

15

L)

!

i

I

1

I

0I

5

10

15

20

25

30

0 5

'

35

0

'!

'

5

i'

i

i

!

I

10

15

20

25

30

MgO (wt %)

35

MgO (wt %)

700

(c)

600

1200

{,i~)A

Q-;

(d)

1000 -

5OO

~'

g

400

800

-

600 -

300 j-

200 100

9

0 0

5

!

I

I

!

l

10

15

20

25

30

400 -

~,

200 -

zx A (!~5

0

35

0

,li-~

!

I

I

I

!

5

10

15

20

25

30

MgO (wt %)

35

MgO (wt %)

Fig. 9. Selected major and trace element variations for the gabbroic and wehrlitic rocks. elements; (2) depletion in Nb; (3) flat patterns of high field strength elements (HFSE) relative to N-MORB (Fig. 11). Th enrichment (together with the LREE) and N b - T a depletion are features of subduction-related volcanic rocks (Wood et al. 1979; Pearce 1983; Arculus & Powel 1986; Yogodzinski et al. 1993; Wallin & Metcalf 1998). The Th enrichment and Nb depletion of the K6miirhan ophiolite rocks imply their formation in a subduction-related tectonic setting. Nb/Th ratio v. Y discriminates between subduction and non-subduction settings based on Nb enrichment or depletion (Jenner et al. 1991). The volcanic rocks, sheeted dykes and metamorphic sole rocks of the K6miirhan ophiolite plot within the arc-related field (Fig. 12a). The Th/Yb v. Ta/Yb plot discriminates between depleted mantle (MORB) and enriched mantle (intraplate) sources (Pearce 1982). Addition of a subduction component from slab-derived fluids or melts results in an increase in Th/Yb in the mantle source, as shown by the arrow (Fig. 12b). On this diagram, all the rocks plot within the volcanic arc field. The T h - H f - N b triangular

diagram (Wood et al. 1979), and the Z r - N b - Y triangular diagram (Meschede 1986) discriminate volcanic rocks erupted in different geotectonic settings. The volcanic rocks, sheeted dykes, isotropic gabbros and metamorphic sole rocks from the K6mfirhan ophiolite plot within the subduction-related field (Fig. 13a and b). Mineral chemistry

Cumulus olivine (unzoned) analyses from the gabbroic cumulate rocks are presented in Table 3. Their Fo contents range from 76.2 to 73.9. NiO content ranges from zero to 0.06% (Table 3). Representative plagioclase analyses from the mafic cumulate rocks are presented in Table 3. The plagioclase has a very wide compositional range: from An94.8 to An92.2 in olivine gabbro, from An77.8 to An5z2 in gabbronorites, and from An56.9 to An43.4 in amphibole gabbro (Table 3). The basic to evolved rock types in the cumulates have resulted in variable An contents. Plagioclases in the amphibole gabbro exhibit both reverse and normal zoning, whereas plagioclases

T. R I Z A O G L U

342

E T AL.

1000

Basic-intermediate volcanic rocks

g

Sheeted dyke rocks

100

~ lO

Metamorphic sole rocks

Acidic volcanic rocks .~ lOO

~

lO

r~

i

i

i

i

I

i

i

i

I

I

i

i

i

i

i

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

i

I

i

i

i

I

/

i

i

i

i

i

i

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 10. R E E d i a g r a m s o f the K 6 m f i r h a n ophiolite rocks ( n o r m a l i z i n g values are f r o m Sun & M c D o n o u g h 1989).

1000

Basic-intermediate volcanic rocks

Sheeted dyke rocks

', ', ', ', ', ', I ', ', ', ~ ', ', ~ ', ~ ', ', ', ', ',

', ', ', ', ', ', ~ ', ', ', ', ', I ', ', ', ', ', ', I ~

Acidic volcanic rocks

Metamorphic sole rocks

O) 100

0.1

100 O~

9

: ~-O

B 0.1

i

Rb

i

i

Th

i

i

Nb

|

i

La

i

i

Pb

i

i

Sr

i

i

Nd

i

i

i

i

Sm Ti

i

i

Y

i

i

Lu

I !

Rb

|

Pr

K i

Th

i

!

Nb

!

!

La

i

!

Pb

i

i

Sr

i

!

Nd

i

1

i

i

Sm Ti

i

i

Y

Fig. 11. S p i d e r d i a g r a m s o f the K 6 m i i r h a n ophiolite r o c k s ( n o r m a l i z i n g values are f r o m Sun & M c D o n o u g h 1989).

!

!

Lu

OPHIOLITES AND GRANITES, SE TURKEY

186](a)

i'(volcanicrockBa~iai"~ea' 1 i"te (a)

ll~.14 NON-ARC l0

t

0

l0

I r'qAcidievoleanierock| + Sheeted dike |

,

-tk

20

30

40

50

60

70

I

/N, / \ A\

343

/

llf/3 A / \ A\

/

A / \ A\

A:N-MORB B:E-MORB C:WPB :

80

Y (ppm) lO

," /

,,,- ~o 0

"0

O

0 ~

~r @trio

:o

f.

o.1.

u

/"

Th

/ ( :~,~ )

9 Mafic & intermediate volcanic rocks [] Acidic volcanic rocks + Sheeted dyke rocks A Isotropic gabbro <>Metamorphic sole rocks

," / . . : /

/

/

~b~ /,oq,-~

, ~ - 4 , ~ , ~-

//~<,""

(b)

Nb/16

Nb*2 ] ~ ~ A :

WPB

i D:N-MORB-VAB

I,+++ . ~ / ~

.r.+ ~,',,-r

v,;@- t 0.01

l

0.01

. . . . . . . .

l

0,1

. . . . . . . .

i

1

........

Ta/Yb Fig. 12. (a) Nb/Th v. Y diagram (after Jenner et al. 1991) and (b) Ta/Yb v. Th/Yb diagram (after Pearce 1982), showing the typically arc-like signature of the K6miirhan ophiolite rocks.

10

in more basic cumulate rocks (gabbronorite and olivine gabbro) are unzoned. Clinopyroxene analyses from the gabbroic cumulate rocks are presented in Table 4. In terms of quadrilateral components, the cumulus clinopyroxene composition is En50.6-39.5Fs23.7_14.6Wo44.926.8 in gabbronorite, and En69.9_69.sF $8.0_3.7W026.8_22.2 in olivine gabbro. The Mg-number of the clinopyroxene ranges from 74 to 68 in gabbronorite, from 95 to 90 in olivine gabbro (Table 4). Representative orthopyroxene analyses from the mafic cumulate rocks are presented in Table 4. The orthopyroxene composition is En61.3 57.9Fs40.4~36.aWo2.3-1.3in the gabbronorite, and Envv.3_76.zFs23.1_22.lWO0.8_0.6 in olivine gabbro. The Mg-number of the orthopyroxene ranges from 63.9 to 60.1 in the gabbronorite, and from 78.2 to 77.3 in the olivine gabbro. Representative analyses of amphiboles from the cumulate rocks are presented in Table 5. The amphiboles in the amphibole gabbro rocks are primary and represented by magnesio-hornblende, whereas the amphiboles in the gabbronorite and olivine

Zr/4

Y

Fig. 13. (a) Th-Hf-Nb (after Wood et al. 1979) and (h) Zr-Nb-Y (after Meschede 1986) tectonomagmatic discrimination diagrams for the rocks of the K6mfirhan ophiolite. N-MORB, normal mid-ocean ridge basalt; E-MORB, enriched MORB; WPB, within-a-plate basalt; IAB, island-arc basalt; VAB, volcanic arc basalt. gabbro are secondary, derived from alteration of pyroxenes and represented by magnesiohornblende in gabbronorite and tschermakite in olivine gabbro (Table 5). The Mg-number of the amphiboles is 56.4-59.4 for amphibole gabbros, 70.3-63.7 for gabbronorites, and 77.9-75.9 for olivine gabbro. Covariation of An content of plagioclase v. Mg-number of orthopyroxene and olivine for the gabbroic rocks is shown in Figure 14a and b, together with results from the Troodos (H6bert & Laurent 1990), Mersin (Parlak et al. 1996), Pozann-Karsantl (Parlak et al. 2000) and Klzdda~ (Ba~cl et al. 2006) ophiolites, and other comparable units from well-documented tectonic settings. The mineral compositions of the gabbroic rocks from the K6miirhan ophiolite show

344

T. RIZAOGLU E T AL.

Fig. 14. (a) Anorthite content in plagioclase (mol%) v. Fo (mol%) content in olivine. (b) Anorthite content in plagioclase (mol%) v. enstatite content in orthopyroxene (mol%) for the K6mtirhan ophiolite. The Troodos ophiolite trend is from H6bert & Laurent (1990). The Mersin ophiolite trend is from Parlak et al. (1996). The Pozantl-Karsantl ophiolite trend is from Parlak et al. (2000). The Klzllda~ ophiolite trend is from Ba~cl et al. (2006). R, Rindjami volcano (Foden 1983); B2, B3a, Boisa volcano (Gust & Johnson 1981); U, Usa volcano (Fujimaki 1986); A, Agrigan volcano (Stern 1979); Data for Lesser Antilles are from Arculus & Wills (1980).

a close similarity to SSZ-type ophiolites of the eastern Mediterranean region and known islandarc settings (Stern 1979; Arculus & Wills 1980; Gust & Johnson 1981; Fujimaki 1986).

Discussion The most important tectonomagmaticstratigraphic units of SE Anatolia in the Late Cretaceous were (1) metamorphic massifs (i.e. Malatya-Keban platform), (2) ophiolites (i.e. G6ksun, [spendere, K6mfirhan and Guleman), (3) volcanic arc units (i.e. Yfiksekova-Elazl(~ magmatics units); (4) granitic rocks (i.e. Baskil). The SE Anatolian orogenic evolution involved progressive relative (southerly) movement of the nappes towards the Arabian plate during Late Cretaceous-Miocene time (Ylldlrlm & Yllmaz 1991; Yllmaz 1993; Yllmaz et al. 1993; Robertson et al. 2006, 2007). The MalatyaKeban platform, belonging to the upper part of the nappe zone of the SE Anatolian orogen (Ydmaz 1993), was amalgamated with the SSZtype ophiolites (G6ksun-K6mfirhan-ispendereGuleman) and their arc-related volcanic succession (Yfiksekova complex-Elazl~ magmatic rocks) around 88-85 Ma (Parlak, 2006; Robertson et al. 2006, 2007).

The nature of the volcano-sedimentary units interbedded with the lavas, the major and trace element geochemistry of the volcanic rocks, as well as their wide compositional range (from basalt to rhyodacite) suggest that this assemblage represents the upper levels of an intra-oceanic volcanic arc formed above of north-dipping subduction zone during Late Cretaceous time. A similar volcano-sedimentary rock association has been reported from the same belt, notably the G6ksun (N Berit) ophiolite (Parlak et al. 2004; Robertson et al. 2006) and Late Cretaceous arc-related rocks around Elazl(g (B61ficek et al. 2004; Robertson et al. 2007). The sheeted dyke and isotropic gabbroic rocks in the K6mfirhan ophiolite are tholeiitic in character (Nb/Y=0.2-0.06). The REE patterns, multi-element and tectonomagmatic discrimination diagrams suggest their formation in a subduction-related environment. The major and trace element geochemistry of the cumulate rocks is similar to that observed in a modern island-arc tholeiite (IAT) sequence. Crystallization of calcic plagioclases in olivine gabbros and amphibole is suggestive of hydrous conditions during magma differentiation. The presence of intercumulus water in the cumulate rocks (Arculus & Wills 1980) may be responsible for the reverse zoning

OPHIOLITES AND GRANITES, SE TURKEY

345

Fig. 15. Tectonic model for the genesis of the ophiolites, related metamorphic units and granitoids in SE Anatolia. (See text for discussion.). of plagioclase observed in the amphibole gabbro. The amphibole gabbros are widespread in central Anatolia and are interpreted as being formed from a wet magma by high-degree partial melting of peridotite in a subductionrelated setting (Kodak et al. 2005). High-level amphibole-bearing gabbros are reported from other eastern Mediterranean ophiolites (H6bert & Laurent 1990) and island-arc settings (Debari

& Coleman 1989). Water-rich magmas are observed in arc regions where amphiboles commonly form (Arculus & Wills 1980). The geochemical evidence from the volcanic and plutonic rocks of the K6miirhan ophiolite shows that these are cogenetic tholeiitic suites formed in an SSZ tectonic setting during the Late Cretaceous in the southern branch of Neotethys. The metamorphic sole rocks of the K6mtirhan ophiolite

346

T. RIZAOGLU E T AL.

exhibit a tholeiitic (Nb/Y=0.07-0.33) nature. The REE patterns, multi-element and tectonomagmatic discrimination diagrams suggest that the protolith is akin to IAT. The granitic rocks related to the evolution of the southern Neotethys in SE Anatolia are observed at three localities the G6ksun-Af~in (Kahramanmara 0, Dofgan~ehir (Malatya) and Baskil (Elazlfg) regions (Asutay 1988; Yazgan & Chessex 1991; Akgiil 1993; Onal 1995; Parlak 2006; Robertson et al. 2006, 2007), as intruding the tectonostratigraphic-magmatic units of the nappe zone (Yllmaz 1993). The most important point is that the granitic rocks are seen to intrude all of the Malatya-Keban platform, the ophiolites and the related metamorphic units, suggesting that the Malatya-Keban platform and ophiolitic units were tectonically juxtaposed before the intrusion took place in the Late Cretaceous (Yazgan & Chessex 1991). The K-Ar isotopic age determinations on the granitic rocks of the region range from 76_+2.45 to 78+2.5 Ma for the Baskil (Elazlfg) area (Yazgan & Chessex 1991) and from 85.76+3.17 to 70.05_+ 1.75 Ma for the G6ksun (Kahramanmara 0 area (Parlak 2006). This suggests that the granite intrusion may be only slightly younger than the formation of the ophiolites (c. 90 Ma) (Mukasa & Ludden 1987; Robertson et al. 2006). Moreover, the field evidence throughout the region shows that the ophiolites and related metamorphic rocks were accreted to the base of the Malatya-Keban platform before intrusion took place. A number of alternative tectonic models have been proposed to explain the genesis and emplacement of the ophiolites and the granitoid magmatism in the region. These are a the 'single subduction zone' model (Hall 1976; Akta~ & Robertson 1984, 1990; Robertson 1998, 2000; Yllmaz 1993; Yllmaz et al. 1993), a 'double subduction zone' model (Robertson 1998, 2000, 2002; Parlak et al. 2004) and a 'multi-phase convergence' model (Robertson et al. 2007). The field, geochemical and geochronological evidence from the Elam~ as well as the Kahramanmara~ (Parlak & Rlzao[glu 2004; Parlak 2006) and Malatya regions are consistent with the 'multiphase convergence' model of Robertson et al. (2006, 2007). The following evolutionary scenario based on this model is proposed for the tectonomagmatic units discussed above. The K6m/irhan ophiolite was formed above a northdipping subduction zone between the Arabian platform to the south and the Tauride platform to the north in the Late Cretaceous (c. 90 Ma) (Fig. 15a). Following this, an ensimatic island-arc assemblage was constructed on SSZ-type crust (Fig. 15b). The metamorphic sole was formed

by the metamorphism of IAT-type basalts that were detached from the front of the overriding K6miirhan ophiolite and then underplated (Fig. 15b). Northward underthrusting of cold oceanic crust (possibly Early Mesozoic in age) was then initiated beneath the Tauride platform to the north (Fig. 15b). The K6miirhan ophiolite, the related metamorphic rocks and the volcano-sedimentary unit were then accreted to the base of the Tauride active margin in the north, where all the units were cut by the Baskil granitic rocks around 88-85 Ma (Fig. 15c).

Conclusions (1) Major, trace and rare earth element geochemistry of the volcanic and subvolcanic rocks, as well as the mineral chemistry of the cumulate rocks from the K6mfirhan ophiolite, suggest that they formed in an SSZ tectonic setting in the southern Neotethys during Late Cretaceous time. The presence of highly evolved rocks in the volcanic (andesite-dacite), sheeted dyke (microdiorite, quartz microdiorite), isotropic gabbro (diorite, quartz diorite) and cumulate gabbro (amphibole gabbro), as well as the high Ca-plagioclase in the mafic cumulate units are an indication of hydrous conditions, related to subduction. (2) The metamorphic sole of the K6mtirhan ophiolite was derived from metamorphism of an IAT-type basaltic protolith during intraoceanic thrusting-subduction. The volcanic rocks were presumably detached from the front of the overriding SSZ-type crust and then underplated and metamorphosed. (3) The widespread volcanic-sedimentary unit (c. 750 m thick) in the Baskil (Elazl~) area exhibits alternations of basic to acidic extrusive rocks, debris flows, volcaniclastic sandstones and pelagic limestones. This unit is interpreted as a tholeiitic ensimatic island-arc assemblage, and has no genetic link with the Baskil granitic rocks. This unit was formed during a mature stage of SSZ spreading as the upper part of the Late Cretaceous K6m/irhan ophiolite. (4) The granitic rocks of the Baskil (Elazl~) region are observed as intruding the ophiolites, the volcanic-sedimentary unit and the Malatya-Keban platform. This suggests that the ophiolites and the volcanicsedimentary unit were accreted to the base of the Malatya-Keban platform before the intrusions of the volcanic arc-type granitic bodies (i.e. Baskil, Do(gan~ehir and G6ksun) around 88-85 Ma in SE Anatolia (Parlak 2006).

OPHIOLITES AND GRANITES, SE TURKEY This work is a part of a PhD study by T. Rlzao~lu. Financial support from TUBITAK (Scientific and Technical Research Council of Turkey, Project No. 102Y041) and Ni~de University Research Foundation (Project No. FBE2002-08) is gratefully acknowledged. We would like to thank F. Capponi for performing major and trace element analyses. We are grateful to M. Delaloye for thoughtful discussions on the evolution of the eastern Mediterranean ophiolites. D. Topa is thanked for his guidance during the microprobe analysis at Salzburg University (Austria). O. Parlak gratefully acknowledges the financial support of TUBA (Turkish Academy of Sciences) in the frame of the Young Scientist Award Programme. A. H. F. Robertson, M. Delaloye and E. Yi~itba~ are thanked for their constructive reviews that improved the quality of the present paper.

References AKGUL, B. 1993. Petrographical and petrological fea-

tures of magmatic rocks in the vicinity of Piran village (Keban-Elazt~,). PhD thesis, Flrat University, Elaine, Turkey. AKTAw G. & ROBERTSON, A. H. F. 1984. The Maden complex, SE Turkey: evolution of a Neotethyan continental margin. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 375-402. AKTA~, G. & ROBERTSON, m. H. F. 1990. Tectonic evolution of the Tethys suture zone in SE Turkey: evolution evidence from the petrology and geochemistry of late Cretaceous and middle Eocene extrusives. In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds)

Ophiolites-Oceanic Crustal Analogues. Proceedings of the Troodos Ophiolite Symposium. Geological Survey, Cyprus 1987, 311-329. 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. AL-R1YAMI, K., ROBERTSON, A. H. F., DIXON, J. & XENOPHONTOS, C. 2002. Origin and emplacement of the Late Cretaceous Baer-Bassit ophiolite and its metamorphic sole in N W Syria. Lithos, 65, 225-60. ARCULUS, R. J. & POWELL,R. 1986. Source component mixing in the regions of arc magma generation. Journal of Geophysical Research, 91, 5913-5926. ARCULUS, R. J. & WILLS, K. J. A. 1980. The petrology of plutonic blocks and inclusions from the Lesser Antilles island arc. Journal of Petrology, 21, 743-799. ASUTAY, H. J. 1988. The geology of Baskil (Elaz:~) vicinity and petrology of Baskil magmatics. Bulle-

tin of Mineral Research and Exploration (Ankara, Turkey), 107, 46-72. BA~CI, U., PARLAK,O. & HOCK, V. 2006. Geochemical character and tectonic environment of ultramafic to mafic cumulates from the Tekirova (Antalya) ophiolite (southern Turkey). Geological Journal, On press).

347

BEYARSLAN, M. 1996. KOmiirhan ofiyolit Biriminin Petrog.rafik ve Petrolojik Ozellikleri. PhD thesis, Flrat Universitesi, Etaz1~, Turkey. BEYARSLAN, M. & BINGOL, A. F. 1996. The geochemistry and petrology of Elaz:~ magmatics. In: KORKMAZ, S. & AKCAY, M. (eds) 30th Annual Symposium Proceedings. Karadeniz Teknik Universitesi, Trabzon, Turkey, 208-224. BEYARSLAN, M. & BINGOL, A. F. 2000. Petrology of a supra-subduction zone ophiolite (Elaza~, Turkey). Canadian Journal of Earth Sciences, 37, 1411-1424. BiNGOL, A. F. 1984. Geology of the Elaz1~ area in the eastern Taurus region. In: TEKELi, O. & GONCrOOdLU, M. C. (eds) The Geology of the

Taurus Belt. International Symposium Proceedings, Ankara, Turkey. MTA, Ankara 209-216. BOLI')CEK, C., AKG1JL, M. & TI3RKMEN,I. 2004. Volcanism, sedimentation and massive sulfide mineralization in a Late Cretaceous arc-related basin, Eastern Taurides, Turkey. Journal of Asian Earth Sciences, 24, 349-360. BORTOLOTTI, V., CHIARI, M., MARCUCCI, M., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2004. Comparison among the Albanian and Greek ophiolites: in research of constraints for the evolution of the Mesozoic Tethys ocean. Ofioliti, 29, 19-35. 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, 4373-4391. DESMONS, J., DELALOYE,M., DESMET, A., GAGNY, C., ROCCI, G. & VOLDET, P. 1980. Trace and REE abundance in Troodos lavas and sheeted dykes, Cyprus. Ofioliti, 5, 35-56. DROOP, G. T. R. 1987. A general equation of estimating Fe 3§ concentrations in ferromagnesian silicates and oxides from microprobe analyses using stoichiometric criteria. Mineralogical Magazine, 51, 431-435. ERDEM, E. & B~NGOL, A. F. 1995. Pfitiirge (Malatya) metamorfitlerinin petrografik 6zellikleri. Fzrat

Universitesi Miihendislik Fakiiltesi Dergisi, 7, 73-85. ERDO~N, B. 1977. Geology, geochemistry and genesis

of the sulfide deposits of the Ergani-Maden region, SE Turkey. PhD thesis, New Brunswick University Brunswick. FLOYD, P. A. & WINCHESTER,J. m. 1975. Magma type and tectonic setting discrimination using immobile elements. Earth and Planetary Science Letters, 27, 211-218. FLOYD, P. m. & WINCHESTER,J. A. 1978. Identification and discrimination of altered and metamorphosed volcanic rocks using immobile elements. Chemical Geology, 21, 291-306. FLOYD, P. A., KEELING, G. & GOK~EN, S. 1991. Geochemistry and tectonic environment of basaltic rocks from the Misis ophiolitic melange, south Turkey. Chemical Geology, 89, 263-280. FODEN, J. D. 1983. The petrology of calcalkaline lavas of Rindjani volcano, East Sunda Arc: model for island arcs. Journal of Petrology, 24, 98-130.

348

T. RIZAO(]LU ET AL.

FUJIMAKI, H. 1986. Fractional crystallization of the basaltic suite of Usa volcano, southwest Hokkaido, Japan, and its relationships with the associated felsic suite. Lithos, 19, 129-140. GUST, D. A. & JOHNSON, R. W. 1981. Amphibole bearing cumulates from Boisa island, Papua New Guinea: evaluation of the role of fractional crystallization in an andesitic volcano. Journal of Geology, 89, 219-232. HALL, R. 1976. Ophiolite emplacement and the evolution of the Taurus suture zone, southeast Turkey. Geological Society of America Bulletin, 87, 10781088. HART, S. R., ERLANK, A. J. & KABLE, E. J. D. 1974. Sea floor basalt alteration: some chemical and Sr isotopic effects. Contributions to Mineralogy and Petrology, 44, 219-230. HI~BERT,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., Moore,s, E. M., PANAYiOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites--Oceanic Crustal Analogues. Proceedings of Troodos Ophiolite Symposium, Cyprus Geological Survey 1987, 149-163. HEMPTON, M. R. 1984. Results of detailed mapping near Lake Hazar (Eastern Taurus Mountains). In: TEKELi, O. & GONCOO~LU, M. C. (eds) Geology of the Taurus Belt. Proceedings of International Symposium, 26-29 September 1983, Ankara, Turkey. MTA, Ankara, 223-228. HEMPTON, M. R. 1985. Structure and deformation history of the Bitlis suture near Lake Hazar, SE Turkey. Geological Society of America Bulletin, 96, 223-243. HUMPRIES, S. E. & THOMPSON, G. 1978. Trace element mobility during hydrothermal alteration of oceanic basalts. Geochimica et Cosmochimica Acta, 42, 127-136. JENNER, G. A., DUNNLNG, 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. KO(TAK, K., Iw F., ARSLAN, M. & ZEDEF, V. 2005. Petrological and source characteristics of ophiolitic hornblende gabbros from the Aksaray and Kayseri regions, central Anatolian crystalline complex, Turkey. Journal of Asian Earth Sciences, 25, 883891. MESCHEDE, M. 1986. A method of discriminating between different types of mid-oceanic ridge basalts and continental tholeiites with Nb-Zr-Y diagram. Chemical Geology, 56, 207-218. MICHARO, A., JUTEAU, T. & WHITECHURCH, H. 1985. L'obduction: revue des modeles et confrontation au cas de l'Oman. Bulletin de la Socidtk Geologique de France, 2, 189-198. MTA. 2002. 11500 000 scale geological maps of Turkey. General Directorate of Mineral Research and Exploration, Ankara. MUKASA, S. B. & LUDDEN, J. N. 1987. Uranium-lead ages of plagiogranites from the Troodos ophiolite,

Cyprus, and their tectonic significance. Geology, 1, 825-828. ONAL, A. 1995. Polat-Be~,re (Do~,an~ehir-Malatya) Cevresindeki magmatik kayaflarm petrografik ve petrolojik 6zellikleri. PhD thesis, Flrat Universitesi, Elazl~, Turkey. PARLAK, O. 1996. Geochemistry and geochronology of the Mersin ophiolite within the eastern Mediterranean tectonic frame. PhD thesis, University of Geneva. PARLAK, O. 2006. Geodynamic significance of granitoid magmatism in southeast Anatolia: Geochemical and geochronological evidence from G6ksun-Af~in (Kahramanmara~, Turkey) region. International Journal of Earth Sciences (in press). PARLAK, O. & ROBERTSON,A. H. F. 2004. Tectonic setting and evolution of the ophiolite-related Mersin M61ange, southern Turkey: its role in the tectonicsedimentary setting of the Tethys in the eastern Mediterranean region. Geological Magazine, 141, 257-286. PARLAK, O. & RIZAO~LU, T. 2004. Geodynamic significance of granitoid intrusions in the southeast Anatolian Orogeny (Turkey) In: CHATZlPETROS, A. A. & PAVLIDES, S. B. (eds) Proceedings of the 5th International Eastern Mediterranean Geology Symposium, Thessaloniki, Greece, 14-20 April 2004, 157. PARLAK, O., DELALOYE, M. & BINGOL, E. 1996. Mineral chemistry of ultramafic and mafic cumulates as an indicator of the arc-related origin of the Mersin ophiolite (southern Turkey). Geologische Rundschau, 85, 647-661. PARLAK, O., HOCK, V. & DELALOYE, M. 2000. Suprasubduction zone origin of the PozantiKarsanti ophiolite (southern Turkey) deduced from whole-rock and mineral chemistry of the gabbroic cumulates. In: BOZKURT, E., WINCHESTER,J. m. & P~PER, J. D. A. (eds) Tectonics and Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 219-234. PARLAK, O., HOCK, V., KOZLU, H. & DELALOYE, M. 2004. Oceanic crust generation in an island arc tectonic setting, SE Anatolian Orogenic Belt (Turkey). Geological Magazine, 141, 583-603. PEARCE, J. A. 1982. Trace element characteristics of lavas from destructive plate boundaries. In: THORPE, R. S. (ed)Andesites: Orogenic Andesites and Related Rocks. Wiley, Chichester, 525-548. PEARCE, J. A. 1983. Role of the subcontinental lithosphere in magma genesis at active continental margins. In: HAWKESWORTH, C. J. & NORRY, M. J. (eds) Continental Basalts and Mantle Xenoliths. Shiva, Nautwich, 230-249. PEARCE, J. A. 1996. A users guide to basalt discrimination diagrams. In: WYMAN, D. A. (ed) Trace Element Geochemistry of Volcanic Rocks." Applications for Massive Sulphide Exploration. Geological Association of Canada, Short Course Notes, 12, 79-113. PEARCE, J. A. & CANN, J. R. 1973. Tectonic setting of basaltic volcanic rocks determined using trace element analysis. Earth and Planetary Science Letters, 19, 290-300.

OPHIOLITES AND GRANITES, SE TURKEY 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., LIPPARD, S. J. • ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR,B. P. t~ HOWELLS, M. F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. PLATE, D. W., PEARCE, J. A., HAWKESWORTH, C. J., COLLEY, H., EDWARDS, C. M. H. & HIROSE, K. 1997. Geochemical variations in Vanuatu island arc lavas: the role of the subducted material and a variable mantle wedge composition. Journal of Petrology, 38, 1331-1358. PERiN~EK, D. 1979. The geology of Hazro-Koruda~iingiir area. Ankara, Turkey. Tfirkiye Jeoloji Kurumu Yayml, Ankara. RIZAOGLU,T., PARLAK,O. & I~LER, F. 2004. Geochemistry and tectonic setting of the K6mtirhan ophiolite in southeast Anatolia. In: CHATZIPETROS, A, A. & PAVLIDES, S. B. (eds) Proceedings of the 5th International Eastern Mediterranean Geology Symposium, 14-20 April 2004, Thessaloniki, Greece, 285. 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, K. C. t~r CAMERLENGHI, A (eds) Proceeding of Ocean Drilling Program, Scientific Results, 160. Ocean Drilling Program, College Stating TX, 723-782. ROBERTSON, A. H. F. 2000. Mesozoic-Tertiary tectonic-sedimentary 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, Special Publications, 173, 97-138. 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. 2004. Development of concepts concerning the genesis and emplacements of Tethyan ophiolites in the eastern Mediterranean and Oman regions. Earth-Science Reviews, 66, 331-387. ROBERTSON, A. H. F., USTAOMER, T., PARLAK, O., (JNL13GENr U. C., TASLI, K. & INAN, N. 2006. The Berit transect of the Tauride thrust belt, S. Turkey: Late Cretaceous-Early Cenozoic accretionary/collisional processes related to closure of the Southern Neotethys. Journal of Asian Earth Sciences (in press). ROBERTSON, A. H. F., PARLAK, O., RIZAOt~LU, T., LINLOGEN~, U. C., iNAN, N., TASLI, K. & USTAOMER, T. 2007. Late Cretaceous-Mid Tertiary tectonic evolution of the eastern Taurus mountains and the southern Tethyan ocean; evidence from the Elazl~ region, SE Turkey. In:

349

RIES, A. C., BUTLER, R. W. H. & GRAHAM, R. H. (eds) Deformation of the Continental Crust, Geological Society, London, Special Publications (in press). ROLLINSON, H. 1993. Using GeochemicalData: Evaluation, Presentation, Interpretation. Longman, Harlow. SEARLE, M. P., LIPPARD, S. J., SMEWING,J. D. & REX, D. C. 1980. Volcanic rocks beneath the Semail ophiolite nappe in the northern Oman mountains and their significance in the Mesozoic evolution of Tethys. Journal of the Geological Society, London, 137, 589-604. SENGOR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: plate tectonic approach. Tectonophysics, 75, 181-241. STERN, R. J. 1979. On the origin of andesite in the northern Mariana island arc: implications from Agrigan. Contributions to Mineralogy and Petrology, 68, 207-219. 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, 313-347. TURAN, M. & BiNGOL, A. F. 1991. Tectonostratigraphic characteristics of the region between Kovanolar-Baskil, Elazl~, Turkey. In: YETIS, C. (ed.) Proceedings of Ahmet Acar Geology Symposium. ffukurova University, Adana, Turkey, 213-227. WALLIN, E. T. & METCALF, R. V. 1998. Suprasubduction zone ophiolites formed in an extensional forearc: Trinity terrane, Klamath mountains, California. Journal of Geology, 106, 591-608. WOOD, D. A., JORON, J. L. & TREUIL, M. 1979. A reappraisal of the use of trace elements to classify and discriminate between magma series erupted in different tectonic settings. Earth and Planetary Science Letters, 45, 326-336. YALINIZ, K. M., FLOYD, P. t~; Gt3NCOO(3LU, M. C. 1996. Supra-subduction zone ophiolites of Central Anatolia: geochemical evidence from the Sarikaraman ophiolite, Aksaray, Turkey. Mineralogical Magazine, 60, 697-710. YALINIZ, K. M., FLOYD, P. & Gt3NC(JOOLU, M. C. 2000. Geochemistry of volcanic rocks from the (~i~kda~ ophiolite, Central Anatolia, Turkey, and their inferred tectonic setting within the northern branch of the Neotethyan ocean. 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, 203-218. YAZGAN, E. 1983. A geotraverse between the Arabian platform and the Munzur mountains. Field Guide Book, Excursion. 5th International Symposium on Geology of the Taurus Belt, Ankara. MTA Ankara. YAZGAN, E. 1984. Geodynamic evolution of the eastern Taurus. In: TEKELI, O. & GONC~OGLU, M. C. (eds) The Geology of the Taurus Belt. Proceedings of the International Symposium, Ankara, 199-208. MTA Ankara.

T. RIZAOGLU ET AL.

350

YAZGAN, E. & CHESSEX,R. 1991. Geology and tectonic evolution of the southeastern Taurides in the region of Malatya. Bulletin of Turkish Association of Petroleum Geologists, 3, 142. Yi~iTBAS, E. & YILMAZ, Y. 1996a. New evidence and solution to the Maden complex controversy of the southeast Anatolian orogenic belt (Turkey). Geologische Rundschau, 85, 250-263. Yi~ITBAS, E. & YILMAZ, Y. 1996b. Post-late Cretaceous strike-slip tectonics and its implications on the Southeast Anatolian orogen, Turkey. International Geology Review, 38(9), 818-831. YILDIRIM, M. & YILMAZ, Y. 1991. Gfineydo~u Anadolu orojenik ku~a~mm ekayh zonu. Bulletin

of Turkish Association of Petroleum Geologists, 3, 57-73. YILMAZ, O.

1 9 7 1 . Etude pktrographique et gkochronologique de la rkgion de Cacas. PhD thesis,

Universit6 Science et M6dicale de Grenoble. YILMAZ, Y. 1978. Bitlis massif and ophiolite relationship around Geva~, Van. Proceedings of 4th

Petroleum Congress of Turkey, Ankara, 88-93. TPJD, Ankara. YmMAZ, Y. 1993. New evidence and model on the evolution of the southeast Anatolian orogen. Geological Society of America Bulletin, 105, 251-271. YILMAZ, Y., GORPINAR, O., KOZLU, H., et al. 1987. Kahramanmara~ Kuzeyinin Jeolojisi ( AndtrmBerit-Engizek-Nurhak-Binbo~a Da~larl). Tfirkiye Petrolleri A.O. Rapor, 2028. YILMAZ, Y., Yi~iTBAS, E. & GENt, $. C. 1993. Ophiolitic and metamorphic assemblages of southeast Anatolia and their significance in the geological evolution of the orogenic belt, Tectonics, 12, 1280-1297. YOGODZINSKI,G. M., VOLYNETS,O. N., KOLOSKOV,A. V., SEL~VERSTOV,N. I. & MATVENKOV,V. V. 1993. Magnesian andesites and the subduction component in strongly calc-alkaline series at Piip volcano, far western Aleutians. Journal of Petrology, 35, 163-204.

Palaeomagnetic insights into the evolution of Neotethyan oceanic crust in the eastern Mediterranean ANTONY

M O R R I S l, M A R K W. A N D E R S O N ALASTAIR

1, J E N N I F E R

I N W O O D 1&

H . F. R O B E R T S O N 2

1School of Earth, Ocean and Environmental Sciences, University o f Plymouth, Drake Circus, Plymouth PL4 8AA, UK (e-mail." antony, morris@plymouth, ac. uk) 2School o f Geosciences, University o f Edinburgh, West Mains Road, Edinburgh EH9 3JW, UK Abstract: A synopsis of palaeomagnetic data from three Late Cretaceous eastern Mediterranean Tethyan ophiolites (Troodos, Hatay and Ba6r-Bassit) and their sedimentary cover sequences is presented. These data provide valuable insights into the role of regional- and local-scale tectonic rotations in the geodynamic evolution of Neotethyan oceanic crust. The geologically earliest phases of tectonic rotation are documented in the Troodos ophiolite, where rotations around both subvertical and subhorizontal axes are readily related to the development of the spreading fabrics and structures during crustal genesis. Subsequent c. 74~ anticlockwise intra-oceanic rotation of a 'Troodos microplate' has been quantified through analysis of the in situ sedimentary cover of the Troodos ophiolite. Results indicate that bulk anticlockwise rotation began soon after the cessation of spreading and ended by the end of the Eocene, with c. 50-60 ~ of microplate rotation being over by the Maastrichtian, the time at which ophiolite thrust sheets were emplaced onto the Arabian continental margin to the east of Troodos. Recent results from the emplaced, structurally dismembered Ba6r-Bassit ophiolite indicate extreme anticlockwise rotations of ophiolitic thrust sheets varying on a kilometre scale. New data from the post-emplacement sedimentary cover confirm that only a small component of these rotations is due to post-emplacement tectonism. Ba6r-Bassit represents the leading edge of the emplaced ophiolitic sheet. New data from the more coherent section preserved in the Hatay ophiolite to the north demonstrate significant anticlockwise rotation. This is equivalent to the rotation of the most northerly part of the Ba6r-Bassit units to the south, and is of the same sense and magnitude as the preMaastrichtian phase of microplate rotation documented in the Troodos. This suggests a common, intra-oceanic origin for the majority of the Troodos and Hatay rotations, and a significant component of the more variable rotations observed in Ba~Bassit. Overall, therefore, the data support a model involving: (1) intra-oceanic rotation of a coherent region of crust within the southern Neotethyan basin; this rotated unit is more areally extensive than has previously been inferred from consideration of data from the Troodos ophiolite alone; (2) emplacement of part of the rotated unit onto the Arabian platform; (3) subsequent localized post-emplacement modification, related to the development of the current plate configuration.

Ophiolitic terranes provide unique opportunities for investigating the structure and kinematics of constructive plate margins, transform faults and oceanic suture zones. The eastern Mediterranean-Middle East segment of the Tethyan orogenic belt is marked by several chains of ophiolites, the most prominent of which are the Troodos (Cyprus), Ba~r-Bassit (Syria), H a t a y (Turkey), Kermanshah and Neyriz (Iran), and Semail (Oman) ophiolites (Fig. 1). These are interpreted as fragments of ocean lithosphere formed in a southern Neotethyan basin during the Late Cretaceous (e.g. Robertson 1998). The aim of this paper is to present a synopsis of the major palaeomagnetic results obtained

from the Troodos, H a t a y and Ba~r-Bassit ophiolites (Fig. 2) over the last 20 years. The Troodos ophiolite of Cyprus, in particular, has played a key role in developing and testing concepts of plate tectonics (Gass 1968, 1980; Moores & Vine 1971; Robertson & Xenophontos 1993; Robertson et al. 1996), because primary sea-floor geometries are preserved owing to an absence of large-scale thrust faulting. Fundamental insights have been provided through a series of palaeomagnetic and structural studies (e.g. Clube 1985; Clube & Robertson 1986; Allerton & Vine 1987, 1990, 1991; Allerton 1988; Bonhommet et al. 1988; MacLeod et al. 1990; Morris et al. 1990, 1998; Hurst et al. 1992) that have highlighted

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 351-372. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

352

A. MORRIS E T AL.

Fig. 1. Major ophiolites of the easternmost Mediterranean and Middle East region. The Troodos ophiolite remains in a pre-emplacement setting, in contrast to ophiolites to the east that were tectonically emplaced onto the Arabian continental margin during the Late Cretaceous.

Fig. 2. Outline tectonic map of the eastern Mediterranean, showing the distribution of Late Cretaceous ophiolites (black) and major structural features.

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS the role of tectonic rotations in oceanic crustal construction and transform tectonism. The importance of such rotations is now firmly established, following their identification in modem spreading systems (e.g. Hurst et al. 1994). More recently, Morris et al. (2002), Inwood et al. (2003) and Inwood (2005) have provided the first palaeomagnetic data from the Ba~rBassit and Hatay ophiolites, which, in contrast to the Troodos ophiolite, were emplaced tectonically onto the Arabian continental margin as a series of thrust sheets. These investigations have discovered ubiquitous large tectonic rotations, only part of which may be attributed to post-emplacement deformation. We show that the combined palaeomagnetic database from all three ophiolites is now sufficiently extensive to provide insights into the range of styles of tectonic rotations that have affected the Neotethyan oceanic crust during formation and later deformation, and to discuss the regional-scale significance of these rotations.

Geological summary of the ophiolites The Troodos ophiolite

The Troodos Complex ophiolite formed during the Late Cretaceous (Cenomanian-Turonian, U-Pb age 90-92 Ma; Mukasa & Ludden 1987) in a suprasubduction-zone (SSZ) setting (Pearce 1975, 1980). It consists of a complete Penrose pseudostratigraphy disposed in a domal structure as a result of Late Pliocene-Recent uplift, giving rise to a broadly concentric outcrop pattern. Mantle and lower crustal (gabbroic) sequences are exposed around the central structural and topographic high. The plutonic section includes layered cumulates cut by gabbroic intrusions, providing clear evidence for the presence of small, multiple magma chambers beneath the Troodos spreading axis (Robinson & Malpas 1990). The contact between the isotropic gabbros and overlying sheeted dykes is commonly a low-angle extensional detachment fault zone (Varga & Moores 1985), providing evidence for amagmatic stretching during crustal formation. The sheeted dyke complex is exposed over an 80 km wide swath and consists of generally north-south-striking dykes (present coordinates) that at some localities (e.g. Lemithou) are rotated to low angles and occasionally cut by later dykes (Dietrich & Spencer 1993). Spreading took place either by steady-state processes (Allerton & Vine 1987), or by formation of discrete, ephemeral, sea-floor grabens (Varga & Moores 1985). The best documented and most distinctive graben runs through the Solea area to the north of

353

the plutonic complex, above the 'Kakopetria Detachment' fault (Verosub & Moores 1981). This graben is interpreted as a fossil spreading axis (the 'Solea axis'; see Fig. 4). Magnetic fabric and dyke surface lineation data (Staudigel et al. 1992; Varga et al. 1998) reveal both upward and lateral magma emplacement through the dyke complex, supporting a model of magma transport along the length of the spreading axis away from isolated magmatic centres beneath axial volcanoes spaced along the ridge crest. The overlying extrusive sequence is best exposed along the northern and SW margins of the ophiolite, providing classic sections where the interplay between magmatic, tectonic and hydrothermal processes during construction of the oceanic crust may be established (Schmincke et al. 1983; Schmincke & Rautenschlein 1987). The southern margin of the main outcrop of the Troodos ophiolite is marked by the eastwest-trending Arakapas Fault Belt, a strike-slip fault system that is interpreted as a fossil oceanic transform fault (Moores & Vine 1971; Simonian & Gass 1978). An anomalous ophiolitic sequence is exposed to the south of this structure within the Limassol Forest Complex. Here, mantle sequence and lower crustal rocks are exposed at high structural levels and are cut by numerous east-west-trending shear zones (Murton 1986; MacLeod 1990). The majority of the Limassol Forest Complex is interpreted to have formed within a leaky (transtensional) 'South Troodos Transform Fault Zone' (STTFZ) whose principal displacement zone is represented by the Arakapas Fault Belt (MacLeod & Murton 1993). A progressive change in orientation of dykes within the Sheeted Dyke Complex is observed as the STTFZ is approached from the north, suggestive of either primary variation in the orientation of the Late Cretaceous stress field adjacent to a sinistrally slipping transform or post-emplacement clockwise tectonic rotations of dykes resulting from dextral shear along the transform. Palaeomagnetic data (described herein) strongly support this latter model. A small area of normal Troodos-type crust exposed in the SE corner of the Limassol Forest Complex is believed to represent a fragment of crust formed at an 'AntiTroodos' spreading axis located to the south of the transform domain (MacLeod 1988, 1990). Together these various structures and ophiolitic regions provide a unique exposure of the Late Cretaceous Neotethyan spreading system that is well preserved because of a lack of large-scale thrust tectonics during emplacement of the Troodos Complex. In the Late Cretaceous, shortly after genesis at the Neotethyan spreading axis, the Troodos

354

A. MORRISE T AL.

oceanic crust became tectonically juxtaposed along its SW margin (present coordinates) with an allochthonous, highly deformed sequence of Upper Triassic to middle Cretaceous deep-sea sedimentary and Upper Triassic volcanic rocks known as the Mamonia Complex. This is interpreted to represent remnants of a passive continental margin and marginal oceanic crust (Robertson & Woodcock 1979; Swarbrick 1979, 1980, 1993). The mode of juxtaposition of the Troodos and Mamonia Complex terranes is under debate, with models invoking either strike-slip-dominated (e.g. Clube & Robertson 1986; Swarbrick 1993) or thrust-dominated (e.g. Malpas et al. 1993) emplacement, or a complex multiphase history of deformation (involving successive strike-slip, transtensional and contractional events; Bailey et al. 2000). Within the suture zone, high crustal level rocks (extrusive rocks and cross-cutting dykes) are exposed in fault-bounded slivers, with faulted contacts marked by discontinuous, steeply dipping strands of serpentinite. These slivers are overlain by an in situ sedimentary cover of Campanian umbers and radiolarites (Perapedhi Formation), and by thick, largely undeformed, successions of bentonitic clays and volcaniclastic sandstones, indistinguishable from the sedimentary cover found overlying the ophiolitic basement of the main Troodos Complex (Robertson 1977; Clube & Robertson 1986). The highly depleted geochemistry of the extrusive rocks (Murton 1990) is distinctly different from the alkaline within-plate basalt (WPB) to tholeiitic mid-ocean ridge basalt (MORB) compositions of the Triassic extrusive rocks of the Mamonia Complex (Malpas et al. 1993). Stratigraphic and petrographic data, therefore, support correlation of these ophiolitic outcrops with the Troodos Complex and its transform fault-related southern margin. This interpretation is also supported by palaeomagnetic data obtained from these slivers by Morris et al. (1998), as discussed below. The H a t a y ( K t z t l D a ~ ) a n d B a ~ r - B a s s i t ophiolites

In contrast to the Troodos Complex, coeval ophiolites to the east have been structurally modified during Late Cretaceous thrustdominated emplacement onto the Arabian continental margin during progressive closure of the southern Neotethyan basin. The most westerly of these emplaced units are the Ba6r-Bassit and Hatay ophiolites. These are closely related spatially (Fig. 2) and represent parts of a single unit emplaced during the Maastrichtian. The Hatay ophiolite to the north forms a relatively

intact sheet (Delaloye et al. 1980), whereas the Bafir-Bassit ophiolite to the south (Kazmin & Kulakov 1968; Parrot 1980) represents the leading edge of the emplaced sheet and is highly deformed by thrust faulting (A1-Riyami et al. 2000). The radiometric age of these ophiolites is poorly constrained, with ophiolitic dykes yielding K-Ar ages in the range 73-99 Ma (Delaloye & Wagner 1984). The Hatay ophiolite is split into a large southwestern massif and a smaller northeastern massif by a high-angle fault (Tahtak6pru Fault). The ophiolite is separated from the underlying Arabian platform by only a thin m61ange and no mdtamorphic sole is preserved (Robertson 2002). The succession in the main Kml Da~ massif (Delaloye & Wagner 1984) begins with serpentinized harzburgite tectonite with local intercalations of dunite, wehrlites, lherzolite and feldspathic peridotites. The ultramafic rocks are separated from the overlying gabbros by a 50100 m thick shear zone that extends upwards into the base of the layered gabbros (Dilek & Thy 1998). The layered gabbros, in turn, pass into isotropic gabbros, intruded by small bodies of plagiogranites, leucocratic gabbro and dolerite. Diabase dykes increase in abundance towards the top of the gabbros, which pass upwards into a sheeted dyke complex. Locally, the gabbrodyke contact is a low-angle shear zone marked by hydrothermal alteration. In the NE, sheeted dykes are unconformably overlain by Maastrichtian non-marine to shallow-marine sediments, presumably after erosion of ophiolitic extrusive rocks (Erendil 1984; Pipkin et al. 1986). The succession in the northeasterly massif (NE of the Tahtak6pru Fault) is exposed at several localities. At one, serpentinized peridotires are tectonically overlain by gabbro, rotated dykes and lavas (Dilek & Thy 1998). Elsewhere, serpentinized peridotites are overlain, above a gently dipping normal fault, by massive and pillow lavas that are rarely interbedded with or overlain by metalliferous sediments (Erendil 1984; Robertson 1986). Pillow flows are steeply dipping to subvertical. Further south (south of Antakya), gabbros or serpentinites are in low-angle faulted contact with overlying pillow lavas. These extrusive rocks include highly magnesian, boninite-type lavas ('sakalavite'; Delaloye & Wagner 1984). The highly dismembered Ba6r-Bassit ophiolite is underlain by a well-developed inverted amphibolite or greenschist-facies metamorphic sole (Whitechurch 1977; AI-Riyami et al. 2002), that has a K-Ar age of 86-93 Ma (Thuizat et al. 1981; Delaloye & Wagner 1984). The ophiolitic outcrop is dominated by two massifs,

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS BaEr in the NE (inland) and Bassit in the N W (near the coast), together with smaller masses of highly dismembered ophiolitic rocks further SE. The Ba~r massif is subdivided into several large thrust sheets, dominated by harzburgites overlain by cumulate ultramafic rocks (A1Riyami et al. 2000). Layered gabbros ( < 1 km thick) are locally cut by dolerite dykes. The Bassit massif comprises a lower sequence of harzburgites and gabbros, which are overthrust by a slice of m61ange and then by thin ( < 100 m thick) imbricate thrust sheets of gabbro, sheeted dykes and pillow lavas. Geochemical analysis of extrusive rocks reveals strongly depleted, highly magnesian boninite types (A1-Riyami et al. 2000). The ophiolite and its metamorphic sole are underlain by the extensive 'Ba~r-Bassit M61ange' (A1-Riyami et al. 2000; AI-Riyami & Robertson 2002), which consists of a Late Triassic to midCretaceous deep-water, passive margin succession (Delaune-May~re 1984; A1-Riyami et al. 2000). All three units (Ba~r-Bassit ophiolite, metamorphic sole and BaEr-Bassit M61ange) were thrust onto the Arabian carbonate platform in the mid-Maastrichtian (c. 70 Ma), based on the biostratigraphic ages of the youngest sequences of the Arabian platform beneath the allochthon and the oldest sedimentary rocks of the post-emplacement cover sequences (Kazmin & Kulakov 1968). Lineations in the metamorphic sole, defined by elongation of amphibole porphyroblasts, together with fold facing and vergence directions within the underlying Ba6r-Bassit M61ange, indicate that thrust sheets were emplaced towards the SE (A1-Riyami et al. 2002; A1-Riyami & Robertson 2002). The disrupted BaEr-Bassit allochthon was later unconformably overlain by a sedimentary sequence of late Maastrichtian to Pliocene age (Boulton et al. 2006). The sedimentary successions are cut by mainly ENE-WSW-trending strike-slip faults that extend offshore (Figs 2 and 7). This fault system represents part of the extension of the plate boundary zone between the African plate and the Turkish microplate, which runs eastwards from south of Cyprus as a zone of distributed deformation and then comes onshore, passing through the BaEr-Bassit region to link with the Dead Sea transform fault system to the east (A1-Riyami et al. 2002; Fig. 2).

The palaeomagnetic database Sources, data selection and reporting

Discussion is restricted to a synopsis of the major palaeomagnetic results obtained to date, and

355

no attempt is made to summarize the growing literature on magnetic fabric (anisotropy of magnetic susceptibility) results from the Troodos ophiolite (e.g. Staudigel et al. 1992; Varga et al. 1998; Abelson et al. 2001; Borradaile & Lagroix 2001). The majority of ophiolite data used in this synopsis were collated from published sources by Morris (2003) for the purpose of assessing the palaeolatitude of the Neotethyan spreading axis. This compilation consisted of results from: (1) 100 palaeomagnetic sampling sites in the Troodos ophiolite, drawn from Clube (1985), Clube et al. 0985) Bonhommet et al. (1988), Allerton (1989a), Morris et al. (1990, 1998) and Hurst et al. (1992); (2) 19 sites in the Ba~r-Bassit ophiolite presented by Morris et al. (2002). Additional data from six sites in the Mandria area of the Troodos ophiolite from MacLeod et al. (1990) are included here. New palaeomagnetic data from 18 sites of the Hatay ophiolite are included from Inwood et al. (2003) and Inwood (2005). Finally, to assess the timing of postcrustal genesis rotations in the ophiolites, data are also drawn from: (1) 26 sites distributed through the in situ Upper Cretaceous to Miocene sedimentary cover sequences of the Troodos ophiolite, reported by Clube (1985), Abrahamsen & Sch6nharting (1987) and Morris et al. (1990); (2) 15 sites within the Tertiary postemplacement sedimentary cover of the Ba~rBassit ophiolite, reported by Morris et al. (2006); (3) 17 sites within the post-emplacement sedimentary sequences overlying the Hatay ophiolite, reported by Inwood et al. (2003), Kissel et al. (2003) and Inwood (2005). The following criteria were used in the selection of data from the source publications: (1) site-level data are based on laboratory cleaned sample remanences with the stability of remanences investigated at each site by using either demagnetization of pilot samples or, preferably, full demagnetization of each sample; (2) structural corrections are reported in the source paper or may be recovered from reported in situ and tilt-corrected remanence data; (3) data fulfil the following statistical constraints: number of samples included in the site mean, n/> 5; cone of confidence, ~95< 20~ and Fisherian precision parameter, k > 10. The reader is referred to the source papers for details of sampling procedures and site locations. Stereographic projections of site-mean remanence data and the associated cones of confidence are included therein. Primary tables of data may be found in the source papers, or have been given by Morris (2003) in the case of data from the Troodos and BaEr-Bassit ophiolitic sites.

356

A. MORRIS ET AL.

Data correction and interpretation techniques Standard palaeomagnetic practice involves structurally correcting in situ (geographical coordinates) remanence data by applying a simple tilt around a strike-parallel axis to restore a palaeohorizontal or palaeovertical surface to the present-day horizontal or vertical. Tilt-corrected vectors may then be compared with an appropriate coeval reference vector, with differences in declination (azimuth) being interpreted in terms of vertical axis rotations, and differences in inclination (dip) being attributed to either palaeolatitudinal movements or to inclination shallowing as a result of compaction (in the case of sedimentary rocks). This tilt correction approach decomposes the total deformation at a site into components of rotation around horizontal and vertical axes. Declination errors may be introduced artificially if deformation involved tilting around inclined axes, if more than one phase of tilting has occurred, or if fold axes are plunging (MacDonald 1980). In the last case, however, declination errors are < 10 ~ for fold plunges of up to 50 ~ if the palaeohorizontal dips at <40 ~ and are < 10 ~ even for vertical beds if the fold plunge is < 10 ~ The most severe source of error in adopting standard tilt corrections in ophiolitic terranes, however, arises from restoration of sheeted dykes to the vertical, as components of tilt around dyke-normal axes are impossible to resolve in the absence of palaeohorizontal markers (Borradaile 2001; Morris & Anderson 2002). Such unresolved tilts may potentially introduce both declination and inclination anomalies. An alternative net tectonic rotation approach that has been widely adopted in these ophiolites (Allerton & Vine 1987; Allerton 1989a; Morris et al. 1990, 1998, 2002; Hurst et al. 1992) is to describe the deformation at a site in terms of a single rotation about an inclined axis, which restores both the palaeohorizontal and palaeovertical to their initial orientation and the site mean magnetization vector to the appropriate palaeomagnetic reference direction. This single rotation may then be decomposed into any number of component rotations on the basis of additional structural data, and/or net tectonic rotation axes may be interpreted directly in terms of structural history. The key assumptions in this method (Allerton & Vine 1987) are that: (1) pre-deformational remanences are preserved; (2) an appropriate (coeval) reference magnetization direction may be found; (3) dyke margins or palaeohorizontal surfaces should be restored to as close to vertical or horizontal as possible;

(4) no internal deformation of sampled units has occurred. Useful insights into the tectonic evolution of the ophiolites under consideration here have been obtained using either one or both of these interpretation methods. For the purposes of this synopsis, however, data are displayed following simple tilt correction only, for the following reasons: (1) this unifies the palaeomagnetic database; (2) primary, detailed interpretations are available in the source papers, where issues related to choice of methods are discussed in detail; (3) the tectonic significance of tilt corrected data is easier to understand intuitively; (4) application of the net tectonic rotation approach requires use of standard or inclination-only tilt tests based on the simple tilt correction approach in order to properly assess the timing of magnetization acquisition relative to deformation, unless magnetizations are merely assumed to be pre-deformational in age; (5) the tectonic significance of the data in terms of rotation patterns and styles is largely unaffected by the choice of technique. Reference is made to the results of net tectonic rotation analyses where appropriate.

Age o f magnetization Detailed discussion of the palaeomagnetic characteristics of the ophiolites is beyond the scope of this paper. In summary, demagnetization of natural remanent magnetizations generally reveals characteristic magnetizations carried by a range of ferrimagnetic phases that can be readily related to assemblages observed within in situ oceanic crust and other ophiolites (Dunlop & Ozdemir 1997). Of prime concern, however, is consideration of the age of magnetization within the sampled ophiolitic lithologies, as this is critical to the tectonic interpretation of the palaeomagnetic data. This is determined with respect to tectonic disruption of sampled units using field tests of palaeomagnetic stability, the most common of which is the palaeomagnetic fold or tilt test (McElhinny 1964; McFadden & Jones 1981). Differential vertical axis rotations, however, invalidate use of area-wide tilt tests based on full remanence vectors (declination and inclination). An alternative approach that is not affected by such rotations is to determine the effect of untilting on the distribution of inclinations only. The angle between the inclination and the palaeohorizontal at a site may be assumed to remain constant during rigid body rotation, regardless of the axis of rotation. Significant improvement in clustering of inclinations upon tilt correction of mean directions from sites

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS

357

Fig. 3. Variation in the Fisher precision parameter with progressive untilting of palaeomagnetic data from sites within the Troodos, Hatay and Ba~r-Bassit ophiolites. These data indicate positive inclination-only tilt tests (Enkin & Watson 1996) in all three cases, with peaked distributions centred on 100% untilting. This demonstrates unequivocally that pre-deformational magnetizations are recorded by the sampled units, and that palaeomagnetic data should therefore be interpreted in tilt-corrected coordinates.

with different structural orientations therefore suggests that a pre-tilt magnetization has been identified (Enkin & Watson 1996). Figure 3 shows the variation of the maximum likelihood estimates of the Fisher precision parameter, k, with degree of untilting for each of the three ophiolites. Strongly peaked distributions with maximum k values at 100% untilting demonstrate unequivocally that pre-deformational remanences are identified within each terrane. The extensive database available for the Troodos ophiolite allows separate consideration of the age of magnetization of the extrusive sequences and sheeted dyke complex. Positive tilt tests are observed in both cases. Close agreement between the tilt-corrected mean inclinations of both units suggest that the sheeted dyke dataset is sufficiently large to ensure that any components of tilting around dyke-normal axes (Borradaile 2001; Morris & Anderson 2002) at individual sites produces little bias in the overall mean inclination for this unit.

Tectonic significance of the data On the basis of the tilt test results described above, palaeomagnetic data from the ophiolites are hereafter described and interpreted in tiltcorrected (stratigraphic) coordinates. Stereographic projections of site-mean remanence data are given in Figure 4 (Troodos) and Figure 7 (Hatay and Bafir-Bassit). Data from the sedimentary

cover sequences are presented in Figure 6 (Troodos) and Figure 7 (Hatay and Ba6r-Bassit). Results are discussed by individual ophiolite below, prior to regional synthesis. Troodos ophiolite

There is a clear distinction within the Troodos palaeomagnetic database (Fig. 4) between regions with magnetization vectors that cluster around westerly declinations and regions where declinations are widely variable from WSW, through northerly to easterly declinations. The former regions (stereonets 1 4 with unshaded ~95 ellipses in Fig. 4) are the northern margin of the main ophiolite (Clube 1985), the Solea region (Allerton & Vine 1987; Allerton 1989a; Hurst et al. 1992), the Akamas peninsula (Clube 1985; Morris et al. 1998) and the western margin of the ophiolite (Morris et al. 1998). These regions are remote from the STTFZ and its inferred westerly along-strike extension, and provide the evidence for the ophiolite-wide bulk anticlockwise rotation of the Troodos microplate first identified from natural remanent magnetization data by Vine & Moores (1969) and Moores & Vine (1971). The mean direction of magnetization of these 29 sites is declination (D) = 272.5 ~ inclination (I) = 38.4 ~ ~95 = 6.5 ~ k = 17.7. This westerly-directed magnetization is also observed at sites (Morris et al. 1990) within the SE part of the Eastern Limassol Forest Complex to the immediate south of the STTFZ (stereonet 5, Fig. 4), providing palaeomagnetic support for the presence of a fragment

358

A. MORRIS E T AL.

Fig. 4. Outline geological map of Cyprus, showing the location of major structural features and the locations of areas that have been investigated palaeomagnetically. Lower hemisphere stereographic projections show tilt-corrected site-levelpalaeomagnetic directions and associated ct95cones of confidence from each area. Shaded ~s ellipses indicate data obtained from ophiolitic crust that is inferred to have experienced significant transform tectonism, whereas unshaded ~s ellipses indicate data obtained from localities outside the transform tectonized zone. Black star in stereographic projection 1 indicates the direction of the present-day geomagnetic field. STTFZ, South Troodos Transform Fault Zone; AFB, Arakapas Fault Belt. Data sources: Clube (1985); Bonhommet et al. (1988); Allerton (1989a); Morris et al. (1990, 1998); Hurst et al. (1992).

of an Anti-Troodos plate in this area, as originally deduced by MacLeod (1988, 1990) from field geological observations. Palaeomagnetic data from the Solea area provide important insights into tectonic processes at slow- to intermediate-spreading axes. The Solea graben is a 15-20 km wide asymmetrical structure defined principally by variations in dyke attitude in the Sheeted Dyke Complex (Varga & Moores 1985), which in this area is separated from the underlying plutonic complex by the low-angle Kakopetria detachment fault zone. An oceanic environment for formation of the graben is demonstrated by the horizontal disposition of the overlying sedimentary sequences (MacLeod et al. 1990). To the east of the inferred spreading axis, dykes are generally steeply dipping to subvertical and trend NNW except for several

small areas where dykes dip more gently (Hurst et al. 1992). The wider western flank is distinctly different, with north-south-trending dykes dipping at low angles of 25-45 ~ to the east. The lowangle orientations of these dykes were attributed by Verosub & Moores (1981) to listric normal faulting associated with the Kakopetria detachment, above a magma chamber at the active spreading axis. In this model, plate separation was at least partly accommodated by tectonic thinning of the upper crust during periods of amagmatic stretching. Palaeomagnetic analysis of these dykes (Allerton & Vine 1987; Allerton 1989a; Hurst et aI. 1992) confirms that they were originally intruded in (sub)vertical orientations. Tilt-corrected vectors (stereographic projection 3, Fig. 4) generally cluster around the mean inclination observed at other localities far

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS from the STTFZ, in contrast to in situ inclinations (not shown) that vary widely from -25 ~ to 75 ~. The data are most informatively analysed using the net tectonic rotation approach (Allerton & Vine 1987). This reveals rotation axes for dykes in the western portion that are subhorizontal and subparallel to the original dyke strikes (Allerton & Vine 1987; Allerton 1989a; Hurst et al. 1992), consistent with rotation (tilting) of dykes by faults above the detachment surface. For the eastern Solea graben, marked by steeper and more variable dyke orientations, structural evidence does not support the existence of a throughgoing detachment fault. Net tectonic rotation analysis here indicates that some dykes appear to have been tilted towards the Solea axis, but the majority show minor to significant rotation about vertical axes. This was attributed by Hurst et al. (1992) to local variations in the amount of extension, related to late-stage episodic intrusions. Extensive palaeomagnetic data (Clube 1985; Bonhommet et al. 1988; Allerton 1989a; MacLeod et al. 1990; Morris et al. 1990) from the southern half of the main Troodos ophiolite and the Limassol Forest Complex have been used to address the debate on the sense of displacement along the STTFZ and its relationship with the Solea axis. A progressive change in dyke trend is observed within the Sheeted Dyke Complex as the transform zone is approached from the north, from a predominant north-south orientation into eventual alignment with the transform lineament (Fig. 5a). This has been interpreted as the result of either dyke injection into a sigmoidal stress field, implying that dykes are in their original orientations relative to the sinistrally slipping transform (Fig. 5b), or clockwise vertical-axis fault block rotations in response to dextral slip (Simonian & Gass 1978; Fig. 5c). Discrimination between these alternative models can be achieved palaeomagnetically, as systematic tectonic rotations would result in systematic variations in magnetization vectors away from the westerly directed vectors observed outside the transform-influenced zone. Data from the region to the north and NE of the Arakapas Fault Belt (Bonhommet et al. 1988; Allerton 1989a; stereonet 6, Fig. 4) show a broad spread of directions that are clearly rotated in a clockwise sense relative to this westerly vector, supporting a dextral shear sense along the transform between sinistrally offset spreading axes. The most unequivocal analysis was presented by Bonhommet et al. (1988), who showed that magnetizations in sampled dykes cluster tightly with

359

westerly declinations after correction of dykes back to the predominant north-south trend, thereby ruling out initial N E - S W trends. Further palaeomagnetic support for dextral shear along the transform zone is provided by data from the SSTFZ itself (Morris et al. 1990; stereonet 7, Fig. 4), the majority of which are also rotated clockwise away from the general Troodos vector. Net tectonic rotation analysis of data from sets of cross-cutting dykes (Allerton & Vine 1990; Morris et al. 1990) demonstrates that clockwise block rotations were actively occurring during crustal genesis, rather than resulting from later reactivation of the fault zone. Analysis of upper crustal rocks of Troodostype exposed in fault-bounded slivers along the suture zone with the Mamonia Complex in SW Cyprus (stereonet 10, Fig 4) reveals significant rotations in a generally clockwise sense away from the Troodos vector (Morris et al. 1998). In particular, differences in remanence directions between cross-cutting units are observed at several localities, interpreted elsewhere as a characteristic of synmagmatic rotation during transform tectonism. Net tectonic rotation analyses allow decomposition of the total rotation at these sites into early and late components. Early rotations are consistently clockwise, in agreement with studies of rotations associated with the STTFZ further to the east. These data, therefore, suggest that transform-tectonized crust is preserved in SW Cyprus, an interpretation supported by stratigraphic and geochemical similarities between these units and the main Troodos ophiolite and its transform fault-related southern margin. The overwhelming palaeomagnetic evidence for clockwise fault block rotations associated with dextral slip along the Southern Troodos transform contrasts with sinistral kinematic data reported by Murton (1986) within the western Limassol Forest Complex (i.e. within the transform zone). This apparent paradox was resolved by MacLeod & Murton (1995), who proposed a model in which sinistral shear developed locally at block boundaries within an overall dextral shear zone. Detailed palaeomagnetic and structural analyses (MacLeod et al. 1990) have identified a limit to the zone of transform-related rotations along the STTFZ (within area 8, Fig. 4; Fig. 5d). The changeover from rotated to unrotated dykes occurs across a complex zone about 2-5 km wide to the west of the village of Mandria. This zone was interpreted by MacLeod et al. (1990) as a fossilized intersection between

360

A. MORRIS E T A L .

Fig. 5. (a) Simplified geological map of the southern margin of the main Troodos ophiolite and the Limassol Forest Complex (after Simonian & Gass 1978), showing the progressive change in dyke trend into near parallelism to the Southern Troodos transform zone over a distance of 10-15 km (from Morris et al. 1990). (b) and (c) show possible alternative settings in which deviations in dyke trend could occur close to the Southern Troodos Transform Fault Zone (from Morris et al. 1990): (b) dyke injection into a sigmoidal stress field operating across a sinistrally slipping transform between dextrally offset spreading axes; (c) rotation of fault blocks related to dextral slip along the active transform domain between sinistrally offset spreading axes. (d) Geological map of the Mandria area (modified from MacLeod et al. 1990), with palaeomagnetic sites shown (in situ declinations indicated by arrows). Dykes in the west have an average north-south trend in contrast to highly rotated dykes to the east. The boundary between these domains represents a fossil ridge-transform intersection. the Solea axis and the SSTFZ. They noted that the radius of curvature of dyke swing remains approximately constant across the entire exposed width of the Troodos massif to the east of the ridge-transform intersection (see Fig. 5a). MacLeod e t al. (1990) concluded that transform-related rotations occurred within the active inside corner of the intersection itself rather than being accommodated progressively with increasing strike-slip displacement along the

transform. This supports a theoretical model (Allerton 1989b) for distortions within weak crust at ridge-transform intersections. Distributed rotational deformation is, therefore, considered to be largely confined to the inside corner of the intersection itself (MacLeod e t al. 1990), and subsequent strain is taken up by strike-slip faulting concentrated almost exclusively in the principal transform displacement zone within the transform valley (Arakapas Fault Belt).

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS The debate over the tectonic interpretation of the STTFZ has recently been reopened by Cann et al. (2001). They noted the presence in the Limassol Forest area of outcrops of deeper mantle or crustal lithologies in low-angle extensional tectonic contact with a range of shallower crustal lithologies, as mapped by B. J. Mutton and C. J. MacLeod (in Gasset al. 1991, 1994). In the eastern Limassol Forest Complex these extensional structures were attributed by MacLeod (1988, 1990) to a sustained period of post-volcanic extensional reactivation of the STTFZ in the Late Cretaceous, related to the initiation of palaeorotation of the 'Troodos microplate'. By contrast, Cann et al. (2001) compared these outcrop patterns and structures with the characteristics of extensional oceanic core complexes developed at inside corners of ridge-transform intersections in modern slowspreading systems (e.g. those associated with the Atlantis Transform Fault, Mid-Atlantic Ridge at 30~ Blackman et al. 1998). In this alternative interpretation (Fig. 6a), the Limassol Forest Complex represents a remnant of a strip of inside-corner crust formed to the present-day south of a sinistrally slipping active transform, with the zone of curved dyke trajectories (described above) lying in crust formed at the outside corner. This model, however, provides no geologically realistic explanation for the progressive change in dyke strike as the STTFZ is approached from the north (Fig. 5a), a pattern that has been unequivocally shown to result from differential vertical-axis tectonic rotations. It is also inconsistent with the dominance of clockwise rotations within the transform zone itself (that demonstrably were synchronous with active magrnatism), and finally cannot be reconciled with the evidence for the presence of a ridgetransform intersection in the Mandria area (Fig. 5d), which indicates that inside-corner crust lies to the present-day east of the Solea axis (Fig. 6b). The large anticlockwise rotation of the 'Troodos microplate' is a regionally significant event within the Neotethys ocean. Initial estimates of the size of the rotated unit (Clube 1985; Clube & Robertson 1986) suggested that rotation was restricted to an oceanic microplate of approximately the same area as the Troodos ophiolite itself, although this now requires re-evaluation in the light of new data from the emplaced ophiolites exposed further east (summarized below). The timing of the rotation may be determined by palaeomagnetic analysis of the continuous upper Cretaceous to Recent in situ sedimentary cover to the ophiolite, on the basis that magnetic declinations within the sediments record rotation

361

of the underlying ophiolitic basement. Several attempts at such analyses have been made, the earliest of which (Shelton & Gass 1980) was hindered by difficulties in measuring the very weak magnetizations of the predominantly carbonate lithologies and a lack of appropriate demagnetization experiments. Subsequent investigations by Clube (1985) and Abrahamsen & Sch6nharting (1987) yielded data distributed through the succession, and further constraints on the early rotation history were provided by Morris et al. (1990). Site-level data from these studies that meet the quality criteria outlined above are shown in Figure 7a. Hydrothermal sediments (umbers) of the Perapedhi Formation share a common direction with the underlying extrusive sequence. The overlying Campanian radiolarian mudstones (Perapedhi Formation) and Maastrichtian-Oligocene chalks (Lefkara Formation) show a general progression from WNW to northerly declinations upwards through the stratigraphy. Data for some time intervals show significant scatter, most notably in the inclination of the Maastrichtian sites and i n b o t h inclination and declination of the Paleocene sites. This may reflect the influence of compaction-related inclination shallowing and/ or potential contamination of site-level remanence vectors by residual normal polarity overprints. This latter effect is the most likely explanation of the inclination difference between the two Maastrichtian sites with the shallowest inclinations (representing inverted reversed polarity sites) and the remaining three (normal polarity) sites of this age. Late Oligocene to Miocene sites exhibit exclusively northerly declinations (within error). Figure 7b shows the variation of rotation angles through time, derived from a comparison of mean remanence data for time periods represented by three or more site mean directions in Figure 7a with expected directions calculated from the African master apparent polar wander path of Besse & Courtillot (2002). These data clearly indicate the progressive and prolonged nature of the rotation of the Troodos microplate during the CampanianEocene interval. However, interpretation in terms of variations in rotation rate through time is made difficult by palaeomagnetic uncertainties relating to limited site numbers and lack of more precise dating of sampled units. Hatay and Bar

ophiolites

Morris & Anderson (2002), Morris et al. (2002), Inwood et al. (2003) and Inwood (2005) have recently provided the first palaeomagnetic data from the emplaced ophiolites directly to the east of Troodos. Layered gabbros, and massive and pillowed lava flows of the Hatay ophiolite

362

A. MORRIS E T A L .

Fig. 6. (a) Summary of the Cann et al. (2001) model for the origin of the Limassol Forest Complex. This model invokes a sinistrally slipping transform between dextrally offset spreading axes to account for an inferred inside-corner setting for the development of extensional structures observed in the Limassol Forest area (modified from Cann et al. 2001). (b) The geometry of the alternative model that is consistent with field and palaeomagnetic evidence for clockwise transform-related rotations and their distribution relative to the Solea spreading axis (as discussed in the text). In this model, extensional structures in the Limassol Forest Complex result from post-volcanic reactivation of the STTFZ during the early stages of microplate rotation, rather than the development of synspreading oceanic core complexes. RTI, ridge-transform intersection; STTFZ, South Troodos Transform Fault Zone.

(Fig. 8a) share a common tilt-corrected inclination with declinations strung out along a partial small circle through this inclination. The mean direction of D = 274 ~ I = 34 ~ is indistinguishable from that commonly reported for non-transform tectonized parts of the Troodos ophiolite (D =273 ~ I = 38 ~ (this paper); D =276 ~ I = 32 ~ (Vine & Moores 1969); or D = 2 7 4 ~ 1 = 3 6 ~ (Clube & Robertson 1986)), and indicates a mean anticlockwise rotation of 73 ~ when compared with an expected direction derived from the apparent polar wander path for Africa (Besse & Courtillot 2002). The distribution of vectors is comparable with that observed in the Solea area

of Troodos (stereonet 3, Fig. 4). The spread of declinations may result from relatively minor vertical-axis rotations during crustal accretion and/or the cumulative effects of minor tilting or rotation during phases of post-emplacement faulting (Inwood 2005). Finally, a contribution to the spread of declinations from transformtectonism cannot be excluded (although crosscutting units, characteristic of such deformation within the STTFZ in Cyprus, were not observed in the field). Data for the post-emplacement sedimentary cover of the H a t a y ophiolite, reported by Kissel et al. (2003) and Inwood (2005), include sites of

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS

363

Fig. 7. (a) Lower hemisphere stereographic projection of tilt-corrected site-level palaeomagnetic data obtained from the in situ sedimentary cover of the Troodos ophiolite. Data from Campanian radiolarites and Maastrichtian and Paleocene carbonates include both normal and reversely magnetized sites, with the latter inverted to the lower hemisphere in this plot. Eocene to Miocene data are all of normal polarity. Data sources: Clube (1985); Abrahamsen & Sch6nharting (1987); Morris et al. (1990). (b) Variation of rotation angle through time derived by comparing mean data for stratigraphic intervals with three or more palaeomagnetic site mean directions with expected directions calculated from the African apparent polar wander path of Besse & Courtillot (2002). These data clearly demonstrate the progressive anticlockwise rotation of the underlying ophiolite during the Late Cretaceous and Palaeogene.

both normal and reverse polarity (Fig. 8c) that pass reversal and tilt tests (Inwood 2005) suggesting that pre-deformational remanences are preserved. These data, including those for the oldest (Palaeocene) sequences, indicate only minor anticlockwise rotation (Fig. 8c). This contrasts with the large anticlockwise rotations observed in the underlying ophiolite, and confirms that most (c. 62 ~ of the rotation of the ophiolite occurred in an intra-oceanic setting and/or during tectonic emplacement in the Maastrichtian. Results from the highly dismembered Ba~rBassit ophiolite (Morris et al. 2002), representing the leading edge of the emplaced ophiolite sheet (Robertson 2002), indicate extreme relative rotations between sampled localities (Fig. 8b), with rotations varying on a kilometre scale. The eastern, Baar massif is dominated by mantle sequence rocks and data are available from only one locality, in the sheeted dyke complex. This is dominated by tilting with only moderate components of anticlockwise rotation about a vertical axis (NW directions, Fig. 8b). Within the Bassit massif, anticlockwise net tectonic rotations increase from c. 90 ~ in the north (westerly directions, Fig 8b) to in excess of 200 ~ in the south (SE

directions, Fig. 8b), and rotation axes determined using the net tectonic rotation approach are steeply plunging to subvertical. The possibility that the largest rotations occurred in a clockwise sense cannot be excluded, but anticlockwise solutions were preferred by Morris et al. (2002), as this results in a systematic pattern of rotations across the Bassit massif. Data from another eight sites at a fifth locality (in the Bassit massif) where subvertical remanence directions resulted from rotation around shallow- to moderately plunging dyke-normal axes are not shown in Figure 8b but were discussed in detail by Morris & Anderson (2002). The increase in rotation angles southwards through the Bassit sheet is most readily explained by increasing proximity to a major strike-slip fault zone that traverses the southern half of the ophiolite, representing the expression of the present-day plate boundary system in this region (A1-Riyami et al. 2000). Palaeomagnetic data for the Palaeogene post-emplacement sedimentary cover sequences within and adjacent to this fault zone (Morris et al. 2006) reveal widely different rotation angles and senses in different fault blocks (Fig. 8d), confirming that significant

364

A. MORRIS E T AL.

Fig. 8. Outline geological map of the Hatay (K~zll Da~) and BaEr-Bassit ophiolites, showing the distribution of lithologies and major structures. Lower hemisphere stereographic projections of tilt-corrected site-level palaeomagnetic data from: (a) and (b) the Hatay and BaEr-Bassit ophiolites, respectively (data from Inwood et al. (2003), Inwood (2005) and Morris et al. (2003)); in (b), reversed polarity data (14 out of 19 sites) have been inverted to allow ready comparison with (a); (c) the post-emplacement sedimentary cover of the Hatay ophiolite (data from Inwood et al. (2003), Kissel et al. (2003) and Inwood (2005)); (d) and (e) the Palaeogene and Neogene post-emplacement sedimentary cover of the Ba~r-Bassit ophiolite (data from Morris et al. (2006)). An inclination-onlytilt test performed on data from the Palaeogene sequences indicates maximum clustering of inclinations at 50% untilting. Black star in stereographic projection (a) indicates the direction of the present-day geomagnetic field.

rotational deformation accompanied the development of these neotectonic structures. The weakly magnetized Paleocene carbonates are gently folded, and an inclination-only tilt test suggests that remanences were acquired during folding. Data from Neogene sequences (Fig. 8e) exposed to the NE are inherently more difficult to interpret, as the distinction between northerly directed viscous overprints and unrotated Neogene magnetizations is impossible in these subhorizontal sequences. Overall, the data obtained to date from the Hatay and Ba~r-Bassit ophiolites and their postemplacement sedimentary cover sequences are most consistent with large-scale, bulk rotation of the ophiolitic sheet, either in an intraoceanic setting or during emplacement. The northern, Hatay, part of the emplaced sheet experienced

only moderate post-emplacement tectonic rotation (Inwood et al. 2003; Kissel et al. 2003; Inwood 2005), in contrast to the more highly deformed southern leading edge in BaEr-Bassit, where subsequent amplification of the amount of rotation resulted from neotectonic modification during development of the modern plate boundary configuration.

Discussion of regional implications Palaeogeographical

implications

Palaeolatitudes of c. 21-24~ derived from inclination-only statistical analysis of the ophiolite palaeomagnetic database are consistent with a Late Cretaceous position for the Neotethyan

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS

365

Fig. 9. End-member tectonic cross-sections along a north-south transect (from the Arabian margin, through the Bafir-Bassit spreading axis and eastern Pontides to the Eurasian margin) that satisfy the palaeomagnetic constraints on the palaeolatitude of the Eurasian and Arabian continental margins, the Ba~r-Bassit spreading axis and the eastern Pontide volcanic arc. Upper and lower sections respectively illustrate the solution that yields the widest and narrowest south Neotethyan ocean (Eur, Eurasian margin; EP, eastern Pontides; Tau, Taurides; BB, Ba~r-Bassit segment of the Neotethyan spreading axis; BS, Black Sea; Arab, Arabian margin).

spreading axis between the Arabian and Eurasian margins (Morris 2003). These data, together with a well-defined palaeolatitude of c. 26~ for the eastern Pontides (Channell et al. 1996) and palaeolatitudes of the Arabian and Eurasian margins derived from appropriate apparent polar wander paths (e.g. Besse & Courtillot 1991, 2002), provide constraints on potential tectonic reconstructions of the eastern Mediterranean Tethys. Two model reconstructions that involve genesis of the ophiolites in a southerly basin (e.g. Robertson & Dixon 1984; Robertson et al. 1991; Robertson 1998) and that satisfy the palaeolatitudinal data are shown in Figure 9, using the data from the Ba~r-Bassit ophiolite as an example. The solution giving the widest southerly Neotethys is obtained by placing the Arabian continental margin at its southernmost limit. This produces a 1000km wide Neotethyan strand to the south of the subduction zone associated with the Ba6r-Bassit spreading axis. Major arc magmatism would result from subduction of this lithosphere during continued plate convergence. Arc-related products are sparse, although in SE Turkey the Baskil arc (Akta~ & Robertson 1984; Yazgan & Chessex 1991; Rizao[glu et al. 2006) 200 km to the NE of Ba~r-Bassit provides evidence for Andean-type arc magmatism along the southern Tauride margin during the late Cretaceous (Robertson 2002). The problem of

the scarcity of arc products in the wider region is potentially avoided in the 'narrowest solution' (and near intermediate solutions), which implies an oceanic tract c. 300 km wide to the south of the subduction zone (Fig. 9). In both solutions, the northern Neotethyan strand was essentially consumed by subduction beneath the Pontides. This is consistent with geological evidence suggesting that northward subduction beneath the Pontides was active from the Late Mesozoic onwards (Usta6mer & Robertson 1993, 1994), providing sufficient time for the northern Neotethys to have been subducted by the Late Cretaceous. The 'narrowest solution' also implies a restricted width of the southern Neotethyan strand following ophiolite emplacement (Fig. 9). In this respect, intermediate solutions that place the Eurasian margin further to the north (thereby permitting a wider oceanic tract between the Taurides and the Neotethyan spreading axis) are more consistent with geological evidence indicating that a southern Tethyan basin persisted to the north of the Arabian margin well into the Tertiary after partial basin closure associated with ophiolite emplacement (Aktas & Robertson 1984; Yflmaz 1993; Robertson et al. 1996). Alternative models (e.g. Ricou 1971; Ricou et al. 1979, 1984; Dercourt et al. 1986, 1993) involve formation of the ophiolites in a northerly Neotethyan basin by spreading at a 'normal'

366

A. MORRIS E T AL. Age implications

Fig. 10. Summary of available radiometric age constraints for the Troodos, Hatay and Ba~r-Bassit ophiolites and the Ba~r-Bassit metamorphic sole (together with biostratigraphic constraints on the timing of emplacement of the Hatay and Bafir-Bassit thrust sheets), in relation to the geomagnetic polarity time scale for the Late Cretaceous (Cande & Kent 1995). Sources are referred to in the text.

oceanic ridge, with subsequent, Late Cretaceous large-scale thrusting (hundreds of kilometres) to the south of emplaced ophiolites over microcontinental fragments to reach their present positions. These cannot be discounted on a purely palaeomagnetic basis in the absence of reliable data from the eastern Taurides (Morris 2003). Such models are not supported, however, by a number of key geological observations (described in detail elsewhere, e.g. Robertson et al. 1996), including the presence of an essentially unbroken Campanian to Lower Tertiary sedimentary sequence overlying the Troodos ophiolite, and also the continuous upper Palaeozoic to Neogene sedimentary sequences exposed in parts of the Tauride Mountains (notably in the Isparta Angle, SW Turkey; e.g. Robertson et al. 2003).

Available radiometric ages for the Troodos, Hatay and Ba6r-Bassit ophiolites, and for the Ba~r-Bassit metamorphic sole, are shown in Figure 10, in relation to the geomagnetic polarity time scale for the Late Cretaceous and the biostratigraphical constraints on the timing of tectonic emplacement of Hatay and Ba~r-Bassit ophiolites. Pre-deformational characteristic remanences in the Troodos ophiolite are ubiquitously of normal polarity, consistent with the U Pb age of 90-92 Ma (Mukasa & Ludden 1987) indicating formation during the Cretaceous Long Normal Period (chron C34N; Fig. 10). Reverse polarity overprints are observed in the Troodos Sheeted Dyke Complex (Gee et al. 1993), but are not ubiquitous. Where they are present, normal polarity characteristic magnetizations are also isolated. These overprints were demonstrably acquired prior to tilting of the dykes (Gee et al. 1993), and are attributed to low-temperature alteration during reversed polarity chron C33R (early Campanian; Gee et al. 1993). Polarities of characteristic magnetizations in the Hatay ophiolite are dominantly normal, in contrast to dominantly reversed polarities observed in the Ba~r-Bassit ophiolite. This suggests acquisition of pre-deformational remanences during at least two polarity chrons and hence significant variations in age within the emplaced ophiolite sheet. Radiometric ages for the Hatay and Ba~r-Bassit ophiolites are at present too poorly constrained to resolve these age differences, with dykes yielding K-Ar ages in the range 73-99 Ma (Delaloye & Wagner 1984) spanning polarity chrons C33N, C33R and part of C34N (Fig. 10). Whether reversed polarity remanences in Bafir-Bassit could be acquired during chron C33R, however, depends critically on the mode of formation of the Bafir-Bassit metamorphic sole, which has a K-Ar age of 86-93 Ma (Thuizat et al. 1981; Delaloye & Wagner 1984), i.e. within chron C34N (Fig. 10). Assuming that this age is reliable, alternative interpretations are as follows: (1) The metamorphic sole may have formed during initial detachment of the oceanic crust near the Neotethyan spreading axis (Whitechurch 1977; Coleman 1981; Boudier et al. 1982), and hence represents the latest age of formation of the ophiolite. The observed reversed polarities would then require a substantial (possibly > 30 Ma) age difference between Ba~r-Bassit and Troodos, with the magnetization of the former being acquired during the Early Cretaceous or within a poorly documented reverse polarity event within chron C34N (Hailwood 1989). (2) The sole may have formed during the initiation of subduction,

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS

367

Fig. 11. Schematic illustration of the motion of the Arabian continental margin (Ar) through the Late Cretaceous. Palaeolatitudinal constraints are derived from the African apparent polar wander path of Besse & Courtillot (2002), after correcting for the effects of Red Sea opening (using the Euler pole of Savostin et al. (1986)), Large grey arrow indicates the relative motion vector of Africa-Arabia relative to a fixed Eurasia (from Dewey et al. 1989). Thick black arrows illustrate the amount of palaeorotation of the Troodos microplate between time frames. The palaeolatitude of the microplate is accurately constrained only in (a), and cannot be determined reliably for subsequent time periods because of the potential effects of sedimentary compaction on palaeomagnetic inclination.

before SSZ spreading of the Bafir-Bassit crust (e.g. Casey & Dewey 1984). This would reconcile the available radiometric and magnetic polarity age constraints, and would require SSZ spreading to have continued over a c. 10 Ma period between the initiation of subduction in the Turonian and the start of microplate rotation in the Campanian. More reliable, higher resolution radiometric dates for the Hatay and Ba6r-Bassit ophiolite and metamorphic sole are clearly required to resolve this debate. Implications for intra-oceanic microplate rotation

Palaeomagnetic data from the Troodos ophiolite and its sedimentary cover are near universally interpreted in terms of intra-oceanic rotation of a 'Troodos microplate'. Data from the sedimentary cover of the Troodos ophiolite indicate that a large component (50-60~ Fig. 7) of intraoceanic anticlockwise rotation had occurred by the Maastrichtian, i.e. by the time of emplacement of the Hatay and Ba6r-Bassit ophiolite sheet onto the Arabian margin. This rotation angle is comparable with the mean rotation observed in the Hatay ophiolite (Inwood et al. 2003; Inwood 2005; Fig. 8a), after removing the post-emplacement rotation recorded by its postemplacement sedimentary cover (Inwood et al. 2003; Kissel et al. 2003; Inwood 2005; Fig. 8c), and also represents a large component of the more extreme rotations observed in Ba~r-Bassit.

Hence, these data are most readily explained by coherent rotation of a Neotethyan oceanic microplate that was more areally extensive than inferred from the Troodos data alone (Inwood et al. 2003; Inwood 2005). The mechanism of microplate rotation is difficult to identify with certainty, particularly at the level of identifying accommodating (bounding) structures. A common feature of existing models (e.g. Clube et al. 1985; Clube & Robertson 1986; Robertson 1990) is that rotation is related to oblique convergence across the southern Neotethyan subduction zone, resulting from NE motion of Arabia relative to Eurasia throughout the Late Cretaceous and Early Tertiary (Dewey et al. 1989). After correcting for the effects of opening of the Red Sea, the African apparent polar wander path (Besse & Courtillot 2002) places the northernmost Arabian continental margin at c. 16~ during the Turonian, several 100 km to the south of the southern Neotethys spreading axis at 21-24~ (Fig. 1 la). Subsequent motion of Arabia to the NE (Fig. 1 lb and c) may then have generated a sinistral sense of motion across the subduction zone separating the SSZ oceanic crust from Arabia. Within this overall plate-scale framework, impingement of the Arabian continental margin with the subduction trench has been invoked as a potentially major contributor to the initiation and progression of microplate rotation (e.g. Clube & Robertson 1986; Robertson 1990). Although further higher resolution palaeomagnetic and

368

A: MORRIS E T AL.

biostratigraphic data from the Troodos sedimentary cover are required to reduce uncertainties in the progressive history of microplate rotation (Fig. 7b), rotation was clearly under way for at least 10-15 Ma (and possibly up to 20 Ma) prior to ophiolite emplacement in the Maastrichtian. The timing of approach of the Arabian continental margin to the trench cannot be accurately determined, but if this impingement acted as a trigger for initiation of rotation then the palaeomagnetic data suggest that rotation progressed for a substantial period of time before eventual emplacement of part of the rotated unit onto the continental margin. Finally, we note that anticlockwise intra-oceanic rotation of the eastern Mediterranean ophiolites considered here contrasts with the clockwise rotation of the Oman (Semail) ophiolite further to the east (e.g. Weiler 2000). This suggests that rotation in both cases was controlled by plate-scale geometry of the Arabian margin during regional convergence. Palaeomagnetic data are now required from the emplaced ophiolites to the east of Hatay and Ba~r-Bassit (i.e. the Kermanshah and Neyriz ophiolites of Iran; Fig. 1) to further constrain the pattern and hence the mechanism of Neotethyan intra-oceanic rotations.

Conclusions

(1) The Late Cretaceous Troodos, Hatay

(2)

and BaEr-Bassit ophiolites of the eastern Mediterranean Tethyan orogenic belt preserve remanent magnetizations of preformational age. Palaeomagnetic data from these units, in conjunction with data from the overlying in situ (Troodos) and postemplacement (Hatay and Ba~r-Bassit) sedimentary cover sequences, therefore provide insights into the styles and timing of rotational deformation that have affected the Neotethyan oceanic crust. Within the Troodos ophiolite, localized rotations are demonstrably related to processes operating during construction of oceanic crust at a Neotethyan spreading axis in close proximity to an oceanic transform fault zone. Rotations around (sub)horizontal, dyke-parallel axes are associated with crustal extension during sea-floor spreading. Rotations around (sub)vertical axes dominate in areas adjacent to the South Troodos Transform Fault Zone, reflecting rotation of kilometre-scale fault blocks in response to dextral shearing within weak crust at the ridge-transform intersection.

(3) The tectonically emplaced Hatay and Bafir-

(4)

Bassit ophiolites record large, and locally extreme, anticlockwise rotations. Within the Baar-Bassit ophiolite, the magnitude of observed rotations increases generally southwards. Analysis of the overlying postemplacement sedimentary successions suggests that this amplification of rotation reflects the development of a strike-slip fault system related to the initiation of the present-day plate configuration. The Hatay ophiolite and the most northerly locality in the Ba~r-Bassit ophiolite share a similar anticlockwise rotation. After removing the post-emplacement component of rotation documented in the Hatay postemplacement sedimentary cover, this rotation angle is directly comparable with the pre-Maastrichtian rotation of the 'Troodos microplate'. At the regional scale, therefore, these data are best explained by intraoceanic anticlockwise rotation of a more areally extensive oceanic microplate than has been considered previously (Inwood et al. 2003; Inwood 2005).

We would like to thank K. AI-Riyami for field assistance in the Ba6r-Bassit ophiolite, and U.-C. r162 for administrative and logistical support during fieldwork in the Hatay ophiolite. Inclination-only tilt tests were performed using software by R. Enkin.

References ABELSON,M., BAER,G. & AGNON,A. 2001. Evidence from gabbro of the Troodos ophiolite for lateral magma transport along a slow-spreading midocean ridge. Nature, 409, 72-75. ABRAHAMSEN,N. & SCHONHARTING,G. 1987. Palaeomagnetic timing of the rotation and translation of Cyprus. Earth and Planetary Science Letters, 81, 409-418. AKTA~,G. & ROBERTSON,A. H. F. 1984. The Maden Complex, S. E. Turkey: evolution of a Neotethyan continental margin. In: DIXON,J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 375-402. ALLERTON, S. 1988. Palaeomagnetic and structural studies of the Troodos ophiolite, Cyprus. PhD thesis, University of East Anglia, Norwich. ALLERTON, S. 1989a. Fault block rotations in ophiolites: results of palaeomagnetic studies in the Troodos Complex, Cyprus. In: KISSEL,C. 8r LAJ,C. (eds) Palaeomagnetic Rotations and Continental Deformation. NATO ASI Series C, 254, 393-410. ALLERTON,S. 1989b. Distortions, rotations and crustal thinning at ridge-transform intersections. Nature, 340, 626-628.

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS ALLERTON, S. & VINE, F. J. 1987. Spreading structure of the Troodos ophiolite, Cyprus: some palaeomagnetic constraints. Geology, 16, 593-597. ALLERTON, S. & VINE, F. J. 1990. Palaeomagnetic and structural studies of the southeastern part of the Troodos complex. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. & XENOPHONTOS,C. (eds) Ophiolites: Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 99-111. ALLERTON, S. & VINE, F. J. 1991. Spreading evolution of the Troodos ophiolite, Cyprus. Geology, 19, 637-640. AL-RIYAMI, K. & ROBERTSON, A. H. F. 2002. Mesozoic sedimentary and magrnatic evolution of the Arabian continental margin, northern Syria: evidence from the Ba~r-Bassit Mrlange. Geological Magazine, 139, 395-420. AL-RIYAMI, K., ROBERTSON, A. H. F., XENOPHONTOS, C., DANELIAN, T. & DIXON, J. E. 2000. Tectonic evolution of the Mesozoic Arabian passive continental margin and related ophiolite in Ba~rBassit region (NW Syria). In: PANAYIDES, I., XENOPHONTOS, C. & MALPAS, S. (eds) Proceedings of the Third International Conference on the Geology of the Eastern Mediterranean. Geological Survey Department, Nicosia, 61-82. AL-RIYAMI, K., ROBERTSON, A. H. F., DIXON, J. E. & XENOPHONTOS, C. 2002. Origin and emplacement of the Late Cretaceous Baer-Bassit ophiolite and its metamorphic sole in NW Syria. Lithos, 65, 225-260. 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. BESSE, J. & COURTXLLOT, V. 1991. Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian plates and true polar wander since 200 Ma. Journal of Geophysical Research, 96, 4029-4050. BESSE, J. & COURTILLOT, V. 2002. Apparent and true polar wander and the geometry of the geomagnetic field over the last 200 Myr. Journal of Geophysical Research, 107, dot: 10.1029/2000JB000050. BLACKMAN, D. K., CANN, J. R., JANSSEN, B. & SMITH, D. K. 1998. Origin of extensional core complexes: evidence from the Mid-Atlantic Ridge at the Atlantis Fracture Zone. Journal of Geophysical Research, 103, 21315-21333. 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, 417426. 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. BOUDIER, F., NICOLAS, A. & BOUCHEZ,J. L. 1982. Kinematics of oceanic thrusting and subduction from basal sections of ophiolites. Nature, 296., 825-828. BOULTON, S. J., ROBERTSON, m. H. F. & UNLOGEN~, U. C. 2006. Tectonic and sedimentary evolution of

369

the Cenozoic Hatay Graben, Southern Turkey: a two-phase model for graben formation. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 613-634. CANDE, S. C. & KENT, D. V. 1995. Revised calibration of the geomagnetic polarity timescale for the late Cretaceous and Cenozoic. Journal of Geophysical Research, 100, 6093-6095. CANN, J. R., PRICHARD, H. M., MALPAS, J. G. & XENOPHONTOS, C. 2001. Oceanic inside comer detachments of the Limassol Forest area, Troodos ophiolite, Cyprus. Journal of the Geological Society, London, 158, 757-767. CASEY, J. F. & DEWEy, J. F. 1984. Initiation of subduction zones along transforms and accreting plate boundaries, triple junction evolution, and fore-arc spreading centres~implications for ophiolitic geology and obduction. In: GASS, I. G., LIPPARD, S. J. & SHELTON,A. W. (eds) Ophiolites and Oceanic Lithosphere. Geological Society, London, Special Publications, 13, 269-290. CHANNELL, J. E. T., TOYSOZ, O., BEKTA~, O. & SENGOR, A. M. C. 1996. Jurassic-Cretaceous paleomagnetism and paleogeography of the Pontides (Turkey). Tectonics, 15, 201-212. CLUBE, T. M. M. 1985. The palaeorotation of the Troodos microplate. PhD thesis, University of Edinburgh. CLUBE, T. M. M., CREER, K. M. & ROBERTSON, A. H. F. 1985. The palaeorotation of the Troodos microplate. Nature, 317, 522-525. CLUBE, T. M. M. & ROBERTSON, A. H. F. 1986. The palaeorotation of the Troodos microplate, Cyprus, in the Late Mesozoic-Early Cenozoic plate tectonic framework of the Eastern Mediterranean. Surveys in Geophysics, 8, 375434. COLEMAN, R. G. 1981. Tectonic setting for ophiolite obduction in Oman. Journal of Geophysical Research, 86, 2497-2508. DELALOYE, M. & WAGNER, J.-J. 1984. Ophiolites and volcanic activity near the western edge of the Arabian plate. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 25-233. DELALOYE, M., PISKIN, Q)., SELCUK, H., VUAGNAT, M. & WAGNER, J. 4. 1980. Geological section through the Hatay ophiolite along the Mediterranean coast, southern Turkey. Ofioliti, 5, 205-216. DELAUNE-MAYI~RE,M. 1984. Evolution of a Mesozoic passive continental margin: Ba~r-Bassit (NW Syria). In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 10-11. 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. DERCOURT, J., RICOU, L. E. & VRIELYNCK, B. (eds) 1993. Atlas Tethys Palaeoenvironmental Maps. Beicip-Franlab, Rueil-Malmaison, France.

370

A. MORRIS ET AL.

DEWEY, J. F., HELMAN, M. L., TURCO, E., HUTTON, D. H. W. & KNOTT, 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. DIETRICH, D. & SPENCER, S. 1993. Spreading-induced faulting and fracturing of oceanic crust: examples from the Sheeted Dyke Complex of the Troodos ophiolite, Cyprus. In: PRICHARD, n. M., ALABASTER, T., HARRIS, N. B. & NEARY, C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 121-139. DILEK, Y. & THY, P. 1998. Structure, petrology and seafloor spreading tectonics of the Klzllda~ ophiolite, Turkey. In: MILLS, R. A. & HARRISON, K. (eds) Modern Ocean Floor Processes and the Geological Record. Geological Society, London, Special Publications, 148, 43-69. DUNLOP, D. J. & OZDEMIR,1~). 1997. Rock Magnetism." Fundamentals and Frontiers. Cambridge University Press, Cambridge. ENKIN, R. J. & WATSON, G. S. 1996. Statistical analysis of palaeomagnetic inclination data. Geophysical Journal International, 126, 495-504. ERENDIL, M. 1984. Petrology and structure of the upper crustal units of the Klzd Da(g ophiolite. In: TEKELI, O. & GONCOO~LO, C. M. (eds) Geology of the Taurus Belt. Proceedings, Mineral Research and Exploration Institute, Ankara, 269-284. GASS, I. G. 1968. Is the Troodos massif of Cyprus a fragment of Mesozoic ocean floor? Nature, 220, 39-42. GASS, I. G. 1980. The Troodos massif: its role in the unravelling of the ophiolite problem and its significance in the understanding of constructive margin processes. In: PANAYIOTOU,A. (ed.) Ophiolites: Proceedings of the International Symposium, Cyprus, 1979. Geological Survey Department, Nicosia, 23-35. GASS, I. G., MACLEOD, C. J., MURTON, B. J., PANAYIOUTOU, A., SIMONIAN, K. O. & XENOPHONTOS, C. 1991. Geological map of the South Troodos Transform Fault Zone at 1:25 000: Sheets 1 (west) and 2 (east). Geological Survey Department, Nicosia. GASS, I. G., MACLEOD, C. J., MURTON, B. J., PANAYIOUTOU, A., SIMONIAN, K. O. & XENOPHONTOS, C. 1994. The Geology of the South Troodos Transform Fault Zone. Geological Survey Department, Nicosia, Memoirs, 9. GEE, J., VARGA, R., GALLET, Y. & STAUDIGEL, H. 1993. Reversed-polarity overprint in dikes from the Troodos ophiolite: implications for the timing of alteration and extension. Geology, 21, 849-852. HAILWOOD, E. A. 1989. Magnetostratigraphy. Geological Society, London, Special Report, 19. HURST, S. D., VEROSUB,K. L. & MOORES, E. M. 1992. Paleomagnetic constraints on the formation of the Solea Graben, Troodos Ophiolite, Cyprus. Tectonophysics, 208, 431-445. HURST, S. D., KARSON, J. A. & VEROSUB, K. L. 1994. Palaeomagnetism of tilted dikes in fast spread

oceanic crust exposed in the Hess Deep Rift: implications for spreading and rift propagation. Tectonics, 13, 789-802. INWOOD, J. 2005. The tectonic evolution of the Hatay (K1zll Da~) ophiolite, southern Turkey. PhD thesis, University of Plymouth. INWOOD, J., MORRIS, A., ANDERSON, M. W., ROBERTSON, A. H. F. & (~INLOGEN~,U. 2003. First palaeomagnetic results from the Hatay (Kazd Da~) ophiolite of Turkey and their implication for the tectonic evolution of the eastern Mediterranean Neotethys. Geophysical Research Abstracts, 5, 09595. KAZMIN, V. G. & KULAKOV, V. V. 1968. The Geological Map of Syria, Scale 1:50000 (Sheet AL Latheqiyeh). Explanatory note. Technoexport, Nedra, Moscow. KISSEL, C., LAJ, C., POISSON, A. & GOROR, N. 2003. Paleomagnetic reconstruction of the Cenozoic evolution of the eastern Mediterranean. Tectonophysics, 362, 199-217. MACDONALD, W. D. 1980. Net tectonic rotation, apparent tectonic rotation, and the structural tilt correction in palaeomagnetic studies. Journal of Geophysical Research, 85, 3659-3669. MACLEOD, C. J. 1988. The tectonic evolution of the eastern Limassol Forest Complex, Cyprus. PhD thesis, Open University, Milton Keynes. MACLEOD, C. J. 1990. Role of the Southern Troodos Transform Fault in the rotation of the Cyprus microplate: evidence from the Eastern Limassol Forest. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites: Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 75-85. MACLEOD, C. J. & MURTON, B. J. 1993. Structure and tectonic evolution of the Southern Troodos Transform Fault Zone, Cyprus. In: PRICHARD, H. M., ALABASTER,Y., HARRIS, N. B. W. & NEARY,C. R. (eds) Magmatic Processes and Plate Tectonics. Geological Society, London, Special Publications, 76, 141-176. MACLEOD, C. J. & MURTON, B. J. 1995. On the sense of slip of the Southern Troodos Transform-Fault Zone, Cyprus. Geology, 23, 257-260. MACLEOD, C. J., ALLERTON, S., GASS, I. G. & XENOPHONTOS, C. 1990. Structure of a fossil ridgetransform intersection in the Troodos ophiolite. Nature, 348, 717-720. 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. MCELHINNY, M. W. 1964. Statistical significance of the fold test in palaeomagnetism. Geophysical Journal of the Royal Astronomical Society, 8, 338-340. MCFADDEN, P. L. & JONES, D. L. 1981. The fold test in palaeomagnetism. GeophysicalJournal of the Royal Astronomical Society, 67, 53-58. MOORES, E. M. & VINE, F. J. 1971. The Troodos Massif, Cyprus and other ophiolites as oceanic

PALAEOMAGNETIC INSIGHTS INTO NEOTETHYS crust: evaluation and implications. Philosophical Transactions of the Royal Society of London, Series A, 268, 433-466. MORRIS, A. 2003. The Late Cretaceous palaeolatitude of the Neotethyan spreading axis in the eastern Mediterranean region. Tectonophysics, 377, 157-178. MORRIS, A. & ANDERSON, M. W. 2002. Palaeomagnetic results from the Bafir-Bassit ophiolite of northern Syria and their implication for fold tests in sheeted dyke terrains. Physics and Chemistry of the Earth, 27, 1215-1222. MORRIS, A., CREER, K. M. ~r ROBERTSON, A. H. F. 1990. Palaeomagnetic evidence for clockwise rotations related to dextral shear along the Southern Troodos Transform Fault, Cyprus. Earth and Planetary Science Letters, 99, 250-262. MORRIS, A., ANDERSON, M. W. & ROBERTSON, A. H. F. 1998. Multiple tectonic rotations and transform tectonism in an intra-oceanic suture zone, SW Cyprus. Tectonophysics, 299, 229-253. MORRIS, A., ANDERSON, M. W., ROBERTSON, A. H. F. & AL-RIYAMI, K. 2002. Extreme tectonic rotations within an eastern Mediterranean ophiolite (BaarBassit, Syria). Earth and Planetary Science Letters, 202, 247-261. MORRIS, A., INWOOD, J., ANDERSON, M. W. & ROBERTSON, A. H. F. 2006. First palaeomagnetic results from Tertiary carbonates of NW Syria, and their implications for the timing of extreme tectonic rotations in the Ba~r-Bassit ophiolite. Earth and Planetary Science Letters (in press). MUKASA, S. B. & LUDDEN, J. N. 1987, Uranium-lead isotopic ages of plagiogranites from the Troodos ophiolite, Cyprus, and their tectonic significance. Geology, 15, 825-828. MURTON, B. J. 1986. Anomalous oceanic lithosphere formed in a leaky transform fault: evidence from the western Limassol Forest Complex, Cyprus. Journal of the Geological Society, London, 143, 845-854. MURTON, B. J. 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. Geological Survey Department, Nicosia, 87-98. PARROT, J. F. 1980. The Baer-Bassit (Northwestern Syria) ophiolitic area. Ofioliti, 2, 279-295. PEARCE, J. A. 1975. Basalt geochemistry used to investigate past tectonic environments in Cyprus. Tectonophysics, 25, 41-67. PEARCE, J. A. 1980. Geochemical evidence for the genesis and setting of lavas from Tethyan ophiolites. In: PANAYIOTOU, A. (ed.) Ophiolites: Proceedings of the International Symposium, Cyprus, 1979. Geological Survey Department, Nicosia, 261-272. PIPKIN, O., DELALOYE, M. & WAGNER, J. J. 1986. Guide to the Hatay geology. Ofioliti, 11, 87-104. RlfOU, L.-E. 1971. Le croissant ophiolitique p6riarabe, une ceinture de nappes mises en place au Cretac6 supOrieur. Revuede Gdographie Physique et de Gkologie Dynamique, 13, 327-349.

371

RICOU, L.-E., MARCOUX, J. & POISSON, A. 1979. L'allochthonie des Bey Da~lan orientaux, Reconstruction palinspastique des Taurides occidentales. Bulletin de la Societk Gkologique de France, 21, 125-134. RICOU, L.-E., MARCOUX, J. & WHITECHURCH, H. 1984. The Mesozoic organisation of the Taurides: one or several ocean basins. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 349-360. RIZAO(3LU, T., PARLAK,O., HOCK, V. & I~LER, F. 2006. Nature and significance of Late Cretaceous ophiolitic rocks and their relation to the Baskil granitic intrusions of the Elam~ region, SE Turkey. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 327-349. ROBERTSON, A. H. F. 1977. The Kannaviou Formation, Cyprus: volcaniclastic sedimentation of a probable Late Cretaceous volcanic arc. Journal of the Geological Society, London, 134, 269-292. ROBERTSON, A. H. F. 1986. Geochemistry and tectonic implications of metalliferous and volcaniclastic sedimentary rocks associated with Late Cretaceous ophiolitic extrusives in the Hatay area, Southern Turkey. Ofioliti, 11, 121-140. ROBERTSON, A. H. F. t990. Tectonic evolution of Cyprus. In: MALPAS, J., MOORES, E. M., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites." Oceanic Crustal Analogues. Geological Survey Department, Nicosia, 235-252. 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, m. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 160, Ocean Drilling Program, College Station, TX, 723-782. 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. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the Eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. ROBERTSON, A. H. F. 8r WOODCOCK, N. H. 1979. The Mamonia Complex, southwest Cyprus: the evolution and emplacement of a Mesozoic continental margin. Geological Society of America Bulletin, 90, 651-665. ROBERTSON, A. n . F. & XENOPHONTOS, C. 1993. Development of concepts concerning the Troodos ophiolite and adjacent units in 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, 85-119. ROBERTSON, A. H. F., CLIFT, P. D., DEGNAN, P. J. & JONES, G. 1991. Palaeogeographic and palaeotectonic evolution of the Eastern Mediterranean

372

A. MORRIS ET AL.

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 Palaeozoic-Early Tertiary development of Tethys in the Eastern Mediterranean region. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBERTSON, A. H. F., POISSON, A. & AK1NC1,O. 2003. Developments in research concerning MesozoicTertiary Tethys and neotectonics in the Isparta Angle, SW Turkey. Geological Journal, 38, 195-234. 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 Crustal Analogues. Geological Survey Department, Nicosia, 13-36. SAVOSTIN, L. A., SIBUET, J.-C., ZONENSHAIN, L. P., LE PICHON, X. & ROULET, M.-J. 1986. Kinematic evolution of the Tethys Belt from the Atlantic Ocean to the Pamirs since the Triassic. Tectonophysics, 123, 1 35. SCHMINCKE, H.-U. & RAUTENSCHLEIN, M. 1987. Troodos extrusive series (Akaki River Canyon) and the Sheeted Diabase. In: XENOPHONTOS, C. & MALPAS, J. G. (eds) Field Excursion Guidebook to Accompany the Troodos 87 Ophiolites and Oceanic Lithosphere Symposium. Geological Survey Department, Nicosia, 36-91. SCHMINCKE, H.-U., RAUTENSCHLEIN, M. ROBINSON, P. T. & MEGEHAN, J. M. 1983. Troodos extrusive series of Cyprus: a comparison with oceanic crust. Geology, 11,405-409. SHELTON, A. W. & GASS, I. G. 1980. Rotation of the Troodos microplate. In: PANAYIOTOU, A. (ed.) Ophiolites: Proceedings of the International Symposium, Cyprus, 1979. Geological Survey Department, Nicosia, 61-65. SIMONIAN, K. O. & GASS, I. G. 1978. Arakapas fault belt, Cyprus: a fossil transform belt. Geological Society of America Bulletin, 89, 1220-1230. STAUDIGEL, H., GEE, J., TAUXE, L. & VARGA, R. J. 1992. Shallow intrusive directions of sheeted dikes in the Troodos ophiolite" anisotropy of magnetic susceptibility and structural data. Geology, 20, 841-844. SWARBRICK, R. E. 1979. The sedimentology and structure of S W Cyprus and its relationship to the Troodos Complex. PhD thesis, University of Cambridge. SWARBRICK, R. E. 1980. The Mamonia Complex of SW Cyprus: a Mesozoic continental margin and its relationship to the Troodos Complex. In:

PANAYIOTOU, A. (ed.) Ophiolites: Proceedings of the International Symposium, Cyprus, 1979. Geological Survey Department, Nicosia, 86-92. SWARBRICK, R. E. 1993. Sinistral strike-slip and transpressional tectonics in an ancient oceanic setting: the Mamonia Complex, southwest Cyprus. Journal of the Geological Society, London, 150, 381-392. THUIZAT, R., WHITECHURCH, H., MONTIGNY, R. & JUTEAtJ, T. 1981. K-Ar dating of some intraophiolite metamorphic soles from the East Mediterranean: new evidence for oceanic thrusting before obduction. Earth and Planetary Science Letters, 52, 302-310. USTAOMER, T. & ROBERTSON, A. H. F. 1993. A Late Palaeozoic-Early Mesozoic marginal basin along the active southern continental margin of Eurasia: evidence from the Central Pontides (Turkey) and adjacent regions. Geological Journal, 28, 219-238. USTAOMER, T. & ROBERTSON, A. H. F. 1994. Late Palaeozoic marginal basin and subductionaccretion: the Palaeotethyan Kfire Complex, Central Pontides, northern Turkey. Journal of the Geological Society, London, 151, 291-305. VARGA, R. J. & MOORES, E. M. 1985. Spreading structure of the Troodos ophiolite, Cyprus. Geology, 13, 846-850. VARGA, R. J., GEE, J. S., STAUDIGEL, H. & TAUXE, L. 1998. Dike surface lineations as magma flow indicators within the sheeted dike complex of the Troodos ophiolite, Cyprus. Journal of Geophysical Research, 103, 5241-5256. VEROSUB, K. L. & MOORES, E. M. 1981. Tectonic rotations in extensional regimes and their paleomagnetic consequences for oceanic basalts. Journal of Geophysical Research, 86, 6335-6349. VINE, F. J. & MOORES, E. M. 1969. Palaeomagnetic results from the Troodos Igneous Massif, Cyprus. Transactions of the American Geophysical Union, 50, 131. WEILER, P. D. 2000. Differential rotations in the Oman Ophiolite: paleomagnetic evidence from the southern massifs. Marine Geophysical Researches, 21, 195-210. WHITECHURCH, H. 1977. Les roches mktamorphiques infrapbridotitiques du Ba~r-Bassit ( N W Syrien), tbmoins de l'kcaillage intraockanique tkthysien. Etude pOtroloqique et structurale. PhD thesis, Nancy University, France. YAZGAN, E. & CHESSEX,R. 1991. Geology and tectonic evolution of Southeastern Taurus in the region of Malatya. Turkish Association of Petroleum Geologists Bulletin, 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.

Tectonic-sedimentary evolution of the western margin of the Mesozoic Vardar Ocean: evidence from the Pelagonian and Almopias zones, northern Greece I A N R. S H A R P 1 & A L A S T A I R H. F. R O B E R T S O N 2

1Norsk Hydro Research Centre, Sandsliveien 90, Bergen, Norway, N-5020 (e-mail." ian. sharp@hydro, com) 2Grant Institute of Earth Science, School of GeoSciences, University of Edinburgh, Edinburgh EH9 3JW, UK The Vardar Zone documents the Mesozoic-Early Cenozoic evolution of several small oceanic basins and a complex history of terrane assembly. Following a Hercynian phase of deformation and granitic intrusion within the Pelagonian Zone to the west, the Vardar Zone rifted in Permian-Triassic time, with the creation of an oceanic basin (Almopias Ocean) during the Late Triassic-Early Jurassic. During the Mid-Jurassic, this ocean subducted northeastwards beneath the Paikon Zone and the Serbo-Macedonian Zone, giving rise to arc volcanism and back-arc rifting. A second ocean basin, the Pindos Ocean, opened to the west of a Pelagonian microcontinent, also during Late Triassic-Early Jurassic time. During the Mid-Late Jurassic, ophiolites were emplaced northeastwards (in present co-ordinates) from the Pindos Ocean onto the Pelagonian microcontinent, forming the Pelagonian ophiolitic m61ange within a flexural foredeep. This emplacement is dated at pre-Late Oxfordian-Early Kimmeridgian from the evidence of corals within neritic carbonates that depositionally overlie the emplaced ophiolitic rocks in several areas. Related greenschist- or amphibolite-facies metamorphism is attributed to deep burial following trench-margin collision and the attempted subduction of the Pelagonian continent. An inferred phase of NNW-SSE displacement, also of pre-latest Jurassic age, imparted a regionally persistent stretching lineation and related ductile fabric, apparently related to post-collisional strike-slip. The Pelagonian Zone and its emplaced ophiolitic rocks then underwent extensional exhumation during Late Jurassic-Early Cretaceous time. The western margin of the Vardar Zone experienced extensional (or transtensional) faulting, neritic carbonate and terrigenous clastic deposition, and intermediate-silicic magmatism during Late Jurassic-Early Cretaceous time. Oceanic crust (Meglenitsa Ophiolite) formed further east in the Vardar Zone during Late Jurassic-Early Cretaceous time, possibly above a subduction zone. A near-margin setting is suggested by the presence of a deep-water terrigenous cover, probably derived from the Paikon continental unit to the east. The Vardar Zone as a whole finally closed related to eastward subduction beneath Eurasia, culminating in collision with the Pelagonian microcontinent during latest Cretaceous-Eocene time, as recorded in foreland basin development, HP-LT metamorphism, ophiolite emplacement and large-scale westward thrusting. In contrast to models that suggest closure of the Vardar Ocean in the Mid-Late Jurassic, followed by reopening of a Cretaceous ocean, we believe that the Vardar Ocean remained partly open from Triassic to Late Cretaceous-Early Cenozoic time. Abstract:

Compared with the westerly, more 'external' tectonic units of the Balkans and Hellenides, the Vardar Zone (Fig. 1) has remained poorly understood despite its critical bearing on the tectonic development of Tethys in the Eastern Mediterranean region (e.g. Smith 1993, 2006; Robertson et al. 1996). Also, its relation to the Pindos (Sub-Pelagonian) Zone to the west is still controversial. The Vardar Zone can be traced eastwards for > 500 k m through Hungary, Croatia, Bosnia, Serbia, Macedonia (former Yugoslavia) and northern Greece until it runs into the northern

Aegean Sea, with only fragments exposed on land further south (Fig. 1). The Vardar Zone of Northern Greece, also known as the Axios Zone, is located between the Pelagonian Zone to the west and the Serbo-Macedonian Zone to the east (Figs 1 and 2b). The Vardar Zone is traditionally divided into a series of tectonostratigraphic zones. F r o m west to east these are the Almopias Zone, the Paikon Zone and the Peonais Zone (Mercier 1966; Fig. 2a). The Almopias Zone is further subdivided into the Western, Central and Eastern Almopias zones (Figs 2b and 3). The

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the EasternMediterranean Region. Geological Society, London, Special Publications, 260, 373-412. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

374

I.R. SHARP & A. H. F. ROBERTSON

Fig. 1. Simplified outline tectonic maps of the Balkan region showing the study area in NE Greece (modified from Robertson & Shallo 2000).

Central Almopias Zone is the most diverse and is further subdivided into several tectonostratigraphic units, each with a distinctive stratigraphic sequence that can be correlated to give an overview of the tectonostratigraphic evolution through time (Fig. 4). Here, we will mainly discuss the traverse shown in Figure 2b, which exposes all of the units of the Almopias Zone and the eastern part of the Pelagonian Zone. We will also take account of correlative units exposed in

the Voras Massif further north (Fig. 3) and relevant units exposed further south, especially in the Vermion Mountains (marked V in Fig. 1, see also Fig. 2a). In general, the Pelagonian Zone and the Western and Central Almopias zones formed parts of the western margin of the Vardar Ocean during the Mesozoic, whereas the Eastern Almopias Zone comprises more oceanic lithologies derived from the Vardar Ocean and has a fundamentally

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE

375

Fig. 2. (a) Outline tectonic map of the Vardar Zone; the small boxed area is shown in Figure 2b; the large boxed area is enlarged in Figure 3. (b) Tectonostratigraphic units of the Almopias Zone (modified from Mercier 1966). different geological history until suturing during Early Cenozoic time. In particular, the Pelagonian Zone and the Western and Central Almopias

zones experienced pre-Cretaceous deformation and metamorphism, which is not represented in the Eastern Almopias Zone. In this paper we use

376

I.R. SHARP & A. H. F. ROBERTSON

Fig. 3. Simplified geological map showing the main tectonostratigraphic units, lithologies and their ages in the Almopias Zone and the Voras Massif to the north, based mainly on mapping by Mercier (1966), Verg61y (1984), Mercier & Verg61y (1984a, b), Brown (1994) and Sharp (1994). Place names mentioned in the text are included.

W MARGIN OF MESOZOIC V A R D A R OCEAN, GREECE

377

,--k

O

O

O

O

< . ,,..~ O

. ,...~

. t:z0 ,...~

378

I.R. SHARP & A. H. F. ROBERTSON

the term Vardar Ocean for oceanic lithosphere that existed within the Vardar Zone, including the Peonais Zone in the east, whereas we use the term Almopias Ocean, more specifically, for oceanic crust that we interpret to have existed between the Pelagonian Zone and the Paikon Zone in the east (Fig. 2a).

Permian-Triassic rifting Pre-rift 'basement' units are exposed within the Pelagonian Zone, where they mainly comprise metasedimentary and recta-igneous rocks (e.g. amphibolites), intruded by granitic rocks of Carboniferous age (Mountrakis 1984). Similar 'basement' rocks are exposed in the Paikon Zone within the Voras Massif, near the border between Greece and Macedonia (former Yugoslavia) (Mercier 1966; Brown & Robertson 2004; Fig. 3). A pre-Mesozoic tectonic fabric survives in the core of the Pelagonian Zone (e.g. in the Vernon Mountains; Mountrakis 1984), but elsewhere the foliation and deformation fabric mainly reflect Late Jurassic orogenesis, with only a minor imprint from the Palaeogene suturing of the Vardar Ocean. The Pelagonian Zone shows evidence of Triassic rifting that could in principle be related to the opening of ocean basins to the west (Pindos Oceafi) or to the east (Vardar Ocean), or both. The western Pelagonian Zone (e.g. in the Vernon Mountains) shows evidence of rifting, as indicated by metaclastic and metavolcanic units that unconformably overlie the metamorphic basement (Mountrakis 1984). In the eastern part of the Pelagonian Zone, as exposed in the Voras Massif in the north (e.g. Kaimaktchalan Unit; see Brown & Robertson 2004; Fig. 3) metasedimentary rocks and amphibolites apparently record rifting to form the Almopias Ocean to the east, although these units remain poorly dated. Within the Voras Massif, a rift-related elastic-volcanic succession overlies metamorphic basement and passes gradationally upwards into a thick Mesozoic carbonate succession further east (Likostomo-Livadia Unit and the basal part of the Loutra Arideas Unit; Fig. 3; Brown & Robertson 2004). Volcanic rocks (Promachi amphibolites) within this sequence exhibit a mid-ocean ridge basalt (MORB) chemical composition (Brown & Robertson 2004). Similar amphibolites of mainly within-plate basalt (WPB) (Fig. 5a) occur within the lower part of the Klissochori Unit of the Central Almopias Zone (Figs 2b, 3 and 4) and may also relate to Triassic rifting, although stratigraphic constraints are poor. The occurrence of rift-related

volcanic rocks and sediments throughout the westerly Vardar units is consistent with the opening of an oceanic basin to the east, within the eastern Almopias Zone. Two important transverse faults, the Kato Loutraki Fault in the north and the Nission Fault in the south (Figs 2b and 3) are believed to have been active during Triassic rifting. Sequences adjacent to these faults are dominated by elastic sediments and appear to separate areas of contrasting palaeogeography from Triassic time onwards (Sharp 1994). These faults trend at a moderate to a high angle relative to the regional N W - S E trend of the Pelagonian and Almopias zones, and are interpreted as transfer faults that subdivided the rifted margin into segments, with contrasting depositional and tectonic histories. Faults parallel to the rifted margin are more difficult to recognize, probably because they developed into thrusts during later compressional deformation. During the Permian-Triassic a thick succession of elastic sediments, bimodal volcanic rocks and neritic carbonates developed along the western margin of the Serbo-Macedonian Zone (Dimitriadis & Asvesta 1993; Stais 1994; Fig. 1). The rift basalts range from WPB, to transitional, to MORB type (Dimitriadis & Asvesta 1993). The Paikon Zone is inferred to have rifted from the Serbo-Macedonian Zone to form a bordering microcontinent. The intervening deep-water basin is preserved within the Peonais Zone and is represented by the Svoula Flysch (Kaufmann et al. 1976; Kockel et al. 1977). A NE-SW facies change is observed, from continental facies within the Serbo-Macedonian continent, to marine units within the Peonais Zone during Triassic time (Stais 1994). The SerboMacedonian Zone was traditionally interpreted as a part of the Eurasian margin (Jacobshagen et al. 1978), but was recently reinterpreted as an exotic terrane that was amalgamated during the Jurassic deformation history (Himmerkus et al. 2006). The eastern margin of the combined Pelagonian-Almopias Zone therefore records rifting to form a subsiding passive margin of Late Triassic-Early Jurassic age, bordering the Almopias (western Vardar) Ocean to the east. The Serbo-Macedonian Zone records coeval rifting. Three possible options for the setting of rifting are: (1) formation within a pre-existing Palaeotethyan Ocean located within the Vardar Zone (Mountrakis 1984; Robertson & Dixon 1984; Karamata & Vujnovid 2000); (2) formation of a back-arc marginal basin related to subduction outwith the Vardar Zone, either southward subduction from a Palaeotethyan Ocean located

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE 9

100.00-

KA311A KA311B

$

KA315 KA444A

Ill

,~ 10.0o o e

379

i.oo

0.10 a

Sr K20 Rb

Ba

Nb

Ce

Nd P205 Zr

Ti

Y

Sc Cr

100.00 9

V1088A

9

V1088B ~ / ~ /

*

V1088C

V1088E V1092A

9

V1092B

\

10.00

1.00

0.10

b

,

K20

,

Rb

,

Ba

Nb

Ce

Nd

P205 Zr

Ti

Y

,

,

Sc

Cr

Fig. 5. (a) MORB-normalized 'spidergrams': (a) amphibolites from the Klissochori Unit (Central Almopias Zone); (b) basalts from the Late Triassic Vryssi Unit, easternmost Central Almopias Zone. Both the amphibolites and basalts are geochemically'enriched' and are attributed to Triassic rifting of the Vardar Ocean. (See Table 1 for representative analyses.)

to the north ($eng6r 1984), or northward subduction related to subduction in the South Aegean region (Stampfli et al. 2001, 2003; Stampfli & Bore12002); (3) formation of a rifted small ocean bordering the Serbo-Macedonian Zone (Stais & Ferri6re 1991; Dimitriadis & Asvesta 1993; Stais 1994; Brown & Robertson 2003, 2004). Option (1) (Palaeotethys within Vardar) now seems unlikely, as there is no obvious evidence of any Palaeotethyan units actually within the Vardar Zone. Option (2), southward subduction from the north, has been tested based on studies of the Pontides in northern Turkey (Usta6mer & Robertson 1997; Usta6mer et al. 2005; see also Okay et al. 2001) and has been found to be invalid. Option (3), northward subduction from the south, has also now been tested based on studies of the South Aegean region and Crete, and is also now known to be invalid (Robertson 2006b; see also Smith 2006). The igneous and

sedimentary evidence from the Vardar Zone as a whole is explicable in terms of the formation of a Triassic small ocean basin. In this interpretation the Pelagonian Zone formed part of Gondwana in Late Palaeozoic time, but later rifted away opening up a small ocean basin bordering both the western (Pindos) and eastern (AlmopiasVardar) margins of a Pelagonian microcontinent. This is comparable with the inferred Triassic rifting of the Tauride-Anatolide microcontinent from Gondwana further east in southern Turkey (Robertson et al. 2004). This interpretation is also consistent with the Serbo-Macedonian Zone being an exotic terrane that was amalgamated to Eurasia only during its Alpine history (Himmerkus et al. 2006). It is possible that a fundamental Palaeotethyan suture is located within units related to the Serbo-Macedonian and Rhodope zones and that all units to the south of this suture are Gondwana-derived.

380

I.R. SHARP & A. H. F. ROBERTSON

Table 1. Major and trace element X R F geochemical analyses of igneous lithologies. Major elements are in weightpercent oxide and trace elements in ppm.

SiO2 A1203 Fe203 MgO CaO Na20

K20 Ti02 MnO

P205 LOI Total Ni Cr V Sc Cu Zn Sr Rb Zr Nb Ba Pb Th La Ce Nd Y

Klissochori 1 KA311B

Klissochori 2 444A

Vryssi 3 Vio88A

M61ange 4 1142C

48.6 16.13 9.09 9.57 7.87 1.57 3.22 0.81 0.15 0.04 2.52 99.59

46.91 15.68 12.08 6.52 9.84 2.51 1.16 2.01 0.15 0.22 2.39 99.49

50.71 17.73 9.01 6.75 2.14 5.51 0.52 1.64 0.44 0.16 4.84 99.37

44.28 14.01 9.44 9.06 17.94 0.22 0.95 1.26 0.21 0.16 3.23 100.76

47.33 14.46 12.43 9.82 7.91 2.61 1.29 1.49 0.19 0.18 2.34 100.05

49.8 9.38 8.01 13.11 16.76 0.55 0.25 0.53 0.19 0.05 1.98 100.65

52.48 14.24 11.97 6.62 5.53 4.48 0.14 1.58 0.17 0.18 3.4 100.48

52.55 14.87 10.66 3.35 5.87 2.39 0.55 0.84 0.18 0.07 9.06 100.39

216 469 267 39 34 87 22 29 89 10 453 3 1 0 19 9 25

85 249 250 40 103 87 179 26 126 7 165 0.6 0.1 1.4 22 17 31

188 679 320 44 18 57 180 4 19 2 165 1 0 1 10 10 14

38 59 376 41 38 137 141 6 155 7 18 1 3 5 16 15 52

1 20 386 36 8 91 115 21 60 3 77 2 3 6 15 9 18

166 501 229 21 30 73 35 95 41 2 364 2.7 b.d. 3 9 11 18

187 465 458 46 47 120 158 27 156 16 132 3 1 0.2 19 19 41

197 250 339 49 81 84 121 22 105 6 91 0 2 5 11 10 32

M61ange 5 12a

M 6 1 a n g e Krania 6 7 11 K1042F

Krania 8 K944B

LOI, loss on ignition; b.d., below detection limit. T r i a s s i c - J u r a s s i c : passive m a r g i n s u b s i d e n c e and o c e a n g en e s is Mid-Triassic to Early Jurassic time was characterized by the formation of oceanic crust within the Vardar Zone. This oceanic crust has since been mainly subducted, but was located within the Eastern Almopias Zone (i.e. Vryssi Unit of Stais et al. 1990; Fig. 2b). Bordering continental units are represented by the Pelagonian Zone to the west and by the Serbo-Macedonian Zone plus the Paikon Zone to the east. However, it should be noted that significant strike-slip may have occurred such that these units did not necessarily face towards each other in the Early Mesozoic, as today. The Pelagonian Zone and the Western and Central Almopias zones were characterized by subsiding carbonate platforms during Late Triassic-Early Jurassic time. Thick supra-, inter- and sub-tidal limestone-dolomite loferite cycles accumulated within the Pelagonian Zone

(e.g. Kaimakchalan Massif; Fig. 3) beginning in Mid-Late Triassic time. Despite deformation and regional greenschist-facies metamorphism, primary facies are locally well preserved. The supratidal facies typically are red marly limestones, whereas the intertidal facies are dolomite and limestone with well-developed laminar algal stromatolites (with fenestral fabrics) and sheetprism shrinkage cracks, whereas the subtidal facies are typically dolomitic with Dasycladacean algae (e.g. Griphoporella curvata) and rare Megalodonts. Evidence from the Klissochori Unit (e.g. the Rhizarion marbles) in the Central Almopias Zone (Figs 2b and 4) and from the Voras Massif further north (Loutra Arideas Unit; Migiros & Galeos 1990; Brown & Robertson 2004; Fig. 3) indicates the presence of an eastward-deepening, mixed carbonate-clastic slope sequence. Triassic?-Jurassic algal and fenestral (loferitic) marbles are also seen in the Livadia Unit of

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE the Voras Massif. The Livadia Unit, together with the 'Rhizarion marbles' further south, formed an isolated area of platform sedimentation. This unit was apparently rifted from the main Pelagonian carbonate platform to the west to form a marginal fault block bounded by a rift basin, represented by the Loutra Arideas Unit of the Voras Massif (Fig. 3). Redeposited and hemipelagic deep-water basinal carbonates and subordinate elastic deposits are also present within the Pelagonian Zone, as represented by the undated Kato Grammatiko Formation of Sharp (1994), and can be interpreted as deposits within an intra-platform basin, located within the Pelagonian platform or along its eastern margin. Comparable intra-platform basins are recognized in the eastern Pelagonian Zone elsewhere in Greece, notably in the Argolis Peninsula (Clift & Robertson 1990a). Triassic oceanic crust is preserved as tiny slices ( < 5 m thick) of MOR-type pillow basalts within the Vryssi Unit in the most easterly part of the Central Almopias Zone, just beneath the basal thrust of the Eastern Almopias Zone (to the east of the Nea Zoi Unit; Figs 2 and 5b). The basalts are depositionally overlain by several metres of ribbon radiolarite of Late Triassic age (Stais et al. 1990). Within the Paikon Zone to the east a metamorphosed Triassic?-Jurassic, mixed carbonateelastic sequence (Gandatch Formation) is interpreted as a deep-water equivalent of elastic sediments within the Peonias Zone (Brown & Robertson 2003). Further east again, successions exposed along the SW margin of the SerboMacedonian Zone and the eastern adjacent Peonais Zone document a subsiding SW-facing carbonate platform during Anisian-Carnian time (Stais & Ferri6re 1991; Dimitriadis & Asvesta 1993; Stais 1994). These platform carbonates are overlain by westward-deepening slope to basinal, mixed carbonate-elastic facies (Stais 1994). Associated volcanic rocks are of WPB to MORB type (Dimitriadis & Asvesta 1993). The sequence extends into the Early Jurassic as thick siliciclastic turbidites (Svoula Flysch; Kaufmann et al. 1976; Kockel et al. 1977). During Late Triassic-Early Jurassic time the floor of the Peonais basin between the Paikon and Serbo-Macedonian continental units is likely to have been represented by stretched continental crust. In summary, the western margin of the Vardar Zone and the adjacent eastern Pelagonian Zone document Triassic rifting, then Jurassic passive margin subsidence related to the opening of a MOR-type oceanic basin to the east (i.e. the Almopias Ocean).

381

Early-Mid-Jurassic: eastward subduction of the Almopias Ocean It is widely believed that the Almopias Ocean was subducted northeastwards beneath the SerboMacedonian margin during Early-Mid-Jurassic time (Verg61y 1984; B6bien et al. 1986, 1987; Brown & Robertson 1994, 2003, 2004). The evidence for this is seen in the more easterly Vardar zones (i.e. the Paikon, Peonais and SerboMacedonian zones), outside the present study area (Fig. 1). This inferred subduction resulted in arc volcanism within the leading edge of the Serbo-Macedonian margin, represented by Paikon Zone (B6bien et al. 1980, 1994; De Wet et al. 1989; Brown & Robertson 1994, 2003, 2004). Contrary to recent suggestions of an oceanic arc origin (Stampfli et al. 2001), these volcanic rocks are seen to overlie continental crust within the Voras Massif (Mercier 1968; Brown & Robertson 2004; Figs 2a and 3). Backarc extension is believed to have reactivated the inferred Peonais rift basin to form an intracontinental back-arc basin in which the Late Jurassic Guevgueli Ophiolite formed (B6bien e t a l . 1987; Mussalam 1991; Danelian et al. 1996; Brown & Robertson 2003, 2004). The Guevgueli Ophiolite retains primary intrusive contacts with adjacent metamorphic rocks, correlated with the Serbo-Macedonian Zone (De Wet et al. 1989; see Smith 1993). This suggests that this ophiolite is para-autochthonous with respect to the SerboMacedonian continental margin to the east.

Mid-Jurassic Pelagonian platform-margin collapse Mid-Late Jurassic time (pre-Oxfordian-Early Kimmeridgian) was characterized by the collapse of the Pelagonian-Almopias carbonate platform and a transition to deeper-water hemipelagic sediments. The top of the Pelagonian platform succession (e.g. at Arnissa; Figs 3, 4 and 6) is gradationally overlain by a sequence (c. 20 m thick) of interbedded siliceous, micaceous and chloritic schists, ribbon cherts, cherty hemipelagic carbonates and siliciclastic sandstones (Arnissa Passage Beds Member of Sharp 1994). A less obvious transition is observed where basinal sediments previously existed (i.e. Kato Gramatiko Formation). Sedimentary structures in the upper part of the Arnissa Passage Beds Member are indicative of deposition by turbidity currents. Petrographic observations reveal an incoming of volcanic quartz, devitrified volcanic rocks and also of detrital chromite near the top of the sequence. X-ray diffraction

382

I . R . SHARP & A. H. F. R O B E R T S O N

rn . .,...~

=,< o q.--, 9"'~ O +..a ~ ,.,,, o

o

,'--..

~-= ,.....,

<.~

9, ~

r.~

O,.~

9- ~

-,--;

O

o

O

r.~ . ,.,,,

9"

I=I

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE analysis of fine-grained sediments revealed an initial dominance of muscovite, chlorite (ripidolite and clinochore), quartz, albite, minor sphene and epidote, followed by an incoming of chromite and talc (serpentinite). This sequence is interpreted to record flexural subsidence and collapse of the Pelagonian platform ahead of emplacing ophiolitic units (Sharp et al. 1991). Similar sequences that can be related to platform collapse are known from the Pelagonian Zone elsewhere in Greece, including Vourinos (Zimmerman 1972; Naylor & Harle 1976; Verg61y 1984), Othris (Smith et al. 1975; Ferri6re 1976), Argolis (Baumgartner 1985; Clift 1992) and Evia (Euboea) (Robertson 1991).

Mid-Late Jurassic emplacement of ophiolitic debris flows and m~lange Within the eastern Pelagonian Zone (e.g. Arnissa area; Fig. 4), the platform-collapse sequence passes gradationally upwards into a largely ophiolite-derived, chaotic sedimentary sequence, the Brown Schist Member (Bijon 1982; Sharp 1994; Figs 6 and 7). This is overlain by ophiolitic m61ange and by dismembered ophiolitic thrust sheets (Mercier & Verg61y 1972; Bijon 1982). The lower part of the succession includes sandstones, mudstones and conglomerates of mainly epiclastic (detrital), basic igneous and subordinate siliciclastic origin. The higher levels are dominated by ophiolitic lithologies (e.g. serpentinite), mainly debris flows and turbidites within an argillaceous sequence, rich in talc, chromite and other ophiolite-derived material. Rare carbonate debris flows (e.g. Mavro Rema section; Fig. 8) associated with blocks of marble (commonly internally brecciated), are interpreted as talus shed from the collapsing Pelagonian carbonate platform. The debris flows as a whole are interpreted as the infill of a foredeep related to downflexure of Pelagonian continental crust ahead of an emplacing ophiolite. Similar collapse and related foredeep sequences are documented elsewhere in the Pelagonian Zone further south (e.g. in Othris, Evia and Argolis: e.g. Robertson et al. 1991). Similar features are also associated with Late Cretaceous Turkish ophiolites (Parlak & Robertson 2004), northern Syrian ophiolites (A1-Riyami & Robertson 2002), and with many other Tethyan settings, including Oman (Lippard et al. 1986; Robertson 1987, 2006a). Upwards, the chaotic foredeep-type debris is covered by an ophiolitic m61ange. This is made up of blocks or sheets of serpentinite ultramafic rocks (mainly harzburgite and dunite), with subordinate gabbro, basalt, diabase, amphibolite and recrystallized red ribbon chert, set within a

383

strongly sheared, scaly serpentine-rich matrix. Good examples of this mdlange are exposed in the Arnissa area and in the Vermion Mountains to the south (Figs 3 and 8). In places (e.g. Arnissa and Reis Tsifliki areas), the m61ange is in direct tectonic contact with large masses of ultramafic rocks (hundreds of metres to several kilometres in size), although the strong Palaeogene deformation makes it difficult to determine original emplacement relationships. The mdlange is locally absent (e.g. Mavri Rakhi region of the Vermion Mountains; Fig. 8), where the Pelagonian carbonate platform is directly overlain by a large slice of serpentinized ultramafic rock, with amphibolite welded to its base. This amphibolite is interpreted as the remnant of a metamorphic sole. Geochemical studies show that the blocks of extrusive rocks within the m61ange range in composition from alkali basalts with WPB signatures (containing titanaugite), to subalkaline tholeiitic basalts with MORB affinities, to islandarc tholeiite (IAT)-type basalts; also, two samples exhibit boninitic affinities (Fig. 9a). This chemical variation suggests that the m~lange basalts were derived from several different tectonic settings and were probably mixed within a subduction complex before being emplaced within the m61ange. The IAT-type basalts and boninitic lavas are similar to those reported from the Vourinos Ophiolite (e.g. Bortolotti et al. 2004; Rassios & Moores 2006), whereas the alkali basalts might record either rift or accreted seamount basalts, and the MORB, remnants of Triassic-Early Jurassic oceanic crust. The predominance of harzburgite in the m61ange blocks and thrust sheets is suggestive of a suprasubduction origin, in common with the Vourinos Ophiolite and many of the other Jurassic Balkan ophiolites (Pearce et al. 1984; Robertson et al. 1991; Rassios et al. 1994; Clift & Dixon 1998; Rassios & Smith 2000). The mdlange includes blocks and dismembered thrust sheets that can be closely compared with ophiolitic m61ange exposed in the Voras Massif further NE (Fig. 3). A mixed carbonatesiliciclastic sequence, of Triassic-Jurassic age, is gradationally overlain by ophiolite-derived turbidites and debris-flow deposits (Loutra Aridea Unit; Migiros & Galeos 1990; Brown & Robertson 2004). In addition, ophiolitic rocks overlie metacarbonate platform rocks within the Western and Central Almopias zones (Fig. 4); these are mainly thin (<200 m) units of harzburgite, dunite, serpentinite and amphibolite (see below). The lithologies within the ophiolitic m61ange in the area studied are closely comparable with

384

I.R. SHARP & A. H. F. ROBERTSON

Q

+,., rt~

E

LU

"r"0

m

r r-

.E/

E

E

~o "e-

.~E -~| n~

Z ~ ~ Q . m

Processes of ophiolite emplacement

E~ ~"0

~~

0 0

E 0

z

the Vourinos Ophiolite and the accretionary Avdella M61ange beneath the Pindos Ophiolite further west (Kostopolous 1989; Jones & Robertson 1991; Fig. 1). The Vourinos Ophiolite and its locally preserved amphibolitic sole (Pichon & Brunn 1985; see Rassios et al. 1994 Rassios & Moores 2006) forms part of a regionally extensive thrust sheet above the Vourinos accretionary m61ange. The m61ange, as exposed in the eastern Pelagonian Zone (e.g. Arnissa area), is a mixture of chaotic m61ange and ophiolitic thrust sheets. A very similar mixture is, for example, found within the m61ange that lies beneath the Late Cretaceous Mersin ophiolite in southern Turkey (Parlak & Robertson 2004). Also, in northern Syria, the Late Cretaceous Bafir-Bassit ophiolite is pervasively imbricated with accretionary material (A1-Riyami & Robertson 2002). In both of these areas this type of m61ange and disrupted ophiolite thrust sheets developed in the frontal zones of an emplacing ophiolite, and a similar setting is inferred for the Pelagonian ophiolitic m61ange. Palaeogene deformation of the Pelagonian and Almopias zones has resulted in thrust imbrication and further disruption of the m61ange, with the serpentinite-rich intervals acting as d6collement horizons (Figs 7 and 10).

O Q

One view is that all of the ophiolites were emplaced westwards from the Vardar Zone, (Mercier et al. 1975); another is that ophiolites were emplaced onto the Pelagonian Zone, towards the NE from a Pindos Ocean (Smith et al. 1975; Robertson et al. 1991; Doutsos et al. 1993), and a further option is that ophiolites were emplaced from both the Pindos Zone (e.g. Vourinos-Pindos) and from the Vardar Zone ('Vardar ophiolites') (Mountrakis et al. 1987). There are two main lines of evidence that could help to discriminate between these alternatives in the area studied: (1) structural data from the Pelagonian platform and the overlying ophiolitic m61ange; (2) evidence from the age of the transgressive sedimentary cover. The effects of Early Cenozoic westward thrusting must also be considered. During this work it was found that the entire Pelagonian platform and counterparts Fig. 7. Schematic cross-section through the uppermost levels of the Pelagonian carbonate platform in the Arnissa area showing the transition to ophiolitic m61ange. The section was taken east of log 3 (Fig. 6). The thrust imbrication is mainly of Early Cenozoic age. Key as in Figure 6. (See text for explanation.)

W M A R G I N OF MESOZOIC V A R D A R OCEAN, G R E E C E

385

Fig. 8. Outline geological map of the Western Vermion Mountains. The locations of logged ophiolitic m61ange in the Western Vermion Mountains (Fig. 10) and at Mavri Rakhi (Fig. 12) are shown. The overall structure relates to westward Palaeogene thrusting. The general location is shown in Figures 1 and 2a. Modified from IGME Arnissa and Pirgi sheets.

386

I.R. SHARP & A. H. F. ROBERTSON

100.00 1112a

=

~

o ,o.oo ~ _

.Jr

1142b

\\

\

--~4~_~

1.00

0.10 ,,

. . . . . . Sr K20 Rb Ba Nb Ce

. . . . . Nd P205 Zr Ti

. Y

,oooo1

! _:-~

/

I --~-

"~176176 ] J~'
12b

I o 20a

9

b

Sc Cr

_

,ooooo]

Sr K20

Rb

Ba

Nb

Ce

Nd P205 Zr

Ti

Y

Sc Cr

. . . . Nd P205 Zr

. Ti

. Y

Sc Cr

~176176176176 1 \ ~

_~ ,ooo/ / 3 ~ ~ ,9 ~

1.oo

0.10 e

. . . . . . Sr K20 Rb Ba Nb Ce

Fig. 9. MORB-normalized 'spidergrams' of lava and amphibolites from the Pelagonian ophiolitic m61ange. ~a) Basalts from the Arnissa area; ~ ) basalts from the Reis Tsifliki (Vermion) area; (c) amphibolites from Reis Tsifliki area. The 'enriched' composition should be noted. (See Table 1 for representative analyses.)

W MARGIN OF MESOZOIC V A R D A R OCEAN, GREECE

387

"~", . ~

rT ~

.~==~ m~o

e .~ .!

"~~ ~-~ . ,,._~

~ o m r~

-,~ ="S N 0

o

~.~

0"~

~ETg

~~

~'~ ~<

O~

~

,.~ . ~ ~

,,.~, ,.~

..~

9

~

~..,~

0

c~

o

~

~ ,-~ ~.~ "~

~ . ~

~=s

;~

o~

~

~

o

388

I.R. SHARP & A. H. F. ROBERTSON

within the Western and Central Almopias zones, up to and including the ophiolitic m61ange have experienced greenschist-facies, to possibly amphibolite-facies metamorphism, indicating that deep burial of the Pelagonian platform has taken place throughout this area. This was associated with ductile-style folding and the development of a penetrative deformation fabric (DO (Mountrakis et al. 1984; Kockel 1986; Mercier 1966). Flat-lying to slightly inclined isoclinal to similar folds are thickened in the hinge zones and thinned or sheared-out parallel to the foliation on the limbs. These folds are associated with the development of an axial planar cleavage, which is typically oriented parallel to bedding, and is also related to the formation of a pronounced stretching lineation that is oriented parallel to the B tectonic axis of folds. The trend of this lineation is remarkably constant throughout the study area with an average orientation of N150 ~ (range N 180~ ~ (Verg61y 1984; Sharp 1994). The vergence of the observed folds is variable, with both SW-WSW and NE-ENE fold vergence commonly being observed. In a small number of exposures in the Arnissa area C-S fabrics are locally indicative of top-to-the-SE motion (Sharp 1994). Previous structural studies suggested that the easterly 'internal' zones of northern Greece as a whole, including the Pelagonian, Almopias and Paikon zones, experienced two major phases of deformation and metamorphism; i.e. JE-1 in the Late Jurassic (155-145 Ma) and JE-2 in the Early Cretaceous (130-110 Ma) (Mercier 1966, 1973; Verg61y 1984). We confirm the presence (but not the vergence) of a JE-l-type event affecting the Pelagonian Zone and the Western and Central Almopias zones. Verg61y (1984) used mainly fold vergence and cross-cutting relations to infer an initial ophiolite emplacement (JE-1) towards the west that was considered to be Late Jurassic in age; this was followed by a second ophiolite emplacement event, also verging towards to the west (JE-2) that was marked by cross-cutting folds and interpreted as being Early Cretaceous in age. During this study we observed that the fold hinges of similar (ductile) folds are commonly oriented parallel to the stretching direction and thus cannot be used as kinematic indicators. Opposing vergence directions were observed more or less randomly even in local outcrops and cannot be interpreted as successive fold phases. In addition, some of the supposedly JE-2 folds deform Cretaceous facies including Albian-Cenomanian transgressive limestones and conglomerates, showing that, within the

regional context, these folds must be postCretaceous in age (i.e. related to Cenozoic deformation). For example, many of the large SW-verging folds in the region (e.g. Messovounon and Ayios Dimitrios area; Figs 2 and 8) involve Late Jurassic and Cretaceous (AlbianCenomanian)-aged sedimentary cover rocks that post-date ophiolite emplacement (Sharp 1994). This contrasts with earlier studies that attributed these particular structures to a regional Late Jurassic westward ophiolite emplacement (e.g. Braud et al. 1984; Verg~ly 1984). Kilias (1991) described a similar NNW-SSE stretching lineation but without a preferential vergence from several localities in the ArnissaEdessa area. Further west, near the Vourinos Ophiolite both SSW and NE (DO vergences were recorded. Kilias (1991) also reported data from quartz c-axis measurements. From the limited data available for this region the positions of the c-axis maxima on cross girdles (with reasonably well-defined outlines) are indicative of sinistral displacement, consistent with an easterly component of tectonic transport. In addition, structural studies of the Vourinos Ophiolite, adjacent small ophiolitic bodies and the underlying Pelagonian platform provide evidence of NE-directed displacement (Naylor & Harle 1976; Rassios et al. 1994; Rassios & Moores 2006). Rassios (pers. com.) noted that structures in the Pelagonian Zone south of the Vourinos Ophiolite (e.g. in the Triassic section south of the Aliakmon River and continuing into the Aliakmon area) exhibit kinematic indicators that indicate movement to the NE. NE-directed emplacement is also inferred for the Avdella M61ange beneath the Pindos Ophiolite further west (Jones & Robertson 1991; Rassios & Moores 2006). Structural evidence also supports generally east-directed emplacement of ophiolites onto the Pelagonian Zone, including Othris (Smith et al. 1979), Evia (Robertson 1991) and Argolis (Clift & Robertson 1990a, b), and also from Albania (e.g. Robertson & Shallo 2000) and former Yugoslavia (see Robertson & Karamata 1994; Karamata 2006). The structural results for the region studied (e.g. Edessa-Arnissa) can also be compared with information from the region further south (south of the Aliakmon River) including the High Pieria, Olympos, Ossa and NW Thessaly area (Fig. 1). A pervasive stretching lineation with a variable trend (NW-SE or NNW-SSE) was reported from the Pelagonian Zone in NW Thessaly (Sfeikos et al. 1991; Sfeikos 1992), High Pieria (Yarwood & Aftalion 1976), Livadi (Nance 1981) and Olympos (Barton 1976; Kilias et al. 1990; Schermer et al. 1990; Doutsos et al. 1993; Schermer 1993; Fig. 1). Based on a kinematic

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE study of the region as a whole, Wallcott (1996) and Wallcott & White (1998) observed that a SSE-directed stretching lineation (DO is well developed in NW Thessaly (as above) but becomes weaker and disappears southwards. Kilias et al. (1990) reported quartz c-axis data (i.e. maxima on crossed girdles) that indicate sinistral (eastward) transport for several localities in the NW Pieria mountain area, whereas the results from areas further south are indicative of dextral (west-directed) tectonic transport. In general, the metamorphic grade (M0 ranges from amphibolite facies at high structural levels to amphibolite or upper greenschist facies at lower structural levels (Schermer et al. 1990; Walcott 1996). Following the D1 event, Walcott (1996) identified a weak east-west-trending fabric (D2) in some areas of Thessaly that shows evidence of east-directed shearing; she related this either to continued thrusting or to an Early Cretaceous extensional event. The subsequent fabric (D3) is indicative of top-to-the-SW tectonic transport, of inferred mid-Cretaceous to Early Cenozoic age according to most workers (e.g. Schermer 1993; Walcott 1996; Walcott & White 1998), punctuated by a phase oftop-to-the-NE shearing (Lips et al. 1998). D3 is associated with regional HP-LT metamorphism (M2) (Schermer et al. 1990; Schermer 1993) and is attributed to generally eastward subduction beneath the Pelagonian continent.

Late Jurassic transgressive deposition The deformed Eastern Pelagonian and Western Almopias zones are typically transgressed by cover sediments of mid-Cretaceous (AlbianCenomanian) age (Mercier 1966; Sharp 1994) thus providing only limited constraints on the timing of the D1 deformation and M~ metamorphism. However, important fossil evidence from the Almopias Zone shows that transgression there began in the Late Jurassic. Specifically, in the Kerassia and Kedronas units of the Westem Almopias Zone (e.g. Nission and Kedronas area; Figs 2b and 3), the basal sediments, which lie unconformably on ophiolitic units in some places, include reef limestone with corals of Late Oxfordian-Early Kimmeridgian age. Coralline fauna present (Fig. 11) include Stylosmilia cf. miehelini, Thecosmilia cf. langi, Cladocoropsis mirabilis, Dermosmilia sp. and Schizosmilia cf. rollieri. Similarly, the ophiolitic m61ange of the Loutra Arideas Unit further NE in the Voras Massif (Fig. 3) is unconformably overlain by shallow-water limestones, containing corals and Cladocoropsis sp. of Late Jurassic

389

(Kimmeridgian?) age (Galeos et al. 1984; Brown & Robertson 2004). These relationships show that ophiolitic rocks were emplaced and eroded subaerially prior to, or during, Oxfordian time. Assuming that the underlying deformed ophiolites, represent part of the regionally emplaced Jurassic ophiolites these results constrain the emplacement of the Pelagonian ophiolitic m6lange and the D~ deformation as pre- to syn-Late Oxfordian-Early Kimmeridgian. Within the Eastern Pelagonian Zone (near Mavri Rakhi; Fig. 8), the ophiolitic m61ange is depositionally overlain by low-grade metamorphosed deep-water carbonates, known as the Mavri Rakhi Formation (Pichon 1976, 1977; Fig. 8). This unit begins with interbedded green arenites, cherts and mudstones (c. 20 m thick), passing gradationally upwards into siliceous carbonates (c. 30 m thick) (Fig. 12). Although lacking age-definitive fossils, calcispheres, aptychi, bivalves and possible calpionellids are present, suggesting a Late Jurassic-Early Cretaceous age. This unit is unconformably overlain, with a minor discordance, by AlbianCenomanian limestones. The facies and stratigraphic position of the Mavri Rakhi Formation are very similar to the well-dated Late JurassicEarly Cretaceous pelagic carbonates, which depositionally overlie the Vourinos Ophiolite (see Rassios & Moores 2006) and the Eastern Albanian ophiolites (e.g. Robertson & Shallo 2000). The Vourinos Ophiolite (e.g. Siatista, Krapa and Zygosti areas) exhibits a Late Jurassic-Early Cretaceous transgressive cover of slope to deep-water carbonate sediments that sit, with a minor angular discordance, on the ophiolitic extrusive rocks (Pichon & Lys 1976; Mavrides et al. 1979). A similar age and setting were inferred for the equivalent, Eastern-type Mirdita ophiolites in Albania (see Robertson & Shallo 2000).

Implications of structural and age information The results from the area studied (Fig. 1) indicate that the pervasive D~ fabric, characterized by the pervasive SSE-NNW stretching lineation, is of pre- to syn-Late Oxfordian-Early Kimmeridgian age. By contrast, in NW Thessaly, south of the Aliakmon River, which marks an important transverse discontinuity, radiometric dating of the early deformation fabric has yielded Early Cretaceous ages (130-100 Ma), i. e. 119__ 15 Ma from granites in the Pieria region; 101 __13 Ma from augen schist from the Infrapierian unit (Yarwood & Dixon 1977); 124__4Ma and 123+_11 Ma from mylonitic Pierian granites

390

I.R. SHARP & A. H. F. ROBERTSON

Fig. 11. Fauna from reefal limestone developed within the Basal Kerassia Unit, Western Almopias Zone, Nission area. (a) Dermosmilia sp. (Late Oxfordian-Early Kimmeridgian); (b) Thecosmilia sp. (Late OxfordianEarly Kimmeridgian); (e) Schizosmilia cf. rollieri (Oxfordian-Kimmeridgian); (d, e) Stylosmilia cf. michelini (Late Oxfordian-Kimmeridgian); (f) Neotrocholina or Trocholina sp. (Late Oxfordian-Early Kimmeridgian); (g) Cladocoropsis mirabilis (lagoonal hydrozoan) (Oxfordian-Early Kimmeridgian); (h) spinose nerinid gastropod with geopetal structure infilled by faecal pellets and micrite at the base and sparite at the top. The samples were identified and dated by B. Rosen, British Museum of Natural History, London.

south of Olympos (Barton 1976) and 98_+2 Ma from the Intrapierian unit (Schermer et al. 1990). Assuming that the D, fabrics are contemporaneous north and south of the Aliakmon River, as is likely (but not proven), it is possible that the age of the D1 deformation in N W Thessaly (e.g. nothern Pieria) was also pre- to syn-Late Oxfordian-Early Kimmeridgian (c. 155Ma); if so, the Early Cretaceous radiometric dates could represent cooling ages. It is notable that the geological evidence from the Almopias Zone in the area studied (Western Vardar Zone margin) is indicative of an important Late Jurassic-Early Cretaceous extensional (or transtensional) phase (see below); this extension could have exhumed the Pelagonian basement and set the Early Cretaceous radiometric ages recorded from the Pelagonian Zone. We are unable to confirm a 'JE-2' Early Cretaceous regional compressional event affecting the area. The Late Jurassic base of the transgressive cover of the Western Almopias Zone passes upwards into a Cretaceous succession

without any intervening contractional or metamorphic event. Also, there is no evidence of the detritus expected if thick ophiolitic or other thrust sheets were emplaced westwards from the Vardar Zone onto the Pelagonian Zone during the Early Cretaceous. In addition, ophioliterelated units are absent from the Paikon Massif within the Vardar Zone. The Paikon Massif experienced ductile deformation and, according to Baroz et al. (1987), HP-LT metamorphism, associated with an early penetrative structural fabric (Verg61y 1984; Brown & Robertson 1994). This fabric is sealed by Kimmeridgian limestones (Kromni Limestones; Brown & Robertson 2003), which pass upwards into Late Jurassic-Early Cretaceous clastic sediments (Ghrammos Formation) and then into platform carbonates (Cretaceous Transgressive Limestones) without any intervening Early Cretaceous JE-2 type event. Further east, the inferred back-arc basin, represented by the Guevgueli Ophiolite in the Peonais Zone, was uplifted and unconformably overlain by Late Jurassic-Early Cretaceous

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE Mavri Rakhi

391

ophiolites, as seen in the High Pieria area (Kilias et al. 1990), are dismembered and thrust over

_TF .ittoral

udites + udist biostromes \lbian-Cenomanian 3M mxite + flaeosol ~conformity .=mipelagic ~rbonates RF

?Late JurassicEarly Cretaceous Mavri Rakhi Fm

)bon cherts, ~.nites, mudstones ~alcturbidites

nsgressive ttact )PH

)PH

(MF

(GF

<MF

Fig. 12. Sketch section of the ? Late Jurassic-Early Cretaceous Mavri Rakhi Formation in the Western Vermion Mountains (see Fig. 8 for location). KMF, Kaimakchalan Marble Fm; KGF, Kato Gramatiko Fro; MRF, Mavri Rakhi Fm; OPH, ophiolite; RBM, Red Bed Member; LTF, Lower Transgressive Fm. sediments (Mercier 1966; B6bien et al. 1987; Stais 1994) again without evidence of Early Cretaceous metamorphism. A number of small meta-ophiolites are distributed throughout north-central Greece, including the High Pieria and Olympos areas (see Pe-Piper & Piper 2002). Schermer (1993) noted that with the exception of the Livadi ophiolitic complex these lack a well-defined D~-type fabric. The

Pelagonian rocks. As they lack the well-defined D~ deformation it is likely that they were emplaced, deformed and metamorphosed postD1. Early to mid-Cretaceous westward emplacement (Jacobshagen et al. 1978) is unlikely, as extension (or transtension) characterized the eastern Vardar Zone during this time (see below). Final emplacement of these ophiolites from the Vardar Zone during Early Cenozoic time is preferred. The presence of a pelagic sediment cover, of probable Late Jurassic-Early Cretaceous age, in the western part of the Pelagonian Zone (Mavri Rakhi Formation) suggests that an implied thick overburden of the preserved metaophiolitic m61ange (more than several kilometres of ophiolite?) was removed after its emplacement. As the transgressive sediments are relatively deep-marine it is unlikely that the structural overburden was removed simply by erosion, as there is no evidence of a non-marine basal conglomerate; it is instead more likely to have been removed by extensional (detachment) faulting, allowing exhumed ophiolitic m61ange to be directly transgressed by relatively deepmarine carbonates. In addition, debris flows ('olistostromes') that are present within the Late Jurassic-Early Cretaceous aged pelagic carbonate succession overlying the Albanian (Mirdita) ophiolites may also relate to exhumation. These debris flows overlie the radiolarian chert cover of the Mirdita ophiolites. They include blocks of ophiolitic, continental margin and basement rocks. This unit is constrained as later than the Mid-Callovian to Early Oxfordian age of the radiolarian cover sediments but earlier than the Late Tithonian-Late Valanginian age of overlying calpionellid limestones (Bortolotti et al. 1996). As a result of the exhumation, material beneath the ophiolite could have been exposed on the seafloor and reworked oceanwards, giving rise to the observed multiple debris flows of Late Jurassic-Early Cretaceous age. This is an alternative to the genesis of these debris flows during the initial ophiolite emplacement onto the Korabi (Pelagonian) continental margin, as suggested by Robertson & Shallo (2000).

Mode of ophiolite emplacement, related deformation and exhumation Taking account of the available evidence we propose the following hypothesis. The elongate Pelagonian micrcontinent collided with a westdipping subduction zone within the Pindos Ocean, following Mid-Jurassic genesis of a suprasubduction-type ophiolite (e.g. Vourinos-Pindos;

392

I.R. SHARP & A. H. F. ROBERTSON

Liati et al. 2004). The ophiolites in the study area were emplaced to the NE and the Pelagonian ophiolitic m61ange was shed from the front of the advancing ophiolite (Fig. 13). This initial ophiolite emplacement did not by itself cause thick-skinned deformation or metamorphism of the underlying Pelagonian platform. With continuing convergence and the attempted subduction of Pelagonian continental crust a vast ophiolitic sheet was emplaced over the Pelagonian Zone. This is generally comparable with the latest Cretaceous attempted subduction of the Arabian continental margin, followed by rapid exhumation, as documented in the Oman Mountains south of the Semail Gap (e.g. Miller et al. 1998; Searle & Cox 2002). We then infer that a regional switch to strike-slip (transpression) took place during pre-Late Oxfordian-Early Kimmeridgian time (c. 155 Ma). This could relate to diachronous trench-margin collision, or to a change in microplate motion triggered by collision. A kilometres-thick competent ophiolitic slab was displaced subparallel to the relatively incompetent Pelagonian platform located beneath, and this was structurally thickened and deformed. This produced the pervasive N N W SSE, D~ stretching lineation and induced the amphibolite- or greenschist-facies metamorphism (M1). We then infer a phase of extensionrelated exhumation during Late Jurassic-Early Cretaceous time (D2). This evolutionary stage

Olistoliths Clasticsinforedeep

can be compared, for example, with the much younger (e.g. Mid-Cenozoic) exhumation of the basement of the Menderes Massif in western Turkey from beneath a thick pile of thrust sheets including ophiolites (Hetzel et al. 1995; Purvis & Robertson 2004). A possible trigger for regional exhumation was a reversal in subduction polarity, from generally westwards (oceanwards) in the MidLate Jurassic, to generally eastwards (towards the continent). The subduction beneath the Pelagonian continent resulted in subsequent mid-Cretaceous to Early Cenozoic greenschistto blueschist-facies metamorphism (M2), associated with dominantly SW displacement (D3) (Schermer et al. 1990).

Latest Jurassic-Early Cretaceous marine transgression An unconformable contact between the underlying Pelagonian units (Triassic-Jurassic marbles and serpentinized ultramafic rocks) and the overlying cover rocks is exposed throughout the Almopias Zone (Fig. 4). Palaeo-karstic weathering is widely developed, with localized bauxite and laterite accumulations on marble and ophiolitic lithologies, respectively. Fissures in serpentinite were commonly infilled with ophicalcite, which is gradationally overlain by

Transform linements

~;~

I ~ ~ ' J / ~ \ \ Existingintra l ~ ~ =-- ,.'// ~=;;=~t::~-L-~;e~ platformbasins ~ i ~ ~ i G ii!i ~r~i !a._-~e~/o~~ _ m m~ a to i k o . Fm) \ Loferiticcycles

_

Member Fig. 13. Schematic block diagram representing the main depositional elements and interpreted depositional setting of the Eastern Pelagonian Zone from the Triassic to Mid-Late Jurassic. Noteworthy features are the northeastward ophiolite emplacement and the setting of the Pelagonian ophiolitic m61angewithin a foredeep above the collapsed Pelagonian platform. This interpretation applies also to the Vermion area further south.

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE serpentinite-derived clastic sediments. Higher in the succession clastic sediments are intercalated with intermediate-silicic volcanic rocks, which are at their thickest (200-250 m) in the Central Almopias Zone (Fig. 14). The intermediate-silicic volcanic flows and water-lain tufts are seen to directly and unconformably overlie deformed ophiolitic lithologies in several sections (e.g. Mauropouli and Petrokorfi; Fig. 14). It is important to note that these volcanic rocks are not part of the underlying Pelagonian ophiolite that was previously emplaced in pre- to syn-Late Oxfordian to Early Kimmeridgian time. Bijon (1982) previously reported the presence of IAT and depleted boninite-type rocks in the Klissochori and Nea Zoi units. During this study basic meta-igneous rocks, interpreted as detached blocks within the Klissochori Unit, were found to be mainly enriched within platetype basalts, which we relate to Triassic rifting (see above). However, chemically depleted basalts with a marked negative Nb anomaly on MORB-normalized plots are also present (Fig. 5a). In places in the Central Almopias Zone (e.g. in the Liki-Margarita and Klissochori units; Figs 2b and 3) the regional unconformity at the top of the deformed and metamorphosed Pelagonian units is dislocated by an important transverse fault zone that trends subparallel to the regionally important Nission Fault (Fig. 3). This area, termed the 'zone de broyage' by Verg61y (1984), is characterized by a highly sheared and deformed unit, known as the Klissochori M61ange. This is a chaotic, mainly sedimentary unit that is strongly deformed in the lower part but much less deformed in the upper part. The lower part is clearly unconformable on the underlying Pelagonian platform (Fig. 15) and contains deformed and metamorphosed clasts of many of the rocks exposed in the Pelagonian platform, its basement and the ophiolite. The m61ange is intersliced with lenticular sheared serpentinite in places. The m61ange clasts are set in a matrix of sericitic and chloritic mudstone and are interpreted as multiple debris flows. The less deformed upper m61ange unit contains similar clasts set in a little-deformed matrix, and grades upwards into the typical shallow-marine mixed terrigenousclastic successionofmid-Cretaceous (BarremianAptian) age. Previously the lower m61ange unit was interpreted as being associated with the initial ophiolite emplacement, possibly as a foredeep sequence, whereas the upper m61ange was interpreted as part of the Cretaceous cover (Verg61y 1984). However, the presence of a definite unconformity between the Pelagonian platform

393

and the ophiolitic m61ange, and the overlying debris flows (m61ange) indicates that the Klissochori Unit as a whole post-dates D~ deformation and M~ metamorphism of the Pelagonian platform and ophiolite. During this work we identified several local successions of the lower and upper m61ange units but no overall intact succession and we were unable to confirm that a stratigraphical unconformity exists between two m61ange units (see Verg61y 1984). The 'zone de broyage' associated with the Klissochori M61ange appears to correlate with an elongate highly deformed unit ('la bande broy6'), which extends NW-SE from the Western Almopias Zone (Liki-Margarita Unit) southwards through the area between Naoussa and Veria (Fig. 1); this comprises sheared, cataclastic and mylonitic serpentinite ('Veria Ophiolite'), together with blocks including diabase, schist and marble (Braud et al. 1984). These units are equivalent to the Western and Central Almopias zones further north. It is important to note that this zone includes reef limestones that were derived from the Late Jurassic cover of the ophiolite, which is locally preserved. This 'zone de broyage' is associated with intense westward thrusting and dextral strike-slip associated with the emplacement of the Vermion nappe over the Pelagonian Zone during Early Cenozoic time (Mercier 1966). The Klisschochori and LikiMargarita units further north experienced similar intense Early Cenozoic deformation. The emplaced serpentinite at the contact between the Pelagonian platform and the overlying Late Jurassic-Cretaceous cover acted as a regional detachment (d6collement) associated with the intense deformation of the lower Klissochori M61ange unit. In this study we relate the Late Jurassic-Early Cretaceous units of the Central Almopias Zone to an important phase of extension (or transtension). Within extensional fault zones, footwall highs underwent subaerial erosion and karst development, whereas hanging-wall depocentres were infilled with coarse clastic sediments, volcanic rocks and minor intrusions (e.g. granophyre) (Fig. 14). Multiple dykes are interpreted as infills of transtensional fissures (e.g. Liki and Klissochori units). The existence of active hydrothermal systems is suggested by epidote mineralization. Some fault zones were apparently exploited by protrusions of ductile serpentinite that flowed onto the sea floor where they were covered by volcaniclastic and hemipelagic sediments, as seen within the lower m61ange unit. Upwards, poorly sorted polymict arenites and rudites were deposited by debris flows and by high- to low-density turbidity currents (e.g. upper

394

I . R . S H A R P & A. H. F. R O B E R T S O N

O

I ~

< o

~

I

9 ,..--.,,

o t'q .,..~ o

<

o

, ,...~

o

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE m61ange unit at Petrokorfi; Fig. 14). Continuing volcanism is indicated by local intercalations of rhyolitic air-fall tuff. The clastic sediments include abundant metamorphic detritus, including augen-gneiss and garnet mica schist (Verg61y 1984). The gneiss was presumably derived from the basement of the Pelagonian Zone or the Almopias Zone, which would require deep exhumation at least locally (e.g. along fault zones). The high-grade metamorphic detritus could, in principle, have been derived from an overriding crystalline thrust sheet, but there is no independent evidence that this existed. Facies trends and limited palaeocurrent data (Sharp 1994) suggest overall supply of the terrigenous sediments eastwards, from the Pelagonian Zone and the Central Almopias Zone, into the Eastern Almopias Zone. The above relationship between the emplaced Pelagonian platform ophiolite nappe and sedimentary cover rocks could, in principle, be explained in two ways: (1) first related simply to a phase of extension (or transtension) along the western Vardar Zone margin; (2) related to more profound rifting to form a new Cretaceous oceanic basin within the Vardar Zone. In option (1) the eastern margin of the Almopias Zone, including the Klissochori M61ange, acted as a zone of extension (or transtension) that was associated with intermediate-silicic volcanism, the emplacement of multiple debris flows, and with neritic carbonate and clastic sedimentation during Late Jurassic-Early Cretaceous time. The silicic composition of the volcanic rocks could reflect partial melting of thick underlying continental basement related to rifting. In option (2) the intersheared serpentinites could be interpreted as emplaced fragments of the lower plate of an asymmetrically rifted continental margin that was associated with the exhumation of continental mantle onto the sea floor during Late Jurassic-Early Cretaceous time. The exhumed material in this model was represented by serpentinized peridotite that was hydrothermally altered to form ophicalcite in fissures, and then covered by silicic volcanic rocks, terrigenous sediments and, locally, by pillow lavas. For example, in the extreme SE of the Klissochori Mdlange serpentinized dunites contain ophicalcite and lenses of pink micritic limestone in their upper part. The dunites are overlain by a thin horizon of serpentinitic or talc-rich mudstones and dunite-derived conglomerates (with reddened clast edges). These sediments are then covered, with a locally preserved primary contact, by little-deformed pillow lavas, lavas breccias and silicic extrusive rocks. This setting is, for example comparable with the exposure of continental mantle and the extrusion

395

of overlying MOR-type extrusive rocks during the final stages of continental break-up to open the Late Jurassic Penninic Ocean in the Western Alps (Manatschal et al. 2003). The main problem with model (2) is that in the Alps such exhumation took place in deep water, associated with radiolarite deposition, and there the volcanic rocks are of MOR type. By contrast, in the western Vardar Zone the associated sediments mainly accumulated in a shallow-water setting (e.g. reef limestones) and the volcanic rocks are mainly of intermediatesilicic composition. However, it is possible that various fragments of the Klissochori M61ange include a now-telescoped proximal to distal continent-ocean transition of Late JurassicEarly Cretaceous age. Figure 16 shows a reconstruction of the edge of the Pelagonian Zone and the Central Almopias Zone between the relatively proximal Klissochori Unit and the more distal Nea Zoi Unit. Further east, the Late Jurassic cover sequence of the Paikon Zone was represented by lagoonal facies, without volcanic rocks (Khromni limestones; Mercier 1966; Brown & Robertson 2003). Transgressive deposits of post-Late Kimmeridgian age in the Peonais Zone, further east again, were deposited in a marginal-marine to locally continental environment (Stais 1994). Early Cretaceous extension and exhumation were also inferred in the Paikon Zone (Brown & Robertson 1994) and elsewhere in the Pelagonian Zone (Doutsos et al. 1993). The extensional (or transtensional) faulting and volcanism within the Western and Central Almopias zones effectively ended prior to Aptian-Albian time. Some areas, especially fault blocks, remained emergent, undergoing redbed deposition and erosion of the metamorphic basement. Breaks in deposition occurred locally. For example, Late Jurassic reef build-ups are overlain by Early Cretaceous red beds and then by Mid-Late Cretaceous neritic carbonates in the Central Almopias Zone (e.g. Liki-Magarita Unit; Figs 2 and 16). The pre-Aptian-Albian time interval was thus marked by continuing tectonic instability along the Pelagonian-Almopias margin.

Late Jurassic-Early Cretaceous oceanic crust genesis The Eastern Almopias Zone is dominated by two large exposures of ophiolite-related extrusive rocks, the Mavrolakkos Unit in the west and the Krania Unit in the east (Mercier 1966; Fig. 2b). Detailed field mapping has allowed the correlation of these two units as a single ophiolite

396

I.R. SHARP & A. H. F. ROBERTSON

KLISSOCHORI UNIT LOWER TECTONIC MELANGE section west of railway bridge

Debris flow clast to matrix support ungraded disorganized fabric clast size 50 cm max clasts of amphibolite, garnet mica schists, foliated micaceous marbles, garnet, lithic greywackes, ? Triassic - Jurassic marbles, gneiss, milky quartz, volcaniclastic arenites Sericite & chlorite mudstone matrix

Erosive contact Debris flow clast to matrix support, crude bedding clasts 10 cm max

Sedimentary serpenitinite Laterite Unconformity Palaeokarstic breccia

Marbles of Pelagonian affinity, showing isoclinal folds and associated axial planar stretching lineation. This fabric is cut by micro extensional faults, brecciation and palaeokarst caverns filled with solution pebbles. The latter are seen reworked in overlying debris flows.

Fig. 15. Sedimentary log of the Klissochori Unit (lower m61ange unit) in the type area (see Fig. 2b). The presence of coarse debris flows should be noted, overlying an erosional remnant of serpentinite with the Pelagonian platform carbonates beneath. These debris flows were emplaced adjacent to the Nission Fault, a major transverse structure. thrust sheet, known as the Meglenitsa Ophiolite (Fig. 2b; Sharp 1994; Sharp & Robertson 1998). The extrusive rocks of both units (c. > 200 m thick) mainly comprise pillow lava and minor hyaloclastite. The lavas are overlain by a sequence ( < 50 m thick) of laminated black to

green, ferruginous and micaceous mudstones, thin turbiditic sandstones and cherts with minor pillow lavas (Black Schist Member). Minor massive sulphides are present along the lavasediment interface in the Krania Unit. A wholerock K - A r age of l 1 0 - 1 3 4 M a was obtained

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE

397 +..a

~N ..= <

. ,..,.~

~D N,-.~

~<

88~

,,-~

0 r~ ~ ,....,

398

I.R. SHARP & A. H. F. ROBERTSON

from basic igneous rocks (basalt or diabase) (Bertrand et al. 1994), in keeping with radiolarian age data. Above are radiolarian sediments, mainly ribbon radiolarites, radiolarian mudstones and siliceous mudstones, up to 2 5 m thick (Radiolarite Member). Pillow lava and thinly bedded sandstone are occasionally present at this level in the Mavrolakkos Unit. Radiolarian determinations (P. De Wever & H. YiLing; in Sharp & Robertson 1998) from the Mavrolakkos Unit indicate ages ranging from Late Jurassic (Callovian), to Early Cretaceous (Neocomian), possibly extending to Barremian-Aptian. Radiolarians from the Krania Unit yielded Late Jurassic (Mid-Oxfordian) to Early Cretaceous ages (Valanginian possibly extending to Berriasian-Turonian). Similar ages were reported by Stais (1994). The radiolarian sediments pass gradationally upwards into a sequence (>200 m thick) of mudstones, siltstones and mixed sandstonecarbonate turbidites (Flysch Member). The Krania Unit in the east is coarser grained, mainly comprising terrigenous and calcareous turbiditic sandstones and debris-flow deposits, with clasts of basalt, neritic limestone, minor volcanic quartz, rhyolite, dolerite, granite, diorite and granophyre. There are also subordinate interbeds of arkosic sandstone and dacitic tuff. The turbidites and debris flows were thus mainly derived from an unmetamorphosed, mostly extrusive igneous terrane, together with smaller amounts of shallow-water carbonate and metamorphic material. The clastic sediments in the Krania Unit in the east are thicker and coarser grained than in the Mavrolakos Unit, suggesting derivation from the Paikon Zone (to the east) where similar lithologies are exposed (Grahmmos Formation; Brown & Robertson 2003). However, within the Mavrolakos Unit, limited palaeocurrent data indicate eastward to southward flow and slump folds are locally NEvergent (Sharp 1994). Intraformational clasts of basalt and radiolarite were probably derived from subjacent oceanic crust. The upper age limit of the turbidites is constrained by the presence of unconformably overlying sediments of Late Cretaceous age along the western margin of the Mavrolakkos Unit. The Krania Unit, including the Flysch Member, is cut by localized granite, granophyre and basaltic sills and also by north-southtrending dykes (B6bien et al. 1980; Sharp & Robertson 1998). Localized amphibole-bearing quartz diorite dykes cutting the clastic sediments were dated at 124 Ma (Bechon 1981), suggesting that the intrusive rocks are approximately

contemporaneous with the underlying ophiolitic lavas. Sharp & Robertson (1998) noted that the basalts of the Mavrolakkos Unit are of nearMORB composition but a few samples are relatively depleted (Fig. 17a~l). Basalts and occasional late-stage dykes from the northern part of the Krania Unit (Mavrolakkos Vodeon) show a spread from near-MORB to relatively enriched with a few relatively depleted samples, plus several samples that show a small but distinct negative niobium anomaly (Fig. 17a). Many samples from the Krania Unit as a whole (e.g. west of Krania village) again show a near-MORB to slightly enriched composition. However, several samples show a pronounced negative Nb anomaly (e.g. from the old KraniaMandalon road-cut; Fig 17b) and are similar to some oceanic arc or back-arc basalts. These results, together with the presence of the local intrusions of granite and granophyre cutting the sedimentary cover, are suggestive of genesis in a subduction-influenced setting. On the other hand, the presence of the terrigenous sedimentary cover, albeit of deep-water origin, shows that this oceanic crust formed in a near-continental margin setting rather than an open-ocean setting. The genesis of the Late Jurassic-Early Cretaceous Meglenitsa Ophiolite post-dates the emplacement and metamorphism of the ophiolitic rocks of the Pelagonian and WesternCentral Almopias zones (pre-Late OxfordianEarly Kimmeridgian). An Aptian-Albian unconformity is developed on the western margin of the Meglenitsa Ophiolite (Sharp & Robertson 1998). A counterpart of the Meglenitsa Ophiolite, the Ano Garefi Ophiolite in the Voras Massif further north (Brown & Robertson 2004), is also unconformably overlain by deep-water sediments of Aptian-Albian age (Mercier 1966; Brown & Robertson 2004). Alternative tectonic settings for the genesis of the Meglenitsa Ophiolite during Late Jurassic-Early Cretaceous time are considered in the Discussion and conclusions section.

Aptian-Cenomanian passive margin subsidence During Aptian-Albian-Cenomanian time the western margin of the Vardar Zone experienced post-rift subsidence. Marine transgression of the combined Pelagonian and the Western and Central Almopias zones culminated in the development of an eastward deepening, mixed carbonate-clastic succession, characterized by fluvial-coastal plain to shelf environments.

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE Summary sedimentary logs of the Pelagonian Zone and the Western Almopias Zone applicable to this time are shown in Figures 4 and 18. Facies trends and palaeocurrent data are generally indicative of eastward sediment supply. Within the Pelagonian Zone, a diachronous marine transgression progressively covered remaining exposed areas, with an upward transition from coastal plain-fluvial to marginalmarine settings. Overlying sequences are fully marine and accumulated mainly within middle to inner shelf settings. Inner shelf areas were characterized by littoral conglomerates, storminfluenced beds and rudist biostromes. Platy bedded, bioturbated carbonates and calcarenites rich in a mixed benthic-planktonic fauna were deposited in deeper-water, more offshore areas. The Western Almopias Zone (i.e. Kerassia and Kedronas units; Fig. 18) represents an eastward continuation of the same east-facing margin sequence. Marked facies variations are evident, especially close to major transverse faults (e.g. Nission Fault) that are interpreted to have been still active. The facies of the Kerassia Unit exhibit an overall east- to SEfacing ramp geometry, from coastal plain to midouter ramp environments. Isolated subaerial highs persisted, for example in the Kedronas Unit, with karstic erosion and non-marine redbed deposition. Within the Central Almopias Zone, the Margarita Unit exhibits mainly neritic accumulation (Fig. 4). An isolated high, characterized by coral-rudist biostromes, developed in the Rhizarion region (e.g. Mavropouli and Skopia sections; Fig. 14), south of the Nission Fault, with contrasting coarse clastic sedimentation in the hanging wall of this fault-controlled basin directly to the north (e.g. Korfyi Pegiou section; Fig. 13). South of the Rhizarion fault block pelagic-hemipelagic deposition dates from Early Cretaceous time. North of the Kato Loutraki Fault (Fig. 3), Triassic-Jurassic marble and serpentinite are unconformably overlain by a mainly carbonate cover of Aptian-Albian age (Livadia Unit). The contact is locally marked by karstic weathering of Triassic-Jurassic marble, bauxite and red-beds (Brown & Robertson 2004). The easternmost part of the Central Almopias Zone, on the other hand, was characterized by deeper-water hemipelagic sediments of AptianAlbian (and younger) age that unconformably overlie a local serpentinite basement (Nea Zoi Unit; Figs 3 and 4). By contrast, in the Eastern Almopias Zone, Aptian-Albian (or younger) sediments have only been recorded as exposures along the western edge of the Meglenitsa Ophiolite.

399

Regional comparisons show that a similar east-facing passive margin developed along the eastern Pelagonian Zone, both to the south (e.g. Sporades, Evia, Argolis; e.g. Clift & Robertson 1990b; Robertson 1990; Clift 1992) and further north in Macedonia and Serbia (see Karamata 2006).

Cenomanian-Turonian: depositional hiatus and relative sea-level rise During Cenomanian-Turonian time, facies evidence points to a relative sea-level rise and deepening of the Pelagonian-Almopias carbonate margin. The west-to-east overall deepening trend persisted, with a few areas remaining subaerially exposed (e.g Arnissa region and Kaimatchalan Massif). Sedimentation during this time appears to have been influenced by a pulse of extension-related subsidence, coupled with eustatic sea-level rise (e.g. Sharland et al. 2001). In the Pelagonian and Western Almopias zones, ramp-interior lagoonal and ramp-margin rudistic carbonates were terminated by intraformational conglomerates that were in turn abruptly overlain by deeper water outer-ramp and hemipelagic facies (Figs 18 and 19). Dramatic drowning of remaining subaerial highs is evident in parts of the Western Almopias Zone south of the Nission Fault, where coarse immature facies were deposited in a littoral setting (i.e. western and eastern sections of the Kedronas Unit; Fig. 18). In the Central Almopias Zone the Klissochori Unit experienced clastic sedimentation (Figs 4 and 14). The more distal Nea Zoi Unit further east (Figs 4 and 16) exhibits the emplacement of a large slide-sheet of mainly neritic carbonates ('Trypia Petra slide' derived from the Klissochori Unit?) and an upward short transition to coarse clastic sediments; some undated radiolarites are also present. A similar relative sea-level rise is well documented further east, in the Paikon Zone, where a short period of subaerial exposure was followed by abrupt drowning of the carbonate platform and a transition to hemipelagic, then turbiditic, deposition (Sharp & Robertson 1993; Brown & Robertson 2003). The Paikon Zone remained an area of deep-water accumulation until Paleocene deformation. Late Cenomanian-Turonian eustatic sea-level rise (Haq et al. 1987; Sharland et al. 2001) does not, by itself, explain features such as synsedimentary faulting, slumping and localized coarse clastic deposition in the area studied, and a tectonic trigger is likely. The CenomanianTuronian boundary was marked by uplift or

400

I.R.

SHARP

& A. H. F. ROBERTSON

6 r O~

CO (3)

r

O~

.m F

I

k-~

I-

F

LO 0r G.

e i , =o

0~

s~

~,~

.

"10 Z

/-..~. ..._.. o

'1~

r

F 0

h Z *'~0

?

O~ -~

~'~

.~ ~1 ~ = I o 0 - - ~ . _ ~ ,_ r I=.~

O0 0 0 0 0

0 0 0

0 0 ~-

0 0 0 0 v.-

0 ~--

o

8~lOIAINOOJ

0

0

0 0 v-

~-

0

~ ' ~ r~

81~OlftPlooJ

/ 0

0 0 0 v-

r

o ;~=

< m O O

0

+ttt n

o

oq

Z

...~

O

.o Z

b

o CO 0 0 ---

8~lOIAIPlOOJ

00

0 ~-CJ

C~

~

~

0

"~ n3 ~

,.0

%,~ 0 ,.-,

0

0 0 0

0

0 0 ,-

8~lO~PlOOJ

I-

t'~

Z

~

~

0

~

.~.~

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE subsidence regionally; for example, the eastern margin of the Pelagonian Zone in Evia (Robertson 1990) and Argolis (Clift 1992) underwent abrupt subsidence at this time. In addition, the south Aegean region was affected by regional subsidence (Harbury & Hall 1988). Late Cretaceous ophiolites of supra-subduction-zone type, based on geochemical evidence, have been identified in the Argolis area (NE Peloponnese) (Clift & Robertson 1989), and in Crete and Karpathos (Koepke et al. 2002). It is possible that the Cenomanian-Turonian hiatus in the area studied was triggered by intra-oceanic subduction and supra-subduction spreading within the Vardar Ocean.

Campanian-Maastrichtian: resumed subsidence Eastward deepening related to subsidence was re-established with the development of thick ramp-shelf margin to outer ramp-shelf carbonates during Late Santonian-Early Campanian time. The dominant control was eustatic sea-level rise (e.g. Hancock & Kauffman 1979). A northsouth-striking, east-facing rudist-dominated ramp-shelf margin complex persisted in the Pelagonian, Western Almopias and parts of the Central Almopias zones during CampanianMaastrichtian time. The northern Pelagonian Zone was finally transgressed, with the faultbounded Kaimaktchalan Massif being the last area to be flooded. Eastwards, the Central and Eastern Almopias zones experienced coarse clastic sedimentation (Klissochori, Nea Zoi and Vryssi units), dominated by siliciclastic turbidites that are interpreted as having accumulated in a deep-water submarine fan or apron setting. In addition, fine-grained hemipelagic sediments accumulated to the south of the Nission Fault (e.g. in the Ano Grammatiko area; Fig. 8).

Late Maastrichtian: transition to a foreland basin Late Maastrichtian time saw dramatic subsidence of remaining areas of the regional eastfacing carbonate ramp-shelf (Figs 19 and 20). In the Pelagonian Zone, Western Almopias Zone and parts of the Central Almopias Zone (i.e. the Margarita & Klissochori Units; e.g. Petrokoryfi region, Fig. 14) there is an abrupt transition from neritic, rudist-bearing carbonates, accumulating within and along the margin of the carbonate platform, to hemipelagic micrites and deep-water terrigenous sediments of an inferred foredeep succession. In the Pelagonian Zone (Fig. 19),

401

iron-phosphate encrusted and bored hardgrounds developed at the top of the platform carbonate succession, overlain by muddy ferruginous carbonates. An iron-stained subaerial emergence horizon developed in the Western Almopias Zone (Kerassia and Kedronas units), associated with fissuring of underlying limestones and the deposition of localized intraformational conglomerates. The base of overlying deeperwater calcareous mudstones includes wellrounded blocks of neritic limestone, derived from subjacent Maastrichtian rudistic carbonates. Late Maastrichtian-Paleocene siliciclastic sequences in the Pelagonian Zone are characterized by hemipelagic sediments and thin-bedded, fine-grained turbiditic sandstones that coarsen upwards, passing into high-density turbidity current and debris-flow deposits, with clasts including fresh basalt and metamorphic detritus. These successions appear to shallow upwards and may be fluvial near the top in places, before being overridden by SW-verging thrust sheets. A similar facies transition to siliciclastic turbidites is known elsewhere along the eastern margin of the Pelagonian Zone, including Evia (Robertson 1990) and Argolis (Clift 1992). The profound facies change is interpreted to mark the transition to a foreland basin ahead of a westward-propagating thrust load (Sharp & Robertson 1993; Fig. 20). The more outboard sequences (i.e. Western Almopias Zone) underwent uplift and erosion, whereas the more inboard (westerly) sequences (i.e. Pelagonian Zone) became the sites of muddy, lagoonal deposition. Later, the entire margin collapsed, becoming a deep-water basin in which siliciclastic turbidites accumulated. This basin filled with coarser-grained, shallower-water sediments derived from the advancing thrust load.

Late Cretaceous eastward subduction Late Cretaceous sediments near the contact between the Central and Eastern Almopias zones (Nea Zoi and Vryssi units; Fig. 4) are dominated by siliciclastic turbidites showing strong layerparallel extension to form classic phacoidal fabrics. Such features could be related to the layer-parallel extension that characterizes some shear zones within thrust sheets but are also reminiscent of the phacoidal fabrics seen in many subduction complexes (e.g. Franciscan Complex, western USA: Cloos 1984). In-situ glaucophane was reported from Late Cretaceous turbidites in the Liki-Margarita Unit (Central Almopias Zone), particularly along the contact between the Liki-Magarita Unit and the Klissochori Unit

402

I . R . S H A R P & A. H. F. R O B E R T S O N

I

~

~-,

,~ <~

8

~+

+1

+

+

~ I.~

,

E

~

A Z

.,m_ .m

-

,

.

"

+~ ~'I

"+ +

<.,+~-0,I.~

:-

~

~,

+-I~,

m.I ~

E~

.

.

+ om-+

.

.

~

.

+- I+

~

I~+l<+ I+

.-_..,++

.

+<,+

. . . . .

~

+i+ "

, ~

<+ ~_ ~

m+ I-+1 +o + + ~<~ m, .gp> +_. .

,

....

~9

,~

~-+

I

--

+

+~-,

,-

+ /++ +1,,,,

~"

.+

.

+<~.++ -1

++

", +

+

++

++

.m

+

\ I= o

\

rj =I o

+

<~I

+ ~\ w

m.

z 0 N

P=

< o. 0 m

+0

.+

~\ ~

tEE

--Jo

~

<+It--+-_+}+,. , ~ \ t , ~ . ,

+

~d

+I_

I=

t~ o

~9 ,+,,~',~i;',;+,i',i~i~+~lm+m ~ H H{"++'+' Oi

m iu I.-

"~

I 1

I I

i=

I=

/,+,~)~i++++ii!iii+ , +++ :++,

+

+

~

,!

m_

~-

,,o

i+

!' (|

o

+', .+

+

+

<

+'+

+ uJ

++++

t

.,+,. .+.%%<++<+>; "..d ."+. :# ,~..... ?:;

=

Z 0 N Z < z 0

,7<>+

+_

,=,++" "+,:+ ?+

. ~

.m

.J

< Z

+.t

~k

<+, <,,l

= +.j,

+.=

~ ++~J

+0,

i=

.m.,m I= o

<.+.+

~ =I

6~

i

rj~

< --1

z

Z

..~ +..,

<.,.. ~ 0 .0. mj

_o rr I-

~-

<

+"

0

~+-k+.

Q

B ~ ~o",

~

z

e~ ul

z

~

z

=E

0

5 <

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE (Verg61y 1984; Sharp 1994). The HP-LT metamorphism is interpreted to relate to eastward subduction of the Vardar Ocean and the eastern edge of the Pelagonian-Almopias continental unit. The siliciclastic turbidites of the Nea Zoi and Vryssi units could thus represent part of an accretionary prism formed along the western margin of a Vardar Ocean rather than merely the distal part of a foreland basin. Similar deformed siliciclastic turbidite facies occur in a comparable tectonic setting further south in Greece (e.g. Argolis Peninsula), where they are interpreted as a latest Cretaceous-Early Cenozoic accretionary prism related to eastward subduction and the related development of a foreland basin ( C l i f t & Robertson 1989; Clift 1992). South of the study area in the Sporades Islands, an ophiolitic thrust sheet is present at the highest structural levels, and is interpreted to be the result of eastward subduction that culminated in westward ophiolite emplacement during latest Cretaceous-Early Tertiary time (Jacobshagen & Wallbrecher 1984). Comparable eastward subduction is inferred to have taken place along the eastern margin of the Pelagonian Zone in Greece and former Yugoslavia (e.g. see Karamata 2006). A number of the metaophiolites of the eastern part of the Pelagonian zone (e.g. High Pieria) that lack D~ deformation are inferred to have been emplaced from the Vardar Zone during latest Cretaceous-Early Palaeogene time.

Palaeogene suturing Regional-scale folding and thrusting took place with a southwesterly to westerly vergence, with the Almopias Zone overthrusting the Pelagonian Zone (Mercier 1966; Verg61y 1984; Mountrakis et al. 1987; Fig. 2b). In the east, the Peonais Zone overthrust the eastern margin of the Paikon Zone, and the Serbo-Macedonian Zone generally overrode the Vardar Zone, as seen in the Voras Massif in the north (Brown & Robertson 2004; Fig. 3). However, the Eastern Almopias Zone (Meglenitsa Ophiolite) was clearly thrust both westwards and eastwards over the Central Almopias Zone and Paikon Zone, respectively (Sharp & Robertson 1993; Brown & Robertson 2003; Fig. 3). Indeed, the Meglenitsa Ophiolite was thrust northeastwards over the Paikon Massif to the highest structural level in the region, consistent with its low metamorphic grade (lower greenschist facies or less), and limited structural deformation (i.e. one phase of Tertiary folding; Sharp 1994; Verg61y 1984;

403

Sharp & Robertson 1998). However, north of the Kato Loutraki Fault the convergence direction was reversed and the equivalent Ano Garefi Ophiolite was thrust southwestwards, together with other units of the Voras Massif (Brown & Robertson 2004). This indicates that the thrust belt was segmented, with the Paikon segment experiencing local back-thrusting (Fig. 3). The Pelagonian Zone was affected by largescale imbrication and the development of regional-scale nappes (e.g. Haut Vermion Nappe), regional greenschist-facies metamorphism (Kockel 1986) and localized dynamic metamorphism, related to regional westward overthrusting (Braud et al. 1984; Fig. 8). The previously emplaced Pelagonian Ophiolite was detached from its Cretaceous cover in many areas, associated with re-thrusting and the development of large-scale SW-verging folds (e.g. Messovounon and Ayios Dimitrios areas; Sharp 1994; Figs 3 and 8). Similarly, in the Western Almopias Zone, the mainly calcareous cover was detached from underlying ophiolitic units, typically exploiting highly incompetent serpentinite. The underlying units were strongly imbricated, increasing the structural complexity of the ophiolitic m61ange (e.g. in the Haut Vermion Nappe; Fig. 8). Within the Pelagonian and Western and Central Almopias zones there is a general increase in the intensity of deformation from the west to the east. This is particularly evident in units of Mid-Late Cretaceous age. The probable explanation is that the downflexed passive margin in the east was located near the axis of the Vardar suture zone, whereas the Pelagonian Zone was located in a more westerly (external) position. The Cretaceous cover of the Pelagonian Zone has generally undergone minimal penetrative deformation, whereas age-equivalent units in the Western and Central Almopias Zones (Kerassia, Margarita, Klissochori and Nea Zoi units) were more pervasively deformed. A similar eastward increase in deformation is seen in the Voras Massif, as far east as, and including, the Livadia Unit (Brown & Robertson 2004; Figs 2b and 3). In addition, as already mentioned, sediments of the Liki-Margarita Unit that post-date the Late Jurassic ophiolite emplacement contain evidence of HP-LT minerals of Late Cretaceous-Early Cenozoic age. Specifically, blue crossite crystals are developed parallel to the main $1 and $2 schistosity within deformed pelitic calcarenites that contain Globotruncana sp., proving a Turonian or younger age (e.g. exposures of Late Cretaceous Flysch exposed in railway sections

404

I.R. SHARP & A. H. F. ROBERTSON TCHOUKA

~o~~

SARAKINI

~u~0,Alll0,te~

1~

-~-

Argillaceous ] carbonates Hardground ~ Orbitoides - ~ packstene

"~ ,.,

.. I I

Upper Maastrichtian .............

Upper

Intra-formational conglomerate& Fe-Phcrust

~

20

M

~ur. ~ Alll

F:0

t~

Hippuritd

Flysch

KERASSIA

s

Turbidites :~--:----1 Upper ~,,- B ~ E ~~ Maastrichtian ~ ~ Intra-formational nglomerate& Q:k/~I~x,-~"' /~ $~/-/-"~-t~"~ ~ P I T Ip,oFe-Phcrust ,_. .:. ,-, .:.,I(p ~ Hippuritid

Pelagoni Arnissa.

,/~~

(kX_~~s

9

a

,~P I, ,I

.~ /

, I

biostrome

29 L ~ . _[e's ~" .'~ NISSION FAULT

~/vv v~/'a/v

I 4 We=

. ~ .

9

[I - ~i - . -~i -

9

'

,o.,s

Zone

9

AI,I

-i- '-n

Turbidites

[~.:~_ ._:~2~"~+ Turbidites

~q-'."

9;

!+:--'k~---Gi ~I ~ i ~ :--Tr

Fe rich

~..:_=:--~.,_----i +

Alll

"~ .F~-~-_.--t

:~--~--~-~_4~"

:~

Upper Maastrichtian '~" Argillaceous carbonates 9 " Hardground --I ~. 17. Orbitoides

upper

~ Maastrichtian P':I ~ , 4 ~l re-P,oust

ii-~i =~ I~,'o'*o~e I cagl

a .tone

All0 j

AI10

"w

aes

Al9g

"%

Rudists Benthic forams F:0 Planktonic forams

~

Bioturbation

~

Nereites Ichnofacies

~

Carbonaceous debris

9

Cor-algae

~

Echinoderms

,J

Shell debris

~2~ Olistoliths

/~

Stylolites

Fe-eh Iron-phosphate

Fig. 19. Sedimentary logs of the top of the Cretaceous carbonate margin and the transition to flysch in the Pelagonian and Western Almopias zones 9(See Fig. 3 for regional map.)

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE Pelagonian Zone

Eroding hinterland (karst) ~

Western Almopias Zone

~

Mature carbonate ramp/rimmed shelf Pelagonian Zone

~

Central Almopias Zone

~ $ Ramp interior to shoal margin -- : Mid-outer ramp + ,.~ , .~__.7..__~.~.~,Rudist_grainston e ~---..~.j.~, , . ~ t ~ ,-"~..~ shoals

Campanian~an

~

~

Hemipelagic & basinal clastics SL

Storm beds & hernipelagic deposits "Scaglia" facies

~

.

.

~

Western Almopias Zone

Basinalclastics Central Almopias Zone

Mixed carbonate/clastic deposits Uplift, erosion, in muddy lagoon hardgrounds, fe-ph crusts rT .'.:=:.'.'.IT ~ ~ Platform/ramp margin~e Olistoliths flexure and collapse ~ ~ -.~ Foreland bao,,,

UpperMaastrichtian

Margin flexure & collapse

405

~sl,cs

~

Sk Advancing thrust stack

~:~::-::::.:..

Fig. 20. Sketch interpretation of the Cretaceous carbonate margin and transition to flysch in the Pelagonian and Western Almopias zones.

at the southeastern end of Krivi Dereki Rema; Fig. 3). In summary, HP-LT metamorphism, intense shearing, tight imbrication and westward nappe emplacement are together seen as the response to the attempted subduction of the Pelagonian microcontinent in a trench, dipping northeastwards beneath the Eurasian active margin and its amalgamated terranes (i.e. the Peonias and Paikon units).

Mid-Cenozoic-Recent: extensional collapse An upper age limit for the regional Early Cenozoic deformation is provided by the presence of subhorizontal Eocene limestones and conglomerates in the Peonais Zone (Mercier 1966). During mid-Cenozoic time the thrust stack in the Vardar Zone underwent extensional collapse, probably related to 'roll-back' of a northward-dipping subduction zone in the south Aegean region (e.g. Kilias et al. 1999), probably beginning in the Oligocene (c. 25 Ma; Burchfiel et al. 2000). The post-Miocene neotectonic phase was also marked by widespread subaerial volcanism and continuing fault-controlled deformation (Mountrakis et al. 2006). This was related to subduction in the south Aegean region and the development of the Aegean arc behind a northward-dipping subduction zone (e.g. Le-Pichon & Angelier 1979).

Discussion and conclusions The units of the eastern Pelagonian Zone and the Western and Central Almopias zones are interpreted as the western margin of the Vardar Ocean from Triassic to latest Cretaceous time. This area experienced rifting during the Triassic associated with alkaline magmatism and terrigenous sedimentation, followed by passive margin subsidence during Late Triassic-Early Jurassic time. During the Late Triassic-Early Jurassic the Pindos oceanic basin formed to the west of the Pelagonian microcontinent. During Mid-Late Jurassic time, a vast ophiolite thrust sheet, including the combined Vourinos-Pindos ophiolite, was emplaced over the Pelagonian microcontinent, probably in response to the collision of the Pelagonian microcontinent with a SW-dipping intraoceanic subduction zone. Northeastward emplacement of this ophiolite from the Pindos Ocean to the west is favoured based on limited structural evidence from the area studied and additional evidence from the Jurassic ophiolites and underlying units throughout Greece, Albania and former Yugoslavia. The Pelagonian ophiolitic mdlange in the area studied is interpreted as a subduction complex that was emplaced onto a regional foredeep following the flexural collapse of the Pelagonian passive margin. The ophiolite and the underlying Pelagonian platform experienced

406

I.R. SHARP & A. H. F. ROBERTSON

greenshist- to amphibolite-facies metamorphism (M1), coupled with ductile deformation. The development of a pervasive NNW-SSE stretching fabric is inferred to relate to lateral (strikeslip) displacement of the emplaced ophiolite. The ophiolite emplacement is dated as preLate Oxfordian-Kimmeridgian, from the age of depositionally overlying coralline limestones (assuming that the ophiolitic rocks of the Western Almopias Zone formed part of this regionally emplaced ophiolite). This has implications for the timing of thrusting (D1) and metamorphism (Ml) further south in the Pelagonian Zone, which is traditionally believed to be no older than Early Cretaceous. The Pelagonian ophiolitic m61ange and the adjacent Vourinos and Albanian (Mirdita) ophiolites were overlain by deep-water carbonates along the western margin of the Pelagonian Zone (e.g. Mavri Rakhi Formation). The emplaced ophiolitic complex draped back into the relict Pindos Ocean to the west, where it was covered by deep-sea carbonate sediments. The Almopias Zone was exhumed by Late Jurassic time; the Pelagonian Zone as a whole was exposed by mid-Cretaceous time and progressively covered by shallow-water carbonate sediments. The probable cause of the exhumation was extensional tectonics, coupled with erosion. During Late Jurassic-Early Cretaceous time, following ophiolite emplacement, the Western and Central Almopias zones experienced extension, or transtension, associated with the emplacement of multiple debris flows ('m61ange'), coarse clastic sedimentation and intermediatesilicic volcanism. This was followed by a resumption of passive margin subsidence along the eastern margin of the Pelagonian microcontinent, interrupted by a Cenomanian-Turonian hiatus that was at least in part tectonically triggered. Further east, the Vardar Zone was the site of a Mesozoic ocean (Almopias Ocean) that opened in response to Triassic rifting. This ocean was subducted northeastwards beneath the Serbo-Macedonian Zone during Early to MidJurassic time associated with opening of the Guevgueli marginal basin in the Peonias Zone. The Vardar Zone (Eastern Almopias Zone) includes large slices of basic extrusive rocks, associated deep-sea terrigenous and pelagic sediments (radiolarites), and minor intrusive rocks that are interpreted as the upper part of a dismembered Late Jurassic-Early Cretaceous ophiolite (Meglenitsa and Ano Garefi ophiolites). The Meglenitsa Ophiolite formed after the regional Mid-Late Jurassic ophiolite emplacement and was not involved in regional deformation (D1) and metamorphism (M0. This relatively

young oceanic crust was later subducted eastwards beneath the Serbo-Macedonian Zone, by then amalgamated to Eurasia, during Late Cretaceous time and finally emplaced onto the Pelagonian-Almopias Zone, which by then had become a foreland basin; any other Vardar oceanic crust was subducted, leaving little trace. HP-LT metamorphism of probable Early Cenozoic age affected the eastern margin of the Pelagonian continent (Central Almopias Zone) related to collison with the subduction zone dipping eastwards beneath Eurasia and its previously accreted units. An outstanding issue is the regional tectonic setting of the Late Jurassic-Early Cretaceous rifting along the western Vardar margin and the related intermediate-composition silicic volcanism and neritic clastic sedimentation, as seen within the Eastern and Central Almopias zones. One option, shown in Figure 21a, is that the Almopias (Vardar) Ocean in northern Greece was completely closed during Mid-Late Jurassic time (pre-Kimmeridgian) (Papanikolaou 1996-1997), followed by the reopening of a new, Cretaceous oceanic basin (e.g. Jacobshagen & Wallbrecher 1984), possibly as a pull-apart basin (Sharp 1994). This model is consistent with the presence of terrigenous turbidites above the Meglenitsa Ophiolite. However, it is generally accepted that a wide Vardar Ocean still existed in the Cretaceous, sufficient to fuel extensive calc-alkaline magmatism along the Eurasian margin, as seen in the Serbo-Macedonian and Peonais zones (Dercourt et al. 1986, 1993, 2000; Ricou et al. 1988). If the Vardar Zone had completely closed in the Late Jurassic this would require a major spreading event during the Cretaceous. A second option, shown in Figure 21b, is that the Meglenitsa Ophiolite formed by suprasubduction-zone spreading adjacent to the, by then, amalgamated Eurasian convergent margin to the NE. The ophiolite was finally emplaced when the Pelagonian microcontinent collided with the subduction trench during latest Cretaceous-Early Cenozoic time. This is consistent with the geochemical evidence of a subduction influence on the Meglenitsa Ophiolite, and the probable derivation of the deep-water clastic sediment cover of the ophiolite from the Paikon Zone to the east. The main problem is to explain the Late Jurassic-Early Cretaceous extensional or transtensional rifting along the western margin of the Vardar Ocean (i.e. within the Western and Central Almopias zones). This could possibly reflect a switch in subduction polarity within the Pindos Ocean to the west, from northeastwards in the Mid-Late

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE

407

•

~

~

~'~

0

o

~N

N

"~.,~ ~ ~ -~ ~ o'-"

o.O ~ ~ ~.: ~o~ ~'~

O~o~ 0

~+..~

.~

"" ~,X::

~ ~~

oo~N

=0o.g

~o~o~ o

9 o

on~

408

I.R. SHARP & A. H. F. ROBERTSON

Jurassic (resulting in c o m p r e s s i o n - t r a n s p r e s s i o n ) to southwestwards in the C r e t a c e o u s - E a r l y Tertiary (resulting in extension-transtension). I. R. S. acknowledges an NERC studentship held at the University of Edinburgh and A. H. F. R. thanks the University of Edinburgh for assistance with the costs of fieldwork. B. Rosen (British Museum of Natural History) kindly identified and dated the corals. We thank S. Brown, J. Dixon, M. R. W. Johnson and A. Rassios for helpful discussions. The manuscript benefited from comments by G. Migiros, A. Kilias and D. Mountrakis. We thank Hydro Media for assistance with drafting several of the diagrams and funding the three colour figures.

References AL-RIYAMI, K. & ROBERTSON, A. H. F. 2002. Mesozoic sedimentary and magmatic evolution of the Arabian continental margin, northern Syria: evidence from the Ba~r-Bassit M61ange. Geological Magazine, 139, 395-420. BAROZ, F., BI~BIEN,J. & IKENNE,M. 1987. An example of high-pressure low-temperature metamorphic rocks from an island-arc: the Paikon series (innermost Hellenides, Greece). Journal of Metamorphic Geology, 5, 509-527. BARTON, C. 1976. The tectonic vector and emplacement age of an allochthonous basement slice in the Olympos area, NE Greece. Bulletin of the Geological Society of Greece, 7(2), 384-496. BAUMGARTNER, P. O. 1985. Jurassic sedimentary

evolution and nappe emplacement in the Argolis Peninsula (Peloponnesus, Greece). M6moires de la Societ6 Helv6tique de Science Naturelle, 99. BI~BIEN, J., BAROZ, J., CAPERDI, S. & VENTURELLI,G. 1987. Magmatisme basique associ6es h l'ouverture d'un basin marginal dans les Hell6nides internes au Jurassic. Ofioliti, 12, 53-70. BI~BIEN,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. BI~BIEN,J., OHNENSTETTER,D., OHNENSTETTER, M. & VERGI~LY, P. 1980. Diversity of Greek ophiolites: birth of oceanic basins in transcurrent systems. Ofioliti, 2, 129-197. BEBIEN, J., PLATEVOET, B. & MERCIER, J.-L. 1994. Geodynamic significance of the Paikon Massif in the Hellenides. Contribution of the volcanic rock studies. Bulletin of the Geological Society of Greece, 30(1), 63-67. BECHON, F. 1981 Caract~res de tholeiites abyssales des formations magmatiques basiques des unit6s orientals de la zone d'Almopias (Mac6doine greque). Comptes Rendus de l'Acadkmie de Sciences, Skrie II, 295, 105-108. BERTRAND, J., FERRII~RE,J. & STAIS,A. 1994. Donn6es petrographique et g6ochronologiques sur des laves des domaines vardariene de P6onais et d'Almopias (Hell6nides orientales). Bulletin of the Geological Society of Greece, 30(1), 213-222.

BIJON, J. 1982. GOologie et gkochemie des formations

volcano-sOdimentaires d'6ge Jurassic supkrieur et CretacO de la region de d'Edessa (Grkce, province de Pella). Th6se de Trois6me Cycle Universit6 de Paris-Sud, Orsay. 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, 2-20. BORTOLOTTI, V., CHIARI, M., MARACUSSI, M., MARRONI, M., PANDOLFI, L., PRINCIPI, G. & SACCANI, E. 2004. Comparison among the Albanian and Greek ophiolites: in search of constraints for the evolution of the Mesozoic Tethys ocean. Ofioliti, 29, 1-94. BRAUD, J., BRUNN, J.-H., CAMPION, G., et al. 1984. La chfiine de Vermion, ses nappes et sa bande broy6e. Bulletin de la Societk GOologique de France, 26, 713-717. BROWN, S. A. M. 1994. The geological evolution

and regional significance of the Mesozoic-Early Tertiary Paikon Massif, Northern Greece. PhD thesis, University of Edinburgh. BROWN, S. & ROBERTSON, A. H. E. 1994. New structural evidence from the Mesozoic-Early Tertiary Paikon unit, north-eastern Greece. Bulletin of the

Geological Society of Greece, 7th International Congress, 30(1), 159-170. BROWN, S. A. M. & ROBERTSON, A. H. F. 2003. Sedimentary geology as a key to understanding the tectonic evolution of the Mesozoic-Early Tertiary Paikon Massif, Vardar suture zone, N Greece. Sedimentary Geology, 160, 179-212. BROWN, S. A. M. & ROBERTSON, A. H. F. 2004. Evidence for the Neotethys ocean rooted in the Vardar zone: evidence from the Voras Mountains, NW Greece. Tectonophysics, 381, 143-173. BURCHFIEL, C. B., NAKOV, R., TZANKOV, T. & ROYDEN, L. H. 2000. Cenozoic extension in Bulgaria and Northern Greece: the northern part of the Aegean extensional regime. In" BOZKURT, E. & WINCHESTER, J. D. A. (eds) Tectonics and

Magmatism in Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 325-352. CAMPION, G. 1966. ;Etudegkologique du massif de Gola Tsouka (chdine de Vermion). Doctoral thesis, Universit6 de Paris X1. CLIFT, P. D., 1992. The collision tectonics of the southern Greek Neotethys. Geologische Rundschau, 86, 669-679. CLIFT, P. D. & DIXON, J. E. 1998. Jurassic ridge collapse, subduction initiation and ophiolite obduction in the southern Greek Tethys. Eclogae Geologicae Helvetiae, 91, 128-139. CLIET, P. D. & 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. D. & ROBERTSON, A. H. F. 1990a. Deepwater basins within Mesozoic carbonate platform of Argolis, Greece. Journal of the Geological Society, London, 147, 825-836.

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE CLIFT, P. D. & ROBERTSON, A. H. F. 1990b. A Cretaceous Neo-Tethyan carbonate margin in Argolis, southern Greece. Geological Magazine, 127, 299-308. CLOOS, M. 1984. Flow m~lange, numerical modeling and geological constraints on the origin of the Franciscan Complex, California. Geological Society of America Bulletin, 93, 330-345. DANELIAN,T., ROBERTSON,A. H. F. & DIMITRIADIS,S. 1996. Age and significance of radiolarian sediments within basic extrusives of the marginal basin Guevgueli Ophiolite (northern Greece). Geological Magazine, 133, 127-136. DERCOURT, J. RICOU, L. E. & VRIELYNCK, B. (eds) 1993. Atlas Tethys Palaeoenvironmental Maps. Beicip-Franlab, Paris. DERCOURT, J., GAETANI, M., VRIELYNCK, B., et al. (eds) 2000. Peri-Tethys Palaeogeographical Atlas. Commission for the Geology Map of the World, Paris. 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. DE WET, A. P., MILLER, J. A., BICKLE, M. J. & CHAPMAN, H. J. 1989. Geology and geochronology of the Arnea, Sithonia and Ouranoupolis intrusions, Chaldiki peninsula, Northern Greece. Tectonophysics, 161, 65-79. DIMITRIADIS, S. & ASVESTA, A. 1993. Sedimentation and magmatism related to the Triassic rifting and later events in the Vardar-Axios zone. Bulletin of the Geological Society of Greece, 28(2), 149-168. DOUTSOS, R., PE-PIPER, G., BORONKAY, K. & KOUKOUVELAS, I. 1993. Kinematics of the Central Hellenides. Tectonics, 12, 936-953. FERRIERE, J. 1976. Sur la signification des s6ries du massif de l'Othrys (Gr6ce continental): la zone isopique maliaque. Annales de la Societk Gdologique du Nord, 96, 121-134. GALEOS, A., POMONI-PAPAIOANNOU, F., TSAILAMONOPOLIS, S., TURNSEK, D. & IOACIM, C. 1994. Upper Jurassic-Lower Cretaceous 'molassic-type' sedimentation in the western part of the Almopia subzone, Aridhea Loutra Unit (northern Greece). 7th Congress of the Geological Society of Greece, ThessalonikL May 1994, Abstracts, 54. Geological Society of Greece. HANCOCK, J. M., & KAUFFMAN, E. G. 1979. The great transgressions of the Cretaceous. Journal of the Geological Society, London, 136, 175-186. HARBURY, A. & HALL, R. 1988. Mesozoic extensional history of the southern Tethyan continental margin in the S.E. Aegean region. Journal of the Geological Society, London, 145, 283-801. HAQ, B. U., HARDENBOL,J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. HETZEL, R., RING, U., AKAL, C. & TROESCH, O. 1995. Miocene NNE-directed extensional unroofing in the Menderes Massif, southwestern Turkey. Journal of the Geological Society, London, 152, 636-654. HIMMERKUS, F., REISCHMANN,T. & KOSTOPOLOUS,D. 2006. Late Proterozoic and Silurian basement units

409

within the Serbo-Macedonian Massif, northern Greece: the significance of terrane accretion in the Hellenides. In: ROBERTSON, A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 35-50. JACOBSHAGEN, V., DURR, S., KOCKEL,F., KoPP, K. O. & KOWALCZYK, G. 1 9 7 8 . Structure and geodynamic evolution of the Aegean region. In: CLOSS, H., ROEDER, D. H. & SCHMIDT, K. (eds) Alps, Apennines, Hellenides. Schweitzerbart, Stuttgart, 537-564. JACOBSHAGEN, V & WALLBRECHER, E. 1984. PreNeogene nappe structure and metamorphism of the North Sporades and the southern Pelion peninsula. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 591-602. JONES, G. & ROBERa'SON, A. H. F. 1991. Tectonostratigraphy and evolution of the Mesozoic Pindos Ophiolite and related units, Northwestern Greece. Journal of the Geological Society, London, 148, 267-288. KARAMATA, S. 2006. The geological development of the Balkan Peninsula related to the approach, collision and compression of Gondwanan and Eurasian units. In: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 155-178. KARAMATA, S., VUJNOVI(), L. 2000. Correlation of Palaeozoic units of the Dinarides and the northern part of the Vardar Zone. In: KARAMATA, S. & JANKOVIC, S. (eds) Proceedings of the International Symposium Geology and Metallogeny of the Dinarides and the Vardar Zone, 1. The Academy of Sciences and Arts of the Republic of Srpska, Banja Luka, 71-78. KAUFMANN, G., KOCKEL, F. & MOLLAT, H. 1976. Notes on the stratigraphic palaeogeographical position on the Svoula Formation in the innermost zone of the Hellenides (Northern Greece). Bulletin de la Societk Gkologique de France, 18(VII), 225-230. KILIAS, A. 1991 Tran.spressive Tektonik in den zentralen Helleniden Anderung der Translationspfade durch die Transpression (Nord-ZentralGriechenland). Neues Jahrbuch fiir Geologische und Paliiontologie, Monatshefte, II 5, 291-306. KILIAS, A., KASSELAS,G. & NASTOS, G. 1990. Quartz e-axis fabrics as a kinematic indicator of sense of nappe emplacement--an example from the N.E. Pieria mountain area (Greece). Zeitsehrift fiir Angewandte Geologische, 36(11), 427-433. KILIAS, A., FALALAKIS,G. & MOUNTRAKIS, D. 1999. Cretaceous-Tertiary structures and kinematics and their relation to the exhumation of the Hellenic hinterland (Macedonian, Greece). International Journal of Earth Science, 88, 513-531. KOCKEL, F. 1986, Die Vardar (Amos-) Zone. In: JACOSSHAGEN, V. (ed.) Geologic yon Griechenland. Borntraeger, Stuttgart, 270-168.

410

I.R. SHARP & A. H. F. ROBERTSON

KOCKEL, F., MOLLAT, H. & WALTHER, H. W. 1977. Erlgiuterungen zur geologischen Karte der Chalkidiki und angrenzender Gebiete 1:100000 (Nord Griechenland). Bundesanstalt ffir Geowissenschaften und Rohstoffe, Hannover. KOEPKE, J., SEIDEL, E. & KREUZER, H. 2002. Ophiolites of the southern Aegean Islands Crete, Karpathos and Rhodes: composition, geochronology and position within the ophiolite belts of the Eastern Mediterranean. Lithos, 65, 183-204. KOSTOPOLOUS, D. K. 1989. Geochemistry and tectonic setting of the Pindos ophiolite. PhD thesis, University of Newcastle upon Tyne. LE-PICHON, X. & ANGELIER, J. 1979. The Hellenic arc and trench system: a key to Neotectonic evolution of the Eastern Mediterranean. Tectonophysics, 60, 1~2. LIATI, A., GERBAUER, D. & FANNING, M. 2004. The age of ophiolitic rocks of the Hellenides (Vourinos, Pindos, Crete): first U-Pb ion microprobe (SHRIMP) zircon ages. Chemical Geology, 207, 171-188. LIPPARD, S. J., SHELTON, A. W. & GASS, I. G. 1986. The Ophiolite of Northern Oman. Geological Society, London, Memoir, 11. LIPS, A. t . W., WHITE, S. H. & WIJBRANS, J. R. 1998. 4~ laserprobe direct dating of discrete deformational events: a continuous record of early Alpine tectonics in the Pelagonian Zone, NW Aegean Sea, Greece. Tectonophysics, 298, 133-153. MANATSCHAL, G., MONTENER, O., DESMURS, L. & BERNOULLI, D. 2003. An ancient oceancontinental transition in the Alps: the Totalp, Err-Plata, and Malenco units in the Eastern Alps (Graubfinden and northern Italy). Eclogae Geologicae Helvetiae, 96, 131-146. MAVRIDES, m., SKOURTSIS-CORONOEOA,V. & TSALIAMONOPOLIS, S. 1979. Contribution to the geology of the Subpelagonian Zone (Vourinos area, West Macedonia). In: 6th Colloquium Geology of the Aegean Region, Athens, 175-195. Geological Society of Greece. MERCIER, J. 1966. Etude gdologique des zones internes des Hellknides en Mackdoine centrale (Grkce). Th6se, MERCIER, J. 1973. Plissements synmetamorphiques d'6chelle kilometrique, d'fige jurassique sup6rieureocretac6 dans les Hell6nides internes (Macedoine, Gr6ce). Cornptes Rendus de l'AcadOmie de Sciences, SOrie D, 276, 2249-2252. MERCIER, J. & VERGELY,P. 1972. Les m61anges ophiolitiques de Mac6donie (Gr6ce): d6crochements d'age ant6 Cretac6 sup6rieur. Zeitschrift der Deutschen Geologischen Gessellschaft, 123, 469489. MERCIER, J. & VERGELY, P. 1984a. Geological Map of Greece 1.'50 000 Sheet Edhessa. Institute of Geology and Mineral Exploration, Athens. MERCIER, J. & VERGELY, P. 1984b. Geological Map of Greece 1:50 000 Sheet Arnissa. Institute of Geology and Mineral Exploration, Athens. MERCIER, J., VERGI~LY, P. & BI~BIEN, J. 1975. Les ophiolites hell6niques obduct6 au Jurassique sup6rieur sont-elles les vestiges d'un oc6an tethysien

ou d'une mer marginale peri-europ6enne? Bulletin de la Societk Gkologique de France, 17, 108-112. MIGIROS, G. & GALEOS, A. 1990. Tectonic and stratigraphic significance of the Ano Garefi ophiolitc rocks (Northern Greece). In: MALPAS, J., MOORES, E., PANAYIOTOU, A. & XENOPHONTOS, C. (eds) Ophiolites; Oceanic Crustal Analogues. Proceedings of the Symposium 'Troodos 1987'. Geological Survey Department, Nicosia, 279-284. MILLER, J. McL., GRAY, D. R. & GREGORY, R. T. 1998. Exhumation of high-pressure rocks in northeastern Oman. Geology, 26, 235-238. MOUNTRAKIS, D. 1984. Evolution of the Palegonian zone in N.W. Macedonia. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 581-590. MOUNTRAKIS, D., TRANOS, M., PAPAEACHOS, C., THOMAIDOU, E., KARAGIANNI,E. & VAMVAKARIS, D. 2006. Neotectonic and seismological data concerning major active faults, and the stress regimes of Northern Greece. In: ROBERTSON,m. n . F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 649-670. MOUNTRAKlS, D., KILIAS, A., PAVLIDES, S., PATRAS, D. & SPYROPOULOS, N. 1987. Structural geology of the internal Hellenides and their role in the geotectonic evolution of the Eastern Mediterranean. Acta Naturalia, 23(4), 147-161. MUSSALAM, K. 1991. Geology, geochemistry and the evolution of an oceanic crustal rift at Sithonia, NE Greece. In: PETERS, T., NICOLAS, m. & COLEMAN, R. G. (eds) Ophiolite Genesis and the Evolution of the Oceanic Lithosphere. Kluwer, Dordrecht, 685-704. NANCE, P. 1981. Tectonic history of a segment of the Pelagonian zone, northeastern Greece. Canadian Journal of Earth Sciences, 18, I 111-1126, NAYLOR, M. A. & HARLE, T. J. 1976. Palaeogeographic significance of rocks and structures beneath the Vourinos ophiolite, northern Greece. Journal of the Geological Society, London, 132, 67-675. OKAY, A. I., SATIR, M., TUYSOZ, O., AKYOZ, S. & CHEN, F. 2001. The tectonics of the Strandja Massifi late Variscan and mid-Mesozoic deformation and metamorphism in the northern Aegean, International Journal of Earth Sciences, 90, 217-233. PAPANIKOLAOU, D. J. 1996-1997. The tectonostratigraphic terranes of the Hellenides. In: PAPANIKALAOU,D. & SASSI, F. P. (eds) Palaeozoic Geodynamic Domains and their Apidic Evolution in the Tethys, 1GCP, 276, 495-514. PARLAK, O. & ROBERTSON, A. H. F. 2004. Tectonic setting and evolution of the ophiolite-related Mersin M61ange, Southern Turkey: its role in the tectonic-sedimentary setting of Tethys in the Eastern Mediterranean Region. Geological Magazine, 141,257-286. PEARCE, J. A., LIPPARD, S. J. & ROBERTS, S. 1984. Characteristics and tectonic significance of

W MARGIN OF MESOZOIC VARDAR OCEAN, GREECE supra-subduction zone ophiolites. In: KOKELAAR, B. P. & HOWELLS, M. F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-89. PE-PIPER, G. & PIPER, D. W. J. 2002. The Igneous Rocks of Greece. The Anatomy of an Orogen. Beitrage zur Regionalen Geologie der Erde, 30, Borntraeger, Stuttgart. PICHON, J. F. 1976. Conditions de giskment des ophiolites sur la bordure occidental du Vermion (zone Pklagonienne, Grkece). Thbse, 3e cycle, Universit6 de Paris Sud, Centre d'Orsay. PICHON, J. F. 1977. Une traversale dans la zone P61agonienne, depuis les collines de Krapa (SW) jusqu'au massif du Vermion (NE): les premibres s6ries transgressives sur les ophiolites. Bulletin of the Geological Society of Greece, 1977, 163-171. PICHON, J. F. & BRIAN, J. H. 1985. An inverted metamorphism under the Vourinos ophiolite suite, Greece. Ofioliti, 10(2-3), 363-374. PICHON, J. F. & LYS, M. 1976. Sur l'6xistance d'une s6rie du Jurassic sup6rieur ~t Cretac~ inf6rieur surmont les ophiolites dans les collines de Krapa (Massif de Vourinos, Gr6ce). Comptes Rendus de l'AcadOmie de Sciences, Sdrie D, 282, 523-526. PURVIS, M. & ROBERTSON,A. 2004. A pulsed extension model for the Neogene-Recent E-W trending Al~ehir (Gediz) Graben, and the Miocene N E SW-trending Selendi and G6rdes Basins, western Turkey. Tectonophysics, 391, 171-201. RASSIOS, A., GRIVAS, E. & VACONDIOS, I. 1994. The geometry of structures forming around the ductilebrittle transition in the Vourinos-Pindos-Othris oceanic slab. Bulletin of the Geological Society of Greece, XXX(2), 109-121. RASSIOS, A. H. E. & MOORES, E. M. 2006. Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 237-266. RASSIOS, A. & SMITH, A. G. 2000. Constraints on the formation and emplacement of western Greek ophiolites (Vourinos, Pindos, 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 fieM studies and the Ocean Drilling Program. Geological Society of America, Special Papers, 349, 473-483. Rlcou, L.-E., BURG, J.-P., GODFRIAUX, I. & IVANOV, Z. 1988. Rhodope and Vardar: the metamorphic and the olistostromic paired belts related to the Cretaceous subduction under Europe. Geodinamica Acta, 11, 285-309. ROBERTSON, m. H. F. 1987. The transition from a passive margin to an Upper Cretaceous foreland basin related to ophiolite emplacement in the Oman Mountains. Geological Society of America Bulletin, 99, 633~553. ROBERTSON, m. H. F. 1990. Late Cretaceous oceanic crust and foreland basin development, Euboea, eastern Greece. Terra Nova, 2, 333-339.

411

ROBERTSON, A. H. F. 1991. Origin and emplacement of an inferred Late Jurassic subduction-accretion complex, Euboea, eastern Greece. Geological Magazine, 128, 27-41. ROBERTSON, A. H. F. 2006a. Contrasting modes of ophiolite emplacement in the Eastern Mediterranean region. In: GEE, D. & STEPHENSON, R. A. (eds) European Lithosphere Dynamics. Geological Society of London, Memoirs 32 (in press). ROBERTSON, A. H. F. 2006b. Sedimentary evidence from the south Mediterranean region (Sicily, Crete, Peloponnese, Evia) used to test alternative models for the regional tectonic setting of Tethys during Late Palaeozoic-Early Mesozoic time. In: ROBERTSON, A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 91-105. ROBERTSON, A. H. E., 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., 1984. Introduction: aspects of the geological evolution of the eastern Mediterranean. In: DIXON J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. ROBERTSON, A. H. F., DIXON, J. E., BROWN, S., et al. 1996. Alternative tectonic models for the Late Palaeozoic-Early Tertiary development of Tethys in the Eastern Mediterranean region. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ROBERTSON, A. H. F. & KARAMATA,S. 1994. The role of the subduction-accretion processes in the tectonic evolution of the Mesozoic Tethys in Serbia. Tectonophysics, 234, 73-94. ROBERTSON, A. H. F. & SHALLO, M. 2000. MesozoicTertiary tectonic evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-214. ROBERTSON, A. H. F., USTAOMER, T., PICKETT, E. A., COLLINS, A., ANDREW, T. & DIXON, J. E. 2004. Testing models of Late Palaeozoic-early Mesozoic orogeny in Western Turkey: support for an evolving open-Tethys model. Journal of the Geological Society, London, 161, 501-511. SCHERMER, E. R., LUX, X. & BURCHFIEL, B. C. 1990. Temperature-time history of subducted oceanic crust, Mount Olympos region, Greece. Tectonics, 9, 1165-1195. SCHERMER, E. R. 1993. Geometry and kinematics of continental basement deformation during the Alpine orogeny, Mt Olympos region, Greece. Journal of Structural Geology, 15, 571-591. 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.

412

I.R. SHARP & A. H. F. ROBERTSON

SENGOR, A. M. C. 1984. The Cimmeride Orogenic System and the Tectonics of Eurasia. Geological Society of America Special Papers, 195. SFEIKOS, A. 1992. Analysis of deformation and kinematics of the Pelagonian nappe system, Kamvounia mountains (North Thessaly, Greece). PhD Thesis, University of Tiibingen. SFEIKOS, A., BOHRINGER, CH., FRISCH, W., KILLIAS, A. & RATSCHBACHER, L. 1991. Kinematics of Pelagonian nappes in the Kranea area, North Thessaly, Greece. Bulletin of the Geological Society of Greece, 25(1), 101-115. Sharland, P., ARCHER, R., CASEY, D., et al. 2001. Arabian Plate Sequence Stratigraphy. GeoArabia Special Publication, 2. SHARP, I. R. 1994. The Triassic-Tertiary tectonic, sedimentary and magmatic evolution of the Pelagonian and Vardar (AMos) zones, Macedonia, Northern Greece. PhD thesis, University of Edinburgh. SHARP, I. R. & ROBERTSON, A. H. F. 1993. Evidence for Turonian rift-related extensional subsidence and eastward early Tertiary thrusting, western Paikon Zone, Northern Greece. Bulletin of the Geological Society of Greece, 28(1), 99-110. SHARP, I. R. & ROBERTSON, A. H. F. 1998. Late Jurassic-Lower Cretaceous oceanic crust and sediments of the Eastern Almopias Zone, NW Macedonia (Greece); implications for the evolution of the Eastern 'Internal' Hellenides. Bulletin of the Geological Society of Greece, 30(1 ), 47-61. SHARP, I. R., ROBERTSON, A. H. F. & DIXON, J. E. 1991. Tectonic and sedimentary history of the Eastern margin of the Pelagonian Zone, NE Greece. European Union of Geosciences, EUG VI, Strasbourg. Terra Abstracts, 3, 323. SMITH, A. G. 1993. Tectonic significance of the Hellenic-Dinaric ophiolites. 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, 213-243. SMITH, A. G. 2006. Tethyan ophiolite emplacement, Africa to Europe motions, and Atlantic spreading. In: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 11-34. SMITH, A. G., HYNES, A. J., MENZIES, M., NISBET, E. G., PRICE, I., WELLAND,M. J. & FERmERE, J. 1975. The stratigraphy of the Othris Mountains, Eastern Central Greece: a deformed Mesozoic continental margin sequence. Eclogae Geologicae Helvetiae, 68, 463-481. SMITH, m. G.,WOODCOCK, N. H. & NAYLOR, M. m. 1979. The structural evolution of a Mesozoic continental margin, Othris Mountains, Greece. Journal of the Geological Society, London, 136, 589-603. STAIS, A. 1994. Evolution gkodynamique des bassins Mksozoiques Vardariens (domaines de Peonias et d'Almopias), Helldnides orientales. PhD thesis. University of Lille. STAIS, A. & FERRII~RE, J. 1991. Nouvelles donn6es sur la pal6og6ographie M6sozoique du domain

Vardarien: les bassins d'Almopias et de P6onias (Macedoine, Hell6nides internes septentrionales). Bulletin of the Geological Society of Greece, 26(1), 491-507. STAlS, A. & FERRIERE, J., CARIDROIT, M., DE WEVER, P., CLEMENT, B. & BERTRAND, J. 1990. Donn6es nouvelles sur l'histoire ante-obduction (TriasJurassique) du domaine d'Almopias (Mac6doine, Gr6ce). Comptes Rendus de l'Acadkmie de Sciences, Skrie II, 310, 1275-1480. STAMPFLI, G. M. & BOREL, G. D. 2002. A plate tectonic model for the Palaeozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrones. Earth and Planetary Science Letters, 169, 17-33. STAMPFLI, G. M., MOSAR, J., FAURE, P., PILLEVUIT,A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethys realm: the Neotethys East Mediterranean basin connection. In: ZIECLER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUINSOLEAU, S. (eds) Peri-Tethys Memoir 5. PeriTethyan Rift~Wrench Basins and Passive Margins. M6moires du Mus6um National d'Histoire Naturelle, 51-108. STAMPFLI, G., M., VAVASSIS, I., DE BOND, A., ROSSELET, F., MATTI, B. & BELLINI, M. 2003. Remnants of the Paleotethys oceanic suture-zone in the western Tethys area. Bolletino della Societgt Geologica Italiana, Special Volume, 2, 1-23. USTAOMER, P. A., MUNOIL, R. & RENNE, P. R. 2005. U/Pb and Pb/Pb zircon ages for arc-related intrusions of the Bolu Massif (W Pontides, NW Turkey): evidence for Late Precambrian (Cadomian) age. Terra Nova, 17, 215-223. USTAOMER, T. & ROBERTSON,A. H. F. 1997. Tectonicsedimentary evolution of the North Tethyan margin in the Central Pontides of Northern Turkey. In: ROBINSON,A. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists, Memoirs, 86, 255-290. VERGELY, P., 1984. Tkctoniques des ophiolites dans les Hgllenides Internes (deformation, mktamorphismes et phenomknes skdimentaires). Consequences sur l'kvolution des region Tkthysiennes Occidentales. PhD thesis, Universit6 de Paris-Sud, Orsay. WALCOTT, C. R. 1996. The Alpine Evolution of Thessaly ( N W Greece) and Late Tertiary Aegean Kinematics. Geologica Ultraiectina (Utrecht), 162. WALCOVT, C. R. & WHITE, S. H. 1998. Constraints on the kinematic of post-orogenic extension imposed by stretching lineations in the Aegean region. Tectonophysics, 298, 155-176. YARWOOD, G. A. & AFTALION, M. 1976. Field relationships and U-Pb geochronology of granite from the Pelagonian zone of the Hellenides. Bulletin of the Geological Society of Greece, 7, 259-264. YARWOOD, G. A. & DIXON, J. E. 1977. Lower Cretaceous and younger thrusting in the Pelagonian rocks of the High Pieria, Greece. 6th Colloquium of the Aegean Region (Athens 1977), 269-280. ZIMMERMAN, J. 1972. Emplacement of the Vourinos Ophiolitic Complex, Northern Greece. Geological Society of America, Memoirs, 132, 225-239.

Late Cretaceous-Early Cenozoic tectonic evolution of the Eurasian active margin in the Central and Eastern Pontides, northern Turkey S A M U E L P. R I C E 1'2, A L A S T A I R

H . F. R O B E R T S O N

1& T I M U R

USTAOMER 3

1Grant Institute o f Earth Science, School of GeoSciences, University of Edinburgh, West Mains Road, Edinburgh EH9 3JW, UK 2present address." CASP, Department of Earth Sciences, West Building, 181a Huntingdon Road, Cambridge CB3 0DH (e-mail." sam.rice@casp, cau. ac. uk) 3Istanbul University, Jeoloji Bgliimii, Mineral@-Petrografi Anabilim Dab, 34850, Avczlar Istanbul, Turkey The Izmir-Ankara-Erzincan suture zone (IAESZ) in the Central and the Eastern Pontides comprises a stack of thrust sheets of mainly Late Cretaceous-Early Cenozoic age that are restored as: (1) a subduction-accretion complex; (2) a continental-margin magmatic arc, plus an associated forearc basin; (3) a back-arc basin and its mainly sedimentary fill. Northward thrusting affected all of the Late Cretaceous units during latest Cretaceous (Campanian-Maastrichtian) time. This was followed by regional southward thrusting to form the present thrust stack during Mid-Eocene time. Alternative tectonic models are considered in the light of sedimentary, igneous geochemical and structural evidence, and global comparisons. We infer that the Northern Neotethys was subducted northwards beneath the Eurasian active margin during the Late Cretaceous. Subduction was associated with the genesis of a magmatic arc and a related forearc basin. The subduction zone retreated oceanwards, associated with the opening of a back-arc basin along the Eurasian margin, floored by oceanic crust and overlain by mixed terrigenous and volcaniclastic deep-marine sediments. Ophiolite genesis in a continental margin back-arc setting is suggested by the presence of screens of basement-type metamorphic rocks within an ophiolite-related sheeted dyke complex in the Eastern Pontides. During the latest Cretaceous closure of the inferred back-arc basin resulted in northward emplacement of ophiolitic and related units onto the Eurasian margin, as well exposed in the Central Pontides. In addition, accretionary m61ange, volcanic arc, forearc and ophiolitic units were emplaced southwards onto the Tauride continent, represented by the Munzur platform in the Eastern Pontides, also during latest Cretaceous time. This incipient ('soft') collision was followed by widespread Paleocene-Early Eocene deposition of Nummulitic shelf carbonates and coarse clastic sediments on deformed and emplaced accretionary m61ange, arc and ophiolitic units. Final closure ('hard collision') of the Northern Neotethys occurred during the Mid-Eocene, resulting in large-scale southward imbrication, together with northward backthrusting in some areas. Suture tightening and Plio-Quaternary strike-slip ensued. Abstract:

During recent years most studies of the origin and emplacement of Tethyan ophiolites have focused on South Tethyan settings, where oceanic lithosphere (e.g. Semail and Troodos ophiolites) was emplaced onto G o n d w a n a and its satellite microcontinents (e.g. Robertson 2002; Garfunkel 2006; Koller et al. 2006; Rassios & Moores 2006; Smith 2006). M a n y of these ophiolites formed in a suprasubduction-zone (SSZ) setting within a Southern Neotethyan oceanic basin (e.g. Parlak et al. 2004; Rlzao~lu et al. 2006). There have been few studies of ophiolites and related units that were formed in more northerly Neotethyan oceanic basins (Northern Neotethys) and emplaced onto the Eurasian continental margin. The Eurasian margin exhibits a long history of tectonic

and magmatic events that can be mainly related to active margin processes, including subduction, accretion and arc volcanism ($eng6r & Ydmaz 1981; Robertson & Dixon 1984; Dercourt et al. 1986; Stampfli et al. 2001). Several workers have recognized that ophiolites were emplaced in the Pontides during Late Cretaceous-Early Cenozoic time and alternative tectonic scenarios have been proposed (Bergougnan 1975; Ydmaz 1985; Ko~yi~it 1990; Tiiysiiz 1990; Okay & Sahintiirk 1997; Usta6mer & Robertson 1997; Ydmaz et al. 1997). However, information on the age, structure and geochemistry of these units has remained sparse. To help address this deficiency we have studied the IzmirAnkara-Erzincan suture zone (IAESZ) in two

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 413-445. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

414

S.P. RICE ETAL.

J

'0o

"0o BLACK SEA

Thrace

" ~

]

_-.----.-.--.~ Studyareas ] @~'!z ~'~Central ~.~"~ ~ ~'~ _'~,,.--,-..-,,_-/~..-'- / ~Pontides/,,,,"/~ _ ~ ..,/./-/.,,-~ ~ - - ~ " ' - ~ ~ / / / F i ~ . 2 ~ ~ & / ~ ~ ~ i ~ e { f 40 East Anatolian < ~'7~lXrr-'.~"l~/'/7* Kirsehir ~ l E r z i n c a n Accretionary .~q. I ' ]" ;' I I ' / ~ • massif' ~ S l-lxa ~u'hzuR' ~ complex e LATFO RM,~/,"~,...eL,l~

i

<

'~L--I[-J"-AntalyaS'uture :" ,--__~ "" ' ~m-' " ' ~ L ~ ' ~ ' - ~ l i ,

-34~ N 0 Km 200 t 9 |

', I i II IIIII

I

i ArabianPlatformlllllllrl'N

~

Mamonia Cyprus IAESZ =lzmir-Ankara-Erzincan Suture Zone m Late Cretaceous-Tertiary suture zones Suture ~ Pontides MEDITERRANEAN SEA , ~~ Taurides Ktr#ehir massif 30~ ~ ,~,, Arabian .Pla.tform I

Fig. 1. Outline geological map of Turkey showing the regional settings of the Central and Eastern Pontide areas studied (marked with boxes). specific areas of the Central and Eastern Pontides (Fig. 1). The IAESZ separates Gondwana-derived (i.e. Tauride-Anatolide) blocks from Eurasia (i.e. Pontides) and records the closure of the main oceanic strand of the Northern Neotethys ocean ($eng6r & Yflmaz 1981; Robertson & Dixon 1984; Dercourt et al. 1986; Okay et al. 2001). This suture zone is well exposed as a dominantly south-vergent thrust belt running across northern Anatolia (Fig. 1). In the north, Late Mesozoic units of the IAESZ are tectonically imbricated with Eurasia-related metamorphic basement rock, represented by the Karakaya Complex and related units. In the south these units structurally overlie continental units that were rifted from Gondwana, including the Munzur Platform (part of the Tauride-Anatolide Platform) in the Eastern Pontides, and the Klr~ehir Massif (probably a separate microcontinent) in Central Anatolia (Fig. 1). In this paper we will combine sedimentary, igneous, structural and palaeontological evidence to infer the tectonic settings of formation of the various units of mainly Late CretaceousEarly Cenozoic age within the suture zone. More detailed information (e.g. detailed maps, sections and logs) has been given by Rice (2005). We will propose an interpretation involving northward subduction beneath the Eurasian margin during

Late Cretaceous time, coupled with arc volcanism and the opening, then closure, of a marginal basin. We will also draw comparisons with modern and ancient marginal basins. The recently published time scale of the International Commission on Stratigraphy (Gradstein et al. 2004) is used throughout. Existing formation names are retained wherever possible but new names are introduced in several cases.

Central Pontides Seven main tectonostratigraphic units of mainly Late Cretaceous age are identified in a wellexposed area of the Central Pontides (Fig. 2), as summarized below in upward structural order. Each of these units is bounded by south-vergent thrusts and is unconformably overlain by Neogene-Recent units. E s k i k g y Formation: cover o f Pontide basement

The Eskik6y Formation (new name; Fig. 3a) (c. 150 m thick) rests with a low-angle unconformity on a thin unit of pisolitic bauxite that caps the schistose metamorphic basement of the Pontides (Kargl Complex; Usta6mer & Robertson 1997). This formation begins with thickly bedded

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

415

Fig. 2. Simplified geological map of the Central Pontides (see Fig. 1 for location). The locations of measured logs are shown (see Fig. 3a-e), also the line of section a-a' (see Fig. 4). The locations of sites sampled for geochemical analysis by XRF are also marked: A, Yaylagayl Formation: B, Iki~am Formation: C, Klzahrmak Ophiolite (see also Fig. 6 and text for discussion).

(50 cm), grey microcrystalline pelagic limestones, c. 10 m thick, that contain planktonic Foraminifera (e.g. Globotruncana sp.) of Campanian age (Tiiysfiz et al. 1988). Sheared and isoclinally folded turbiditic sandstones with shaly partings, c. 150 m thick, include occasional poorly sorted, matrix-supported conglomerates (debrites). The sandstones exhibit flute casts and Thalassinoides

sp. bioturbation. The conglomerates contain well-rounded pebbles of paraquartzite, presumably derived from the Pontide basement to the north. In thin section, the sandstones contain polycrystalline quartz, schistose lithoclasts and detrital white mica, indicating a metamorphic source. The succession passes upwards into thinbedded red shale and radiolarian chert. Towards

416

S.P. RICE E T A L .

d o

2 .9 O

g ea)

2

o

9

o

~.,eq

O

o

/::~0 o o ~

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY the top of the formation the sediments are thrustimbricated with ophiolitic m61ange (Kirazba~l M61ange; see below). The EskikSy Formation is also exposed as thin (c. 40 m) thrust slices within more northerly outcrops of the m61ange (e.g. near Ilgaz; Fig. 2). The pelagic lower part of the Eskik6y Formation is interpreted to record subsidence of the Eurasian margin during Campanian time (83.5-70.6 Ma). The terrigenous gravitydeposited sediments were shed from the Pontide basement into a subsiding deep-water basin. This was probably a flexurally controlled foredeep, associated with the northward emplacement of ophiolites and related units onto the Eurasian margin.

Upper Cretaceous Kzzzhrmak Ophiolite The K1zlllrmak Ophiolite (Ttiystiz 1990), c. 3.5 km thick, occurs at several different structural levels within the thrust stack (Figs 2 and 4). Outcrops of deformed ophiolitic rocks, 2-5 km thick, are found near Bayat and Eldivan, and also along the banks of the Klzlhrmak River, at Pelitcik, near Kargl (Fig. 2). The ophiolitic complex includes serpentinized harzburgite, cumulate pyroxenite, dunite, isotropic and layered gabbro, and deformed greenschist-facies metabasaltic pillow lava (Fig. 3b). A Campanian-Maastrichtian age (83.5-65.5 Ma) for the ophiolite is indicated by the presence of Globotruncana linneiana (d'Orbigny) within pelagic limestones interbedded with basaltic pillow lavas (Tfiysiiz 1990) that are well exposed along the Klzlhrmak river valley, north of Kamil (A~ikbfikfi-Yukan Zeytin section). Basalt from this locality was collected for analysis (see below). The Klzlhrmak Ophiolite is interpreted as an incomplete section of oceanic lithosphere of Late Cretaceous age. Supporting geochemical evidence is presented later in the paper.

Upper Cretaceous ikifam Formation." marginal basin sediments A very thick sedimentary unit, the iki~am Formation (new name; > 3000 m thick), is found within the middle to upper structural levels of the imbricate thrust stack (Fig. 4). A Late Cretaceous (Campanian-Maastrichtian) age is indicated by the presence of several species of Globotruncana (i. Ongen, pers. comm.). The lower part (e. 2 km) of the formation comprises thinly bedded (beds < 4 cm), buff-coloured micritic, to muddy limestones with thin grey pelitic schist partings and tuff (Fig. 3c). Higher

417

levels of the succession comprise coarser-grained quartzo-feldspathic sandstones, volcaniclastic sandstones and sericitic shales. There are also interbeds of turbiditic calcarenites that contain volcanic grains and feldspars of probable magmatic origin (Fig. 5). Grains of undeformed quartz, polycrystalline quartz (i.e. quartzite) and schistose lithoclasts are also present. In the north the succession includes occasional igneous sills, massive lava flows and rare pillow lavas, as seen near Tosya (see below for chemical analysis). The lavas contain large ( < 2 cm) phenocrysts of biotite, analcite and salite (alkali pyroxene), suggesting a markedly alkaline composition. In the south the iki~am Formation interdigitates with andesitic lavas and coarse andesitic volcaniclastic conglomerates and sandstones, correlated with the Yaylaqayl Formation, an Upper Cretaceous inferred volcanic arc unit, also exposed near Tosya (see below; Fig. 2). The more basic extrusive rocks within the lkigam Formation were collected for chemical analysis (see below). However, extrusive rocks in the south are chemically too evolved to allow geochemical discriminant analysis, although their tectonic setting of eruption can be inferred from an interfingering relationship with the Yaylagay~ Formation inferred arc unit. The ikigam Formation is unconformably overlain by sediments of Late Eocene or or younger age, which were deposited after the inferred suturing of the Northern Neotethys. The Upper Cretaceous lki~am Formation represents a volcanically active deep-water slope setting, with an increasing abundance of texturally immature volcanic and terrigenous sediments upwards. The turbiditic sandstones range from terrigenous, to volcaniclastic and calcareous in composition. The Pontide metamorphic basement to the north is the obvious source for the terrigenous material. The polycrystalline quartz is interpreted as metachert derived from pre-Jurassic accretionary complexes in the Pontides (UstaSmer & Robertson 1997). Deposition was accompanied by sparse alkalineperalkaline volcanism in the north that could be extension related. In addition, the basalticandesitic composition of the interfingering volcanic rocks and coarse sediments in the south suggests that this material was derived from an adjacent magmatic arc unit. Overall, the iki~am Formation is interpreted as the emplaced sedimentary-volcanic fill of a deep-water backarc basin that formed between the Pontide continental margin to the north and an active volcanic arc to the south.

418

S.P. RICE ETAL.

Fig. 4. Cross-section showing the main structural and stratigraphic relations in the Central Pontides. (See Fig. 2 for line of section). It should be noted that some variation is present along strike (see Fig. 2).

Upper Cretaceous Yaylaqayt Formation: volcanic arc unit Two major thrust slices of volcanic and volcaniclastic rocks, each up to c. 4 km thick, occur at two different structural levels (Figs 2-4). The presence of several species of Globotruncana

Fig. 5. Photomicrograph of Upper Cretaceous sandstone from the Iki~am Formation. The presence of terrigenous material (white mica, quartzite), glassy basic lava and chert should be noted.

within interbedded sediments indicates a Late Cretaceous (Campanian-Maastrichtian) age for this unit (Tiiysiiz et al. 1995). The Yayla~;ayl Formation (Yolda~ 1982) begins with basaltic pillow lavas and pelagic interpillow sediments and passes gradually upwards into a very thick succession (up to 3500m in apparent thickness) of andesitic lava and coarse matrix-supported volcaniclastic conglomerates (Fig. 3d). Both the clasts and the matrix of these conglomerates are of intermediate composition, based on petrographic evidence. Felsic volcanic rocks, intrusive rocks and altered tuff are present in lesser amounts. Stratigraphically higher levels are dominated by volcaniclastic sandstones and shales that grade into pale grey thinly bedded shaly and micritic limestones. At higher levels of the thrust stack volcaniclastic and metavolcanic schists, also attributed to the Yayla~ayl Formation, exhibit well-developed metamorphic fabrics and a greenschist-facies mineral assemblage (i.e. epidote, talc, quartz, albite). The highest stratigraphic levels of the formation are transitional to pelagic limestones with Globotruncana in the iskilip area, whereas further west near (~anklrl (Fig. 2) the formation is unconformably overlain by a shallow-marine sedimentary cover (Yaprakh Formation; see below).

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY Within the Yaylagayl Formation, the volcanic rocks become generally more evolved stratigraphically upwards, from basaltic, to andesitic, then felsic (e.g. rhyodacite). Pelagic sediments are present between pillow lavas low in the succession but above this texturally immature volcaniclastic sediments predominate. The formation is interpreted to record a fragment of a volcanic arc, which developed away from a supply of terrigenous sediment. A gradual passage from volcanic rocks to volcaniclastic sediments records a waning of volcanism, or a switch in the locus of volcanism to a more distal location. The arc edifice was eroded following cessation of volcanism. The pelagic limestones at the top of the Campanian-Maastrichtian succession indicate a reduced supply of volcaniclastic material, possibly caused by cessation of arc magmatism, tectonic subsidence, or a eustatic relative sea-level rise.

Upper Cretaceous Yaprakh Formation." arc apron-forearc basin A sedimentary cover unit up to 500 m thick, known as the Yaprakh Formation (Birgili et al. 1975), unconformably overlies the arcrelated Yayla~ay~ Formation, as exposed near ~ a n k m (Figs 2 and 3e). A Late Cretaceous (Campanian-Maastrichtian; 83.5-65.5 Ma) age is indicated by the presence of planktonic foraminifera, including Globotruncana linneiana (d'Orbigny) (l. Ongen, pers. com.). The formation begins with a thin basal conglomerate (r 3 m), followed by grey volcaniclastic shales. The shale passes upwards into thick-bedded, coarse-grained calcarenites, volcanogenic shales, thick-bedded volcaniclastic sandstones and conglomerates, with minor grey fissile shaly partings. These sediments contain poorly sorted grains of mafic- and intermediate-composition volcanic rocks, quartz, feldspar, glauconite and abundant calcareous shell fragments. The detrital grains within individual samples range from subangular to well rounded and show evidence of textural inversion (i.e. well-rounded but poorly sorted grains). Large bivalves ( < 10 cm; probably rudists) are locally present. Individual beds, up to 1.5 m thick, are commonly massive or graded and contain subrounded pebbles and boulders of feldspar-phyric andesite (up to 40 cm in diameter). The unconformable base of the succession and the thin basal conglomerate together indicate that at least some erosion of the underlying volcanic arc unit has occurred, probably in a subaerial setting. This was followed by re-submergence

419

and fine-grained deposition. The upward change from homogeneous grey shales to texturally immature lithologies suggests a relatively proximal source for the volcaniclastic and carbonate material. The large bivalves, well-rounded grains and the presence of glauconite suggest a shallowmarine setting. However, the existence of textural inversion confirms that some redeposition has occurred. The clastic sediments probably formed from sheet-like density flows within broad (c. 30 m) submarine channels. By contrast, the nature of the fine-grained shaly partings suggests low-energy background accumulation in a relatively deep-water setting. Measurements of crosslaminations and pebble imbrication yielded (dip-corrected) palaeocurrent directions towards the NE (Fig. 3e). The Upper Cretaceous Yaprakh Formation is interpreted as part of the northern edge of a forearc basin that is mainly buried beneath the younger (~ankm Basin to the south (Kaymak91 2000). Contemporaneous volcanogenic material (e.g. air-fall tuff) is absent. This unit probably records reworking of arc-derived material after arc volcanism had ended (at least locally), but prior to final tectonic emplacement. The shallowwater carbonate, including large bivalves, was derived from carbonate build-ups on, or around, arc edifices that were partially eroded after volcanism ended. Terrigenous sediment (e.g. metamorphic quartz) is notably absent, in common with the underlying arc unit.

Upper Cretaceous Kirazba~;z Mklange: ophiolitic mdlange An ophiolitic m61ange unit, the Kirazbasl M61ange (Tfiysfiz 1990), is widely distributed throughout several structural levels of the suture zone (Figs 2 and 4). In the north the m61ange tectonically overlies the Upper Cretaceous Eskik6y Formation, interpreted above as a foredeep succession. The m61ange is overlain unconformably by Eocene Nummulitic limestones and fluviodeltaic sandstones; these belong to the Kadlklzl Formation, which postdates suturing. The m61ange exhibits a block-against-block fabric without any matrix of sedimentary origin. The most common lithologies of the m61ange are serpentinized ultramafic rocks, basalt-metabasalt, dolerite, red radiolarian chert, pelagic-hemipelagic limestone, shale, volcaniclastic-terrigenous sandstone and neritic limestone. The size of the blocks ranges from centimetres to hundreds of metres. There are also dismembered thrust sheets up to several kilometres long. Locally, a red-brown matrix is

420

S.P. RICE ET AL.

present, which is poorly sorted and contains fragments of all of the above lithologies; this matrix is interpreted as of tectonic origin. The extrusive and sedimentary blocks in the m61ange occur as two main lithological associations: (1) basalt-chert; (2) basalt-volcaniclastic sediment-neritic carbonate. Calcareous microfossils from several of the blocks yielded ages ranging from Albian to Maastrichtian based on microfossils including Pseudosiderolites vidali Douville and Rotalipora sp. (1. Ongen, pers. comm.). The formation of the m61ange is assumed to be approximately coeval with the youngest known age of the blocks in the m61ange (i.e. Late Maastrichtian). In addition, the exposed m61ange predates unconformably overlying Eocene sediments (Kadlkxzl Formation; see below). The m61ange is interpreted as an accretionary prism related to northward subduction of the Northern Neotethys. The pelagic sediments (radiolarites and rare pelagic carbonates), now present as blocks, were originally deposited on oceanic crust and were later accreted into the m61ange. Accordingly, the age of the underlying oceanic crust was at least Albian-Late Maastrichtian (112-65.5 Ma). The blocks and slices of the basalt-chert lithological association and also serpentinite were accreted from Neotethyan oceanic lithosphere, whereas the blocks and slices of the basalt-volcaniclastic sediment-neritic carbonate association are interpreted as fragments of emplaced oceanic seamounts (Rojay et al. 2001). The dominance of oceanic material and the lack of a terrigenous matrix suggests that the accretionary wedge developed some distance from the Pontide continental margin to the north.

Middle Eocene Kadtktzt Formation: post-suture sediments The oldest rocks above the Neotethyan IAESZ are represented by the KadlklZl Formation of Mid-Eocene age. This formation overlies the Kirazba~l M61ange with an irregular unconformity, as exposed north of Tosya (Fig. 2). The succession begins with Nummulitic limestones grading into calcarenites and shales, and then passes upwards into sandstones and lenticular conglomerates, which increase in abundance towards the top of the formation. The sandstones are poorly sorted, terrigenous litharenites with well-rounded clasts, indicating reworking in a fluvial or shoreface environment. In addition, the presence of plant-derived material within these sediments suggests the proximity of fluvial input. The presence of coarse-grained Nummulitic calcarenites further indicates that this formation was deposited in a shelf-type

setting. The conglomerates exhibit a lenticular geometry and an imbricated clast-supported texture, dominated by metamorphic lithoclasts, which suggests deposition in channels mainly fed from the Pontide continental margin to the north. The overall regressive nature of the succession suggests a transition from a shallow carbonatedepositing shelf ( < 200 m: lower shoreface) to a proximal subaqueous delta that was constructed directly on the Kirazba~l M61ange during Mid-Eocene time. Unlike the Upper Cretaceous units described above, the Kad~km Formation is relatively undeformed. It lacks a penetrative cleavage and does not exhibit north-vergent deformational fabrics, as seen in the underlying units. This suggests that the emplacement of these underlying units took place prior to the Mid-Eocene.

Geochemistry of Central Pontide basaltic rocks and peridotites Analytical methods Samples of relatively unaltered basaltic rocks were collected from the Kazfllrmak Ophiolite, the Kirazba~l M61ange, the Yaylaqayl volcanic arc unit and from the ikiqam, inferred back-arc unit and were analysed for major and trace elements by X-ray fluorescence (XRF) at the School of GeoSciences, University of Edinburgh, using the method of Fitton et al. (1998). In addition, chrome spinel grains from serpentinized harzburgites taken from both the m61ange (several samples) and the K m h r m a k Ophiolite (one sample) were analysed using a Cameca SX100 electron microprobe at the School of GeoSciences, University of Edinburgh, using the method of Reed (1975). A focused beam of 20 nA was used. The instrument was fitted with five wavelength-dispersive spectrometers and operated with a gun potential of 20 kV. Probe current was measured using a Faraday cup. The analytical standards were a selection of Specpure metals, simple synthetic oxide crystals and simple silicates. Fe 3+ concentrations were calculated stoichiometrically following the method of Droop (1987).

Results o f basalt chemical analysis The extrusive rocks are andesite, andesite-basalt, sub-alkali basalt and alkali basalt, as shown in Figure 6a. The compositions of the ophiolitic and volcanic arc basalts are also illustrated using normal-mid-oceanic ridge basalt (N-MORB)normalized 'spidergrams' (Fig. 6b) (Pearce et al. 1984). Selected analyses are shown in Table 1.

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY (a)

c.... d,toIt

1

421

b(i) Yayla(;ayl Formation

~llerite/~onolit

100]

e

""y~ \ / 0.1

~ / X ~ Trachyte Rhyodacite] ~ Dacite I Tra.chy:~ I* Yaylac~lylFormation .~ ancles,te / / ~ ~ Ikit;am Formation Antic'ire +J S - ' - , / I" KEihrrrek Ophiolite zx ~ - i - ~ ' ~ " * [ n DI Basanite A n d ~ I I nephelinite . . basalt I Alkali-basalt ._ Sub-alkaline

._ ~q

10

"~_

0.01

1

-~ ~

~/"~

0.1

i

OOOl O.Ol

0.01 o.1

1

~o

Sr

K

Rb Ba Nb Ce Nd

P

Zr

Ti

Y

Sc Cr

Nb/Y

b(ii)

b(iii) K i z d l r m a k Ophiolite 100

ikis

Formation

100 10 1

0.1

0.1

0.01

I

Sr

K

I

I

I

I

I

Rb Ba Nb Ce Nd

I

I

I

I

P

Zr

Ti

Y

I

Sc Cr

0.01

i

Sr

K

,

,

i

i

,

Rb Ba Nb Ce Nd

i

,

,

J

P

Zr

Ti

Y

,

,

Sc Cr

Fig. 6. Chemistry of Central Pontide basaltic rocks. (a) Zr/Ti v. Nb/Y diagram showing the range of lithologies present; (b) MORB-normalized trace-element patterns of basaltic rocks from the Central Pontides. (i), Yayla?ayl Formation (volcanic arc); (ii), Iki?am Formation (marginal basin sediments); (iii), Klzd]rmak Ophiolite (SSZ-type oceanic crust). Normalizing values: St, 120 ppm; KzO, 0.15%; Rb, 2.0 ppm; Ba, 20 ppm; Nb, 3.5 ppm; La, 3 ppm; Ce, 10 ppm; Nd, 8 ppm; P205, 0.12% ; Zr, 90 ppm; TiO2, 1.5%; Y, 30 ppm; Sc, 40 ppm; Cr, 250 ppm (Pearce 1982). (See Fig. 2 for sampling locations.)

Most of the basalts of the Yayla?ayl Formation, inferred arc unit are enriched in large ion lithophile elements (LILE; e.g. Sr, K, Rb, Ba) relative to the less mobile high field strength elements (HFSE; e.g. Ti, P, Zr, Nb). This enrichment is accompanied by a marked negative Nb anomaly in most samples (Fig. 6b, i). These features are characteristic of a mantle source that was chemically affected by subduction fluids (e.g. Pearce & Cann 1973; Pearce et al. 1984). The extrusive rocks of the Yayla?ay~ Formation are chemically similar to those of modern subduction-related volcanic arcs (e.g. Mariana, SW Pacific; Pearce 1982). The LILE enrichment is attributed to the effects of fluids derived from the downgoing slab as it underwent P T controlled phase changes and associated dehydration reactions (Anderson et al. 1978; Saunders et al. 1980). By contrast, the samples o f the Iki?am Formation, from the more northerly part of the

inferred back-arc unit, do not exhibit any identifiable subduction-influenced geochemical signature and are compatible with a within-plate setting (e.g. rift-related or seamount; Pearce 1982; Fig. 6b, ii). MORB-normalized plots of basaltic rocks from the Klzdlrmak Ophiolite (Fig. 6b, iii) are slightly enriched relative to MORB, and show a slight negative Nb anomaly. The low Cr values suggest a fractionation effect. Chemically similar basalts are known from many Tethyan ophiolites (e.g. Pearce et aL 1984; Robertson 2002; Parlak et al. 2004). Basalts exhibiting similar MORBnormalized plots have been dredged from modern back-arc basins (e.g. Weaver et al. 1979; Taylor et al. 1992). The basaltic rocks from the Kirazba~l M61ange exhibit trace-element signatures that suggest a range of mid-oceanic ridge to withinplate-type settings not influenced by subduction (Rice 2005).

S . P . R I C E E T AL.

422

~

.

~

~

~

.

m

~

o

~

~

~

~

8o

0

z

o o ,..~i ,_o

~g el,.-.-'

t',l

o

8~ ?-

g

~z oo

r

{

o ~ , . . ~~

...~ oO

~0

a r

.~el

~Z

~== ~o~ ~Z

M

e,i

~

ki o -~'~

t~

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

Results of chrome spinel analysis

423

Six tectonostratigraphic units of Late Cretaceous-Early Cenozoic age were identified in the well-exposed area of the IAESZ in the Eastern Pontides (Fig. 8). As for the Central Pontides, these units will be described from north to south in generally downward structural order. The Munzur Mountains (Tauride) platform unit in the far south, described last, is located at the lowest structural level (Figs 8 and 9).

Erzincan, mainly at a high structural level (Figs 1 and 8). Smaller ophiolitic exposures also occur at a much lower structural level, near the Munzur platform (Fig. 8). The base of the Refahiye Complex is a north-dipping thrust (Fig. 9); its upper boundary is an unconformity with overlying Eocene sediments (Sipik6r Formation) or younger units. The complex, with an estimated apparent thickness of c. 8 km, is composed of >75% (by volume) serpentinized harzburgite, c. 20% diabase dykes and < 5% trondhjemite (plagiogranite) dykes. The diabase includes thrust slices of sheeted dykes up to 1000 m thick. The boundaries of the individual thrust sheets of sheeted dykes are commonly zones of sheared serpentinite, up to c. 30 m thick. There are also isolated dykes, individually < 2 m thick, within serpentinized harzburgite (Fig. 10a). An important observation is that the sheeted dykes include numerous elongate screens, each up to c. 50 m thick, that are composed of highly strained metamorphic rocks. These screens locally dominate the dyke section in the SW of the complex (Fig 11) and include epidote-actinolite schist, metabasite, metaserpentinite and massive marble. The individual screens are intruded by swarms of diabase dykes, together with rare plagiogranite dykes and late-stage aplite dykelets, up to 30 cm wide (Fig. 11). Undisturbed primary igneous contacts are preserved between many of the dykes and the host rocks. The exposure of the Refahiye ophiolitic complex in the north is interpreted as a dismembered section of oceanic lithosphere, of which the upper, extrusive levels are not now preserved. The metamorphic host rocks within the Refahiye Complex are lithologically comparable with the Late Palaeozoic-Early Mesozoic metamorphic basement of the Pontides (e.g. Domuzda~ Unit; Topuz et al. 2004), and are seen as fragments of country rocks that were rifted from the Pontide basement and incorporated into the ophiolite complex. An alternative origin as fragments of an older dyke-rich metamorphic basement, as known from some parts of the Pontides (eg. Artvin region; T. Usta6mer, pers comm.), is unlikely in view of the close association of the metamorphic rock screens with the nearby 100% sheeted dyke sections of the Refahiye Complex. An important implication of the dyke-rich metamorphic rock screens is that the Refahiye ophiolitic complex formed in a rifted continental margin, rather than an oceanic setting.

Upper Cretaceous Refahiye Complex: Neotethyan oceanic crust

Upper Cretaceous Karada~ Formation." oceanic arc

An ophiolitic unit termed the Refahiye Complex (Yflmaz 1985) crops out north and NW of

This volcanic and volcaniclastic unit crops out at two structural levels, located west and SE of

The harzburgitic composition of the ophiolitic peridotites from the Klzlhrmak Ophiolite and from the m61ange beneath implies the presence of highly depleted mantle, possibly resulting from hydrous melting related to a subduction zone (Pearce et al. 1984). In this context, the ratios of Cr-number (Cr • 100/(Cr + A1)) and Mg-number (Mg • 100/(Mg+Fe 2+) in spinels allow peridotites that formed in a MOR-type setting to be distinguished from those formed in an SSZ-type setting (Dick & Bullen 1984). The main constituents of spinel (Mg,Fe 2+) (Cr,A1,Fe3+)204 behave differently during partial melting or crystallization, with Cr and Mg partitioning into solid phases, and A1 into the melt. The results of 273 analyses of 127 spinel grains from three samples of harzburgite from the Central Pontides (Table 2) were plotted on a Crnumber v. Mg-number diagram (Fig. 7a). In general, the observed high Mg-number values of the ophiolitic samples relative to abyssal peridotites could reflect low-temperature re-equilibration with olivine (Dick & Bullen 1984). The grains analysed from the single sample from the Klzdlrmak Ophiolite (OCP1) exhibit Cr-number values within the range for abyssal peridotites and are consistent with either MORtype or a back-arc marginal basin setting. The two samples of serpentinized harzburgite from the m61ange (Fig. 7; MCP1, MCP2) plot in the higher Cr-number group, implying a higher degree of partial melting of the mantle source. The high Cr-number values are similar to those for SSZ boninite-type settings, including the Upper Pillow Lavas of the Troodos ophiolite (Dick & Bullen 1984). The most likely origin is that these harzburgites originated in a forearc setting and were later incorporated into the accretionary m61ange beneath.

Eastern Pontides

424

S.P. RICE E T AL.

e-: r oO o(3

Z

oO

9 .~

r

9 oo -.,..,.

9 .~ U'-

r.~

t'q oo

oo -.,..,..

9 e~

O

9 eq >

9

:m

t"q

9 .<

.~ C"q

eq

c'Q

9 9

. ,,...~

. ,,,,~

8

C E N T R A L & E A S T E R N P O N T I D E S ACTIVE M A R G I N , T U R K E Y

425

~ ~:'-'~

w ~ w w w

ooooo~ O

~

~

X

I-I

+

~m

0.0 ~

o

"~'

c O.

o~'o S~'~

:..,~:

,,,.,.

,.

"

,.

,..

,

-.'-

-

l'l~

I

~

<

~

~=._=~

~.~ ~

m

I

!

:.'.--

~-~z" -=~o~ '~

~lU

.~

~1I ~

~i

~'~

~.o

~

<~I

~

~

~--~

~9

<

",~ I

I

"~I

I

o

8

8 o

~

.~

~

N

"~ ~ ' ~

=.==~

0_

I~.

n

0 0 0 o

0 ~

~

O

!

u~ ~H rJ ~ +~-

o

i

-'~o

..,~,

e'"

,

I

0 0

~ ~

~I~

~19 ~ 1 n,

0

x ' ~

0

N ~ O

o

<+~-~

~

m<

)

~ .J .=_~ _o..~

~-~

~ ~

o. o o o ._~

"9

rT1

..

~,"'":':"":-~ ............ "~:.,-: '

i~~~ _ . . ~ ~ "

I

I

!

I

I

i

0

0

0

0

0

0

CO

"...........

I

0

I

0

I

0

~

--~ 0

~l

0

=9 ~

0

,=

g~..=

426

S.P. RICE ET AL.

Fig. 8. Simplified geological map of the Eastern Pontides (see Fig. 1 for location). Line of section b-b' (see Fig. 9). The locations of sites sampled for geochemical analysis by XRF are marked: A, Karada~ Formation; B, Refahiye Complex (see Fig. 12). The Aylkayas~ Formation is too thin and localized to show here (see Fig. 10).

Erzincan (Figs 8 and 9). The lower contact is a north-dipping thrust, whereas the upper boundary is an unconformity overlain by sediments of Early Eocene age or younger. The formation takes the form of large wedge-shaped thrust slices, individually up to 20 km long and c. 3 km thick. The total outcrop in the Erzincan area is estimated as c. 250 km 2. The presence of microfossils, including Globotruncana, within thin limestones interbedded with volcanic rocks

indicates a Late Cretaceous (CampanianMaastrichtian) age (K. Ta~h & N. inan, pers. comm.). The Karada~ Formation comprises a thick (c. 3 km), disrupted succession of hornblendeand plagioclase-phyric basaltic to andesitic lavas, interbedded with coarse-grained volcaniclastic conglomerates (Fig. 10b). In addition, volcaniclastic sandstones, shales and tuffs are present in lesser amounts. Higher-level thrust sheets are

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

427

Fig. 9. Cross-section showing the main structural and stratigraphic relations in the Eastern Pontides. (See Fig. 8 for the line of section.) mainly schistose and include metabasic and meta-andesitic flows (>70% by volume), individually 2-3 m thick, interbedded with volcaniclastic, tuffaceous and calcareous schists. Rare hemipelagic Globotruncana-bearing limestones are also present. The mineral assemblage quartz + epidote + plagioclase + chlorite + calcite is indicative of greenschist-facies metamorphism. The Karada~ Formation is interpreted as part of a Late Cetaceous volcanic arc, dominated by andesitic volcanic rocks and volcaniclastic sediments, with an open-marine pelagic microfauna and little or no terrigenous input.

Upper Cretaceous-Palaeogene Siitpznar Formation." forearc basin This coarsening-upward sedimentary unit, c. 1500 m thick, crops out in the mid-levels of the thrust stack (Figs 8 and 9), mainly east of Erzincan. The stratigraphic base is not exposed. North-dipping thrust faults imbricate the formation with other units. Planktonic Foraminifera (e.g. Globotruncana) in the lower part of the succession indicate a Late Cretaceous age, whereas benthic Foraminifera (e.g. Alveolina) in the higher part of the formation are indicative

of an Early Eocene age (K. Tash & N. inan, pers. comm.). The Sfitpmar Formation (Fig. 10c) passes conformably, and probably diachronously, into facies similar to the Sipik6r Formation in some areas (see below). The lower c. 800 m of the succession comprises thin- to medium-bedded ( 1 0 c m - l . 5 m ) quartz- and feldspar-bearing calcareous sandstones. Individual beds are typically amalgamated and exhibit planar and convolute lamination, with partings of mud, shale and marl. In places, the succession is dominated by thinly bedded, fine-grained, dark volcaniclastic shale, or by more thickly bedded, pale calcarenites with rare andesitic lava flows and volcaniclastic debris-flow conglomerates. The upper c. 600 m of the succession exhibits an increase in the abundance and thickness of volcaniclastic sandstones and conglomerates. The lower part of the Siitpmar Formation formed in a deep-marine setting, marked by lowenergy background deposition of volcaniclastic mud and pelagic carbonate. Subordinate intercalations of coarser-grained clastic sediments were deposited from turbidity currents. The volcaniclastic lithologies higher in the succession are

428

S.P.

RICE

E T AL.

9

9

13)

.,.~

"--r

o

o0

(0

0

-=

0

0

0

-~

r"

o ~

~l ~[9I I ~~ " _ ~ -

~-I~ a

~

~ .

~

~

~

.:~Y.c~,~s-

~] r ~-~ ~~

0

.r~ "~ = ~0

( w ) ssou~lo!q_L 0

w

~

~

"0

.,~

.-~

0

"~ _ ~~

~

BIll1

(w) sseu~0!ql J

~,.

g o

:::3

i

0 .,..~ 0

]

~Oll~ ::~Z) 0

(o

o ~

E

:o ~

~

H

lH~l~i!!~l~

~

0

MI

.( w ) sseu~lo!ql

~.~

0

~N

l,~ ._m

I

~-~

-I

~_~

I

,

I

I"

,

(w)

I

,

ssaU~lO!qZ

x 9 o r~ Q)

~

~N

44

N

44

4 4

N~ "~ 0 0 ~ "' ~~

~a

(w)sseu~o!q•

9-~ ~ ~ m

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

429

Eastern Pontides W Marble lozenges in sheared schist Dolerite dykes

Doleritic dykes and intermediate-composition Doleritic sheeted dykes hornblende-phyricdykes \ Plagiogranitedykelets Aplite dykes \ and serpentinized curing microgabbro, \ cumulatepyroxenite dolerite and pegmatite _ , ~,~.~...~, ~.% ~ ~

b 0 I

20 m I

I'~ I'~

Schist and gneiss

~

Serpentinizedperidotite

Dolerite

~

Plagiogranite

Marble

~.~ Thrust contact

Hornblende-phyricdyke ~

Igneous contact

Fig. 11. Screens of marble, schist and gneiss intruded by doleritic dykes within the Refahiye ophiolitic complex. (See text for explanation.)

very similar to those of the Karada~, inferred arc unit, suggesting that these were intergradational. Volcanogenic material is mainly detrital and coeval lava flows rarely occur. The coarse carbonate sediment was redeposited from a shelfdepth setting, rich in Nummulites sp. Increasing amounts of coarse, reworked ophiolitic and metamorphic material appear towards the top of the formation, grading into the Sipik6r Formation (see below). An increase in depositional energy and source proximity is confirmed by reduced textural maturity upwards. Overall, the Siitpmar Formation is interpreted as a Late Cretaceous-Early Cenozoic regressive forearc basin, in which detritus was being supplied from an uplifted m61ange, a volcanic arc, an ophiolite and continental basement units. By this time only very minor arc-related volcanism persisted, at least in this area.

Palaeogene SipikSr Formation: post-emplacement sediments In the area studied this varied sedimentary unit rests unconformably on all of the Upper Cretaceous units (including ophiolitic rocks) with the sole exception of the inferred forearc basin unit (Siitpmar Formation), with which it is intergradational and partly coeval (see above). The Sipik6r Formation is widely exposed around Erzincan, especially to the north of the city (Fig.

8). In the south, the formation unconformably overlies the Munzur Limestone Formation and to the north it unconformably overlies the Pontide metamorphic basement (Topuz et al. 2004; Figs 8 and 9). The Sipik6r Formation, up to 500 m thick, is dated as Paleocene-Eocene based on planktonic and benthic Foraminifera (e.g. Nummulites) (K. Ta~h & N. Inan, pers. comm.), in agreement with previous workers (Okay & ~ahintiirk 1997; Topuz et al. 2004). In the area studied, the Sipik6r Formation (Fig. 10) comprises lenticular polymict conglomerates with well-rounded clasts, together with poorly bedded to massive, fine- to coarse-grained sandstones. These lithologies interdigitate with very fine-grained, dark grey, hard, massive foraminiferal limestones, forming lenses up to c. 50 m thick. The clastic content is varied and includes metamorphic lithoclasts. In places, medium- to thick-bedded oncoidal and Nummulites-bearing bioclastic limestones are present, together with rudist bivalves, microbial carbonate and large sponges. The limestones are interbedded with massive, poorly sorted, lenticular sandstones, conglomerates and thin (c. 2 m) layers of anhydrite. There also local interbeds of red siltstones that contain thin (c. 5 cm) layers of caliche (palaeosols). The Sipik6r Formation accumulated in varied shallow-marine, lagoonal, deltaic to fluvial settings and was sourced from a wide variety

430

S.P. RICE ET AL.

of lithologies in the area, including ophiolitic, arc-related and terrigenous (continental) units. The Sipik6r Formation in the area studied lacks north-vergent tectonic structures, as seen in all of the structurally underlying units. The formation in this area is interpreted as a transgressive cover, deposited after north-vergent deformation and emplacement of the underlying Upper Cretaceous units. The timing of northvergent deformation in this area is, therefore, constrained to predate the Sipik6r Formation (i.e. pre- or syn-Paleocene-Eocene). Elsewhere (e.g. 90km westwards), similar coarse clastic sediments of Paleocene age (Ta~demir Formation) are unconformably overlain by Paleocene-Eocene Nummulitic limestones. This unit is structurally overlain by a thrust sheet of limestone (i.e. ~imendafg Nappe) that was presumably emplaced northwards in this area. Similar clastic sediments structurally underlie another thrust sheet to the east of the study area (Imalida~ Nappe; Okay & Sahintiirk 1997). Also, further east again in the Artvin area, Mid-Eocene, coarse non-marine clastic sediments are structurally overlain by Upper Cretaceous and older units. It is, therefore, probable that additional northward thrusting took place during Mid-Eocene time, approximately contemporaneously with the development of the south-vergent thrust wedge as a whole.

Karayaprak MOlange: ophiolitic mdlange This ophiolitic m61ange unit occurs at several different levels within the thrust stack, forming discrete slices up to 4 km thick (Figs 8 and 9). In the south, near Kemah (Fig. 8), the m61ange tectonically overlies the basinal latest Cretaceous Aylkayasl Formation above the Munzur carbonate platform. The m61ange is unconformably overlain by Paleocene-Eocene clastic sediments of the Sipik6r Formation (Fig. 10d) that postdate Mid-Eocene suturing, as seen in small exposures to the north and south of Erzincan (Figs 8 and 9). The m61ange is a tectonized mixture of blocks and slices that individually are up to c. 2 km in size. The blocks exhibit three main lithological associations: (1) altered basaltic pillow lavas commonly interbedded with red radiolarian chert, pelagic limestone and mudstone (i.e. basalt-pelagic sediment association); (2) very large blocks ( > 1 km) of pale grey massive crystalline limestone commonly associated with basalt (i.e. basalt-neritic limestone association); (3) blocks of serpentinite, gabbro and diabase (i.e. ophiolitic association). Less common lithologies include volcaniclastic shale, volcaniclastic

sandstone and rare amphibolite. Fissile, sheared serpentinite also occurs along shear zones and as a matrix to mafic and ultramafic blocks in some areas. Pelagic carbonate blocks in the m61ange yielded planktonic foraminifera and calpionellids (e.g. Globotruncana linneiana (d'Orbigny), Archaeoglobigerina sp., Calpionella alpina (Lorenz), Calpionella elliptica (Cadish)) of Early Cretaceous (Berriasian) age (K. Ta~li & N. inan, pers. comm.). The basalt-pelagic sediment association is interpreted as accreted abyssal sediments that originally accumulated on oceanic crust. By contrast, the basalt-neritic limestone association is indicative of deposition above the carbonate compensation depth (CCD), probably on seamounts that were later detached from the ocean floor and accreted into the m61ange. In addition, relatively rare deformed blocks of volcaniclastic and polymict sediments possibly represent accreted trench-fill sediments. The absence of terrigenous sediments suggests that the accretionary wedge developed away from a supply of continentally derived sediment.

Upper Cretaceous Aytkayasl Formation." foredeep unit At the base of the thrust stack in the south (Fig. 9), Triassic-Cretaceous shallow-water limestones of the Tauride Munzur platform (Munzur Dafg~ Unit) are overlain, with a sharp, but conformable, contact by a relatively thin (tens to several hundred metres thick) unit of pelagic carbonates, known as the Aylkayasl Formation (C)zgiil & Tur~ucu 1984). Small exposures in the south (too small to show on the regional map) reveal that this unit unconformably overlies the northern margin of the Munzur Dafgl Unit. A Late Cretaceous (Campanian-Maastrichtian) age for the Aylkayasa Formation is indicated by the presence of Late Cretaceous planktonic foraminifera (e.g. Globotruncana sp.) (K. Ta~h & N. inan, pers. comm.). The upper stratigraphic boundary of this formation has been cut out by south-vergent thrusting of Early Cenozoic age. The Aylkayasl Formation includes buff-grey, thin-bedded pelagic limestones, interbedded with redeposited oolitic and bioclastic calcarenites (Fig. 10e). The limestones include scattered lithoclasts of basic igneous rocks. The succession passes upwards into poorly sorted, coarse, clast-supported, lenticular graded conglomerates and breccias, containing angular to sub-rounded clasts of neritic limestone, red mudstone and chert within a matrix of calcarenite.

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY Conglomerate beds exhibit scoured erosive bases and appear to be channelized. The sharp stratigraphic base of the formation implies an abrupt change from a shallow-water carbonate platform to a deeper marine setting during Campanian-Maastrichtian time. The foundering of the carbonate platform is interpreted as the result of flexural loading of the Munzur carbonate platform, ahead of southward emplacing Neotethyan units, including mdlange and ophiolite (Ozer et al. 2004). This subsidence was accompanied by an increased input of coarse detrital sediments as high-energy gravity flows derived from both the advancing Neotethyan units (e.g. basic igneous and ultramafic material) and the collapsing Munzur carbonate platform (e.g. redeposited carbonate material).

431

Comendite A

R.yo,,}/X'~

1

pantellerite I\

0.,

~_/X~..

Phonolite

Trachyte

Rhyodacite]

~ _

Dacite

I Trachy-~ I andesite />-I*Karada0 Formation

Andesite

~lA,.....~-~/

~

/

[] Refahiye ophiolite dykes

[3

.... A

n

[ I

~

_ . ,~b-alkaline

I

Basanite [ nephelinite

,_ } Alkali-basalt

basalt

D

o.oo1 O.Ol

, o.1

, 1

, lO

Nb/Y

(a) 100

Geochemistry of the Eastern Pontide basalts and peridotites As for the Central Pontides, samples of relatively unaltered basaltic rocks were collected from the ophiolitic, arc-type and m~lange units in the Eastern Pontides. The Karada~ arc-related basalts plot in the andesite-basalt field on the Zr/Ti v. Nb/Y plot (Fig. 12a). On MORB-normalized plots (Fig. 12b, i) these basalts show enrichments in the more incompatible LILE relative to HFSE, and a distinct negative Nb anomaly in several samples. The patterns are comparable with those of modern magmatic arc basalts (Pearce 1982). Several samples show strong depletion of immobile elements (Nd, P, Zr, Ti, Y; see Table 3) suggestive of high-degree melting (Saunders & Tarney 1984). When plotted on a MORB-normalized diagram (Fig. 12b, ii) the dykes from the Refahiye ophiolitic complex range from near-MORB with slight Nb depletion, to more depleted patterns with a distinct negative Nb anomaly. The compositions of spinels from six samples of serpentinized harzburgite in the Eastern Pontides were plotted on a Cr-number v. Mg-number diagram (Fig. 7b). Multiple analyses were made of the cores and the rims of 12 or more individual spinel grains in each sample (Table 2). One sample of peridotite from the m61ange (MEP1), as for the m61ange from the Central Pontides (see above), exhibits a high Crnumber typical of Alpine-type peridotites. Of five samples from the intact Refahiye Ophiolitic Complex, four (OEP1, OEP2, OEP4 and OEP5) fall within or close to the field of abyssal spinel peridotites; one other sample from the ophiolite (OEP3) exhibits a much higher Cr-number.

0.1

0.01

I

Sr

K

t

I

I

I

I

Rb Ba Nb Ce Nd

I

I

I

I

I

F

P

Zr

Ti

Y

Sc

Cr

(bi) K a r a d a c j a r c V o l c a n i c s 100

0.1 0.01 ~ l Sr

I

K

I

I

L

I

Rb Ba Nb Ce Nd

I

I

I-

I

t ~

P

Zr

Ti

Y

Sc

Cr

(bii) R e f a h i y e o p h i o l i t e d y k e s Fig. 12. Whole-rock geochemistry of basaltic rocks from the Eastern Pontides: (a) Zr/Ti v. Nb/Y diagram; (b) MORB-normalized trace-element patterns of basaltic rocks from the Eastern Pontides: (i) volcanic arc (Karada~ Formation); (ii) dolerite dykes from the ophiolitic Refahiye Complex. Normalizing values: Sr, 120 ppm; K20, 0.15%; Rb, 2.0 ppm; Ba, 20 ppm; Nb, 3.5 ppm; La, 3 ppm; Ce, 10 ppm; Nd, 8 ppm; P205, 0.12%; Zr, 90 ppm; TiO2, 1.5%; Y, 30 ppm; Sc, 40 ppm; Cr, 250 ppm (Pearce 1982). (See Fig. 8 for sample locations.) Samples OEP1, OEP3 and OEP5 were collected from a 3 km long transect of a single peridotite thrust sheet. Multiple grains were analysed from samples from this unit. Each sample clearly falls within one of the two petrogenetic types (i.e. Alpine-type or abyssal peridotite). A marked

432

S.P. RICE E T AL.

~o g~ On

o;Z

t~

o--

O0

~ d ~ d ~ d ~ d d d ~ d _~eq

-~o O ~ o~

o

oo~

8~

oo.~1

9

~q o~M o

H o

~

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY compositional variation is thus present within this one thrust sheet. Such variation is consistent with variable depletion and enrichment trends found in chromites within peridotites dredged from back-arc basins (e.g. Dietrich et al. 1978; Saunders & Tarney 1984; Barker et al. 2003).

Structural vergence of units The structural vergence recorded within individual units is critical to an understanding of the emplacement history of both the Central and the Eastern Pontides. All of the Upper Cretaceous units within the IAESZ in both of these areas preserve shear fabrics, folds and faults exhibiting a relatively early top-to-the-north kinematic sense of movement. Selected, representative outcrops are illustrated in Figure 13a-d. For example, in both regions the serpentinized harzburgites show a strong sigmoidal shear fabric indicating top-tothe-north movement. In the Eastern Pontides, near Akbudak (Fig. 8), the tectonic contacts between lithological units are locally southward dipping but with top-to-the-north shear fabrics (e.g. small sigmoidal shear-pods) (Fig. 13a). The Upper Maastrichtian, inferred forearc succession (Sfitpmar Formation) exhibits north-vergent asymmetrical folds (Fig. 13b) within large north-dipping thrust slices. In the Central Pontides, at Akkaya (Fig. 2), similar structural relationships are seen. For example, the Kirazbasl ophiolitic m61ange exhibits a strong top-to-the-north shear fabric (Fig. 13c). Downward-facing asymmetrical folds with north-dipping axes are also locally preserved, also suggesting top-to-the-north vergence, as seen near Ilgaz, at Yuvasaray (Figs 2 and 13d). Following north-vergent displacement, units in both the Central and the Eastern Pontides were regionally deformed associated with the development of south-vergent thrusts, folds and shear zones. The initial vergence was, therefore, northwards, followed by southward vergence. The transgressive Sipik6r Formation of Paleocene-Eocene age is the oldest unit in the Eastern Pontides to lack evidence of northvergent deformation in the area studied. We, therefore, see the north-vergent deformation as a mainly Late Cretaceous event in the area studied although, as noted earlier, there is evidence from adjacent areas of further north-verging displacement in the Mid-Eocene. In general, the southvergent thrusts are capped by clastic units of Late Eocene age ( ~ a n k m Basin; Central Pontides), which are interpreted to post-date final suturing of the Northern Neotethys. These Eocene and

433

younger sediments were further deformed by further south-vergent thrusting of pre-PlioQuaternary age, which is seen as the result of post-collisional suture tightening.

Comparison of the Central and Eastern Pontides It is important to decide whether the Central and Eastern Pontide units of the IASZ exhibit a similar tectonic evolution. Two large slices of inferred Upper Cretaceous volcanic arc units are present in both areas (Fig. 14a and b), associated with emplaced accretionary complexes and dismembered ophiolites. However, the Upper Cretaceous mixed terrigenous-volcaniclastic-volcanic unit (Iki~am Formation), interpreted as a back-arc marginal basin, is exposed only in the Central Pontides. On the other hand, a thick succession of inferred forearc basin sediments (Sfitpmar Formation) is widely exposed only in the Eastern Pontides, probably because its Upper Cretaceous counterpart in the Central Pontides (Yaprakh Formation) has been largely concealed by the post-suturing ~ a n k m Basin. Ophiolitic units crop out far more extensively in the Eastern Pontides (Refahiye unit) than the Central Pontides (Kmllrmak units). Ophiolitic volcanic rocks are well represented in the Central Pontides but not exposed in the Eastern Pontides. Each of the above Upper Cretaceous units occupies a distinct structural position within the suture zone (Fig. 14a and b). The ophiolitic units in both the Central and Eastern Pontides occur towards the top of the thrust stack. As an exception, relatively small exposures of ophiolitic rocks are present in the southern part of the Eastern Pontide area; these could represent large tectonic blocks within accretionary m61ange. In both areas two thrust slices of inferred volcanic arc rocks occur towards the middle of the thrust stack. The upper of these two slices is more deformed in both areas. Metamorphic minerals and textures are better developed in the upper slice, suggesting a slightly higher metamorphic grade, although a greenschist-facies mineral assemblage is present in both of the inferred arc units. In addition, the ophiolitic m61ange is distributed throughout the thrust stack in both areas. In the Central Pontides, thrust slices of the thick, Upper Cretaceous, inferred marginal basin unit 0ki~am Formation) mainly occur towards the top of the structural pile. The Upper Cretaceous forearc-type sediments (Yaprakh Formation) that locally unconformably overlie the inferred arc unit (Yaylagayl Formation) occur at

434

S.P. RICE E T AL.

a) Eastern Pontides - Akbudak Melange

Oligocene-Miocene sediments

N

Serpentinite [~

Dolerite Metamorphic basement

0

!

[=~

1 km

Upper Cretaceous-Lower Tertiary sediments

i

~

s

Anticline, syncline Thrust or shear-zone

sandstone succession

~

~ a~-3"~-~.~~

0

lm

I

I

_I

Stratigraphic way-up

--1 m'-'~-

c) Central Pontides, Akkaya

Pillow lavas

downward-facing asymmetrical folds d) Central Pontides, Yuvasaray Upper Cretaceous sediments (Ikigam Fro.) N

Tectonic transport directi

~

~

"" '

0

10m

Fig. 13. Selected exposures from the Eastern Pontides (a, b) and the Central Pontides (c, d) showing the northward kinematic vergence within Upper Cretaceous-Lower Cenozoic units. (See text for explanation.)

a low level in the thrust stack (Fig 14a). The thick Upper Cretaceous inferred forearc basin succession in the Eastern Pontides (Sfitplnar Formation) is also located at a low to mid-level in the thrust stack, where it is tectonically imbricated as two slices of the volcanic arc unit. Basalts from the volcanic arc units in both areas (Karada(~ and Yaylagayl Formations) fall into the fields of island-arc and calc-alkali basalts, as summarized in Figure 15. A single sample from the Central Pontides (Yaylaqay~ Formation) plots just within the field of within-plate type basalts. The ophiolitic basaltic dykes from the Eastern Pontides (Refahiye Complex) lie within the fields of ocean-floor basalts and island-arc basalts. The basaltic rocks

analysed from the K~zlhrmak ophiolite in the Central Pontides are ocean-floor basalts or island-arc basalts. Basalts from the m61ange (not shown) exhibit MORB and WPB geochemical signatures in both areas (Rice 2005). In summary, the above comparisons indicate that there are sufficient similarities between the Late Cretaceous-Early Cenozoic tectonostratigraphy of the Central and the Eastern Pontides to suggest that both originated in a similar overall tectonic setting, c. 600 km apart along strike from each other. The two areas can thus be taken together when considering possible tectonic models (see below). Little useful information is available from the intervening areas, which are generally less well exposed.

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

435

Fig. 14. Interpretative cross-sections of tectonostratigraphic units within (a) Central Pontides, and (b) Eastern Pontides. (See text for explanation.)

Comparison with adjacent areas of the Eurasian margin The evolution of the Eurasian continental margin can be documented from the relatively autochthonous Pontide basement to the north. In the Eastern Pontides, Cretaceous-Eocene

calc-alkaline volcanic rocks of the Eastern Pontide arc (up to 4 km thick) are exposed c. 50 km north of the suture zone, extending eastwards. These volcanic rocks are traditionally interpreted as part of a continental margin-type magmatic arc related to northward subduction of the Northern Neotethys ($en~or & Yllmaz 1981;

436

S.P. RICE E T A L .

Till00 9 Kizlltrmak Ophiolite ,x Yaylas volcanic unit [] ikigam Formation o Ophiolitic Refahiye Complex o KaradaO volcanic unit Zr

Yx3

Fig. 15. Summary of the compositions of Central and Eastern Pontides basalts plotted on a Ti-Zr-Y diagram (Pearce & Cann 1973). Ocean-floor basalts, field B; island-arc basalts, fields A and B; calc-alkaline basalts, fields B and C; within-plate basalts, field D.

A k m o 1984; Robertson & Dixon 1984; Okay & ~ahintiirk 1997; Ydmaz et al. 1997). Subductionrelated volcanism in the Eastern Pontides arc began in the Turonian (Taner& Zaninetti 1978), or at the Late Coniacian-Santonian boundary (Ydmaz et al. 2003) according to different views. The arc-related volcanic rocks are overlain by Campanian pelagic limestones (Akmo 1984), which in turn are overlain by Eocene volcanic rocks including high-K andesites (Bekta~ et al. 1999). It was recently suggested that the Eastern Pontide arc might instead relate to southward subduction of oceanic crust within the Black Sea region to the north, based on geochemical evidence (Bektas et al. 1999). Southward subduction of Western Black Sea lithosphere has also been proposed based on the interpretation of seismic data (Hossack 2004). However, this interpretation is yet to be supported by independent structural or stratigraphical field evidence. Also, it is not clear whether the Eocene volcanic rocks relate to a subduction-related setting or a syn- to post-collisional tectonic setting. Another relevant unit, located much further west, is the Galatean Volcanic Province in the Central Pontides. The lower part of this unit, known as the Sara~k6y Volcanic Suite, is located c. 80 km SW of the area studied. This suite of lavas was radiometrically dated at 76.4_+ 2.4 Ma (Late Campanian), and inferred to represent an extensional back-arc setting (Kogyi~it et al.

2003). By contrast, most of the overlying Galatean Volcanic Province represents postcollisional magmatism of Oligocene-Miocene age. The Saraqk6y Volcanic Suite is geochemically similar to the Campanian-Maastrichtian Yaylaqay~ Formation and might represent a more southwesterly extension of this inferred volcanic arc unit. By contrast, a back-arc marginal basin, interpreted by us as a Late Cretaceous marginal basin, restores to a more northerly position behind the Yaylaqayl volcanic arc unit. In addition, kilometres-thick sedimentary basins of latest Cretaceous to Mid-Eocene age, for example in the Haymana-Polath region, near Ankara (e.g. Koqyi~it 1991) are widely interpreted as forearc basins related to northward subduction of the Northern Neotethys that persisted until Mid-Eocene time (G6rfir et al. 1984). However, there is little regional evidence of steady-state subduction after latest Cretaceous time when regional northward thrusting took place; for example, an Early Cenozoic accretionary prism is not known to exist. It is, therefore, likely that the Haymana-Polath basin and other parts of the Central Anatolian basin complex developed in a regional setting of incipient collision ('soft collision') in the latest Cretaceous (Campanian-Maastrichtian), followed by final collision ('hard collision') during Mid-Eocene time (Clark & Robertson 2002).

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

Comparisons with modern settings Useful insights into the tectonic development of the Pontides can also be gained from comparisons with a range of modern and ancient convergent margin settings, including accretionary prisms, arcs and back-arc basins. The Pontide m61anges are comparable with many examples of accretionary m61ang~, including those associated with ophiolite err.placement (e.g. Late Cretaceous Oman m61~nge; Searle & Cox 1999; Robertson 2002, 2006), and those formed in active continental margin settings (e.g. Franciscan Complex, USA; Cloos 1982). The Pontide inferred arc and back-arc units are generally similar to other examples including the Dras-Kohistan arc (Robertson & Degnan 1994; Clift et al. 2002) and the associated Shyok backarc basin in the Himalayas (Robertson & Collins 2002), although these units appear to have formed in a more oceanic setting, far from the Eurasian margin. Useful comparisons can also be made with both ancient (e.g. Western USA; Eastern Mediterranean; Southern Andes) and extant back-arc basins (e.g. Japan Sea; South Atlantic Bransfield Strait). Several Eastern Mediterranean examples are directly relevant. The Late Palaeozoic-Early Mesozoic Kiire marginal basin in the Central Pontides opened by rifting of the Eurasian continental margin in response to inferred northward subduction during Late Palaeozoic-Early Mesozoic time (Usta6mer & Robertson 1994, 1997), and later collapsed prior to Late Jurassic time related to southward underthrusting. Opposing subduction also features in our preferred tectonic model for the areas studied (see below). The ophiolitic extrusive rocks of the Kiire Basin are of near-MORB composition, with a small negative Nb anomaly (Usta6mer & Robertson 1999). Further west, in northern Greece, a Guevgueli ophiolite formed within a rifted continental unit (Serbo-Macedonian zone) during Mid-Late Jurassic time (B6bien et al. 1987). This ophiolite retains primary intrusive contacts with metamorphic rocks, comparable with the screens of country rock within the Eastern Pontide ophiolitic dykes. The ophiolitic extrusive rocks are mainly of MORB-type but locally show a minor subduction influence. The back-arc ophiolite was bordered on its oceanic side by a rifted continental fragment, capped by a Jurassic magmatic arc (Brown & Robertson 2003). These comparisons indicate that subduction-related processes were active along the Eurasian margin prior to Late Mesozoic time.

437

Within the circum-Pacific region, the Jurassic Josephine marginal basin rifted along a preexisting accretionary margin during Mid-Late Jurassic time. The back-arc basin was bordered on its oceanward side by an active arc (Chetco arc). The back-arc crust was overlain by mainly terrigenous but locally volcaniclastic sediments (Galice unit). This marginal basin closed, oceanwards, beneath the arc, resulting in the emplacement of oceanic lithosphere (Josephine Ophiolite) onto the continental margin (Harper 1984). A similar model is favoured for the Pontide areas studied. Another comparable ancient continental margin back-arc basin setting is the Rocas Verdes marie complex (Sarmiento and Tortuga ophiolites) in southern Chile (Dalziel et al. 1974; Rabinowitz & La Breque 1979; Weaver et al. 1979; Dalziel 1986; Dilek & Flower 2004). This comprises discontinuous exposures of gabbros, sheeted dykes and pillow basalts ( > 3 km thick), interpreted as deformed, uplifted (but relatively autochthonous) marginal basin crust. The presence of pillow basalts overlain by tufts, volcaniclastic sediments and cherts suggests proximity to an active volcanic arc. The Andean ophiolites formed between Palaeozoic basement to the east and an Early Cretaceous andesitic arc to the west (Patagonian Batholith). The ophiolitic rocks locally intrude metamorphic basement, supporting an intra-continental origin (Dalziel et al. 1974). Deformational structures within the Andean ophiolites indicate eastward displacement, towards the continental interior. NMORB-normalized trace-element patterns from the ophiolitic basalts (Sarmiento and Tortuga) show a general enrichment in LILE relative to HFSE and a negative Nb anomaly, suggestive of a subduction influence. In contrast to the Pontide examples, the Southern Andean marginal basin formed well within the continental borderland and a large continental fragment was rifted. A comparable modern-day intracontinental marginal basin is the Bransfield Strait, South Atlantic. This is inferred to have formed in a suprasubduction-zone setting, with the development of a back-arc spreading centre related to trench rollback (e.g. Keller & Fisk 1992). The arc rocks are exposed on the South Shetland Islands, where the oldest extrusive rocks are mostly low-K, high-alumina basalts, basaltic andesites and low-silica andesites of Aptian age. Cogenetic gabbros, tonalites and granodiorites are also exposed. Samples dredged from the centre of the Bransfield Strait are geochemically variable and plot in the combined field of ocean-floor, islandarc and calc-alkaline basalts (Keller & Fisk 1992). The Upper Cretaceous ophiolites of the

438

S.P. RICE E T AL.

Central and Eastern Pontides are likely to have formed in a similar setting to the Bransfield Strait. In general, oceanic subduction zones may either retreat (roll-back) towards the ocean associated with an extensional stress field in the upper plate (Mariana-type subduction; Uyeda & Kanamori 1979; Uyeda 1982), or instead advance towards the hinterland, resulting in compression. Intra-oceanic trench retreat can accommodate the opening of back-arc basins (e.g. Mariana and Lau basins). 'Retreating accretionary orogens' characterize the Western Pacific region; e.g. Mariana (Karig 1971), Sea of Japan (Uyeda 1982), Scotia Sea (Saunders & Tarney 1984) and the Bransfield Strait (Keller & Fisk 1992). A change from a retreating orogen to an advancing one can be triggered by local, regional, or global factors, including: (1) changes in the forces acting on the subducting lithosphere as its angle of descent changes (Royden 1993); (2) changes in regional-scale relative plate motions (Dalziel et al. 1974; Smith 2006); (3) lateral mantle flow (Flower 2003); (4) arrival of a large bathymetric feature (e.g. seamount) at the trench (Cadet et al. 1987). In the Pontides, a switch from a 'retreating orogen' to an 'advancing orogen' was triggered by the arrival of the Tauride continental margin (e.g. Munzur Platform) at the subduction trench, as discussed below.

Alternative tectonic models We now consider alternative tectonic interpretations for the Late Mesozoic-Early Cenozoic development of the Central and Eastern Pontides in the light of the modern and ancient comparisons. Previous models

In a simple tectonic model, envisaged by Seng6r & Yllmaz (1981; Fig. 16a) a single north-dipping subduction zone consumed MOR-type Northern Neotethyan oceanic lithosphere. This subduction generated the Eastern Pontide arc and eventually resulted in southward ophiolite emplacement onto the Tauride-Anatolide platform related to trench-margin collision. This model does not, however, explain the emplacement of Neotethyan ophiolitic units northwards onto the Eurasian margin. Also, the available geochemical evidence suggests that the Pontide ophiolites formed in a Late Cretaceous SSZ setting rather than at a mid-ocean ridge. Furthermore, the presence of screens of dyke-intruded metamorphic basement rock within the Eastern Pontide

ophiolite suggest that this unit formed by rifting of the Eurasian continental margin. A second model (Fig. 16b) envisages two subduction zones, one developing beneath the Pontide margin and another within the ocean to the south (Tiiysfiz 1990); however, the polarity and timing of this intra-oceanic subduction were not clearly specified. A third model postulates a subduction polarity reversal (Okay & Sahintfirk 1997): ophiolites were first emplaced northwards onto the Pontide basement as a result of trench-margin collision during Cenomanian-Turonian (c. 93 Ma; Fig. 16c, i); subduction then flipped to consume remaining oceanic crust beneath the Eurasian margin, creating the Eastern Pontide magmatic arc during the Palaeogene (Fig. 16c, ii). The main problems here are that (1) Eastern Pontide arc volcanism began as early as the Turonian (Taner & Zaninetti 1978), or Coniacian (Yllmaz et al. 1997), rather than Palaeocene as would be expected in this interpretation; (2) the model does not explain the southward emplacement over the collapsed Munzur Da~l carbonate platform during the latest Cretaceous; and (3) the model assumes that the Eocene volcanic rocks of the Eastern Pontides relate to normal subduction when they may instead have erupted in a syn- or post-collisional setting. In a fourth alternative (Fig. 16d), Neotethyan oceanic crust was subducted northwards related to opening of a marginal basin along the Eurasian margin; this basin later collapsed and ophiolites were emplaced northwards onto the Pontide margin (Usta6mer & Robertson 1997). However, the timing and processes involved were not clearly specified. Proposed new model

The interpretation that best fits our new information is shown in Figure 17. Northward subduction is seen as initiating within the Northern Neotethys adjacent to the Eurasian margin during the Late Cretaceous (CenomanianTuronian?). This led to the construction of a Late Cretaceous (Santonian?-Campanian) marginal volcanic arc (Fig. 17a) and an accretionary prism made up of mainly Cretaceous-aged fragments of oceanic crust, deep-sea sediments and seamounts. In addition, tectonic erosion of the forearc could explain the presence of SSZ-type harzburgite blocks and slices in the m61ange (e.g. Beccaluva et al. 2004). It is possible that the subduction zone trended at an oblique angle to the Eurasian margin. As a result, the belt of arc volcanism was located well inboard to the east within the Eastern Pontides (Artvin region), but then intersected with and

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY

439

Fig. 16. Published tectonic models for the Late Cretaceous-Early Cenozoic tectonic assembly of the suture zone in the Pontides, generally: (a) single northward-dipping subduction zone (~eng6r & Yllmaz 1981); (b) two northward-dipping subduction zones (polarity of oceanic arc not specified; Tfiysiiz 1990); (e) southward-dipping subduction, followed by reversal of subduction direction (Okay & Sahintfirk 1997); (d) single northwarddipping subduction zone with the genesis and emplacement of a marginal basin (timing not specified; Usta6mer & Robertson 1997). (See text for discussion.) straddled the continental margin further west in the Eastern and Central Pontide areas studied. Further west the arc was possibly located some distance out into the Northern Neotethys, which would explain the apparent absence of Late Cretaceous arc volcanic rocks on the Eurasian margin in the Western Pontides. A back-arc basin (mainly Campanian) rifted along the south Eurasian margin within the Pontide continental basement (Fig. 17b), explaining the inclusions of metamorphic rocks within

the Refahiye ophiolite in the Eastern Pontides. As the back-arc basin opened (Fig. 17c) the active arc migrated oceanwards, switching off the Eastern Pontide arc prior to the Campanian. Subduction of the Northern Neotethys continued in latest Cretaceous time (CampanianMaastrichtian) until the trench began to collide with the leading edge of the Tauride continent, represented in the Eastern Pontides by the northfacing Munzur platform. The resulting collision drove the southward emplacement of the

440

S.P. RICE E T AL .

Fig. 17. Proposed new tectonic model for the development of the suture zone in the Pontides. (See text for explanation.)

Neotethyan accretionary mdlange, arc and ophiolitic units onto the Munzur carbonate platform, which by then had collapsed (Fig. 17d). The Late Cretaceous oceanic crust within the back-arc marginal basin was then subducted southwards until the convergence zone collided with the Eurasian margin, still during

Campanian-Maastrichtian time, as best documented in the Central Pontides. This explains the initial northward emplacement of ophiolites and related units onto the Pontide basement prior to the Paleocene in the areas studied. It is interesting to note that seismic tomographic studies reveal a high-velocity slab (i.e. a high-Q body) dipping

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY southwards beneath the region (Koulakov et al. 2002), which could relate to a south-dipping subduction zone. Because allochthonous oceanic units were emplaced onto both the Eurasian and Tauride margins during the Campanian-Maastrichtian (e.g. Eastern Pontides) it is likely that the Northern Neotethys was in the process of closing completely by this time ('soft collision'). However, during the Paleocene-Early Eocene some oceanic crust persisted between the Taurides and Pontides, especially within embayments, and this was presumably subducted northwards. During the Mid-Eocene the Tauride margin underwent attempted subduction beneath the Eurasian margin. During the collision that ensued ('hard collision') the entire thrust stack was re-imbricated and thrust southwards (Fig. 17e). The northward thrusting in some area can be seen as a related phase of backthrusting. Further south-vergent folding and thrusting, documented within Oligocene and Miocene sedimentary basins, are seen as a response to post-collisional suture tightening. The main difficulty we see with the above tectonic model is that much of the tectonic development took place during CampanianMaastrichtian time; a period of c. 18 Ma that at present cannot be adequately resolved. Thus, tectonic processes that appear to be broadly contemporaneous may in reality have been discrete, sequential events (e.g. back-arc opening and closure).

441

Santonian-Campanian (85. 5-70 Ma). A volcanic arc was constructed bordering the Eurasian continental margin in the Central and Eastern Pontides. Subduction zone 'rollback' triggered back-arc rifting, giving rise to subductioninfluenced volcanism and minor extensionrelated alkaline magrnatism in the north (Central Pontides). Metamorphic basement was incorporated into an extension-related dyke complex (Eastern Pontides). A back-arc marginal basin opened, floored by oceanic lithosphere and overlain by redeposited terrigenous, volcaniclastic and pelagic sediments. The activity of the Eastern Pontide arc ceased prior to the Campanian, whereas arc volcanism continued in a more outboard location. Deep-water pelagic and volcaniclastic sediments accumulated in associated fore-arc basins to the south. Campanian-Maastrichtian (85.8~55. 5 Ma). The Tauride continent (e.g. Munzur carbonate platform) collided with the subduction trench. With continued convergence the leading edge of the Tauride continent entered the trench, and the accretionary complex, the arc and its related forearc basin were thrust southwards over the collapsed platform margin. In response, the forearc basin rapidly shallowed and filled. Further north, the inferred back-arc marginal basin was subducted (underthrust) southwards, resulting in the marginal basin ophiolite and its deepsea sedimentary and volcanogenic cover being thrust northwards onto the Eurasian margin during Campanian-Maastrichtian time (e.g. Central Pontides), initiating a 'soft collision'.

Conclusions

Paleocene-Mid-Eocene (65. 5-48.6 Ma). Shallow-

The Izmir-Ankara-Erzincan suture zone in the Central and Eastern Pontide regions exposes Upper Cretaceous units that record the development of an accretionary complex, a volcanic arc, a forearc basin and a rifted back-arc basin. The following Late Cretaceous-Early Cenozoic stages of tectonic development are inferred.

marine to non-marine clastic and carbonate sediments accumulated on deformed and emplaced units. There is little evidence of convergence in the areas studied during this time. However, it is assumed that some Northern Neotethyan oceanic lithosphere persisted during this time and was subducted until a regional 'hard collision' took place.

Late Cretaceous ( Cenomanian-Coniacian; c. 99.6-85.8Ma). Neotethyan oceanic crust was subducted northwards beneath the Eurasian continental margin, represented by the Pontide metamorphic basement, leading to the initiation of the Eastern Pontide arc. A frontal accretionary prism developed composed of fragments of pelagic sediments, basalt and seamounts (e.g. basaltlimestone-chert). Serpentinite was possibly derived from the overriding forearc. Subductionaccretion persisted until CampanianMaastrichtian time, but there is no evidence of Palaeogene accretion.

Mid-Late Eocene (c. 48.6-37.4 Ma). In response to the attempted northward subduction of the Tauride continental margin beneath the Eurasian margin the entire Northern Neotethyan thrust stack was re-imbricated and thrust southwards. Northward backthrusting also affected some areas. Late Eocene-Late Miocene (37.4-5.33 Ma). The suture zone was largely emergent during the Oligocene, and then transgressed by shallow-water carbonates in some areas during the Miocene. Further compression and southward thrusting relates to post-collisional suture tightening.

442

S.P. RICE ET AL.

Plio-Quaternary. Segments of the suture zone experienced left-lateral displacement (by up to 80 km) along the North Anatolian Fault Zone in both the Central and the Eastern Pontides, and this must be taken into account in any tectonic reconstruction. S. R. acknowledges the financial support provided by an NERC PhD studentship. He also acknowledges 0. Karsho~lu, Y. Aydar and L. Meston for invaluable assistance in the field. A.H.F.R. acknowledges the Carnegie Trust for the Scottish Universities for financial assistance with fieldwork. Palaeontological determinations for this work were kindly provided to A.H.F.R. by N. inan and K. Ta~h (Mersin University), mainly for the Eastern Pontides. Additional data were provided to T.U. by i . Ongen 0stanbul University) mainly for the Central Pontides.

References AKINCI, O. T. 1984. The Eastern Pontide volcanosedhnentary belt and associated massive sulphide deposits. In: ROBERTSON, A. H. F. & DIXON, J. E. (eds) The Geological Evolution of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 17, 415-428. ANDERSON, R. N., DELONG, S. E. & SCHWARZ,W. H. 1978. Thermal model for subduction with dehydration in the downgoing slab. Journal of Geology, 86, 731-739. BARKER, D. H. N., CHRISTESON,G. L., AUSTIN, J. A., JR & DALZIEL, I. W. D. 2003. Backarc basin evolution and cordilleran orogenesis: insights from new ocean-bottom seismograph refraction profiling in Bransfield Strait, Antarctica. Geological Society of America Bulletin, 31(2), 107-110. BI~BIEN, J., BAROZ, J., CAPERDI, S. 8z VENTURELLI,G. 1987. Magmatisme basique associ6es/t l'ouverture d'un basin marginal dans les Hell6nides internes au Jurassic. Ofioliti, 12, 53-70. BECCALUVA, L., COLTORTI, M., GIUNTA, G. & SIENA, F. 2004. Tethyan vs. Cordilleran ophiolites: a reappraisal of distinctive tectono-magmatic features of supra-subduction complexes in relation to the subduction mode. Tectonophysics, 393(1-4), 163-174. BEKTA~, O., ~EN, C., ATICI, Y. & KOPRI]BA~I,N. 1999. Migration of Upper Cretaceous subduction-related volcanism towards the back-arc basin of the Eastern Pontide magmatic arc (NE Turkey). Geological Journal, 34, 95-106. BERGOUGNAN, n . 1975. Relations entre les 6difices pontique et taurique dans le Nord-Est de l'Anatolie. Bulletin de la Societk Gdologique de France, XVII, 7(6), 1045-1057. BiRGiLI, ~., YOLDA~, R. & UNALAN,G. 1975. Canklrz(orum havzasmm jeolojisi ve petrol olanaklarl. MTA Enstittisfi, Rapor, 5621. BROWN, S. A. M. & ROBERTSON, A. H. F. 2003. Sedimentary geology as a key to understanding the tectonic evolution of the Mesozoic-Early Tertiary Paikon Massif, Vardar suture zone, N Greece. Sedimentary Geology, 160, 179-212.

CADET, J.-P., KOBAYASHI,K., AUBOUIN,J., et al. 1987. The Japan Trench and its juncture with the Kuril Trench: cruise of the Kaiko project, Leg 3. Earth and Planetary Science Letters, 83, 267-284. CLARK, M. S. & ROBERTSON,A. H. F. 2002. The role of the Early Tertiary Uluka~la Basin, southern Turkey in suturing of the Mesozoic Tethys ocean. Journal of the Geological Soctety, London, 159, 673-690. CLIFT, P. D., HANNIGAN, R., BLUSZTAJN, J. & DRAUT, A. E. 2002. Geochemical evolution of the Dras-Kohistan Arc during collision with Eurasia: evidence from the Ladakh Himalaya, India. The Island Arc, 11,255-273. CLOOS, M. 1982. Flow m61ange, numerical modelling and geological constraints on the origin of the Franciscan Complex, California. Geological Society of America Bulletin, 93, 330-345. DALZlEL, 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, 389404. DALZIEL, I. W. D., DEWIT, M. J. & PALMER, K. F. 1974. Fossil marginal basin in the southern Andes. Nature, 250, 291-294. 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. DICK, H. J. B. & BULLEN, Z. 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpinetype peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, 86, 54-76. DIETRICH, V., EMMERMANN, R., OBERHANSLI, R. PUCHELT, H. 1978. Geochemistry of basaltic and gabbroic rocks from the West Mariana basin and the Mariana trench. Earth and Planetary Science Letters, 39, 127-144. DILEK, Y. & FLOWER, M. F. S. 2004. 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. DROOP, G. T. R. 1987. A general equation for estimating Fe 3+ concentrations in ferromagnesian silicates and oxides from microprobe analyses, using stoichiometric criteria. Mineralogical Magazine, 51, 431-435. FITTON, J. G., SAUNDERS, A. D., LARSEN, L. M., HARDARSON, B. S. & NORRY, M. J. 1998. Volcanic rocks from the southeast Greenland margin at 63~ composition, petrogenesis and mantle sources. In: SAUNDERS, A. D., LARSEN, H. C. & WISE, S.W. JR (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 152. Ocean Drilling Program, College Station, TX, 331-350. FLOWER, M. F. J. 2003. Ophiolites, historical contingency, and the Wilson cycle. In: DILEK, Y. & NEWCOMa, S. (Eds) Ophiolite Concept and the Evolution of Geological Thought. Geological Society of America, Special Papers, 373, 111-136.

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY GARFUNKEL, Z. 2006. Neotethyan ophiolites: formation and obduction within the life cycle of the host basin In: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) Tectonic' Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 301-326. GORUR, N., OKTAY, F. Y., SEYMEN, i. • ~ENGOR, A. M. C. 1984. Palaeotectonic evolution of the Tuzg61fi Basin complex, central Turkey: sedimentary record of a Neotethyan closure. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 467-482. GRADSTEIN, F. M., OGG, J. G., SMITH, A. G., et al. 2004. A Geological Time Scale 2004. Cambridge University Press, Cambridge. HARPER, G. D. 1984. The Josephine ophiolite, northwestern California. Geological Society of America Bulletin 95, 1019-1026. HOSSACK, J. 2004. The tectonic evolution of the Black Sea. In: BEACH, A., BUTLER, R., GRAHAM, R., KNIPE, R., MCCLAY, K., RIES, A. 8r STEWART, S. (eds) Continental Tectonics." Discussion Meeting in Memory of the Life and Work of Mike Coward. Geological Society, London (abstracts). KARIG, D. E. 1971. Origin and development of the marginal basins in the Western Pacific. Journal of Geophysical Research, 76, 2542-2561. KAYMAK~I, N. 2000. Tectono-stratigraphical evolution of the Canktrt Basin (Central Anatolia, Turkey). PhD thesis, University of Utrecht. KELLER, R. A. & FISK, M. R. 1992. Quaternary marginal basin volcanism in the Bransfield Strait as a modern analogue of the southern Chilean ophiolites. In: PARSON, L. M., MURTON, B. J. & BROWNING, P. (eds) Ophiolites and their Modern Oceanic Analogues. Geological Society, London, Special Publications, 60, 155-169. KOr A. 1990. Structural relationships of three suture zones to the west of Erzincan (NE Turkey): Karakaya, Inner Tauride and Erzincan sutures. In: 8th Petroleum Congress of Turkey, Turkish Association of Petroleum Geologists UCTEA Chamber of Petroleum Engineers, 152-161. KO~Y~JIT, m. 1991. An example of an accretionary forearc basin from northern Central Anatolia and its implications for history of subduction of Neo-Tethys in Turkey. Geological Society of America Bulletin, 103, 22-36. KOffYIGiT, A., WINCHESTER, J. A., BOZKURT, E. & HOLLAND, G. 2003. Saraqk6y volcanic suite: implications for the subductional phase of arc evolution in the Galatean Arc Complex, Ankara, Turkey. Geological Journal, 38, 1-14. KOLLER, F., HOECK, V., MEISEL, T., IONESCU, C., ONUZI, K. & GHEGA, D. Cumulates and gabbros in southern Albanian ophiolites: their bearing on regional tectonic setting. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 267-299.

443

KOULAKOV, I., TYCHKOV, S., BUSHENKOVA, N. & VASILEVSKY, A. 2002. Structure and dynamics of the upper mantle beneath the Alpine-Himalayan orogenic belt, from teleseismic tomography. Tectonophysics, 358, 77-96. OKAY, A.I. & ~AHiNTURK, (~). 1997. Geology of the Eastern Pontides. In: ROBINSON, G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists, Memoirs, 68, 291-311. OKAY, A. I., TANSEL, I. & T~YSf2Z, O. 2001. Obduction, subduction and collision as reflected in the Upper Cretaceous-Lower Eocene sedimentary record of western Turkey. Geological Magazine, 138(2), 117-142. 0ZER, E., KOC, H. & OZSAYAR, T. Y. 2004. Stratigraphical evidence for the depression of the northern margin of the Menderes-Tauride Block (Turkey) during the Late Cretaceous. Journal of Asian Earth Sciences, 22(5), 401-412. 6ZG~3L, N. & TUR~UCU, A. 1984. Stratigraphy of the Mesozoic carbonate sequence of the Munzur Mountains (Eastern Taurides). In: TEKELI, O. & GONCOO~LU, C. (eds) Geology of the Tauride Belt, International Symposium, Ankara, 173-180. PARLAK, O., HOCK, V., KOZCU, H. & DELALOYE, M. 2004. Oceanic crust generation in an island arc tectonic setting, SE Anatolian Orogenic Belt (Turkey). Geological Magazine, 141, 583-603. PEARCE, J. A. 1982. Trace element characteristics of lavas from destructive plate boundaries. In: THORPE, R. S. (ed.) Andesites. Wiley, New York, 525-548. PEARCE, J. A. & CANN, J. R. 1973. Tectonic setting of basic volcanic rocks determined using trace element analysis. Earth and Planetary Science Letters, 19, 290-300. PEARCE, J. A., LIPPARD, S. J. & ROBERTS, S. 1984. Characteristics and tectonic significance of suprasubduction zone ophiolites. In: KOKELAAR,B. P. 8t; HOWELLS, M. F. (eds) Marginal Basin Geology. Geological Society, London, Special Publications, 16, 77-94. RABINOWITZ, P. D. 8r LA BREQUE, J. 1979. The Mesozoic South Atlantic Ocean and evolution of its continental margins. Journal of Geophysical Research, 84, 5973-6002. RASSIOS, A. H. E. & MOORES, E. M. 2006. Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab (northern Greece). In: ROBERTSON, A. H. F. 8r MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 237-266. REED, S. J. B. 1975. Electron Microprobe Analysis. Cambridge University Press, Cambridge. RICE, S. P. 2005. The Role of an Upper Cretaceous Volcanic" Arc in the Tectonic Assembly of the Tethyan Suture Zone: Pontides, Northern Turkey. PhD thesis, Unversity of Edinburgh. RIZAO~LU, T., PARLAK, O., HOCK, V. & [~LER, E. 2006. Nature and significance of Late Cretaceous

444

S.P. RICE ET AL.

ophiolitic rocks and their relation to the Baskil granitic intrusions of the Elazl~ region, SE Turkey. In: ROBERTSON, A. H. F. & MOUNTRArdS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 327-349. 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. 2006. Contrasting modes of ophiolite emplacement in the Eastern Mediterranean region. In: GEE, D. & STEPHENSON,R. A. (eds) European Lithosphere Dynamics. Geological Society of London, Memoir 32 (in press). ROBERTSON, A. H. F. & COLLINS, A. S. 2002. Shyok Suture Zone: late Mesozoic-Tertiary evolution of a critical suture zone separating the oceanic Ladakh arc from the Asian continental margin. Journal of Asian Earth Sciences, 20, 309-351. ROBERTSON, m. H. F. & DEGNAN, D. J. 1994. The Dras arc Complex: lithofacies and reconstruction of a Late Cretaceous oceanic volcanic arc in the Indus Suture Zone, Ladakh Himalya. Sedimentary Geology, 92, 117-145. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the Eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. ROJAY, B., YALINIZ, K. M. & ALTINER, D. 2001. Tectonic implications of some Cretaceous pillow basalts from the North Anatolian ophiolitic m61ange (Central Anatolia-Turkey) to the evolution of Neotethys. Turkish Journal of Earth Sciences, 10, 93-102. ROYDEN, L. H. 1993. The tectonic expression of slab pull at continental convergent boundaries. Tectonics, 12(2) 303-325. SAUNDERS, A. D. & TARNEY, J. 1984. Geochemical characteristics of basaltic volcanism within backarc basins. In: KOKELAAR,B. P. & HOWELLS,M. F. (eds) Marginal Basin Geology. Geological Society, London, Special Publicatios, 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. SEARLE, M. P. & Cox, J. 1999. Tectonic setting, origin and obduction of the Oman ophiolite. Geological Society of America Bulletin, 111, 104-122. ~ENGOR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 181-241. SMITH, A. G. 2006. Tethyan ophiolite emplacement, Africa to Europe motions, and Atlantic speading. In: ROBERTSON, A. H. F. & MOUTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 11-34. STAMPFLI, G., MOSAR, J., FAURI~, P., PILLEVUIT, A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution

of the western Tethys realm: the Neotethys East Mediterranean basin connection. In: ZIEGLER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUINSOLEAU, S. (eds) Peri-Tethys Memoir No. 5 Per# Tethyan Rift~Wrench Basins and Passive Margins. M6moires du Mus6um National d'Histoire Naturelle, 182, 51-10. TANER, M. F. & ZANINETTI, L. 1978. Etude pala6ontologique dans le Cretac~ volcanosedimentaire de Giineyce (Pontides orientales, Turquie). Rivista Italiana di Paleontologica e Stratigrafica, 84, 187-198. 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. TOPUZ, G., ALTHERR, R., SATIR, M. & SCHWARZ,W. H. 2004. Low-grade metamorphic rocks from the Pulur complex, NE Turkey: implications for the pre-Liassic evolution of the Eastern Pontides. International Journal of Earth Science (Geologische Rundschau), 93, 72-91. TOYsOz, O., 1990. Tectonic evolution of a part of the Tethyside orogenic collage: the Kargl Massif, northern Turkey. Tectonics, 9, 141-160. T(/YSOZ, O., Y|~iTBA~, E. & SERDAR, H. S. 1988. Vezirk6prii-Boyabat dolaymm jeolojisi. Turkish Petroleum Company (TPAO) Report, 55. TI)YSOZ, O., DELLALOGLU, A. A. & TERZIOG-LU, N. 1995. A magmatic belt within the Neo-Tethyan suture zone and its role in the tectonic evolution of northern Turkey. Tectonophysics, 243, 173-191. USTAOMER, T. & ROBERTSON, A. H. F. 1994. Late Palaeozoic marginal basin and subductionaccretion: evidence from the Palaeotethyan Kfire Complex, Central Pontides, N. Turkey. Journal of the Geological Society, London, 151, 291-305. USTAOMER, T. & ROBERTSON,A. H. F. 1997. Tectonicsedimentary evolution of the north Tethyan margin in the Central Pontides of northern Turkey. In: ROBINSON, A. G. (ed.) Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists, Memoirs, 68, 255-290. USTAOMER, T. & ROBERTSON, A. H. F. 1999. Geochemical evidence used to test alternative plate tectonic models for pre-Upper Jurassic (Palaeotethyan) units in the Central Pontides, N. Turkey. Geological Journal, 34, 25-54. UVEDA, S. 1982. Subduction zones: an introduction to comparative subductology. Tectonophysics, 81, 133-159. UYEDA, S. & KANAMORI, H. 1979. Back-arc opening and the mode of subduction. Journal of Geophysical Research, 84, 1049-1061. WEAVER, S. D., SAUNDERS, A. D., PANKHURST, R. J. & TARNEY, J. 1979. A geochemical study of magmatism associated with the initial stages of back-arc spreading. The Quaternary volcanics of

CENTRAL & EASTERN PONTIDES ACTIVE MARGIN, TURKEY Bransfield Strait, from South Shetland Islands. Contributions to Mineralogy and Petrology, 68, 151-169. YILMAZ, A. 1985. Basic geological characteristics and structural evolution of the region between the Upper Kelkit Creek and the Munzur Mountains. Bulletin of the Geological Society of Turkey, 28, 79-92. YILMAZ, Y., T~YSOZ, O., Yi~iTBA~, E. CAN GENt, $. & SENGOR, A. M. C. 1997. Geology and tectonic evolution of the Pontides. In: ROBINSON,A. G. (ed.)

445

Regional and Petroleum Geology of the Black Sea and Surrounding Region. American Association of Petroleum Geologists, Memoirs 68, 183-226. YILMAZ, C., SEN, C. & OZGOR, S. 2003. Sedimentological, palaeontological and volcanic records of the earliest volcanic activity in the Eastern Pontide Cretaceous volcanic arc (NE Turkey). Geologica Carpathica, 54(6), 377-384. YOLDA~, R. 1982. Tosya (Kastamonu) ile Bayat (Corum) arasmdaki bOlgenin jeolojisi. PhD thesis, Istanbul University.

The wide distribution of HP-LT rocks in the Lycian Belt (Western Turkey): implications for accretionary wedge geometry G A E T A N R I M M E L I ~ 1, R O L A N D O B E R H A N S L I 2, O S M A N C A N D A N 3, B R U N O G O F F I ~ 1& L A U R E N T

JOLIVET 4

1Laboratoire de Gkologie de l'Ecole Normale Sup&ieure, C N R S , U M R 8538, 24, rue Lhomond, 75005 Paris, France (e-mail." rimmele@geologie, ens.fr) 2Institut fiir Geowissenschaften, Universitiit Potsdam, Karl Liebknechtstrasse 24-25, D-14476 Potsdam-Golm, Germany 3Dokuz Eyliil Oniversitesi, Miihendislik-Mirmalik Fakiiltesi, Jeoloji Miih. B6liimii, TR-35100 Bornova-Izmir, Turkey 4Laboratoire de Tectonique, U M R 7072, Universitd Pierre et Marie Curie, Tour 26-0, Etage 1, case 129, 4, place Jussieu, 75252 Paris Cedex 05, France In SW Turkey, Fe-Mg-carpholite has recently been recognized in the basal metasediments of the Lycian Nappes, which overthrust the Menderes Massif on its southern flank. This high-pressure-low-temperature (HP-LT) metamorphic index mineral was widely found in the Bodrum peninsula region. Our new metamorphic and structural data on similar carpholite-bearing rocks found farther north in several klippen of the Lycian Nappes located on top of the Menderes Massif show that HP-LT rocks in SW Turkey occur over a distance of > 200 km in both north-south and east-west directions, thus indicating a wide HP-LT metamorphic belt. The deformation pattern from the Bodrum peninsula to ~ivril, all along the contact between the Lycian Nappes and the Menderes Massif, reveals the role played by major top-to-the-NE shear zones contemporaneous with exhumation of the Lycian HP-LT rocks. This deformation shows an oblique direction of opposite shear sense relative to the earlier southward translation of the Lycian Nappes over the Menderes Massif, for which top-to-the-south displacements are preserved in the upper units of the Lycian Nappes on the Bodrum peninsula, as well as at the base of the Lycian nappe klippen located farther north. The widespread distribution of well-preserved Fe-Mg-carpholite-bearing rocks in the Lycian Nappes has implications for the geometry of the accretionary wedge responsible for HP-LT metamorphism in SW Turkey. Abstract:

Many high-pressure-low-temperature (HP-LT) metamorphic terranes have been described from the internal zones of the Alpine belt in the Mediterranean region. Blueschists and eclogites followed exhumation paths along cold P - T gradients allowing the preservation of H P - L T parageneses. Petrographic and structural studies of these metamorphic rocks throughout the Mediterranean region provide crucial information on subduction tectonics and the dynamics of accretionary complexes in which H P - L T metamorphic rocks were formed and exhumed (e.g. Jolivet et al. 2003). Western Turkey comprises several tectonic units in which H P - L T metamorphic rocks have been described. Some of these metamorphic rocks are vestiges of the Pan-African orogeny (Oberh/insli et al. 1997; Candan et al. 2001; Warkus 2001); others are located in the Sakarya Zone (Fig. la), formed during the Cimmerian orogeny related to closure of the Palaeotethys

Ocean (Okay & Moni6 1997; Okay et al. 2002). South of the Izmir-Ankara suture (Fig. 1a), other H P - L T metamorphic domains have resulted from the Late Cretaceous northward subduction of the Neotethys Ocean and subsequent Early Tertiary collision between the Sakarya microcontinent and the Anatolide-Tauride platform ($eng6r & Yllmaz 1981; Okay et al. 2001). These Alpine metamorphic terranes form several tectonometamorphic units: the Late Cretaceous blueschist belt of the Tavsanh Zone (Okay & Kelley 1994; Okay et aL 1998; Sherlock et al. 1999; Okay 2002), the recently described H P - L T rocks of the Afyon Zone (Candan et al. 2002, 2005), the Eocene Cycladic blueschists and eclogites of the Dilek peninsula region in the westernmost part of the Menderes Massif area (Candan et al. 1997; OberhS,nsli et al. 1998; (~etinkaplan 2002), and the H P - L T metasediments of the Lycian Nappes (Oberh~insli et al. 2001; Rimmel6 et al. 2003a, 2005) and the

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 447-466. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

448

G. RIMMELI~ E T AL.

27~E

28~

2~~

Fig." 5

39~ I

d

J

/FIg. 2 (

)

~

30~

25~

Klippen of

,**~

\

II

NA

PTES G

I

~

36~

_... ,,'~

B

~ o ~ . , , ....... II

- ~,--...,--:.-~r

SEA

~

Klippen of I I V Lycian Nappes

=(~2.~

"

~.,]~

FLYSCH

ZONE

~

II ' '

BORNOVA :lrum

t

.,-01

'

n

_

I , :

~L

CYCLADIC

30OE

t

I

\ \

C:3

CYCLADIC

i

COMPLEX

s

I

(E~ ""

Klippen of Lycian Nappes

ALPINE HP-L T METAMORPHISM (Carpholite-bearlng assemblages)

in the Lycian Nappes in the Menderes Massif

=~~~as

LYCIAN

REPRESENTA"i IVE SENSES OF SHEAR in the lowermost metasecliments of the Lycian Thrust Sheets (Karaova I=:::~ Formation) and in the uppermost metasediments of the Menderes Massif.

v..

NAPPES ~37ON

in the uppermost levels of the Lycian Thrust Sheets (KarabS~Qrllen wildflysch) and in the metamorphic sole of the Lycian Peridotite Thrust Sheet. in the uppermost levels of the Lycian m = l ~ Thrust Sheets, in the Lycian Mt!lange and in the Lycian Peridotite Thrust Sheet (data of Collins and Roberlson 2003). 27~

Z~~

2;~

36~ 30~

Fig. 1. (a) Simplified tectonic map showing the main tectonometamorphic units of western Turkey (modified after Okay & Tiiysiiz 1999; Bozkurt & Oberh~insli 2001); (b) metamorphic map showing the occurrences of carpholite-bearing rocks in the Lycian Nappes (sensu stricto and klippen) and in the southern Menderes Massif; (e) structural map showing the representative senses of shear deciphered in the Lycian Nappes and in the uppermost levels of the southern Menderes Massif. Menderes Massif (Rimme16 et al. 2003b) (Fig. l a). In SW Turkey, the Lycian Nappes overthrust the Menderes Massif on its southern flank (Fig. l b). The recent discovery of H P - L T rocks in both tectonic units together with detailed

structural work led us to reconsider the Alpine tectonic evolution of the whole area (Rimmel6 2003). However, up to now, the H P - L T assemblages have mainly been described in the Bodrum peninsula region, between Milas and

HP-LT ROCKS LYCIAN BELT, TURKEY KarabS~iirtlen (location shown in Fig. l b). We expand on this by providing new petrological and structural data for areas farther north, within several klippen of Lycian Nappes cropping out above the Menderes metamorphic rocks in the Dilek peninsula area to the west, near Borlu in the north, and in the region of ~ivril in the east. After presenting new field-based information, we discuss the metamorphic and structural data in the setting of western Turkey, and, comparing the Lycian Belt with other Mediterranean metamorphic orogens, we highlight the implications of the particularly wide distribution of these HP-LT rocks for the geometry of the accretionary complex responsible for the metamorphism in the Lycian Nappes.

Geological setting The Menderes Massif

The Menderes Massif basically consists of a Pan-African augen gneiss 'core' and an overlying metasedimentary 'cover' made up of Palaeozoic schists and Mesozoic to Cenozoic marbles (Schuiling 1962; de Graciansky 1966; Diirr 1975; Akk6k 1983; ~eng6r et al. 1984; Satlr & Friedrichsen 1986; Konak et al. 1987; Hetzel & Reischmann 1996). This description of the Menderes Massif was recently considered as being over-simplified and some workers have argued that the massif comprises a nappe pile (Partzsch et al. 1997; Hetzel et al. 1998; Ring et al. 1999a, 2001; Gessner 2000; Gessner et al. 2001a,b,c, 2004) that was assembled during the Alpine orogeny. In the central part of the massif, eclogite occurrences within the core have been reported (Oberh~insli et al. 1997; Candan et al. 2001). The HP-LT metamorphic event has been dated as Neoproterozoic and is related to crustal thickening during the Late Precambrian-Early Palaeozoic orogeny (Candan et al. 2001; Warkus 2001; Oberh~insli et al. 2002). At the base of the metasedimentary cover of the Menderes Massif, in the region of Milas and Mu~la (Fig. l b), magnesiocarpholite-kyanite-chloritoid assemblages have recently been described within synfolial quartz veins in an Upper Triassic metaconglomerate (Rimmel6 et al. 2003b). These findings led to the first consideration of a major Alpine HP-LT metamorphic event in the southern massif (12-14 kbar and 470-500 ~ Rimmel6 et al. 2003b, 2005). Before this discovery, the whole massif was thought to have recorded only a Barrovian-type metamorphism (the 'Main Menderes Metamorphism' of Seng6r et al. 1984) with greenschist-facies to amphibolite-facies

449

conditions (Diirr 1975; Akk6k 1983; Ashworth & Evirgen 1984a; Seng6r et al. 1984; Satxr & Friedrichsen 1986; Konak et al. 1987; Okay 2001; Ring et al. 2001; Whitney & Bozkurt 2002; R6gnier et al. 2003). The magnesiocarpholitebearing metaconglomerate is overlain by a thick Mesozoic envelope composed of massive dolomitic marbles with diaspore- and corundumbearing metabauxites, rudist-bearing marbles and reddish cherty marbles (Diirr 1975; Konak et al. 1987; Yalqln 1987; C)zer 1998; Ozer et al. 2001). This marble sequence is overlain by a Paleocene metaolistostromal formation, which consists of metaserpentinite and marble blocks within a schist matrix (Dfirr 1975; Gutnic et al. 1979; ~a~layan et al. 1980; Konak et al. 1987; Ozer et al. 2001). This metaolistostrome crops out along the contact with the overlying Lycian Nappes sporadically and is weakly metamorphosed (Gutnic et al. 1979; Rimmel6 2003). In the southern part of the Menderes Massif, the metasedimentary series are intensely deformed. Northward-verging kilometre-scale folds are oriented parallel to the contact with the Lycian Nappes (Bozkurt & Park 1999; Whitney & Bozkurt 2002; Rimmel6 et al. 2003a,b). Associated with the HP-LT metaconglomerate of the Menderes cover series, stretching lineations trend NE-SW (Rimmel6 et al. 2003b), and ENE-WSW stretching with top-to-the-ENE shear sense has been described in the upper levels of the marble envelope, approaching the contact with the Lycian Nappes (Collins & Robertson 2003; Rimmel6 et al. 2003a) (Fig. lc). In the augen gneisses and in the Palaeozoic schists, top-to-thenorth kinematic indicators overprinted by top-tothe-south fabrics have been recognized in the southern Menderes Massif, although the age of these deformation patterns is controversial. It has been claimed that the top-to-the-north fabrics formed during the Alpine orogeny, these kinematic indicators being overprinted by top-to-thesouth displacements (Bozkurt & Park 1994, 1999; Hetzel et al. 1998; Lips et al. 2001; Whitney & Bozkurt 2002). In contrast, other workers have proposed that the top-to-the-north displacements correspond to a pre-Alpine deformation and that the top-to-the-south shear senses were recorded during the main Alpine contractional episode (Ring et al. 1999a; Gessner 2000; Gessner et al. 2001a,b, 2004). The Cycladic Complex

in w e s t e r n T u r k e y

The Cycladic Blueschist Complex crops out in the Dilek Peninsula region (Fig. 1). This complex is made up of the 'Selguk Formation' overlying

450

G. RIMMELI~ E T AL.

the 'Kayaaltx Formation' (Giing6r & Erdo~an 2001; Fig. 2). The former consists of an olistostromal unit in which eclogite and eclogitic metagabbro have been found as blocks NE of Sel~uk (Candan et al. 1997; Oberh/insli et al. 1998; Cetinkaplan 2002) (Fig. 2a). These blocks, showing HP-LT assemblages, are surrounded by serpentinites and garnet micaschists (Giing6r & Erdo~an 2001). A correlation between the Selguk unit and a similar HP-LT metaolistostrome on Syros Island (Ridley & Dixon 1984; Okrusch & Br6cker 1990) has been proposed (Candan et al. 1997). The latter is composed of Mesozoic marbles intercalated with chloritoid-kyanite schists, blue amphibole-beating metabasites, and corundum- and diaspore-bearing metabauxites (Candan et al. 1997; Oberh/insli et al. 1998) (Fig. 2a). Candan et al. (1997) estimated P - T conditions for this blueschist-facies metamorphism of 10 kbar minimum and 470 ~ maximum; and 4~ dating on phengite yielded a Mid-Eocene (40 Ma) age for this HP-LT metamorphic event (Oberh/insli et al. 1998). In the westward lateral continuation of the Dilek Peninsula, the occurrence of similar blueschists on Samos Island (Mposkos & Perdikatsis 1984; Okrusch et al. 1984; Chen 1995; Will et al. 1998) led to a correlation of both blueschist terranes with the Cycladic Complex, with the Cycladic blueschists resting on top of the Menderes Massif (Candan et al. 1997; Oberh/insli et al. 1998; Will et al. 1998; Ring et al. 1999a,b; Gessner et al. 2001a,b,c; Okay 2001; Rimmel6 2003) (Fig. 1). In the Cycladic blueschist unit of the Dilek peninsula, Gessner et al. (2001b) described a topto-the-NE shearing deformation that aided early exhumation of HP-LT assemblages, and a subsequent top-to-the-south greenschist-facies event that led to the formation of the contact between the Cycladic blueschists unit and the Menderes nappes, the 'Cyclades-Menderes thrust'. The Lycian Nappes

The Lycian Nappes are exposed south of the Menderes Massif (Brunn et al. 1970; de Graciansky 1972; Poisson 1977, 1984; ()zkaya 1990; Ersoy 1993). They overthrust the metasedimentary sequence of the massif (de Graciansky 1972). As termed by Collins & Robertson (1997, 1998, 1999, 2003), from base to top, the 'Lycian Allochthon' is made up of the 'Lycian Thrust Sheets' composed of Upper Palaeozoic to Cenozoic sediments, the thick 'Lycian M61ange' unit, and the 'Lycian Peridotite Thrust Sheet' consisting of serpentinized peridotites with a metamorphic sole (Celik & Delaloye 2003). Overlying the HP-LT rocks of the Menderes 'cover', the basal 'Lycian Thrust

Sheets' are widely exposed on the Bodrum peninsula. These sediments recorded a continuous sedimentation from the Late Palaeozoic to the Late Cretaceous. They consist of Permo-Triassic reddish to greenish metapelites (the 'Karaova Formation' of Phillippson 1910-1915), overlain by a thick sequence of Triassic to Upper Cretaceous massive limestones and dolomites grading upwards to cherty limestones. The limestone sequence is topped by the Campanian to Maastrichtian 'Karab6~firtlen wildflysch' (de Graciansky 1972; Bernoulli et al. 1974; Cakmako~lu 1985; Okay 1989). Between Bodrum and Karab6~iirtlen (Fig. l b), Fe-Mgcarpholite-bearing rocks have been found in the basal Karaova Formation throughout the Bodrum peninsula, thus documenting an HP-LT metamorphic event recorded in the sediments of the Lycian Nappes (Oberh~insli et al. 2001; Rimmel6 2003; Rimmel6 et al. 2003a, 2005). Before this description of HP-LT metamorphism in the Lycian Nappes, the Karaova series was considered to have recorded only low-grade greenschist-facies conditions (Ashworth & Evirgen 1984b). P - T conditions for the HP-LT metamorphic peak are of 10-12 kbar and a maximum of 400 ~ and carpholite-bearing rocks recorded various retrograde paths depending on their structural position in the HP-LT unit (Rimmel6 et al. 2005). Based on stratigraphic observations, an age between the Late Cretaceous and Eocene has been suggested for the HPLT metamorphic event (Rimmel6 2003; Rimmel6 et al. 2003a,b; Jolivet et al. 2004). Recent a~ dating of phengite from two samples of the red-green phyllites of the Karaova series revealed Late Cretaceous ages for the HP-LT metamorphism (c. 70-90 Ma; Ring & Layer 2003). The palaeogeographical origin of the Lycian Nappes in western Turkey, although debated for many years, is now widely agreed to be in the Izmir-Ankara suture zone, north of the Menderes Massif (de Graciansky 1972; Diirr 1975; Diirr et al. 1978; Gutnic et al. 1979; Seng6r & Yllmaz 1981; Okay 1989; Collins & Robertson 1997, 1998, 1999, 2003; Giing6r & Erdo~an 2001; Rimmel6 2003; Rimmel6 et al. 2003a). On the Bodrum peninsula, Rimmel6 et al. (2003a) described NE-SW- to ENE-WSWtrending stretching with top-to-the-NE shear sense in the lowermost HP-LT metasediments of the Lycian Thrust Sheets (Karaova series) and in the uppermost metasediments of the southern Menderes Massif (Fig. lc). This deformation is coeval with exhumation of HP-LT rocks that was aided by the activation of major top-tothe-NE to top-to-the-east shear zones that postdated the southward translation of the Lycian

HP-LT ROCKS LYCIAN BELT, TURKEY

451

Fig. 2. (a) Geological map of the Cycladic Complex in the Dilek peninsula region (modified after 0nay 1949; Ozer 1993; Candan et al. 1997; Oberh~insli et al. 1998) showing the location of the two klippen of Lycian Nappes. Location of this map in SW Turkey is shown in Figure lb. (b) Map of the northern klippe (Kirazh area) and (c) map of the southern klippe (Tirhak6y area) (modified after Gfing6r & Erdo~an 2001), showing the location of HP-LT relics and associated deformation.

452

G. RIMMELI~ E T A L .

Allochthon (Rimmel6 et al. 2003a, 2005). Collins & Robertson (2003) also observed some top-tothe-east kinematic indicators in this area, at the Lycian-Menderes contact, but emphasized that this shearing deformation was recorded during the overall Late Cretaceous-Late Miocene top-to-the-east to top-to-the-SE translation of the Lycian Nappes over the Menderes Massif. In the Bodrum peninsula, earlier top-to-the-south structures contemporaneous with the southward transport of the nappe complex are preserved only in the uppermost levels of the Lycian Thrust Sheets (Fig. lc; Rimmel6 et al. 2003a). East of the village of Karab6~iirtlen, similar southeastward transport-related fabrics have been deciphered in the Lycian mdlange, as well as in peridotite slices (Fig. lc; Collins & Robertson 1998, 2003). Alternatively, a recent study concludes that all the top-to-the-NE directions observed in the southern Menderes Massif and at the base of the Lycian Nappes are related to a major shear zone located south of the Bodrum peninsula (south of the Lycian-Menderes contact), the so-called 'Dat~a-Kale main breakaway fault' along which the main exhumation of the Menderes Massif occurred (Seyito~lu et al. 2004). Up to now, the description of HP-LT assemblages and associated deformation features in the Lycian Nappes has been restricted to the Bodrum peninsula region (Rimmel6 et al. 2003a), although sporadic carpholite occurrences have been recognized in tectonic slices of the Lycian Nappes, north of the Bodrum peninsula (Oberhfinsli et al. 2001). During our investigations farther north, we found new localities of HP-LT rock occurrences from Karab6~firtlen to

~ivril along the contact between the Menderes Massif and the Lycian Nappes, and in several klippen of Lycian metasediments located on top of the Menderes Massif and the Cycladic Complex (Fig. lb). This paper focuses on three regions in which we report new outcrops of HP-LT rocks and describes the associated ductile deformation. From west to east, the study area encompasses the Dilek peninsula region, the Borlu area and the (~ivril region (Fig. 1b).

The Lycian Nappe Klippen of the Dilek peninsula In the Dilek peninsula region, south of Sel~uk (Fig. 2a), the Lycian Nappes are found as two tectonic slices (Giing6r & Erdo~an 2001). The base of both klippen is marked by rare outcrops of the typical red-green phyllites of the Karaova Formation, widely exposed farther south on the Bodrum peninsula (Fig. 2b and c). The northern klippe of Lycian metasediments is the largest one (Fig. 2b). It consists mainly of greyish and yellowish limestones and dolomites resting tectonically on top of the Sel~uk metaolistostromal formation (Giing6r & Erdo~an 2001). The red-green phyllites of the Karaova Formation are exposed only south of Kirazh (Fig. 2b). Pseudomorphs after carpholite have been found (Fig. 3a; Table 1). They show a total retrogression of Fe-Mg-carpholite to chlorite. The foliation is made of chlorite and phengite, and c. 100 ~tm long chloritoid crystals are recognized in thin sections (Fig. 3b). Pyrophyllite and kaolinite are found in quartz

Fig. 3. Photomicrographs showing the occurrence of pseudomorphs after Fe-Mg-carpholite (a; plane-polarized light) and chloritoid (b; crossed polars) within the Karaova red-green phyllites of the northern klippe of the Lycian Nappes in the Dilek peninsula region (south of Sel~uk). Both photomicrographs are from the sample KIRAZ2D (location shown in Fig. 2b). Car pseud, pseudomorph after carpholite; Cld, chloritoid; Chl, chlorite; Phg, phengite; Qtz, quartz.

HP-LT ROCKS LYCIAN BELT, TURKEY X

X

X o

X

X

X

X

X

X

.o +

+

o

x

x

+

+

x

o=

e~

x

x

x

-','e

X

+x

+

x

+x

+

x

X

X-X-

2

453

segregations that contain the pseudomorphs after carpholite (Table 1). The metasediments are intensively deformed. The stretching lineations roughly trend N010 ~ and shear senses are topto-the-SSW (Fig. 2b), as also reported by Gfing6r & Erdo~an (2001). The southern klippe is located in the eastern part of the Dilek peninsula, between Davutlar and S6ke) (Fig. 2a). As in the northern klippen, the same lithologies of the Lycian Nappes crop out in this region, south of Tirhak6y (Fig. 2c). The red-green phyllites of the Karaova Formation are also exposed in a very small area (a few hundred square metres) in the eastern part of the klippen. Similarly to the klippe near Kirazh, they tectonically overlie the Selguk Formation. Centimetre-scale fibres of Fe-Mg-carpholite are recognized within quartz segregations (Fig. 4a). At the microscopic-scale, Fe-Mg-carpholite appears as hair-like fibres in quartz (Fig. 4b) or forms assemblages with phengite and chlorite (Fig. 4c; Table 1). Whereas pyrophyllite and sudoite occur in the metapelites, chloritoid has not been found in the metapelites. In this area, the foliation of the strongly deformed metapelites is subvertical and stretching lineations trend approximately N030 ~ (Fig. 2c).

o

The Lycian Nappe klippe of Borlu L) ~ O

I

2

~o

L)

N

N;-~ N;-.~ N

o~ o~

ak ~

-,~ .*~~ ?~ o ~?.~ ,o ~ ~ o N ~ ,.~ ~ ,-~ .~

i ]

o o

o

:0

+.~

0

In the northern part of the Menderes Massif, a small relic of Lycian metasediments was identified by Oberhfinsli et al. (2001), NE of Borlu (see location in Fig. lb). This klippe is the northernmost known tectonic slice of Lycian Nappes. In this region, greyish-bluish schists of the Karaova Formation and thin beds of Triassic limestones of the Lycian Nappes overlie thin slices of Menderes 'cover' sequence rocks (marbles and staurolite-garnet schists), which overthrust highgrade garnet-staurolite-kyanite-bearing schists of the Menderes 'core' (Figs 5 and 6). In this area, carpholite has been recognized as hair-like fibres in quartz segregations (Fig. 7a; Table 1). The metasedimentary rocks of the Karaova Formation are severely deformed. Stretching lineations trend N120 ~ and metre-scale quartz segregations as well as S-C structures in the chloritoidbearing schists attest to top-to-the-ESE shearing (Fig. 7b~l).

The Lycian Nappes in the region of (~ivril . ,....~

0

.-~

"go

x~

In the eastern part of the Menderes Massif, in the region of (~ivril (location shown in Fig. lb), the Lycian Nappes sensu stricto and a klippe of Lycian metasediments crop out (Fig. 8). In this

454

G. RIMMELIS E T AL.

Fig. 4. (a) Fe-Mg-carpholite-bearing quartz segregation in the Karaova Formation of the southern klippe of the Lycian Nappes in the Dilek peninsula region (south of Tirhak6y). Photomicrographs showing hair-like fibres of Fe-Mg-carpholite within quartz (b; crossed polars) and Fe-Mg-carpholite + phengite + chlorite assemblages (c; crossed polars). Both photomicrographs are from the sample DAV 1B (location shown in Fig. 2c). Car, Fe-Mg-carpholite; Chl, chlorite; Phg, phengite; Qtz, quartz. area, the Lycian Thrust Sheets are composed of the Karaova Formation and the overlying Jurassic to Cretaceous limestones. They tectonically overlie the metasedimentary sequence of the Menderes Massif. Southwest of (~ivril, near G6mce, the klippe of Lycian Nappes overlies the thick Jurassic to Cretaceous marbles, as well as the typical Upper Cretaceous reddish cherty marbles of the Menderes Massif (Figs 8 and 9a). The basal Karaova Formation contains HP-LT assemblages such as Fe-Mg-carpholite + quartz. Fe-Mg-carpholite was also found in calcite crystals (Fig. 10a; Table 1), and chloritoid abundantly occurs within the schists of the Karaova Formation (Fig. 10b). South of the Lycian slice, the underlying metaolistostrome of the Menderes Massif crops out. As on the Bodrum peninsula, it comprises blocks of cherty marbles and metavolcanic lenses surrounded by a schist matrix in which blue and green amphiboles occur (Fig. 8; Rimmel6 2003). The whole formation is highly deformed. Stretching lineations trend N020-N030 ~ and S-C structures show top-to-the-north shear sense (Figs 8 and 9b). This metaolistostromal unit unconformably overlies the Upper Cretaceous

reddish marbles, which show eastward overturned folds. These folded structures, roughly trending north-south, were observed at the kilometre scale (Fig. 9a), as well as at the metre scale (Figs 9b and 10c). Southwest of the (~ivril region, similar kilometre-scale eastward overturned folds were described by Okay (1989), who interpreted them as a Late Eocene-Oligocene, post-nappe emplacement, deformation. Northeast of (~ivril, Fe-Mg-carpholite and chloritoid also occur near the Akda~ and I~lkh villages (location in Fig. 8). West of I~lkh, shearing deformation indicates movement towards the NNE, whereas between I~lkh and Akda~, the Fe-Mg-carpholite-chloritoid-bearing rocks from the Karaova Formation display intense deformation that is characterized by NW-SE-trending stretching lineations and top-to-the-NW shear sense (Figs 8 and 10d,e). Chloritoid occurs in the foliation and Fe-Mg-carpholite is nearly completely retrogressed to chloritoid+quartz (Fig. 10f; Table 1). The significance of the NW-SE-trending stretching and displacements towards the N W in this area is questionable because of the many surrounding neotectonic faults, which may have tilted and rotated the various blocks (Fig. 11).

HP-LT ROCKS LYCIAN BELT, TURKEY

455

Fig. 5. Geological map of the Borlu region (modified after Candan 1993) showing the location of HP-LT assemblages from a klippe of the Lycian Nappes on top of the rocks of the Menderes Massif. Location in SW Turkey is shown in Figure lb.

Farther south, in the N W of ~al (not on the map shown in Fig. 8; see location in Fig. lb), similar H P - L T assemblages have been found in the Karaova Formation. Relics of Fe-Mgcarpholite occur in quartz segregations within the chloritoid-bearing schists, and Fe-Mgcarpholite is commonly retrogressed to chlorite and pyrophyllite.

Chemistry of HP-LT index minerals Mineral analyses were performed with two electron microprobes, a Cameca SX50 at the

Universit6 Pierre et Marie Curie in Paris and a Cameca SX100 at the GeoForschungsZentrum in Potsdam. Both units were operated under standard conditions (15 kV accelerating voltage, 10-20nA current, PAP correction procedure, analytical spot diameter between 3 and 5 gm keeping the same current conditions), using natural and synthetic standard minerals (in Paris: Fe203 (Fe), MnTiO3 (Mn, Ti), diopside (Mg, Si), CaF2 (F), orthoclase (A1, K), anorthite (Ca), and albite (Na); in Potsdam: Fe203 (Fe), rhodonite [Mn], rutile (Ti), MgO (Mg), wollastonite (Si,

456

G. RIMMELI~ E T AL.

Fig. 6. Panorama showing the klippe of Borlu overlying the rocks of the Menderes Massif (northern part of the Menderes Massif). Location of sample BORLU001B is shown.

Fig. 7. (a) Photomicrograph showing hair-like fibres of Fe-Mg-carpholite within quartz in the area of Borlu, northern Menderes Massif (crossed polars, sample BORLU001B). Block diagram (b) and field photographs (c, d) showing the characteristics of the deformation in the metapelites of the Karaova Formation.

Ca), fluorite (F), orthoclase (A1, K), and albite (Na)). Structural formulae as well as the Fe 3§ and Fe 2§contents in Fe-Mg-carpholite and chloritoid were calculated following Goff6 & Oberh~insli (1992) and Chopin et al. (1992), respectively.

In the Dilek peninsula region, the composition of Fe-Mg-carpholite from both klippen is given by XMg [XMg= Mg/(Mg + Fe ~2+)+ Mn)] values ranging between 0.6 and 0.7; chloritoid compositions show XMgvalues of about 0.15 (Fig. 12;

HP-LT ROCKS LYCIAN BELT, TURKEY

457

Fig. 8. Simplified geological map of the ~ivril region (modified after Konak 1993; Ozer et al. 2001) showing distribution of metamorphic minerals and kinematic indicators. Location of the ~ivril area is shown in Figure lb. Tables 2 and 3). In the ~ivril region, chemical compositions of Fe-Mg-carpholite (0.6<XMg <0.65) and chloritoid (XMg= 0.2 on average) are roughly in the same ranges as these from the Dilek klippen (Fig. 12; Tables 2 and 3). For both minerals, the Mn-content is very low (XMn< 0.03). Similar compositions of carpholite and chloritoid have been reported in the same Karaova red-green metasedimentary rocks from the Bodrum peninsula, south of the Menderes Massif (0.55 < XMg(carpholite) < 0.7 and 0.1 < XMg(Chloritoid) < 0.2; Rimmel6 et al. 2005).

Discussion and conclusions Evidence f o r widespread H P - L T metamorphism Fe-Mg-carpholite assemblages are found widely in the metasedimentary rocks of the Karaova

Formation, which forms the base of the Lycian Nappes. As shown in Figure lb, they are found in the Lycian Nappes sensu stricto continuously from the Bodrum peninsula region to the region of ~ivril along the contact with the Menderes Massif, as well as in several klippen of Lycian Nappes located at the top of the Menderes Massif (Dilek peninsula region in the west; Borlu area in the north; ~ivril region in the east). Fe-Mgcarpholite-bearing rocks, therefore, occur over a distance of > 200 km in both north-south and east-west directions, thus documenting an extensive metamorphic belt. The similar chemical composition of the H P - L T index minerals over the whole of SW Turkey thus indicates that the metasedimentary rocks of the Lycian Nappes everywhere record similar metamorphic peak P - T conditions to those recently estimated in the Bodrum peninsula region, i.e. 10-12 kbar and 400~ (Rimmel6 et al. 2005), which corresponds to a burial of c. 30 km.

458

G. RIMMELI~ ET AL.

Fig. 9. Cross-sections showing the main structures in the southwestern part of the (~ivril area (location of cross-sections AB and CD is shown in Figure 8).

D e f o r m a t i o n related to H P - L T metamorphism In the Bodrum peninsula region, major top-to-the-NE and top-to-the-east shear zones are contemporaneous with exhumation of the Lycian H P - L T rocks (Rimmel6 et al. 2003a). In the t~ivril region, along the Lycian-Menderes contact, this study shows similar top-to-the-NE shear sense in the uppermost metaolistostromal unit of the Menderes Massif and in the Karaova Formation of the Lycian Nappes. The relevance of the few occurrences of top-to-the-NW shear sense near I~lkh must be questioned, because of the presence of recent faults, which encircle outcrops of the Karaova metasedimentary rocks and could have induced block rotations. This main deformation towards the NE, observed all along the contact between the Lycian Nappes and the Menderes Massif, from the Bodrum peninsula towards the (~ivril region (Fig. lc), shows an opposite direction of shear relative to the southward translation of the Lycian Nappes over the Menderes Massif. Along this contact, the redgreen phyllites of the Karaova Formation show neither top-to-the-south shear sense, as observed in the same basal lithologies from the klippen of the Dilek region, nor top-to-the-SE shear sense as deciphered in the Borlu klippen (Fig. l c). In the

Bodrum peninsula region, several similar top-tothe-south to top-to-the-SE movements are preserved only in the Karab6~firtlen wildflysch that constitutes the upper levels of the Lycian Thrust Sheets (Rimmel6 et al. 2003a), in the Lycian M61ange and in the overlying metamorphic sole of the Lycian Peridotite (Collins & Robertson 1998, 2003; Rimmel6 2003). It is here suggested that this deformation pattern with two opposite directions of kinematic indicators favours the idea of reactivation of the Lycian-Menderes contact as a major top-to-the-NE shear zone during exhumation of the Lycian HP-LT rocks, obscuring the evidence of the earlier deformation coeval with southward nappe transport. The idea of a northward backthrusting of the Lycian Nappes subsequent to their southward translation over the Menderes Massif has also been proposed, by Bozkurt & Park (1999). The Lycian Nappes, therefore, underwent a complex tectonometamorphic history from the Late Cretaceous (age of the HP-LT metamorphic imprint; Ring & Layer 2003) to the Miocene (age of final exhumation of H P - L T rocks; G. Rimmel6, unpublished apatite fission-track data). The timing and the cause of the episode during which the Lycian-Menderes contact was reactivated as a major top-to-the-NE shear zone still remain enigmatic.

HP-LT ROCKS LYCIAN BELT, TURKEY

459

Fig. 10. (a) Photomicrograph showing Fe-Mg-carpholite occurrence within calcite crystals in the Karaova metasediments of the Lycian klippen, in the ~ivril region (south of Grmce, sample CAL0104, crossed polars). (b) Photomicrograph of chloritoid occurrence in the Karaova Formation (south of Grmce, sample CAL0103A, plane-polarized light). (c) Eastward overturned folds in the Upper Cretaceous reddish marbles of the Menderes Massif. (d) S-C structures in the schists of the Karaova Formation showing top-to-the-NW shearing deformation (north of I~lkh). (e) Photomicrograph of an asymmetric quartz crystal and chloritoid-bearing pressure-shadows also indicating top-to-the-NW shear sense (north of I~lkh, sample CIVRIL001A2, crossed polars). (t) Photomicrograph of chloritoid grown from the breakdown of carpholite (north of I~lkh, sample CIVRIL001A1, plane-polarized light). Photographs (d), (e) and (f) are taken from samples collected at the outcrop shown in Figure 11. For photographs (e) and (f), thin sections were prepared from oriented samples where sections were cut parallel to the lineation and perpendicular to the foliation. Location of samples is shown in Figure 8.

G. RIMMELI~ ET AL.

460

Fig. 11. Panorama showing outcrops of the Fe-Mg-carpholite-chloritoid-bearing Karaova metasediments surrounded by active faults, north of I~lkh (NE of ~ivril). Location of I~lkh is shown in Figure 8. Mg

considerable implications for the geometry of the accretionary wedge in which the original sediments of the future Lycian Thrust Sheets were buried. At least 30 km of overburden are required to reach the P-T conditions of 10-12 kbar and 400 ~ as estimated by Rimmel6 et al. (2005). To 40 ~ 60 account for the palaeogeography of the various tectonic units of SW Turkey during the Cretaceous times, one has to restore the sediments, now forming the Lycian Thrust Sheets, to a position north of the present-day position of the Menderes Massif (Fig. 13). These units are interpreted as continental slope deposits forming part of the Anatolide-Tauride passive margin before closure of the Neotethyan Ocean (Collins Fe 20 40 60 80 Mn & Robertson 1997, 1998, 1999, 2003). A northward-dipping intra-oceanic subduction zone [] DILEKPENINSULAREGION I~DILEKPENINSULAREGION formed within the Neo-tethys Ocean during the A (~IVRILREGION A (~IVRILREGION Cretaceous, and in the latest Cretaceous times the southward-advancing ophiolitic complex was Fig. 12. Fe-Mn-Mg ternary diagram showing the obducted onto the passive margin (Collins & compositions of Fe-Mg-carpholite (white symbols) Robertson 1997) (Fig. 13). Probably during this and chloritoid (grey symbols) from the klippen of the Lycian Nappes located in the Dilek peninsula region, stage, the Lycian metasedimentary rocks were and from the (~ivril region. imbricated, metamorphosed under HP-LT conditions, and thrust southwards, at a depth of about 30 km (Fig. 13). Implications for the accretionary wedge One remarkable point that arises from this geometry study is the very wide distribution of wellpreserved Fe-Mg-carpholite throughout the hinThis discovery of widely distributed carpholitebearing assemblages in the basal metased- terland of the Lycian Allochthon, indicative of a imentary rocks of the Lycian Nappes has 200 km long (from Gfillfik to (~ivril) and 200 km

HP-LT ROCKS LYCIAN BELT, TURKEY

461

Table 2. Selected electron microprobe analyses of Fe-Mg-carpholite (location of samples & shown in Figs 2 and 8)

Sample: SiO2

A1203 FeO MnO MgO F Sum " Si A1 Fe 3+ Fe 2+ Mn Mg F XMg

CAL0104

CAL0104

CAL0104

DAV 1B

DAV 1B

DAV 1B

38.74 31.23 8.44 0.13 8.07 0.45 87.07 2.04 1.97 0.03 0.35 0.01 0.64 0.08 0.64

38.44 31.71 8.70 0.05 7.93 0.35 87.18 2.02 1.98 0.02 0.37 0.00 0.63 0.06 0.63

38.67 31.64 8.08 0.13 7.87 0.32 86.71 2.04 2.00 0.00 0.36 0.01 0.63 0.05 0.63

38.41 31.24 7.69 0.12 8.22 0.54 86.21 2.04 1.99 0.01 0.33 0.01 0.66 0.09 0.66

39.53 32.12 8.13 0.08 8.45 0.30 88.60 2.04 1.98 0.02 0.34 0.00 0.66 0.05 0.66

39.46 32.08 8.19 0.04 8.35 0.29 88.41 2.04 1.99 0.01 0.34 0.00 0.65 0.05 0.65

Table 3. Selected electron microprobe analyses of chloritoid (location of samples is shown in Figs 2 and 8) Sample: SiO2 TiO2 A1203 FeO MnO MgO Sum Si Ti A1 Fe 3+ Fe 2+ Mn Mg X(Mg)

KIRAZ2D

KIRAZ2D

KIRAZ2D

24.79 0.03 40.11 25.13 0.22 2.67 93.00 2.04 0.00 3.89 0.11 1.62 0.02 0.33 0.17

24.72 0.00 40.35 24.58 0.31 2.62 92.59 2.04 0.00 3.92 0.08 1.62 0.02 0.32 0.16

24.87

25.32

24.85

0.00 39.90 25.44 0.27 2.47 93.04 2.05 0.00 3.88 0.12 1.63 0.02 0.30 0.16

0.06 39.95 22.16 0.15 3.23 90.96 2.10 0.00 3.90 0.10 1.43 0.01 0.40 0.22

0.14 39.53 23.32 0.19 2.95 90.98 2.07 0.01 3.89 0.11 1.51 0.01 0.37 0.19

wide (from the Bodrum peninsula to Borlu) metamorphic belt. To restore the original geometry of the accretionary wedge in which H P - L T metamorphism is recorded, one has first to remove the effects of post-orogenic extension that segmented the Menderes Massif into several Neogene grabens. Closing these grabens still yields a significant distance of > 100 km separating the Fe-Mg-carpholite-bearing Lycian rock localities in a north-south direction, implying a particular geometry of the accretionary complex responsible for such widespread H P - L T metamorphism. Indeed, the width of the H P - L T Lycian Belt is systematically greater than that of similar carpholite-bearing metamorphic belts in the Mediterranean realm, for instance in the Betic Cordilleras (Goff6 et al. 1989; Azafion & Goff6

CIVRIL001A 1

CIVRIL001A 1

1997), in the Alps (Goff6 & Chopin 1986; Goff4 & Oberh~insli 1992; Bousquet et al. 1998; Agard 1999; Bousquet et al. 2002), in Crete and the Peloponnese (Theye et al. 1992, Jolivet et al. 1996; Trotet 2000) or in Oman (Goff6 et al. 1988; Michard et al. 1994). The accretionary complex could thus have been particularly wide and m a n y imbricated tectonic units, now composing the Lycian Nappes, formed quasi-simultaneously at depth during obduction. This accretionary wedge must have maintained a particularly cold environment (below 400 ~ under high-pressure conditions to explain such widely distributed well-preserved carpholite ('fresh' carpholite), the stability of carpholite being very sensitive to temperature changes. In the basal parts of a shallow-dipping accretionary complex, isotherms

462

G. RIMMELI~ E T AL.

roughly trend parallel to the tectonostratigraphy, and the pressure and temperature conditions are similar throughout a given tectonostratigraphic unit such as the Karaova unit. In this particular setting, narrow temperature ranges could be recorded at the same levels over wide areas. This could explain such large amounts of fresh carpholite and the wide mineral distribution in the basal metasediments of the Lycian Nappes. A similar setting has already been described for carpholite-bearing metasediments in the Alps (Bousquet et al. 2002; Goff6 et al. 2003). In contrast, a setting of a steep downgoing slab may not explain such a wide distribution or preservation of huge quantities of fresh carpholite, as the implied obliquity of isotherms relative to the lithostratigraphy would preclude the record of narrow temperature ranges at the same tectonostratigraphic levels over a wide zone. Therefore, it is here proposed that the construction of the Lycian H P - L T Belt, probably the widest carpholite-bearing metamorphic belt in the Mediterranean region, could be the result of the evolution of a wide southward-moving accretionary wedge over a nearly horizontal slab of continental material (Fig. 13). The first stages

of orogeny in SW Turkey were thus comparable with the tectonic scenario responsible for the H P LT metamorphism in Oman by the obduction of the doubled(?) oceanic crust onto the Arabian continental margin (Goff6 et al. 1988; Michard et al. 1994). The Lycian Nappes must then have maintained their H P - L T metamorphic imprint during the nappe translation towards the south over a relatively cold passive continental margin (the future Menderes Massif), and significant quantities of fresh carpholite survived this transport. However, well-preserved Fe-Mg-carpholite occurrences in some areas, compared with relict carpholite occurrences elsewhere, imply different exhumation histories after a common H P - L T metamorphic peak. O. 6. Dora is thanked for his support during our fieldwork in this region of western Turkey. We acknowledge C. Fischer for the quality and number of rock thin sections prepared. The Deutsche Forschungsgemeinschaft (DFG-project OB80/21-2), the Deutsch-Franz6sische Hochschule (DFH), the CNRS and the Volkswagen Stiftung are thanked for their financial support. A. Collins, T. Theye and O. Monod are thanked for their valuable comments on the first version of the manuscript.

Fig. 13. Schematic north-south cross-section showing the tectonic setting of the accretionary wedge during the latest Cretaceous time. This shows the southward obduction of the ophiolitic complex onto the sediments of the future Lycian Thrust Sheets, which were part of the Anatolide-Tauride passive margin. HP-LT metamorphism in the Lycian Thrust Sheets was probably recorded during this obduction episode.

H P L T ROCKS LYCIAN BELT, TURKEY

References AGARD, P. 1999. Evolution mktamorphique et structurale des mOtap~lites ocbaniques dans l'orogkne Alpin. l'exemple des Schistes LustrOs des Alpes occidentales (Alpes Cottiennes). PhD thesis, Universit6 Pierre et Marie Curie, Paris. AKKOK, R. 1983. Structural and metamorphic evolution of the northern part of the Menderes massif: new data from the Derbent area and their implication for the tectonics of the massif. Journal of Geology, 91, 342-350. ASHWORTH, J. R. & EVIRGEN, M. M. 1984a. Garnet and associated minerals in the southern margin of the Menderes Massif, southwest Turkey. Geological Magazine, 121(4), 323-337. ASHWORTH, J. R. & EVIRGEN, M. M. 1984b. Mineral chemistry of regional chloritoid assemblages in the Chlorite Zone, Lycian Nappes, southwest Turkey. Mineralogical Magazine, 48, 159-165. AZAlqON, J.M. & GOFFE, B. 1997. Ferro- and magnesiocarpholite assemblages as record of high-P, low-T metamorphism in the Central Alpujarrides, Betic Cordillera (SE Spain). European Journal of Mineralogy, 9, 1035-1051. BERNOULLI, D., DE GRAC1ANSKY,P. C. & MONOD, O. 1974. The extension of the Lycian Nappes (SW Turkey) into the southeastern Aegean Islands. Eclogae Geologicae Helvetiae, 67, 39-90. BOUSQUET, R., OBERHANSLI, R., GOFFI~, B., JOLIVET, L. & VIDAL, O. 1998. High-pressure-lowtemperature metamorphism and deformation in the Bfindnerschiefer of the Engadine window: implications for the regional evolution of the eastern Central Alps. Journal of Metamorphic Geology, 16, 657q574. BOUSQUET, R., GOFFI~, B., VIDAL, O., OBERHANSLI,R. & PATRIAT, M. 2002. The tectono-metamorphic history of the Valaisan domain from the Western to the Central Alps: new constraints on the evolution of the Alps. Geological Society of America Bulletin, 114(2), 207-225. BOZKURT, E. & OBERHANSLI,R. 2001. Menderes Massif (Western Turkey): structural, metamorphic and magmatic evolution--a synthesis. International Journal of Earth Sciences, 89(4), 679-708. BOZKURT, E. & PARK, R. G. 1994. Southern Menderes Massif--an incipient metamorphic core complex in Western Anatolia, Turkey. Journal of the Geological Society, London, 151, 213-216. BOZKURT, E. & PARK, R. G. 1999. The structures of the Palaeozoic schists in the southern Menderes Massif, western Turkey: a new approach to the origin of the Main Menderes metamorphism and its relation to the Lycian Nappes. Geodinamica Acta, 12(1), 2542. BRUNN, J. H., DE GRACIANSKY, P. C., GUTNIC, M., et al. 1970. Structures majeures et corr61ations stratigraphiques dans les Taurides occidentales. Bulletin de la Socidtk Gkologique de France, 12(3), 515-556. (~A~oLAYAN, A. M., Q)ZTURK, E. M., OZTURK, Z., SAY, H. & AKAT, U. 1980. Structural observations on the southern Menderes Massif. Publications

463

of The Chamber of Geological Engineers of Turkey, 10, 9-17 (in Turkish with English summary). ~AKMAKO~LU, A. 1985. Aydin N19-d3, Marmaris 019-a2 ve Denizli M21-d3-c4. M.T.A. report ('Paftasina ait genellestirilmis dikme kesit'). M.T.A. enstitfisfi Jeoloji, Bornova, Izmir. CANDAN, O. 1993. Petrography, petrology and metamorphism of the region between Demirci and Borlu at the northern flank of the Menderes Massif. Turkish Journal of Earth Sciences, 2, 69-87 (in Turkish with English summary). CANDAN, O., DORA, O. (~)., OBERH~.NSLI, R., OELSNER, F. & DISRR, S. 1997. Blueschist relics in the Mesozoic cover series of the Menderes Massif and correlations with Samos Island, Cyclades. Schweizerische Mineralogische und Petrographische Mitteilungen, 77, 95-99. CANDAN, O., DORA, O. (~)., OBERH.ANSLI, R., ~ETINKAPLAN, M., PARTZSCH, J. H., WARKUS, F. C. & D~RR, S. 2001. Pan-African high-pressure metamorphism in the Precambrian basement of the Menderes Massif, western Anatolia, Turkey. International Journal of Earth Sciences, 89(4), 793-811. CANDAN, O., (~ETINKAPLAN, M., OBERHA,,NSLI, R. & RIMMEL~, G. 2002. Fe-Mg-carpholitepyrophyllite--chloritoid-bearing Triassic metapelites from Afyon Zone, Turkey: first evidence for Alpine low-grade, high-Pllow-T metamorphism. First International Symposium of the Faculty of Mines (Istanbul Technical University) on Earth Sciences and Engineering, Istanbul, Turkey, 107. CANDAN, O., ~ETINKAPLAN, M., OBERHANSLI, R., RIMMELE, G. & AKAL, C. 2006. Fe Mg carpholite occurrence as a record of Alpine high-P/low-T metamorphism of Afyon Zone and implication for metamorphic evolution of western Anatolia, Turkey. Lithos, 84, 102-124. ~ELIK, (~). F. & DELALOYE,M. F. 2003. Origin of metamorphic soles and their post-kinematic mafic dykes swarms in the Antalya and Lycian ophiolites, SW Turkey. Geological Journal, 38, 235-256. ~ETINKAPLAN, M. 2002. Tertiary high pressure~low temperature metamorphism in the Mesozoic cover series of the Menderes Massif and correlation with the Cycladic Crystalline Complex. PhD thesis, Dokuz Eyliil Universitesi, Izmir. CHEN, G. 1995. Evolution of the high- and mediumpressure metamorphic rocks on the island of Samos, Greece. Annales Gkologiques des Pays Hell~niques, 36(0, 799-915. CHOPIN, C., SEIDEL, E., THEYE, T., FERRARIS, G., IVALDI, G. & CATrI, M. 1992. Magnesiochloritoid, and the Fe-Mg series in the chloritoid group. European Journal of Mineralogy, 4, 67-76. COLLINS, A. S. & ROBERTSON, A. H. F. 1997. Lycian mtlange, southwestern Turkey: an emplaced Late Cretaceous accretionary complex. Geology, 25, 255-258. COLLINS, A. S. & ROBERTSON, A. H. F. 1998. Processes of Late Cretaceous to Late Miocene episodic thrust-sheet translation in the Lycian Taurides, SW Turkey. Journal of the Geological Society, London, 155, 759-772.

464

G. RIMMELI~ ET AL.

COLLINS, A. S. & ROBERTSON, A. H. F. 1999. Evolution of the Lycian Allochthon, western Turkey, as a north-facing Late Palaeozoic to Mesozoic rift and passive continental margin. Geological Journal, 34, 107-138. COLLINS, A. S. & ROBERTSON, A. H. F. 2003. Kinematic evidence for Late Mesozoic-Miocene emplacement of the Lycian Allochthon over the Western Anatolide Belt, SW Turkey. Geological Journal, 38, 295-310. DE GRACIANSKY, P. C. 1966. Le massif cristallin du Menderes (Taurus occidental, Asie Mineure). Un exemple possible de vieux socle granitique remobilisd. Revue de G~ographie Physique et de Gkologie Dynamique, SGrie 2, 8(4), 289-306. DE GRACIANSKY, P. C. 1972. Recherches gdologiques dans le Taurus Lycien occidental. PhD thesis, Universitd Paris Sud. DI3RR, S. 1975. (Jber Alter und geotektonische Stellung des Menderes-Kristallins/SW-Anatolien und seine Aequivalente in der mittleren Aegaeis. Habilitation Thesis, Univesity of Marburg/Lahn. DURR, S., AETHER, R., KELLER, J., OKRUSCH, M. & SEIDEL, E. 1978. The median Aegean crystalline belt: stratigraphy, structure, metamorphism and magmatism. In: CLOSS, H., ROEDER, D. R. & SCHMIDT, K. (eds) Alps, Apennines, Hellenides. Schweizerbart, Stuttgart, 455-477. ERSOY, S. 1993. The geological setting of the tectonic units situated on the SW Anatolia (Turkey) and their geodynamic development. Bulletin of the Geological Society of Greece, 28, 617-628. GESSNER, K. 2000. Eocene nappe tectonics and lateAlpine extension in the central Anatolide belt, western Turkey. Structure, kinematics and deformation history. PhD thesis, Johannes-Gutenberg Universit/it Mainz. GESSNER, K., PIAZOLO, S., GI3NGOR, T., RING, U., KRONER, A. & PASSCHIER, C. W. 2001a. Tectonic significance of the deformation patterns in granitoid rocks of the Menderes nappes, Anatolide belt, southwest Turkey. International Journal of Earth Sciences, 89(4), 766-780. GESSNER, K., RING, U., PASSCHIER, C. W. 8~ GI3NGOR, T. 2001b. How to resist subduction: evidence for large-scale out-of-sequence thrusting during Eocene collision in western Turkey. Journal of the Geological Society, London, 158, 769-784. GESSNER, K., RING, U., JOHNSON, C., HETZEL, R., PASSCHIER, C. W. & G13NGOR, T. 2001c. An active bivergent rolling-hinge detachment system: Central Menderes metamorphic core complex in western Turkey. Geology, 29(7), 611-614. GESSNER, K., COLLINS, A. S., RING, U. & G1JNGOR, T. 2004. Structural and thermal history of polyorogenic basement: U-Pb geochronology of granitoid rocks in the southern Menderes Massif, Western Turkey. Journal of the Geological Society, London, 161, 93-101. GOFFI~, B. & CHOPIN, C. 1986. High-pressure metamorphism in the Western Alps: zoneography of metapelites, chronology and consequences. Schweizerische Mineralogische und Petrographische Mitteilungen, 66, 41-52.

GOFF1~, B. & OBERHANSLI,R. 1992. Ferro- and magnesiocarpholite in the 'Bfindnerschiefer' of the eastern Central Alps (Grisons and Engadine window). European Journal of Mineralogy, 4, 835-838. GOFFt~, B., MICHARD,A., KIENAST,J. R. & MER, O. L. 1988. A case of obduction-related high pressure, low temperature metamorphism in upper crustal nappes, Arabian continental margin, Oman: P - T paths and kinematic interpretation. Tectonophysics, 151, 363-386. GOFFI~, B., MICHARD, A., GARCIA-DUENAS,V., et al. 1989. First evidence of high pressure, low temperature metamorphism in the Alpujarride nappes, Betic Cordillera (SE Spain). European Journal of Mineralogy, 1, 139-142. GOFFI~, B., BOUSQUET,R., HENRY, P. & LE PICHON,X. 2003. Effect of the chemical composition of the crust on the metamorphic evolution of orogenic wedges. Journal of Metamorphic Geology, 21, 123-141. GI3NGOR, T. & ERDOGAN, B. 2001. Emplacement age and direction of the Lycian Nappes in the SGke-Selguk region, western Turkey. International Journal of Earth Sciences, 89(4), 874-882. GUTNIC, M., MONOD, O., POISSON, A. & DUMONT, J. F. 1979. GGologie des Taurides occidentales (Turquie). Mdmoires de la Sociktd Gdologique de France, 137, 1-112. HETZEL, R. & REISCHMANN, T. 1996. Intrusion age of the Pan-African augen gneisses in the southern Menderes Massif and the age of cooling after Alpine ductile extensional deformation. Geological Magazine, 133(5), 565-572. HETZEL, R., ROMER, R. L., CANDAN, O. & PASSCHIER, C. W. 1998. Geology of the Bozda~ area, central Menderes Massif, SW Turkey: Pan-African basement and Alpine deformation. Geologische Rundschau, 87, 394-406. JOLIVET, L., GOFEI~, B., MONII~,P., TRUFFERT-LUXEY, C., PATRIAT, M. & BONNEAU, M. 1996. Miocene detachment in Crete and exhumation P-T-t paths of high-pressure metamorphic rocks. Tectonics, 15(6), 1129-1153. JOLIVET, L., FACENNA, C., GOFFI~, B., BUROV, E. & AGARD, P. 2003. Subduction tectonics and exhumation of high-pressure metamorphic rocks in the Mediterranean orogens. American Journal of Science, 303, 353-409. JOLIVET, L., RIMMELI3,G., OBERH~,NSLI,R., GOFFI~, B. 8,~ CANDAN, O. 2004. Correlation of syn-orogenic tectonic and metamorphic events in the Cyclades, the Lycian Nappes and the Menderes Massif. Geodynamic implications. Bulletin de la Sociktk Gkologique de France, 175(3), 217-238. KONAK, N. 1993. Structural characteristics of the Menderes Massif around Cal-Civril-Karahalli Tfirkiye Jeoloji Kurultayi, Bildiri 6zleri. Geological Congress of Turkey, Ankara, Abstracts, p. 32. (in Turkish). KONAK, N., AKDENIZ, N. & ()ZTI3RK, E. M. 1987. Geology of the south of Menderes Massif. In: 'Correlation of Variscan and pre-Variscan events of the Alpine Mediterranean Mountain Belt. Field Meeting, IGCP Project 5. Publications of the Mineral Research and Exploration Institute of Turkey', pp. 42-53.

HP-LT ROCKS LYCIAN BELT, TURKEY LIPS, A. L. W., CASSARD,D., SOZBILIR,H., YILMAZ,H. & WIJBRANS, J. R. 2001. Multistage exhumation of the Menderes Massif, Western Anatolia (Turkey). International Journal of Earth Sciences, 89(4), 781-792. MICHARD, A., GOFFI~, B., SADDIQI, O., OBERH.~NSLI, R. & WENDT, A. S. 1994. Late Cretaceous exhumation of the Oman blueschists and eclogites: a two-stage extensional mechanism. Terra Nova, 6, 404-413. MPOSKOS, E. & PERDIKATSlS, V. 1984. Petrology of glaucophane metagabbros and related rocks from Samos, Aegean Island (Greece). Neues Jahrbuch fiir Mineralogie, Abhandlungen, 149(1),..43-63. OBERH,~NSLI, R., CANDAN, O., DORA, O. O. & Dt3RR, S. H. 1997. Eclogites within the Menderes Massif, western Turkey. Lithos, 41, 135-150. OBERH,~NSLI, R., MONII~,P., CANDAN,O.:. WARKUS, F. C., PARTZSCH, J. H. & DORA, O. O. 1998. The age of blueschist metamorphism in the Mesozoic cover series of the Menderes Massif. Schweizerische Mineralogische und Petrographische Mitteilungen, 78, 309-316. OBERH,~NSLI, R., PARTZSCH, J., CANDAN, O. & @ETINKAPLAN, M. 2001. First occurrence of Fe-Mg-carpholite documenting a high-pressure metamorphism in metasediments of the Lycian Nappes, SW Turkey. International Journal of Earth Sciences, 89(4), 867-873. OBERH.~NSLI, R., WARKUS, F. & CANDAN, O. 2002. Dating of eclogite and granulite facies relics in the Menderes Massif. In: First International Symposium of the Faculty of Mine (Istanbul Technical University) on Earth Sciences and Engineering, Istanbul, Turkey, 104. OKAY, A. I. 1989. Geology of the Menderes Massif and the Lycian Nappes south of Denizli, western Taurides. Bulletin of Mineral Research and Exploration (Turkey), 109, 37-51. OKAY, A. I. 2001. Stratigraphic and metamorphic inversions in the central Menderes Massif: a new structural model. International Journal of Earth Sciences, 89(4), 709-727. OKAY, A. I. 2002. Jadeite-chloritoid-glaucophanelawsonite blueschists in northwest Turkey: unusually high PIT ratios in continental crust. Journal of Metamorphic Geology, 20, 757-768. OKAY, A. I. & KELLEY, S. P. 1994. Tectonic setting, petrology and geochronology of jadeite + glaucophane and chloritoid +glaucophane schists from north-west Turkey. Journal of Metamorphic Geology, 12, 455-466. OKAY, A. I. & MONIE, P. 1997. Early Mesozoic subduction in the Eastern Mediterranean: evidence from Triassic eclogite in northwest Turkey. Geology, 25(7), 595-598. OKAY, A. & T(;rvst3z, O. 1999. Tethyan sutures of northern Turkey. In: Durand, B., Jolivet, L., Horvath, F. & S6ranne, M. (eds) The Mediterranean Basins: Tertiary Extension within the Alpine Orogen. Geological Society, London, Special Publications, 156, 475-515. OKAY, A. I., HARRIS, N. B. W. & KELLEY, S. P. 1998. Exhumation of blueschists along a Tethyan suture in northwest Turkey. Tectonophysics, 285(3-4), 275-299.

465

OKAY, A. I., TANSEL, 1. & TOYS~Z, O. 2001. Obduction, subduction and collision as reflected in the Upper Cretaceous-Lower Eocene sedimentary record of western Turkey. Geological Magazine, 138(2), 117-142. OKAY, A. I., MONOD, O. & MONII~, P. 2002. Triassic blueschists and eclogites from northwest Turkey: vestiges of the Paleo-Tethyan subduction. Lithos, 64, 155-178. OKRUSCH, M., RICHTER, P. & KATSIKATSOS, G. 1984. High-pressure rocks of Samos, Greece. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 529536. OKRUSCH, M. & BROCKER, M. 1990. Eclogites associated with high grade blueschists in the Cyclades archipelago, Greece: a review. European Journal of Mineralogy, 2, 451-478. ONAY, T. S. 1949. Uber die Smirgelgesteine SWAnatoliens. Schweizerische Mineralogische und Petrographische Mitteilungen, 29, 359-484. I~)ZER, S. 1993. Upper Cretaceous rudists from the Menderes Massif. Bulletin of the Geological Society of Greece, 28(3), 55-73. OZER, S. 1998. Rudist bearing Upper Cretaceous metamorphic sequences of the Menderes Massif (Western Turkey). Geobios, 22, 235-249. OZER, S., SOZBILIR, H., (~)ZKAR, I., TOKER, V. & SARI, B. 2001. Stratigraphy of Upper CretaceousPalaeogene sequences in the southern and eastern Menderes Massif (Western Turkey). International Journal of Earth Sciences, 89(4), 852-866. 0ZKAYA, I. 1990. Origin of the allochthons in the Lycian belt, Southwest Turkey. Tectonophysics, 177, 367-379. PARTZSCH, J. H., OELSNER, F. & OBERH,~NSLI,R. 1997. The Menderes Massif, W Turkey: a complex nappe pile recording 1.0 Ga of geological history? Terra Abstract, 9, 394. PHILLIPPSON, A. 1910-1915. Reisen und Forschungen imwestlichen Kleinasien. P etermanns Geographische MitteilungenlErg~inzungsheft, 167-183. POISSON, A. 1977. Recherches gbologiques dans les Taurides occidentales ( Turquie) . PhD thesis, Universit6 Paris Sud. POISSON, A. 1984. The extension of the Ionian trough into southwestern Turkey. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 241249. RI~GNIER, J. L., RING, U., PASSCHIER,C. W., GESSNER, K. & GONGOR, T. 2003. Contrasting metamorphic evolution of metasedimentary rocks from the ~ine and Selimiye nappes in the Anatolide belt, western Turkey. Journal of Metamorphic Geology, 21, 699-721. RIDLEY, J. & DIXON, J. E. 1984. Reaction pathways during the progress of deformation of a blueschist metabasite: the role of chemical disequilibrium and restricted range equilibrium. Journal of Metamorphic Geology, 2, 115-128.

466

G. RIMMELI~ ET AL.

RIMMELI~, G. 2003. Structural and metamorphic evolution of the Lycian Nappes and the Menderes Massif ( S W Turkey): geodynamic implications and correlations with the Aegean domain, PhD thesis, Universit~it Potsdam/Universit6 d'Orsay-Paris XI. RIMMELI~, G., JOLIVET,L., OBERHANSLI,R. & GOFFI~, B. 2003a. Deformation history of the high-pressure Lycian Nappes and implications for tectonic evolution of SW Turkey. Tectonics, 22(2), 1007, doi: 10.1029/2001TC901041. R1MMELI~,G., OBERHANSLI,R., GOFFI~,B., JOLIVET,L., CANDAN, O. & tTETINKAPLAN, M. 2003b. First evidence of high-pressure metamorphism in the 'Cover Series' of the southern Menderes Massif. Tectonic and metamorphic implications for the evolution of SW Turkey. Lithos, 71, 19-46, doi: 10.1016/S0024-4937(03)00089-6. RIMMELI~, G., PARRA,T., GOFFI~,B., OBERH~i.NSLI,R., JOLIVET, L. & CANDAN,O. 2005. Exhumation paths of high pressure-low temperature rocks from the Lycian Nappes and the Menderes Massif (SW Turkey): a multi-equilibrium approach. Journal of Petrology, 46(3), 641-669, doi: 10.1093/petrology/ egh092. RING, U. & LAYER, P. W. 2003. High-pressure metamorphism in the Aegean, eastern Mediterranean: underplating and exhumation from the Late Cretaceous until the Miocene to Recent above the retreating Hellenic subduction zone. Tectonics, 22(3), doi: 10.1029/2001TC001350, 2003. RING, U., GESSNER, K., GUNGOR,T. & PASSCHIER,C. W. 1999a. The Menderes Massif of western Turkey and the Cycladic Massif in the A e g e a n ~ o they really correlate? Journal of the Geological Society, London, 156, 3-6. RING, U., LAWS, S. & BERNET, M. 1999b. Structural analysis of a complex nappe sequence and lateorogenic basins from the Aegean Island of Samos, Greece. Journal of Structural Geology, 21, 15751601. RING, U., WILLNER, m. P. & LACKMANN, W. 2001. Stacking of nappes with different pressuretemperature paths: an example from the Menderes Nappes of western Turkey. American Journal of Science, 301, 912-944. SATIn, M. & FRIEDRICHSEN, H. 1986. The origin and evolution of the Menderes Massif, W Turkey: a rubidium/strontium and oxygen isotope study. Geologische Rundschau, 75, 703-714.

SCHUILING, R. D. 1962. On petrology, age and structure of the Menderes migmatite complex (SW Turkey). Bulletin of the Institute for Mineral Research and Exploration, Turkey, 58, 71-84. ~ENGOR, A. M. C. & YILMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 181-241. ~ENGOR, A. M. C., SATIn, M. & AKKOK, R. 1984. Timing of tectonic events in the Menderes Massif, western Turkey: implications for tectonic evolution and evidence for Pan-African basement in Turkey. Tectonics, 3, 693-707. SEYITO6LU, G., I~lK, V. & ffEMEN, I. 2004. Complete Tertiary exhumation history of the Menderes massif, western Turkey: an alternative working hypothesis. Terra Nova, 16(6), 258-364. SHERLOCK, S., KELLEY,S. P., INGER, S., HARRIS,N. & OKAY, A. I. 1999. 4~ and Rb-Sr geochronology of high-pressure metamorphism and exhumation history of the Tavs.anli Zone, NW Turkey. Contributions to Mineralogy and Petrology, 137, 46-58. THEYE, T., SEIDEL, E. & VIDAL, O. 1992. Carpholite, sudoite and chloritoid in low-grade high-pressure metapelites from Crete and the Peloponnese, Greece. European Journal of Mineralogy, 4, 487 507. TROTET, F. 2000. Exhumation des roches de haute pression-basse temperature le long d'un transect des Cyclades au Pkloponnkse, implications g~odynamiques. PhD Thesis, Universit6 Paris Sud. WARKUS, F. C. 2001. Untersuchungen an Hochdruckrelikten im zentralen Menderes Massiv, W Tiirkei. PhD thesis, Universit~it Potsdam. WHITNEY, D. L. & BOZKURT, E. 2002. Metamorphic history of the southern Menderes massif, western Turkey. Geological Society of America Bulletin, 114(7), 829-838. WILL, T., OKRUSCH,M., SCHM,~DICKE,E. & CHEN, G. 1998. Phase relations in the greenschist-blueschistamphibolite-eclogite facies in the system Na20CaO-FeO-MgO-AI:O3-SiO2-H20 (NCFMASH), with application to metamorphic rocks of Samos, Greece. Contributions to Mineralogy and Petrology, 132, 85-102. YAL~IN, l~l. 1987. Petrologie und Geochemie der Metabauxite S W-Anatoliens. PhD thesis, Universit~it Bochum.

Synthesis of the tectonic-sedimentary evolution of the Mesozoic-Early Cenozoic Pindos ocean: evidence from the NW Peloponnese, Greece P A U L J. D E G N A N 1 & A L A S T A I R

H . F. R O B E R T S O N 2

1UK Nirex, Curie Avenue, Harwell, Didcot O X l l ORH, U K (e-mail: paul. degnan@nirex, co. uk) 2Grant Institute o f Earth Science, University o f Edinburgh, Edinburgh E H 9 3JW, U K The tectonic development of the western part of the Pindos ocean in southern Greece is exemplified by the mountainous Pindos thrust belt in the NW Peloponnese. A Late Triassic-Early Cenozoic succession exposed within imbricate thrust sheets records a range of deep-water siliciclastic, redeposited carbonate and siliceous sediments, which in general become more distal oceanwards towards the east. Igneous rocks, locally dated as Triassic, occur within a m61ange that is entrained beneath and within the Pindos thrust stack; these igneous rocks and related sediments are interpreted as remnants of a continent-ocean transition zone. 'Immobile' element geochemistry is explicable by rifting of a compositionally heterogeneous subcontinental mantle, possibly related to pre-existing Hercynian subduction, although coeval Triassic subduction cannot be excluded based on evidence from this area alone. Localized, 'enriched' basalts are interpreted as fragments of oceanic seamounts formed in a relatively distal setting. Late Paleocene-Early Eocene (locally MidEocene) siliciclastic turbidites, derived from the north, record the latest deposition prior to incorporation of the sedimentary succession into a westward-migrating accretionary wedge during post-Early Eocene time in the NW Peloponnese. Structural restoration of the wellordered thrust stack indicates a minimum of 201 km (55%) of shortening at an average rate of 5.8 mm a-1. As the Pindos allochthon approached the Apulian continent, the GavrovoTripolitza foreland underwent flexural upwarp during the Mid-Eocene, followed by collapse to create a foreland basin by the Late Eocene. This basin was infilled with generally upwardthickening and -coarsening deep-water turbiditic sediments of Late Eocene-Early Oligocene age. The foreland was, in turn, overthrust by the Pindos accretionary prism during post-Early Paleocene time, and was then imbricated and thrust over the Ionian foreland basin to the west by Pliocene time. Abstract:

Forming the backbone of mainland Greece and extending into Albania, the Pindos Mountains have attracted considerable interest since the classic regional mapping and stratigraphical studies of Dercourt (1964), Aubouin et al. (1970), British Petroleum (1971), Jacobshagen et al. (1978), Fleury (1980) and Dercourt et al. (1986, 1993) (Fig. 1). Western Greece is traditionally subdivided into a series of tectonostratigraphic units, originally termed isopic zones (Aubouin et al. 1970). The more westerly of these include the Gavrovo-Tripolitza zone and the Pindos (or Pindos-Olonos) zone (Pindos suture in Fig. 1). The Gavrovo-Tripolitza zone forms the foreland to a thrust belt represented by the Pindos zone. The Pindos zone has been variously interpreted as a Triassic rift located close to Gondwana (Aubouin et al. 1970; Dercourt et al. 1986; Yllmaz et al. 1996), a mid-ocean ridge-type basin (Smith et al. 1975; Robertson & Dixon 1984; Robertson et al. 1991, 1996; Smith 1993),

or a marginal basin variously related to southward subduction (~eng6r et al. 1984), northward subduction (Stampfli & Borel 2002; Stampfli et al. 1998, 2001), or partially related to westward intra-oceanic subduction (Pc-Piper & Piper 2002). The main objective here is to synthesize the constructive and destructive evolution of the Pindos ocean based mainly on evidence from the N W Peloponnese (Fig. 2). First, we summarize the rift and passive margin evolution of the western margin of the Pindos ocean. We then discuss the importance of a m61ange unit beneath and within the thrust stack for understanding the nature of the continent-ocean transition zone, utilizing previously unpublished igneous geochemical data. We then present detailed structural information for a well-exposed structural traverse in the N W Peloponnese and interpret this as a westward-migrating accretionary prism. We also outline evidence for the emplacement

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 467491. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

468

P.J. DEGNAN & A. H. F. ROBERTSON

Moesian Platform

Scutari Pe~ line

,) NW Peloponnese

\ ;

\

8 r

,,~'~.o(3.

20 ~ I Fig. 1. Setting of the Pindos suture and the location of the study area in the NW Peloponnese.

of the Pindos allochthon over the GavrovoTripolitza platform to the west following the development of a related foreland basin. Finally, we summarize the history of opening and closure of the Pindos ocean in southern Greece in the light of the wider regional tectonic setting. R i f t - p a s s i v e margin development The rifted passive margin of the Pindos ocean is recorded by the Triassic-Early Cenozoic deepwater sediments of the Pindos Group that are exposed as a stack of imbricate thrust sheets (Robertson 1994; Degnan & Robertson 1998; Fig. 2). The differential thicknesses and facies of successions exposed in more proximal (westerly) to more distal (easterly) areas have influenced the style of subsequent deformation, as discussed later in the paper. The oldest sediments, mainly exposed in the westerly, structurally lower thrust sheets, known as the Priolithos Formation (Degnan & Robertson 1998) are mainly sandstones (litharenites),

siltstones, shales, calcilutites and nodular limestones. The sandstones contain abundant detritus from metamorphic (e.g. quartzite, mica schist) and plutonic igneous (e.g. granitic units) sources and include sparse to locally abundant volcanic material of mainly intermediate to silicic composition. The Triassic sandstones plot in the litharenite field of the Q:F:L diagram (McBride 1963) and in the recycled orogen field on the Qm:F:L diagram (Dickinson & Suczek 1979). The Triassic limestones contain Holobia sp. and conodonts, indicative of a Late Triassic (Carnian-Norian) age (Flament 1973). The Triassic mixed carbonate-siliciclastic sediments pass depositionally upwards into debris flows, carbonate turbidites and hemipelagic carbonates of the Carnian-Liassic Drimos Formation. In the Early Jurassic there was a regional change to contrasting argillaceous sediment deposition, known as the Kastelli Mudstone Member (Leesteena Formation). These sediments pass upwards into reddish, mainly siliceous facies, including well-developed ribbon radiolarites

PINDOS OCEAN NW PELOPONNESE, GREECE

469

GULF OF CORINTH

PATRAS Panahaiko Mtns. Glafkos

:i::-:::::.:/-i...::i:........,. i":"...i:"i.......-,-::"i !i..-i.i

KALAVRITA

" . : : : . : . : : i : . : " i i " . :.'.:. ". .-i..: .'. : -:; Alepohorio 9".~i i i .." -' STRUCTURAL

9".'." .TRAVERSE

..~ '". " c

.,,. ::..,,....: .,-..v.-.-..-: vl

.~

....... .,, " i,-." ..,,..' . 9,,, ~..~

KLITORIA w

*Larnbia

~ari '

Kondovazena ~ ;~'~--

~_.~. 0

10 km

20

/

Thrust fault

/

Normal fault

t

N

KEY ~ Quaternary-Holocene,unconsolidated E F T ] Plio-Quaternary, consolidated ~ I

Eocene-Oligocene

clastics

I Mesozoic Pindos Group

Fig. 2. Outline tectonic map of the NW Peloponnese, showing approximate trajectory of the structural traverse studied and location of places discussed in this paper.

of the Middle-Upper Jurassic Aroania Chert Member, associated with manganese of probable hydrothermal origin (Pc-Piper & Piper 1989; Robertson & Degnan 1998). Local developments of fine-grained calpionnellid limestones of Late Jurassic (Tithonian) age mainly occur in the south Peloponnese. Manganese-rich cherts of Late Jurassic age are locally interbedded with volcaniclastic sediments and overlie strongly altered volcanic rocks (e.g. at Aroania, Kombigadi and Drimos; Robertson & Degnan 1998). By the Tithonian, there was a swich to the accumulation of fine-grained carbonates with varying amounts of pelitic material in relatively distal settings represented by the Paos Limestone Member (Lambia Formation; Neumann et al. 1996; Neumann & Zacher 2004). There are occasional dark, locally organic-rich intervals (Neumann et al. 1996; Wagreich et al. 1996; Neumann 2003). During Cenomanian-Turonian

time terrigenous turbidites, known as the Klitoria Sandstone Member (Premier Flysch de Pinde), are exposed within thrust imbricates located in the central parts of the thrust traverse. These sediments lack sand-sized ophiolitic debris as seen in equivalent sediments north of the Gulf of Corinth, but clay mineral studies of interbedded fine-grained sediments suggest an ophiolitic source within the Peloponnese (Thi6bault & Fleury 1980). A thick sequence of Late Cretaceous pinkish-greyish pelagic carbonates, represented by the Erymanthos Limestone Member (Lambia Formation) follows, interspersed with minor greyish organic-rich layers (Neumann et al. 1996; Neumann 2003). Around the Cretaceous-Cenozoic boundary there was a return to periodic siliceous and organic-rich sedimentation, known as the Kataraktis Passage Member of the Lambia Formation (Robertson & Degnan 1998), or the 'Couches de

470

/

P . J . D E G N A N & A. H. F. R O B E R T S O N

<

..~z

,o

<

~

~z

.-

LU

~ !

0

.

tU

"

-=.

UJ

o ~

/kU_~

"'''

":=

hOj!

;0i

s

I -

~

i

~

I~,

= I

I.u

i

%

i ,,

.o~;

~,

PINDOS OCEAN NW PELOPONNESE, GREECE Passage' (Flament 1973; Fleury 1980). The age of the base of this member ranges from Mid- to Late Maastrichtian or even Campanian in different areas (see Piper 2006). Siliceous sediments of the Kataraktis Passage Member are interbedded with, and overlain by, relatively fine-grained carbonate conglomerates derived from the Gavrovo-Tripolitza platform to the west. More proximal successions of Cretaceous age, located in structurally lower thrust sheets, contain abundent channelized limestone conglomerates, interpreted as a base-of-slope apron (Degnan & Robertson 1998). Piper (2006) confirmed that successions become more distal from west to east across the Peloponnese as a whole. He also showed that palaeocurrents are generally towards the SE, although data are sparse. Results obtained during this study support this interpretation (Fig. 3). Sedimentation culminated in the deposition of siliciclastic turbidites known as the Pindos Flysch Formation ('Flysch de Pinde'), of mainly Late Paleocene-Early Eocene age, but extending to Mid-Eocene age in the extreme SE of the Peloponnese (see Piper 2006). Degnan & Robertson (1998) reported mainly southward palaeocurrents in the N W Peloponnese (Fig. 3). These results are based on measurements of micro-cross-lamination, primary current lamination, groove casts and sole marks. Differences in current orientation were occasionally noted between sole marks and bedding-internal structure; sole-mark data were used preferentially in such cases. Piper (2006) noted generally SE-directed palaeocurrents in the N W Peloponnese. Southdirected palaeocurrents were reported from the west central and eastern Peloponnese (Richter 1993). By contrast, sparse palaeocurrents in the SW Peloponnese are to the NE (see Piper 1984, 2006 for literature review). Degnan & Robertson (1998) interpreted the Pindos flysch in the NW Peloponnese as turbiditic sedimentation in a subduction trench setting with sand supply mainly from the north. Transitional facies represented by carbonate debris flows in the west record a continuation of input from the adjacent Gavrovo-Tripolitza carbonate platform (Degnan 1992). Piper (2006) envisaged sea-level fall rather than tectonics as the dominant control of Pindos flysch deposition in the Peloponnese. Sediment was potentially derived from both the Apulian foreland and the overthrusting Pelagonian active margin in northern Greece, followed by southward transport into the Peloponnese area. Both tectonic setting and sea level are likely to have contributed to the deposition.

471

Nature o f the continent-ocean transition zone

Clues to the nature of the basement beneath the deep-sea sediments of the Pindos Group are provided by the presence of igneous rocks that, where dated, are mainly of Triassic age (see Pe-Piper & Piper 2002, for literature review). Some of these volcanic rocks occur as local thrust sheets intercalated with shallow-water to deepwater sediments in different areas. They are also more extensively represented by a volcanicsedimentary m61ange that is entrained beneath and within the Pindos thrust stack. In several areas, the Pindos thrust sheets include coherent units of igneous rocks. For example, north of the Gulf of Corinth (near Mesophyton) Triassic Halobia limestones include a Jurassic? dolerite sill (Green 1982). Further west in the Lakmon Mountains shallow-marine sedimentary rocks are interbedded with lavas, agglomerate and tuff; extrusive rocks include andesites and shoshonites (Pc-Piper 1983; Pe-Piper & Piper 2002). Scattered throughout the Peloponnese the Pindos thrust sheets are underlain by, or intercalated with, a thin (tens of metres) unit of m61ange, termed the Formation ~t Blocs by De Wever (1976a). This typically occurs between the Gavrovo--Tripolitza foreland unit and the overlying Pindos thrust sheets at a number of localities (Fig. 4) throughout the Peloponnese and also to the north of the Gulf of Corinth (De Wever 1975, 1976a,b; Richter & Lensch 1977, 1989; Dercourt et al. 1978; Pe-Piper & Piper, 1998). In the central Peloponnese the m61ange includes basalt blocks (near Methedrio and Lagadi), and in the southern Peloponnese, a local m61ange exposure (at Sellas) includes basic extrusive blocks resembling pillow basalt. In the Peloponnese, most of the igneous rocks occur as blocks in a m61ange, although primary sedimentary contacts with deep-sea sediments are preserved locally. In the far SW Peloponnese, at Kokino, blocks of volcanic and volcaniclastic rocks (up to 20 m thick) are associated with hyaloclastites and sills. Above this, a sequence of volcanic rocks, volcanogenic sediments and minor pink Halobiabearing limestones passes depositionally upwards into well-bedded Halobia-bearing hemipelagic limestones of Late Triassic age. This unit can be correlated with the locally exposed base of the Pindos Group (Degnan & Robertson 1994), and is important as it establishes that the oldest sediments exposed in the Pindos thrust sheet (Priolithos Formation) accumulated on a volcanic basement.

472

P.J. DEGNAN & A. H. F. ROBERTSON 9" 9: 9- ' 9.-.'

i:.

I

!

\

~

so,

o

i)

"

,

/:~:~:/;:

~

.

.

l'

."/.

- ]" "

~ 9 . K.okmO

~I,~

. ]

" ""~-

: /:Jz " " " :: " ~

"~

"~

!

~7: :!";(:~\ ."."',

.'." :'." ~.".'.':-b'.:','.......

9 SoIIos

~ ,,

eohouni!>)

/-

-~,o

213o

: : <: :"

Z

'h

" '. ' -

.

.'.:.'.

'"] .' -.'.' ,'. ".'.ii'. " .' : :-k:.'.':."''.'.:.'.

.... : .

'. -.-

~,-,~-,~. . . . . . . . . . . . . . . .

.

F~,o~,Nzo,~ ~r

~,~

GAVROVO-

t

~

~

I=~:~I PELAGONIAN zONE

TRIPOLITZA ZONE

Fig. 4. Outline map of part of the Peloponnese showing the main locations of the m61ange that is entrained at the base of and within the Pindos thrust sheets. From Degnan & Robertson (1994).

Where locally present beneath the Pindos thrust sheets in the N W Peloponnese, the area of main concern here, the m61ange typically includes large volumes of strongly sheared sedimentary rocks, especially near the structural base. The extrusive blocks (mostly < 1 m 3 in size) are mainly basic-intermediate-siliceous volcanic rocks, volcaniclastic rocks and rarely intrusive igneous rocks (e.g. dolerite). Plagiogranite has also been reported (Pe-Piper & Piper 1991,2002). Most of the m61ange blocks are undated, but at one locality (Mati) limestone crusts attached to an igneous block contain Triassic fossils (De Wever 1975). Sheared sedimentary rocks occur in the matrix of the m61ange and as strongly dismembered thrust sheets in some places. These sediments can be correlated with lithologies of both the Gavrovo-Tripolitza zone beneath (e.g.

sandstones, redeposited limestones with benthic Foraminifera, brown shales) and with the Pindos zone above (e.g. pelagic and redeposited limestones, quartzose sandstones, radiolarian cherts, red shales). The matrix of the m61ange was created by tectonic fragmentation of mainly sedimentary material from both the Gavrovo and Pindos units. The igneous lithologies (e.g. basalt) are preserved as relatively large competent blocks (up to 20 m in size), whereas the less competent sedimentary rocks (e.g. volcaniclastic sandstone) have experienced pervasive layer-parallel extension on all scales, giving rise to phacoid-shaped blocks. A prominent m61ange exposure in the eastern part of the N W Peloponnese, at Drakovouni (Fig. 4) includes numerous metre-scale blocks of basalt that were created by the fragmentation of pre-existing basalt flows. Pillow interstices

PINDOS OCEAN NW PELOPONNESE, GREECE

473

Drakovouni again plot mainly in the WPB field. In the Ti/Cr v. Ni diagram (Beccaluva et al. 1979) samples are spread over the MORB and IAT fields. The least fractionated basalts were also plotted on MORB-normalized 'spidergrams' (Pearce 1980) and were found to exhibit nearMORB compositions of 'immobile' trace elements, commonly with a negative Nb anomaly (Degnan 1992). In several samples (sample 605, central Peloponnese and sample 421, eastern Peloponnese (Table 1)) no relative Nb depletion is apparent. Many, but not all of the basalts from the m61ange exhibit an apparent subduction influence, as shown by a depletion of high field strength elements (HFSE), relative to an enrichment in large ion lithophile elements (LILE). The basalts show a wide range of largely overlapping compositions. Overall, two types of basalts can be recognized: one is similar to MORB, but typically with a Nb depletion suggestive of a subducGeochemical fingerprinting o f basaltic rocks tion influence; the other, as at Drakovouni, is The geochemistry of the extrusive rocks within more WPB-like. the m61ange may help to indicate the nature and The limited available evidence suggests that tectonic setting of the crust beneath the deep-sea many of the basalts are Triassic in age, although Pindos sediments. As the rocks are all altered some could be younger. For example, sparse only 'immobile' major elements and trace ele- extrusive igneous and pyroclastic rocks directly ments were used for tectonic discrimination underlie Late Jurassic radiolarian cherts and (Pearce & Cann 1973; Pearce 1980). The litholointerbedded hydrothermal manganese deposits, gies are mainly basaltic andesites but range from notably at Aroania (Robertson & Degnan 1998); basalt, to andesite, to rhyodacite. A previous these volcanic rocks could be post-Triassic. study of the stable trace elements (c. 20 samples) Noting the common subduction influence and suggested that most are subalkaline; a few are the rare presence of plagiogranites, Pe-Piper & enriched, akin to within-plate ocean-island Piper (1991) suggested eruption in an intrabasalts (OIB) (Pe-Piper & Piper 1991). The oceanic subduction setting within the Pindos extrusive rocks in the m61ange fall within the ocean. Pe-Piper & Piper (1991) suggested that the range of compositions of Triassic rocks from Triassic volcanic rocks formed in a back-arc numerous other areas of Greece, as summarized marginal basin related to a generally eastwardby Pe-Piper & Piper (2002). dipping subduction zone located in the south During this work a large number of lava Aegean region (present coordinates). However, samples (c. 100) were collected to determine if they suggested more recently (Pe-Piper & Piper any significant geochemical differences exist 2002) that some of the Triassic volcanic rocks along or across the strike of the m61ange outcrops (e.g. high-Mg extrusive rocks and shoshonites) throughout the Peloponnese as a whole. Sixty- could relate to an additional westward-dipping two samples remained after removal of markedly intra-oceanic subduction zone, located within the fractionated and highly altered rocks (i.e. SiO2 eastern part of the Triassic Pindos ocean, near the outside the range 35-80%; MgO outside 3-9% Pelagonian zone (i.e. two opposing subduction and CaO outside 5-15%). In a Ti v. Zr diagram zones would have existed). As a simpler alternative we suggest that the most samples plot in the overlapping mid-ocean ridge basalt (MORB), volcanic arc basalt (VAB) geochemical variation in the Triassic basalts and within-plate basalt (WPB) fields (Fig 5a). relates to variable-degree melting of heterogeHowever, samples from Drakovouni plot in the neous subcontinental mantle that was modified WPB field. In the Zr/Y v. Zr diagram (Pearce by the addition of subduction fluids, possibly 1980) many sample from Drakovouni also related to Hercynian subduction (Robertson & plot in the WPB field (Fig. 5b). In the C r v . Y Dixon 1984; Dixon & Robertson 1993). This, diagram unfractionated basalts plot in the over- however, would not apply to the WPB-like lapping VAB-MORB and WPB-MORB fields, basalts (e.g. at Drakovouni). These are assoor the VAB field alone (Fig. 5c). Samples from ciated with oceanic (rather than rift-related)

are infilled by pelagic sediment, either carbonate or manganiferous chert. Generally smaller blocks of pyroclastic tuff, lava breccia and lava conglomerate are also present. Within several of the larger igneous blocks, short intact successions can be recognized, e.g. basalt passing into dolerite and tuff or volcaniclastic sediment. This locality is notable as successions within the thrust stack are mainly restricted to Cretaceous and Palaeogene ages; also, the volcanic rocks (e.g. at Drakovouni) are commonly of enriched, within-plate-type (see below). Elsewhere in this eastern area the m61ange occurs extensively around Chelmos (Fig. 4), where it contains blocks derived from the Gavrovo footwall, the overlying Pindos thrust sheets and from a basicintermediate volcanic sequence associated with volcaniclastic sediments (Degnan & Robertson 1994).

474

P.J. DEGNAN & A. H. F. ROBERTSON

/wO

10

105

%

E Q.

O

~" I~

104

~

~

O 9R

B

F-

~ 10

103 101

102

a

103

Zr (ppm)

A O

.... ,

i , . . ,,,i

,, i,,,, . . . . .

IO0 Zr (ppm)

b 104

Cr

103

500-

~\~

:

MORB

102 I-

(ppm)

101 100-

VA

10 ~

i

|

i

i

iiii

I

i

l

i

H

10 50-

L I i 1 i

9 0

oi

d

:

o"

~

WPB~

: :

C

5

ii

I

'

I

100

~

i l i

1000

Ni (ppm)

o!

KEY

~0

o

~

Drakovouni

Central Peloponnese ~, Southern Peloponnese 9 Eastern Peloponnese 9 Central Imbricates o Frontal Imbricates

0

t

I0

[i

1;

~o Y (ppm)

Fig. 5. 'Immobile' element geochemical plots of basic lava blocks from the m61ange beneath the Pindos thrust in different areas; (a) Ti v. Zr; (b) Zr/Y v. Zr; (c) Crv. Y; (d) Ti/Cr v. Ni. (See text for discussion.)

sediments and are found in a relatively easterly, distal location, and so these basalts might record fragments of accreted oceanic seamounts. Assuming the above interpretation of the Triassic subduction-influenced volcanic rocks is correct, a restoration of the Pindos basin represented by the imbricate thrust slices containing m61ange at the base (see below), suggests that these basalts are likely to have formed within 60 km oceanwards of the rifted continental margin edge. These basalts could, therefore, have erupted within a continent-ocean transition zone. Modern continent-ocean transition zones remain poorly understood, except for several

examples such as the Atlantic margin of Iberia (e.g. Whitmarsh et al. 2001) and the Alps (Manatschal et al. 2003). This has been mainly explored by the drilling of topographic highs and it is not known whether extrusive igneous rocks are present in the intervening basins.

Genesis and emplacement of the Pindos thrust sheets The Pindos ocean was consumed in an eastdipping subduction zone during Late Cretaceous-Palaeogene time (Robertson et al. 1991). The subduction zone reached the westernmost

PINDOS OCEAN NW PELOPONNESE, GREECE

475

Table 1. Selected chemical analyses of Triassic basic extrusive igneous rocks from the m61ange beneath and within the Pindos thrust sheets at several localities in the Peloponnese. Analysis was by X-ray fluorescence according to the method of Fitton & Dunlop (1985). The data are given in full in Degnan (1994). LOI, loss on ignition. Locality: Sample no.:

Drakovouni

Central Pelop.

276

453

454

TiO2 MnO P205 LOI Total

51.36 18.3 8.09 8.5 2.17 6.39 0.12 1.37 0.15 0.35 4.18 101.6

49.09 17.12 10.09 5.74 8.45 4.77 0.15 2.01 0.2 0.56 3.83 99.99

Ni Cr V Sc Cu Zn Sr Rb Zr Nb Ba Pb Th La Ce Nd Y

114 157 152 24 34 64 324 0 183 9 53 1 0 13 32 15 23

SiO2 A1203 Fe203. MgO CaO Na203

K20

38 107 248 38 36 79 166 1 237 10 83 0 0 22 50 28 35

E. Pelop.

S. Pelop.

588

605

421

423

356

399

50.13 16.02 9.42 5.62 7.07 5.5 0.19 2.07 0.18 0.64 3.24 100.08

46.64 14.81 8.74 4.21 11.46 4.81 0.95 1.06 0.11 0.18 7.45 100.42

48.7 18.1 6.19 9.05 7.32 3.06 0.63 1.18 0.15 0.18 4.9 99.46

45.56 14.8 7.85 6.36 11.16 4.34 0.16 0.79 0.15 0.16 9.3 100.62

52.28 15.4 4.46 2.54 9.97 4.08 2.69 0.75 0.1 0.3 7.64 99.95

50.65 17.48 5.42 8.38 7.96 2.97 0.26 1.21 0.12 0.23 5.57 100.24

50.19 16.55 7.33 3.04 8.41 6.92 0.06 0.69 0.23 0.23 6.25 99.9

20 73 210 28 59 69 388 2 297 14 28 3 3 9 51 36 43

136 328 310 37 20 53 403 13 82 75 3 3 1 3 28 19 21

part of the P i n d o s o c e a n in the N W P e l o p o n n e s e soon after the a c c u m u l a t i o n o f the Pindos flysch during Late Paleocene to Early or M i d - E o c e n e . The process o f s u b d u c t i o n - a c c r e t i o n , m o d i f i e d by post-accretionary tightening, gave rise to the present P i n d o s thrust stack. Previously, the structure o f particularly the G a v r o v o - T r i p o l i t z a footwall in the N W P e l o p o n n n e s e was considered in detail by Xypolias & D o u t s o s (2000). Also, within the Pindos thrust stack of the s o u t h e r n Peloponnese, Skourlis & D o u t s o s (2003) recognized a dense array o f m o d e r a t e - a n g l e imbricate thrusts in the west, with broader, upright to m o d e r a t e l y inclined t h r u s t sheets in the centre, a n d inclined to r e c u m b e n t folds in the east o f the thrust stack. In the orogenic m o d e l o f D o u t s o s et al. (2006), the P i n d o s thrust stack was subdivided into a

85 201 244 40 60 73 397 7 89 4 93 4 3 5 12 16 23

141 475 267 42 4 61 145 3 58 3 42 2 1 8 15 9 17

25 41 194 16 44 54 307 22 101 7 53 1 2 16 40 24 19

83 149 233 46 85 67 357 0 122 6 99 0 2 8 26 15 21

10 7 172 37 18 107 156 1 73 4 15 5 1 11 30 15 55

'pro-wedge' in the west, an 'uplifted plug' in the central P e l o p o n n e s e (i.e. C h e l m o s massif) a n d a 'retro-wedge' in the east. D u r i n g this w o r k ( D e g n a n 1992) a detailed structural transect was carried out across the N W Peloponnese, c o m b i n e d with reconnaissance o f adjacent areas (Fig. 2). This s h o w e d that the thrust belt in this area can be usefully subdivided into three subareas: the F r o n t a l Imbricates, the Central Imbricates a n d the E a s t e r n Imbricates, based on distinctive structural features. A detailed cross-section given in Figure 6.

Frontal Imbricates T h e F r o n t a l Imbricates are b o u n d e d to the west by the Pindos basal thrust, w h i c h at A l e p o h o r i o (Figs 2 and 7) dips at 20-30 ~ eastwards, a l t h o u g h

476

P . J . D E G N A N & A. H. F. R O B E R T S O N

~ 9

~

o

0

0 I

o

I

E J

Iv.."

~J

"~z~7

0 .. r i

p,,"

<12

0

5 0

~"

0

el)

0

o4

= o

! p/"

E_

~

o ".

,'".

~s

]

j"

.o

o'

5

5

o .~

e i

c~ "~

9 =

i

o

~ D.,

o

0

=

g ~

~

'-

~

"E

9

PINDOS OCEAN NW PELOPONNESE, GREECE a discrete thrust plane is rarely exposed. The structurally lowest strata of the Pindos Group comprise a 'broken formation', but form more coherent thrust sheets upwards. There are marked variations in age of the stratigraphic units exposed in the lowermost thrust units. For example, Triassic sediments (Priolithos Formation) are exposed at Alepohorio in the north, whereas Upper Triassic-Liassic (Drimos Formation) and Upper Jurassic (Lesteena Formation) sediments are seen further south. A possible reason is that the basal thrust cuts variably up and down section within the Pindos hanging-wall section along strike. Exposures in the frontal region are dominated by Cretaceous limestones (Lambia Formation). Throughout the frontal imbricates, bedding is generally moderately to steeply eastward dipping. The typical fold style varies from open concentric folds to medium-scale chevron folds, with gently plunging fold hinges (both NE- and SW-trending). The axial planes are moderately to steeply dipping. The trend of the fold-hinge lineations implies that the mean emplacement direction in the frontal imbricates was towards the N N W on the transect studied (Fig. 8). The strata are disharmonically folded and numerous highly contorted folds and small thrusts are seen in the thicker-bedded units. Numerous fold trains of low-amplitude concentric folds are exposed. Fold inter-limb angles are generally wide, with wavelenghs of up to c. 5 m; axial surfaces are steeply inclined eastwards, to vertical. The thicknesses of largely intact sequences in the Cretaceous interval (Erymanthos Limestone Member) are estimated at 250-300 m. The apparent thicknesses of Cretaceous limestone reaches 500-600 m and reflect thickening as a result of folding and the development of intraformational duplexes. The structural style of the frontal imbricates is exemplified at Alepohorio, where a large-scale synform (amplitude > 300 m) is present in the footwall to a major thrust at the head of the Alepohorio valley (Fig. 7). The axial surface of this structure is horizontal to weakly eastward dipping and strikes towards 020 ~. Both limbs of the major synform are preserved on the southern side of the valley. The oldest sediments are exposed beneath the lower limb, whereas equivalent lithologies of the eastern limb may be dissected by a later thrust. Medium-scale folds are preserved in the lower limb. By contrast, an east-vergent, isoclinal fold with chevron folds is developed on the upper limb. On the north side of the valley, the upper limb of the macroscopic fold

477

has been largely thrust-out, but the enveloping surfaces of a fold train (a series of chevron folds) define part of the synform. Mapping of the thrust sheets in the eastern part of the Frontal Imbricates demonstrates that a splay of minor thrusts is connected by branch lines to a major thrust fault in the footwall that converges to the south and north. The smaller thrust sheet corresponds to a hanging-wall cut-off at the present level of erosion.

Central Imbricates Within the Central Imbricates (from just east of Platanitza to the Aroania thrust; Figs 2 and 6) the sediments are generally thinner bedded than in the Frontal Imbricates, reflecting more distal deposition (Degnan & Robertson 1998). The Central Imbricate folds range in amplitude and wavelength from < 1 cm to > 200 m. Disharmonic, medium-scale folds dominate the Jurassic interval (Lesteena Formation) and the overlying Cretaceous interval (Erymanthos Limestone Member). Chevron folds exhibit c. 60 ~ interlimb angles, although some folds are in places strongly flattened. Larger-scale folds are well exposed in thickly bedded, competent, intraformational limestone beds within the Jurassic siliceous interval (Aroania Chert Member; e.g. at Livardzi and Drimos, Fig. 2). Folding there is highly irregular, with wavelengths of c. 100-300 m. Small-scale chevron folds and local box folds apparently formed prior to larger-scale folding, as the enveloping surfaces of fold trains are themselves folded. The longer-wavelength folds correspond to tightening of a 'single layer' to form irregular buckle folds. Major thrust planes that duplicate large parts of the stratigraphy are generally not well exposed in the Central Imbricates. Mapping indicates that most of the major thrust faults are hangingwall and footwall flats. Local hanging-wall or footwall cut-offs probably represent thrust deformation of pre-existing folds rather than thrusts cutting up through an undeformed 'layercake' sequence. Thrusting of previously folded strata also accounts for local younger-on-older stratigraphic juxtaposition. Many of the thrust faults die out along strike as doubly plunging structures. The maximum stratigraphic separation is in the centre of the mapped thrust traces, with displacement disappearing at both ends. Asymmetrical antiformal folds are seen at the surface (e.g. west of Livardzi). Shortening directions derived from mean fold-hinge lineations indicate movement towards 297-298 ~ in the west and 284-293 ~ in the east (Fig. 8).

478

P.J. D E G N A N & A. H. F. ROBERTSON

o

.....,4

. ,,,,,q

2 ~o az~:

o ~

~N

9

0

0 N

~.~

~~ o; o

o

s~

e

PINDOS OCEAN NW PELOPONNESE, GREECE

Eastern Imbricates The Eastern Imbricates (from east of Aroania to west of the Chelmos antiform; Figs 2 and 6) differ from those further west as the spacing between major thrust faults that duplicate the stratigraphy is wider there. Bedding flattens on a map scale into more gently folded antiform-synform couplets, and higher stratigraphic levels (Cretaceous) are more widely exposed. Thrust planes are steep in the west, but become shallower eastwards (e.g. 20-30~ Fig. 6). Medium-scale folds again form fold trains, with enveloping surfaces connecting fold inflection points and surfaces that were seen to be folded at several localities. Folded enveloping surfaces define the limbs of outcrop-scale folds, as in the Central Imbricates, although some folds may have resulted from later south-to-north directed compression; this would account for the observed wide spread of bedding and hinge lineation orientations. In the eastern part of the Eastern Imbricates (Fig. 7) bedding is inclined both ESE and also westwards. Shortening directions derived from fold-hinge orientations imply a mean emplacement direction of 270 ~ Further west there is a 23 ~ change in the inferred emplacement direction (Fig. 9). Minor thrusts show a wide spread of fault plane orientations, but with a concentration dipping ESE. Normal fault trends are also widely dispersed, although WSW-ENEstriking, high-angle fault planes are statistically significant. One other notable feature is that Jurassic siliceous sediments (Aroania Chert Member) are generally the oldest unit exposed in the Eastern Imbricates (east of Kato Klitoria), compared with the Western and Central Imbricates in which the complete Triassic-Early Cenozoic stratigraphy is eposed.

A ccretionary processes The style of deformation records the processes of detachment of the sedimentary cover from the transitional-oceanic crust that was being subducted eastwards. We interpret the sequence of events as indicated below. Initial compression within the Pindos ocean resulted in the generation of medium-scale folds. Continued shortening produced serial sinusoidal folds, which tightened into chevron folds within competent units. By contrast, incompetent units experienced largely bedding-parallel shear. Fold amplification and tightening continued until strain accommodation was achieved by flattening of sedimentary multilayers, and meso-scale folds

479

were incorporated into larger-scale folds. During decoupling of the sedimentary cover from oceanic crust, thrust faults propagated upwards from a basal detachment along axial-planar lines of weakness, exploiting the previous large-scale folding. In places, it can be seen that transverse ramps and high-angle faulting have controlled the local structural style. These faults reflect differential stresses within the thrust sheets and are clearly coeval with the main phase of deformation. Some additional deformation may relate to fault-propagation folding. A good example of this sequence of events is suggested by the thrust/ fold relations near Drimos (Fig. 2) in the Eastern Imbricates (Fig. 9). It should be noted that more recent N W - S E normal faulting is also present, reflecting neotectonic extension focused on the Gulf of Corinth. A minimum estimate of the shortening across the Pindos thrust-belt can be made. To achieve an effective restoration, the base of the Cretaceous limestones (Lambia Formation) is extrapolated to link hanging-wall and footwall cut-offs (with minimum values). The restored line-length also takes into account the shortening caused by mesoscale folding and the volume loss through dissolution (estimated from stylolite occurrences). Medium-scale folding over a 100m section was measured and shortening was estimated at three representative localities in each of the three imbricate regions (Alepohorio, Livardzi and Aroania). Taking account of the observed variation in pressure solution across the thrust belt the volume loss is estimated as 3.5% (with a possible range from 1 to 18%). The overall results suggest that there has been at least 60% shortening in the 43 km-wide traverse of the imbricate thrust sheets (Degnan 1992). The Pindos allochthon extends eastwards from the Eastern Imbricates (east of Chelmos) for a further 99 km. Of this, the 56 km wide 'Table d'Arcadie' (Dercourt 1964) consists almost entirely of the flat-lying, strongly deformed, Late Cretaceous pelagic carbonates (Erymanthos Limestone Formation). It is difficult to evaluate the shortening in this area; 50% was suggested by Fleury (1980) but it may be much more. The total shortening across the exposed Pindos thrust belt is estimated as at least 55% and more probably c. 67%, suggesting that this part of the pre-existing Pindos basin was at least 300 km wide. Green (1982) estimated orogenic contraction at 63% for the Pindos thrust stack north of the Gulf of Corinth. A similar result was obtained by Xypolis & Doutsos (2000). In addition, an unknown amount of oceanic crust was subducted, and so the Pindos ocean basin

480

P . J . D E G N A N & A. H. F. R O B E R T S O N

o,I Zr7 111

\ j z ~o~ z

II CI ZW co o41 r7

ILl

~r7 II W

z~

o~

z~

y

r,., a II W

,~

z~

.~

8~ 0 N N

~~

o

~,_~ ~ ON co r7

Z

~r7 r II w

b~

g~

0

~w

.~

z~

-,o

I

PINDOS OCEAN NW PELOPONNESE, GREECE was probably much wider (Robertson et al. 1991). The period of deformation when the preserved part of the Pindos ocean represented by the thrust belt in the N W Peloponnese was telescoped from > 300 km to 99 km is equivalent to an average shortening rate of c. 5.8 mm a -~. This rate is similar to that for some other mountain systems; e.g. 5 mm a -1 in the Canadian Rockies (Elliott 1976); or 6 mm a -1 in the Pyrenees (Williams & Fischer 1984). The development of the imbricate fan in the west in contrast to the 'Table d'Arcadie' region in the east probably relates to the presence of relatively competent thick and homogeneous Cretaceous pelagic limestones in the east. This compares with the more heterogeneous, generally older sequence further west; in these, incompetent layers (e.g. shale) provided the theological contrasts necessary for the development of a linked thrust stack system. In the frontal region of the thrust belt (Western Imbricates) there is a curvature of the thrust traces and a radial arrangement of both transverse faults and emplacement vectors (Fig. 8). A possible explanation is that the orogenic front was buttressed by a promontory of the Apulian margin, now located in the Chelmos region. This collapsed when it was overridden by the Pindos thrust sheets; insufficient topographical relief perhaps remained to influence thrust geometry. Emplacement structures were influenced by similar continental margin promontories elsewhere, including the Appalachians, the Rocky Mountains and Oman.

Evolution of the foreland The Pindos thrust complex was emplaced over the eastern margin of the Apulian continent in post-Early Oligocene time, as dated by the youngest sediments exposed in a thick underlying clastic foreland basin succession of the GavrovoTripolitza zone (Kamberis et al. 2006; Fig. 1). The underlying carbonate platform documents a Triassic rift history, coeval with the opening of the Pindos ocean, and a passive margin history contemporaneous with the deep-sea sedimentation in the Pindos ocean. The Gavrovo subzone forms the footwall of the Pindos thrust sheets in the N W Peloponnese, whereas the Tripolitza subzone is exposed in tectonic windows through the Pindos thrust stack further east (e.g. Chelmos). During the Mesozoic the GavrovoTripolitza zone was bordered to the west by the Ionian zone, an intracontinental rift basin (Blumenthal 1933; Renz 1955; Dercourt 1964; De

481

Wever 1975; Fleury 1980; Thirbault 1982; Clews 1989).

Rift-passive margin evolution In the N W Peloponnese the lowest exposed stratigraphic unit is composed of Triassic low-grade siliciclastic metamorphic rocks, rich in mica (Zarouhla Group; De Wever 1975). This sequence includes metavolcanic rocks (tens to several hundred of metres thick), mainly composed of trachyandesite, metabasalt and metasiliceous tufts (Aghios Ilias Eruptive Formation; De Wever 1975). Chemically, the extrusive igneous rocks typically show a trace-element composition suggestive of a subduction influence, similar to those of the mrlange, discussed above, although again interpretations vary (e.g. Pe-Piper 1982, 1998; Dornsiepen & Manutsoglu 1996). The low-grade metamorphic sequence in the N W Peloponnese can be generally correlated with the mainly Triassic Tyros Beds exposed in the southern and central Peloponnese (Ktenas 1926; Lekkas & Papanikolaou 1980; Dittmar & Kowalezyk 1991). The Mesozoic carbonate sequence of the overlying Gavrovo carbonate platform in the N W Peloponnese begins with poorly dated shallow-water limestones, gypsum and brecciated dolomite ('gargneoule'), commonly folded and thrust imbricated. This is then overlain by a well-documented shallow-water carbonate platform sequence, several kilometres thick (Dercourt 1964; De Wever 1975). A Late Triassic-Early Eocene sequence (Tripolitza subzone) is well exposed east of Chelmos Mountain. This accumulated on a carbonate platform, with numerous breaks in deposition, erosion surfaces and bauxite development, especially during the Late Jurassic (De Wever 1975) and Cretaceous (Tsalia-Monopolis 1977). Hardgrounds and bauxites that developed in Mid-Eocene time in some areas indicate emergence at the highest levels of the platform sequence (Fleury 1980).

Foreland basin development A thick Upper Eocene-Lower Oligocene clastic sedimentary sequence overlying the GavrovoTripolitza zone documents the development of a foreland basin related to flexural warping and then collapse of the carbonate platform as the Pindos allochthon was overthrust westwards. Within the more westerly areas of the carbonate platform (Gavrovo subzone), neritic carbonates of Late Mesozoic-Early Cenozoic age pass through an interval of redeposited carbonate breccias and then into a thick

482

P.J. D E G N A N & A. H. F. R O B E R T S O N

o

r~

//

4~'

"/ ~"

~ 0

e~ !

9,---

9.,-.~

.

~

.~ e

e

"g~

9',...~

~

o

'..

oJ z'~

'~

.

.... ~ ::.!..'......~

!i ~ .:i.==================================== -:.-;::i:::::-.~.i " " ::"i..... !::.i!:..:::::.: ::.i-:..:-." ..

j

,',,

~ ~' *'J

~

e~ m

~

e~ ~,_~ ~_~m ~,

o

~ .~,.~

~

~,.__<

.j,~

.,

~g N~

PINDOS OCEAN NW PELOPONNESE, GREECE 50

,~o~

100

,

,/ .

.

.e..,

.

r

....

=.,

1

--

200

25C

--]

,-

.

.

.

.

-

9

.

N

~ - , _

. -

O, ' :

.

40

483

' ,

. 9

.

. ..

9 ,

.

.

,.

:i

,.

9

~

"

9 .i

..., . .

..

,

.kbZ37

.-----

--__--! .

--.._~

0

~9 : ; o . - ~ o ~

:..~...o..-~.~,:~ 1

0

:;~..;o:.<~:.~9;~::I ~.-~.: .':.~..-;:O.J

9

~ 1 0 _

.

7

5

.

9' . ~ ) , "

0

9 .

, ".0

'.'.~.,D, .

,/Lo;

:~:,o'.

--

-.,'

..

,

. - . .

o-% -

,

, . ~ . . "

,

--._.-, ~..'.-o.'"

,4

.

-.

, '

. ;

:

,

,..

i i | f w

Fig. 10. Sedimentary log of the Early Tertiary succession of the highest levels of the Gavrovo-Tripolitza zone beneath the Pindos thrust in the Alepohorio area. Key: mudstones (broken horizontal lines) pass upwards into sandstones (stipple) then into conglomerates. Clasts range from clast-supported (densely packed) to matrix sorted (with a sandy matrix). Depositional units are tabular or channelled (with scoured bases). The conglomerate matrix towards the top is calcareous in places.

siliciclastic sequence, known as the G a v r o v o Tripolitza Flysch, of Eocene-Oligocene age (Lekkas 1980; Bassias & Lekkas 1988). Near the Pindos thrust front (e.g. at Alepohorio; Fig. 2), a thick siliciclastic sequence above platform carbonates (Fig. 10) begins with mudstones, in beds up to 1.2 m thick, interbedded with sandstone, minor siltstone and rare pebblestone with well-rounded limestone clasts. Sandstone turbidites exhibit well-developed grading, flute casts, groove casts and microcross-lamination. The main lithology is litharenite, with abundant grains of quartz, carbonate and subordinate feldspar, chert and volcanic rocks. The siliciclastic sequence culminates in a 15 m thick interval of mainly medium-grained, thick-bedded sandstone, in beds up to 2.7 m thick. The turbiditic sequence passes upwards into channelized conglomerates and breccias (c. 230 m

thick). Individual beds are mainly massive (up to 14 m thick) and include large clasts. Three endmember conglomerate types are present. First, there are clast-supported conglomerates, in beds up to 6.2 m thick with well-rounded clasts (up to 90 cm x 50 cm in size), composed ofmicritic limestone and carbonate grainstone. Second, there are matrix-supported conglomerates, in beds up to 4 m thick, that contain large clasts (up to 50 cm in size), set in a fine-grained carbonate matrix. Third, there are matrix-supported conglomerates that are lithologically variable with clasts of sandstone, carbonate, shale and chert, set within a green pelitic matrix. A Late Eocene, or younger, age for these conglomerates is indicated by the presence of Actinocyclina sp. and Nummulites sp. within a clast of neritic limestone (C. Betzler, pers. comm.). This sequence (Fig. 10) terminates with > 200 m of strongly deformed siliciclastic turbidites beneath the Pindos thrust.

484

P.J. DEGNAN & A. H. F. ROBERTSON

Similar turbidites elsewhere in the N W Peloponnese were dated as Oligocene (Izart 1976). In the central Peloponnese, other conglomerates are dated as Late Oligocene (Mansy 1971). The massive conglomerates in the N W Peloponnese contain few palaeocurrent indicators. However, facies equivalents elsewhere in western Greece typically indicate north-south paleocurrents (Izart 1976), as seen north of the Gulf of Corinth (Lash 1988; Leigh 1991). The redeposited carbonate breccias can be explained by mass wasting related to break-up and subsidence of the platform that was probably fault controlled. Breccia clasts locally contain a Triassic or Jurassic fauna (Fleury 1980), suggesting that relatively deep levels of the platform were exposed and eroded. Elsewhere, within the more distal Tripolitza subzone (e.g. at Chelmos), a transition from neritic carbonate to sandstone turbidites is marked by alternations of carbonates and siltststones-fine sandstones ('couches de passage') of Eocene (upper Lutetian-lower Priabonian) age (Dercourt 1964). The uppermost neritic carbonates contain only benthic Foraminifera (e.g. Discocyclina sp.), whereas the overlying siltstones contain a mixture of benthic and planktonic Foraminifera, indicating deeper-water deposition. In some areas neritic carbonates of MidEocene age are directly overlain by medium- to coarse-grained turbiditic sandstones and mudstone, up to several hundred metres thick, e.g. around Drakovouni. Recently, Kamberis et al. (2006) have carried out a detailed micropalaeontological and sedimentological study of the turbiditic sediments of the both the Gavrovo-Tripolitza and Ionian basins in the N W Peloponnese using a combination of field, well and seismic data. The thickness of the Gavrovo foreland basin sediments is estimated as up to 3.7 km from subsurface information. Deposition of the Gavrovo turbiditic sediments was found to range from Late Eocene (locally), to Early Oligocene. As noted above, deposition began with relatively fine-grained deep-water sediments, followed by a much coarser succession of channelized conglomerates and sandstones, disorganized conglomerates and massive sandstones; an upper finergrained interval is also reported across the NW Peloponnese. Following a prolonged period of passive subsidence, the Gavrovo-Tripolitza carbonate platform emerged during Mid-Eocene-Late Eocene time, probably reflecting westward migration of a flexural forebulge ahead of the advancing Pindos allochthon. Erosion locally reached deep levels of the platform that were

locally exposed, possibly accompanied by faulting. Carbonate breccias were shed into deep water, especially near the former shelf-slope break. The appearance of deep-water siliciclastic turbidites in Late Eocene-Early Oligocene time indicates that flexural collapse of the platform had by then taken place to create a foredeep (Leigh 1991; Kamberis et al. 2006). The main sediment source was the Pindos allochthon in view of the abundance of chert and other lithologies typical of the deep-water Pindos sequences. The coarsening- and thickening-upward trend of the siliciclastic turbidites as a whole reflects the advance of the overthrust load. The coarse massive conglomerates rich in neritic carbonate clasts in the upper part of the sequence probably record advanced break-up and erosion of the carbonate platform prior to overthrusting. A similar flexural response to overthrusting is seen elsewhere, including many comparable Tethyan settings (e.g. Oman; Glennie et al. 1973; Robertson 1987).

Deformation o f the Gavrovo- Tripolitza foreland The high-level structure of the Tripolitza subzone in the east is well exposed in a large tectonic window on Chelmos Mountain where the Pindos basal thrust is broadly folded (Fig. 2). This large-scale fold developed adjacent to an extensional fault in the platform, possibly related to flexural collapse; this was later inverted prior to, or during, emplacement of the Pindos thrust sheets. The Pindos units were, therefore, finally transported over an already folded and internally thrust-imbricated foreland, which could have helped trigger the development of frontal ramps within the overriding Pindos allochthon. Following emplacement, large-scale folding of the Pindos and Tripolitza units probably reflects tightening of the Pindos suture zone. The Pindos basal thrust almost directly overlies the platform carbonates, as at Chelmos, or is first underlain by several tens of metres of intervening turbiditic sediments, as at Drakovouni, ~20 km to the south. The basal d~collement forms a thrust-flat in the south, and cuts up-section further north. Further west, deep-level thrust imbrication of the Gavrovo carbonate platform and its overlying foreland basin has been revealed by well and seismic reflection data (Sotiropoulous et al. 2003; Kamberis et al. 2006). Large listric thrusts are inferred to root towards the basal d6collement of the underlying Ionian zone sediments. Thrusting of the Pindos thrust stack over the foundered Gavrovo foreland basin took place following

PINDOS OCEAN NW PELOPONNESE, GREECE Early Oligocene turbidite deposition within this basin. The Gavrovo-Tripolitza zone was, in turn, thrust over the Ionian zone by the Pliocene following the termination ofclastic sedimentation in the Ionian basin (Underhill 1989; Clews 1989).

Discussion: regional evidence for the Pindos ocean The Pindos ocean rifted in the Early-MidTriassic and began to spread in Late Triassic time bordered to the west by the Gavrovo-Tripolitza carbonate platform. The continental-ocean transition zone comprised mainly subductioninfluenced basic-intermediate-acidic lavas and volcaniclastic sediments in our interpretation. Deep-water sedimentation was in progress from the Early-Mid-Triassic onwards, as documented by well-dated sediments from, for example, Vardoussia north of the Gulf of Corinth (Celet 1962; Bernoulli & Laubscher 1972; Ardaens 1978). In Albania (Krasta-Cukali zone), Permian clastic sediments (Verucanno facies) and Lower Triassic pelagic carbonates-volcanic rocks occur at the base of a Mesozoic-Lower Cenozoic succession, similar to the Pindos Group in the Peloponnese (see Robertson & Shallo 2000). In former Yugslavia, correlative units, notably the East Bosnian-Durmitor unit in northen Montenegro, Serbia-Kosovo, eastern Bosnia and Croatia, again show evidence of Triassic rifting, with a peak of volcanism in Ladinian time (Karamata & Vujnovi6 2000; Knezevid & Cvetovid 2000; Karamata 2006). In Croatia, Late Permian rifting, with block faulting and localized evaporite deposition was followed by mainly shallow-water carbonate deposition. The eruption of rift volcanic rocks (basalts, andesites and rhyolites) climaxed in Ladinian time, and had mainly ended prior to Norian time (Pami6 1984). Platform rocks are locally cut by rift-related syenite, diorite and gabbro (Pami6 et al. 2002). A subduction influence on volcanism is widely inferred (see Karamata 2006, and references therein), as for the Triassic rocks of Greece (Pe-Piper & Piper 2002). In Greece, large blocks of ammonite-rich pelagic limestone occur beneath the Pindos thrust front, at Glafkos in the N W Peloponnese and Meghdovas to the north of the Gulf of Cornith (Fleury 1980). This Ammonitico Rosso has been interpreted as accumulating on marginal highs (fault blocks) located along the western margin of the Pindos ocean, within the continent-ocean transition zone (Degnan & Robertson 1998; Pe-Piper & Piper 2002). In

485

former Yugoslavia, Upper Triassic shallowwater carbonates are overlain by Lower Jurassic condensed Ammonitico Rosso in a similar setting. Sedimentation in the western Pindos ocean became generally more distal eastwards, which, combined with palaeocurrent data, indicates sediment derivation from Apulia. Widspread Mn deposition within non-calcareous cherts of Late Jurassic-Early Cretaceous age (Leestena Formation) possibly relate to hydrothermal activity at a spreading centre (Pe-Piper & Piper 1989). Suprasubduction spreading is inferred by some workers to have taken place above a westward-dipping subduction zone in the Mid-Jurassic, followed by northeastwards emplacement of ophiolites (e.g. Pindos, Vourinos) onto the Pelagonian continent by Late Jurassic time (Robertson et al. 1991; see also Rassios & Moores 2006; Smith 2006, and references therein). However, other workers continue to believe that the Jurassic ophiolites were emplaced from the Vardar zone further east (e.g. Dercourt et al. 2000; Stampfli et al. 2001). The width of the Pindos ocean remaining after the ophiolite emplacement depends on the relative width of the Vardar ocean in the NE versus the Pindos ocean in the SW, both being located between the Eurasian and North African continents. The Pindos ocean remaining after the Mid-Late Jurassic ophiolite emplacement was relatively narrow in the north but widened southwards into southern Greece (Robertson et al. 1991). Support for this includes the apparent absence of a Pindos ocean west of the Scutari-Pe6 transverse lineament, the presence of ophiolitic detritus (e.g. chrome spinel) in Upper Cretaceous (Cenomanian-Turonian) turbidites associated with the Apulian margin (e.g. Klitoria Sandstone Member or Premier Flysch de Pinde), north of the Gulf of Corinth, and the apparent absence of Late Cretaceous-Early Cenozoic arc magmatism related to subduction of the Pindos ocean. Assuming the Pindos ocean was relatively narrow by the Early Cretaceous, most of the remaining separation between Eurasia and North Africa was accommodated by the Vardar ocean (see Sharp & Robertson 2006). The Pindos ocean was closing in Albania and northern Greece in the Late CretaceoousPaleocene, giving rise to the Pindos Flysch in these northerly areas. This sediment contains material derived from an exhumed accretionary wedge (i.e. blue amphibole) and the overriding Pelagonian continent (e.g. ophiolitic debris) (Faupl et al. 1996, 1998, 2002; Richter & Mtiller 2002). The timing of onset of eastward subduction in southern Greece (south of the Gulf of Corinth) remains poorly constrained but it is likely that there too subduction was active at least

486

P . J . D E G N A N & A. H. F. R O B E R T S O N

from Mid-Cretaceous time. We infer that the Pindos ocean was subducted during an eastward, scissors-like closure of an originally southwardwidening ocean, and thus the time of final collision and suturing decreased southwards (Robertson et al. 1991). In the Peloponnese there is little record of the eastern Pindos ocean, possibly because the subduction d6collement was located at a high level resulting in complete subduction, or an early accretionary wedge was overridden by the Pelagonian continent, a form of subduction erosion. As subduction continued the d6collement level lowered to near, or above, the igneous basement-sediment interface, allowing the Cretaceous pelagic carbonates of the Eastern Imbricates (Lambia Formation) to be accreted while oceanic basement was subducted. A thickening

Tripolitza carbonate platform

W --

------

,/

. . . . .

wedge of passive margin sediments, located c. 60 km from the rifted margin, then entered the trench and the d6collement level moved to a relatively lower stratigraphical level again, allowing the entire sedimentary succession and fragments of 'transitional' igneous crust and possible seamounts (e.g at Drakovouni) to be accreted within the m61ange. There is a surprising absence of thick clastic sediments (turbidites and debris flows) derived from the overriding accretionary prism in a subduction-trench setting, in contrast to the Pindos flysch in northern Greece. Such sediments were perhaps originally deposited in the subduction trench but were later detached and bulldozed ahead of the advancing accretionary prism and later reworked as Pindos-derived sediment into the very thick Gavrovo foredeep.

Fault blocks with condensed ~ facies

9

Cont,nental / - - . ~ - - - ~ crust / ~ ~ ~ _ - - - ,'..'-,...~" [=o~
-

_

~

E: Abyssal IF=,

Volcanic seamounts

.,~-

_

plain / -

-

[

~ - - - - - - - - ~ " ' - - 2,1=2~- ~ ~ - --. . . . . /_ ---.-',",~'-",*.~'~.--~..... ....... ~ . . . . . \/-,-;.Z.z_- : ~ " -" '.:.: .." . : - . ~ ,.,--.' . . . . . . , { ~ ~ ~ . ~ ~ . _ ~ . . : ,_:_ _ . . . . . . . . . . .

I

z

L

--

z ~" - ~

~ ~-T-.-::T--"

,

"

"

"

/,

-

,.~

--

" ~ . L " / S p r e ading " ,,"J-'l axis ;,~,.v ,: ~::/ -7

..;.:^ ~z.z ---p~,ow

" " ~ "~':':':':':':':':':'-'~'-'r'-":':':':-~':':':':':':':':':":':::::':::l':/

/

Transitional crust

,-A ,.~L.I

lavas

N

Oceanic crust

a. J u r a s s i c

20 - 4 0 k m L,=,,,-.,,,_I

Break-up of platform related to advancing orogenic wedge

Sand supply from suture zone in north

/.

-2

! '

.

.....

cretio~ary wedge

..

b. P a l e o c e n e - E o c e n e

Tectonic-sedimentary accreted melange

Fig. 11. Schematic reconstruction of the western margin of the Pindos ocean in the N W Peloponnese.

(a) Mesozoic rifted margin of the Pindos ocean; (b) Early Cenozoic eastward subduction. (See text for explanation.)

PINDOS OCEAN NW PELOPONNESE, GREECE

487

Conclusions

References

The Pindos zone in the N W Peloponnese is interpreted as an oceanic basin that rifted during the Triassic, reaching its maximum width in Jurassic-Cretaceous time (Fig. 11). This was followed by eastward subduction beneath the Pelagonian continent during Mid-Cretaceous?Palaeogene time. Triassic to Palaeogene sedimentary successions of the rifted passive margin become more distal towards the east away from the Apulian continent. The latest sediments to accumulate, turbidites of Paleocene-Late Early Eocene age (Pindos flysch), were mainly derived from north of the Gulf of Corinth. The geochemistry of Triassic basaltic rocks within the m61ange, like the Triassic volcanic rocks in many adjacent areas, including the foreland, is indicative of a subduction influence, related to either contemporaneous or pre-existing (Hercynian?) subduction. We suggest that this volcanism was related to variable melting of a heterogeneous subcontinental mantle located within a c. 60 km wide continent-ocean transition zone. Igneous crust and deep-sea sediments were detached from the continent-ocean transition zone, along with probable seamount fragments, and incorporated as blocks into a m61ange entrained at the base of, and within, individual thrust sheets. The deep-sea sedimentary cover of the Pindos ocean was accreted to form a imbricate stack of thrust sheets in post-Early to Mid-Eocene time. As the Pindos allochthon approached the Apulian continental margin, represented by the MesozoicEarly Cenozoic Gavrovo-Tripolitza carbonate platform, the footwall flexed upwards in the MidEocene, then collapsed beneath the advancing thrust load. Siliciclastic turbidites accumulated in a foredeep constructed on the downflexed Gavrovo carbonate platform during the Late Eocene-Early Oligocene, prior to overthrusting by the Pindos allochthon. The GavrovoTripolitza zone was itself imbricated and thrust westwards over the former Ionian rift basin by the Pliocene. Post-accretion compression and strike-slip related to post-accretion tightening of the Pindos thrust stack was followed finally by neotectonic extensional faulting.

ARDAENS, E. 1978. GOologie de la chdine du Vardoussia, comparaison avec le Massif du Koziakas (Grkce continentale). Thesis 36me cycle, University of Lille. AUBOUIN, J. 1959. Contribution ~ l'ktude gkologique de la Grkce septentrionak les confins de l'Epire et de la Thessalie. Annales G6ologiques des Pays Hell6niques, 9. AUBOUIN, J. BONNEAU,M., CELET, P., et al. 1970. Contribution h la g6ologie des Hell6nides: le Gavrovo, le Pinde et la Zone Ophiolitique Subpelagonian. Annales de la Societk Gkologique du Nord, 90, 277-306. BASSIAS, Y. & LEKKAS, S. 1988. La s6rie de transition entre les zones de Tripolitza et du Pinde. Annales de la Societk Gkologique du Nord, 107, 297-304. BECCALUVA,L., OHNENSTETTER,D. & OHNENSTETTER, M. 1979. Geochemical discrimination between ocean-floor and island are tholeiites-application to some ophiolites. Canadian Journal of Earth Sciences, 16, 1874-1882. BERNOULLI, D. & LAUBSCHER, H. P. 1972. The palinspastic problem of the Hellenides. Eclogae Geologicae Helvetiae, 65, 107-118. BLUMENTHAL, M. 1933. Zur Kenntnis der Querprofils des zentralen und nordlichen Peloponnes. Neues Jahrbuch fiir Geologie und Paliiontol6gie, Monatshefte, 70, 449-514. BR~TISH PETROLEUM COMPANY LTD. 1971. The Geological Results of Petroleum Exploration in Western Greece. Institute for Geology and Subsurface Research, Special Report, 10. CELET, P. 1962. Contribution h l'6tude g6ologique du Parnasse-Kiona et d'une partie des r6gion m6ridionale de la gr6ce continentale. Annales Gkologiques des Pays Hellbniques, 13. CLEWS, J. E. 1989. The Mesozoic and Cenozoic evolution of the Ionian zone, Western Greece. PhD thesis, University of Wales, Cardiff. DEGNAN, P. J. 1992. Tectono-sedimentary evolution of a passive margin: the Pindos Zone of the N W Peloponnese, Greece. PhD thesis, University of Edinburgh. DEGNAN, P. J. & ROSERTSON, A. H. F. 1994 Early Tertiary m61ange in the Peloponnese (S. Greece) formed by subduction-accretion processes. Bulletin of the Geological Society of Greece, 30(2), 93-107. DEGNAN, P. J. & ROBERTSON, A. H. F. 1998. Mesozoic--early Tertiary passive margin evolution of the Pindos ocean (NW Peloponnese, Greece). Sedimentary Geology, 117, 33-70. DERCOURX, J. 1964. Contribution ~ l'6tude g6ologique d'un secteur du Peleponn6se septentrional. Annales Gkologique des Pays Hellknique, 15. DERCOURT, J., MEILLIEZ, F., FLAMENT, J. M. & DE WEVER, P. 1978. Kertezi sheet.l:50 000 Geological Map of Greece. Institute of Geological and Mineral Exploration, Athens. DERCOURT, J., ZONENSHA~N,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.

This work was financially supported by an NERC PhD studentship to P. J. D., with additional support from the University of Edinburgh to A. H. F. R. for fieldwork in the Peloponnese. We are grateful to J. E. Dixon, S. P. Varnavas, P. D. Clift, G. Jones and T. Danelian for discussion. D. James assisted with X-ray fluorescence analysis and D. Baty helped draft some of the figures. The manuscript benefited from comments by G. Piper, D. W. J. Piper and D. Mountrakis.

488

P.J. D E G N A N & A. H. F. ROBERTSON

DERCOURT, J., RICOU, L. E. & VRIELYNCK, B. (eds) 1993. Atlas Tethys Palaeoenvironmental Maps, Gauthier-Villars, Paris. DERCOURT, J., GAETANI, M., VRIELYNCK, B., et al. (eds) 2000. Peri-Tethys Palaeogeographical Atlas (2000). DE WEVER, P. 1975. Etude gdologique des skries apparaissant en fenOtre sous l'allochtone pindique (skrie de Tripolitza et s&ie epimetamorphique de Zarouchla). Peloponnkse septentrional, Grkce. Th6se 36me cycle, Universit6 de Lille. DE WEVER, P. 1976a. La 'formation it blocs': olistostrome chevauche par la nappe du PindeOlonos (Peloponn+se, Grace). Comptes Rendus de l'Academie des Sciences, 282, 21-24. DE WEVER, P. 1976b. Mise en evidence d'importants affleurements de roches 6ruptives h la base de la nappe du Pindos-Olonos au s6rie de la 'formation /t blocs' (Peloponn6se, Grace). Annales de la Societ~ Gdologique du Nord, 97, 123-126. DICKrNSON, W. R. & SUCZEK, C. A. 1979. Plate tectonics and sandstone composition. American Association of Petroleum Geologists Bulletin, 93, 2164-2182. DITTMAR, U. 86 KOWALEZYK, G. 1991. Die Metaklastite im liegenden der Plattenkalk-Karbonate der stidlichen Peloponnese. Zeitschrift der Deutschen Geologischen Gesellschaft, 209-227. DIXON, J. E. 86 ROBERTSON, m. H. F. 1993. Arc signatures in Mediterranean Triassic rift basalts: a lithosphere-hosted inheritance from Hercynian subduction. Journal of Conference Abstracts, 10, 314. DORNSIEPEN, U. F. 86 MANUTSOGLU, E. 1996. Die Vulkanite der Tyros-Schichten Kretas und des Peloponnes-orogene Andesite oder anorogene Trapp-Basalte. Zeitschrift der Deutschen Geologischen Gesellschaft, 147, 101-123. DOUTSOS, T., KOUKOUVELAS, I. K. & XYPOLIAS, P. 2006. A new orogenic model for the External Hellenides. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 507-520. ELLIOT, D. 1976. The energy balance and deformation mechanisms of thrust sheets. Philosophical Transactions of the Royal Society of London, 283, 289-312. FAUPL, P., PAVLOPOULOS, A., WAGRE1CH, M. & MIGIROS, G. 1996. Pre-Tertiary blueschist terrains in the Hellenides: evidence from detrital minerals of flysch successions. Terra Nova, 8, 186-190. FAUeL, P., PAVLOPOULOS,A. & MIGIROS, G. 1998. On the provenance of flysch deposits in the External Hellenides of mainland Greece: results from heavy mineral studies. Geological Magazine, 135(3), 421-442. FAUPL, P., PETRAKKIS, K., MIGIROS, G. & PAVLOPOULOS, A. 2002. Detrital blue amphiboles from the western Othris Mountain and their relationship to the blueschist terranes of the Hellenides (Greece). International Journal of Earth Sciences (Geologische Rundschau), 91,433-444. FITTON, J. G. & DUNLOP, H. M. 1985. The Cameroon line, West Africa, and its bearing on the origin of

oceanic and continental alkali basalts. Earth and Planetary Science Letters, 72, 23-38. FLAMENT, J. M. 1973. De l'Olonos au Chelmos. Etude gkologique d'un secteur de la nappe du Pinde-Olonos. Th&e 36me cycle, Universit6 de Lille. FLEURY, J. J. 1980. Evolution d'une platforme et d'un bassin darts leur cadre alpin: les zones de GavrovoTripolitza et du Pinde-Olonos. Annales de la Societk G~ologique du Nord, 4, 1-651. GAETANI, M. & GARZANTI, E. 1991. Multicycle history of the Northern India continental margin (Northwestern Himalayas). American Association of Petroleum Geologists Bulletin, 75, 1427 1446. GLENNIE, K. W., BOEF, M. G. A., HUGHESCLARKE,M. W, MOODY-STUART, M., PILAAR, W. F. H. & REINHARDT, B. M. 1973. Late Cretaceous nappes in the Oman Mountains and their geologic significance. American Association of Petroleum Geologists Bulletin, 57, 5-27. GREEN, T. J. 1982. Structural and sedimentological studies of the Pindos zone, central Greece. PhD thesis, University of Cambridge. IZART, A. 1976. Etude gOologique d'un secteur du Pdloponnese nord-occidental (Grkce): la carte de Goumeron. Th6se 36me cycle, Universit6 de Lille. JACOBSHAGEN,V., DORR, S., KOCKEL, F., KoPP, K. O. & KOWALCZYK, G. 1978. Structure and geodynamic evolution of the Aegean region. In: CLOSS, H., ROEDER, D. & SCHMIDT, K. (eds) Alps, Apennines, Hellenides. Schweitzerbart, Stuttgart, 537-564. KAMBERIS, E., PAVLOPOULOS,A., TSAILA-MONOPOLIS, S., SOTIROPOULOS, S. 86 IOAKIM, C. 2006. Flysch facies and paleogeographic evolution of foreland basins in NW Peloponnese (Greece). Geologica Carpathica (in press). KARAMATA, S. 2006. The geological development of the Balkan Peninsula related to the approach, collision and compression of Gondwanan and Eurasian units. In: ROBERTSON, A. H. F. 86 MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 155-178. KARAMATA, S. 86 VUJNOVI~, L. 2000. Correlation of Palaeozoic units of the Dinarides and the northern part of the Vardar Zone. In: KARMATA, S. 86 JANKOVIC S. (eds) Proceedings of the International Symposium of Geology and Metallogeny of the Dinarides and the Vardar Zone. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 150-160. KNEZEVI~, V. 86 CVETOVI(~, V. 2000. Triassic rifting magmatism of the Dinarides. In: KARMATA, S. 86 JANKOVlC S. (eds) Proceedings of the International Symposium of Geology and Metallogeny of the Dinarides and the Vardar Zone. Academy of Science and Arts of the Republic of Srpska, Collections and Monographs, 1, 150-160. KTENAS, C. 1926. Sur le d6veloppement du primaire au Peloponn6se central. Praktika Academia, Athens, 1, 53-59. LASH, G. G. 1988. Along-strike variations in foreland basin evolution: possible evidence for continental collision along an irregular margin. Basin Research, 1, 71-83.

PINDOS OCEAN NW PELOPONNESE, GREECE LEIGH, S. P. 1999. The sedimentary evolution of the Pindos foreland basin, Western Greece. Unpublished PhD Thesis, University of Wales, Cardiff. LEKKAS, S. 1980. Les phyllades du Peloponn6se: un metaflysch Ionien chevauche par le s6rie de Gavrovo-Tripolitza. Comptes Rendus de l'AcadOmie des Sciences, 291, 21-24. LEKKAS, S. & PAPANIKOLAOU,O. 1980. On the phyllite problem in the Peloponnese. Annales de la SocietO GOologique de Pays Hellknique, 29, 395410. MANTSCHAL, G., MONTENER, O., DESMURS, L. & BERNOULLI, D. 2003. An ancient ocean-continent transition in the Alps: the Totalp, Err-Plata, and Malenco units in the Eastern Central Alps (Graubunden and northern Italy). Eclogae Geologicae Helvetiae, 96, 131-146. MANSV, J. L. 1971. Etude g6ologique des monts de Kyparisssia (Messenie, Gr6ce). Annales de la Societd GOologique du Nord, 91, 57-63. MCBRIDE, E. F. 1963. A classification of common sandstones. Journal of Sedimentary Petrology, 49, 837-868. MECHEDE, M. 1986. A method of discriminating between different types of mid-ocean ridge basalts and continental tholeiites with the N b - Z r - Y diagram. Chemical Geology, 56, 207-218. NEUMANN, P. 2003. Ablagerungsprozesse, Event und Biostratigraphie kreidezeitlicher Tiefwassersedimente der Tethys in der Olonos-Pindos-Zone Westgriecenlands. Miinchner Geowissenschaftliche Abhandlungen, Reine A, Geologie und Palaeontologie, 40. NEUMANN, P. & ZACHER, W. 1998. New results on radiolarian biostratigraphy and sedimentology of the early Cretaceous to Turonian of the Pindos Zone in the Central Pindos Mountains (Mainland Greece). Bulletin of the Geological Society of Greece, 31(2), 59-65. NEUMANN, P. & ZACHER, W. 2004. The Cretaceous sedimentary history of the Pindos Basin (Greece). International Journal of Earth Sciences ( Geologische Rundschau) , 93, 119-131. NEUMANN, P., RISCH, H., ZACHER,W. & FYTROLAKIS, N. 1996. Die stratigraphische und sedimentologische Entwicklung der Olonos-Pindos-Serie zwischen Koroni und Finkounda (SW-Messenien Griechenland). Neues Jahrbuch fiir Geologic und Paliiontologue, abhandlungen, 200, 405424. PAMI~, J. 1984. Triassic magmatism of the Dinarides in Yugoslavia. Tectonophysics, 109, 273-307. PAMIt~, J., TOMLJENOVI~ B. & BALEN, D. 2002. Ophiolites of the central and northwestern Dinarides. Lithos, 65, 113-142. PEARCE, J. A. 1980. Geochemical evidence for the genesis and eruptive setting of lavas from Tethyan ophiolites. In: PANAYIOTOU, A. (ed.) Proceedings of the International Symposium 'Troodos 1979'. Geological Survey Department, Nicosia, 261-272. PEARCE, J. A. & CAYN, J. R. 1973. Tectonic setting of basic volcanic rocks determined using trace element analysis. Earth and Planetary Science Letters, 19, 290-300. PE-PIPER, G. 1982. Geochemistry, tectonic setting and metamorphism of the mid-Triassic volcanic rocks of Greece. Tectonophysics, 85, 153-272.

489

PE-PIPER, G. 1983. The Triassic volcanic rocks of Tyros, Zarouchla, Kalamae, and Epidavros, Peloponnese, Greece. Schweitzerische. Mineralogische und Petrographische. Mitteilungen, 63, 249-266. PE-PIPER, G. 1998. The nature of Triassic extensionrelated magmatism in Greece: evidence for Nd and Pb isotope geochemistry. Geological Magazine, 13, 331-348. PE-PIPER, G. & PIPER, D. W. J. 1989. The geological significance of manganese distribution in JurassicCretaceous rocks of the Pindos basin, Peloponnese, Greece. Sedimentary Geology, 65, 127-137. PE-PIPER, G. & PIPER, D. W. J. 1991. Early oceanic subduction-related volcanic rocks, Pindos basin, Greece. Tectonophysics, 192, 273-292. PE-PIPER, G. & PIPER, D. W. J. 2002. The Igneous Rocks of Greece. The Anatomy of an Orogen. Beitrage zur Regionalen Geologic der Erde, 30. PIPER, D. W. J. 1984. Tectonic setting of the Mesozoic Pindos basin of the Peloponnese, Greece. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 563-568. PIPER, D. W. J. 2006. Sedimentology and tectonic setting of the Pindos Fysch of the Peloponnese, Greece. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 493-505. RASSIOS, A. H. E. & MOORES E. M. 2006. Heterogeneous mantle complex, crustal processes, and obduction kinematics in a unified Pindos-Vourinos ophiolitic slab (northern Greece). In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 237-266. RENZ, C. 1955. Die vorneogene Stratigraphie der normassedimentaren Formationen, Griechenlands. Institute of Geology and Subsurface Research, Athens. RICHTER, D. 1993. Die Flyschzonen Griechenlands VII. Sedimentstrukturen, Ablagerungsart und Stromungrichtungen im Flysch der Pindos-Zone (Griechenland). Neues Jahrbuch fiir Geologic und Palaeontologie, Monatshefte, 1993, 513-544. RICHTER, D. & LENSCH, G. 1977. Die Ophiolitischen Vulkanit-Vorkommen der Olonos-Pindos Decke im Zentralpeloponnese (Griechenland). Neues Jarbuch fiir Mineralogie, Abhandlungen, 129, 312-332. RICHTER, D. & LENSCH, G. 1989. Neue VulkanitVorkommen der Pindos-Decke auf dem Peloponnese (Griechenland). Zeitschrift fiir Geologische Wissenschyten, 17, 947-961. RICHTER, D. & Mt0LLER, C. 1992. Die Flysch-Zonen Griechenlands, VIII. Neue Vorkommen von BSotischem Flysch im nSrdlichen Pindos-Gebirge (Griechenland). Zeitschrift der Deutschen Geologischen Gessellsehaft, 143, 87-94.

490

P.J. D E G N A N & A. H. F. ROBERTSON

ROBERTSON, A. H. F. 1987. The transition from a passive margin to an Upper Cretaceous foreland basin related to ophiolite emplacement in the Oman Mountains. Geological Society of America Bulletin, 99, 633-653. ROBERTSON, A. H. F. 1994. Role of the tectonic facies concept in orogenic analysis and its application to Tethys in the Eastern Mediterranean region. Earth-Science Reviews, 37, 139-213. ROBERTSON, A. & DEGNAN, P. 1998. Significance of modern and ancient oceanic Mn-rich hydrothermal sediments, exemplified by Jurassic Mn-cherts from Southern Greece. In: MILLS, R. A. & HARRISON, K. (eds) Modern Ocean Floor Processes and the Geological Record. Geological Society, London, Special Publications, 148, 217-240. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the Eastern Mediterranean. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 17, 1-74. ROBERTSON, A. H. F. & SHALLO, M. 2000. Mesozoic tectonic development of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-214. ROBERTSON, A. H. F., CLIFT, P. D., DEGNAN, P. J. & JONES, G. 1991. Palaeographic 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 Palaeozoic-Early Tertiary development of Tethys in the Eastern Mediterranean Region. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. ~ENGOR, A. M. C., YILMAZ, Y. & SUNGURLU, O. 1984. Tectonics of the Mediterranean Cimmerides: nature and evolution of the western termination of Palaeo-Tethys. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 77-112. SHARP, I. A. & ROBERTSON, A. H. F. 2006. Tectonicsedimentary evolution of the western margin of the Mesozoic Vardar Ocean: evidence from the Pelagonian and Almopias zones, northern Greece. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 373-412. SKOURLIS, K. & DOUTSOS, T. 2003. The Pindos foldand thrust belt (Greece): Inversion kinematics of a passive continental margin: International Journal of Earth Sciences, 92, 891-903. SMITH, A. G. 1993. Tectonic significance of the Hellenic-Dinaric ophiolites. 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, 213-244. SMITH, A. G. 2006. Tethyan ophiolite emplacement, Africa to Eruope motions, and Atlantic spreading. In: ROBERTSON, A. H. F. & MOUNTRAKIS,D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 11-34. SMITH, A. G., HYNES, A.J., MENZIES, A. J., NISBET, E. G., PRICE, I., WELLAND, M. J. P. & FERRIERE, J. 1975. The stratigraphy of the Othris Mountains, eastern central Greece. Eclogae Geologicae Helvetiae, 68, 463-481. SOTIROPOULOS, S. P., KAMBERIS, E., TRIANTAFYLLOY, M. V. & DOUTSOS, T. 2003. Thrust sequences in the central part of the External Hellenides. Geological Magazine, 140, 661-668. STAMPFLI, G. M. & BOREL, G. D. 2002. A plate tectonic model for the Palaeozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrones. Earth and Planetary Science Letters, 169, 17-33. STAMPFLI, G., MOSAR, J., DE BONO, A. & VAVASSIS,I. 1998. Late Palaeozoic, early Mesozoic plate tectonics of the western Tethys. Bulletin of the Geological Society of Greece, 32, 113-120. STAMPFLI, G . , MOSAR, J., FAURE, P., PILLEVUIT,A. & VANNAY, J.-C. 2001. Permo-Mesozoic evolution of the western Tethys realm: the Neotethys East Mediterranean basin connection. In: ZIE6LER, P., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUINSOLEAU, S. (eds) Peri-Tethys Memoir 5, Peri-Tethys Rift/Wrench Basins and Passive Margins. M6moires du Mus6um National d'Histoire Naturelle, 51-108. TAYLOR, B., HUCHON, P. & KLAUS, A., et al. (eds) 1999. Proceedings of the Ocean Drilling Program, Initial Reports, 180, National Science Foundation, Joint Oceanographic Institutions Inc. Texas A. M University, College Station, Texas. THiI~BAULT, F. 1982. Evolution GOodynamique des HellOnides externes en Peloponnkse meridionale (Grkce). Societ6 G6ologique du Nord, Special Publications, 6. THII~BAULT, F. & FLEURY, J.-J. 1980. The clay sedimentation in the Olonos-Pindos basin during the Mesozoic: an approach for the source identification and processes of transport. In: 5th Geological Congress of Greece, ThessalonikL May 1990, Abstracts, 62. TSALIA-MONOPOHS, S. 1977. Study of the Tripolitza Zone: micropalaeontology and stratigraphy. Institute for Geology and Subsurface Research, Special Report, 20, 96-99 (in Greek with English summary). UNDERHILL, J. 1989. Late Cenozoic deformation of the Hellenide foreland, western Greece. Geological Society of America Bulletin, 101,613-634. WAGREICH, M., PAVLOPOULOS, A., FAUPL, P. & MIGIROS, G. 1996. Age and significance of Upper Cretaceous siliciclastic turbidites in the central Pindos Mountains, Greece. Geological Magazine, 133, 325-331.

PINDOS OCEAN NW PELOPONNESE, GREECE WILLIAMS, G. D. & FISCHER, M. W. 1984. A balanced section across the Pyrenean orogenic belt. Tectonics, 3, 773-780. WHITMARSH, R. l . , MANATSCHAL, G. 8r MINSHULL, T. A. 2001. Evolution of magma-poor continental margins from rifting to seafloor spreading. Nature, 413, 150-154. XYPOLIAS, P. 8~ DOUTSOS, T. 2000. Kinematics of rock flow in a crustal-scale shear zone: implications

491

for the orogenic evolution of the southwestern Hellenides. Geological Magazine, 187, 81-96 YILMAZ, P. O., NORTON, I. O., LEARLY, D. & CHUCHLA, R. A. 1996. Tectonic evolution and palaeogeography of Europe. In: ZIEGLER, P. A. & HORVARTH, F. (eds) Peri-Tethys Memoir 2: Structure and Prospects of Alpine Basins and Forelands. M6moires du Mus6um National d'Histoire Naturelle, 48-60.

Sedimentology and tectonic setting of the Pindos Flysch of the Peloponnese, Greece D A V I D J. W. P I P E R

Geological Survey of Canada (Atlantic), Bedford Institute of Oceanography, P.O. Box 1006, Dartmouth, N.S. B2 Y 4A2, Canada (e-mail." [email protected]) Abstract: The Palaeogene Pindos Flysch of the Peloponnese shows important differences from the flysch of northern Greece. Stratigraphic sections and palaeocurrent indicators were measured in the Pindos Flysch Formation and the underlying Kataraktis Passage Member throughout the Peloponnese. The Kataraktis Passage Member records carbonatedominated sedimentation from the Apulian continental margin to the west, with intercalated terrigenous sediment also derived from the west. Variations in thickness and turbidite facies show that the overlying Pindos Flysch Formation was deposited in channels with levees and in channel-termination lobes in the western Peloponnese and in a distal basin plain, locally ponded, in the east. At least in the central Peloponnese, facies variation, palaeocurrents and detrital petrology show that the Pindos Flysch was derived from the Apulian margin. The Pindos Flysch of northern Greece, of late Paleocene to Oligocene age, was deposited in a foreland basin and derived from the rising Pelagonian nappes to the east. A younger microcontinental collision south of the Gulf of Corinth line resulted in the Pindos Flysch of the Peloponnese being incorporated in the accretionary prism by Mid-Eocene time.

The Hellenide orogen of western Greece resulted from progressive convergence of the Apulian and Pelagonian microcontinents through the Cretaceous and Cenozoic (Robertson et al. 1991; Pe-Piper & Piper 2002). A series of isopic zones, or terranes, are recognized in western Greece (Fig. 1; Aubouin 1959). The Pelagonian terrane is a Hercynian continental fragment of Gondwanan affinity, with a Mesozoic platform sedimentary succession onto which ophiolites were obducted from the Pindos ocean to the west in the Late Jurassic. Compressive deformation and metamorphism of the Pelagonian terrane began in the mid-Cretaceous (Schermer et al. 1990). The Pindos zone (or terrane) marks a Mesozoic ocean basin that rifted in the Triassic and was largely consumed by later Mesozoic ocean-ocean subduction (Pe-Piper & Piper 1991). Preserved sedimentary rocks of the Pindos zone consist principally of calc-turbidites and pelagic sediments that accumulated on the passive margin of the Apulian microcontinent. The GavrovoTripolitsa zone consists of Mesozoic platform sedimentary rocks that accumulated on a marginal high at the eastern edge of the Apulian microcontinent, whereas the more westerly Ionian zone experienced Triassic rifting and includes deeper-water late Mesozoic sedimentary rocks. Farther west, the Paxos or pre-Apulian zone represents another Mesozoic platform sedimentary succession. Continental subduction of Apulia beneath Pelagonia probably began in the mid -Cretaceous

and, by the Paleocene, the westward-moving Pelagonian nappe pile shed terrigenous sediment onto the old Apulian margin of the remnant Pindos ocean, where flysch sediments accumulated. Internal thrusting of Apulian continental crust throughout the Palaeogene resulted in westward thrusting of the sedimentary pile that had accumulated on the Apulian passive margin of the Pindos zone, to form the Pindos nappes. These in turn created a foreland basin in which the Tripolitsa flysch accumulated (GonzalezBonorino 1996). Thrusting propagated westward in the Apulian continent, resulting in the Pindos nappes overriding the outer high of the GavrovoTripolitsa zone and creating a foreland basin coincident with the former rift of the Ionian zone, within which Oligocene to Miocene Ionian flysch accumulated. Thrust deformation in the Quaternary was concentrated farther west along the pre-Apulian zone in the Ionian islands (Underhill 1989). The Palaeogene Pindos Flysch has been classically interpreted as a foreland basin succession deposited after final closure of the Mesozoic Pindos ocean (Gonzales-Bonorino 1996). It is best known from northern mainland Greece, where it is hundreds of metres thick and apparently derived from the advancing Pelagonian nappes. In contrast, the Pindos Flysch is thin in the Peloponnese and may have a rather different origin (Piper & Pe-Piper 1980). This study presents sedimentological observations and interpretations of the Pindos Flysch Formation and

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 493-505. 0305-8719106l$15.00 9 The Geological Society of London 2006.

494

D.J.W. PIPER rifted in the Triassic, underwent sea-floor spreading at least in the Late Triassic and Early Jurassic, experienced ophiolite obduction in the Late Jurassic, but continued to accumulate deep-water sediments throughout the Cretaceous (Robertson et al. 1991). The basin was bounded to the SW by the Apulian promontory of Gondwana and to the NE by the Pelagonian microcontinent.

Stratigraphy

Fig. 1. General geological map of western Greece, showing Pindos zone sedimentary rocks and ophiolites derived from the Pindos Ocean. Pa, Parnassos zone; Serbo-Mac., Serbo-Macedonian massif. Map modified from Jacobshagen (1986).

the underlying Kataraktis Passage Member of the Peloponnese (Fig. 2). The Kataraktis Passage Member is the highest unit of the Lambia Formation (Degnan & Robertson 1998), a carbonatedominated passive margin succession of Cretaceous age on the Apulian margin of the Pindos ocean (Piper & Pe-Piper 1980; Neumann et al. 1996; Neumann & Zacher 2004), now preserved in a series of imbricated thrust slices (Degnan & Robertson 1998). The purpose of the study is to use the stratigraphy and sedimentology of the Pindos Flysch of the Peloponnese to constrain tectonic interpretations of closure of the Pindos ocean. The Pindos Flysch Formation is the youngest deposit of the Pindos ocean basin. This basin

In this study, we use the stratigraphic nomenclature for the northwestern Peloponnese of Degnan & Robertson (1998); alternative nomenclature for the southern Peloponnese has been provided by Neumann & Zacher (2004). The base of the Kataraktis Passage Member (Couches de Passage of Aubouin 1959; Rigani Member of Neumann & Zacher 2004) was defined by Degnan & Robertson (1998) as the first occurrence of black chert or terrigenous clastic sediment within the upper part of the Lambia Formation (Platy Limestone Formation of Neumann & Zacher 2004). It overlies the Erymanthos Limestone Member (Calcaires en Plaquettes of Aubouin 1959). Its top was defined as where terrigenous clastic sediment first consistently dominates carbonate and it passes into the Pindos Flysch Formation (Flysch du Pinde of Aubouin 1959). The Kataraktis Passage Member straddles the Cretaceous-Palaeogene boundary (Fleury 1970) and its base is probably diachronous, in places as old as Mid- or Late Maastrichtian (Neumann 2003) or even Campanian (Richter & Mfiller 1993) (Fig. 3). Nannofossil data of Richter & Mfiller (1993) show that the base of the Pindos Flysch Formation was of latest Paleocene age in the N W Peloponnese (between NP1 and NP9 at our locality 61; between NP8 and NP9 at 111; above NP9 near 247) and in this area, microfossils within the flysch date from earliest Eocene time (NP10). Foraminiferal biostratigraphy by Dercourt (1964) in general confirms the nannofossil evidence, with the Pindos Flysch yielding principally Late Paleocene ages, with possible Early Eocene ages at Kaledsi (south of 62). A tectonized calcirudite bed at Divri (north of 183), possibly within the Pindos Flysch but more probably part of the Tripolitsa zone, yielded faunas of late Lutetian to Priabonian age (Dercourt 1964). In the extreme eastern Peloponnese, Richter & Mfiller (1993) showed that the base of the Pindos Flysch Formation lies between NP6 and NP9 (71) and in or above NP9 (near 141), with microfossils from within the flysch of zone NP9. In the extreme SW Peloponnese (near 207) the Kataraktis Passage Member extends to the

PINDOS FLYSCH, PELOPONNESE, NW GREECE

495

Age (Ma) uJ z uJ

2O ~: 9~ 4o

PeloponneSeNw sw w ~._0

~

60

___

:::::::::::::::::::: _--_--- -

- - ,

- -

- - ,

Kataraktis Passage Mbr

Erymanthos

,o

Limestone Mbr . . . . . . .

Klitoria Sandstone Mbr i'l'l'r'li

100

~

Paos

Limestone

Mbr

Fig. 2. Index map of the Peloponnese showing location of measured sections. Grey tone indicates distribution of Pindos zone rocks. Bold numbers refer to sections shown in Figures 4, 5 and 7.

Early Eocene (zones from NP9 to NP13) and the Pindos Flysch locally reaches the late MidEocene in age (Bfichl et al. 2004). In contrast, in northern Greece, the Pindos Flysch extends from Late Paleocene to Mid- to Late Oligocene time (zone NP24-25) (Richter et al. 1993).

Methods Lithological sections were measured in the Kataraktis Passage Member and the Pindos Flysch Formation throughout the Peloponnese. Where tectonic deformation was too severe, observations on lithological types were made. Palaeocurrent directions were measured based on sole marks and ripple cross-lamination in turbidites and were corrected in the field for one phase of folding. Most measurements were made on rocks dipping < 30 ~ Quoted stratigraphic thicknesses are based on measurement of continuous sections, with careful correlation across faults, and tend to be a little thinner than the more general thicknesses estimated by Richter (1993) and Richter & Mfiller (1993).

Fig. 3. General stratigraphic nomenclature for the Late Cretaceous and Palaeogene of the Pindos basin (after Degnan & Robertson 1998). Biostratigraphy from Richter & Miiller (1993) and Neumann (2003).

Lithologies The Kataraktis Passage Member consists of alternating terrigenous and calcareous rocks (Figs 4 and 5). In the lower part of the Kataraktis Passage Member in the N W margin of the basin (e.g. section 51, Fig. 5), graded calcarenitecalcilutite-micrite beds pass up into green shales and interbed with spotty pelagic limestone and grey marls. Calcirudites are also exposed. The graded calc-turbidites have bioturbation restricted to the top of the micrite, similar to calc-turbidites in the underlying Erymanthos Limestone Member. A few terrigenous graded sandstone beds occur, showing typical turbidite sedimentary structures, including sole marks and Bouma sequences of sedimentary structures. In the north-central Peloponnese (e.g. section 93, Fig. 4), the succession is rather similar, with interbedded calc-turbidites, minor terrigenous turbidites, and some pelagic limestones. However, the calc-turbidites have a higher proportion of micrite to calcarenite. Thick beds of grey

496

D . J . W . PIPER

,.6 "O r

o a,2

,.8

r~

O q'3

o'3

O

"O

_g

PINDOS FLYSCH, PELOPONNESE, NW GREECE marlstone are common, in places with a coarsergrained base and a shaly top. Greenish shales in places form the top of graded terrigenous and calc-turbidite beds and appear to be of turbidite origin. In the central Peloponnese (e.g. section 105, Fig. 4) and the eastern Peloponnese (e.g. section 71, Fig. 5), similar lithologies are present, but the proportion of calcarenite is much lower. In the eastern Peloponnese, the catc-turbidite beds tend to be thinner, passing up into green shales, and the proportion of thick marlstone and mudstone beds is higher. Grey marlstones have been found in beds up to 4.5 m thick: some of these beds grade up from calcarenite, and only their extreme tops are bioturbated. In addition, red shale beds are present, which in some sections form thick sharp-based beds of probable turbidite origin. Thus there is a gradual progression eastward from successions with more proximal sedimentological character to more distal facies. In the southwestern Peloponnese, the upper part of the Kataraktis Passage Member (e.g. section 197, Fig. 5) consists of bioturbated sandy turbidites (some of which are red) that interbed with red siltstone and shale. Grey-green shale is less abundant. Fine-grained conglomerate occurs locally. Calcareous lithologies appear minor, although some massive calcarenite beds, up to 3 m thick, suggest a relatively proximal environment. The degree of bioturbation of turbidite sandstones is unusual. In the Koroni Peninsula in the SW Peloponnese, red shales are absent from sections with calcirudites, thick calcarenites and grey marls. The Pindos Flysch Formation everywhere consists of terrigenous turbidite sandstones of variable thickness that interbed with grey and greenish shales. Lithologies have been previously described in detail by Richter & Mfiller (1993) and Degnan & Robertson (1998).

Palaeocurrents and facies distribution Relatively few palaeocurrent measurements are available from the Kataraktis Passage Member and consist principally of ripple drift directions, which are known to be more variable than sole markings. At single localities, interbedded calcareous and terrigenous turbidites generally show similar palaeocurrent indicators (e.g. locality 51, Fig. 5). Palaeocurrents in the northern, central and eastern Peloponnese are generally to the SE (Fig. 6). The sparse palaeocurrents in the southwestern Peloponnese are more variable, but range from north to southeasterly. Facies distribution in the Kataraktis Passage Member confirms the general eastward dispersal

497

system inferred from palaeocurrents. Calcirudites are restricted to localities near the western margin of the basin (Fig. 6). The thickness of the thickest calcarenite beds tends to decrease from west to east, although beds up to 1.5 m thick found in the east are of fine to very fine sand size. In the western part of the basin, the distribution of red siltstones and shales is distinctive. There appears to have been a major source of terrigenous sediment near Kiparissia, locally represented by deposits of red sandstones and elsewhere by a high proportion of red shales. Some red shales are present near the top of the Kataraktis Passage Member in the Lambia area, but are absent in areas to the NE. They are also absent in proximal sites in the Koroni peninsula. The multi-metre thick turbidite marlstones in the eastern Peloponnese are similar to thick fine-grained turbidites occurring on modern abyssal plains (see Piper & Stow 1991). The rarity and thinness of calcarenites in this area confirms that this is a very distal environment. High-quality palaeocurrent measurements in the Pindos Flysch Formation are found principally in the N W Peloponnese (Figs 7 and 8), where flows were to the SE. Consistent southward directions of sole marks were reported by Richter (1993) from the west-central Peloponnese and eastern Peloponnese. In the SW Peloponnese, palaeocurrent indicators are sparse but consistently indicate flow to the NE. Three zones can be recognized from west to east on the basis of stratigraphic thickness and turbidite facies (Fig. 8). In the Western zone, the total thickness of flysch is generally between 10 and 15 m, with typically 70-80% sandstone and the maximum bed thickness is 0.5-1.5 m. Most beds have Bouma T, divisions at their bases, and the coarsest sediment is generally medium sandstone, although locally channels with conglomerate and coarse sandstone are recognized. Two rather thicker sections (98 and 109; see Fig. 2) are 20-30 m thick, but have a much higher shale percentage (60%) and sandstones typically have Bouma Tb divisions at their base. The Central zone lies further east, and has flysch sequences more than 20 m thick, with a maximum measured thickness of 47 m. Fine conglomerates are locally present. In several sections, the basal few metres consist of shaly flysch, but otherwise shale percentage does not exceed 10% and the thickest sandstone beds are more than 2 m thick. Although many beds have T~ division at their base, some have basal Tb divisions. The Eastern zone occurs in the eastern Peloponnese. The measured thickness of the Pindos Flysch is less than 10 m. Sandstone makes up less than 30% of the section, is fine grained, in

498

D . J . W . PIPER

t-.-

t~

,.Q

,.Q

o

,....-t

O ~

~ ,,...w

o ,...4

r~

(D

r~

PINDOS FLYSCH, PELOPONNESE, NW GREECE

499

Fig. 6. Map showing facies variations in the Kataraktis Passage Member. Thickness of calcarenite: A, thickest beds > 1 m; B, thickest beds 0.5-1 m; C, thickest beds <0.5 m. 'Mean palaeocurrents' based on at least five consistent measurements; means of < 5 measurements are shown as 'isolated palaeocurrent measurements'. It should be noted that the map shows the present geography after substantial mid-Cenozoic compression.

beds less than 40 cm thick, and most beds have Tb divisions at their bases.

Palaeogeographical interpretation Although previous workers (Richter 1993; Gonzales-Bonorino 1996) have compared the Pindos Flysch of the Peloponnese with that of northern Greece, this concept needs reexamination. The similarity in facies distribution

and palaeocurrents in the Kataraktis Passage Member and Pindos Flysch Formation indicates that, unlike in northern Greece, the two stratigraphic units reflect the same palaeogeographical setting. In the southern Peloponnese, most sediment is derived from the western margin of the basin. Terrigenous sediment was dispersed to the N E from probably two or more sources near Kiparissia. In the Kataraktis Passage Member, distinctive red terrigenous sediment was supplied. In the Pindos Flysch, a fine

500

D . J . W . PIPER

cq

id~?.:
ii

"'"'"~ I ~ i ~ / ~ ! ~

<'~,'~'~'l'k:::, . . . . . . . . .

I,,Is~,~i:~,,II, , '"" .........

e-

~.

~

',', ',,',

_

~> < ~

,,, ~,,, ,, ,,, ~ !~i~1,,,,,,,,;;

o~ ,,,,~

~

zg_ o ~E

,,,~,,,I,,,j,,,,,, . . . . . . . . t2--~!:1, . . . I II" ' I ~................ '"'" ' " i "I~,iil ' ........ l~/tiI'"'"'" ~1~ ~,~,~,~,~,~,~,~,1,~,~,1,1,1,1,1,1,1,1,1, i' " 1.... ' " " "il,," ' I,,I ' ~~176176 r I~1~1 ' '" 1' : '""'~

o~

|

,l

I

tl

IIIIIIIIlo:o:o:o:o:.:o:ollllllllllllllllllllllllllll

Ill I

I,,

....

t--

;

. . . . . . . . .

-

. . . . . . . . .

I I i~176

.......

.......... o: . . . . . . . . .

,~ . . . . . . . . .

,'.'.',

........ ,~,~,~,~,~,~,~:~ ~ . . . . . . . . . .r

N

o

I

I

I I I II

5:~176176 I o:o:o:.:,:o:,

I i I I ~

I~:. ~]~::~1,,,I,,,, , I:111 III, ,, ,, ,, ,,

IIIIII

I

,,,,,,,,,r~~,,,,,,,,~,,

N tu ~ ' ~ i . o0!

r-

....

~ . . . . . . . . .

,:._ . . . . . . . . .

~o . . . . . . . . .

~ . . . . . . . . .

~ . . . . . . . . .

-

~"

z~

o

On

~-E

~-~

,

~'"'"""""

"" o

,.b., o",, co

~

to

l-

• liliiiiiiii,i~iiiiiiiiiiiiiiiiiiiiiiiiiiiii~i~i~i~i~ iiiiiiiiiiililili!i ,,,,,,,,,,,,,,~i~i~i~i~i~o~

o~ z~ Oa~-E 0 r LLI

~

e-

r

0 i-

o - -

~

o,i

r

.q-

~o

PINDOS FLYSCH, PELOPONNESE, NW GREECE

501

Fig. 8. Map showing total thickness variations and facies variations in the Pindos Flysch. Character of litharenite: A, thickest beds > 1 m, mostly T,; B, thickest beds 0.5-1 m, mostly Ta or Tab; D, thin bedded, mostly Tb, Tbc, T~. 'Mean palaeocurrents' based on at least five consistent measurements; means of < 5 measurements are shown as 'isolated palaeocurrent measurements'. Numbers in circles are heavy mineral localities of Faupl et al. (2002). It should be noted that the map shows the present geography after substantial mid-Cenozoic compression.

quartzitic conglomerate is found (locality 165). In the northern Peloponnese, sediment dispersal was to the SE, and the shales overlying the Kataraktis Passage Member (section 71, Fig. 5) of the eastern Peloponnese are the most distal exposed sediments of this system. Fleury (1970) and Richter (1993) argued that this sediment was transported axially along the north-south Pindos Trough. There might be a

component of axial transport from northern Greece in the younger rocks of the Central and Eastern zones of the basin, indicated by sparse south-directed sole marks reported by Richter (1993, Fig. 19; see also Fig. 8). Nevertheless, at the western margin of the basin several lines of evidence point to a partial western source of sediment. Calcirudites are present in the Kataraktis Passage Member. Calcturbidites in the

502

D.J.W. PIPER

Kataraktis Passage Member are capped with green shales or grey marls, most easily interpreted as deposited from the same turbidity current, and thus providing evidence for a common western source for calcareous and terrigenous sediment. Red shales are present in sections west of Lambia, but absent in sections to the NE, suggesting a western source. The detailed measured sections show that the Western zone of the Pindos Flysch was deposited in an upper- to mid-fan setting, with the thick shaly sections representing levee deposits adjacent to broad channels, some of which contain conglomerate. In the SW, the palaeocurrents clearly indicate derivation from this western margin. The similar thickness and sedimentological character of the flysch along the entire western margin of the Pindos Basin of the Peloponnese suggests that even in the N W it is a predominantly marginal and not an axial deposit. The Central zone is where most sand was deposited and probably represents a mid- to lower-fan depositional setting, which may have had a contribution of sediment from axial turbidity currents flowing southward. By analogy with modern turbidite systems, the thick conglomeratic sequences fill channels, with thick sandstones representing channel-termination lobe deposits. The shaly flysch in the east represents distal muds. It is possible that the Pindos basin of the Peloponnese was already segmented by internal thrusting during flysch deposition, as has been suggested for the younger Ionian basin to the west (Avramidis et al. 2000), but we have seen no evidence to support this hypothesis. The biostratigraphy of Richter & Miiller (1993) indicates that all the Pindos Flysch Formation of the Peloponnese is of the same latest Paleocene to earliest Eocene age, except for younger MidEocene strata in the extreme SW. Ponded turbidite facies are found only in the Eastern zone and the presence of some sandstone beds, together with very thick mud beds, indicates that this zone did not represent a basin perched above the general level of the Central zone.

Detrital petrology The heavy mineral composition of Pindos Flysch determined by Faupl et al. (2002) from five localities in the Peloponnese places further constraints on palaeogeography. Their locality 3 (Fig. 8, corresponding approximately to locality 197 of this study) includes 5% chrome spinel and lacks blue amphibole. This is from an area in the southwestern Peloponnese that from facies distribution and palaeocurrents clearly consists of sediment

derived from the Apulian margin. This suggests that chrome spinel cannot be used as a provenance indicator for the internal Hellenides. Significant amounts of blue amphibole were identified at two localities. At locality 4 of Faupl et al. on the Koroni peninsula, where the thin Pindos Flysch extends to Mid-Eocene age, blue amphibole makes up 1.1% of the heavy mineral assemblage and is consistently present in all seven samples. Blue amphibole was found in one of four samples (1% of that sample) at locality 5 of Faupl et al. (locality 69 of this study) in the eastern Peloponnese, where the sparse palaeocurrents support a derivation of some sediment from the north (Fig. 8). Blue amphibole was absent from seven samples at locality 1 of Faupl et al. (locality 245 of this study) in zone B of the northern Peloponnese. At their locality 2 (locality 98 of this study), 0.5% blue amphibole was found in one of three samples. The regional age of blueschist belts within the Pelagonian zone (Faupl et al. 1998; Br6cker & Enders 1999) makes this area by far the most likely source of blue amphibole.

Regional tectonic setting The Pindos Flysch Formation was deposited either in a trench basin formed during subduction of the Pindos ocean beneath the Pelagonian microcontinent, or a foreland basin resulting from the westward movement of the Pelagonian nappes. In northern Greece, the palaeocurrent data of Richter (1993) and Gonzales-Bonorino (1996) and petrographic studies of conglomerate (Richter et al. 1993) and sandstone (Faupl et al. 1998) clearly indicate a sediment source to the east, within the Pelagonian nappes that had been advancing since the mid-Cretaceous (Schermer et al. 1990). The consistent southward direction of axial palaeocurrents in the Pindos basin of northern Greece, however, indicates a regional deepening of the basin to the SE. The Pelagonian nappes shed much more sediment to the northern Pindos basin than to the south, suggesting that their topographical expression and loading capacity was greatest in the north, to the north of the Gulf of Corinth transverse lineament. Evidence is lacking for nappe stacking in the Pelagonian zone during the Cretaceous in southern Greece: in Argolis, Cliff (1992) argued that collision did not begin until the Palaeogene, although Schwandner (1998) showed that ophiolite emplacement, probably from the Vardar ocean (Cliff 1992), took place as early as Turonian. Blueschist metamorphism of oceanic crustal rocks in parts of the Cyclades (Br6cker & Enders 1999) indicates late Cretaceous subduction of the Pindos ocean. Some ophiolites of Late

PINDOS FLYSCH, PELOPONNESE, NW GREECE

503

Fig. 9. Schematic illustration of the Paleocene palaeogeography and tectonics of the Pindos basin.

Cretaceous extrusion age, such as at Arvi in Crete (Bonneau 1984), probably formed within the Pindos ocean. High Fe/Mn ratios in midCretaceous manganiferous rocks of the Pindos basin suggest distal drifted deposits from 'black smokers' (Pe-Piper & Piper 1989). We can therefore conclude that in the later part of the Cretaceous, northern Greece was experiencing continental collision along the Pindos suture, whereas ocean crust was still forming in the Pindos basin of southern Greece and was being subducted beneath Pelagonia (Fig. 9). The Eastern zone of the Pindos Flysch Formation could have been ponded outboard from the deformation front at a slow subduction zone of the Pindos ocean beneath the Pelagonian microcontinent, as originally suggested by Robertson et al. (1991). Sediment supply from the east, where collisional deformation and uplift were only just beginning, may have been trapped on the shelf. The sparse occurrence of blue amphibole and the south-directed palaeocurrents suggest that some sediment from northern Greece was transported axially to this zone. Uplift of the Eastern and Central zones into the accretionary prism would have lifted these zones out of the reach of turbidites by the basal Eocene (NP10). As a result, these zones were not sites of deposition for Eocene turbidites derived from the southwestern (Apulian) margin of the Pindos basin or for turbidites transported axially down the Pindos basin from northern Greece. Flysch

sedimentation continued in the extreme southwestern Peloponnese until the Mid-Eocene, and the presence of common blue amphibole in the younger sediments here suggests a Pelagonian source by the Eocene. Continued collision and overthrusting of the Apulian microcontinent south of the Gulf of Corinth line resulted in the development of a foreland basin in the GavrovoTripolitsa and Ionian zones beginning in the Oligocene, within which the more outboard flysch accumulated. The interpretation of the palaeogeography of the Pindos Flysch in the northwestern Peloponnese remains unresolved. The gradual upward passage from the Kataraktis Passage Member, which is clearly derived from the Apulian margin on the basis of palaeocurrents and facies variation, suggests continued supply from Apulia, as in the central Peloponnese. On the other hand, palaeocurrents are directed SSE or SE, rather than ESE as in the Kataraktis Passage Member, and one out of 10 samples analysed for heavy minerals by Faupl et al. (2002) contains a trace of blue amphibole, which might suggest supply of sediment from northern Greece. It could be argued that all the original Pindos Flysch of the Peloponnese was similar to that of northern Greece and the upper part of the formation was removed by erosion (Faupl et al. 2002). Although this possibility cannot be excluded, it is unlikely given the rather uniform age and

504

D.J.W. PIPER

thickness of the Pindos Flysch in much of the Peloponnese. Erosional processes, particularly in orogens, rarely affect large areas uniformly. Furthermore, the very thin flysch succession extending to Mid-Eocene age in the extreme SW Peloponnese (Biichl et al. 2004) clearly demonstrates that there are along-strike changes in the style of flysch deposition. The timing of terrigenous flysch deposition in the Peloponnese from the Apulian margin might be a consequence of either tectonic uplift or sea-level fall. The Turonian 'First Flysch' of northern Greece (derived from the Pelagonian nappes: Wagreich et al. 1996) and the apparently correlative Klitoria Member of the Peloponnese (Degnan & Robertson 1998), derived from the Apulian margin, correspond to the prominent late Turonian fall in sea level (Haq 1993), and Neumannn (2003) also recognized a terrigenous interval in mid-Cenomanian time in both northern Greece and the western Peloponnese. In the Peloponnese, the Pindos Flysch Formation of the Peloponnese is of comparable thickness to the Klitoria Member and corresponds to the prominent Thanetian (Late Paleocene, Ta) lowstand of sea level. Stratigraphic evidence for foreland uplift on the Apulian margin in the GavrovoTripolitsa zone is not found until the MidEocene, with the development of bauxites: this postdates all but the youngest Pindos Flysch of the southwestern Peloponnese. Thus the development of the Pindos Flysch Formation appears to be principally the consequence of eustatic sea-level fall. Sediments that resemble the Kataraktis Passage Member in northern Greece, close to the Albanian border, are termed the 'BuntpelitKalk-Sandstein-Folge' (Richter et al. 1993). In the western thrust slices, they are the oldest known part of the Pindos Flysch and are of Early to Mid-Eocene age. In the east, they are of Mid-Eocene age and overlie Paleocene to Early Eocene terrigenous flysch. They consist of a succession of multicoloured shales, terrigenous turbidites and calc-turbidites that underlie thicker more continuous Late Eocene terrigenous flysch. The 'Buntpelit-Kalk-Sandstein-Folge' lacks blue amphibole (Faupl et al. 1998) and palaeocurrent indicators are not known. By analogy with the Kataraktis Member of the Peloponnese, the Buntpelit-Kalk-Sandstein-Folge might also be derived from the Apulian margin. This is an issue for further study.

Conclusions The Pindos Flysch Formation of the Peloponnese differs from that of northern Greece in being

much thinner and of only Late Paleocene to Early Eocene age, with rocks of Mid-Eocene age in the extreme SW. Facies changes and palaeocurrents indicate that the Pindos Flysch Formation of at least the central Peloponnese, together with the underlying Kataraktis Passage Member, forms a passive margin sediment prism derived from the Apulian microcontinent to the west. Most of the Pindos Flysch of northern Greece, of Late Paleocene to Oligocene age, was deposited in a foreland basin and derived from the evolving Hellenide mountain chain to the east. The younger microcontinental collision south of the Gulf of Corinth line resulted in much less sediment supply from the east to the Peloponnese, and the Pindos Flysch of the Peloponnese was incorporated in the accretionary prism by the Mid-Eocene. Fieldwork was supported by an NSERC Discovery Grant. I thank G. Pe-Piper, P. Degnan, M. Zelilidis and I. Vakalas for helpful discussion in the field, and G. Pe-Piper, P. Faupl and P. Neumann for improving the manuscript.

References AUBOUIN, J. 1959. Contribution ~ l'6tude g6ologique de la Gr6ce septentrional: les confins de L'Epire et de la Thessalie. Annales Gkologiques des Pays Hellkniques, 10, 1~84. AVRAMIDIS, P., ZELILIDIS, A. & KONTOPOULOS, N. 2000. Thrust dissection control of deep-water clastic dispersal patterns in the Klematia-Paramythia foreland basin, western Greece. Geological Magazine, 137, 667-685. BONNEAU, M. 1984. Correlation of the Hellenide nappes in the south-east Aegean and their tectonic reconstruction. In: DIXON, J. E. & ROBERTSON,A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 517-526. BROCKER, M. & ENDERS, M. 1999. U-Pb zircon geochronology of unusual eclogite facies rocks from Syros and Tinos (Cyclades, Greece). Geological Magazine, 136, I 1l-118. BOCHL, W. NEUMANN, P. & ZACHER,W. 2004. New aspects of the Early Tertiary Pindos basin evolution in Messenia, Greece. Extended Abstracts, lOth International Congress of the Geological Society of Greece, 15-17 April 2004. Geotechnical Service of Greece, Thessaloniki, 361-362. CLIFT, P. n. 1992. The collision tectonics of the southern Greek Neotethys. Geologische Rundschau, 81, 669-679. DEGNAN, P. J. & ROBERTSON, A. H. F. 1998. Mesozoic-early Tertiary passive margin evolution of the Pindos Ocean (NW Peloponnese, Greece). Sedimentary Geology, 117, 33-70.

PINDOS FLYSCH, PELOPONNESE, NW GREECE DERCOURT, J. 1964. Contribution ~ l'6tude g6ologique d'un secteur du P61oponn6se septentrional. Annales Gkologiques des Pays Hellbniques, 15, 1418. FAUPL, P., PAVLOPOULOS,A. & MIGIROS, G. 1998. On the provenance of flysch deposits in the External Hellenides of mainland Greece: results from heavy mineral studies. Geological Magazine, 135, 421442. FAUPL, P., PAVLOPOULOS, A. & MIGIROS, G. 2002. Provenance of the Peloponnese (Greece) flysch based on heavy minerals. Geological Magazine, 139, 513-534. FLEURY, J.-J. 1970. Sur les modalit6s d'installation du flysch du Pinde au passage Cretace-Eoc6ne (Gr6ce continentale et P61oponn6se septentrionale). Bulletin de la SocidtO Gdologique de France, 12, 1110-1117. GONZALES-BONORINO, G. 1996. Foreland sedimentation and plate interaction during closure of the Tethys ocean (Tertiary, Hellenides, western continental Greece). Journal of Sedimentary Research, 66, 1148-1155. HAQ, B. U. 1993. Deep-sea response to eustatic change and significance of gas hydrates for continental margin stratigraphy. In: POSAMENTIER, H. W., SUMMERHAYES, C. P., HAQ, B. U. & ALLEN, G. P. (eds) Sequence Stratigraphy and Facies Associations. International Association of Sedimentologists, Special Publications, 18, 93-106. JACOBSI-IAGEN, V. 1986. Geologie von Griechenland. Borntraeger, Berlin. NEUMANN, P. 2003. Ablagerungsprozesse, Event- und Biostratigraphie kreidezeitlicher Tiefwassersedimente der Tethys in der Olonos-Pindos-Zone Westgriechenlands. Miinchner Geowissenschaftliche Abhandlungen, A, 40, 1-156. NEUMANN, P. & ZACHER, W. 2004. The Cretaceous sedimentary history of the Pindos Basin (Greece). International Journal of Earth Sciences, 93, 119-131. NEUMANN, P., RISCH, H., ZACHER,W. & FYTROLAKIS, N. 1996. Die stratigraphische und sedimentologische Entwicklung der Olonos-Pindos-Serie zwischen Koroni und Finikounda (SW-Messenien).

Neues Jahrbuch fiir Geologie und Paliiontologie, Abhandlungen, 200, 4054-424. PE-PIPER, G. & PIPER, D. J. W. 1989. The geological significance of manganese distribution in JurassicCretaceous rocks of the Pindos Basin, Peloponnese, Greece. Sedimentary Geology, 65, 127-137.

505

PE-PIPER, G. & PIPER, D. J. W. 1991. Early Mesozoic oceanic subduction-related volcanic rocks, Pindos Basin, Greece. Tectonophysics, 192, 273-292. PE-PIPER, G. & PIPER, D. J. W. 2002. The Igneous Rocks of Greece. Borntraeger, Stuttgart. PIPER, D. J. W. & PE-PIPER, G. 1980. Was there a western (external) source of terrigenous sediment to the Pindos Zone of the Peloponnese (Greece)?

Neues Jahrbuch fur Geologie und Paliiontologie, Monatshefte, 1980(2), 107-115. PIPER, D. J. W. & STOW, D. A. V. 1991. Fine-grained turbidites. In: EINSELE, G., SEILACHER, A. & RUCKEN, A. (eds) Sequence and Event Stratigraphy. Springer, Berlin, 360-376. RICHTER, D. 1993. Die Flysch-Zonen Griechenlands VII. Sedimentstrukturen, Ablagerungsart und Schfittungsrichtungen im Flysch der Pindos-Zone (Griechenland). Neues Jahrbuch fiir Geologie und Paldontologie, Monatshefte, 1993(9), 513-544. RICHTER, D. & MrOLLER, C. 1993. Die Flysch-Zonen Griechenlands VI. Zur Stratigraphie des Flysches der Pindos-Zone zwischen der Querzone von Kastaniotikos und dem Sfidpeloponnes (Griechenland).

Neues Jahrbuch fiir Geologie und Paldontologie, Monatshefte, 1993(8), 449476. RICHTER, D., MULLER, C. & MIHM, A. 1993. Die Flysch-Zonen Griechenlands V. Zur Stratigraphie des Flysches der Pindos-Zone im n6rdlichen Pindos-Gebirge zwischen der albanischen Grenze und der Querzone von Kastaniotikos (Griechenland). Neues Jahrbuch fiir Geologie uncl Paliiontologie, Monatshefte, 1993(5), 257-291. ROBERTSON, A. H. F., CLIFT, P. D., DEGNAN, P. & JONES, G. 1991. Palaeoceanography of the Eastern Mediterranean Neotethys. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 289-343. SCHERMER, E., Lux, D. & BURCHFIEL, B. C. 1990. Temperature-time history of subducted continental crust, Mount Olympos region, Greece. Tectonics, 9, 1165-1195. SCHWANDNER, F. M. 1998. Polyphase Meso- to Cenozoic structural development on Poros island (Greece). Bulletin of the Geological Society of Greece, 32(1), 129-136. UNDERHILL, J. R. 1989. Late Cenozoic deformation of the Hellenide foreland, western Greece. Geological Society of America Bulletin, 101, 613-634. WAGREICH, M., PAVLOPOULOS, A., FAUPL, P. & MIGIROS, G. 1996. Age and significance of Upper Cretaceous siliciclastic turbidites in the central Pindos Mountains, Greece. Geological Magazine, 133, 325-331.

A new orogenic model for the External Hellenides *T. D O U T S O S , I. K. K O U K O U V E L A S

& P. X Y P O L I A S

University o f Patras, Department o f Geology, Division o f Physical Geology, Marine Geology and Geodynamics, 265 O0 Patras, Greece (e-maik [email protected]) *T. Doutsos, deceased Abstract: In the context of the External Hellenides, 'pro'-lithosphere, corresponding to the Apulian microcontinent, converges on 'retro'-lithosphere, corresponding to the Pelagonian microcontinent. Structural and stratigraphic data in the External Hellenides suggest that the convergence at this margin is fairly well described by the conceptual doubly vergent accretionary wedge model. This new orogenic model for the External Hellenides differs from the classical west-verging assumption and emphasizes that the retro-mass flux is critical for the pro-mass flux. Our model is primarily 2D, and is described in terms of three system components: an accretionary wedge (or pro-wedge), an uplifted plug and a retro-wedge. Three 'isopic' zones (Pindos, Gavrovo-Tripolitsa and Ionian) are included in the pro-wedge. The uplifted plug in the north (Epirus area) includes the Pindos ocean ophiolitic rocks and the Pindos zone, the Parnassos zone in central Greece, and the HP belt of the External Hellenides in the Peloponnese. The retro-wedge includes the Mesohellenic Trough in the north and the Argos plain in the south.

The records of structures and stratigraphy preserved at convergent margins have attracted considerable interest because they offer a reliable method for understanding the cumulative deformation in complex geotectonic settings such as the Tethys Ocean in the Eastern Mediterranean and surrounding areas (Fig. 1). The Hellenides is considered to be a site of complex ocean destruc= tion. The External Hellenides resulted from the Early Tertiary destruction of a Neotethyan oceanic strand known as the Pindos Ocean (Smith et al. 1979; Robertson et al. 1991), which led to the collision between the Apulian and Pelagonian microcontinents (Mountrakis 1986; Robertson et al. 1991; Doutsos et al. 1993). Relicts of the Pindos Ocean are preserved along the suture zone between the External and Internal Hellenides (Fig. 1) (Robertson 2004). The eastern part of the Apulian microcontinent represented a passive continental margin, which was divided, in Late Jurassic time, into lithotectonic zones commonly referred to as the 'isopic' zones of the External Hellenides (Brunn 1956; Aubouin 1959; Bernoulli & Laubsher 1972). During this period, two shallow-water carbonate platforms, the Gavrovo-Tripolitsa and the Pre-Apulian zones, were separated by the Ionian zone, a deep-water basin filled by evaporites, cherts and limestones (Karakitsios 1995). To the east, the Gavrovo-Tripolitsa zone passed gradually into the Pindos zone, which consists of Mesozoic deep-water carbonates, and siliciclastic and siliceous rocks (Smith et al. 1979; Pe-Piper & Piper 1991; Robertson et al. 1991; Pe-Piper

& Koukouvelas 1992). The Pindos Mesozoic sequence was accumulated on a transitional-type crust, which passed gradually eastwards into the Pindos oceanic crust (Degnan & Robertson 1998; Pe-Piper & Piper 2002). Structurally below the Gavrovo-Tripolitsa zone are exposed two metamorphic units known as the Phyllite-Quartzite and Plattenkalk units (Bonneau 1973; Figs 1 and 2). The Phyllite-Quartzite unit tectonically rests over the lowermost Plattenkalk unit, and together they constitute the HP belt of the External Hellenides, which extends from the northern Peloponnese to Crete (Fig. 1). The tectonic evolution of the metamorphic rocks of the External Hellenides began in the Oligocene, involving the intracontinental subduction of the PhylliteQuartzite unit protolith and its basement beneath the Gavrovo-Tripolitsa basement (e.g. Xypolias & Doutsos 2000; Kokkalas & Doutsos 2004). From the north at the Greek-Albanian border to the southern Peloponnese this passive continental margin was irregular and included several microcontinental fragments (e.g. Orliakas, Ultrapindic, Parnassos, inner Pindos; Fig. 2). These fragments were separated by embryonic or welldeveloped strands of the Pindos Ocean that was situated between the Apulian and Pelagonian microcontinents. This original palaeogeographical configuration led to a complex pattern along the Apulian-Pelagonian suture zone (Skourlis & Doutsos 2003). Orogenic movements began in the Eocene first affecting the innermost area of the Pindos zone, and were associated with eastward subduction of

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 507-520. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

508

T. DOUTSOS

ET AL.

Fig. 1. Map of the External Hellenides showing the 'isopic' zones and significant structural elements including the Mesohellenic Though, Pindos and Orthris ophiolites, and the tectonic windows in the Peloponnese. Inset shows the location of the study area within the eastern Mediterranean region.

EOCENE Northwestem

Greece Gavrovo-Tdpolitsa zonek,,

Ionian zone

\

-....4.--

Pindos

Orliakas unit or Ultrapindic zone

-"1--

APULIA

Central G r e e c e Gavrovo-Tripolitsa zone

Ionian

--r_

.... .

.

.

Pindos zone

Pamassos zone

*~

~

.

---'F--

--'F--

~

"""t---

~ ~ - C ~ l

--4--

t

I

I

~

APULIA OCEANIC CRUST

Peloponnese Ionian/Plattenkalk .... ~ ~ -.~ .-t-~

APULIA ~

Phyllite-Quartzite Gavrovo-Tripolitsa unit zone outer Pindos inner Pindos _.,..---r- ~ i ~ . ~ ~ ~ ~ ::_. ,.-,----.~. ~ .....~ , , , , ~ ~ , _ _ - - . r ~ ~_.~: __ ______.......~ "-~-~ - ~ f ~ PE LA GONIA N

~ - - ~ _ _ _ ~ CONTINENTALCRUST

/ TRANSITIONAL CRUST

~

/ OCEANIC CRUST

"'K,'~'J-- I I I -"'---J---L

Fig. 2. Series of schematic east-west cross-sections across the eastern Apulian margin during the Eocene time at three locations (northwestern and central Greece and the Peloponnese), illustrating major Mesozoic rift structures and the 'isopic' zones of the External Hellenides. The margin was irregular, including Apulianderived microcontinental fragments (i.e. the Parnassos zone). Orogenic movements along the margin begun in the Eocene and involved the eastward subduction of the Pindos Ocean beneath the Pelagonian zone.

NEW OROGENIC MODEL, EXTERNAL HELLENIDES the Pindos oceanic crust beneath the Pelagonian microcontinent (Doutsos et al. 1994; Degnan & Robertson 1998). In this tectonic context, the Pindos zone rocks were emplaced westwards onto the Apulian margin, forming the Pindos fold-and-thrust belt (Skourlis & Doutsos 2003, and references therein). Throughout Oligocene and Early Miocene times compressional movements progressively migrated westwards (Aubouin 1959; Jacobshagen 1986; Doutsos et al. 1993, 2000) causing tectonic thickening of the margin. A result of this compression-related deformation was the formation of a foreland flysch basin, which evolved into several 'piggyback' basins. These basins were isolated or semiisolated during Maastrichtian to Burdigalian times (Richter 1976; Doutsos et al. 1987; Underhill 1989; Richter et al. 1992; GonzalesBonorino 1996; Bellas 1997; Avramidis et al. 2002). In the course of this tectonism, a molasse basin of 130 km length and 40 km width known as the Mesohellenic Trough (Fig. 1) was also developed as a 'piggyback' basin along the suture zone between the Apulian and the Pelagonian microcontinents (Doutsos et al. 1994). At present the inner parts of the External Hellenides are undergoing extension (Pavlides et al. 1995; Doutsos & Koukouvelas 1998; Doutsos & Kokkalas 2001), whereas compression is restricted to the most external parts of the orogen, along the Ionian Islands and the western Epirus area (Doutsos et al. 1987; Hatzfeld et al. 1993). The compressional field in the western Epirus area is related to the Adriatic plate collision with northern Greece (Anderson & Jackson 1987). A series of seismic profiles have been shot across the External Hellenides, particularly in the area between the Ionian Islands and the western coast of the Peloponnese, indicating an array of west-directed thrust sheets. Along these profiles, the major thrust faults deform MesozoicTertiary rocks and sole out in a low-angle sub-evaporitic detachment located at a depth of 3-5 km below the Mesozoic carbonate sequence of the Ionian zone. A regionally significant detachment at a depth of 10-15 km below the Ionian and Gavrovo-Tripolitsa zones has also been recorded (BP 1971; Jenkins 1972; Monopolis & Bruneton 1982; Hirn et al. 1996; Sotiropoulos et al. 2003). A remarkable feature of the orogen anatomy is the Moho depth, which reaches its maximum (c. 45-50 km; Makris 1978; Tsokas & Hansen 1997) west of the ApulianPelagonian suture zone, below the Ioannina area. In the present study, we re-evaluate existing structural and stratigraphic data in four key

509

areas along the External Hellenides to determine if similarities exist between the Hellenic orogenic belt and the 'doubly vergent accretionary wedge model' ofWillett et al. (1993). For this purpose, a cross-section trending parallel to the transport direction of major structures was constructed for each key area. Selected sections will be called hereafter the Epirus, Mouzaki, Nafpaktos and Peloponnese sections. The construction of these cross-sections was based on a detailed database of structural data and stratigraphic records mainly within the flysch basins. The database was developed throughout the last decade and was complemented by seismic profiles.

Accretion at convergent margins, the doubly vergent model The doubly vergent accretionary wedge model of Willett et al. (1993) is a simple conceptual model that classifies deformation at convergent margins (Fig. 3). Basic geometric and mechanical components of this model are: an accretionary wedge (Fig. 3; P) on the outboard side of the subduction zone, an uplifted area called an uplifted plug (Fig. 3; U), and a retro-wedge (Fig. 3; R) behind the subduction zone (located within the retrolithosphere). Retro-step-up shear separates the retro-side from the uplifted plug, and the latter is separated from the accretionary wedge (or prowedge) by a pro-step-up shear (WiUett et al. 1993; Beaumont et al. 1999). The conduit (Fig. 3; C) and the subduction channel are formed between the subducted and obducted plates (Fig. 3). The presence or absence of a subduction conduit or channel in this tectonic frame has important implications for the geometry of the wedge (Beaumont et al. 1999). Several other geometric components within the wedge could be related to the rigidity of the retro-lithosphere and the prolithosphere. It is important to note that most of the geometric and mechanical components of these models cannot be confirmed directly in the field without deep seismic profiles.

The Epirus section The ENE-trending Epirus section extends from the Mesohellenic Trough in the east to the frontal parts of the Ionian zone (Fig. 4). The construction of the cross-section was based on structural data collected throughout the area, supplementary seismological data west of the Ioannina city, and a new 70 km long balanced cross-section within the Ionian zone. Three major structural provinces can be distinguished in this section: (1) the Mesohellenic Trough to the east; (2) the

510

T. DOUTSOS E T AL.

Fig. 3. Basic geometric and mechanical components at convergent plate margins. The nomenclature follows the doubly vergent model, proposed by Willett et al. (1993) and Beaumont et al. (1999). The end-member of the model in (a) shows an inactive subduction conduit whereas the end-member in (b) is a fully active subduction conduit.

Pindos fold-and-thrust belt in the central part; (3) a broadly spaced array of thrusts within the Ionian zone in the west (Fig. 4). The length of the section is 185 km. R e t r o - w e d g e p r o vince

The evolution of this province began with the progressive obduction of the relict intervening Pindos Ocean, both eastwards and westwards onto the Pelagonian microcontinent and the Apulian passive margin, respectively (Figs 2 and 4). During Early Eocene time, this process involved westward underthrusting of the Pelagonian microcontinent beneath the Apulian microcontinent (Doutsos et al. 1994; Beccaluva et al. 2004). This westward underthrusting was coeval with the formation of the Pindos

and Krania flysch basins (Fig. 4). During the Early Oligocene this westward subduction was succeeded by the eastward underthrusting of Apulian beneath the Pelagonian microcontinent (Doutsos et al. 1994). Oligocene indentation processes caused overall crustal thickening and formation of the Mesohellenic Trough. The Mesohellenic Trough developed along the suture zone between the Apulian and the Pelagonian microcontinents. It is primarily floored by Pindos ophiolitic rocks and is filled by the Middle Eocene Krania flysch (Figs 4 and 5a) and Oligocene to Lower Miocene molasse sediments (Fig. 5b) (Doutsos et al. 1994; Ferri6re et al. 2004). The trough developed ahead of three hinterland-propagating thrust sheets, which formed in the Early Eocene and propagated fully throughout the Oligocene. At the western border of the Mesohellenic Trough a high-angle reverse fault carries the ophiolitic rocks over the Krania flysch (Fig. 5). The flysch includes ophiolitic detritus, suggesting that obduction was almost complete during the Eocene. East-directed contractional structures affect the Krania flysch, indicating that these structures control the early stages of the trough formation (Fig. 5a; Doutsos et al. 1994). Mesoscopic cross-sections published by Doutsos et al. (1994; p. 259, fig. 2), also provided clear evidence that the ophiolitic rocks occupying the western border of the trough were thrust over the Oligocene Eptachori Formation (Fig. 5b). Mesoscopic east-verging folds and fault-related fold structures are dominant close to the western border of the basin but their occurrence progressively declines toward the east. Sediments caught in the core of some anticlines have a well-defined axial-planar solution cleavage that dips steeply to the SW. Based on these structural data, it seems that the western flank of the Mesohellenic Trough corresponds to a map-scale east-vergent thrust system that resulted in general uplift and thickening of the crust (Fig. 4). Therefore, adopting the nomenclature of Beaumont et al. (1999), the marginal east-directed thrust faults controlling the evolution of the trough can be interpreted as a 'retro-step-up shear' in a retro-wedge (R). According to this interpretation the Pelagonian microcontinent is the retro-lithosphere. Pro-wedge province

The area west of the Mesohellenic Trough to Corfu is mainly occupied by calcareous rocks of the Ionian zone and flysch basins. The deformation in this area appears to be complex, involving flysch-filled piggyback basins in the east and a system of thrust faults to the west (Fig. 4).

NEW OROGENIC MODEL, EXTERNAL HELLENIDES

511

Fig. 4. Simplified geological-tectonic map and an ENE-WSW cross-section in the Epirus area. The crosssection is based on new structural data and published data of Doutsos et al. (1994), Skourlis & Doutsos (2003), Kostakioti et al. (2004), Ferri+re et al. (2004) and Robertson (2004). Age ranges of structures are based on stratigraphic data from the Mesohellenic Trough and flysch deposits throughout the External Hellenides (for further details see the text). E-O, Structures active in Eocene-Oligocene times; M-p, structures active from the Miocene to the present.

For the purposes of this study, we adopted the Mesozoic stratigraphy of the Pindos and Ionian zones at the rifted eastern margin of the Apulian microcontinent, as described by Robertson (2004, and references therein) and Tsikos et al. (2004), as well as the tectonic evolution of the zones into fold-and-thrust belts as proposed by Underhill (1989) and Skourlis & Doutsos (2003). Flysch deposition in the Pindos zone started during the Late Paleocene to Early Eocene (see Piper 2006), whereas in the Ionian zone it began later, in Late Eocene time (e.g. Richter 1976; Richter et al. 1992; Bellas 1997; Faupl et al. 1998). Flysch was deposited in a piggyback fashion, coeval with west-propagating thrusting in the area (Skourlis & Doutsos 2003). Thus, two main flysch basins, the Pindos and Ionian, developed in the Epirus area, accumulating deposits almost 3000 m thick (Xypolias & Koukouvelas 2006; Fig. 4). The carbonates of the G a v r o v o Tripolitsa zone are not exposed in the north

Epirus area. However, given that the closest exposure of the zone is about 30 km south of Ioannina, it can be assumed that in the Epirus section of the Gavrovo-Tripolitsa zone can be included within the nappe stack below the Pindos thrust (Fig. 4). Therefore, we suggest that the thrusting of the Gavrovo-Tripolitsa zone over the inner Ionian zone to the west began at the Eocene-Oligocene boundary, as documented by the biostratigraphy of the flysch in the inner Ionian zone (Bellas 1997). Further west, on Corfu, the west-directed Ionian thrust carries Ionian zone Mesozoic rocks over the Pre-Apulian zone (Fig. 4). Structural studies on the Ionian Islands have shown that Miocene-Pliocene sediments there are deformed by the Ionian thrust (Underhill 1989; Doutsos & Frydas 1994). In the hanging wall of the Ionian thrust, the thrust system is characterized by broadly spaced arrays of both west-verging (fore-) and east-verging (back-) thrusts, which are

512

T. DOUTSOS E T AL.

Fig. 6. Quarry face showing the Solopoulo backthrust (highlighted by dashes). The Solopoulo backthrust carries Lower Jurassic rocks over Eocene-Oligocene flysch with a total offset of c. 1000m.

Fig. 5. Photographs of key structural outcrops showing ophiolites thrust onto the Krania flysch (a) and the Eptachori Formation (b). In both outcrops, the flysch and molasse contain ophiolite detritus.

rooted in a gently dipping detachment at the base of Triassic evaporites (IGRS-IFP 1966; Underhill 1989). Major tip anticlines formed in the hanging wall of both fore- and backthrusts. Imbricate forethrust sheets mainly occur in the outer parts of the Ionian zone. Backward movements are restricted to the middle and the inner parts of the Ionian zone. Prominent features of these movements are the Mitsikeli anticline, which represents a complex fault-propagationfold with an overturned backlimb, and the Soulopoulo backthrust (Kostakioti et al. 2004), which is associated with the development of an east-verging tip anticline on its hanging wall (Fig. 4). The Soulopoulo backthrust carries Lower Jurassic rocks over Eocene-Oligocene flysch deposits, implying an offset of the order of 1000 m (Fig. 6). The abundance of backthrusts in this area implies that the progressive westward advance of the Ionian zone was impeded (blocked) in the middle part of the section by an elevated sub-evaporite structure or an abrupt change in the inclination of the subducting plate. Using diagnostic criteria for blind foreland thrust systems (in the sense of Ferrill & Dunne 1989), we propose the presence of a localized basement

duplex in the middle part of the zone. The presence of such a crustal-scale blind thrust here is also supported by a linear concentration of seismic activity along a subduction zone dipping gently eastward (Martakis 2004). This blind megathrust fault potentially represents a reactivated Mesozoic rift-related fault zone or a bend of the subducting plate. Mean shortening of the Mesozoic cover (Ionian zone) is c. 35%. Furthermore, the concentration of microseismic events at a depth of 5-10 km (Martakis 2004) indicates that this detachment is seismically active. The absence of seismicity eastwards from the Mitsikeli area and in the Mesohellenic Trough suggests that this part of the orogen is inactive. Thus, it is reasonable to assume that the deformation in the frontal part of the orogen (west of Ioannina city) is related to modern subduction. In this study, we examine the reliability of the 'doubly vergent accretionary wedge model' for the apparently fossilized part of the orogen, located east of Ioannina. Summarizing, based on the deformation pattern in the area between Ioannina and the western part of the Mesohellenic Trough, as well as on application of the 'doubly vergent accretionary wedge model', we distinguish the following structural provinces: the pro-wedge area, occupied by the inner Ionian zone, and the uplifted plug, which coincides with the area between the Pindos thrust and the western boundary of the Mesohellenic Trough. According to this interpretation the uplifted plug in the Epirus section was formed during the Eocene-Oligocene period.

The Mouzaki section To investigate further the complex pattern of deformation along the Apulian-Pelagonian

NEW OROGENIC MODEL, EXTERNAL HELLENIDES suture we constructed a 10 km long cross-section, which is located close to Koziakas Mountain and describes the style of deformation in the inner parts of the Pindos zone that have also been referred to as the 'Ultrapindic' zone (Richter e t al. 1992) (Fig. 7, B1-B2). Along the Mouzaki section the thrust system within the Pindos zone is characterized by a branching array of westverging thrusts. A remarkable feature of the deformation in this section is the pronounced increase in thrust fault dip towards the contact between the UltraPindic zone and the ophiolitic complex. The internal deformation of thrust

513

sheets close to the contact is intense and characterized by a dense pattern of tight to isoclinal upright folds, which occasionally appear to be overturned towards the east. However, the clear formation of a basin in a retro-position in this section is missing and the main structural feature of the section is a vertical contact between Pindos zone and ophiolitic rocks.

The Nafpaktos section The Nafpaktos section was constructed with structural data, which were collected along a

Fig. 7. Simplified tectonic map and cross-sections for the Sterea Hellas. The Nafpaktos section (A1-A2) shows structural differentiation across the Pindos fold-and-thrust belt and the role of the Vardousia and Parnassos zones as a back-stop. The Mouzaki section (B1-B2) shows the vertical contact between the Ultrapindic zone and the ophiolites. Simplified and modified from Skourlis & Doutsos (2003) and Sotiropoulos et al. (2003). Age ranges of structures are based on stratigraphic data from flysch deposits throughout the External Hellenides (for further details see the text). E-O, Structures active in Eocene-Oligocene times; O-eM, structures active from Oligocene to Early Miocene times.

514

T. DOUTSOS E T AL.

110 km long cross-section extending from the Parnassos microcontinent to the Ionian zone. This was integrated with subsurface data from a 25 km long seismic profile covering the westernmost end of the section (Sotiropoulos et al. 2003). The section (Fig. 7; AI-A2) shows a pronounced orogenic polarity of structures in the west. In this area, the collision started in the Early Eocene after the closure of a small oceanic strand of the Pindos Ocean lying between the Parnassos and Apulian mircocontinents (Fig. 2). During collision, the Pindos zone was detached along the crust-sediment interface by the Pindos thrust and overthrust westwards onto the GavrovoTripolitsa zone (Degnan & Robertson 1998; Fig. 2). The Pindos zone can be structurally differentiated into a rear domain to the east and a frontal domain to the west, which are characterized by discrete internal deformation styles (Skourlis & Doutsos 2003). The rear domain is characterized by a dense pattern of duplexes at depth, causing folding of earlier low-angle roof thrusts. The resultant open, upright synclines close to the surface are cored by relatively thick flysch deposits. The deformation in the frontal domain is characterized by an imbricate system of moderate- to high-angle thrust faults, which are associated with the formation of thin piggyback flysch basins. Some of the frontal thrust faults are passively rotated backwards attaining a nearly vertical dip. This thrust-fault steepening can be related either to a local emplacement of other faults in their footwall or to impediment of westward movements as a result of an elevated structure at depth. Further west, a Mesozoic normal fault zone, which separated the Ionian zone from the Gavrovo-Tripolitsa zone, was reactivated during Early Oligocene time as a crustal-scale thrust fault, forming the Gavrovo-Arakynthos thrust (Sotiropoulos et al. 2003). The deformational history of this thrust is complex and includes outof-sequence thrusting and flexural bending. In this area, there are structural and stratigraphic records suggesting that the Pindos and GavrovoArakynthos thrusts operated simultaneously until Late Oligocene time (Sotiropoulos et al. 2003). In addition, the absence of seismicity in this area suggests that almost all parts of the Nafpaktos section have not been affected by the present-day subduction, as recognized in the Ionian Islands (Laigle et al. 2002). Summarizing, the geometric and mechanical provinces recognized in this section include the Pindos fold-and-thrust belt as a pro-wedge and the Parnassos zone, possibly part of an uplifted plug.

The Peloponnese section The Peloponnese section, with a total length of 140 km, trends across the southern part of the External Hellenides and extends from the Argos plain to the east through the Plattenkalk zone in the central part into the Ionian zone to the west. The Peloponnese section is mainly characterized by west-directed thrust faults in the external part and the presence of two tectonic windows in the central part, which are cored by HP metamorphic rocks (Fig. 8). Major east-verging structures are recognized at the eastern border of both the Taygetos and the Parnon windows as well as in the Argos area. Based on these data we distinguish the following structural components in the Peloponnese cross-section: a retro-wedge province, an uplifted plug and a pro-wedge province. Retro-wedge province

The retro-wedge province is flanked by the eastern border of the Parnon window and extends eastward to the Argos plain (Fig. 8). The Parnon window represents a NNW-trending anticline that deforms the early ductile thrust contact between the Phyllite-Quartzite unit and the Plattenkalk unit. The geometry of the Parnon anticline also resembles the box-fold geometry of the Taygetos anticline. Particularly important for the structure of the Parnon window is its eastern margin, where a steeply west-dipping backthrust carries the Plattenkalk unit over the Phyllite-Quartzite unit (Doutsos et al. 2000). This marginal backthrust operated under brittleductile conditions and was coeval with the formation of the Parnon anticline, causing the observed nearly vertical dip of the eastern flank of the anticline on its hanging wall (Fig. 8). We interpret this backthrust as a retro-step-up shear in retro-position. Further east, 2.5 km from the eastern border of the Parnon window, folds and shear zones within the Pindos zone show a bimodal west and east vergence. West-directed structures in the form of upright and tight to isoclinal folds are mainly restricted to the region close to the contact between the Pindos and Gavrovo-Tripolitsa zones (Fig. 8). East-directed folds are widespread throughout the rock succession of the inner Pindos zone (Fig. 8) but become progressively dominant as the eastern coast of the Peloponnese is approached. They range in wavelength and amplitude from metres to hundreds of metres. East-verging folds are close to tight and their axial planes are inclined, to recumbent, and asymptotic to bedding (Xypolias & Doutsos 2000).

NEW OROGENIC MODEL, EXTERNAL HELLENIDES

515

Fig. 8. Simplified structural map of the SW Peloponnese (for legend see Fig. 7). The cross-section shows the Parnon and Taygetos structural windows, two crustal-scale backthrusts and the Pindos fold-and-thrust belt. Age ranges of structures are based on stratigraphic data from flysch deposits throughout the External Hellenides and the age of metamorphism in the phyllites (for further details see the text). E-O, Structures active in Eocene-Oligocene times; M, structures active during the Miocene.

A spectacular overturned map-scale fold showing clear east-directed movement is recorded 2 km west of Argos city (Fig. 8; inset).

The uplifted plug This structural area includes the Taygetos and the Parnon tectonic windows as well as the intervening area where HP rocks crop out. The tectonothermal evolution of the HP rocks began in the Oligocene. During this time the deformation was caused mainly by the eastward subduction of the Apulian continental margin beneath the Pelagonian zone in geometric continuity with the preceding subduction of the Pindos Ocean (Degnan & Robertson 1998). Subduction along

the Pelagonian margin was progressively hindered and convergence continued further west along the boundary between the protoliths of the Phyllite-Quartzite unit and the GavrovoTripolitsa zone, resulting in an intracontinental subduction within the Apulian crust. In the course of this tectonism, the Plattenkalk and Phyllite-Quartzite units were buried and underwent HP-LT metamorphism (Katagas 1980; Thiebault & Triboulet 1984). The later stages of intracontinental subduction potentially resulted from the reactivation and conversion of a Mesozoic normal fault zone to a thrust fault, associated with the accumulation of the Oligocene Plattenkalk (Doutsos et al. 2000). In this interpretation the Plattenkalk unit represents

516

T. DOUTSOS

the southern prolongation of the Ionian zone, whereas the protoliths of the Phyllite-Quartzite unit are a Permo-Triassic rift sequence. The exhumation history of the deeper parts of the External HeUenides began at the OligoceneMiocene boundary with the progressive arrival of low-density material in the subduction

ETAL.

channel, which resulted in the initiation of ductile extrusion of the Phyllite-Quartzite unit (Xypolias & Koukouvelas 2001; Xypolias & Kokkalas 2006, and references therein). As horizontal shortening proceeded, blocking of the thrust movement and simultaneous duplexing of the Apulian basement (Fig. 9) occurred, as a result of the

Fig, 9. Interpreted tectonic styles across the Apulian margin during the Eocene, showing basic geometric and mechanical components of the doubly vergent model, and the relative position of major lithotectonic units of the External Hellenides. Dark grey shading indicates undeformed retro-lithosphere; light grey shading indicates the pro-wedge, uplifted plug and retro-lithosphere. (a) Interpretative diagram for the Epirus cross-section, showing major lithotectonic units of each geometric and mechanical component. The dip of the subducting plate is based on seismological (Martakis 2004) and geophysical data (Tsokas & Hansen 1997). (b) Interpretative diagram for the Nafpaktos section, showing the uplifted plug and accretionary wedge. The dip of the subducting slab is from geophysical data of Him et al. (1996), Laigle et aL (2002) and Sotiropoulos et aL (2003). (e) The geometric and mechanical components in the Peloponnese. The dip of the subducting slab is from geophysical data of Monopolis & Bruneton (1982), seismological data of Hatzfeld (1994) and structural data of Doutsos et al. (2000).

NEW OROGENIC MODEL, EXTERNAL HELLENIDES resistance to underthrusting (Xypolias & Doutsos 2000). The progressively increased buoyancy forces during the Early to Mid-Miocene caused vertical expulsion of the orogenic wedge (uplifted plug) as well as the formation of two pop-up structures mapped as the Taygetos and Parnon windows (Doutsos et al. 2000). Pro-wedge province

The pro-wedge province extends from the western flank of the Taygetos window and includes the Pindos zone which was emplaced westwards over the Gavrovo-Tripolitsa zone during the late Eocene to Oligocene (Doutsos et al. 2000). The Pindos zone is characterized by across-strike changes in the style of deformation in this area. The frontal domain of the zone was internally deformed by a dense array of moderate-angle imbricate thrusts, whereas more broadly spaced and gently dipping thrusts occur in the eastern parts (Skourlis & Doutsos 2003). Variation in deformation pattern is also clear on a mesoscopic scale. The western and central parts of the zone contain upright to moderately inclined folds, whereas the eastern part is deformed into inclined and recumbent folds (Xypolias & Doutsos 2000). Recumbent folds, with overturned limbs and west-dipping normal faults, are observed within several klippen located on the western flank of the Taygetos window. Limited gravity movements within the belt possibly occurred during the Early to Mid-Miocene (Xypolias & Doutsos 2000). The pro-step-up shear controlled the rear end of the pro-wedge in the Peloponnese section and borders to the west the core of the Taygetos window. The role of the subduction channel is significant for the Peloponnese section, where the crust between the Pindos Ocean and the Phyllite-Quartzite unit was subducted coevally with tectonic emplacement of the Pindos zone over the Gavrovo-Tripolitsa zone.

Discussion and conclusions Three major Mesozoic rift structures within the eastern margin of the Apulian continent are recognized. From west to east these are located: (1) in the area between the Ionian and GavrovoTripolitsa zones; (2) in the area between the Gavrovo-Tripolitsa zone and the OrliakasUltrapindic, or Parnassos, or the inner Pindos continental fragment (Fig. 2); (3) in the region corresponding to the Pindos Ocean. These Mesozoic rift structures were arranged almost in north-south trending straight lines within the External Hellenides, and were reactivated in the

517

Tertiary to form intracontinental thrusts. Along the pre-existing faults a lithosphere strength reduction took place, as is the case in many intraplate basins in Europe (van Wees & Stephenson 1995; Ziegler et al. 1995). Strain localization along these zones caused strong uplift and crustal thickening, which resulted in a maximum crustal thickness of 50 km, as indicated by estimations for the Peloponnese (Makris 1978; Tsokas & Hansen 1997). In contrast, the suture zone between the eastern parts of the Apulian microcontinent and the Pindos Ocean remained below sea level throughout the collisional stage, and the crustal thickness in this area does not exceed 40 km. The inversion of the intracontinental rift zone located between the Ionian-Plattenkalk and Gavrovo-Tripolitsa zones, as is indicated by the flysch basins, occurred almost synchronously throughout the External Hellenides. This inversion varies considerably in terms of slip rate from 7 mm a -I in the Peloponnese (Doutsos et al. 2000) to c. 1 mm a -1 in central Greece (Sotiropoulos et al. 2003). In the Epirus section, the absence of outcrop data, seismic profiles and boreholes prevents us from determining the role of this rift zone in the evolution of the area, or its slip rate (Fig. 9a). The observed alongstrike differences in slip rate are potentially controlled by different mechanical properties of the crust along this rift zone (e.g. Ziegler et al. 1995; Thompson et al. 2001). Also, particularly important is the role of microcontinental fragments and oceanic strands located east of the Apulian margin (see Robertson 2004). From north to south, the oceanic basins were wider and remained open for a longer time (Jones & Robertson 1991; Degnan & Robertson 1998; B6bien et al. 2000). The microcontinental fragments also became wider southwards (Clift & Dixon 1998). Of these fragments, the most important was the Parnassos microcontinent, an area dominated by frontal accretion (Fig. 9b). In this process the Parnassos zone acted as a strong back-stop (see also Skourlis & Doutsos 2003). According to this interpretation, the buoyancy or the flexing of the downgoing plate possibly blocked backthrusting and the formation of a retro-wedge. In the north, along the Epirus section, the tectonic model specifies an accreted side including the area from the western border of the Mesohellenic Trough and the Pindos thrust (Fig. 9). This area is the pro-wedge, and the opposite, non-accreting side, located within the Mesohellenic Trough, is the retro-wedge. Significant for the Epirus area is the fact that the uplifted plug and related

518

T. DOUTSOS ET AL.

structure were formed during the EoceneOligocene period. The formation of the retroand pro-wedge and the uplifted area between is also well recognized in the Peloponnese section (Fig. 9c). Accordingly, the Argos plain is located in the retro-position, whereas the area west of the Taygetos window belongs to the pro-wedge province. Based on our structural data we can draw the following conclusions. (1) The External Hellenides corresponds well to a doubly vergent orogenic model (Fig. 9). Geometric and mechanical retro-wedge elements are the Mesohellenic Trough and the lowlying area near Argos. The mountain ranges of Smolikas (Pindos zone), Parnassos (Parnassos zone), Parnon and Taygetos (HP belt of the External Hellenides) correspond to the uplifted plug. Pro-wedge elements include the Pindos fold-and-thrust belt and a series of piggyback flysch basins. (2) The inherited intracontinental Mesozoic rift zones and the oceanic strands played a crucial role in the formation of the doubly vergent geometry within the External Hellenides. During this process most of the geometric and mechanical components of the doubly vergent model were formed during the Eocene-Oligocene stages of the orogenic evolution. (3) The uplifted plug in the External Hellenides appears to be physically and mechanically linked with the retro-lithosphere. We acknowledge the editorial assistance and valuable comments of D. Mountrakis. Thanks are extended to the referees, A. Kilias and J. Ferrirre, for constructive criticism.

References ANDERSON, H. & JACKSON,J. 1987. Active tectonics of the Adriatic region. Geophysical Journal of the Royal Astronomical Society, 91, 937-983. AUBOUIN, J. 1959. Contribution /: l'rtude grologique de la Grrce septentrionale: le confins de l'Epire et de la Thessalie. Annales GOologiques des Pays Hellkniques, 10, 1-525. AVRAMIDIS, P., ZELILIDIS, A., VAKALAS, I. & KONTOPOULOS, N. 2002. Interactions between tectonic activity and eustatic sea-level changes in the Pindos foreland and Mesohellenic piggyback basins, NW Greece: basin evolution and hydrocarbon potential. Journal of Petroleum Geology, 25, 53-82. BEAUMONT, C., ELLIS, S. & PFIFFNER, A. 1999. Dynamics of sediment subduction-accretion at convergent margins: short-term modes, long-term deformation, and tectonic implications. Journal of Geophysical Research, 140(B8), 17573-17601.

BI~BIEN, J., DIMO-LAHITFE, A., VERGt~LY, P., INSERGUEIX-FILIPPI, I. & DUPEYRAT, L. 2000. Albanian ophiolitics: I. Magmatic and metamorphic processes associated with the initiation of subduction. Ofioliti, 25, 47-54. BECCALUVA, L., COLTORTI, M., GIUNTA, G. & SIENA, F. 2004. Tethyan vs. Cordilleran ophiolitics: a reappraisal of distinctive tectono-magmatic features of supra-subduction complexes in relation to the subduction mode. Tectonophysics, 393, 163-174. BELLAS, S. 1997. Calcareous nannofossils of the Tertiary flysch (Post Eocene to Early Miocene) of the Ionian Zone in Epirus, NW Greece: taxonomy and biostratigraphical correlations. Berliner Geowissenschaftliche Abhandlungen, E22, 1-173 BERNOULI, D. & LAUBSCHER, H. P. 1972. The palinspastic problem of Hellenides. Eclogae Geologicae Helveticae, 65, 107-118. BONNEAU, M. 1973. Sur les affinitrs ioniennes des 'calcaires en plaquettes' 6pimetamorphiques de la Cr&e, le chariage de la srrie de Gavrovo-Tripolitza et la structure de l'4gren. Comptes Rendus de rAcadkmie des Sciences, 277, 2453-2456. BP (BRITISH PETROLEUM COMPANY LIMITED) 1971. The geological results of petroleum exploration in Western Greece. Institute of Geological and Subsurface Research, 10, 1-73. BRUNN, J. H. 1956. Contribution ~i l'6tude grologique du Pindos septentrional et d'une partie de la Macrdoine occidentale. Annales Gkologiques des Pays Hellkniques, 7, 1-358. CLIFT, P. D. & DIXON, J. E. 1998. Jurassic ridge collapse, subduction initiation and ophiolitics obduction in the southern Greek Tethys. Eclogae Geologicae Helvetiae, 91, 128-139. DEGNAN, P. J. & ROBERTSON, A. H. F. 1998. Mesozoic-early Tertiary passive margin evolution of the Pindos ocean (NW Peloponnese, Greece). Sedimentary Geology, 117, 33-70. DOUTSOS, T. & FRYDAS, D. 1994. The Corfu Thrust (Greece). Comptes Rendus de l'Acad~mie des Sciences, Skrie II, 318, 659-666. DOUTSOS, T. & KOUKOUVELAS,I. 1998. Fractal analysis of normal faults in northwestern Aegean area, Greece. Journal of Geodynamics, 26, 197-216. DOUTSOS, T. & KOKKALAS,S. 2001. Stress and deformation in the Aegean region. Journal of Structural Geology, 23, 455-472. DOUTSOS, T., KONTOPOULOS, N. & FRYDAS, D. 1987. Neotectonic evolution of northwestern continental Greece. Geologische Rundschau, 76, 433-452. DOUTSOS, T., KOUKOUVELAS, I., POULIMENOS, G., KOKKALAS,S., XYPOLIAS, P. • SKOURLIS,K. 2000. An exhumation model of the south Peloponnesus, Greece. International Journal of Earth Science, 89, 350-365. DOUTSOS, T., KOUKOUVELAS, I., ZELILIDIS, A. & KONTOPOULOS, N. 1994. Intracontinental wedging and post-orogenic collapse in the Mesohellenic Trough. Geologische Rundschau, 83, 257-275. DOUTSOS, T., PIPER, G., BORONKAY, K. & KOUKOUVELAS, I. 1993. Kinematics of the Central Hellenides. Tectonics, 12, 936-953. FAUPL, P., PAVLOPOULOS,A. & MIGIROS, G. 1998. On the provenance of flysch deposits in the External

NEW OROGENIC MODEL, EXTERNAL HELLENIDES Hellenides of mainland Greece: results from heavy mineral studies. Geological Magazine, 135, 421-442. FERRII~RE, J., REYNAUD,J.-Y., PAVLOPOULOSA., et al. 2004. Geologic evolution and geodynamic controls of the Tertiary intramontane piggyback MesoHellenic basin, Greece. Bulletin de la SociktO G~ologique de France, 175, 361-381. FERRILL, D. A. & DtrNNE, W. M. 1989. Cover deformation above a blind duplex: an example from West Virginia, USA. Journal of Structural Geology, 11, 421-431. GONZALES-BONORINO, G. 1996. Foreland sedimentation and plate interaction during closure of the Tethys Ocean (Tertiary; Hellenides; Western Continental Greece). Journal of Sedimentary Research, B66, 1148-1155. HATZFELD, D. 1994. On the shape of the subducting slab beneath the Peloponnese, Greece. Geophysical Research Letters, 21, 173-176. HATZFELD,D., BESNARD,M., MAKROPOULOS,K., et al. 1993. Subcrustal microearthquake seismicity and fault plane solutions beneath the Hellenic Arc. Journal of Geophysical Research, 98, 9861-9870. HIRN, A., SACHPAZI, M., SILIQI, R., MCBRIDE, J., MARNELIS, F., CERNOBORI, L. & the STREAMERS-PROFILES Group 1996. A traverse of the Ionian islands front with coincident normal incidence and wide-angle seismics. Tectonophysics, 264, 35-49. IGRS-IFP (INSTITUTEFOR GEOLOGYAND SUBSURFACE RESEARCH OF GREECE-INSTITUT FRANt~MS DU PETROLE) 1966. Etude gdologique de l'Epire (Grkce nord-occidentale). Technip, Paris. JACOBSHAGEN, V. 1986. Geologie yon Griechenland. Gebruder Borntraeger, Berlin. JENKINS, D. A. L. 1972. Structural development of Western Greece. American Association of Petroleum Geologists Bulletin, 56, 128-149. JONES, G. & ROBERTSON, A. H. F. 1991. Tectonostratigraphy and evolution of the Mesozoic Pindos Ophiolitic and related units, Northwestern Greece. Journal of the Geological Society, London, 148, 267-288. KARAKITSIOS, V. 1995. The influence of preexisting structure and halokinesis on organic-matter preservation and thrust system evolution in the Ionian basin, Northwest Greece, American Association of Petroleum Geologists Bulletin, 79, 960-980. KATAGAS, C. 1980. Ferroglaucophane and chloritoidbearing metapelites from the phyllite series, southern Peloponnese, Greece. Mineralogical Magazine, 43, 975-978. KOKKALAS, S. & DOUTSOS, T. 2004. Kinematics and strain partitioning in the southeast Hellenides (Greece). Geological Journal, 39, 121-t40. KOSTAKIOTI, A., XYPOL1AS, P., KOKKALAS, S. & DOUTSOS, T. 2004. Thrust fault damage zones in carbonate rocks: an example from the external Hellenides. Bulletin of the Geological Society of Greece, 36, 1643-1651. LAIGLE, M., HIRN, A., SACHPAZI,M. & CLEMENT, C. 2002. Seismic coupling and structure of the Hellenic subduction zone in the Ionian Islands region. Earth and Planetary Science Letters, 200, 243-253.

519

MAKRIS, J. 1978. A geophysical study of Greece based on: deep seismic soundings, gravity, and magnetics, In: Closs, H., ROEDER, D. & SCHMIDT, K. (eds) Alps, Apennines, Hellenides. Schweizerbart, Stuttgart, 392-401. MARTAKIS, H. 2004. Passive seismic tomographic survey ofEpirus. PhD thesis, University of Patras. MONOPOLIS, D. & BRUNETON, A. 1982. Ionian Sea (western Greece): its structural outline deduced from drilling and geophysical data. Tectonophysics, 83, 227-242. MOUNTRAKIS, D. 1986. The Pelagonian zone in Greece: a polyphase-deformed fragment of the Cimmerian continent and its role in the geotectonic evolution of the eastern Mediterranean. Journal of Geology, 94, 335-347. PAVLIDES, S. B., ZOUROS, N. C., CHATZIPETROS,A. A., KOSTOPOULOS, D. S. & MOUNTRAKIS,D. M. 1995. The 13 May 1995 western Macedonia, Greece (Kozani Grevena) earthquake; preliminary results. Terra Nova, 7, 544-549. PE-PIPER, G. & KOUKOUVELAS, I. 1992. Petrology, geochemistry and regional significance of igneous clasts in Parnassos flysch, Amphissa area, Greece.

Neues Jahrbuch ffir Mineralogie, Abhandlungen, 164, 94-112. PE-PIPER, G. & PIPER D. W. J. 1991. Early Mesozoic oceanic subduction-related volcanic rocks, Pindos Basin, Greece. Tectonophysics, 192, 273-292. Pc-PIPER, G. & PIPER, D. J. W. 2002. The Igneous Rocks of Greece. Gebrfider Borntr~eger, Stuttgart. PIPER, D. J. W. 2006. Sedimentology and tectonic Setting of the Pindos Flysch of the Peloponnese, Greece. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 493-505. RICHTER, D. 1976. Das Flysch-Stadium der Helleniden--Ein Uberblick. Zeitschrift der Deutschen Geologischen Gesellschaft, 127, 467-483. RICHTER, D., MIHM, m. & MOLLER, C. 1992. Die faziellen und pal/iogeographischen Beziehungen zwischen Pindos-Zone und Koziaks-Einheit sowie der Ophiolith-Komplex in West-Thessalien (Griechenland). Zeitschrift der Deutsehen Geologischen Gesellschaft, 143, 67-85. ROBERTSON, A. H. F. 2004. Development of concepts concerning the genesis and emplacement of Tethyan ophiolites in the Eastern Mediterranean and Oman regions. Earth-Science Reviews, 66, 331-387. ROBERTSON, A. H. F., CLIFT, P. D., DEGNAN, P. J. & JONES, G. 1991. Palaeogeographical and palaeotectonic evolution of the Eastern Mediterranean Neotethys. Palaeogeography, Palaeoclimatology, Palaeoecology, 87, 289-343. SKOURLIS, K. & DOUTSOS, T. 2003. The Pindos Fold and Thrust Belt (Greece): inversion kinematics of a passive continental margin. International Journal of Earth Sciences, 92, 891-903. SMITH, A. G., WOODCOCK, N. H. & NAYLOR, M. A. 1979. The structural evolution of a Mesozoic continental margin, Othris Mountains, Greece. Journal of the Geological Society, London, 136, 589-603. SOTIROPOULOS, S., KAMBERIS, E., TRIANTAPHYLLOU, M. & DOUTSOS, T. 2003. Thrust sequences at the

520

T. DOUTSOS ET AL.

central part of the External Hellenides. Geological Magazine, 140, 661-668. THIEBAULT, F. & TRIBOULET, T. 1984. Alpine metamorphism and deformation in Phyllite nappes (external Hellenides, southern Peloponnesus, Greece): geodynamic implication. Journal of Geology, 92, 185-199. THOMPSON, A. B., SCHULMANN, K., JEZEK, J. & TOLAR, V. 2001. Thermally softened continental extension zones (arcs and rifts) as precursors to thickened orogenic belts. Tectonophysics, 332, 115-141. Tsmos, H., KARAKITSIOS,V., VAN BREUGEL, Y., et al. 2004. Organic-carbon deposition in the Cretaceous of the Ionian Basin, NW Greece: the Paquier Event (OAE lb) revisited. Geological Magazine, 141, 401416. TSOKAS, G. N. & HANSEN, R. O. 1997. Study of the crustal thickness and the subducting lithosphere in Greece from gravity data. Journal of Geophysical Research, 102(B9), 20585-20597. UNDERHILL, J. R. 1989. Late Cenozoic deformation of the Hellenide foreland, western Greece. Geological Society of America Bulletin, 101,613-634. VAN WEES, J. D. & STEPHENSON,R. A. 1995. Quantitative modelling of basin and theological evolution of the Iberian Basin (Central Spain): implications for lithospheric dynamics of intraplate extension and inversion. Tectonophysics, 252, 163-178.

WILLETT, S., BEAUMONT, C. & FULLSACK, A. 1993. A mechanical model for the tectonics of doublyvergent compressional orogens. Geology, 21, 371-374. XYPOLIAS, P. & DOUTSOS, T. 2000. Kinematics of rock flow in a crustal-scale shear zone: implication for the orogenic evolution of the southwestern Hellenides. Geological Magazine, 137, 81-96. XYPOLIAS, P. & KOKKALAS, S. 2006. Heterogeneous ductile deformation along a mid-crustal extruding shear zone: an example from the External Hellenides (Greece). In: LAW, R.D., SEARLE, M. & GODIN, L. (eds) Extrusion, Channel Flow and Exhumation in Continental Collision Zones. Geological Society, London, Special Publications (in press). XYPOLIAS, P. & KOUKOUVELAS, I. 2001. Kinematic vorticity and strain rate patterns associated with ductile extrusion in the Chelmos Shear zone (External Hellenides, Greece). Tectonophysies, 338, 59-77. XYPOLIAS, P. & KOUKOUVELAS, I. 2006. Paleostress magnitude in a fold-thrust belt (External Hellenides, Greece): evidence from twinning in calcareous rocks. Episodes, 28, 245-251. ZIEGLER, P. A., CLOETINGH, S. & VAN WEES, J. D. 1995. Dynamics of intra-plate compressional deformation: the Alpine foreland and other examples. Tectonophysics, 252, 7-22.

Geometry and structural evolution of the Mesohellenic Trough (Greece): a new approach A. V A M V A K A , A. K I L I A S , D. M O U N T R A K I S & J. P A P A O I K O N O M O U D e p a r t m e n t o f Geology, A r i s t o t l e University, G R - 5 4 1 2 4 , Thessaloniki, Greece (e-mail: a g n e s _ v a @ y a h o o , co. u k ) The Mesohellenic Trough (MHT) is an elongate basin parallel to the Hellenide isopic zones that extends from southern Albania through northern Greece. The basin developed from Mid-Late Eocene to Mid-Late Miocene time related to Alpine orogenic processes. Structural and kinematic evidence shows that the MHT developed in response to successive tectonic events, involving isostatic crustal flexure, strike-slip and normal faulting, all related to inferred oblique convergence of the Apulian and Pelagonian microcontinents. The Mesohellenic Trough evolved as a piggyback basin above westward-emplacing ophiolites and higher Pelagonian units. This differs from previous interpretations that envisaged foreland flexure related to backthrusting, or subsidence associated with asymmetrical flexure, or normal faulting. The first stage of basin development during the Mid-Late Eocene was contemporaneous with the final emplacement of Pindos oceanic units and culminated in deformation and uplift of Eocene strata. The second phase was dominated by strikeslip faulting during Oligocene-Early Miocene time. The third stage was characterized by low-angle normal faulting at the eastern boundary of the MHT during the Early-Late Miocene. The evolution of the sedimentary basin ended around Late Miocene time, followed by rapid uplift and marine regression. A compressional event occurred during the latest Miocene. Finally, extensional tectonics affected the area from the Late Miocene to .the present. Abstract:

The Mesohellenic Trough (MHT), the largest and most important late orogenic 'molasse-type' basin of the Hellenides, formed during the latest stages of Alpine orogenesis and was filled by marine turbidites and siliciclastic shelf deposits. The basin, c. 200 km long by 30-40 km wide, extends with a NW-SE trend from southern Albania in the north through Greece, passing southwards by the cities of Kastoria, Grevena and Kalambaka and finally beneath the younger Neogene and Quaternary deposits of the Thessaly plain (Fig. 1). The basin is characterized by sediments up to 4 km thick that vary along the axis of the MHT and include fan-delta conglomerates, alluvial fans, turbiditic sandstones and shales, deltaic and flood-plain sandstone and siltstones, and sandy shelf sediments (Zelilidis et al. 1997, 2002). Brunn (1956) was the first to map and distinguish distinctive sedimentary formations within the basin. Subsequent studies focused on mapping (Brunn 1956, 1960; Savoyat & Lalechos, 1969, 1972; Savoyat et al. 1971a, b; Savoyat & Monopolis 1972; Mavridis & Matarangas 1979; Koumantakis 1980; Mavridis & Kelepertzis 1985), sedimentary analysis and the nature of depositional systems (Papanikolaou & Sideris 1977; Papanikolaou et al. 1988; Desprairies 1979; Ori & Roveri 1987; Wilson 1993; Zelilidis et al.

1997) and palaeontological evidence (Soliman & Zygogiannis 1980; Zygogiannis & Sidiropoulos 1981; Zygogiannis & Mtiller 1982; Barbiery 1992). In addition, seismic data (Kontopoulos et al. 1999; Zelilidis et al. 2002) and structural data were also utilized in a few recent studies (Doutsos et al. 1994; Ferri+re et al. 1998, 2004). The Mesohellenic Trough developed from the Late Eocene to the Late Miocene in the area of the suture located between the Apulian microplate and the Pelagonian continental block. The basin formed in several stages as successively overlapping basins (Ferri6re et al. 1998). Previously it was suggested that a foreland depression developed in front of backthrust faults, dipping to the west (Doutsos et al. 1994), or that an asymmetrical flexure depressed the eastern side of the basin (Ferri6re et al. 1998), controlled by normal faulting during eastward subduction of the Pindos basin and collision of the GavrovoTripolitsa platform unit (Ferri6re et al. 2004). Recently, Zelilidis et al. (2002) proposed that the basin formed as a strike-slip half-graben, based on seismic data. A connection with strikeslip faulting was also suggested by Vamvaka et al. (2004). The aim of this study is to determine the structural evolution of the Mesohellenic Trough through time and to suggest a new tectonic model for its formation.

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 521-538. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

522

A. V A M V A K A E T AL.

or

O

o ..,.4,

~z

[-. ,,,,~

,aZ

o . ,..., +.~ o

MESOHELLENIC TROUGH, GREECE

523

Fig. 2. Geological map of the Mesohellenic Trough (based on Brunn 1956; Savoyat & Lalechos 1969, 1972; Savoyat et al. 1971a, b; Savoyat & Monopolis 1972; Vamvaka et al. 2004). A-A' is the cross-section shown in Figure 12.

Geological setting The Mesohellenic Trough developed parallel to the isopic zones of the Hellenides, and today is sited between the external and internal Hellenide zones and has a NW-SE trend (Aubouin 1959; Mountrakis 1986; Fig. 1). Thick sediments within the basin overlie ophiolitic rocks and Cretaceous limestones. Brunn (1956) divided the sedimentary fill of the basin into five main siliciclastic formations (see Fig. 2), which are, from bottom to top: the Krania Formation (of Mid-Late Eocene age); the Eptahori Formation (of Mid-Late Oligocene

age); the Pentalophos Formation (of Aquitanian age); the Tsotyli Formation (of Late AquitanianTortonian age); the Ondria Formation (of Mid-Late Miocene age). These stratigraphically defined formations are retained here, although more detailed studies of facies, lateral lithostratigraphic relations and the internal unconformities have been carried out more recently (e.g. Desprairies 1979; Papanikolaou e t al. 1988; Wilson 1993; Zelilidis e t al. 1997, 2002). The biostratigraphy of the basin is based mainly on planktic Foraminifera and nannoplankton (Zygogiannis & Sidiropoulos 1981;

524

A. VAMVAKA E T AL.

Zygogiannis & Mfiller 1982; Barbiery 1992; Kontopoulos et al. 1999). The Krania Formation, with an estimated maximum thickness of 1500 m (Brunn 1960; Ferri6re et al. 2004), is characterized by various facies including coarse breccias and olistolithic blocks, fan delta deposits and turbiditic siltstones and fine-grained sandstones (Wilson 1993; Zelilidis et al. 2002). The Eptahori Formation, with an estimated thickness of 1000m (Brunn 1960; Savoyat & Monopolis 1972), consists of conglomerates and sandstones that are overlain by marine turbiditic shales with lignitic horizons (Zelilidis et al. 2002). Nannofossils indicate a Ruppelian age in the north, whereas benthic Foraminifera suggest a water depth of around 600 m (Zygogiannis & Mfiller 1982; Barbiery 1992). Southwards, the basin thickness, as estimated by seismic data, increases to c. 1200 m (Zelilidis et al. 2002). In the southern part of the basin the formation is more sandy, consisting of marine sandstones and some pebbly conglomerates, suggesting southward shallowing. The base of the Pentalophos Formation consists of conglomerates, followed by alternating turbiditic sandstones and shales, with minor conglomerates (Brunn 1956, 1960; Zelilidis et al. 1997, 2002). The estimated thickness is 2500 m. Near the centre of the southern part of the basin (Meteora area), the formation is conglomeratic and is characterized by several unconformities. The palaeo-bathymetry of the Pentalophos and Tsotyli formations is unknown, but the facies types of the Pentalophos Formation suggest water depths of 300-700 m (Zelilidis et al. 2002). The base of the Tsotyli Formation, estimated as 1500 m thick (Mavridis & Matarangas 1979; Mavridis & Kelepertzis 1985), is characterized by conglomerates that are mainly ophiolite-derived in the northern part of the basin and polygenic, derived from the Pelagonian continent, in the south. The conglomerates pass upwards into alternating turbiditic conglomerates, sandstones and shales. In the southern part of the basin, the Tsotyli Formation lies unconformably on the Pentalophos Formation, although this unconformity is not observed in the northern part of the basin. In the outer Theotokos village area (Fig. 2) the Tsotily Formation is faulted against the Eptahori Formation, whereas in the southernmost part of the MHT, east of Vassiliki village (Fig. 2), the Tsotily Formation directly overlies Eocene strata. In places, the Tsotyli Formation is overlain by sandy shelf deposits (i.e. sandstone, marl and limestone) of the Ondria Formation that accumulated in a shallow-water setting (Savoyat

et al. 1971a). The Ondria Formation remains in only a few places of the MHT (Fig. 2) probably because of erosion (Papanikolaou et al. 1988). The shallow-water Ondria Formation may relate to rather rapid uplift of the basin, contemporaneous with marine regression during the Tortonian, but without completely filling the basin with clastic material (Papanikolaou et al. 1988). With the exception of the Krania Formation in the westernmost part of the basin (termed the 'Gulf of Krania' by Brunn 1956), the other four formations were deposited parallel to one another from west to east, respectively (Fig. 2). They show an eastward migration within time, as shown by their location and orientation on the map in relation to their age (Brunn 1956; Zygogiannis & Mfiller 1982; Barbiery 1992). At the western edge of the basin, the strata dip towards the ENE at steep angles; dips decrease progressively away from this basin margin, whereas in the centre and along the eastern margin of the basin the strata dip with a low angle towards the WSW. As a result an asymmetrical syncline formed, controlled by structural and depositional processes. The MHT splits into two narrower synclines in the south separated by an uplifted structure (Theotokos and Theopetra village areas).

Geometry and kinematics of deformation Compressional structures are evident only in the Eocene strata of the Krania Formation. Reverse faults trending NW-SE are associated with nearly NE-SW-trending dip-slip striations on fault planes. Asymmetric folds, also NW-SE trending, form the main compressional structures within the Krania Formation (i.e. Krania village area). The sense of movement shown by both faults and folds is towards the SW and NE (Figs 3a and 4). As previously documented, the northern and southern margins of the 'Gulf of Krania' are bounded by strike-slip faults (Papanikolaou et al. 1988; Ferri6re et al. 1998). The contact between the Krania Formation and Mesozoic ophiolites beneath was described in previous studies as a thrust (Wilson 1993; Doutsos et al. 1994; Ferri6re et al. 2004). The ophiolites and the Krania strata are almost concordant, dipping at a high angle (c. 80 ~ to the west or east. The western boundary of the MHT is largely defined by a steeply dipping fault forming an impressive morphology in many places, as well observed near the villages of Spileo, Filippei and Alatopetra, areas where dextral strike-slip faults trend NW-SE, and exhibit both normal and

MESOHELLENIC TROUGH, GREECE B

+

i (a)

~+ ~

(b) Fig. 3. (a) Palaeostress analysis diagram of the reverse faults observed in the Krania Formation (T1 event). (b) Palaeostress analysis diagram of dextral and sinistral strike-slip faults (T 2 event). Stress axes: circle, cyl;diamond, ~2; square, ~3. Lower hemisphere, equal-area stereographic projection. The fault planes and slip direction are shown.

reverse dip-slip components (Figs 4-6). Dextral strike-slip faults with slightly different orientations ( N N W - S S E ) and small reverse dip-slip components also occur towards the centre of the southern part of the M H T (e.g. Theotokos village area; Figs 2 and 4), where an uplifted flower-shaped structure is formed that exposes

525

basement rocks (i.e. ophiolites and Cretaceous limestones) and the Eptahori Formation (Fig. 7). West of this structure, strata dip to the WSW, whereas on the eastern side of the basin they dip to the ENE. Dextral strike-slip faults in this area trend parallel to each other and have formed a flower structure that controlled sediment deposition during Oligocene-Early Miocene time (Figs 5 and 7). A small reverse component in the strike-slip faults implies that these faults developed under a transpressional regime. A positive flower structure was also recognized by Zelilidis et al. (2002), based on study of a seismic profile of a specific area that shows the strike and dip of strata, as well as several unconformities. A second dip-slip striation overprinting the strike-slip one is observed on many fault surfaces. Dip-slip striations on cataclastites are superimposed on strike-slip striations, showing that strike-slip preceded normal faulting. The strike-slip faults of the western boundary and in the centre of the M H T (e.g. Eptahori and Theotokos faults) were interpreted as thrust faults by Doutsos et al. (1994) and as faulted flexures by Ferri~re et al. (1998), although normal faulting was reported in a more recent paper (Ferri6re et al. 2004). Previous workers recognized strike-slip movement in some places, but considered this to be of minor importance compared with reverse or normal faulting (Doutsos et al. 1994; Ferri6re et al. 2004). Sinistral strike-slip faults, striking N E - S W to E N E - W S W , are documented in many parts of the M H T with the same kinematic relations and relative ages as the N W - S E dextral strike-slip faults; these are interpreted as antithetic Riedel faults to the main dextral faults (Fig. 3b). Low-angle normal faults with a small sinistral component (Figs 4, 8 and 9) were observed at the eastern boundary of the MHT. These faults occur at the contact between the Tsotyli Formation and the Pelagonian basement (Figs 2 and 4); they show synsedimentary activity but do not affect the younger, Pliocene deposits. These normal faults exhibit a N W - S E strike (e.g. Pylori and Kerasoula village areas) and a southwestward sense of movement (Fig. 8). They have contributed to the subsidence of the eastern part of the basin where the Tsotyli Formation was deposited. A small number of dip-slip reverse faults, striking N W - S E , were also observed in the Miocene strata of the Tsotyli Formation. High-angle normal faults that strike in several different directions (Figs 4, 10 and 11) cut the basement rocks, the M H T formations and Plio-Quaternary deposits; they also overprint all

526

A. VAMVAKA E T AL.

Fig. 4. Tectonic map of the MHT showing the main faults developed during the different tectonic events and palaeostress analysis diagrams for each event and region. Diagrams: 1 for Tj event; 2, 3 and 4 for T 2 event; 5 for T3 event; 6, 7 and 8 for T5 event. previous structures. Some of these normal faults, generally those oriented east-west, are believed to be still active (Chatzipetros 1998; Chatzipetros et al. 2005). Two high-angle normal faults with a N N W SSE orientation dip to the east some distance south of Theotokos village (Fig. 10). These faults cut the Pentalophos and Eptahori Formations, and account for the direct juxtaposition of the Tsotyli Formation with the Eptahori Formation (Fig. 2). Further north, following the strike of these faults, the contact between the Pentalophos Formation and the Tsotyli Formation, along

which the Aliakmonas River runs, may be also characterized as a normal fault. Tectonic events

Several sets of structures, as described above, record a complex deformational history under brittle conditions, from Late Eocene to Quaternary time. To assess the stress regime governing each deformational event, we have calculated its stress tensor from a large number of fault-slip data. For this palaeostress analysis we used both the P - T method (after Turner 1953) and the Angelier (1979) method.

MESOHELLENIC TROUGH, GREECE

527

Fig. 5. Cross-section of the Theotokos village area where parallel dextral strike-slip faults occur bounding the Eptahori Formation.

Fig. 6. (a) Marginal strike-slip fault of the western basin boundary (Phillipei-Alatopetra villages area). The arrows show two movements that occurred in different periods (see text for explanation). (b) Nearly horizontal striation on the dextral strike-slip fault plane on Cretaceous limestones at Spileo village area, close to the contact with the Eptahori Formation.

1"1 event. The Mid-Eocene to Early Oligocene period corresponds to the first tectonic event (TO; this resulted in the deformation and uplift of the Eocene strata, producing folds and reverse faults. Palaeostress analysis indicates on almost horizontal maximum principal stress axis (C~l),

trending N E - S W , and on almost vertical minim u m principal stress axis (~3) (Figs 3a and 4). 7"2 event. Strike-slip faults are assigned to a second tectonic event (T2), dating from the Early Oligocene to Early Miocene, as they occur

528

A. VAMVAKA E T A L .

Fig. 7. (a, b) Schematic cross-sections across the MHT, showing the marginal strike-slip faults and a flower structure in the Theotolos village area (T2 event). (e) Low-angle and high-angle normal faults of the T 3 event (see text for explanation).

MESOHELLENIC TROUGH, GREECE

529

A comparable compressional event of the same period was documented by Kilias et al. (2001) further north, within Albania, where the Mirdita ophiolites overthrust Upper Miocene molasse-type sediments along reverse faults, giving rise to similar kinematic features to the 7 4 event. Palaeostress analysis by Kilias et al. (2001) suggested that the stress regime that produced these structures was characterized by a subhorizontal NE-SW-oriented c~ and a subvertical 0-3.

Fig. 8. Low-angle shear zone observed in the ophiolites, close to their contact with the Tsotyli Formation (eastern margin of the MHT, close to Pilori village).

between basement rocks (Cretaceous limestones and ophiolites) and the Eptahori Formation (Figs 2, 4-7). These faults occur along the western MHT boundary and in the Theotokos village area, and appear to affect the Eptahori and Pentalophos formations as well as the Krania deposits, whereas the Tsotyli Formation does not appear to be affected. The palaeostress analysis indicates an almost horizontal ~l, with a N N E - S S W strike (Fig. 3b), showing a small change in direction compared with the first tectonic episode (TO. In contrast to %, which retains almost the same orientation, ~3 is horizontal with an E S E - W N W orientation. 7"3 event. The third tectonic event (T3) is related to extensional tectonics and is characterized by low-angle normal faulting with a small sinistral component. These faults are assigned to the Early Miocene, as they relate to the subsidence of the eastern part of the basin where the Tsotyli Formation was deposited. Palaeostress analysis shows that 0-3 was almost horizontal with a N E SW direction, and Cyl was nearly vertical (Figs 4 and 9), indicating an extensional stress regime (Fig. 3b). T4 event. The NW-SE-trending reverse faults that cut Miocene strata of the Tsotyli Formation are inferred to relate to a compressional event (T4), of relatively local importance. The T4 event is assigned to a Late Miocene age, as it affects the Tsotyli Formation but not younger (Pliocene) deposits. In addition, in the cross-section of the I G M E Knidi Sheet (Mavridis & Kelepertzis 1985), reverse faults are shown between the basement and the Tsotyli Formation near Knidi (close to Grevena); their identification was based on seismic data.

T5 event. The last tectonic event (Ts), assigned to post-Late Miocene time, comprises high-angle normal faults that affect all of the formations. Two high-angle NW-SE-trending normal faults south of Theotokos village are inferred to mark the contact between the Eptahori and the Tsotyli formations (Fig. 10), and are assigned to the T5 event. ENE-WSW-trending normal faults of this stage are responsible for the elongate topography that is today followed by the Aliakmonas (north of Mount Vourinos), Ionas (southern of Theotokos village) and Pineos Rivers (Trikala region). The orientation of these faults suggests that some of them utilized pre-existing weak zones of Oligocene age at a time when sinistral faults of the same trend developed under a different stress regime (i.e. T~ and T2 events). It is also possible that the rivers were controlled by even earlier structures. The Plio-Pleistocene age and the rectangular shape of the small Karperou basin (directly south of the Vourinos Massif and the Aliakmonas River) reflect a possible control by young E N E - W S W normal faults (at its northern and southern margins); these faults were active after the end of the Miocene, causing subsidence of specific areas. It is noteworthy that the basin's northern margin coincides with the Aliakmonas River (Fig. 2). Palaeostress analysis indicates 0-3 oriented subhorizontaUy from N E - S W to c. north-south; the younger orientation (north-south) coincides with the extension direction of recent seismic activity of the area, including an earthquake of magnitude 6.5 Richter in 1995 (Papazachos & Papazachou 1997). This earthquake caused great damage to the city of Kozani and many villages in the area. The 0-1 of this fault is vertical (Fig. 4). The T5 event marks a generally extensional period, which started in the Early Miocene and continues today (i.e. cy3is north-south).

Discussion of structural evolution The formations of the MHT were deposited subparallel to one another (Fig. 12), showing a

530

A. VAMVAKA E T AL.

Fig. 9. Palaeostress analysis diagram of low-angle normal faults (T 3 event). Stress axes: circle, cyl; diamond, cy2;square, o3. Lower hemisphere, equalarea stereographic projection. The fault planes and slip directions are shown. migration of depocentres from west to east through time. The major geodynamic control of basin development is attributed to underthrusting of the External Hellenides beneath the

Pelagonian microcontinent (Mountrakis 1986; Ferri6re et al. 2004). We have recognized five main stages in the evolution of the MHT region in relation to regional tectonic events (see Figs 13-15). The basin evolution was initiated in the MidEocene, nearly contemporaneously with the thrust imbrication of Pindos oceanic units (Jones & Robertson 1991) and the deformation of the Pelagonian upper plate (Mountrakis 1986; Kilias et al. 1991a). A compressional regime with cr1 oriented N E - S W dominated this period (Figs 3a and 4). The first sub-basins that have Eocene deposits developed on an ophiolitic basement affected by flexural subsidence beneath an advancing load made up of the thick Pindos thrust-fold belt (Fig. 13; for examples of flexural processes, see e.g. (Karner & Watts 1983; Royden & Karner 1984; Moxon & Graham 1987). Passive isostatic subsidence became active as the basin began to infill with sediments. Complicated stratal relationships developed in places between the basin fill and the evolving structural high above the Pindos accretionary complex. Transverse faults with an ENE-WSW trend and an inferred lateral slip, as observed along the margins of the Krania basin, may have occurred during this period. The first stage ended in the Early Oligocene with uplift and deformation of the two sub-basins (Fig. 13).

Fig. 10. Photograph and schematic cross-section of two normal faults observed south of Theotokos village, which cut the Eptahori and Pentalophos formations, juxtaposing the Tsotyli Formation directly with the Eptahori Formation. The three formations are marked on the photograph and the cross-section.

MESOHELLENIC TROUGH, GREECE

531

Fig. 11. Cross-section of the foothills of Mount Vounassa (eastern MHT margin). Low-angle (F0 and highangle (F2) normal faults related to T3 and T5 tectonic events, respectively, are shown (see text for explanation). The fault that occurs between the ophiolites and Triassic limestones was related to the T3 event according to Mountrakis et al. (1992).

The second stage was associated with the subsidence of a narrow, elongate basin in which the Eptahori and Pentalophos formations were successively deposited from Early Oligocene to Early Miocene time (Figs 2 and 14). This stage was controlled by strike-slip faulting bounding the basin, and the main locus of contraction during the continuing convergence migrated progressively westwards, cy~ shows only a small change in orientation during the previous stage, whereas, in contrast cy3 was almost horizontal (Fig. 4). Dextral strike-slip N W - S E faults developed under this stress regime, which tended to be transpressional as indicated by small reverse dipslip components. In this sense, the basin could be seen as a type of pull-apart basin that developed between strike-slip faults. The strike-slip fault at the western margin of the basin acted as a master fault that controlled the basin development, whereas the strike-slip fault that bounded the eastern margin of the basin was apparently cut by normal faults during the following tectonic event (T3) and later covered by the Tsotyli Formation. Alternative models for the evolution of the M H T relate the main subsidence to reverse faulting (Doutsos et al. 1994) or normal faulting (Ferri6re et al. 2004). Doutsos et al. (1994) related the evolution of the MHT to compression and backthrusting. However, we have not identified clear compressional structures, specifically thrusts dipping to the west (except within the Krania Formation). In contrast, Ferri6re et al. (2004) related the main subsidence of the basin

to normal faulting after Early Oligocene time. Although normal faulting was also recognized by us, this did not start until the Miocene. Variable water depths along the axis of the basin, as shown in a palaeo-bathymetry map based on seismic data (Zelilidis et al. 2002), and an uplifted structure in the middle of the basin, can both be related to strike-slip faulting; strike-slip basins commonly experience localized episodes of rapid subsidence or uplift, resulting in unconformities (Figs 7 and 14). The synclinal structure preserved by the Eptahori and Pentalophos formations (Figs 2 and 12) can be related to synsedimentary tectonics and continuing sedimentary loading of the area during the T2 stage. The next deformational stage (T3) involved low-angle normal faulting and took place from Early to Mid-Late Miocene (Fig. 15). Subhorizontal N E - S W oriented extension dominated this period (Figs 4 and 9). The compression moved further west and the region experienced late orogenic collapse under a plate convergence regime (Kilias et al. 1991a, b; Mountrakis et al. 1992) (Fig. 15). This change in tectonic framework resulted in some deformation of the eastern margin of the inferred Oligocene strike-slip basin. The marginal low-angle N W - S E normal faults with a small sinistral component cut the previous eastern margin of the MHT, and caused subsidence and some widening of the basin where the Tsotyli Formation was deposited (Figs 7,

532

A. V A M V A K A E T AL.

2

O

az

C'q

0 .,~

0

e,i

MESOHELLENIC TROUGH, GREECE

533

Fig. 13. Schematic cross-sections and map view showing the first stage of the MHT tectonic evolution during the Mid-Eocene to Early Oligocene (T~ event). 12 and 15). Low-angle normal faulting utilized pre-existing low-angle extensional shear zones of similar kinematics, which also affected the surrounding ophiolitic and Pelagonian basement rocks (Kilias et al. 1991a, b) (Fig. 8). In the southeasternmost part of the MHT, the Tsotyli Formation was deposited directly on Eocene strata (Fig. 2), suggesting that this region might not have been an active depocentre during Oligocene-Early Miocene time, perhaps because

of uplift and erosion at the beginning of the Oligocene. This area started receiving sediment again after Early Miocene time, when low-angle normal faulting affected the area. The deposition of the Ondria Formation followed during Mid-Late Miocene time in rather shallow water. Today, this formation remains only in a few places in the MHT (Fig. 2), following rapid uplift, contemporaneous with marine regression around Tortonian time.

534

A. VAMVAKA E T AL.

Fig. 14. Schematic cross-section and map view showing the second stage of the MHT tectonic evolution during the Oligocene to Early Miocene (Y 2 event).

During the Late Miocene the compressional event caused local thrusting of the MHT and in some places overthrusting of the ophiolites onto Miocene sediments. T4 compression occurred in a generally extensional period, characterized by orogenic collapse and uplift of the Hellenides after Eocene crustal overthickening (Lister et al. 1984; Kilias et al. 1991a, b, 2001). This shows that orogenic extension can be occasionally interrupted by compressional events. Finally, the late stage of the MHT evolution is connected with the Ts deformational event, which Y4

affected the MHT and the younger deposits from the Late Miocene to the present day. This period is characterized by high-angle normal faulting with variable orientations (Fig. 4). Some of these faults are still active. According to the classification of basin types of Busby & Ingersoll (1995), the basin geometry, stratigraphy, stratal relationships and kinematics of deformation suggest that the MHT can be considered as a polyhistory strike-slip basin. Changing structural settings and repeated episodes of rapid subsidence and uplift characterize

MESOHELLENIC TROUGH, GREECE

535

Fig. 15. Schematic cross-section and map view showing the latest stages of the MHT tectonic evolution during the Early to Late Miocene (T3 and T4 events). long-lived strike-slip zones, such as the San Andreas Fault (Crowell 1974a, b, 1987). The MHT developed as a result of different tectonic events that include isostatic crustal flexure, strike-slip and normal faulting; this corresponds to the pattern for polyhistory strike-slip basins (Busby & Ingersoll 1995). Supporting criteria are as follows: (1) asymmetry and the length-to-width ratios (4:1), typical of strike-slip basins; (2) axial infill, subparallel to the principal displacement zones; (3) lateral migration of the depocentres, parallel to the principal bounding faults; (4) the presence of

diverse depositional facies including landslide, alluvial-fan, fan-delta and turbidites; (5) the presence of thick but laterally restricted sedimentary sequences characterized by high sedimentation rates; (6) abrupt lateral and vertical facies variations; (7) localized uplift and erosion (e.g. Theotokos area), resulting in unconformities of the same age; (8) a strike-slip fault, which certainly exists along the western side of the MHT and possibly is present also along the eastern side, although obscured by subsequent normal faulting. The basin developed during convergence of the Apulian and Pelagonian

536

A. VAMVAKA E T AL.

microcontinents that was possibly oblique. The M H T can also be seen as a piggyback basin, as it developed above migrating ophiolites and the Pelagonian upper plate, simultaneously with the underthrusting of the External Hellenides (Ferri6re et al. 1998, 2004).

Conclusions (1) This study emphasizes the role of strike-slip in the structural evolution of the M H T as a polyhistory strike-slip basin. (2) The overall evolution of the basin took place under a plate convergence regime during a time when compression migrated westwards to the more external area. The plate convergence was possibly oblique, based on our kinematic analysis. (3) In agreement with Ferri~re et al. (2004), the M H T can be characterized as a piggyback basin, as it developed on top of the migrating ophiolites and Pelagonian upper plate, simultaneously with the underthrusting of the External Hellenides. (4) The first sub-basin developed during the M i d - L a t e Eocene by crustal flexure and subsidence as the result of loading of the overthickened Hellenide accretionary prism (Fig. 12). During the ensuing basin closure, intense deformation and uplift, the Eocene sediments at the western basin margin were tilted to a high angle or locally inverted, becoming concordant with the adjacent ophiolitic basement. (5) The Oligocene-Early Miocene period was characterized by strike-slip faulting, which controlled the subsidence and evolution of the basin during this time (Fig. 13). Strikeslip faults are today recognized along the western margin and near the centre of the MHT. (6) Extensional tectonics dominated the latest stages of evolution of the M H T , from Early Miocene time (Fig. 14). Extension was responsible for the subsidence of the eastern part of the basin along low-angle normal faults related to late orogenic collapse and the uplift of the Olympos window (Kilias et al. 1991a, b; Schermer 1993). (7) The latest phase of extensional faults was interrupted by local compression at the end of the Miocene. We would like to thank G. Migiros for the helpful comments on an earlier version of the manuscript, and A. H. F. Robertson for his support and time

spent in providing us with instructive and thorough reviews that greatly improved the final version. We also acknowledge the State Scholarship Foundation of Greece for its financial support to A.V.

References ANGELIER, J. 1979. Determination of the mean principal direction of stresses for a given fault population. Tectonophysics, 56, T17-T26. AUaOUIN, J. 1959. Contribution h l'6tude gdologique de la Gr+ce septentrionale: les confins de l'Epire et de la Thessalie. Annales Gdologiques des Pays Helldnique, 10, 1-483. BARBIERY, R. 1992. Foraminifers of the Eptahori Formation (early Oligocene) of the Mesohellenic basin, northern Greece. Journal of Micropalaeontology, II, 73-84. BRUNN, J. H. 1956. Contribution ~t l'6tude g6ologique du Pinde septentrional et d'une parfie de la Mac6doine occidentale. Annales GOologiques des Pays HellOnique, 7, 1-346. BRUNN, J. H. 1960. Geological Map of Greece, Pentalophos Sheet, scale 1:50 000. IGEY, Athens. BUSBY, C. J. & INGERSOLL, R. V. 1995. Tectonics of Sedimentary Basins. Blackwell Science, Cambridge, MA. CHATZIPETROS, A. 1998. Palaeoseismology and morphotectonics of the Mygdonia, eastern Chalkidiki and Kozani-Grevena active fault systems. PhD thesis, Aristotle University of Thessaloniki. CHATZIPETROS, A., KOKKALAS, S., PAVLIDES, S. KOUKOUVELAS, I. 2006. Palaeoseismic data and their implication for active deformation in Greece. Journal of Geodynamics, 40, 170-188. CROWELL, J. C. 1974a. Sedimentation along the San Andreas fault, California. hi: DOTT, R. H. & SHAVER, R. H. (eds) Modern and Ancient Geosynclinal Sedimentation. Society of Economic Paleontologists and Mineralogists, Special Publications, 19, 292-303. CROWELL, J. C. 1974b. Origin of late Cenozoic basins in southern California. DICKINSON, W. R. (ed.) Tectonics and Sedimentation. Society of Economic Paleontologists and Mineralogists, Special Publications, 22, 190-204. CROWELL,J. C. 1987. Late Cenozoic basins of onshore southern California: complexity is the hallmark of their tectonic history. In: INGERSOLL,R. V. & ERNST, W. R. (eds) Cenozoic Basin Development of Coastal CaliJbrnia ( Rubey Volume VI). PrenticeHall, Englewood Cliffs, N J, 207-241. DESPRAIRIES, A. 1979. l~tude sedimentologique des formations ~ caract6re flysch et molasse, Mac6doine, l~pire (Gr6ce). MOmoires de la SociOtb Gkologique de France, 136, 1-80. DOUTSOS, T., KOUKOUVELAS, I., ZELILIDIS, A. 8r KONTOPOULOS, N. 1994. Intracontinental wedging and post-orogenic collapse in the Mesohellenic trough. Geologische Rundschau, 83, 257-275. FERRII~RE, J., REYNAUD, J.-Y., MIGIROS, G., PROUST, J.-N., BONNEAU,M., PAVLOPOULOS,A. & HOUZE, A. 1998. Initiation d'un bassin transport&

MESOHELLENIC TROUGH, GREECE l'exemple du <<sillon m6so-hell6nique>) au Tertiaire (Grbce). Comptes Rendus de l'AcadOmie des Sciences, Skrie II, 326, 567-574. FERRII~RE,J., REYNAUD,J.-Y., PAVLOPOULOS,A., et al. 2004. Geologic evolution and geodynamic controls of the Tertiary intramontane piggyback MesoHellenic Basin, Greece. Bulletin de la SociktO Gkologique de France, 175, 361-381. JONES, G. & ROBERTSON, A. H. F. 1991. Tectonostratigraphy and evolution of the Mesozoic Pindos ophiolite and related units, northwestern Greece. Journal of the Geological Society, London, 148, 267-288. KARNER, G. D. & WATTS, A. B. 1983. Gravity anomalies and flexure of the lithosphere at mountain ranges. Journal of Geophysical Research, 88, 10449-10477. KILIAS, A., FASOULAS, C., PRINIOTAKIS,M., SFEIKOS, A. & FRISCH, W. 1991a. Deformation and HP-LT metamorphic conditions at the tectonic window of Kranea (W Thessaly, Northern Greece). Zeitschrift der Deutschen Geologischen Gesellschaft, 142, 87-96. KILIAS, A., FRISCH, W., RATSHBACHER,L. & SFEIKOS, A. 1991b. Structural evolution and metamorphism of blueschists, Ampelakia nappe, eastern Thessaly, Greece. Bulletin of the Geological Society of Greece, 25(1), 81-89 (in Greek). KILIAS, A., TRANOS, M., MOUNTRAKIS, D., SHALLO, M., MARTO, A. & TURKU, I. 2001 Geometry and kinematics of deformation in the Albanian orogenic belt during the Tertiary. Journal of Geodynamics, 31, 169-187. KONTOPOULOS, N., FOKIANOU, T., ZELILIDIS, A., ALEXIADIS, C. & RIGAKIS, N. 1999. Hydrocarbon potential of the middle Eocene-middle Miocene Mesohellenic piggyback basin (central Greece): a case study. Marine and Petroleum Geology, 16, 811-824. KOUMANTAKIS, J. 1980. Geological Map of Greece, scale 1:50 000, Panagia Sheet. Institute of Geology and Mineral Exploration, Athens. LISTER, G. S., BANCA, G. & FEENSTRA, A. 1984. Metamorphic core complexes of Cordilleran type in Cyclades, Aegean Sea, Greece. Geology, 12, 221-225. MAVRIDIS, A. & KELEPERTZIS, A. 1985. Geological Map of Greece, scale 1:50 000, Knidi Sheet. Institute of Geology and Mineral Exploration, Athens. MAVRIDIS, A. & MATARANGAS, D. 1979. Geological Map of Greece, scale 1:50 000, Agiofvllon Sheet. Institute of Geology and Mineral Exploration, Athens. MOtrNTRAKIS, D. 1986. The Pelagonian zone in Greece: a polyphase deformed fragment of the Cimmerian continent and its role in the geotectonic evolution of the Eastern Mediterranean. Journal of Geology, 94, 335-347. MOUNTRAKIS, D., KILIAS, A. & ZOUROS, N. 1992. Kinematic analysis and Tertiary evolution of the Pindos-Vourinos, ophiolites (Epirus-Western Macedonia, Greece). Bulletin of the Geological Society of Greece, 28(1), 111-124.

537

MOXON, I. W. & GRAHAM, S. A. 1987. History and controls of subsidence in the Late CretaceousTertiary Great Valley forearc basin, California. Geology, 15, 626-629. ORI, G. G. & ROVERI, M. 1987. Geometries of Gilbert-type deltas and large channels in the Meteora Conglomerate, Meso-Hellenic basin (Oligo-Miocene), Central Greece. Sedimentology, 34, 845-859. PAPANIKOLAOU,D. t~ SIDERIS,CH. 1977. Contribution to the knowledge of Greece molasse. I. Preliminary study at Karditsa's Kanalia region. Annales GOologiques des Pays Helldnique, 28, 387-417. PAPANIKOLAOU, D., LEKKAS, E., MARIOLAKOS, E. & MIRKOU, R. 1988. Contribution on the geodynamic evolution of the Mesohellenic trough. Bulletin of the Geological Society of Greece, 20, 17-36. PAPAZACHOS, B. 8~ PAPAZACHOU, C. 1997. The Earthquakes of Greece. Ziti, Thessaloniki. ROYDEN, L. H. & KARNER, G. D. 1984. Flexure of lithosphere beneath Apennine and Carpathian foredeep basins: evidence for an insufficient topographic load. AAPG Bulletin, 68, 704--712. SAVOYAT, E. 8~;LALECHOS,N. 1969. Geological Map of Greece, scale 1:50 000, Trikala Sheet. Institute of Geology and Mineral Exploration, Athens. SAVOYAT, E. ~; LALECHOS,N. 1972. Geological Map of Greece, scale 1:50 000, Kalambaka Sheet. Institute of Geology and Mineral Exploration, Athens. SAVOYAT, E. 8~ MONOPOLIS, D. 1972. Geological Map of Greece, scale 1:50 000, Grevena Sheet. Institute of Geology & Mineral Exploration, Athens. SAVOYAT, E., MONOPOLIS, D. 8z BIZON, G. 1971a. Geological Map of Greece, scale 1:50 000, Nestorion Sheet. Institute of Geology and Mineral Exploration, Athens. SAVOYAT, E., VIERDIER,A., MONOPOLIS, D. & BIZON, G. 1971b. Geological Map of Greece, scale 1.'50 000, Argos Orestikon Sheet. Institute of Geology and Mineral Exploration, Athens. SCHERMER, E. R. 1993. Geometry and kinematics of continental basement deformation during the Alpine orogeny, Mt. Olympos region, Greece. Journal of Structural Geology, 15, 571-591. SOLIMAN, H. A. ~I; ZYGOGIANNIS,N. 1980. Geological and paleontological studies in the Mesohellenic Basin, Northern Greece. I. Oligocene smaller Foraminifera; II. Eocene smaller Foraminifera. Geological and Geophysical Research, Institute of Geology and Mineral Exploration, Athens, XXII(1) 1-66. TURNER, F. 1953. Nature and dynamic interpretation of deformation lamellae in calcite of three marbles. American Journal of Science, 293, 463-495. VAMVAKA, A., KILIAS, A. & MouYrRArdS, D. 2004. Geometry and structural evolution of the Mesohellenic Trough. A new approach In: CHATZIPETROS, A. & PAVLIDES, S. (eds) 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 1,209-212 (extended abstract).

538

A. VAMVAKA ET AL.

WILSON, J. 1993. The anatomy of Krania basin, northwest Greece. Bulletin of the Geological Society of Greece, 28(1), 361-368. ZELILIDIS, A., KONTOPOULOS, N., AVRAMIDIS, P. & BOUZOS, D. 1997. Late Eocene to early Miocene depositional environments of the Mesohellenic basin, North-Central Greece: implications for hydrocarbon potential. Geologica Balcanica, 27, 45-55. ZELILIDIS, A., PIPER, D. J. W. & KONTOPOULOS, N. 2002. Sedimentation and basin evolution of the Oligocene-Miocene Mesohellenic basin, Greece.

American Association of Petroleum Geologists Bulletin, 86(1), 161-182. ZYGOGIANNIS, N. & M~3LLER, C. 1982. Nannoplankton-Biostratigraphie der terti/iren Mesohellenischen Molasse (Nordwest-Griechenland).

Zeitschrift der Deutschen Geologischen Gesellschaft, 133, 445-455. ZYGOGIANNIS, N. & SIDIROPOULOS,D. 1981. Schwermineralverteilungen und pal/iogeographische Grundziige der terti~iren Molasse in der Mesohellenischen Senke, Nordwest-Griechenland.

Neues Jahrburch fiir Geologie und Paliiontologie, Monatshefte, 100-128.

First results of fission-track thermochronology in the Albanides BARDHYL

M U C E K U 1'2, G E O R G E S H. M A S C L E 1 & A R T A N T A S H K O 2

1Laboratoire de Gkodynamique des Chafnes Alpines ( L G C A U M R 5025 CNRS/UJF/USavoie), Observatoire des Sciences de l'Univers de Grenoble ( OSUG), Universitd Joseph Fourier ( UJF), Maison des Gkosciences, B P 53, 38041 Grenoble Cedex, France (e-mail." Georges. [email protected]) 2polytechnic University o f Tirana, Rruga ElbasanL Tirana, Albania

Abstract: Albania, situated at the boundary between the Dinaric and the Hellenic branchs of the Dinaro-Hellenic fold belt, has experienced a multiphase geodynamic evolution. The internal zones show a Mid-Jurassic episode of deformation characterized by ophiolite obduction, followed by development of a fold-and-thrust belt in the external zones during the Cenozoic. More recently, Albania has experienced a tensional regime. We present apatite and zircon fission-track (AFT and ZFT) measurements for 22 samples, and seven measurements of track-length distributions to elucidate the thermal evolution. AFT ages vary from 10.8 +0.7 Ma to 50.5 ___5.7 Ma. The oldest ages (Eocene) occur in the western Albanides, corresponding to Eocene emplacement of the internal zones over the external ones. Neogene ages in the eastern Albanides suggest rapid exhumation, which we relate to an extensional regime. The ZFT ages show that the internal Albanides did not reach temperatures > 200 ~ during the Cenozoic.

Albania occupies a critical position within the Dinaro-Hellenic Alpine fold belt, at the boundary between the Dinarides and HeUenides (Fig. 1). The Dinaro-Hellenic orogen is characterized by three fundamental components: a western (external) fold-and-thrust belt, a central belt characterized by ophiolitic nappes, and an eastern (internal) complex (Aubouin et al. 1970; Memo & Aliaj 2000; Robertson & Shallo 2000). Some key points of the geodynamic evolution of the Albanides remain controversial, partly because of limited well-constrained geochronological data, mainly concerning Mid-Jurassic ophiolite obduction, which was dated using the 4~ method on the amphibolitic metamorphic sole of the ophiolitic nappe (Dimo 1997; Dimo-Lahitte et al. 2001). Apatite and zircon fission-track (AFT, ZFT) thermochronology is an invaluable tool to decipher the lowtemperature history of orogenic belts (Gallagher et al. 1998). Here, we report 18 A F T ages and four ZFT ages, together with seven measurements of track-length and track-width distributions to help determine the low-temperature history of the Albanides.

Geological setting of Albania Present-day structure o f Albania Geological and gravimetric data, combined with velocity determination for P and S waves,

indicate a thickening of the Albanian crust (Fig. 2a and b), from a normal thickness of about 30 km in western Albania, to 45-50 km in the eastern part, near the Macedonian and Greek borders (Frasheri et al. 1996; Papazachos et al. 2002; Cavazza et al. 2004). Seismological data (Aliaj 1991; Muqo 1994; Frasheri et al. 1996; Louvari et al. 2001) characterize a gently eastdipping slab with compressional mechanisms for up to 50 km located beneath the AlbaniaMacedonia border (Fig. 2c). Eastern Albanian is characterized by extensional mechanisms down to 15 km (Fig. 2c). Tomographic imagery ONortel & Spakman 1992, 2000; Cavazza et al. 2004) shows a cold lithospheric slab dipping gently eastward below the Dinaro-Hellenic belt (Fig. 2c); this represents the subducting Apulian lithosphere. Modern stress field data in the Dinaric belt (Mariucci & Miller 2003; Cavazza et al. 2004), indicate a more or less N E - S W oriented compressional stress field in the external zones and a tensional one in the internal areas. Global motion vectors (DeMets et al. 1990), as well as more recent kinematic models (Altamimi et al. 2002; Sella et al. 2002), are compatible with the existence of a Dinaric compressive boundary. Published global positioning system (GPS) data for the Dinaric and northern Hellenic areas are sparse (Khale et al. 2000; McClusky et al. 2000; Anzidei et al. 2001; Bertran 2003; Hollenstein et al. 2003), but show, in a N o r t h European fixed frame, a NE-oriented displacement of the

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 539-556. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

540

B. M U C E K U E T AL.

Fig. 1. Setting of the Albanides in the Mediterranean and simplified geological map of Albania. After ISPGJ-IGJN (1982, 1985, 2003); the cross-sections 2A-C and 2B are shown in Figure 2.

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES

541

Fig. 2. (a) General cross-section of the Albanides (modified after Collaku et al. 1990). (b) Geodynamic section of the Helleno-Dinaric belt at the latitude of central Albania (modified after Transmed 2004; Carazza et al. 2004). I, Ionian; KG, Kruja-Gavrovo; KP, Krasta-Pindos; PK, Korabi-Pelagonian; S, Sazani-Preapulian; V, Vardar; f~, Mirdita. (e) The subducting Apulian lithosphere from tomographic imagery and seismicity (data from Mugo 1994; Frasheri et al. 1996; Wortel & Spakman, 2000; Louvari et al. 2001; Papazachos et al. 2002).

external Dinaric units (Fig. 1) at a velocity of 5 mm a -~, whereas the internal Dinaric units move in the same direction but slightly faster, in good agreement with the existing tensional regime of both areas. For example, at the Ohrid station (Macedonia), the displacement

is eastward, at a velocity of 2 mm a -1 (Fig. 1). Therefore, all the present-day data suggest the existence of a compressional regime in western Albania, related to the subduction of the Apulian lithosphere and a tensional regime in eastern Albania.

542

B. MUCEKU E T AL.

G e o l o g i c a l subdivision o f A l b a n i a The external fold-and-thrust belt. This covers (Figs 1 and 2) about half of the surface of Albania. The thrust belt is located west of a line joining Skodra to Elbasani and to Permeti, near the Greek border, and reappears in eastern Albania as indicated by the Peskopi tectonic window (Collaku et al. 1990). The westernmost unit (Sazani) is characterized by a neritic platform succession (of Late Triassic to Oligocene age) and a foreland complex (of Early Miocene to Pliocene age; mainly redeposited carbonate facies), deformed into large ramp anticlines with westward displacement (Frasheri et al. 1996); this unit is correlated with the Apulian carbonate platform (Meqo & Aliaj 2000; Robertson & Shallo 2000; Kilias et al. 2001; Cavazza et al. 2004). The Ionian zone constitutes a thin-skinned fold-and-thust belt, overthrusting the Apulian unit aided by an evaporitic basal d6collement (ISPGJ-IGJN 1982, 1985, 2003; Frasheri et al. 1996; Kilias et al. 2001; Cavazza et al. 2004). The stratigraphical section consists of an evaporitic Permo-Triassic sole, an Upper Triassic-Middle Liassic carbonate platform, a pelagic basinal sequence (of Dogger-Late Eocene age), and an Oligocene-Miocene foreland complex. This unit is also mainly redeposited carbonate facies, showing westward progradation with progressive unconformities; thrusting occurred during the Messinian. According to Collaku et al. (1990), the evaporitic diapirs of the Peskopia tectonic window represent the eastern prolongation of the Ionian Zone, which reappears some 60 km east of the Kruja thrust. The Ionian Zone is overthrust by the Kruja unit, corresponding to the Greek Gavrovo and Dalmatian zones (Me9o & Aliaj 2000; Robertson & Shallo 2000). This unit is characterized by Mid-Upper Cretaceous platform carbonates, Upper Cretaceous-Palaeocene pelagic facies, and a thick (up to 5 km) Upper Eocene-Miocene turbiditic sequence. Thrust sheets of a similar turbiditic sequence are observed in the Peshkopi tectonic window (Collaku et al. 1990; ISPGJ-IGJN 2003). The Kruja Zone is itself overthrust by the Pindos nappe (Krasta Zone) represented by Cretaceous turbiditic sandstones and mudstones, followed by Upper Cretaceous pelagic facies (scaglia), and overlain by Maastrichtian-Eocene turbidites (Pindos flysch) (Memo & Aliaj 2000; Robertson & Shallo 2000). In northern Albania, the Maastrichtian-Eocene flysch sequence overlies a thin pelagic radiolarite-siliceous-carbonate sequence of Mid-Triassic-Late Cretaceous age (Cukali Zone, Me9o & Aliaj 2000).

The central belt. This shows a very complex structural arrangement (Figs 1 and 2). North of the SW-NE Shkodra-Pe6 line, the Albanian Alps represent the southern continuation of the Dinaric nappe system (Mego & Aliaj 2000); the lowermost nappe (Malesia e Madhe; High Karst) shows a Permian-Middle Triassic terrigenous formation (Verrucano), a thick Middle TriassicCretaceous platform carbonate sequence, and Paleocene-Lower Eocene flysch. The second unit (Valbona; pre-Karst) is similar up to the Upper Jurassic sequence, followed by a mixed turbiditic pelagic Kimmeridgian-Cretaceous sequence and Maastrichtian flysch. The third unit (Vermoshi; Bosnian) shows a strongly folded TithonianValanginian flysch sequence. South of the Shkodra-Ped line, the Mirdita Zone is characterized by a huge ophiolitic nappe (Mirdita ophiolite), up to 13 km thick in the Tropoja massif (Llangora & Bushati 1990), which represents the largest European ophiolitic complex. Between the ophiolitic sequence and the Krasta (Pindos) Zone there exists a strongly deformed tectonic complex, variously interpreted (and named as) the peripheral complex by Robertson & Shallo (2000); or the Hajmeli, Querreti-Miliska and Gjallica unit of Kodra et al. (1993) and Meqo & Aliaj (2000). This tectonic complex may be subdivided into three main units. The lowermost one is characterized by a thick sequence of Triassic platform carbonates (Hajmeli in western Mirdita and Gjallica in eastern Mirdita following Kodra et al. 1993). In our opinion these units belong to the Pelagonian Korabi Zone. The Triassic platform is overthrust by a Permo-Triassic pelagic and volcanic complex, termed the Rubik complex, which is well dated in various places by microfauna (Kodra et al. 1993; Mego & Aliaj 2000). This unit appears not only on both sides of the ophiolitic nappes (Rubik and Mirake on the western side; Gjegjan on the eastern one), but also in several tectonic windows below the ophiolitic pile (Fushe Arresi, Blinishti-Reps). Copper mineralization is associated with Triassic alkali lavas (Gjegjan, Rubik). The unit is strongly tectonized, forming numerous thin thrust sheets. The Rubik complex is itself overthust by a thin metamorphic unit, which constitutes the amphibolitic metamorphic sole of the ophiolite nappe, dated as Mid-Jurassic in age using the 4~ method (Dimo 1997; Dimo-Lahitte et al. 2001). The Mirdita ophiolitic complex is itself subdivided into two belts, the western (WOM) and the eastern (EOM) belts (Shallo et al. 1987; Beccaluva et al. 1994; Tashko 1996), with mainly tectonic relationships. However, B6bien et al. (1998) reported a possible continuity between the two belts. The WOM is characterized

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES by a lherzolitic mantle sequence, followed by a thin gabbroic troctolitic sequence and pillow lavas of normal mid-ocean ridge basalt (NMORB) type (Beccaluva et al. 1994; Tashko 1996; Robertson & Shallo 2000, and references therein). Accociated pelagic sediments have yielded a Bathonian age (Marcucci et al. 1994). The EOM is thicker and is characterized by a harzburgitic mantle sequence, well-developed gabbronoritic plutonic sequence, a dyke complex, and island are tholeiite OAT) to boninite extrusive rocks (Shallo et al. 1995; Tashko 1996; B6bien et al. 1998; Robertson & Shallo 2000, and references therein). Pelagic sediments have yielded a Late Bathonian to Mid-Callovian age (Marcucci et al. 1994). As shown by dating their tectonic sole, the Mirdita ophiolitic nappes were emplaced during the Mid-Jurassic (Dimo 1997; Dimo-Lahitte et al. 2001). After tectonic emplacement, the ophiolites underwent erosion, as shown in the EOM by intense lateritic alteration, and in the WOM by the existence of a regional unconformity below the post-obduction sediments (ISPGJ-IGJN 1982, 1985, 2003). The post-obduction sedimentary cover includes a succession of 'chaotic' sediments that rework the internal units including these beneath ophiolite. Ophiolitic clasts are, however very uncommon, probably as a consequence of the prevalent climatic conditions, which caused lateritization of the EOM. The chaotic sequence is followed by turbidites of Tithonian-Early Cretaceous age (ISPGJ-IGJN 1982, 1985, 2003), and then by shallow-water carbonates of HauterivianBarremian to Late Cretaceous age (ISPGJ-IGJN 1982; Peza 1985; Shallo 1990). The eastern internal complex. North of the Shkodra-Pe6 line (Figs 1 and 2), this corresponds to the Gashi Zone, characterized by a SiluroDevonian terrigenous formation, intruded by the large Trokuzi granodioritic batholith, and followed by a succession of dacitic and andesitic rocks, with limestone intercalations (of Late Permian to Early Triassic age), and ending with a conglomeratic sequence (Verrucano) (Mego 1991; Meqo & Aliaj 2000). This Unit is correlated with the Durmitor Zone of Montenegro. South of the Shkodra-Pe6 line, the internal complex corresponds to the Korabi Zone, which is correlated with the Golia Zone in the Dinarides, or the Drina Zone of former Yugoslavia, and the Pelagonian Zone in the Hellenides (Robertson & Shallo 2000). The section, strongly deformed in several tectonic slices, shows a succession of quartzites, shales and minor carbonates, with some volcanic intercalations of Ordovician to

543

Devonian age (Melo 1970; Mego 1988, 1991; Memo & Aliaj, 2000). These sequences underwent low-grade metamorphism and were intruded by monzosyenites and lamprophyres, dated by the K/Ar method at 373 + 50.7 Ma, 294+47.04 Ma and 241 +28.9 Ma, respectively (Shallo 1992). A weakly metamorphosed sequence of sandstones and conglomerates, with typical Verrucano facies, unconformably overlies the Palaeozoic succession. This passes upwards into a calcalkaline volcano-sedimentary unit of Early-MidTriassic age, and then into platform carbonates of Mid-Triassic to Early Jurassic age. This platform sequence is identical to the Gjallica sequences, which form the lowermost tectonic slice below the EOM; therefore, following Kilias et al. (2001), we interpret the Gjallica sequence as the uppermost tectonic slice of the Korabi Zone, and adopt the same interpretation for the Hajmeli sequence, situated below the WOM. The platform sequence is overlain by a pelagic sequence of Late Liassic-Late Jurassic age (Shallo, 1992). An erosion surface truncates the Mesozoic sequence with local bauxitic pockets. The erosion surface is transgressed by a chaotic and turbiditic sequence of Tithonian-Early Cretaceous age (Shallo 1992) that reworks the ophiolites and their tectonic substratum (Rubik unit). The section continues with shallowwater carbonates of Barremian to Albian age, which are locally transgressed by Palaeogene terrigenous turbidites. The Albano-Thessalian depression. The N N W SSE-oriented Albano-Thessalian depression (Fig. 1) crosscuts both the Korabi and the Mirdita zones. It shows a shallow-marine and continental clastic sequence of late Eocene to Tortonian age. The basin represents the northern continuation of the Meso-Hellenic Trough of northern Greece, interpreted as a piggyback basin developed behind the compressional front of the external fold-and-thrust belt (Ferri6re et al. 2004). The Neogene-Quaternary graben. A northsouth-oriented Neogene-Quaternary graben system crosscuts the entire regional structure (Korabi, Mirdita and Albano-Thessalian basin) from Korca to Progradec lake and continues into Macedonia (Fig. 1). The fault system has been activated several times, involving Late Tortonian SE-NW extension, Early Pliocene ENE-WSW compression, Late Pliocene east-west extension, early Pleistocene east-west transpression and SE-NW to east-west Quaternary extension (Tagari et al. 1993).

544

B. MUCEKU E T AL.

G e o d y n a m i c evolution o f A l b a n i a

Although there is a general consensus as to a westward transported fold-and-thrust belt, a controversy exists concerning the deep structure of the ophiolite, which is considered either as a far-travelled nappe originating in the Vardar Zone (Collaku et al. 1990), as a locally rooted zone reversely faulted on both sides (Kodra et al. 1993), or as a twice-deformed moderately displaced unit (Robertson & Shallo, 2000). For Collaku et al. (1990), the existence of the Peshkopia tectonic window indicates the allochthony of the ophiolite. For Kodra et al. (1993), the thickness of the Tropoja ophiolite is not compatible with an allochthonous massif, and there exist kinematic indicators of reverse faulting on both sides of the ophiolite, west-directed on the western side and east-directed on the eastern side. The model of Robertson & Shallo (2000), is based on the petrological and geochemical differences between the WOM, seen as a MOR-type ophiolite, and the EOM, interpreted as a supra-subduction-type ophiolite. Robertson & Shallo inferred two-phase emplacement history in which eastward-dipping Jurassic subduction was followed by westward transport related to Early Tertiary collision. In our opinion, the structure of Albania has resulted from several structural episodes. The first well-characterized one is the obduction of the Mirdita ophiolite, well dated as Mid-Jurassic, either by geochronology (Dimo 1997; Dimo-Lahitte et al. 2001) or by the sedimentary evolution of the internal units (ISPGJI G J N 1982, 1985, 2003; Shallo 1992; Kodra et al. 1993; Robertson & Shallo 2000). The ophiolites were thrust over the Rubik complex, locally metamorphosed (metamorphic sole), and emplaced over the Korabi sequences. Some kinematic data from the ophiolite (Tashko et al. 1996; Robertson & Shallo 2000), or the metamorphic sole (Dimo 1997; Dimo-Lahitte et al. 2001) suggest a northeastward transport direction (in present-day orientation). However, this model does not explain the existence of calc-alkaline volcanic rocks in the Korabi and Gashi Triassic units, which possibly related to an earlier tectonic regime characterized by northward subduction. A second well-defined major tectonic episode resulted in the construction of the fold-and-thrust belt, characterized by west-southwestward transport, beginning in Eocene time with deformation of the Krasta (Pindos) Zone, then progressively affecting the more external domains. In our opinion, the previously structured internal complex (Mirdita ophiolite, Rubik complex and Korabi unit) was passively transported on top of the

Krasta (Pindos) nappes at the start of this tectonic episode, resulting in the complete uprooting of the ophiolites (Fig. 2a and b).

Fission track data Fission-track thermochronology

Apatite and zircon fission-track (AFT, ZFT) thermochronology is widely used for reconstruction of low-temperature thermal histories of upper crustal rocks. This method allows one to estimate the temperature history and long-term denudation rates in orogenic mountain belts, rifted margins and more stable continental areas (e.g. Green et al. 1989; Wagner & Van den Haute 1992; Gallagher et al. 1994; Fitzgerald et al. 1995; Carter 1999; Zarki-Jakni et al. 2004). The apatite partial annealing zone (PAZ) is considered to extend from 120 to 60 ~ (Gleadow & Fitzgerald 1987; Green et al. 1986, 1989). Confined tracks formed below 60 ~ are characterized by a mean track length (MTL) of c. 15 ~tm and a standard deviation (SD) of their distribution < 1 ~tm. Within the PAZ, tracks shorten at highly temperature-dependent rates. The relationship between track shortening, time and temperature has been quantified by laboratory experiments (e.g. Laslett et al. 1987; Carlson et al. 1999). Therefore, the track-length distribution within an apatite sample can be inverted to determine its thermal history experienced (e.g. Gallagher 1998; Ketcham et al. 1999). However, the annealing kinetics is dependent on apatite chemistry; an efficient measure of annealing kinetics is obtained by measuring the width of fission-track etch pits parallel to the c-axis (Dp,r; Carlson et al. 1999; Barbarand et al. 2003). The PAZ of zircon is in the range of 200250 ~ and the temperature of 90% track retention is c. 240 ~ in most cases (Brandon & Vance 1992; Brandon et al. 1998). Annealing of fission tracks in zircon during reheating is partially a function of alpha damage in the zircon. Highly damaged zircons will anneal at lower temperatures, whereas more pristine crystals may anneal only at temperatures of > 250 ~ depending on heating time. Sampling and analytical procedures

Lithologies suitable for FT thermochronology are present in the external fold-and-thrust belt (clastic sequences of the foreland complex and flysch sequences), in the central complex (basement of the Rubik nappes, metamorphic sole and Mirdita ophiolite), in the internal complex (magmatic intrusions and Verrucano), and in

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES the clastic sequences of the Albano-Thessalian depression. After initial wide-mesh sampling, we concentrated our attention on the central belt and the internal complex. Twenty-eight samples were collected from different magmatic bodies and terrigenous formations of the Korabi, Rubik and Gashi zones, 19 from the amphibolitic metamorphic sole and 15 from gabbroic and plagiogranitic bodies within the Mirdita ophiolite. Twenty-two samples were collected from clastic layers of the external flysch units and the foreland complex and from the Albano-Thessalian depression. Apatite and zircon were separated using standard magnetic and heavy liquid separation techniques. After separation, apatites were mounted in epoxy, polished and etched in 5M HNO3 solution at 20 ~ for 20 s. All samples were dated by the external detector method, using a zeta calibration factor for Fish Canyon Tuff (FCT) and Durango Tuff age standards (Hurford 1990). Samples were irradiated at the well-thermalized ORPHEE facility of the Centre d'Etudes Nucldaires in Saclay, France, with a nominal fluence of 5 x 1015 neutrons cm -2. Neutron fluence was monitored using CN5 and NBS962 dosimeter glasses. For calibration of confined track length measurements, we measured confined track lengths in apatite from the Durango and FCT age standards. We obtained an MTL of 14.2 and 14.4 gm for Durango and FCT, respectively, with standard deviations (SD) of 1.0 and 1.1 gm, respectively (see Fig. 6).

Results We have so far dated 22 samples from the inner Albanides. AFT and ZFT data are summarized in Tables 1 and 2. All AFT ages are quoted as central ages (Hurford 1990) with_+ lcy uncertainties throughout and range from 10.8_+0.7 to 50.5 +_5.7 Ma. All samples show a very low age dispersion (D < 6%, P(Z 2) _>90%), suggesting that chemical heterogeneity of the apatite is not a problem in the crystalline rocks that we sampled (Fig. 3). The MTLs for our samples vary between 10.2_+0.3 gm (AM12-02) and 13.0_+0.2gm (AM13-02), with standard deviations (SD) between 1.3 and 3.0 gm. The youngest AFT ages ( < 20 Ma) are along the eastern border of the Mirdita Zone and in the Korabi Zone. Amphibolites, from the base of the ophiolites in the eastern Mirdita Zone, and micaschists from Gjegjani have AFT ages between 20 and 15 Ma (Table 1 and Fig. 3). In the Korabi Zone we dated monzonite, lamprophyre

545

granite and Palaeozoic sandstones, and found AFT ages between 17 + 1 and 11 + 1 Ma. We also analysed one granitoid from the Trokuzi massif in the Gashi Zone, which yielded an AFT age of c. 40 Ma. This age is close to the range of ages between 40 and 50 Ma that we observed in the western Mirdita Zone from amphibolites at the base of the ophiolitic nappe, and from plagiogranites in the western part of the ophiolites, and granite in the Rubik nappe. All AFT ages are significantly younger than the 4~ ages of 165-175 Ma determined from the metamorphic base and some intrusions into the ophiolites (Dimo-Lahitte et al. 2001). We also dated four zircon samples from different locations: the Trokuzi massif (Gashi Zone); one granite in the Rubik area, and two samples from the Korabi Zone. The zircon FT ages (Table 2) range between 100 and 174 Ma, a considerably older age spectrum than the AFT ages.

Discussion Our AFT ages show a clear regional trend. They are very similar from north to south but change significantly from west to east in the inner Albanides (Figs 3 and 4). The samples with young AFT ages ( < 15 Ma) are located in the eastern part of Albanides (Korabi and eastern Mirdita). They show an interesting age-elevation relationship; the ages are constant and independent of elevation (Fig. 5). Such an age-elevation trend is generally considered as being developed during a period of very fast tectonic denudation (exhumation) or erosion. These ages are significantly younger than the results (35-40 Ma) obtained in the Pelagonian domain of Macedonia and northern Greece, east of Korabi (Most et al. 2001); this suggests that relatively recent exhumation was localized near the western boundary of the Korabi Zone. The MTL of the young samples in the eastern Albanides are relatively short, around 10-12 ~tm. One sample from the Peladhi granite (AM 12-02), situated on the western border of the Korabi zone, shows a bimodal track-length distribution with an MTL of 10.2 _+0.31xm (SD=3.0~tm) (Fig. 6). Other samples from the Korabi Zone do not show bimodal distributions, but do have similarly short MTL and wide track-length distributions. This suggests that all of the Korabi samples were maintained for a long time at a temperature of 90_+10 ~ within a PAZ, and that fast exhumation, as suggested by the ageelevation relation of the samples, affected the region recently.

546

ETAL.

B. M U C E K U

,.0

9

c~

+1 +I +1 +I +1 +1 +1 +1 +1 +1 +1 +1 +1

+1 +1 +1 +1

+1 +1 +1 +1 +1

+1 +1

~ 0 0 0 0

~AAAA

A~A~

t,r

~

t"-~ .-~

tr

P~

z

~

0

~'~.Zn

-~'.~ o ~ ~'~ .~ .,, ~

.,,~

,.~

?

N o ~ g ~

~

0

~

.< 0

8 9..~ .,'-~ ~

~.~ ~ ~ 0 ~ 0

~M~M

I

<<<

<

<<<<

I

I

<<<<

I

~~o

~

FISSION-TRACK T H E R M O C H R O N O L O G Y , A L B A N I D E S

+1+1+1

+1

/X

/x

t'q

t"q

' , ~ 9 ~,O. t " -.

t "~

O'~

,,-,~

~

,---~

~,-~

,'-~ t ~

Our samples collected from west of the Mirdita Zone, with AFT ages of 40 Ma or older, do not show any significant age-elevation relationship (Fig. 5). There, we were able to measure confined track lengths on one sample (AM 13-02), and found an MTL of 13.0 ~tm with SD 1.3 ~tm (Fig. 6). The inferred cooling history appears to be different. Our zircon FT results indicate that the internal Albanides have in general not reached a temperature high enough ( > 200 ~ to reset zircon. The zircon FT ages of the granite from the Rubik area (174 Ma) are compatible with K/Ar ages of 165 to 175_+6 Ma for the same rock, as reported by Castorina et al. (1995). This suggests that the maximum temperatures in the Albanides remained well below 200 ~ since the Mesozoic, and particularly during the Eocene deformation (see Fig. 8).

+l

t"-

t'r

t'-

Thermal modelling

OO

O O t ~

t'~

ott'5~ t - -

e.i

c,i r

r

r

tt~

.<

t-q

,-2

r ,.<3

[tt~

b~ ~4

t'q

tr~

547

To explore in more detail the cooling history, we used the multi-kinetic model of AFTsolve of Ketcham et al. (1999, 2000). The chemical composition of apatite may vary strongly, even for crystalline basement samples from the same lithology (Crowley et al. 1991; O'Sullivan & Parrish 1995). As a control of chemical variability we measured Dpar (Table 1). Opar is the mean fission-track etch pit diameter parallel to the crystallographic c-axis for each apatite grain, and is specified in units of microns (e.g. Carlson et al. 1999; Donelick et al. 1999). We also measured the orientation of confined track length to the crystallographic c-axis for each track length, and corrected for this in the thermal history modelling (see Donelick et al. 1999). To test this modelling approach, we first used grain-age, track-length (MTL 14.4 ~tm) and Dpar (2.1 ~tm) measurements of FCT apatite. The aim was to compare the modelling results with a sample with a known cooling history. The results are shown in Figure 7. The thermal model of FCT shows very fast cooling (Fig. 7d) through the PAZ between 27 and 28 Ma; this corresponds well to the known thermal history of these volcanic rocks. Encouraged by these results, we modelled two samples from the Korabi Zone (Fig. 7a and b), and one sample of amphibolite from the western Mirdita Zone (Fig. 7c). For these samples we assumed that they cooled from temperatures of c. 120 ~ since the last heating event. To translate these cooling histories into exhumation rates, we needed to have an estimate of the geothermal gradient. Frasheri et al. (1996)

548

B. MUCEKU E T A L . AM30-02

AM1-00

50.5 + 5.7 Ma

11.3 _ 0.6 Ma

17

o--:'-ts,

7o

27 crystals++

60 12 50 45

-1

10

40 7O 60 5o

6040 30

4O

20

15

P,el. error [%]

Rel. error [%]

AM13-02

AM9-00

41.4 + 4.2 Ma

10.8 + 0.7 Ma 20 crystals

4-

~~.~

+ "k~

12

++

35 30

* ii

80 70 60

i

!

i

I

70 50 40 3o

50

Rel. error [%]

Rel. error [%]

AM 18-03

AM8-00

40.2 + 5.3 Ma

11.9 + 0.7 Ma

19

10

70 ' 60

c ~ s t a ~

2 1

'

/

........

,+

,

12

-1 -2

, /

70 60 50 40 Rel. error [%]

45 40 35

0

40 30 20 15 Rel. error [%]

Fig. 3. Apatite fission-track (AFT) ages: radial plots of some representative samples from the western Mirdita and Korabi zones.

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES

549

Fig. 4. Apatite fission-track (AFT) and zircon fission-track (ZFT) ages for the Albanides. reported two different geothermal gradients: 15 ~ -~ for the external Albanides and 20-30 ~ km -~ for the internal Albanides. We, therefore, believe 25 ~ -~ is a reasonable approximation for modelling the pre-fast exhumation cooling history of the Korabi Zone samples.

At about 30 Ma, sample AM12-02 was near or even at the surface (Fig. 7b). Subsequently the Korabi Zone was buried to about 4 km depth (assuming a palaeo-geothermal gradient of 25 ~ km -1 and 10 ~ surface temperature) to reach temperatures of 110-120 ~ allowing FT annealing. In contrast, sample AM8-00 experienced

550

B. MUCEKU E T A L . 2500

Korabi Zone

W Mirdita

2000 E

1500

m

c.O m

1000

I'--

~

I

uJ I

9

I

500 i

0

10

i

i

i

i

i

i

i

20

30

40

50

60

70

80

AFI a g e [Ma)

Fig. 5. Relationship between AFT age and sample elevation. slow cooling from about 16 to 15 Ma (Fig. 7a), at an average rate of 5 ~ Ma -1, suggesting rapid exhumation after 2.3 Ma. Comparing our result from multi-kinetic modelling of sample AM8-00 (Ketcham et aL 1999) with the results we obtained using monokinetic modelling (Laslett et al. 1987) for four other monzonite and lamprophyre samples of the Korabi Zone (Muceku et al. 2003), we then found a similar evolution. We believe that this is caused by the similarity of the chemical composition of these samples to the Durango standard, as we measured Dpa r values of 1.7 pm and 1.8 gm, respectively, in these samples. The oldest AFT ages (40-50 Ma) are situated in the western part of the Mirdita Zone. They document an Eocene cooling event that corresponds well in time to the Eocene emplacement of the internal zones onto the external zones. Nowadays, the region has a geothermal gradient of 15-20 ~ km -~ (Frasheri et al. 1996). However, for the Cenozoic deformation history we assume that the palaeo-geothermal gradient was probably between 20 and 25 ~ at 40-50 Ma. Therefore, the 40-50 Ma AFT cooling ages of the western Mirdita rocks suggest that these rocks were not buried deeper than c. 4-5 km since that time (Fig. 8). The multi-kinetic thermal modelling of sample AM 13-02, even if not very well constrained by a sufficient number of tracklength measurements, suggests a fast cooling since 49 Ma (Fig 7c).

Geodynamic interpretation Figure 8 is a tentative synthesis based on our FT data and the small amout of reliable published Albanian geochronological data. First, our

zircon FT results indicate that maximum temperatures in the central and eastern Albanides stayed well below 200 ~ since Mesozoic time; in particular, there is no indication of deep burial despite the Eocene deformation, supporting the idea that the tectonic pile created during the Jurassic ophiolitic obduction (ophiolitic nappemetamorphic sole-Rubiku nappe-Korabi Zone) was more or less passively transported, with only minor reworking of some tectonic contacts. Similarly, deformation in the Albanian Alps took place under upper crustal conditions. Second, the AFT results document two very different vertical evolutions. In the western part of the belt (western Mirdita and Albanian Alps) the rocks were uplifted above 4-2.5 km depth during the Early to Mid-Eocene and possibly exposed at the surface before the Oligocene. The uplift rate was of the order of 0.1-0.2 mm a -1. For this area the chronology agrees well with the Mid-Eocene age for the uppermost layers of the Pindos flysch, representing the tectonic substratum just below the Mirdita nappes. Therefore, we relate the uplift of the western area to the crustal thickening that occurred as a consequence of the first step of the fold-and-thrust belt construction. The eastern part of the belt shows a very different evolution. The rocks underwent slow cooling during Mid-Late Miocene time, reaching a 4-2.5 km depth only in the Late Pliocene, at an uplift rate of the order of 0. I mm a-L Since the Late Pliocene the uplift rate has accelerated about 12 times. The first episode, well documented in sample 12-02, may correlate with the emplacement of the Albano-Thessalian inferred piggyback basin at the rear of the propagating fold-and-thrust belt. Therefore, we consider this episode to be a consequence of the crustal

FISSION-TRACK T H E R M O C H R O N O L O G Y , ALBANIDES

Fig. 6. Distribution of A F T ages and representative horizontal confined track-length distributions for some selected samples.

551

552

B. M U C E K U E T A L .

Fig. 7. Time-temperature histories of selected samples. Time-temperature histories were calculated by the inversion of track-length distribution (using AFTsolve software). The shaded areas on the time-temperature plots represent solutions that statistically fit the observed data (sample age and track distribution); the continuous black line is the overall 'best fit' solution for the sample. The plots at the right show the observed track-length distribution as a histogram and the modelled distribution as a continuous line. The 'K-S test' is the Kolmogorov-Smirnov test; the age 'GOF' is the goodness of fit between the age data and age predicted by the model. FCT is the Fish Canyon Tuff standard; N, number of measured track lengths.

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES

553

Fig. 8. Thermal evolution of Albanides: tentative synthesis based on our AFT and ZFT results and ages in literature. thickening; the preliminary results of Most et al. (2001) for the internal Pelagonian domain suggest that the whole area underwent general uplift at this time. The Late Pliocene episode corresponds well in time to the evolution of the Neogene-Quaternary graben system, and specifically the Late Pliocene east-west extension event ofTagari et al. (1993). We propose that the recent acceleration of uplift at the eastern boundary of

the Mirdita belt can be related to a recent tensional regime of crustal thinning, as documented by the formation of the graben system and the tensional focal mechanisms in the upper crust.

Conclusions Our A F T data provide clear evidence for differential cooling and erosional denudation of the

554

B. MUCEKU E T A L .

internal Albanides. Near their frontal thrust, the internal Albanides record cooling and denudation during Late Eocene-Early Oligocene times, contemporaneously with their tectonic emplacement onto the external fold-and-thrust belt. In this area, the cooling and denudation are clearly related to isostatic uplift as a consequence of crustal thickening in relation to subduction of Apulian lithosphere. The internal part of the internal Albanides records a more complex cooling evolution: a first stage of early Late Eocene-Early Oligocene cooling was followed by Early Neogene burial, also related to a crustal thickening regime; a second stage of Late Neogene cooling, characterized by recent acceleration, can be related to a still-active period of crustal thinning. Thinning is a c o m m o n tendency for over-thickened crust. In the present case this could either be a local effect or a northward extension of the regional Aegean thinning regime. Our zircon fission track results indicate that the at present exposed internal Albanides did not reach high temperatures during their Tertiary deformation. This study was a part of a NATO-supported programme (Science for Peace). Thanks go to A. Kodra (Geological Survey of Albania) for providing information and some samples; to M. Bernet (LGCA Grenoble) for discussion and improving the manuscript; to V. Gardien (G6odynamique, Lyon) and P. Van der Beck (LGCA Grenoble) for discussions; and to E. Labrin (LGCA Grenoble) for help with some measurements and discussions. M. Brunel (Montpellier), A. H. F. Robertson (Edinburgh) and P. Verg61y(Orsay) provided very helpful suggestions to improve the manuscript. B. Muceku was supported by a grant from the French Ministry of Foreign Affairs.

References ALIAJ, S. 1991. Neotectonic structure of Albania. Albanian Journal on Natural and Technological Science, 4, 79-98. ALTAM~NL Z., SILLARD, P. & BOUCHER, C. 2002. ITRF2000: a new release of the International Terrestrial Reference Frame for Earth science applications. Journal of Geophysical Resesearch, 107, 2214, doi: 10.1029/2001JB000561. ANZIDEI, M., BALDI, P., CASULA, G., et aL 2001. Insights into present-day crustal motion in the central Mediterranean area from GPS surveys. Geophysical Journal International, 146, 98-110. AUBOUIN, J., BONNEAU, M., CELET, P., et al. 1970. Contribution h la g6ologie des Hellenides: le Gavrovo, le Pinde et la zone ophiolitique Subp61agonienne. Annales de la Socidtk GOologique du Nord, 90, 277-306. BARBARAND,J., CARTER, A., WOOD, I. & HURFORD, A. J. 2003. Compositional and structural control

of fission track annealing in apatite. Chemical Geology, 198, 107-137. BgBIEN,J., SHALLO,M., MANIKA,K. & GEGA,D. 1998. The Shebenik massif (Albania): a link between MOR and SSZ-type ophiolites? Ofioliti, 23, 7-15. BECCALUVA,L., COLTORTI, M., PREMTI, I., SACCANI, E., S~ENA, F. & ZEDA, O. 1994. Mid-ocean ridge and supra-subduction affinities in ophiolitic belts from Albania. Ofioliti, 19, 77-96. BERTRAN, A. 2003. Quantification de la dkformation actuelle en Mkditerran& et dans les Alpes occidentales par mesures GPS. M6moire DEA, Universit6 de Chamb6ry. BRANDON, M. T. & VANCE, J. A. 1992. Fission track ages of detrital zircons: implications for the tectonic evolution of the Cenozoic Olympic subduction complex. American Journal of Science, 292, 565636. BRANDON,M. T., RODEN-TICE, M. R. & CARVER,J. I. 1998. Late Cenozoic exhumation of the Cascadia accretionary wedge in the Olympic Mountains, northwest Washington State. Geological Society of America Bulletin, 100, 985-1009. CARLSON, W. D., DONELICK,R. A. & KETCHAM,R. A. 1999. Variability of apatite fission track annealing kinetics I: Experimental results. American Mineralogist, 84, 1213-1223. CARTER, A. 1999. Present status and future avenues of source region discrimination and characterisation using fission-track analysis. Sedimentary Geology, 124, 3145. CASTORINA, F., GARBARINO,C., MASI, U. & BEQIRAJ, A. 1995. The granitic rocks from Fierza (NorthEastern Albania): geochemical evidence of the Jurassic margin of the Balkan continent. In: Proceedings of" the X V Congress of the CarpathoBalkan Geological Association, 17-20 Septemter, 1995. Special Publication of the Geological Society of Greece, 4(2), 436-442. CAVAZZA, W., ROURE, F. M., SPAKMAN, W., STAMPFLI, G. M. & ZIEGLER, P. A. (eds) 2004. The T R A N S M E D Atlas--the Mediterranean Region from Crust to Mantle. Springer, Berlin. COLLAEU, A., CADET,J. P., MELO, V. & BONNEAU,M. 1990. Sur l'allochtonie des unit6s Albanaises: mise en 6vidence de fenetres tectoniques ~ l'arri6re de la nappe ophiolitique de la Mirdita (Albanie). Comptes Rendus de l'Acad~mie des Sciences, 311, 1251-1258. CROWLEY, K. D., CAMERON, M. & SCHAEFER,R. L. 1991. Experimental studies of annealing of etched tracks in fluoapatite. Geochimica and Cosmochimica Acta, 55, 1449-1465. DEMETS, C., GORDON, R. G., ARGUS, D. F. & STEIN, S. 1990. Current plate motions. Geophysical Journal International, 101,425-478. DIMO, A. 1997, Le mkcanisme de mise en place des ophiolites d'Albanie. PhD thesis, Universit6. Paris-Sud. DIMO-LAHITTE, A., MONIE, P. & VERGELY, P. 2001. Metamorphic soles from the Albanian ophiolites: petrology, geochronology, and geodynamic evolution. Tectonics, 20, 78-96. DONELICK, R. A., KETCHAM,R. A. & CARLSON,W. D. 1999. Variability of apatite fission-track annealing

FISSION-TRACK THERMOCHRONOLOGY, ALBANIDES kinetics II: Crystallographic orientation effects. American Mineralogist, 84, 1224-1234. FERRII~RE, J., REYNAUD, J. Y., PAVLOPOULOS, A., et al. 2004. Geologic evolution and geodynamic controls of the Tertiary intramontane piggy-back Meso-Hellenic basin, Greece. Bulletin de la Sociktk Gkologique de France, 175, 361-380. FITZGERALD, P. G., SHORKHABI, R. B., REDFIELD, T. F. & STUMP, E. 1995. Uplift and denudation of the central Alaska Range: a case study in the use of apatite fission track thermochronology to determine absolute uplift parameters. Journal of Geophysical Research, 100, 20175-20191. FRASHERI, A., NISHANI, P., BUSHATI, S. & HYSENI, A. 1996. Relationship between tectonic zone of the Albanides, based on results of geophysical studies. In: ZIEGLER, P. & HORVATH, F. (eds) Peri-Tethys Memoir 2." Structure and Prospects of Alpine Basins and Forelands. Mkmoires du Muskum National d'Histoire Naturelle, 170, 485-511. GALBRAITH, R. F. & LASLETT, G. M. 1993. Statistical models for mixed fission track ages. Nuclear Tracks and Radiation Measurements, 21,459-470. GALLAGHER, K., BROWN, R. & JOHNSON, C. 1998. Fission track analysis and its implications to geological problems. Annual Review of Earth and Planetary Sciences, 26, 519-572. GALLAGHER, K., HAWKESWORTH, C. & MANTOVANI, M. 1994. The denudation history of the onshore continental margin of SE Brazil inferred from fission track data. Journal of the Geological Society, London, 154, 451-457. GLEADOW, A. J. W. & FITZGERALD, P. G. 1987. Tectonic history and structure of the Transantarctic Mountains: new evidence from fission track dating in the Dry Valleys area of southern Victoria Land. Earth and Planetary Science Letters, 82, 1-14. GREEN, P. F., DUDDY, I. R., GLEADOW, m. J. W., TINGATE, P. R. & LASLETT, G. M. 1986. Thermal annealing of fission tracks in apatite 1. A qualitative description. Chemical Geology, 59, 237-253. GREEN, P. F., DUDDY, I. R., LASLETT, G. M., HEGARTY, K. A., GLEADOW,A. J. W. & LOVERING, J. F. 1989. Thermal annealing of fission tracks in apatite 4. Quantitative modelling techniques and extension to geological timescales. Chemical Geology, 79, 155-182. HOLLENSTEIN, C., KHALE, H. G., GEIGER, A., JENNY, S., GOES, S. & GIARDINI, O. 2003. New GPS constraints on the Africa-Eurasia plate motion boundary zone in southern Italy. Geophysical Research Letters, 30(18), 1935, doi: 10.1029/2003 GL017554. HURFORD, A. J. (1990). International Union of Geological Sciences Subcommission on Geochronology recommendation for the standardization of fission track dating and calibration and data reporting. Nuclear Tracks Radiation Measurements, 17, 233236. ISPGJ-IGJN 1983. Geological map of Albania. Scale 1/200 000. ISPGJ-IGJN, Tirana. ISPGJ-IGJN 1985. Tectonic map of Albania. Scale 1/200 000. ISPGJ-IGJN, Tirana. ISPGJ-IGJN 2003. Geological map of Albania. Scale 1/200 000. 2nd edn. ISPGJ-IGJN, Tirana.

555

KETCHAM, R. A., DONELICK,R. A. & CARLSON,W. D. 1999. Variability of apatite fission track annealing kinetics 3: Extrapolation to geological time scales. American Mineralogist, 84, 1235-1255. KETCHAM, R. A., DONELICK,R. A. & DONELICK,M. B. 2000. AFTSolve: a program for multi-kinetic modeling of apatite fission-track data. Geological Materials Research, 2(1); http://gmr.minsocam.org/ Papers/v2/v2n 1/v2n 1abs.html. KHALE, H. G., COCARD, M., PETER, Y., GEIGER, A., REITLINGER, R., BARKA,A. & VEIS, G. 2000. GPSderived strain rate field within the boundary zones of the Eurasian, African and Arabian plates. Journal of Geophysical Research, 105, 23353-23370. KILIAS, A., TRANOS, M., MOUNTRAKIS, D., SHALLO, M., MARTO, A. & TURKU, I. 2001. Geometry and kinematics of deformation in the Albanian orogenic belt during the Tertiary. Journal of Geodynamics, 31, 169-187. KODRA, A., GJATA, K. & BAKALLI, F. 1993. Les principales &apes de l'&olution pal6og6ographique et g6odynamique des Albanides internes au cours du M6sozoique. Bulletin de la Socidt~ Gkologique de France, 164, 61-67. LASLETT, G. M., GREEN, P. F., DUDDY, I. R. & GLEADOW, A. J. W. 1987. Thermal annealing of fission tracks in apatite 2. A quantitative analysis. Chemical Geology, Isotope Geosciences Section, 65, 1-13. LLANGORA, L. & BUSHATI,S. 1990. Geotectonical position of the ophiolites in inner Albanides according to gravity field data. Ofioliti, 5, 52-64. LOUVARI, E., KIRATZI, A., PAPAZACHOS, B. & HATZIDIMITRIOU, P. 2001. Fault-plane solutions determined by waveform modeling confirm tectonic collision in the Eastern Adriatic. Pure and Applied Geophysics, 158, 1613-1637. MARCUCCI, M., KODRA, A., PIRDENI, A. & GJATA,TH. 1994. Radiolarian assemblage in the Triassic and Jurassic cherts of Albania. Ofioliti, 19, 110-115. MARIUCCI, M. T. & MILLER, B. 2003. The tectonic regime in Italy inferred from borehole breakout data. Tectonophysics, 361, 21-35. McCLUSKY, S., BALASSANIAN, S., BARKA, A., et al. 2000. Global Positioning System constraints on plate kinematics and dynamics in the Eastern Mediterranean and Caucasus. Journal of Geophysical Research, 105, 5695-5719. MECO, S. 1988. Konodontet dhe stratigrafia e depozitimeve paleozoike e triasike ne zonen e Korabit. PhD Dissertation, University of Tirana. ME~O, S. 1991. Mbi karakterin mbulesor t~ structures s~ malit t~ Korabit. Buletini i Shkencave Gjeollogjike, 1, 229-234. ME~O, S. & ALIAJ, S. 2000. Geology of Albania. Beitr/ige zur Regionalen Geologie der Erde, 28. MELO, V. 1970. Mbi p6rhapjen e silurian-devonianit n~ zon~n e Korabit. Buletini USHT i Shkencave Naturore, 2, 61-70. MOST, T., FRISCH, W., DUNKL, I., KADOSA, B., BOEV, B., AVGERINAS,A. & KILIAS, A. 2001. Geochronological and structural investigations of the northern Pelagonian crystalline zone. Contraints from K/Ar

556

B. M U C E K U E T AL.

and zircon and apatite fission track dating. Bulletin of the Geological Society of Greece, 34, 91-95. MUCEKU, B., TASHKO, A. & MASCLE, G. 2003. T6 dhana paraprake mbi termokronologjin~ e disa shk~mbinj~ve magmatik t~ zonave Korabi, Gashi e Mirdita. Buletini i Shkencave Gjeollogjike, 2, 17-24. Muco, B. 1994. Focal mechanism solutions for Albanian earthquakes for the years 1964-1988. Tectonophysics, 231, 311-323. O'SULLIVAN, P. B. & PARRISH, R. R. 1995. The importance of apatite composition and single grain ages when interpreting fission track data from plutonic rocks: a case study from the Coast Ranges, British Columbia. Earth and Planetary Science Letters, 132, 213-224. PAPAZACHOS, C., SCORDILIS, E. & PECI ,V. 2002. P- & S-deep velocity structure of the southern AdriaticEurasia collision obtained by robust non-linear inversion of travel times. 50th ESC Assembly, Genoa, Extended Abstracts. PEZA, L. 1985. Depozitimet e Neokomianit ne nenzonen e Malsise se Madhe. Buletini i Shkencave Gjeollogjike, 1, 41-53. ROBERTSON, A. & SHALLO, M. 2000. MesozoicTertiary evolution of Albania in its regional Eastern Mediterranean context. Tectonophysics, 316, 197-254. SELLA, G. F., DIXON, T. H. & MAO, A. 2002. REVEL: a model for recent plate velocities from space geodesy. Journal of Geophysical Research, 107, 2081, doi: 10.102912000JB000033. SI-IALLO, M. 1990. Ophiolitic m61ange and flyschoidal sediments of the Tithonian-Lower Cretaceous in Albania. Terra Nova, 2, 470-488. SHALLO, M. 1992. Geological evolution of the Albanian ophiolites and their platform periphery. Geologische Rundschau, 81, 681-694.

SHALLO, M., KOTE, DH. & VRANAJ, A. 1987. Geochemistry of the volcanics from ophiolitic belts of Albanides. Ofioliti, 12, 125-136. SHALLO, M., CINA, A. & TURKU, I. 1995. Outline of the metallogeny of the Albanian MORB and SSZ-type ophiolites. In: OHNENSTETTER,M. (ed.) Substances, Minerals and Energetics. Documents du BRGM, 244, 27-46. TAGARI, D., VERGELY, P. & ALIAJ, S. 1993. Tectonique polyphas6e plio-quaternaire en Albanie orientale (r6gion de Kor9a-Progradeci). Bulletin de la SociktO Gkologique de France, 164, 727-737. TASHKO, A. 1996. Diversit6 g6ochimiques, p6trologiques et structurales des ophiolites des Albanides. Bulletin de la Socibtk Gkologique de France, 167, 335-343. TASHKO, A., ROUER, O. & TERSHANA,A. 1996. Biotitegranitic dikes with accessory monazite cutting the mantle ultramafic rocks of Bulqiza ultramafic massif (Albania): witness of a ~hot)) obduction. Buletini i Shkencave Gjeollogjike, 1-2, 115-125. WAGNER, G. & VAN DEN HAUTE, J. 1992. Fission Track Dating. Kluwer, Dordrecht. WORTEL, M. J. R. & SPAKMAN, W. 1992. Structure and dynamics of subducted lithosphere in the Mediterranean region. Proceedings, Koninklijke Nederlandse Akademie van Wetenschappen, 95, 325-347. WORTEL, M. J. R. & SPAKMAN, W. 2000. Subduction and slab detachment in the MediterraneanCarpathian region. Science, 290, 1910-1917. ZARKI-JAKN1, B., VAN DER BEEK, P., POUPEAU, G., SossoN, M., LABRIN, E., ROSS~ PH. • FERRANDINI, J. 2004. Cenozoic denudation of Corsica in reponse to Ligurian and Tyrrhenian extension: results from apatite fission track thermochronology. Tectonics, 23, TC1003, doi:1029/2003TC001535.

Late Cenozoic extension in SW Bulgaria: a synthesis ROB WESTAWAY

Faculty o f Mathematics and Computing, The Open University, Eldon House, Gosforth, Newcastle upon Tyne NE3 3P W, UK (e-mail." r. w. c. westaway@ncl, ac. uk) Also at: School o f Civil Engineering and Geosciences, University of Newcastle-upon-Tyne NE1 7R U, UK Abstract: Southwest Bulgaria forms the northern margin of the Aegean extensional province. Since the Early Pliocene (c. 4 Ma), this region has accommodated southward or SSE extension at several millimetres per year, superimposed on c. 400 m of post-Early Pliocene regional uplift. This sense of deformation superseded earlier extension, oriented ENEWSW, which is estimated to have begun in the early Late Miocene (c. 10-9 Ma) and lasted until c. 4 Ma. The regional topography is dominated by NNW-SSE-striking grabens and normal fault escarpments, relics from this time. Normal faults that are now active cut across these older structures, although in some localities normal faults that were oriented obliquely to the earlier extension have remained active, also oblique to the modern extension sense. It is suggested that this present phase of extension relates to the modern sense of deformation throughout the Aegean region and to the modern geometry of the North Anatolian Fault Zone (NAFZ), which is independently inferred to have existed since c. 4 Ma. The earlier ENE-WSW extension is inferred to have involved two phases, the first predating the NAFZ and the second synkinematic with its initial phase of slip during c. 7-4 Ma, when its geometry and the overall sense of deformation in the Aegean region were different from at present. Some previous studies have inferred that SW Bulgaria experienced large-scale extension on low-angle normal faults in the Mid-Miocene or earlier. However, the limited evidence in support of this view is open to other interpretations, and after due consideration can be discounted.

Southwest Bulgaria forms the northern part of the Aegean extensional province (Figs 1 and 2), but has featured in the international literature much less than other parts of this region, such as central Greece and western Turkey. This study will concentrate on two Neogene terrestrial depocentres in SW Bulgaria with general N N W SSE trends, which are now drained by the Struma (Strimon) and Mesta (Nestos) rivers, separated by the Rila and Pirin massifs (Fig. 2). This region thus resembles western Turkey, with subparallel Late Cenozoic depocentres transecting ancient metamorphic massifs (analogous to the Alasehir and Biiyfik Menderes grabens and central Menderes Massif; Fig. l a). This study will thus summarize information regarding timings, rates, senses and amounts of extension in SW Bulgaria, and where possible will draw analogies with western Turkey. One difference concerns topography. The land surface reaches 2925 and 2915 m in the Rila and Pirin massifs (Fig. 2), c. 1 km above the highest points in the Menderes Massif. These high altitudes mean that, unlike most mountain ranges in western Turkey (see Erinq 1978; Demir et al. 2004), Rila and Pirin were extensively glaciated during cold stages of the Pleistocene (e.g. Velchev

1995; Zagorchev 1995). Another difference concerns climate. Although southern Bulgaria is barely 100 km from the Aegean coast, the characteristic Mediterranean climate, with hot dry summers and warm wet winters, gives way abruptly northward to a different regime, with maximum rainfall (locally > 2 0 0 0 m m ; Furlan 1977) in summer (notably, in June). One consequence of this rainfall seasonality is much more profuse vegetation than in most of western Turkey, limiting field exposures of normal faults and sediments. One aspect in common between western Turkey and SW Bulgaria is that active normal faulting is superimposed on a background of regional uplift. This idea is well established in Bulgaria (e.g. Zagorchev 1992a) but has developed only recently in western Turkey (e.g. Westaway et al. 2004). It is thus of interest to compare the uplift histories of the two regions. Another is that investigation of young crustal extension, taken up on initially steep normal faults, has become mixed up in both regions with hypothetical earlier low-angle normal faulting. This association makes it difficult to disentangle the two forms of evidence in some past publications (e.g. that by Burchfiel et al. 2000). Although

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 557-590. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

558

R. WESTAWAY

i

"~

sE 3~

I"m&g~lI ~ '

t "~

I \M

' 5

'-0~'"

J

s s''"~w

l Ishkur

~

km

i'" ~

43~

\

0

.... " Sofia ) ~ 1.2:1:14",L 9 " oE,SOFI( ~\ \ , S8::1:68

/

-MK - ~ ' . ~ ,'

~

.....

/

I

100

L.,~

, !

Maritsa

,

..... BOGS~' t YBlag. . . . / 2.5::L1.1('~ "*" '" 2.0+1,0 " ~" grad /DOBR S15+25~ $6• ', \ ...a...~3.5+1.1 PLO2 "..~4.f .~LSI9+I8OE 2 3+1 1 /- ~ t'~ "h 3.9• c'A-;~7o= i \ ~ J ,SAT1 S23+16~ ~=x"n,~-3... ..~ ' BG

Vardar~

~

-

"" ~ Q - 4 ~ L '"'GR

/ < Key:

~

I~.~,~l Actlvestrike.sllpfault

~

.....

poc....

Subducil. . . . .

~ ~

oAnCt/~o rn~ ifangUJt:lithepoc. . . . . .........................

Z g

I(b)

/

~

. j ~ :' ] ~

~

~

j.:

~-~ r

.:i < ~::

Fig. 1. (a) Map of the Aegean extensional province, showing active faulting and related sedimentation, adapted from Westaway (1994, 2002c). The right-lateral North Anatolian Fault Zone enters the study region from the NE, its strands terminating against NE-dipping normal fault zones bounding the NE coasts of Evvia and adjacent islands. Thess, Thessaloniki; GD, Po and WP, Gotse Delchev, Podgorie and West Pirin normal faults in SW Bulgaria; D, L, Se and X, Drama, Langadas, Serrai (Serres) and Xanthi grabens in northern Greece; P and S, Parnassos and Sperchios normal faults in central Greece. (b) Crustal velocity field across SW Bulgaria, measured by GPS surveys in 1994-1998 (data from Kotzev et al., 2001, fig. 4; McClusky et al. 2000), illustrating a gradual increase in southward velocity relative to Eurasia, starting south of point BERK. Kotzev et al. (2001) inferred from this dataset that a single notional zone of NW-SE extension at up to c. 4 mm a-] transects this region, located between the Vitosha Mountains and the Rila Massif (i.e. south of PLAI and north of BOGS and PLO2. However, the geological evidence favours more widely distributed southward extension instead.

low-angle normal faulting is mechanically feasible (e.g. Westaway 1999, 2005), to cause the stress field at depth to rotate sufficiently to allow a planar normal fault to form with an initial dip as low as c. 30 ~ requires a demanding set of physical conditions, and no-one has ever demonstrated that such conditions have existed in the Late Cenozoic in either Bulgaria or Turkey. Claims regarding low-angle normal faulting in western Turkey have anyway been considered mistaken, because thermochronological evidence of cooling attributed to this process may instead represent erosion (e.g. Westaway 1996, 2006) and because initially steep normal faults may develop lowangle dips by back-tilting (e.g. Bozkurt 2000; Purvis & Robertson 2004). Criticism of the idea of Late Cenozoic low-angle normal faulting in Bulgaria (see Zagorchev 1994, 1998a, 2001b) has been strident (e.g. Zagorchev 200 l b).

This study will attempt a synthesis of evidence, regarding both steep and possible lowangle normal faulting, for the Late Cenozoic sedimentary basins in SW Bulgaria. Having summarized the evidence, it will estimate rates and senses of extension and rates of regional uplift and consider the regional significance of this normal faulting. One difficulty is that the existing literature on SW Bulgaria reports many apparently 'neotectonic' faults, as in Figure 2. However, it is evident that this region has experienced a long and complex deformation history (see Zagorchev 1992a, b, 2001a), and that in addition to young normal faults this map shows faults associated with earlier (?Mid-Miocene) NNW-SSE-directed right-lateral strike-slip and even earlier (?Early Miocene) reverse faulting. Furthermore, many of the faults shown affect only metamorphic

LATE CENOZOIC EXTENSION, SW B U L G A R I A

559

--.J

L. x

~ +

+

+

4` +

/ SERBIA

+

t' "

+ +

4+

4-

+ 4`

)

Musala A

2925m

/

Delchev,

/ Normal faults Thrust

~,~ \

faults

Pliocene - Quaternary sediments

\

|

\

Late Miocene - P l i o c e n e sediments

~-

\

\ Middle Oligocene - early L a t e Miocene sediments ~

"~ ~9 ) ;'

\

r.re

/

\ Ograzhden (-~'J"

Early Oligocene

Mts.

)~ l~

volcanics

Early Oligocene granites

) (.

\

E o c e n e - Early Oligocene marine sediment

Late

~

~

E o c e n e - Early Oligocene volcano-sedimentary deposits

.)

Late

.

Late Eocene terrestrial

sediment

/~T"I 'IL~..L~ ' ~'~"

MACEDONIA

Late Cretaceous

1"

granites

i GREECE

f

0

, ....

, ....

10 ,

20 , km

30 i

40 J

Fig. 2. Map of the present study region in SW Bulgaria, adapted from Zagorchev (1992b, fig. 9), who described it as a 'simplified Late Alpine neotectonic map'. Precambrian, Palaeozoic and most Mesozoic rocks are not shaded. The area around Sofia is illustrated only schematically, having been omitted from the map on which this figure is based. The Struma river rises NW of Pernik, from where it flows initially SW through Radomir and Zemen to Kyustendil, then SE to Boboshevo, then SSE past Blagoevgrad, Simitli, Kresna and Sandanski. The Mesta rises south of Kostenets, flowing initially SW to Razlog then SSE past Gotse Delchev. After passing into Greece, both rivers reach the north coast of the Aegean Sea.

560

R. WESTAWAY

basement, and so could be ancient, and others indicate how such old faults have been inferred to project beneath young depocentres. Moreover, as will later be discussed in more detail, some of the faults shown are notional constructs to provide boundaries between rock units in areas of limited exposure. As a result, to keep this study focused and of manageable length, it will concentrate on faults that show clear relationships to Late Cenozoic extension, and subsequent maps will be simplified to emphasize this evidence. Older faults, or faults that have been inferred as structural constructs, will be discussed only where relevant to this history of young extension.

The Blagoevgrad Basin and surroundings The Blagoevgrad Basin is a c. 40 km by 10 km terrestrial depocentre bounding the western and northern margins of the Rila Massif between Blagoevgrad and Sapareva Banya (Fig. 3). Exposure is very limited except in the badland landscape of the 'Stobski Piramidi' (in the footwall of the Rila Fault; Fig. 3; see below) and in occasional roadcuts. The local stratigraphy, after Zagorchev (1992a, 2001a), begins with a basal red polygenetic conglomerate, overlain by polygenetic conglomerate interbedded with sand and clay, the Pokrovnik Formation. This is followed by, or interbedded with, stacked white, yellow and green fluvial sand and clay, interbedded with gravel lenses, of the Dzherman Formation. Sites in this formation at Kocherinovo and Mursalevo have yielded characteristic Turolian mammals, including the giraffe Helladotherium sp., the hipparion Cremohipparion mediterraneum, the hyaena Adcrocuta eximia, the gazelle Gazella brevicornis, and the suid Microstonyx major (Nikolov 1985). In modern nomenclature (see Westaway et al. 2004, fig. 6), these species coexisted during biozones M N l l (c. 9.0-8.2 Ma) and MN12 (c. 8.2-7.1 Ma), in the Meiotian stage of the Late Miocene. This unit is followed unconformably by stacked cross-bedded white and yellow fluvial gravels and sands, of the Barakovo Formation, of inferred (?)Pontian to Late Pliocene (Romanian) age. Finally, at the western margin of the Rila Massif, the eroded top of the Barakovo Formation is covered by slope wash, of inferred Eopleistocene (i.e. Early Pleistocene) to Pleistocene (i.e. Mid-Late Pleistocene) age, known as the Badino Formation.

The Saparevo norrnal fault Arguably the clearest indication of active extension in this region is provided by the NW-dipping

Fig. 3. Map of the Blagoevgrad Basin, adapted from Zagorchev (1992a, fig. 9), with additional information from Shipkova & Ivanov (2000, fig. 1) and my own field observations. Dupnitsa was known as Stanke Dimitrov in the communist era and is still labelled as such on some modern maps. Estimated ages of the Late Cenozoic formations defined within this basin are discussed in the text. The Kalin pluton forms the footwall of the Saparevo Fault in the NW part of the Rila Massif. However, it is not individually labelled, as no pre-Cenozoic rocks are differentiated on this map.

Saparevo normal fault near Dupnitsa (Figs 3 and 4a,b), which has uplifted the West Rila horst in its footwall by c. 2 km relative to the depositional surface in its hanging wall (c. 2600 against c. 600 m above sea level (a.s.1.)). The fault plane is well exposed at the base of this footwall escarpment (Fig. 4c), as a smooth, relatively fleshlooking rock surface, locally in granite of the Kalin pluton, which intrudes Late Precambrian basement (see Zagorchev 2001a). This is one of many geochemically similar plutons, designated as the 'South Bulgarian granitoids' (see Zagorchev 1998a). Although the Kalin pluton is not dated, others in this grouping have been Rb-Sr dated to the Hercynian orogeny

LATE CENOZOIC EXTENSION, SW BULGARIA

561

Fig. 4. The Saparevo normal fault between Dupnitsa and Sapareva Banya. (a) View looking east obliquely towards this fault from [FM 77398 81787], c. 1 km east of Samoranovo, with Resilovo in foreground. View shows characteristic footwall morphology of a steep planar normal fault. Viewpoint is c. 600 m a.s.1.; summit skyline in footwall is c. 2600 m a.s.1.. (b) View looking S20~ at the footwall escarpment from [FM 81655 82272], between Resilovo and Ovchartsi, showing characteristic 'flat-iron' facets being incised by small streams to form 'wineglass canyons'. (c) Close-up view of the base of the footwall escarpment at Ovchartsi, at [FM 83571 82236], showing a relatively fresh smooth rock surface (in Late Palaeozoic granite) degrading upwards into a more eroded rock face and buried downslope by a pediment of slope scree. This site is located just east of the point where the Goritsa River crosses the Saparevo Fault in Figure 3. (d) Interpretation of the locality illustrated in (c), from Shipkova & Ivanov (2000, fig. 3). The interpreted 'chlorite breccia' corresponds to the slope talus depicted in (c), and the interpreted 'mylonite' corresponds to the brittle joint set oriented subparallel to the fault plane in (c). Universal Transverse Mercator (UTM) position fixes such as [FM 77398 81787], indicating the co-ordinates of sites to the nearest metre, were obtained in the field using a handheld GPS receiver. The letters indicate the 100 km x 100 km UTM quadrangle; the five-digit numbers indicate distance east and north, respectively, from the SW corner of this quadrangle. (See text for discussion.)

562

R. WESTAWAY

(c. 340-240 Ma; Zagorchev & Moorbath 1986; Zagorchev et al. 1989a). The geochemically similar Kavala (Symvolon) granite in northern Greece, also assigned to this grouping (see Zagorchev 1998a), has also yielded U-Pb dates on zircons of c. 335 Ma (Kokkinakis 1980; Dinter et al. 1995; see also below). The Kalin granite has experienced pervasive brittle deformation leading to the formation of a fracture set oriented subparallel to the escarpment face, plus other sets at different orientations (some visible in Fig. 4c), which contribute to producing the irregular appearance of this fault escarpment. In places, the fault surface is preserved only as small facets, with dimensions of tens of centimetres, bounded by the intersection lineations between these fracture sets. This fault surface is not precisely planar: measured up a c. 5 m face its orientation varied upward from dip 60 ~ towards N53~ to dip 40 ~ towards N60~ However, it seems clear that the escarpment rising c. 2 km above this faceted fault surface represents the uplifted footwall of this fault, its degree of weathering increasing upslope. No striations were found that could indicate the precise extension sense. However, the heave on this normal fault (between its footwall and hanging-wall cutoffs) indicates several kilometres of extension. Shipkova & Ivanov (1999, 2000, 2001) have proposed a radically different interpretation of this structure, which they call the 'Dzherman detachment'. In their view (Fig. 4d), the Kalin pluton was intruded during Late Cenozoic extension, and became mylonitized at mid-crustal depths before passing upward into the brittle regime and receiving what they interpreted as a 'cataclastic overprint' oriented subparallel to the fault plane. They interpreted what seemed to me to be slope scree in the hanging wall (Fig. 4c) as 'chloritic breccia' (Fig. 4d), its thickness providing an indication to them of the magnitude of the slip on this fault while in the brittle regime. Although no dipping Neogene sediments were visible to me in the Holocene alluvial plain adjoining this normal fault, they concluded that these sediments dip subparallel to the fault plane, thus suggesting that it formed at a low-angle dip, and must thus have accommodated many tens of kilometres of extension in order to exhume material that was initially at mid-crustal depths. This interpretation seems unlikely but is impossible to test in detail because necessary information (e.g. thin sections indicating mylonitization, exposures of Neogene sediment, dates indicating a Neogene intrusion age) are lacking. I return in the Discussion section to the wider question of granite intrusion ages in this region.

The Kyustendil normal fault West of Kyustendil, beyond the western margin of the Blagoevgrad Basin (Fig. 2), the northern margin of the Osogovo mountains is abruptly truncated by a planar escarpment, dropping from c. 1800 m to c. 900 m a.s.1.. This feature continues in line with $75~ strike for c. 20 km to the border with Macedonia at Gyushevo (Fig. 2), then for c. 30 km farther inside Macedonia. It has been mapped (Fig. 2) as the footwall escarpment of a significant active normal fault (see Zagorchev 1992a). Kyustendil town occupies the SW corner of the Kyustendil Basin, another (?)Mid-Miocene to (?)Early Pliocene terrestrial depocentre now c. 500m a.s.1. (see Vatsev & Bonev 1994, or Zagorchev 2001 a, for details of its stratigraphy). The Struma enters this basin at the outlet of the Zemen gorge at Razhdavitsa, then flows across this basin for c. 15 km to Nevestino, where it enters the Skrino gorge that leads to the Blagoevgrad Basin at Boboshevo (Figs 2 and 3). There is abundant evidence throughout this area, including within the Skrino gorge (notably at Pastuh [FM 57452 74256]), of a staircase of Struma terraces, indicating significant young fluvial incision. The southern margin of the Kyustendil Basin was inspected for evidence of normal faulting for c. 15 km distance between Kyustendil and Nevestino (at [FM 51991 79937]). The abrupt footwall escarpment of the Kyustendil normal fault seems to follow the margin of this basin for a few kilometres directly south of Kyustendil town. However, farther east it appears to die out, as the basin margin becomes much more subdued. The suggestion in Figure 2 that it continues ENE beneath this basin seems conjectural, as no supporting evidence (such as warped or offset river terraces) could be observed.

The Rila normal fault Unlike farther NE, the contact between basement and Neogene sediments at the western margin of the Rila Massif between Dupnitsa and Rila town (Fig. 3) shows no clear geomorphological evidence of active normal faulting. However, a clear active normal fault zone, which I call the Rila Fault, strikes SW, oblique to the margin of this massif (Figs 3 and 5a). This fault is clearest just west of Rila town (Fig. 6a), where its c. 300 m high footwall escarpment is formed in basement (Fig. 5a and b). Farther SW (Fig. 6b), the fault cuts through the Neogene sequence of the Blagoevgrad Basin, creating the 'Stobski Piramidi' badland landscape in deposits of the

LATE CENOZOIC EXTENSION, SW BULGARIA

563

responsible for the local downthrow of conglomerates of the Palaeogene Padala Formation (see Zagorchev et al. 1999; see also below) relative to the basement farther south, which has facilitated preservation of these sediments. The Rilska gorge section

Fig. 5. (a) Map of the SE Blagoevgrad Basin, showing the locations of the Rilska river gorge and the Rila normal fault. (b) Inset showing the dissected landscape of the Stobski Piramidi nature reserve in the footwall of the Rila normal fault. (See (a) for location.) (e) Map of the lower part of the Rilska gorge through the Palaeogene Padala Formation. (See (a) for location.) Adapted from Zagorchev (1995, fig. 40) and Zagorchev et al. (1999, fig. 3). (See text for discussion.)

Barakovo and Badino formations in its footwall (Fig. 5a and c). In contrast, its hanging wall, drained axially by the Rilska River, is covered by inset Pleistocene or Holocene sediments (Figs 3 and 5a). The height of this footwall escarpment decreases gradually southwestward from Stob, becoming minimal by its intersection with the Struma River (Fig. 3). An approximate in-line continuation of this fault zone ENE of Rila town can be traced to the vicinity of Padala (Fig. 5c), and may be

The stratigraphic section exposed in the gorge of the Rilska River (Fig. 6c) has recently been the subject of controversy regarding possible reinterpretation of it as evidence of Late Cenozoic lowangle normal faulting (see Shipkova & Ivanov 1999, 2000, 2001). As already noted, the local outcrop has been regarded as a pocket of Palaeogene conglomerate, which overlies metamorphic basement and has been fortuitously preserved as a result of young normal faulting that has downthrown it relative to the adjoining basement (see Zagorchev et al. 1999). Coal beds and other fragments of plant remains have long been reported in this deposit (e.g. Bonchev 1912; Konyarov 1932; Moskovski 1983; Marinova 1991; Zagorchev 1998a; Zagorchev et al. 1999), suggesting a sedimentary origin, presumably marsh at the foot of a hillslope at the margin of the basement upland. Shipkova & Ivanov (1999, 2000) have proposed, on the contrary, that this conglomerate is a fault breccia and the adjacent basement surface is mylonitized (Fig. 6d), the contact between the two thus indicating a lowangle normal fault surface with a c. 20 ~ dip, along which the basement to the east has uplifted from mid-crustal depths during large-scale Late Cenozoic extension. However, Shipkova & Ivanov (1999, 2000) offered no explanation for the plant material found in this sediment, and their interpretation that it is a cataclasite was based on sedimentary characteristics (notably, angularity of clasts; fining-upward characteristics; clasts and matrix have same composition, matching the local basement; micas in the matrix show chloritization) that could also be expected in a slope deposit derived from the local basement. The contact between Padala conglomerate and basement is well exposed as a result of recent widening of the road to Rila monastery, east of Orlitsa monastery at [FM 79016 66340]; it seemed to me to be an unconformity surface, dipping west at c. 30 ~ Moreover, I could see no evidence of mylonitization of the underlying basement, whose upper surface appeared weathered, not faulted or deformed in a ductile manner. However, this conglomerate is broken up by minor normal faulting, which post-dates its deposition; for instance, one fault at [FM 78953 66350] dips at 72 ~ towards $80~ I did not observe any clear evidence of plant remains in this conglomerate (several of the sites depicted

564

R. WESTAWAY

Fig. 6. (a) Oblique view looking SSE from [FM 75370 65032], just SW of Rila town, towards the S35~ Rila normal fault. Alluvial fans prograde from wineglass canyons, through the faceted range front; the downthrow from the tops of the facets shown to the foreground hanging-wall alluvial plain, at c. 500 m a.s.l., is c. 300 m. (b) View looking ESE at the Rila normal fault from [FM 73198 63489], on the road between Rila town and Stob (the village in the right middle distance), c. 3 km SW of the viewpoint in (a). The footwall escarpment is now only c. 100 m high, compared with c. 300 m; its height continues to decrease to the SW, dying out completely c. 5 km farther SW. Erosion of the Neogene sediment that has been uplifted in this footwall has created the erosional landscape known as the 'Stobski Piramidi' (Fig. 5b). The summit in the left background, at 2386 m a.s.l., is c. 15 km away, in the interior of the Rila Massif. The bulk of the local relief shown is thus unrelated to the small amount of throw indicated by the height of the footwall escarpment along the range front. (e) View looking south, from [FM 77958 66740], across the Rilska river gorge between Rila town and Orlitsa, showing Palaeogene conglomerate of the Padala Formation dipping at c. 10~ towards $56~ on the south flank of the gorge. (d) Interpretation of the section along the Rilska gorge through the locality illustrated in (c), from Shipkova & Ivanov (2000, fig. 2). The interpreted 'autoclastic breccia' corresponds to the conglomerate depicted in (e). (See text for discussion.)

LATE CENOZOIC EXTENSION, SW BULGARIA

565

in Fig. 6c are now inaccessible because roadwidening has converted the gorge flanks into vertical cliffs), but the literature reports (dating back to the early 19th century) of coal, plant leaves and stems, and pollen indicating an Oligocene age (summarized by Zagorchev et al. 1999) seem clear enough.

across the Blagoevgrad Basin and to have tilted to a low-angle dip to achieve the steep dip of the sequence in the Padezh Basin. I return to these views, which seem conjectural, in the Discussion section.

Other inferred low-angle n o r m a l f a u l t i n g

South of Blagoevgrad (Fig. 7), the Struma flows for c. 10 km, in a gorge through basement, to the Simitli Basin. This c. 20 km by 5 km basin is in the hanging wall of the Krupnik normal fault to the SSE (Fig. 8). On leaving it, the Struma flows into this footwall, then along the Kresna gorge for c. 15 km to the Sandanski Basin. Kotzev et al. (2001) suggested that the Krupnik Fault is left-lateral, but this view seems to have no basis. The detailed literature on it (e.g. Botev et al. 2001; Meyer et al. 2002) makes clear that it is an active normal fault, as seems obvious from the local structure and geomorphology (Fig. 8). It was indeed the probable source of the 4 April 1904 Krupnik earthquake (e.g. Zagorchev 1992a, 1995; Meyer et al. 2002; see Ambraseys 2001), the largest event known in the Balkan region (magnitude c. 7.8). The stratigraphy of this basin (see Marinova & Zagorchev 1990; Zagorchev 2001a) begins with a coal-bearing unit, the Oranovo Formation, dated to the Sarmatian from pollen. This is overlain in the stratigraphic scheme used by Meyer et al. (2002) by a stacked sequence of inferred Meiotian age, comprising the Chernichevo, Gradevo and Simitli Formations: the first two consisting of conglomerate and the last of fluvial and lacustrine sandstone and siltstone with conglomerate interbeds. In contrast, rather than describing a layer-cake sequence, in the Zagorchev (2001a) scheme these deposits are interpreted as marking a lateral facies variation: the coarser Chernichevo Formation at the northern and western margins of the basin with the finer deposits in more distal areas. These deposits have been dated to the Meiotian using pollen (Zagorchev 2001a). They are overlain by stacked fluvial sand and conglomerate that is assigned to the Kalimantsi Formation (defined within the Sandanski Basin farther south; see below). Outliers of this deposit crop out on the crest of the Krupnik Fault footwall (Fig. 8a and b), suggesting that when deposited it was continuous between the Simitli and Sandanski basins. Near Brezhani, this uppermost member of the Kalimantsi Formation has yielded fossil evidence of the bovid Gazella (Tragoportax) aft. gaudryi (Nikolov 1985). Assuming correct species identification, this is a Turolian (latest Miocene) taxon that spanned biozones MN11-12 (the Meiotian)

West of the Blagoevgrad Basin is an elongated (c. 10 km by 3 km) S20~ outcrop of Palaeogene sediment, the Padezh Basin (Figs 2 and 3). Now situated c. 600 m a.s.l., this is dated to the Priabonian to Early Oligocene, and records a transition from terrestrial to marine conditions (Zagorchev et al. 1989b; Zagorchev 2001a). These only partly lithified sediments, which attain a thickness of many hundreds of metres, are strongly tilted ENE and broken up by postdepositional faulting; they are exposed along the roads from Blagoevgrad to Padezh and the nearby villages of Leshko and Gabrovo; for instance, marine limestone (Padezh Formation) dipping at 50 ~ towards N65~ at [FM 64951 42801] and the underlying fluvial sand, conglomerate and lacustrine claystone (Logodazh Formation) dipping at 55 ~ towards N74~ at [FM 62288 44446]. This marine basin is thought to have been linked to the open sea to the NW, and is the farthest south where marine sediments from this time are documented (see Zagorchev 1992a). This Palaeogene sequence thickens westward across this basin, suggesting that the clastic input was from the Vlahina mountain range to the west (Fig. 2). These sediments are locally capped by up to c. 50 m of stacked fluvial sand and gravel assigned to the Barakovo Formation (see above), which has also been derived from Vlahina. The literature is unspecific as to the tectonic setting during deposition in this basin. There is of course no requirement for any contemporaneous crustal deformation (independent of rock uplift caused erosion, and subsidence caused by deposition); these sediments could simply have accumulated in a sag basin adjacent to the eroding Vlahina mountains. Their subsequent deformation has been attributed (e.g. by Zagorchev 1992a) to a combination of thrusting and NNW-SSEdirected right-lateral strike-slip in the EarlyMid-Miocene, between the end of local deposition and the start of crustal extension. Burchfiel et al. (2000, 2003) suggested, on the contrary, that these sediments were deposited during Eocene to Oligocene extension, and attributed their strong eastward tilting to downthrow to the west on a major normal fault system, which they inferred to be associated with extension

The Simitli Basin

566

R. WESTAWAY o, .... Massif

X x

+

~

*

+o

x

~

Dautov pluton " 9 +, + =

4.,

X

X

BASIN' x x

+,4.

x

x x

x

X x

x x

x x

x x

9+

pluton

4-

+/~+

4- ,

x

x

X Rila-West Rhodope x x batholith x

2915m A

2,0

X x

', : : : ; R A Z L O G

kp ....

x x

t-

x

x

x

x

x x

x

x

x x

x

x x

x

x

9

x

KREBNA + +

+

Central Pirin + pluton +~

4-

4-

+

4.

+

44-

+

+

, 4-

+

I

§

+

.':~+

+~+4

4+

+

Massif

+

+ 4-

Rhodope

4+

+

+I +

+

+

| + li

GOTSE DELCHEV

~9

Ograzhden

Massif

./~_

Igralisht xXJ ~~

'

~

pluton

+ Ograzhdel Fault

+

t+

Teshovo

+ pluton

+

4- Tesho~o +

~,... Bulgaria 9~ - . ~ "~.. ~ . - - .

STRUMESHNITSA~.I'-:::<: o,-"

.:::.!::~A.s ! N ~ c )

~ O D G O R I E FAULT Belasitsa Mts.

'Greece

( ) ,,,f

i

N e o g e n e and Quaternary sediment

~+'--'~

sediment

~

Palaeogene granites

(?) Paleocene - E o c e n e Early Oligocene sediment

-x--~

Palaeogene volcanlcs

[

S

Nevrokopi

Normal fault associated with earlier (Late Miocene - Early Pliocene)

Hercynian granites IMet .... phicb ......

,/

Active normal fault

Late Cretaceous granites

Middle-Late Ollgocene

t" ,j

phase of extension

t

Inferred or notional normal fault

Fig. 7. Map of the southern part of the study region, showing the Simitli, Razlog, Sandanski and Gotse Delchev basins, simplified from Zagorchev (1995, fig. 2). The distinction between Hercynian, Late Cretaceous and Palaeogene granites should be noted; this is supported by dating (e.g. Zagorchev & Moorbath 1986; Zagorchev et al. 1989a) and geochemical and isotopic analysis (e.g. Zagorchev 1995). Faults older than the Late Cenozoic are not shown, except for the reverse fault along the eastern margin of the Palaeogene Brezhani Basin, which is labelled with ticks on its uplifted hanging wall.

and MN13 (the Pontian) (see Gentry & Heizmann 1996). However, in the northernmost Sandanski Basin farther south, near Kresna, the uppermost member of the Kalimantsi Formation has yielded fossil evidence of the mastodon A n a n c u s arvernensis (Nikolov 1985), a Pliocene species (see Athanassiou & Kostopoulos 2001).

From the evidence in this area this sedimentary unit is thus not reliably dated and may be diachronous. The Brezhani Basin comprises a c. 2 km by 8 km pocket of Late Oligocene terrestrial sediment also in the footwall of the Krupnik Fault, between the Simitli and Sandanski basins (Figs 7

LATE CENOZOIC EXTENSION, SW BULGARIA

567

Fig. 8. (a) Map of the Simitli Basin. (b) Cross-section through the Krupnik Fault. Adapted from Meyer et al. (2002, fig. 3), with additional information from Marinova & Zagorchev (1990). Points A, A', etc. are labelled to facilitate the discussion in the text of amounts of vertical slip on the Krupnik Fault and amounts of erosion from its footwall.

568

R. WESTAWAY

and 8a). Its succession consists of five formations: two (the Goreshtitsa and Loulevska) contain lignite, the others being siltstone, sandstone and conglomerate (see Marinova & Zagorchev 1990; Zagorchev 2001a). After deposition, this basin was partly buried by (?)Early Miocene overthrusting from the east (see Zagorchev 1995); in the south it was later unconformably overlain by the Kalimantsi Formation (Figs 8 and 9). Previous studies (e.g. Meyer et al. 2002) have inferred that the whole Simitli Basin succession

Fig. 9. Map of the Sandanski Basin, also summarizing detail for the Simitli Basin, adapted from Zagorchev (1992a, fig. 8). B, Brezhani; D, Novo Delchevo; I, Ilindentsi; Ka, Kalimantsi; Kr, Krupnik; S, Strumyani. 1 and 2, alternative positions for the southern end of the Melnik Fault, from Kojumdgieva et al. (1982) and Zagorchev (1992a). It should be noted that, as drawn here, by appearing to juxtapose older sediment to the west against younger sediment to the east, this southern end of the Melnik Fault appears to show downthrow to the east, not the downthrow to the west illustrated in Fig. 10b, ii. As noted in the text, there is ambiguity here regarding the precise position of this fault and also regarding the correlation of the sediment west of it (illustrated here as the Delchevo Formation): whether it correlates laterally with the Katuntsi Formation, and is thus older than the Sandanski Formation, or not.

accumulated during extension, in the hanging wall of the Krupnik Fault, thus placing the start of extension in the Sarmatian (c. 13 Ma). Meyer et al. (2002) thus estimated a time-averaged vertical slip rate of c. 0.1 m m a -1 and a total of c. 1.3 k m of vertical slip ( C - C " in Fig. 8b), whereas Zagorchev (1992a) estimated c. 0.25 m m a -~ and c. 3.5 km instead. In the east (P-Q in Fig. 8a), Meyer et al. (2002) estimated the offset of the base of the Kalimantsi Formation as c. 600 m (from c. 450 to c. 1050 m a.s.1.), implying a vertical slip rate of c. 0.1 m m a -1 if its age is 6 Ma (or c. 0.09 m m a -1 if 7 Ma), suggesting that the slip has been uniform at this rate. However, the greater altitude of the basement in the footwall farther west ( > 1100 m, west of the Kresna gorge; Fig. 8b) suggests that this part of the Krupnik Fault has slipped faster. Given that deposition of the Kalimantsi Formation straddled the eastern part of the Krupnik Fault that appears to have slipped least, all the earlier formations may have also been deposited across this fault farther west, but have since been removed from its footwall by erosion. If so, and all its slip is further assumed to post-date the Kalimantsi Formation, then an estimated thickness of 950 m (A'-C" in Fig 8b) could have been eroded from the footwall, in which case the vertical slip on the western Krupnik Fault could have been as much as C"-B (c. 1600 m) plus A : C " , or c. 2550 m. Furthermore, if all this slip occurred after deposition of the Kalimantsi Formation (since a nominal time of c. 3 Ma; see further discussion below) then the vertical slip rate could be as high as c. 2550 m in c. 3 Ma or c. 0.85 m m a -~. If so, then the c. 250 m of incision by the Struma river from c. 550 m (A' in Fig. 8b) to its present level of c. 3 0 0 m a.s.1, occurred at a timeaveraged rate of c. 0.08 mm a -1, with a timeaveraged incision rate of c. 0.62 m m a -1 at the footwall cutoff (for c. 1850 m of incision since 3 Ma). Conversely, if one takes the estimate of c. 0.1 mm a -~ from Meyer et al. (2002) as a lower bound to the vertical slip rate on the Krupnik Fault, then the c. 0.08 m m a -~ time-averaged rate of incision by the Struma in the hanging wall requires an incision rate of c. 0.18 m m a -~ in the footwall cutoff, requiring c. 540 m of incision since c. 3 Ma, placing the contemporaneous land surface at the footwall cutoff at c. 840 m a.s.l.. This is significantly higher than the modern flanks of the Kresna gorge, which only reach to c. 600 m a.s.1., indicating that the present morphology of this gorge significantly underestimates the amount of incision that has occurred. Presumably, when this gorge began to develop the Struma cut into the lowest point in the former

LATE CENOZOIC EXTENSION, SW BULGARIA depositional surface of the Kalimantsi Formation, causing it to become localized in its present position, where it subsequently incised through the base of the Kalimantsi Formation into the underlying metamorphic basement; it is thus an example of superimposed drainage.

The Sandanski Basin The Late Cenozoic sequence in the Sandanski Basin (Figs 7 and 9), with total thickness > 1600 m, begins with the Delchevo Formation, of Badenian to Sarmatian or Meiotian age, consisting of red to green silty fluvial sandstone with siltstone, shale, clay and fine conglomerate, plus occasional tuff and limestone beds (e.g. Kojumdgieva et al. 1982; Zagorchev 1992a). It has yielded, at Levunovo, teeth of the antelope Micromeryx flourensianus (e.g. Kojumdgieva et al. 1982; Nikolov 1985), an Astaracian (late Mid-Miocene; biozones MN6-8) taxon (see Kojumdgieva et al. 1982; Gentry & Heizmann 1996) and, at Novo Delchevo, a tooth of the Vallesian elephant Choerolophodon serridentinoides (e.g. Kojumdgieva et al. 1982; Nikolov 1985). Similar sediments in the adjacent Serrai Basin in northern Greece (Fig. la) are late Vallesian (biozone MN10) from their mammal faunas (e.g. Karistineos & Georgiades-Dikeoulia 1986). The Katuntsi Formation, consisting of red conglomerate, has been designated (e.g. Zagorchev & Dinkova 1990) near the eastern margin of the Sandanski Basin (Fig. 9) and interpreted as a proximal facies that interfingers with the more distal Delchevo Formation (Zagorchev 1992a), but is not directly dated. The Sandanski Formation, of reported Meiotian age (e.g. Zagorchev 1992a), consists of coarser fluvial sandstone, sometimes crossbedded, with silty interbeds. The Kalimantsi Formation, of reported Pontian to (?)Romanian or (?)Dacian age (e.g. Zagorchev 1992a), consists mainly of conglomerate, typically formed of rounded clasts of granite from the Palaeogene Central Pirin pluton to the east (Fig. 7). However, its lowest member, the Ilindentsi Member, comprises carbonate-cemented clasts of the (?)Precambrian Dobrostan Marble. It is found on the eastern flank of the Sandanski Basin, adjacent to outcrop of this marble in the western Pirin Massif (Figs 9 and 10a), the clast size ranging up to huge olistostromes with dimensions of many tens of metres. The above sediments typically dip east at e. 5-20 ~ (Figs 9 and 10b), indicating significant downthrow since their deposition on normal faults bounding the eastern basin margin (see Zagorchev 1992a). North of Lyubovka, this margin is bounded by the S25~ West

569

Pirin normal fault (Figs 9 and 10a). At Lyubovka, this fault is considered to splay, with one strand (the Melnik Fault) continuing $25~ into the basin interior, and the other (the Gorno Spanchevo Fault) (Fig. 10c and d) branching eastward then bending progressively to the right as it follows the basin margin, reaching c. 6 km from the Melnik Fault before rejoining it near the border with Greece (Fig. 9). In this southern part of the basin, the lower part of the Sandanski Formation contains the Kromidovo mammal site (Fig. 9), with a diverse Turolian or 'Pikermian' fauna (see, e.g. Kojumdgieva et al. 1982, or Nikolov 1985, for species lists). Species include the hipparions Cremohipparion mediterraneum and C. matthewi, indicative of biozones MN11-13 (c. 9.0-5.4 Ma) and MNI2-13 (c. 8.2-5.4 Ma), respectively (e.g. Bernor et al. 1996), and the primate Mesopithecus pentelici, also from MN12-13 (e.g. Andrews et al. 1996). From its stratigraphic position, this site is assigned to biozone MN12 (c. 8.2-7.1 Ma) (e.g. Kojumdgieva et al. 1982). The upper part of the Sandanski Formation has yielded fish fossils indicative of brackish conditions (e.g. Kojumdgieva et al. 1982; Palmarev 1982), suggesting a correlation with the latest Meiotian marine transgression that reached as far north as the Serrai Basin (e.g. Gramman & Kockel 1969; Karistineos & Georgiades-Dikeoulia 1986). Higher up the sequence, near Melnik (Figs 9 and 10b, i), the Kalimantsi Formation has yielded the Pliocene mastodon Anancus arvernensis (e.g. Kojumdgieva et al. 1982; Nikolov 1985), indicating a significantly younger age. In the extreme south of this basin the stratigraphy becomes more problematic. Kojumdgieva et al. (1982) designated the Kalimantsi Formation stratotype along the road from Kalimantsi to Belyovo, in the Melnik Fault footwall and the Gorno Spanchevo Fault hanging wall (Fig. 10b, ii). This sequence passes upward from fluvial sands, similar to the Sandanski Formation elsewhere in the basin, to conglomerates, with four mammal sites (Kalimantsi 1 4 ; Fig. 10b, ii). Site Kalimantsi-3 has yielded Cremohipparion matthewi, again suggesting biozone MN12. However, Kalimantsi-1, with several characteristic Vallesian species (see, e.g. Kojumdgieva et al. 1982, or Nikolov 1985, for lists) including the rhinoceros A ceratherium zernowi (from MN9-10 according to Heissig 1996; see also Geraads et al. 2001), is considered significantly older. Kojumdgieva et al. (1982) deduced from the biostratigraphic ages that the sediments designated as the Sandanski and Kalimantsi Formations overlap, but Zagorchev (1992a) disputed this. The biostratigraphic evidence suggests that

570

(d)

R. W E S T A W A Y

LATE CENOZOIC EXTENSION, SW BULGARIA they do overlap in the extreme south of this basin, making it unfortunate that this is where the latter formation has been defined. As it would now appear that much of its stratotype lies within the age span of the Sandanski Formation, redefinition seems essential. However, the uppermost Kalimantsi Formation (which is important to the present study for dating the end of extension and/ or the start of regional uplift across this basin) seems to have been defined in an unambiguous manner, permitting correlation of it throughout the basin. The West Pirin Fault

At Ilindentsi (Figs 7 and 10), outcrop of the Ilindentsi member is observed (e.g. from [FM 86418 13202]) to be back-tilted east at c. 10 ~ A traverse is possible from here, at c. 400 m a.s.l., up the footwall escarpment of the West Pirin Fault to the source area of its clasts of Dobrostan Marble in a quarry at [FM 91449 16729], at c. 1100 m a.s.1. This c. 6 km traverse passes upward through fluvial deposits of the Kalimantsi Formation, with clasts predominantly of marble lower down and of Palaeogene granite higher up. Locally, fluvial units are observed to be separated by palaeosols; and their sequence is capped by gravel of inferred 'Eopleistocene' age that has not been offset by normal faulting (see Zagorchev 1992a). This pervasive cover of sediment makes the precise location of the footwall cutoff difficult to establish, and there is locally no geomorphological evidence of active slip on the West Pirin Fault. Zagorchev (1992a) estimated the total vertical slip on this fault as c. 3500 m, of which c. 100m was considered to post-date the

571

Kalimantsi Formation. However, it is unclear how this c. 100 m post-depositional offset was established, as these sediments are reportedly present only in the hanging wall of the fault. Between Ilindentsi and the Struma, the Sandanski Formation typically dips east at c. 10-15 ~ (e.g. Zagorchev & Dinkova 1990; Zagorchev 1992a, 1995; Fig. 9). As shown in Figure 10a, its eroded surface is terraced at m a n y different levels, designated as pediments or river terraces (see Zagorchev 1992a, 1995). No evidence could be observed of any faulted offsets of these levels. Near Ilindentsi, the footwall cutoff of the West Pirin Fault is observed c. 900 m a.s.1. Zagorchev (1992a, 1995) estimated by projection that the base of the sequence in the Sandanski Basin beside it is c. 1000 m below sea level; he correlated this palaeo-land surface with an erosion surface now c. 2500 m a.s.1, in the Pirin Massif. As a result, he partitioned the vertical slip on this fault with c. 1900 m of hanging-wall subsidence (below the present footwall cutoff) and c. 1600 m of footwall uplift. However, this estimate is clearly an upper bound to the total slip, as the correlation of erosion surfaces (see Zagorchev 1992a) seems intuitive and difficult to test. This fault was also examined farther south at Lyubovka (Fig. 7), around [FM 97208 05650]. It is locally mapped (Zagorchev & Dinkova 1990) as a c. 500 m wide fault zone separating the uppermost member of the Kalimantsi Formation from Palaeogene granite of the Central Pirin Pluton, with a local W S W dip of 40-50 ~ (Zagorchev 1992a, 1995; Fig. 10). There is locally a rather subdued escarpment bounding the granite, against which the sediments dip at c. 10~ This granite is highly weathered, suggesting that it has

Fig. 10. (a) View ENE across the Sandanski Basin from its western margin at [FM 82536 09606] near Mikrevo. Light patches near the skyline in the centre of the field of view are in situ exposures of the Dobrostan Marble in quarries. Light patches lower down the slope are olistostromes of the same marble within the Ilindentsi Member of the Kalimantsi Formation; they surround Ilindentsi village, which is barely visible. Village lower down to left of field of view is Strumjani. A gravel pediment (of inferred 'Eopleistocene' age; Zagorchev 1992a) also passes downslope onto the Ilindentsi Member. It is not cut by normal faulting, indicating that this part of the West Pirin Fault is no longer active. Subhorizontal benches visible at a variety of levels below Ilindentsi are presumed to be terraces of the Struma; no evidence was observed that they are separated by fault scarps. This photograph was kindly provided by I. Zagorchev. (b) Cross-sections through the southern Sandanski Basin, adapted from Kojumdgieva et aL (1982, figs 4 and 5). (i) Hotovo-Melnik; (ii) Kalimantsi-Belyovo (see Fig. 7 for locations). Both sections are along roads; their average orientations are indicated. In (ii) the conglomerate in the hanging wall of the Melnik Fault was assigned to the Kalimantsi Formation by Kojumdgieva et al. (1982) (K) and to the Katuntsi Formation in later studies by Zagorchev (Z) (see Fig. 9). (e) View northward to the roadcut exposure of the Gorno Spanchevo Fault at [GL 10582 98279]. Subhorizontally bedded conglomerate of the Kalimantsi Formation is juxtaposed, across a surface that dips west at c. 25~ against mylonitized and folded gneiss and amphibolite (with pegmatite veins) of the Vucha Formation of the Rhodopian Supergroup. This contact is sealed by subhorizontal gravel of inferred 'Eopleistocene' age, which is not visible in this view. (d) Field sketch (i) and interpretation (ii) of the locality in (b), adapted from part of Zagorchev (1995, fig. 25). Zagorchev interpreted the gneiss near the contact with the sediments as pervasively deformed by brittle fractures typically dipping to the west. My alternative impression was that fragments of the gneiss were becoming progressively looser nearer the contact, before weathering out into the overlying regolith. However, the section was evidently much less fresh at the time of my visit than when first examined by Zagorchev.

572

R. WESTAWAY

been locally exposed at the Earth's surface for a significant time. This evidence suggests that this fault is not active at present, although it was presumably important at an earlier stage when slip on it caused the observed tilting of the basin sediment.

The Melnik Fault The Melnik Fault is readily identifiable in the Melnishka river gorge c. 1 km N E of Melnik (Fig. 9), marked by a transition from Kalimantsi Formation conglomerate in its hanging wall to Sandanski Formation sandstone in its footwall. The beds on both sides of it locally dip consistently E N E at c. 5-10 ~ (see Fig. 10b, i), indicating back-tilting accompanying downthrow to the west. Both footwall and hanging wall are locally highly dissected by badland erosion, producing the 'Melnishki Piramidi' erosional landscape. The absence of clear differential dissection between the footwall and hanging-wall blocks suggests that little or no active slip is occurring on this fault at present. Although the absence of basement exposure makes it impossible to estimate the total slip on the Melnik Fault, its slip since the end of deposition of the Sandanski Formation can be roughly estimated. Mapping by Zagorchev & Dinkova (1990) indicates that c. 2 km east of this fault near Melnik, the top of this formation is at c. 500 m a.s.1, and dips away from this fault at c. 5 ~ It thus projects back to c. 700 m a.s.1, at the footwall cutoff (500 m + 2 0 0 0 m • tan(5~ Its c. 300 m a.s.1, altitude at the hanging-wall cutoff (Fig. 10b, i) thus indicates that the throw since this time has been c. 400 m. West of Melnik, the outcrop passes progressively down-section from the Kalimantsi Formation to the Sandanski Formation (Fig. 10b, i) and then the Delchevo Formation, with no evidence of further disruption by normal faulting. Numerous sections through palaeo-channels are evident locally, at up to c. 200 m above the Struma, inset into the Sandanski Formation and overlain by slope deposits, for instance at [FL 98268 97924] near Lozenitsa, c. 3 km SW of Melnik and c. 100 m above the present level of the Struma (c. 340 m a.s.1, against c. 240 m). These appear to typically trend SSE, subperpendicular to the modern tributary drainages (along which these channel sections are exposed), and evidently record progressive incision by the Struma to its present level. In the south, the stratigraphic relationships across this fault are not well resolved. Near Kalimantsi, Kojumdgieva et al. (1982) mapped it (1 in Fig. 9) with the Kalimantsi Formation in

both footwall to the east and hanging wall to the west (Fig. 10b, ii), although as noted above the former sediments may be part of the Sandanski Formation. Zagorchev & Dinkova (1990) and Zagorchev (1992a) located this fault several kilometres farther east (2 in Fig. 9) and interpreted it with the Katuntsi Formation to the west. This requires a complex pattern of local vertical crustal motions, which are inferred to have occurred on a mosaic of normal faults across this area, at positions indicated in Figures 7 and 9. It was not possible to resolve this discrepancy during my fieldwork in the area.

The Gorno Spanchevo Fault As already noted, this fault is interpreted along the eastern margin of the Sandanski Basin south of Lyubovka. I investigated the contact between Kalimantsi Formation and metamorphic basement of the Rhodopian Supergroup at several localites, it being most accessible near Gorno Spanchevo at c. 700 m a.s.1. [GL 10582 98579] on the road from Katuntsi to Gotse Delchev (Figs 9 and 10c,d). This basement is mylonitized, but the mylonitic fabric can be seen to not be parallel to the contact and has also been inferred to have a top-to-the-east sense of shear (e.g. Zagorchev 1995), so it could not be claimed as evidence of Late Cenozoic exhumation of the footwall from mid-crustal depths as a result of slip on a low-angle normal fault dipping to the west. Although this contact has been described as a normal fault plane in previous studies (with a dip of 25-30 ~ according to Zagorchev 1992b, or c. 15-20 ~ according to Burchfiel et al. 2000), it seemed to me to be more appropriately interpreted as an unconformity. The gneiss near this contact is highly weathered, the dark (biotite-rich) bands evidently more so than the light (quartzo-feldspathic) bands. Its top indeed seems to be a gradation, through progressively loose pieces of quartzo-feldspathic gneiss into a 'regolith' of fragments of this material. This is overlain by a c. 0.5 m thickness of poorly sorted slope wash, including clasts of granite and amphibolite, before passing into the uppermost Kalimantsi Formation. This contact is capped with gravel of inferred 'Eopleistocene' (i.e. Early Pleistocene) age, leading Zagorchev (1992a, 1995) to infer that slip had ended by then. Similar 'Eopleistocene' gravel was also noted by Zagorchev (1992a), sealing the southern part of this fault near the border with Greece (Fig. 9). However, from the evidence exposed at this roadcut site I could see no basis that it is a faulted contact at all; instead it seemed more likely that

LATE CENOZOIC EXTENSION, SW BULGARIA deposition of the Kalimantsi Formation progressively buried the weathered surface of the metamorphic basement. It is clearly not an active fault, and its appearance suggests that slip on it ceased before the end of deposition of the Kalimantsi Formation. The c. 25-30 ~ dip of this surface may simply indicate the profile of a weathered slope, with no relationship to the dip of any normal fault plane. About 800m farther NW, the contact between gneiss and Kalimantsi Formation (locally, red-weathered coarse conglomerate of rounded granite clasts) is also accessible at a lower level, c. 400 m a.s.l., in the Pirinska Bistritsa gorge just upstream of Gorno Spanchevo village (at [GL 09691 98880]). This contact is locally picked out by gulley erosion; the fault plane is not clearly visible, but must be steep (estimated >45 ~ dip), because if it had a low-angle dip it would be traceable upslope to the east given the extent of local dissection of the landscape. These red-weathered sediments (of the Upper Gravel Member of the Kalimantsi Formation) are not tilted consistent with normal faulting at this margin; they instead dip at c. 5~ towards the west, suggesting that slip on this fault had ended before they were deposited. This stratigraphic relationship can be observed in sections exposed as a result of recent incision by the Pirinska Bistritsa river, for instance at [GL 09471 97533] near Gorno Spanchevo, c. 1.3 km W N W of the locality in Figure 10c. Kojumdgieva et al. (1982) described a similar exposure c. 5 km farther south, where the road from Kalimantsi to Belyovo, which follows a tributary gorge of the Pirinska Bistritsa, crosses the Gorno Spanchevo Fault at c. 450 m a.s.1. (Fig. 10b, ii). This contact between gneiss and Kalimantsi Formation conglomerate (which I have not visited) was reported by Kojumdgieva et al. (1982) as a normal fault plane dipping west at c. 50~ However, the westward dip of this conglomerate is not apparent in Fig. 10b, ii that illustrates their interpretation of this area.

The Podgorie Fault Just north of the border with Greece, a right bank tributary, the Strumeshnitsa, joins the Struma near Petrich (Figs 7 and 9). This tributary valley, typically c. 4 km wide, is bounded to the south by the c. 2000 m high northern escarpment of the Belasitsa mountain range. This escarpment is c. 60 km long (east-west), its western half being in Macedonia. This valley has formed a depocentre for sediments that have been correlated with the Sandanski and Kalimantsi Formations, these

573

deposits now tilting to the south. This Belasitsa escarpment has been interpreted as the footwall of a significant active normal fault, the Podgorie Fault (see Zagorchev 1992a).

The Ograzhden Fault Along most of the western margin of the southern Sandanski Basin the Struma follows the contact between Neogene sediments and metamorphic basement (Figs 7 and 9). This basin margin has been depicted in many studies (e.g. Zagorchev 1992a,b) as an ENE-dipping antithetic fault. Around [FL 89780 96456], beside the Struma west of Nova Delchevo, the basement surface (interpreted as a gently basinward-sloping erosion surface of Mid-Miocene age; see Zagorchev 1992a,b, 1995) steepens at the basin margin to produce an escarpment up to c. 50 m high. Zagorchev (1992a) estimated the total slip on this fault as c. 500 m, of which c. 50-100 m was considered post-Pliocene. However, this escarpment is not particularly fresh-looking, and its height is less than the > 100 m by which the Struma, which flows at its base, is estimated to have incised since the Early Pleistocene (see below). It is thus possible that this escarpment has been formed by recent fluvial erosion, or is an inactive fault line scarp exposed by this fluvial incision, rather than indicating an active antithetic normal fault with a very low slip rate. This locality also provides clear exposures of the silt and tuffite of the Delchevo Formation, locally dipping at up to 25 ~ towards N52~ This disposition confirms that the main normal faulting affecting this basin has occurred along its eastern, or ENE, margin. The presence of these fine-grained sediments beside the Ograzhden Fault suggests that this fault was not yet active at the time, consistent with a later start of local extension, a view reinforced by the pockets of similar sediment on the eastern flank of the Ograzhden Massif in the footwall of this fault (see Kojumdgieva et al. 1982; Fig. 9).

Interpretations o f low-angle normal faulting Burchfiel et al. (2000) used evidence from the Sandanski Graben to infer large-scale extension in the Mid-Late Miocene. With regard to Ilindentsi, they wrote (pp. 332-333): 'A prominent layer of limestone breccia [was] emplaced in Pontian time . . . Limestones similar to those within the breccia.., are not present directly east of the Sandanski Graben and probably have a source many kilometres to the east. This suggests that the normal faults that bound the east side of the Sandanski Basin have many kilometres

574

R. WESTAWAY

of Middle-Late Miocene displacement.' These remarks are problematic; as well as being clear in the field (Fig. 10a), it is evident from geological maps (Marinova & Zagorchev 1990; Zagorchev & Dinkova 1990) and the local literature (e.g. Nedjalkov et al. 1986; Zagorchev 1992a, 1995, 2001a) that the Dobrostan Marble directly upslope from Ilindentsi was the source of the marble clasts Ilindentsi Member of the Kalimantsi Formation. Rather than requiring many kilometres of normal slip, the close proximity of source area and depocentre preclude this. In describing the Gorno Spanchevo Fault, Burchfiel et al. (2000) wrote (p. 333) 'the basin is bounded by a gently (c. 15-20 ~ west dipping normal fault that juxtaposes Middle Miocene sediments ... against ... basement of the Rhodope M a s s i f . . . Slickensides ... indicate an east-northeast-west-southwest direction of extension.' It is not clear how they identified slickensides; none were evident to me. Nor is it clear how they interpreted Middle Miocene sediment juxtaposed against basement, as the available mapping (e.g. Kojumdgieva et al. 1982; Zagorchev 1992a, 1995; Fig. 9) shows the uppermost Kalimantsi Formation in contact with basement throughout this structure.

The Razlog and Gotse Delchev Basins The Razlog and Gotse Delchev Basins are east of the the Rila and Pirin Massifs and are thus drained by the Mesta River (Fig. 7). They both cover part of the larger Palaeogene Mesta Basin (see Zagorchev 1995; Burchfiel et al. 2003), and the same stratigraphic terms are used for both these basins). The designated Neogene sequence begins with the Valevitsa Formation (basal conglomerate and sandstone). This is overlain by the Baldevo Formation, comprising interbedded siltstone, silty clay, fine sandstone, diatomite and lignite, in turn overlain by the Nevrokop Formation of fluvial conglomerate and sandstone (see Vatsev 1980). The Valevitsa Formation has yielded pollen of Pontian age (e.g. Vatsev 1980), and the Baldevo Formation has yielded Pontian plant remains (e.g. Palmarev 1970; Ivanov 1995) and diatoms (e.g. Temniskova & Ognjanova 1983) but its upper part has yielded algae of inferred Pliocene age (Zagorchev 1995). Conglomerates of the Nevrokop Formation, unlike those in the older formations, contain clasts of Palaeogene granite from the Pirin Massif (from the Teshovo and Central Pirin plutons), suggesting that it postdates their unroofing and thus correlates with the Kalimantsi Formation of the Sandanski Basin (Zagorchev 1995).

At Hadzhidimovo, the lower Nevrokop Formation has yielded characteristic Turolian mammals, including the hipparion Cremohipparion mediterraneum, the antelope Paleoreas lindenmayeri, and the four-tusked elephant Tetralophodon longirostris (Nikolov 1985). These species all spanned biozones M N l l - 1 3 (see Bernor et al. 1996; Gentry & Heizmann 1996; Lungu & Obada 2001). Zagorchev (1995) inferred a Pontian age (MN13; 7.1-5.4 Ma) for this site, whereas Geraads et al. (2001) placed it in early MN12 (c. 8.2-7.1 Ma). At Borovo, the upper Nevrokop Formation has yielded the mastodon Anancus arvernensis (see Nikolov 1985). Zagorchev (1995) considered this stratigraphic level to be Early Pliocene (Ruscinian), but this was a long-lived species spanning biozones MN14 to MN17 (i.e. the whole Pliocene) (see Athanassiou & Kostopoulos 2001; Lungu & Obada 2001) so it provides no strong constraint on the end of deposition. After sedimentation ceased (before the Early Pleistocene), the Mesta began to incise into the basin floor, forming a staircase of river terraces (see Nenov et al. 1972). Several studies (e.g. Ivanov 1995; Zagorchev 1995) have suggested that the three 'Formations' defined for the Gorse Delchev Basin interfinger with each other or pass laterally into each other, indicating a lateral facies variation from typical coarser sediment in the west adjacent to the sediment source in the Pirin Massif to finer sediment in more distal localities. The overlap in dates between sites assigned to the Baldevo and Nevrokop Formations (noted above) would seem to confirm this. The Razlog Basin has an irregular shape (Fig. 7); only its SSW margin appears to be bounded by a major normal fault (the Predela Fault; see below), elsewhere, the eroded margins of its Neogene sequence appear to lap onto Eocene terrestrial sediments of the Mesta Basin and older metamorphic basement. The Gorse Delchev Basin is more regular, c. 25 km long (north-south) and c. 8 km wide. However, many studies (e.g. Nenov et al. 1972; Ivanov 1995) have noted that its depocentre is typically not normal fault bounded: these sediments instead lap onto the basement at the basin margins. The maximum overall thickness of the Neogene deposits in this basin is c. 600 m (e.g. Ivanov 1995; Zagorchev 1995), rather less than in the Sandanski Basin. The Predela Fault

This fault (named by Zagorchev 1995; Meyer et al. 2002 called it the Bansko Fault) bounds

LATE CENOZOIC EXTENSION, SW BULGARIA the southern margin of the Razlog Basin, with typical N70~ strike. Its c. 800 m high footwall escarpment, rising to c. 2100 m a.s.l., is prominent in the field (Fig. 7) and on satellite images (Meyer et al. 2002). Its overall along-strike length is c. 30 km (see Meyer et al. 2002); its hanging wall forms the Predela col between the Pirin and Rila massifs; its western end adjoins the eastern end of the Krupnik Fault (see above). South of Bansko (Fig. 7), the moraine of an ice lobe emanating NE from Mt Vihren can be seen to be offset c. 10 m by this fault. As well as proving its Holocene activity and indicating a slip rate approaching 1 mm a -1, this locality thus provides a rare instance of interaction between glaciation and active normal faulting in the Aegean region. The Gotse Delchev Fault

No clear normal fault escarpment bounds the eastern margin and much of the western margin of the Gotse Delchev Basin. The clearest instance where its western margin is normal fault bounded is between Gotse Delchev and Musomishta. Here, as documented by Zagorchev (1995), two N60~ normal faults are arranged en echelon: the Musomishta Fault, dipping N N E at c. 50-70 ~ and with a c. 300 m high footwall escarpment, forms the contact between the Nevrokop Formation in its hanging wall and the Dobrostan Marble in its footwall. Approximately 1 km farther NNE, the subparallel Gotse Delchev fault offsets the Nevrokop Formation, with a r 100 m high footwall escarpment (Fig. 7). The fresh appearance of this escarpment (for instance at [GM 29891 05108], c. 1 km south of Gotse Delchev town), including characteristic faceted spurs and incised wineglass canyons, suggests that this normal fault is active, especially as the sediment exposed in this uplifting footwall is not fully lithified; this faulting clearly post-dates the deposition of this sediment.

Discussion Correlations between s e d i m e n t a r y sequences

Figure 11 indicates schematically how the sedimentary sequences in different basins in SW Bulgaria correlate. Its main differences compared with similar diagrams published previously (e.g. Burchfiel et al. 2000, fig. 6; Zagorchev 2001a, fig. 15) are the adoption of a clear time scale and the placing of the Meiotian-Pontian boundary for the Paratethyan realm at the TortonianMessinian boundary, consistent with most

575

modern quantitative chronostratigraphies (e.g. Steininger et al. 1996). One clear feature shown is the ending of sedimentation in all basins at c. 3 Ma. Since this time, with minor exceptions this part of Bulgaria has been subject to erosion, with stacked depositional sequences no longer developing. The principal exception is the stacked Late PliocenePleistocene sedimentation south and west of Sofia, around Radomir and Trun (Fig. 2). This part of Bulgaria is at the drainage divide between the south-flowing Struma, the Ishkur that flows NE from the Sofia area to the Danube, and the Nishava that flows northwestward across Serbia to the Morava, then northward to the Danube; being in headweaters, where rivers have little erosional power, this area has not experienced the hundreds of metres of recent fluvial incision typical elsewhere. This synchronous ending of sedimentation was previously noted by Zagorchev (2001a), although he assigned it a nominal age of c. 2 Ma rather than c. 3 Ma. It differs from the depiction by Burchfiel et al. (2000), who inferred (without providing supporting evidence) different timings for the cessation of sedimentation in different basins. As is clear from the descriptions above, this ending of sedimentation is not precisely dated in any basin, but there is no basis for assigning it a different age in different basins. As is discussed in more detail below, a nominal age of c. 3 Ma, rather than c. 2 Ma, is favoured for this event here, to match the Late Pliocene fluvial incision that is widely observed across western and central Europe (see Westaway 2002a). The observation that the extensional region of SW Bulgaria is almost entirely erosional marks one clear difference relative to western Turkey, where the hanging walls of the principal onshore active normal faults (bounding the Ala~ehir and Biiyiik Menderes grabens; Fig. 1a) are active depocentres. In both western Turkey and SW Bulgaria, vertical crustal motions caused by active normal faulting are superimposed onto regional uplift at comparable rates (see, e.g. Westaway et al. 2004; also below, where the regional uplift rate in SW Bulgaria is estimated as c. 0.15 mm a-l). This difference presumably relates to slower extension in SW Bulgaria. Extension rates are estimated below as no more than c. 1 mm a-~ on any of the active normal faults in SW Bulgaria. Assuming a 45 ~ dip, 1 mm a -1 of slip on a normal fault in an erosional region could be partitioned with 90% footwall uplift and 10% hanging-wall subsidence relative to the uplifting regional reference frame. So, with regional uplift at c. 0.15 mm a -~, the footwall and hanging wall would uplift at 1.05 mm a -1

576

R. WESTAWAY Sandanski Basin

Simitli Basin

Blagoevgrad Basin

Kyustendil Basin

Sofia Basin

G o t s e Delchev Basin

0-

I

Pleistocene

2_

MN17

1.8

oooooo.....,

-R-o.m a n~i a6n --3 MN16 ~o 4-

.M. .m. 4

Dacian 4

e

Pontian

MN13

7.1

w i11 r3- 8 -

.,12

--

9.9

[-

O~

9

9 ~ 9

9

Kalimants', F m

Barakovo Fm.

Koilishka Fro.

O

O

9

9

O

O

O

O

O

O

9

~

~-~ r-~ C::

1:2 1=3 i:~ r~ . . . . . . . . . . . . . "

"

" , ~

" ~

"

.

.

9

9 9 " " " . ". ". ". O O O j'." ." ..... I~ . . . . . . ~" . . . . . I ~_~it. Fm ~ D z b e . , a n F m ,

a ~ ' - :

9 ~

"~

14MN7

9

9

9 9 -- N-~i -I-'~u~-'#'m.-" " " . ". ". ". ". "." ..... Gniljane Fm.m~ . . . . . -------" " "

O

O

o

o

~ w ~u~.r

--

O

'~,ariegated

-~d-evo-~m.-

terrigenous"

9

9

9

~/

~/

~/

."

~

9149149 Fm.

Basal

9

Red

9

Cglt 9

.

.

.

9

.

.

.

.

.

9 9 9 Valevitsa Fin.

.

.

9 .

9

:~

13.6 ~

9

9

9 9 " " PPokrovnik o k r o v n i k FFm m . ~ L. I . . . . . . Chernichevo Fm.~. 9 9 ~

"--

o !li

9

SkrinyanoFm.

9

MN8

9

O ~ . . ~ . _ _ _ _ ~ ! - ~ - ~

_Sandanskil=m--lGradevo~'m~--~_ --"

9

//eez - - ~ n .e n ~ ' ~ ~ ' ~9

9

O

.

Sarmatian

9

O

MN9 ~ - - - - ~ ' - ~ - I ~ D e l c h e v o Fm.

--

O~

O

9 9 . 0 MN10 ~ " " . .~ . 9.5~ . . . . " " 10-

~ O

O

"

MN11

/ O

Kalimantsi Fm.

8.2-MeioUan

/O

Spasovitsa

Katuntsi F m . ~ _

9

e

Fro.

9

Badenian MN6 9

Silt,

marl,

diatomite

o

o

O

Conglomerate, with c l a s t s of Palaeogene

o

o

o

granite

Lignite

9 9

9 9

9 9

Other conglomerate

9

.

9

. . . . . . . 9 ~ ~ .

.

~

.

.

1::2 I:= I:= [:2 I::= [:= [:~ I:~

Sandstone

Coarse

marble

breccia

Fig. 11. Stratigraphic correlation diagram for the Late Cenozoic terrestrial sedimentary basins of SW Bulgaria. Slope deposits and river terraces are omitted. Sources of information for most of these basins are discussed in the text. The stratigraphy for the Sofia Basin is from Kamenov & Kojumdgieva (1983), with mammalian biostratigraphic control from Nikolov (1985). The stratigraphy for the Kyustendil Basin is from Vatsev & Bonev (1994), also with mammalian biostratigraphic control from Nikolov (1985). As noted in the text, the Katuntsi Formation of the Sandanski Basin is problematical; it may be indistinguishable from the Kalimantsi Formation (see Kojumdgieva et al. 1982). The age bounds for the terrestrial Paratethyan stages and mammalian biozones, which are constrained by magnetostratigraphy and Ar-Ar dating, are from Steininger et al. (1996), and are thus unaffected by the revision of the 'marine' Paratethyan chronology proposed by Vasiliev et al. (2004).

and 0.05 mm a -~, respectively. In contrast, using geodetic data from McClusky et al. (2000), Westaway et al. (2004) estimated the extension rate across the Ala~ehir graben as c. 6 mm a -~, superimposed onto regional uplift at c.0.2 mm a -~. Partitioning the normal fault-related vertical motions as before would now indicate absolute hanging-wall subsidence. However, on other, less active, normal faults in western Turkey, hanging walls are experiencing absolute uplift (e.g. Westaway 1993; Purvis & Robertson, 2004; Westaway et al. 2004, 2005), as in SW Bulgaria.

In contrast, the starts of sedimentation differ widely between basins. Several sequences start with thin basal conglomerates that seem to have accumulated slowly over long periods of time before sedimentation rates increased significantly; others begin with stacked sequences of silt (the Delchevo Formation in the Sandanski Basin) or lignite (the Oranovo Formation of the Simitli Basin). Although a general coarsening upward is apparent (Fig. 11), some sequences are dominated by clastic input, whereas others, notably in the Gotse Delchev and Sofia basins, were

LATE CENOZOIC EXTENSION, SW BULGARIA lacustrine basins, where rhythmic alternations of deposition of lignite, diatomite and other sediments indicate fluctuations in environmental conditions. Thorough high-resolution chronostratigraphic studies have been carried out in the apparently analogous rhythmic Messinian-Early Pliocene sequence in the Servia-Ptolemais Basin of NW Greece (Fig. la) (e.g. van Vugt et al. 1998, 2001; Steenbrink et al. 1999, 2000). These studies indicate Milankovitch forcing of the sedimentary rhythmicity, under the dominant influence of c. 20 ka precession cyclicity, with lignite deposition at times of cool summers (i.e. summer insolation minima, when the Earth's orbit was oriented with aphelion during the northern hemisphere summer) and marl or diatomite deposition at times of warmer summers. This implies that palaeo-lakes were typically deeper when summers were warmer, implying higher summer precipitation, as is generally accepted for the northern margin of the eastern Mediterranean (e.g. Rohling & Hilgen 1991). Rhythmicity in the Bulgarian lacustrine sequences has instead been explained in terms of alternations between 'silting up', marked by lignite deposition, and renewed subsidence, marked by deepening of water and deposition of diatomite (e.g. Ognjanova & Yaneva 2001). These Bulgarian lake basins would be good targets for future high-resolution cyclostratigraphic studies. Structural and geodetic estimates o f slip rates for the present phase o f extension This description indicates that the principal active normal faults in SW Bulgaria strike west (between WSW and NW; Figs 3, 7 and 9). If this region accommodates uniaxial extension on these faults, then the extension direction is roughly north-south. This set of subparallel, en echelon normal faults continues farther north, a notable additional member being the north-dipping Sofia fault (Fig. 2), whose footwall forms the c. 1200 m high northern escarpment, rising from c. 700 m to c. 1900 m a.s.l., of the Vitosha mountain range south of Sofia. Burchfiel et al. (2000) regarded the Sofia Basin as effectively marking the northern limit of the Aegean extensional province. Figure lb illustrates the crustal velocity field in SW Bulgaria, measured by Kotzev et al. (2001) using the Global Positioning System (GPS). GPS point BERK, north of the Sofia Basin (Fig. 2), is the most southerly site in stable Eurasia, delimiting the northern margin of the Aegean extensional province. The progressive increase in southward velocity that is observed geodetically

577

to occur southward from this point is evidently the result of the cumulative slip on the various east-west-striking normal faults in the region, including the Sofia, Kyustendil, Saparevo, Rila, Krupnik, Predela, Podgorie and Gotse Delchev faults. The cumulative extension across this array of en echelon normal faults can thus be estimated to account for the observed (Fig. lb) c. 3 mm a -1 of southward motion of the southern margin of western Bulgaria relative to stable Eurasia. Rates of southward motion increase dramatically farther south across Greece, as illustrated by the velocity vector at SOXO in Figure lb and by the data presented by McClusky et al. (2000, fig. 2). Taking account of the heights and dips of footwall escarpments and the thicknesses of hanging-wall fill, the most important active normal faults in this array can be estimated to have taken up 2-3 km of extension. As a result, the total extension along a north-south line across the Sofia, Saparevo, Krupnik and Podgorie faults can be estimated as c. 10 km. Dividing this into the geodetic extension rate gives an estimated age of this phase of faulting of 3-4 Ma. As already noted, the start of this phase of extension is not well constrained directly, largely because of the uncertainties over dating the Pliocene sediments in the region. Burchfiel et al. (2000) estimated a similar (c. 3.5-4 Ma) timing of the start of this phase of extension, but this seems to have been based on arguments regarding a preceding phase of hypothetical lowangle normal faulting (see Dinter & Royden 1993), which no longer appear appropriate (see below); its agreement with the numerical age estimate in the present study may be coincidental. Recent syntheses (Westaway 2003, 2004a) place the start of the present phase of right-lateral slip on the North Anatolian Fault Zone (Fig. 1a), during which it has been conjugate to the leftlateral East Anatolian Fault Zone (EAFZ), around 4 Ma. The NAFZ is estimated in these schemes to have first developed around 7 Ma, but was initially conjugate to the left-lateral Malatya-Ovaclk Fault Zone (MOFZ) located north of the modern EAFZ. During these two phases of slip, the Euler pole to the N A F Z seems to have been located in different places (see Westaway 2004b), so one may well expect the kinematics of regions near its western end to have changed significantly around 4 Ma. Constraints on the earlier extension In the Burchfiel et al. (2000) scheme, extension is presumed to have begun in the southern Sandanski Basin in the early Badenian stage of

578

R. WESTAWAY

the Mid-Miocene (c. 15 Ma), the deposits of the Delchevo Formation being presumed by them to be synextensional. This phase of extension was considered to be oriented NE-SW, by analogy with apparent synchronous low-angle normal faulting in northern Greece (see Dinter & Royden 1993; Dinter et al. 1995). Burchfiel et al. (2000) deduced that the zone of extension expanded in the Early Meiotian (c. 9 Ma) to affect the whole Sandanski Basin, the Blagoevgrad, Gotse Delchev, Razlog and Sofia basins, and other basins located outside the present study region. However, as already noted there is no direct evidence that the Delchevo Formation was deposited during crustal extension, and some evidence that it was not. The Delchevo Formation is tilted eastward at up to 25 ~ (see above). From Zagorchev & Dinkova (1990), the Sandanski Formation dips at up to 25 ~ near its base and typically at 10-15 ~ near its top (Fig. 9), whereas the Kalimantsi Formation typically dips at up to c. 10~ near its base but can be subhorizontal at its top. The similarity in tilt between the Delchevo Formation and the lower Sandanski Formation suggests that extension began early during deposition of the latter. Evidence already discussed (e.g. Fig. 10c and d) suggests that the end of slip on the Gorno Spanchevo Fault preceded the end of deposition of the Kalimantsi Formation. The fluvial gravels of inferred 'Eopleistocene' age, which 'seal' the normal faults at the eastern margin of the Sandanski Basin, confirm that these faults ceased to be active before the Early Pleistocene. Zagorchev (1992a) suggested an alternative explanation: that so much extension occurred on the Gorno Spanchevo Fault that it became back-tilted to such an extent that it was no longer suitably oriented relative to the stress field, so the Melnik Fault formed in its hanging wall with a steeper dip. This adjustment process is observed in many parts of the Aegean region that have extended significantly, such as along the Ala~ehir and Biiyiik Menderes fault zones in western Turkey (e.g. Westaway 1998; Purvis & Robertson 2004). However, from Zagorchev's (1992a) estimate, the heave across the West Pirin Fault is only moderate, at most c. 3.5 km x cot(50 ~ or c. 3 km. It is unlikely to have been much greater at the southern end of the Sandanski Basin, so severe back-tilting on any normal fault seems unlikely. Slip on this set of faults is now presumed to have ended by c. 4 Ma given the timing of the start of the younger phase that superseded it (see above). This normal fault system, oriented NNW-SSE, was thus unsuited to take up N N W - S S E extension when this began at c. 4 Ma.

Regional kinematics

The evidence suggests that three distinct phases of extension have occurred in SW Bulgaria from the Late Miocene to the present day. The start of sedimentation in several basins (Fig. 11) suggests that in each case local extension began in the early Meiotian stage of the Late Miocene. Extension at this time is presumed to have been towards the WSW or SW, accommodated on NNW-SSE-striking normal faults, such as the West Rila (Fig. 3) and West Pirin (Figs 7 and 9) faults. This timing matches the earliest evidence for the 'present' phase of extension in NW Turkey from this region's earliest alkali basaltic volcanism, which is dated to c. 10 Ma in the Tekirdafg and (~anakkale areas (Fig. l a) in the vicinity of the modern Sea of Marmara (see review of this dating evidence by Westaway et al. 2005). At this time, the Sea of Marmara itself did not exist, because the N A F Z had not yet developed. Moreover, before the subsequent many tens of kilometres of right-lateral slip accumulated on the NAFZ, Tekirdafg and (~anakkale would have been closely juxtaposed. At this time, it thus appears that extension may have affected only a limited part of the present Aegean region, in N W Turkey and southern Bulgaria (Fig. 12b). It is suggested that this initial phase of extension was caused by incipient 'rollback' of the oceanic slab that was subducting beneath the southern margin of the Aegean region; beforehand, it is presumed that the downdip length of this slab was insufficient to drive this process (Fig. 12a). As noted above, it is inferred that the various pre-Meiotian sedimentary deposits in SW Bulgaria (Fig. 11) are not extension related. Some of these deposits (e.g. the Oranovo Formation) are localized, and the depocentre of the Delchevo Formation was evidently cut through by later normal faulting. It thus evident that when active normal faults began to develop in this region some of them cut through pre-existing depocentres that existed for other reasons. Such a geometry, of sediments accumulating in depocentres that are unrelated to normal faulting, but that later become overlain by hanging-wall sedimentary fill, is widely observed in western Turkey (e.g. Ko~yi[git et al. 1999; Bozkurt 2000; Yllmaz et al. 2000). It has indeed caused many problems with trying to date the start of extension, as studies (e.g. Seyitofglu & Scott 1992; Seyito~lu et al. 1992) have regarded this older sediment as part of the synextensional sequence, causing the start of extension to be placed too early in the record. By analogy, it seems likely that the start of extension in SW Bulgaria was no earlier than the Meiotian.

LATE CENOZOIC EXTENSION, SW BULGARIA early Middle Miocene to J

579

/

Early Tortonian to

-e -e

X\

"-'-.

%

x\

\

I )

Messinian to Early

Pliocene/

~

, :

-

"

..

"

\

"~,-

\

~

EarlyPliocene to present /

Fig. 12. Schematic crustal velocity fields at key stages in the evolution of the Aegean region during the Late Cenozoic, consistent with the present study. (a) In the Mid-Miocene, with no extension yet occurring in the Aegean region. Northward relative motion of the Arabian plate relative to Africa and Eurasia was already occurring, but the Dead Sea Fault Zone died out into a complex zone of transpression in Syria and SE Turkey and a broad zone of distributed shortening in eastern and central Turkey (see Westaway 2003, 2004a). (b) In the early Late Miocene, with slow extension inferred in parts of NW Turkey (from volcanism; see Westaway et al. 2005) and analysis of SW Bulgaria (this study). (e) The deformation accompanying the initial phase of NAFZ slip, during c. 7-4 Ma, accommodated by NNW tapering of WSW extension across Bulgaria. The timing of this phase is constrained by arguments in the main text. Velocities during this phase have been scaled to the same rate of relative motion between the Turkish plate and Eurasia as at present (see McClusky et al. 2000), although as discussed in the text the typical contemporaneous velocities may have been lower. (d) The deformation accompanying the present phase of NAFZ slip, since c. 4 Ma, accommodated by southward extension across Bulgaria and by westward tapering of southward extension across central Greece. This velocity field is essentially a schematic version of the results of McClusky et al. (2000). Their results are well known to be consistent with the NAFZ kinematics; they have been shown by Westaway (2003, 2004a) to be consistent with the active strike-slip faulting in SE Turkey; and are now shown to also be consistent with the active faulting in Bulgaria. (See text for discussion.)

In the Early Pontian (c. 7 Ma), extension in SW Bulgaria seems to have spread more widely (for instance, the deposition of the lacustrine Gniljane Formation suggests that extension began at this time in the Sofia Basin; Fig. 11). This effect was noted by Burchfiel e t al. (2000), who inferred much more widespread extension in

the Pontian (their fig. 9) than in the Meiotian (their fig. 8). Extension also evidently spread westward, as indicated by the 6.9 Ma start of lacustrine deposition (Steenbrink et al. 2000) in the Servia-Ptolemais basin of northern Greece (Fig. la). The typical coarsening of the sediment at this time in the depocentres of SW Bulgaria

580

R. WESTAWAY

(Fig. 11) also suggests a higher-energy environment, with faster erosion in footwall localities, implying an increase in structural relief and thus in slip rates (see Zagorchev 1992a). However, this time marked the onset of the Messinian salinity crisis in the Mediterranean region, so some change in sedimentary environments may reflect contemporaneous climate change (for instance, increased aridity may have reduced the vegetation cover, leading to increased rates of erosion). However, it seems obvious that the emplacement at this time of the marble olistostromes of the Ilindentsi Member of the Kalimantsi Formation in the Sandanski Graben required significant local topographic gradients for the first time, from which it can be inferred that the footwall of the West Pirin normal fault was uplifting at a significant rate relative to the adjacent hanging wall. At this time, extension in SW Bulgaria seems to have continued, as before, in the ENE-SSW direction. Recent interpretations (Westaway 2003, 2004a; Westaway et al. 2005) regard c. 7 Ma as a key point in the tectonic evolution of Turkey, marking the starts of the initial phase of slip on the N A F Z (conjugate to the MOFZ in the east; see Westaway & Arger 1996, 2001) and of the 'present' phase of north-south extension across most of western Turkey. Robertson et al. (2004) also deduced a c. 7 Ma switch from crustal shortening to strike-slip in the NE corner of the Mediterranean Sea, in good agreement with these estimates. As discussed by Westaway. (2006), the clearest evidence now available for this timing of extension in western Turkey has been developed from the dating to c. 6.7 Ma of the extension-related volcanism in the Denizli region (see Westaway et al. 2005) and from thermochronological evidence (from Lips et al. 2001) indicating a c. 7 Ma start of rapid slip on normal faults bounding the Ala~ehir graben (Fig. l a). Westaway (2003, 2004a) has suggested that the start of this phase of extension was synkinematic with the start of slip on the NAFZ, both processes having possibly been triggered by the change in state of stress in the crust that accompanied the drawdown in sea level in the Mediterranean basin at the start of the Messinian salinity crisis (see calculations by Westaway 2003). It follows that, at this time, coupling via the rightlateral slip on the N A F Z for the first time caused kinematic linkage between the pre-existing convergent zone in eastern Turkey (Fig. 12a and b) and the Aegean extensional province (Fig. 12c). Finally, at a later stage, estimated above as c. 4 Ma, the extension in SW Bulgaria changed from E N E - W S W to NNW-SSE or northsouth (Fig. 12d). The major NNW-SSE-striking

normal faults in this region, such as the West Rila and West Pirin faults, were no longer suitably oriented to accommodate the extension, and became superseded by more optimally oriented normal faults, such as the Kyustendil, Rila, Krupnik and Podgorie faults. However, some pre-existing normal faults that were oblique to the earlier ENE-WSW extension were evidently also oblique to the subsequent NNW-SSE extension, and so could remain active; these include the Sofia, Saparevo, Predela and Musomishta faults (Figs 3 and 7). This array of roughly west-eaststriking active normal faults seems to persist southward into northern Greece: similarly oriented faults there bound the Langadas graben (having slipped in the M = 6.4 Thessaloniki earthquake sequence in 1978, accommodating northsouth extension; Soufleris et al. 1982; Tranos et al. 2003) as well as the Serrai, Drama, and Xanthi grabens (Fig. la). The M = 6 . 6 Grevena earthquake sequence in 1995, farther west in northern Greece (Fig. l a), also involved northsouth extension (e.g. Rigo et al. 2004). By analogy, the poorly documented Plovdiv (Fig. 2) earthquake sequence in central Bulgaria in 1928 (mainshock M = 7 ; see Richter 1958) probably also involved north-south extension. This estimated c. 4 Ma timing corresponds to the end of slip on the MOFZ, when the N A F Z propagated eastward and the EAFZ developed conjugate to it (see Westaway & Arger 1996, 2001; Westaway 2003, 2004a). Its timing is estimated by dividing the total slip on the EAFZ by its slip rate; Westaway et al. (2006) have constrained this timing to 3.73 +_0.05 Ma by this method. As Westaway (2004b) noted, it is difficult to make predictions for how the NAFZ behaved before this time, because of the possibility that the Euler vector for the motion relative to Eurasia of the Turkish plate to the south of it may have differed for its two phases of slip. At present, this Euler pole is located near the SE corner of the Mediterranean Sea near the Suez Canal (see McClusky et al. 2000). If during the previous phase it was several hundred kilometres farther south and west, making it more distant from the NAFZ, the resulting adjustment in deformation sense can explain the change in the extension direction in SW Bulgaria (Fig. 12c and d). Recent studies of the Sea of Marmara pullapart basin, where, during its present slip phase, the main active strand of the N A F Z steps to the right (Fig. 1a), shed some light on the duration of its present geometry. Localized right-lateral slip occurs on the N A F Z at a rate of 18 ___4 mm a-' (Hubert-Ferrari et al. 2002). Probably 80% of this, or c. 15 mm a -~, is taken up across the Sea

LATE CENOZOIC EXTENSION, SW BULGARIA of Marmara, the rest occurring on subparallel strands farther south (Fig. la; see Armijo et al. 1999). Armijo et al. (1999) estimated that this northern N A F Z strand has slipped by c. 60 km since its present geometry developed, which would suggest initiation at c. 60/c. 15 or c. 4 Ma. However, using different reasoning, Okay et al. (2004) obtained a revised lower bound to its estimated slip of c. 40 km, whereas using a different argument Seeber et al. (2004) estimated that it has slipped no more than 28 kin. Attempts to date the start of this slip directly using local sedimentary evidence have led to further controversy, because the ages of the sedimentary units and their relationships to this faulting have been disputed (see Tfiysfiz et al. 1998; Armijo et al. 1999; Yaltlrak et al. 2000). It is thus clear that no consensus yet exists regarding which local evidence provides the best estimate of the start of the present phase of N A F Z slip, but an age of c. 4 Ma cannot be precluded. The present geometry (Fig. 12d) achieves kinematic consistency between the slow southward or SSW velocities (at c. < 10 mm a q) that develop across Bulgaria and northern Greece and the 35-40 mm a -1 SSW velocities observed south of the western end of the N A F Z (see McClusky et al. 2000; Kotzev et al. 2001). Such consistency requires rapid southward or SSW extension across the en echelon set of major active normal faults in central Greece, including the faults bounding the Gulf of Corinth, the Parnassos mountain range, the Sperchios basin and adjacent Gulf of Evvia, the north coast of Evvia, and the SE end of the Thermaic Gulf and SW end of the North Aegean Trough (Fig. 1a). The required extension rate increases eastward from zero in the west to c. 25-30 mm a -~ along a line between the eastern Gulf of Corinth and the intersection between the Thermaic Gulf and North Aegean Trough. This westward tapering in extension rates will require clockwise rotation of the Peloponnese block to the south (Fig. la) relative to regions farther north. The importance of the active normal faulting in this region for maintaining kinematic consistency between Aegean extension and right-lateral slip on the NAFZ, and for generating clockwise rotations that are observed palaeomagnetically, was recognized long ago (e.g. McKenzie & Jackson 1983). However, in their scheme the normal-fault-bounded blocks were envisaged as like slats attached to pivots at both ends, which means that the predicted extension and rotation do not vary laterally. In contrast, the present scheme resembles the opening of a fan about a pivot in the west, with extension increasing from west to east and clockwise rotation increasing from north to south.

581

Between c. 7 and c. 4 Ma the right-lateral slip on the N A F Z is inferred to have been accommodated by N N W tapering in the SSW extension across Bulgaria (Fig. 12c). This geometry would also result in clockwise rotation, which would have increased from east to west; it indeed resembles the 'classic' geometrical interpretation of such palaeomagnetic evidence for the western Aegean region (see Kissel & Laj 1988). The observed palaeomagnetic dataset indicating systematic clockwise rotation across the western half of this extensional province (see Kissel & Laj 1988) would thus appear to relate in part to each of the deformation senses in Figure 12c and d, rather than requiring a single mechanism. In contrast, in western Turkey the crustal velocity field is predicted to have remained essentially the same after 4 Ma as before (Fig. 12c and d), consistent with the absence of evidence for a change in the deformation sense at this time. During the initial phase of N A F Z slip, the geometry (Fig. 12c) suggests that the N A F Z slip rate should equal the maxinmm rate of WSW extension along a line directly north of the N A F Z and the maximum rate of WSW rollback (relative to Eurasia) of the surface trace of the Hellenic subduction zone. In contrast, during the present phase, the maximum rate of SSW extension across Bulgaria and northern Greece plus the N A F Z slip rate should roughly equal the ma• mum rate of WSW rollback (relative to Eurasia) of the surface trace of the Hellenic subduction zone. From two points of view the present deformation sense can be considered more 'effective' than its predecessor. First, it allows faster rollback of the Hellenic subduction zone for a given N A F Z slip rate. As the length of subducted slab increases, the dynamics favours faster rollback (see Meijer & Wortel 1997), potentially forcing this change in deformation sense as a mechanically 'easier' alternative than forcing a faster N A F Z slip rate. Second, the geometry in Figure 12d avoids the requirement in Figure 12c for rapid E N E - W S W crustal shortening north of the western end of the subduction zone. Such shortening would lead to crustal thickening and, thus, growth of topography, affecting the regional stress field so as to ultimately oppose the driving mechanism. However, detailed calculations regarding both these potential causes of the c. 4 M a reorganization of the kinematics are beyond the scope of this study. Local versus regional vertical crustal motions

Across most of the Aegean region, the view became established in the 1980s (see Jackson

582

R. WESTAWAY

et al. 1982; Jackson & McKenzie 1988) that verti-

cal crustal motions during extension have been caused only by active normal faulting. This is despite an abundance of evidence to the contrary, notably from Turkey, for regional uplift, onto which local effects of active normal faulting have been superimposed (see summary of this evidence by Demir et al. 2004). It is now clear that c. 400 m is a representative typical value for the regional uplift in western Turkey since the Early Pliocene, of which c. 150 m has occurred since the start of the Mid-Pleistocene (see Westaway 1993; Westaway et al. 2003, 2004). The early MidPleistocene marked a general increase in uplift rates at localities in temperate latitudes worldwide (e.g. Kukla 1975, 1978; Westaway 2002a), apparently linked to coupling between surface processes (e.g. increased erosion rates, cyclic loading by ice sheets) caused by long time-scale climate change (as the climate system adjusted from predominant c. 40 ka to c. 100 ka climate cyclicity; see Mudelsee & Schulz 1997) and the isostatic uplift response that is mediated by flow in the lower continental crust. An apparently similar increase in uplift rates also occurred around 3 Ma (see van den Berg & van Hoof 2001; Westaway 2001, 2002a), but is less well resolved because of the more limited datasets from that time. However, despite the evidence to the contrary, other studies (e.g. Bunbury et al. 2001) continue to repeat the view that vertical crustal motions in western Turkey relate only to active normal faulting. The recent publication (by Allen et al. 2004a) of the extraordinary claim that there is no evidence of uplift in Turkey since the Miocene provoked a strong reaction (Westaway 2004b) but even so was not retracted (Allen et al. 2004b). Given this history of dispute in western Turkey, it is noteworthy that it is well established that in SW Bulgaria local vertical crustal motions as a result of Late Cenozoic normal faulting have been superimposed onto regional uplift (e.g. Zagorchev 1992a,b). The local literature indeed contains extensive discussion of erosion surfaces in the mountain massifs, which have been warped and offset by normal faulting and dissected to progressively lower levels by fluvial incision in response to this regional uplift. The youngest part of this history of regional uplift is revealed by the terrace staircases of the Struma and Mesta rivers (Table 1), which are well developed in the Sandanski and Gotse Delchev basins and elsewhere. These terraces seem to correlate well with the principal cold stages from the latest Early Pleistocene (oxygen isotope stage (OIS) 22; Shackleton et al. 1990) to

the Late Pleistocene. These terrace staircases thus resemble those of the major rivers in Turkey (see Demir et al. 2004), but differ from those in central and western Europe where often every cold stage is represented, sometimes with multiple terraces per climate cycle (e.g. Westaway 2002a). Of the two, the Struma terrace staircase is taken as a better proxy for regional uplift, indicating c. 110 m of uplift since OIS 22. This is, first, because the Struma is a significantly larger river and so likely to be better able to incise fully in response to regional uplift and, second, because of the absence of slip on the normal faults bounding the Sandanski Basin since c. 4 Ma. To constrain the less well-resolved earlier part of the uplift history, additional data points are added. The first comes from the pediment of 'Eopleistocene' (i.e. Early Pleistocene) gravel that seals the West Pirin normal fault above Ilindentsi (see above). At this fault, this tributary gravel is at c. 940 m a.s.l., but it descends towards the Struma at a gradient of c. 4 ~ reaching as low as c. 500 m a.s.1. (c. 300 m above present river level) (Zagorchev 1992a). Zagorchev (1995) also reported lower pediments inset into it, at c. 480400, 360-320 and c. 270-220 m a.s.1. The last of these presumably grades into one of the youngest Struma terraces and the third into the terrace level that has been assigned to OIS 22 (Table 1) (see Galabov 1982; Zagorchev 1992a). However, the c. 400 m pediment provides a second additional data point, indicating c. 200 m of incision. Next, it was estimated above that near the active Krupnik normal fault the Struma has incised since 3 Ma at a time-averaged rate of c. 0.08 mm a -1 in the hanging wall and that the lower bound to the footwall incision rate has been c. 0.18 mm a -~. The average of these two values, c. 0.13 mm a -l, is taken as representative of the component of regional uplift, implying c. 400 m of uplift since 3 Ma. It is clear that this is a very crude calculation, but nothing better seems possible at this stage given the extent of uncertainty regarding the incision history of this footwall (discussed earlier). Finally, the average altitude, between footwall and hanging-wall cutoffs, of the top of the Sandanski Formation at Melnik is c. 500 m (see above). Local evidence (the brackishwater sedimentation in close proximity to the marine sedimentation in the Serrai Graben in the Meiotian, when these two depocentres were evidently interconnected; see above) indicates that this sediment was deposited near sea level. Allowing for possible net glacioeustatic sea-level fall, c. 450 m of net uplift can thus be estimated since c. 7 Ma. This is broadly consistent with the estimate from the Kresna gorge, also suggesting that regional uplift was slow between c. 7 and c.

LATE CENOZOIC EXTENSION, SW BULGARIA

583

Table 1. Altitudes of Struma and Mesta river terraces Terrace T1 T2 T3 T4 T5 T6 T7

Nominal age

Struma altitude (m)

Late Pleistocene Late Pleistocene Mid-Pleistocene Mid-Pleistocene Early Pleistocene Early Pleistocene Eopleistocene

5-7 8-12 20-22 40-45 60-65 85-100 -

Mesta altitude (m) 8-12 18-24 28-30 40-45 60 80-90 100-110

Nominal altitude (m)

Preferred OIS

6 10 21 40 63 90 110

2 4 6 8 12 16 22

Terrace altitude data (above present river level) are from the compilation by Zagorchev (1995), based on Nenov et al. (1972) and Galabov (1982). Previously assigned terrace 'ages' use the Russian definition of the Pleistocene.

In this scheme, the Eopleistocene is equivalent to the international Early Pleistocene (i.e. from c. 1.8 Ma to c. 780 ka or OIS 19), the Early Pleistocene is equivalent to the international early Mid-Pleistocene (i.e. to OIS 12), and the Mid-Pleistocene is equivalent to the international late Mid-Pleistocene (i.e. to OIS 6). Nominal altitude means the altitude considered representative for each terrace, used in the uplift modelling in Figure 13. Preferred OIS is the preferred oxygen isotope stage to which each terrace is assigned as a result of this modelling.

3 Ma. The high degree of erosion since the Late Miocene-Early Pliocene and the resulting obliteration of so much former sedimentary evidence clearly makes it difficult to estimate precise amounts of incision, and thus uplift, on this time scale. This is another point of similarity to recent investigations of this topic in western Turkey (see Westaway et al. 2004). The possibility was also considered of using the c. 600 m a.s.1, present-day altitude of the Oligocene shallow marine sediment in the Padezh Basin (Fig. 2) as an uplift constraint. However, the strong tilt of these sediments makes it difficult to select any particular altitude datum for them. None the less, these sediments do suggest much slower uplift rates during the Miocene than since, consistent with the evidence from the Sandanski Basin. Their low altitudes also preclude the idea, held by some workers, that before its extension the Aegean region was a high plateau analogous to modern Tibet. To constrain the associated uplift history, these data are modelled using the technique of Westaway (2001) (see also Westaway et al. 2002). This calculates the isostatic response to forcing of lower-crustal flow by cyclic loading at the Earth's surface. It is used here to model, as an approximation, an isostatic uplift response that is probably mainly the result instead of variations in erosion rates; but, as noted by Westaway (2002b), these two distinct processes can induce very similar uplift responses. The results (Fig. 13) can be compared with modelled uplift histories for western Turkey, such as Westaway et al. (2004, fig. 21). The total uplift of c. 400 m estimated since the Early Pliocene is similar in both regions. However, in western Turkey this seems to be partitioned with c. 150 m

of uplift since the late Early Pleistocene and c. 200-250 m of uplift during the Late Pliocene and early Early Pleistocene. In SW Bulgaria, the proportion during the later of these two phases is lower, c. 110 m, and that in the earlier phase correspondingly higher. A priori, faster erosion was expected in SW Bulgaria than in western Turkey, as the former has been more severely glaciated during cold stages of the Pleistocene (see Demir et al. 2004). The modern profuse vegetation in SW Bulgaria, which is expected to inhibit erosion of the unlithified Late Cenozoic sediments, will of course have died back during cold stages. The lower uplift rates estimated for the Mid-Late Pleistocene thus at first sight appear surprising. This modelling indeed suggests substantial erosion rates; for instance, the Sandanski Basin has been incised by c. 300 m since c. 2 Ma at a time-averaged rate of c. 0.15 mm a -1, whereas the spatial average erosion rate for western Turkey seems to be c. 0.1 mm a -1 (see Westaway 1994; Westaway et al. 2004). However, only a small proportion of the present study region in SW Bulgaria, perhaps c. 20%, is occupied by eroding Late Cenozoic sedimentary basins (Fig. 2), whereas the proportion is rather higher in western Turkey (see Westaway et al. 2004). It can thus be inferred that the overall spatial average erosion rate for SW Bulgaria, calculated as a weighted average of c. 0.15 mm a -1 for the basins and a rather lower value for the mountain massifs, is less than the c. 0.1 mm a -1 spatial average value for western Turkey; hence the lower uplift rates in the Mid-Late Pleistocene. The c. 3 Ma start of this initial phase of uplift is considered to reflect global climate change and to thus have no connection with the c. 4 Ma

584

R. WESTAWAY Struma river terraces, SW Bulgaria: Uplift history

50o-

~

Kresna gorge

400 300

schematically in Fig. 11). This is consistent, for instance, with the deduction that the uppermost Kalimantsi Formation post-dates the end of slip on the Gorno Spanchevo fault (Fig. 10c and d).

River terraT:: pediment' ~ 8 ~ ~

~9 200 ~ ~ _ [

[]

~

'

~

T h e p o s s i b l e role of l o w - a n g l e n o r m a l

'400m pediment'

faulting

:~ 100

0.0

0.5

1.0 1.5 2.0 2.5 Time t before present (Ma)

Ca)

3.0

3.5

$truma river terraces, SW Bulgaria: Uplift history ?50 t E

-1001 50

o 0.0

lT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.2

(b)

0.4 0.6 Time t before present (Ma)

0.8

Struma river terraces, SW Bulgaria: Predicted uplift rates

~,~0.20 E ~0.16 0.12

~

0.08

0.0tl

0.0

(C)

0.5

1.0 1.5 2.0 2.5 Time t before present(Ma)

3.0

3.5

Fig. 13. Uplift histories for the Struma in and around the Sandanski Basin, matched to the observational evidence from river terraces, pediments and gorge incisxon. Calculations follow the method of Westaway (2001) and Westaway et al. (2002), and are based on the following parameter values (defined in these references): Zb 15 km; z~27 km; c 20~ km-~; 1.2 mm2 s - l ; to,~ 18 Ma, ATe.E -10~ to,23.1 Ma, ATe,2-15.3~ to,3 0.9 Ma, ATe,3-6.5~ (a) Predicted uplift history and supporting data for the Pliocene and Quaternary; (b) enlargement of (a) showing the late Early Pleistocene onwards; (e) predicted variation in uplift rates for the same time scale as in (a). Solution predicts 112 m of uplift since 870 ka (OIS 22), 119 m since OIS 25, and 400 m since 3.1 Ma. Peak uplift rates are 0.194 mm a-~ at 2.45 Ma and 0.149 mm a-1 at 0.35 Ma. reorganization of Aegean kinematics. However, the relative timing of these two events is not well constrained. None the less, the evidence suggests that the incision post-dates the kinematic reorganization, implying that sedimentation continued in the Sandanski Basin and elsewhere after slip ceased on the bounding normal faults (as shown

The possibility of large-scale extension in SW Bulgaria before or during the Late Miocene, on normal faults that formed at dips of c. 30~ or less and took up tens of kilometres of extension, exhuming the mountain ranges in their footwalls from mid-crustal depths as metamorphic core complexes, has been debated. On the one hand, studies such as those by Dinter & Royden (1993), Dinter et al. (1995), Shipkova & Ivanov (1999, 2000, 2001) and Burchfiel et al. (2000, 2003) have argued this, based on diverse structural and thermochronological evidence. On the other hand, their claims have been repeatedly disputed, notably by Zagorchev (1994, 1998a, 2001b). Part of this dispute relates to granites. Zagorchev (1995, 1998a,b) has distinguished three characteristic granite intrusion ages in SW Bulgaria: Hercynian, Late Cretaceous (e. 90 Ma), and Oligocene (c. 35 Ma), each with distinct geochemical, petrological and structural characteristics (for instance, the ones considered younger show less evidence of deformation), supported by isotopic dating (Fig. 7). However, others (e.g. Dinter & Royden 1993; Dinter et al. 1995) have argued that intrusion of all these granites was synkinematic with large-scale Oligocene or Miocene crustal extension. In such a view, the presence of Hercynian or Cretaceous zircons in some granites can be explained by reworking of this stable mineral into Cenozoic magmas. Exactly the same disputes have occupied the recent literature on western Turkey: whether ancient zircons are in situ or reworked (see Logs & Reischmann 1999, 2001); whether Ar-Ar dates for minerals with low closure temperatures in granite plutons indicate intrusion ages or cooling ages (see Westaway 1996, 2005); and whether such cooling was caused simply by erosion or by 'tectonic denudation' by low-angle normal faulting (see Hetzel et al. 1995; Lips et al. 2001; Ring et al. 2003; Westaway 2006). My own recent modelling (Westaway 2006) suggests that the available thermochronologic dataset for western Turkey can be well explained as a result of perturbations to the geothermal gradient caused by erosion and by past changes to the geometry of subduction. However, although such results are consistent with the available evidence, they cannot be proven: one has no way of knowing

LA 1tz CENOZOIC EXTENSION, SW BULGARIA what rates of erosion or geometries of subduction were in the Early-Mid-Cenozoic; one can only make estimates, for input into numerical models. The arguments by Shipkova & Ivanov (1999, 2000, 2001), that Late Cenozoic low-angle normal faulting on their 'Dzherman detachment' was synkinematic with intrusion of the Kalin pluton, depend entirely on the assumption that this normal fault zone is analogous to others where this combination of processes has been claimed (e.g. by Dinter & Royden 1993; Dinter et al. 1995) to have occurred; no new evidence has been offered. Their arguments seem in my view to not be worth considering further (see Zagorchev 2001b). The papers by Dinter & Royden (1993), Dinter et al. (1995) and Burchfiel et al. (2000) initially developed the idea of low-angle normal faulting from evidence in Greece, and then applied this idea to SW Bulgaria. Of the two sites in the Sandanski Basin considered by Burchfiel et al. (2000) to reflect low-angle normal faulting, at one (Ilindentsi) their interpretation is clearly wrong, being based on a mistake regarding the local geology (Fig. 10a). The second, the Gorno Spanchevo Fault, is clearly steep for much of its length (see above, also Fig. 9). Moreover, given the standard vertical shear construction (Westaway & Kusznir 1993), the c. 25 ~ tilting of the oldest sediments in the Sandanski Basin means that fault surfaces with present-day dips of 30 ~, 40 ~ or 50 ~ would restore, respectively, to 46 ~ 53 ~ or 59 ~ plausible initial dips of a conventional steep normal fault. The roadcut site at Gorno Spanchevo (Fig. 10c and d) is admittedly problematic, but it does seem inappropriate for Burchfiel et al. (2000) to have emphasized the apparent low-angle fault dip inferred at this one site rather than the preponderance of other evidence that contradicts the inference of such a low-angle dip. Arguably the most compelling reasoning for low-angle normal faulting in the present study region comes from the analysis by Dinter et al. (1995) of the cooling history of the Symvolon pluton near Kavala (see Kyriakopoulos et al. 1996; Fig. la). They regarded this cooling as accompanying large-scale SW extension on a low-angle normal fault, their 'Strymon detachment', which they inferred as synkinematic with the Gorno Spanchevo Fault. Those workers deduced rapid cooling of this inferred footwall from c. 750~ to c. 150~ during c. 25-15 Ma from U-Pb dating of zircon and titanite and ArAr dating of hornblende, biotite and K-feldspar. However, their dates for zircon (closure temperature Tc 750~ of c. 25 Ma depend on the isotope ratio analysed; one sample yielded a z~ age as low as 24.4+0.7 Ma but several yielded

585

2~176 ages of up to c. 300 Ma, consistent with a Hercynian intrusion age (see Kokkinakis 1980; Zagorchev, 1998a). The U-Pb dating of titanite includes even greater systematic errors between numerical ages from different isotope ratios. Furthermore, K-feldspar has complex closure behaviour as a result of its complex microstructure; the Tc of 150~ adopted by Dinter et al. (1995) is a lower bound that is appropriate only for very slow cooling (see McDougall & Harrison 1999). If one removes the data for these three isotopic systems one is left with a dataset indicating cooling from c. 500~ at c. 20 Ma (Ar closure in hornblende) to c. 350~ at c. 15 Ma (Ar closure in biotite). This cooling history closely resembles what is observed (in datasets by Hetzel et al. 1995; Lips et al. 2001; Ring et al. 2003) for the central Menderes Massif in western Turkey (between the Alasehir and Biiyfik Menderes grabens; Fig. l a). Such a cooling history can be explained as a consequence of exhumation by erosion while the region was simultaneously being cooled from below by incipient subduction at a low angle (Westaway 2006), thus removing any basis for inferring low-angle normal faulting in the first place. Recent studies (e.g. Kounov et al. 2001, 2004; Burchfiel et al. 2003) have begun to focus on the possibility of large-scale Palaeogene (EoceneOligocene) extension by low-angle normal faulting in SW Bulgaria. This is in accordance with a recent trend in western Turkey, whereby researchers (e.g. Purvis & Robertson 2004) have accepted that the Late Cenozoic extension occurred on initially steep normal faults, but have allowed the possibility of a different geometry of normal faulting at an earlier stage. Kounov et al. (2001) suggested from thermochronological data that the Osogovo Mountains (Fig. 2) experienced Palaeogene extension by low-angle normal faulting. However, their own data have no simple interpretation in terms of this process. In the 'conventional' geological literature, these mountains are interpreted (see Zagorchev 1995, 200 l a) as having been intruded by Hercynian granite, then overthrust probably in the Late Cretaceous, then subjected to prolonged erosion. Burchfiel et al. (2003) argued for a phase of Eocene to Early Oligocene ENE-WSW extension on their 'Mesta Detachment', which they regarded as contemporaneous with their inferred extension across the Padezh Basin (see above) and the Oligocene extrusive volcanism in the Mesta Basin (Fig. 7). However, the supporting evidence that they presented was rather limited; it amounts to another claim that angular breccia marks a lowangle normal fault rather than possibly indicating slope processes (see Shipkova & Ivanov 1999,

586

R. WESTAWAY

2000, 2001; see above), and an assertion that tilting of beds must be due to extension, when local literature (e.g. Zagorchev 1992a) includes multiple phases of deformation in different senses, which could have caused this tilting. As this literature already includes a range of possible interpretations of the Palaeogene evolution of this area (see Zagorchev 1992a, 1998b, 2001a), it would seem more productive to begin further investigation of this topic by testing these existing hypotheses rather than proposing entirely new ones on the basis of limited evidence.

Conclusions Since the Early Pliocene (c. 4 Ma), SW Bulgaria has accommodated southward or SSE extension at several millimetres per year, superimposed on c. 400 m of post-Early Pliocene regional uplift. This sense of deformation superseded earlier extension, oriented E N E - W S W , which is estimated to have begun in the early Late Miocene (c. 10-9 Ma) and lasted until c. 4 Ma, the regional topography being dominated by N N W - S S E striking normal fault escarpments and grabens that are relics from this time. Normal faults that are now active cut across these older structures, although in some localities normal faults that were oriented obliquely to the earlier extension have remained active, also oblique to the modern extension sense. It is suggested that this present phase of extension relates to the modern sense of deformation throughout the Aegean region and to the modern geometry of the N A F Z , which is independently inferred to have existed since c. 4 Ma. The earlier E N E - W S W extension is inferred to have involved two phases, the first pre-dating the N A F Z and the second synkinematic with its initial phase of slip during c. 7-4 Ma, when its geometry and the overall sense of deformation throughout the Aegean region were different from at present. Some previous studies have inferred that SW Bulgaria experienced large-scale extension on low-angle normal faults in the Mid-Miocene or earlier. However, the limited evidence in support of this view is open to other interpretations, and after due consideration can be discounted. I thank I. Zagorchev and R. Nakov for helpful discussions and guidance in the field, and M. Coltorti and M. Tranos for thoughtful and constructive reviews. This study contributes to IGCP 449 'Global correlation of Late Cenozoic fluvial sequences' and to IGCP 518 'Fluvial sequences as evidence for landscape and climatic evolution in the Late Cenozoic'.

References ALLEN, M., JACKSON,J. & WALKER,R. 2004a. Late Cenozoic reorganization of the Arabia-Eurasia collision and the comparison of short-term and long-term deformation rates. Tectonics, 23(1), TC2008, doi 10.1029/2003TC001530. ALLEN, M., JACKSON, J. • WALKER, R. 2004b. Reply to comment by R. Westaway on 'Late Cenozoic reorganization of the Arabia-Eurasia collision and the comparison of short-term and long-term deformation rates' by M. Allen, J. Jackson, & R. Walker. Tectonics, 23(5), TC5007, doi 10.1029/ 2004TC001695. AMBRASEYS,N. N. 2001. The Kresna earthquake of 1904 in Bulgaria. Annali de Geofisica, 44, 95-117. ANDREWS, P., HARRISON, T., DELSON, E., BERNOR, R.L. & MARTIN, L. 1996. Distribution and biochronology of European and southwest Asian Miocene catarrhines. In: BERNOR, R. L., FAHLBUSCH, V. & MITTMANN, H.-W. (eds) The Evolution of Western Eurasian Neogene Mammal Faunas. Columbia University Press, New York, 168-207. ARMIJO, R., MEYER, B., HUBERT, A. & BARKA, A. 1999. Westward propagation of the North Anatolian fault into the northern Aegean: timing and kinematics. Geology, 27, 267-270. ATHANASSIOU, A. & KOSTOPOULOS, D. S. 2001. Proboscidea of the Greek Pliocene-Early Pleistocene faunas: biochronological and palaeoecological implications. In: CAVARETTA,G., GIOlA, P., MUSSl, M. & PALOMBO,M. R. (eds) The Worm of Elephants: Proceedings of the 1st International Congress, Rome, 16-20 October 2001. Consiglio Nazionale delle Ricerche, Rome, 85-90. BERNOR,R. L., KOUFOS,G. D., WOODBURNE,M. O. & FORTELIUS,i . 1996. The evolutionary history and biochronology of the European and Southwest Asian Late Miocene and Pliocene hipparionine horses. In: BERNOR, R. L., FAHLBUSCH, V. & MITTMANN, H.-W. (eds) The Evolution of Western Eurasian Neogene Mammal Faunas. Columbia University Press, New York, 307-338. BONCHEV, G. 1912. Prinos kum petrografyata i mineralogyata na Rila Planina. Spisanie na Bulgarska Academiya na Naukite, 2, 1-176. BOTEV, E., DIMITROV,D. & GEORGIEV,1. 2001. Principal tectonic stress tensor in the region of the Kroupnik Fault from seismic and geodetic data. Geologica Balcanica, 31, 92-93. BOZKURT,E. 2000. Timing of extension on the Btiytik Menderes Graben, western Turkey, and its tectonic implications. In: BOZKURT, E., WINCHESTER,J. A. & PIPER, J. D. A. (eds) Tectonics and Magmatism of Turkey and the Surrounding Area. Geological Society, London, Special Publications, 173, 385-403. BUNBURY, J. M., HALL, L., ANDERSON, G. J. & STANNARD, A. 2001. The determination of fault movement history from the interaction of local drainage with volcanic episodes. Geological Magazine, 138, 185-192. BURCHFIEL, B. C., NAKOV, R., TZANKOV, T. & ROYDEN, L. H. 2000. Cenozoic extension in Bulgaria and northern Greece; the northern part of

LATE CENOZOIC EXTENSION, SW BULGARIA the Aegean extensional regime. 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, 325-352. BURCHFIEL, C. B., NAKOV, R. & TZANKOV, T. 2003. Evidence from the Mesta half-graben, SW Bulgaria, for the Late Eocene beginning of Aegean extension in Central Balkan Peninsula. Tectonophysics, 375, 61-76. DEMIR, T., YE~iLNACAR, I. & WESTAWAY, R. 2004. River terrace sequences in Turkey: sources of evidence for lateral variations in regional uplift. Proceedings of the Geologists' Association, 115, 289-311. D1NTER, D. & ROYDEN, L. 1993. Late Cenozoic extension in north-eastern Greece: Strymon valley detachment system and Rhodope metamorphic core complex. Geology, 21, 45-48. DINTER, D., MACFARLANE,A., HAMES,W., ISACHSEN, C., BOWRING, C. & ROYDEN, L. 1995. U-Pb and 4~ geochronology of the Symvolon granodiorite: implications for the thermal and structural evolution of the Rhodope metamorphic core complex, north-eastern Greece. Tectonics, 14, 886-908. ERiN~, S. 1978. Changes in the physical environment in Turkey since the end of the last glacial. In: BRICE, W. C. (ed.) The Environmental History of the Near and Middle East since the Last Ice Age. Academic Press, London, 87-110. FURLAN,D. 1977. The climate of southeast Europe. In: WALLEN, C. C. (ed.) Climates ofCentraland Southern Europe. World Survey of Climatology, Volume 6. Elsevier, Amsterdam, 185-235. GALABOV, Z. 1982. Geography of Bulgaria. Physical Geography. Bulgarian Academy of Sciences, Sofia. GENTRY, A. W. & HEIZMANN,E. P. J. 1996. Miocene ruminants of the central and eastern Tethys and Paratethys. In: BERNOR, R. L., FAHLBUSCH,V. & Mn-TMANN, H.-W. (eds) The Evolution of Western Eurasian Neogene Mammal Faunas. Columbia University Press, New York, 378-391. GERAADS, D., SPASSOV,N. & KOVACHEV, D. 2001. New Chalicotheriidae (Perissodactyla, Mammalia) from the Late Miocene of Bulgaria. Journal of Vertebrate Paleontology, 21, 596-606. GRAMMAN, F. & KOCKEL, F. 1969. Das Neogen im Strimonbecken (Griechisch-Ostmazedonien)--I, Lithologie, Stratigraphie und Pal~iogeographie. Geologisches Jahrbuch, 87, 445-484. HEISSIG, K. 1996. The stratigraphical range of fossil rhinoceroses in the Late Neogene of Europe and the Eastern Mediterranean. In: BERNOR, R. L., FAHLBUSCH, V. & MITTMANN, H.-W. (eds) The Evolution of Western Eurasian Neogene Mammal Faunas. Columbia University Press, New York, 339-347. HETZEL, R., RING, U., AKAL, C. & TROESCH, M. 1995. Miocene NNE-directed extensional unroofing in the Menderes Massif, southwestern Turkey. Journal of the Geological Society, London, 152, 639-654. HUBERT-FERRARI, A., ARMIJO, R., KING, G., MEYER, B. & BARKA,A. 2002. Morphology, displacement, and slip rates along the North Anatolian Fault,

587

Turkey. Journal of Geophysical Research, 107(B10), JB2235, doi: 10.1029/2001JB000393. IVANOV, D. 1995. Palynological data on fossil flora from the village of Ognjanovo (southwestern Bulgaria). Phytologia Balcanica, 2, 3-14. JACKSON, J. m. & MCKENZIE, D. P. 1988. Rates of active deformation in the Aegean Sea and surrounding regions. Basin Research, 1, 121-128. JACKSON, J. A., KING, G. C. P. & VITA-FINZI,C. 1982. The neotectonics of the Aegean: an alternative view. Earth and Planetary Science Letters, 61, 303-318. KAMENOV, B. & KOJUMDGIEVA,E. 1983. Stratigraphy of the Neogene of Sofia Basin. Palaeontologiya, Stratigrafiya i Lithologiya, 18, 69-85 (in Bulgarian with English summary). KARISTINEOS, N. K. & GEORGIADES-DIKEOULIA,E. 1986. The marine transgression in the Serres Basin. Annales Gdologiques des Pays Helleniques, 33, 221-232. KlSSEL, C. & LAJ, C. 1988. The Tertiary geodynamical evolution of the Aegean arc: a palaeomagnetic reconstruction. Tectonophysics, 146, 183-201. Ko(~Yi~iT, A., YUSUFO~LU,H. & BOZKURT, E. 1999. Evidence from the Gediz graben for episodic twostage extension in western Turkey. Journal of the Geological Society, London, 156, 605-616. KOJUMDGIEVA, E., NIKOLOV, I., NEDJALKOV, P. & BUSEV, A. 1982. Stratigraphy of the Neogene in Sandanski Graben. Geologica Balcanica, 12, 69-81. KOKKINAKIS, A. 1980. Altersbeziehungen zwischen Metamorphosen, mechanischen Deformationen und Intrusionen am Sfidrand des RhodopeMassivs (Makedonien, Griechenland). Geologische Rundschau, 69, 726-744. KONYAROV, G. 1932. Kafyavite vuglisha v Bulgaria. Durzhavni Mill, Pernik. KOTZEV,V., NAKOV,R., BURCHFIEL,B. C., KING, R. & REILINGER, R. 2001. GPS study of active tectonics in Bulgaria; results from 1996 to 1998. Journal of Geodynamics, 31, 189-200. KOUNOV,A., SEWARD,D., BERNOULLI,D., BURG,J.-P. & IVANOV,Z. 2001. Timing of Cenozoic extension in the Kraishte region (SW Bulgaria): evidence from fission track analysis. Geologica Balcanica, 31, 112-113. KOUNOV, A., SEWARD, D., BERNOUILLI, D., BURG, J.-P. & IVANOV,Z. 2004. Thermotectonic evolution of an extensional dome; the Cenozoic OsogovoLisets core complex; Kraishte Zone, Western Bulgaria. International Journal of Earth Sciences, 93, 1008-1024. KUKLA, G. J. 1975. Loess stratigraphy of central Europe. In: BUTZER, K. W. & ISAAC, G. L. (eds) After the Australopithecines. Mouton, The Hague, 99-188. KUKLA, G. J., 1978. The classical European glacial stages: correlation with deep-sea sediments. Transactions of the Nebraska Academy of Sciences, 6, 57-93. KYRIAKOPOULOS, K., MAGGANAS, A., NORELLI, O., BIGAZZI, G., DEE MORO, A. & KOKKINAKIS, A. 1996. Thermochronological evolution of Symvolon and Pangeon plutons and their country rocks, Kavala area, N. Greece: an apatite fission track

588

R. WESTAWAY analysis. Neues Jahrbuch fiir Monatshefte, 1996(11), 519-529.

Mineralogie,

LIPS, A. L. W., CASSARD,D., SOZBILiR,H., YILMAZ,H. & WIJBRANS, J. R. 2001. Multistage exhumation of the Menderes Massif, western Anatolia (Turkey). International Journal of Earth Sciences, 89, 781-792. Loos, S. & REISCHMANN, T. 1999. The evolution of the southern Menderes Massif in SW Turkey as revealed by zircon dating. Journal of the Geological Society, London, 156, 1021-1030. t o o s , S. & REISCHMANN,T. 2001. Reply to comment by Bozkurt, E. and Park, R.G., on 'The evolution of the southern Menderes Massif in SW Turkey as revealed by zircon dating'. Journal of the Geological Society, London, 158, 394-395. LUNGU, A. & OBADA, T. 2001. Contributions to the study of the Neogene representatives of ordo Proboscidea (Mammalia) from eastern Europe. ln: CAVARETTA, G., GIOIA, P., MussI, M. & PALOMBO, M. R. (eds) The Worm of Elephants: Proceedings of

the 1st International Congress, Rome, 16-20 October 2001. Consiglio Nazionale delle Ricerche, Rome, 119-121. MARINOVA, R. 1991. Geological Map of the People's

Republic of Bulgaria, 1:100 000 series, Blagoevgrad sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MAR1NOVA, R. & ZAGORCHEV,I. 1990. Geological Map

of the People's Republic of Bulgaria, 1:100 000 series, Razlog sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MCCLUSKY, S., BALASSANIAN, S., BARKA, A., et al. 2000. Global Positioning System constraints on plate kinematics and dynamics in the eastern Mediterranean and Caucasus. Journal of Geophysical Research, 105, 5695-5719. McDOUGALL, I. & HARRISON, T. M. 1999. Geochro-

nology and Thermochronology by the 4~ Method, 2nd ed. Oxford University Press, Oxford. MCKENZIE, D. & JACKSON, J. 1983. The relationship between strain rates, crustal thickening, palaeomagnetism, finite strain and fault movements within a deforming zone. Earth and Planetary Science Letters, 65, 182-202. MEIJER, P. T. & WORTEL, M. J. R. 1997. Present-day dynamics of the Aegean region: a model analysis of the horizontal pattern of stress and deformation. Tectonics, 16, 879-895. MEYER, B., ARMIJO, R. & DIMITROV, D. 2002. Active faulting in SW Bulgaria; possible surface rupture of the 1904 Struma earthquakes. Geophysical Journal International, 148, 246-255. MOSKOVSKI, S. 1983. Certains particularit~s des sediments pal6og6nes et plio-pl6istoc~nes dans les parties moyennes de la vall6e de Struma. Rkunion

Extraordinaire de la Sociktd GOologique de France, Guide de l'Excursion, 105-108. MUDELSEE, M. & SCHULZ, M. 1997. The MidPleistocene climate transition: onset of 100 ka cycle lags ice volume build-up by 280 ka. Earth and Planetary Science Letters, 151, 117-123. NEDJALKOV, P., TCHEREMISIN,N., KODUMDGIEVA,g., TZATZEV, B. & BUZEV, A. 1986. Facial and paleogeographic features of Neogene deposits

in the Sandanski graben. Geologica Balcanica, 18, 61-66. NENOV, T., SLAVOV, I. & STOYKOV, S. 1972. Pliocene and Quaternary in the Gotse Delchev depression and principal stages in its neotectonic development. Review of the Bulgarian Geological Society, 33, 195-203 (in Bulgarian with English summary). NIKOLOV, I. 1985. Catalogue of the localities of Tertiary mammals in Bulgaria. Palaeontologiya, Stratigrafiya i Lithologiya, 21, 43-61 (in Bulgarian with English summary). OGNJANOVA, N. & YANEVA, M. 2001. New data about Baldevo Formation, Gotse Delchev Basin, based on sedimentological and biostratigraphical evidences. Geologica Balcanica, 31, 12%128. OKAY, A. I., TI)YSOZ, O. & KAYA, ~. 2004. From transpression to transtension: changes in morphology and structure around a bend on the North Anatolian Fault in the Marmara region. Tectonophysics, 291, 259-282. PALMAREV, E. 1970. Fossile floren aus drei Braunkohlenbecken in Sfidwestbulgarien. Izvesti na

Botanicheskiya Institut Bulgarska Academiya na Naukite, 20, 35-79 (in Bulgarian with German abstract). PALMAREV, E. 1982. Fosilnata flora na Melnishkiya basein. Palaeontologiya, Stratigrafiya i Lithologiya, 16, 3-44. PURVIS, M. & ROBERTSON, A. H. F. 2004. A pulsed extension model for the Neogene-Recent E-W trending Ala~ehir Graben and the NE-SW trending Selendi and G6rdes Basins, western Turkey. Tectonophysics, 391, 171-201. RICHTER, C. F. 1958. Elementary Seismology. Freeman, San Francisco, CA. RIGO, A., DE CHABALIER,J.-B., MEYER, B. & ARMIJO, R. 2004. The 1995 Kozani-Grevena (northern Greece) earthquake revisited; an improved faulting model from synthetic aperture radar interferometry. Geophysical Journal International, 157, 727-736 RING, U., JOHNSON, C., HETZEL, R. & GESSNER, K. 2003. Tectonic denudation of a Late CretaceousTertiary collisional belt: regionally-symmetric cooling patterns and their relation to extensional faults in the Anatolide Belt of western Turkey. Geological Magazine, 140, 421-441. ROBERTSON, A. H. F., ()NLI3GENG, U. C., iNAN, N. & TA~LI, K. 2004. The Misis-Andmn Complex: a Mid-Tertiary m61ange related to late-stage subduction of the Southern Neotethys in S Turkey. Journal of Asian Earth Sciences, 22, 413-453. ROHLING, E. J. & HILGEN, F. J. 1991. The eastern Mediterranean climate at times of sapropel formation: a review. Geologie en Mijnbouw, 70, 253-264. SEEBER, L., EMRE, O., CORMIER, M.-H., et al. 2004. Uplift and subsidence from oblique slip: the Ganos-Marmara bend of the North Anatolian transform, western Turkey. Tectonophysics, 391, 239-258. SEYITOGLU, G. & SCOTT, B. 1992. The age of the Biiyfik Menderes graben (west Turkey) and its tectonic implications. Geological Magazine, 129, 239-242. SEY]TO~LU, G., SCOTT, B. & RUNDLE, C. C. 1992. Timing of Cenozoic extensional tectonics in west

LATE CENOZOIC EXTENSION, SW BULGARIA Turkey. Journal of the Geological Society, London, 149, 533-538. SHACKLETON, N. J., BERGER, A. & PELTIER, W. R. 1990. An alternative astronomical calibration of the lower Pleistocene timescale based on ODP site 677. Transactions of the Royal Society of Edinburgh, Earth Sciences, 81, 251-261. SHIPKOVA, K. & IVANOV, Z. 1999. The Djerman detachment fault; a result of late Tertiary extension in the north-western parts of the Rhodope Massif, Bulgaria. European Union of Geosciences Conference Abstracts, EUG 10. Journal of Conference Abstracts, 4, 470. SHIPKOVA, K. & IVANOV, Z. 2000. The Djerman detachment fault; an effect of the late Tertiary extension in the north-west part of the Rhodope Massif. Dokladi na Bulgarskata Akademia na Naukite, 53, 81-84. SHIPKOVA, K. & IVANOV, Z. 2001. Effects of Late Alpine extension in the northwestern foot of Rila Mountain. Geologica Balcanica, 31, 138-139. SOUFLERIS, C., JACKSON, J. A., KING, G. C. P., SPENCER, C. P. & SCHOLZ, C. H. 1982. The 1978 earthquake sequence near Thessaloniki (northern Greece). Geophysical Journal of the Royal Astronomical Society, 68, 429-458. STEENBRINK, J., VAN VUGT, N., HILGEN, F. J., WIJBRANS, J. R. & MEULENKANP, J. E. 1999. Sedimentary cycles and volcanic ash beds in the lower Pliocene lacustrine succession of Ptolemais (NW Greece); discrepancy between 4~ and astronomical ages. Palaeogeography, Palaeoclimatology, Palaeoecology, 152, 283-303. STEENBRINK, J., VAN VUGT, N., KLOOSTERBOER-VAN HOEVE, M. L. & HIL6EN, F. J. 2000. Refinement of the Messinian APTS from sedimentary cycle patterns in the lacustrine Lava section (Servia Basin, NW Greece). Earth and Planetary Science Letters, 181, 161-173. STEININGER, F. F., BERGGREN, W. A., KENT, D. V., BERNOR, R. L., SEN, S. ~; AGUSTI, J. 1996. CircumMediterranean Neogene (Miocene and Pliocene) marine-continental chronologic correlations of European mammal units. In: BERNOR, R. L., FAHLBUSCH, V. & MITTMANN, H.-W. (eds) The Evolution of Western Eurasian Neogene Mammal Faunas. Columbia University Press, New York, 7-46. TEMNISKOVA, D. t~ OGNJANOVA, N. 1983. Siliceous algae from fresh-water Neogene diatomites in the Gotse Delchev area. Fitologiya, 22, 29-45 (in Bulgarian with English summary). TRANOS, M. D., PAPADIMITRIOU, E. E. & KILIAS, A. A. 2003. Thessaloniki-Gerakarou Fault Zone (TGFZ): the western extension of the 1978 Thessaloniki earthquake fault (northern Greece) and seismic hazard assessment. Journal of Structural Geology, 25, 2109-2123. TI)YSI3Z, O., BARKA, A. & Yi~ITBA~, E. 1998. Geology of the Saros Graben and its implications fror the evolution of the North Anatolian Fault in the Ganos-Saros region, northwestern Turkey. Tectonophysics, 293, 105-126. VAN DEN BERG, M. W. & VAN HOOF, T. 2001. The Maas terrace sequence at Maastricht, SE Netherlands: evidence for 200 m of late Neogene and

589

Quaternary surface uplift. In: MADDY, D., MACKLIN, M. G. & WOODWARD, J. C. (eds) River Basin Sediment Systems: Archives of Environmental Change. Balkema, Rotterdam, 45-86. VAN VUGT, N., STEENBRINK, J., LANGEREIS, C. G., HILGEN, F. J. • MEULENKAMP, J. E. 1998. Magnetostratigraphy-based astronomical tuning of the early Pliocene lacustrine sediments of Ptolemais (NW Greece) and bed-to-bed correlation with the marine record. Earth and Planetary Science Letters, 164, 535-551. VANVUGT, N., LANGEREIS,C. G. & HILGEN, F. J. 2001. Orbital forcing in Pliocene-Pleistocene Mediterranean lacustrine deposits; dominant expression of eccentricity versus precession. Palaeogeography, Palaeoclimatology, Palaeoecology, 172, 193-205. VASILIEV, I., KRIJGSMAN, W., LANGEREIS, C. G., PANAIOTU, C. E., MATENCO, L. & BERTOTTI, G. 2004. Towards an astrochronological framework for the eastern Paratethys Mio-Pliocene sedimentary sequences of the Foc~ani Basin (Romania). Earth and Planetary Science Letters, 227, 231-247. VATSEV, M. 1980. Lithostratigraphy of the Neogene sedimentary rocks of the Gotse Delchev Basin. Annals of the Higher Institute of Mining and Geology, Sofia, 25, 103-115 (in Bulgarian with English summary). VATSEV, M. & BONEY, P. 1994. Lithostratigraphy of the Neogene from the Kyustendil coal basin. Reviews of the University of Mining and Geology (St. Ivan Rilski University, Sofia), 40, 43-50. VELCHEV, A. 1995. Pleistocene glaciation of the Bulgarian Mountains. Annuaire de l'Universitk de Sofia, Livre 2, GOographie, 87, 53-65. WESTAWAY, R. 1993. Neogene evolution of the Denizli region of western Turkey. Journal of Structural Geology, 15, 37-53. WESTAWAY, R. 1994. Evidence for dynamic coupling of surface processes with isostatic compensation in the lower crust during active extension of western Turkey. Journal of Geophysical Research, 99, 20203-20223. WESTAWAY, R. 1996. Comment on 'Bivergent extension in orogenic belts: the Menderes massif (southwestern Turkey)' by R. Hetzel, C.W. Passchier, U. Ring & O.O. Dora. Geology, 24, 93-94. WESTAWAY, R. 1998. Dependence of active normal fault dips on lower-crustal flow regimes. Journal of the Geological Society, London, 155, 233-253. WESTAWAY, R. 1999. The mechanical feasibility of low-angle normal faulting. Tectonophysics, 308, 407-443 (Correction: Tectonophysics, 341, 237-238). WESTAWAY, R. 2001. Flow in the lower continental crust as a mechanism for the Quaternary uplift of the Rhenish Massif, north-west Europe. In: MADDY, D., MACKLIN, M. & WOODWARD, J. (eds) River Basin Sediment Systems: Archives of Environmental Change. Balkema, Rotterdam, 87-167. WESTAWAY, R. 2002a. Long-term river terrace sequences: evidence for global increases in surface uplift rates in the Late Pliocene and early Middle Pleistocene caused by flow in the lower continental crust induced by surface processes. Netherlands Journal of Geosciences, 81, 305-328.

590

R. WESTAWAY

WESTAWAY, R. 2002b. Geomorphological consequences of weak lower continental crust, and its significance for studies of uplift, landscape evolution, and the interpretation of river terrace sequences. Netherlands Journal of Geosciences, 81, 283-304. WESTAWAY,R. 2002c. The Quaternary evolution of the Gulf of Corinth, central Greece: coupling between surface processes and flow in the lower continental crust. Tectonophysics, 348, 269-318. WESTAWAY, R. 2003. Kinematics of the Middle East and Eastern Mediterranean updated. Turkish Journal of Earth Sciences, 12, 5-46. WESTAWAY, R. 2004a. Kinematic consistency between the Dead Sea Fault Zone and the Neogene and Quaternary left-lateral faulting in SE Turkey. Tectonophysics, 391, 203-237. WESTAWAY, R. 2004b. Comment on 'Late Cenozoic reorganization of the Arabia-Eurasia collision and the comparison of short-term and long-term deformation rates' by M. Allen, J. Jackson, and R. Walker. Tectonics, 23(5), TC5006, doi: 10.1029/ 2004TC001674. WESTAWAY, R. 2005. Active low-angle normal faulting in the Woodlark Extensional Province, Papua New Guinea: a physical model. Tectonics, 24, TC6003, doi: 10.1029/2004TC001744. WESTAWAY, R. 2006. Cenozoic cooling histories in the Menderes Massif, western Turkey, may be caused by erosion and flat subduction, not low-angle normal faulting. Tectonophysics, 412, 1-25. WESTAWAY, R. & ARGER, J. 1996. The G61ba~l basin, southeastern Turkey: a complex discontinuity in a major strike-slip fault zone. Journal of the Geological Society, London, 153, 729-743. WESTAWAY, R. & ARGER, J. 2001. Kinematics of the Malatya-Ovaclk fault zone. Geodinamica Acta, 14, 103-131. WESTAWAY, R. & KUSZNIR, N. J. 1993. Fault and bed 'rotation' during continental extension: block rotation or vertical shear? Journal of Structural Geology, 15, 753-770 (Correction: Journal of Structural Geology, 15, 1391). WESTAWAY, R., MADDY, D. & BRIDGLAND, D. 2002. Flow in the lower continental crust as a mechanism for the Quaternary uplift of south-east England: constraints from the Thames terrace record. Quaternary Science Reviews, 21,559-603. WESTAWAY, R., PRINGLE, M., YURTMEN, S., DEMIR, T., BRIDGLAND, D., ROWBOTHAM, G. & MADDY, D. 2003. Pliocene and Quaternary surface uplift of western Turkey revealed by long-term river terrace sequences. Current Science, 84, 1090-1101. WESTAWAY, R., PRINGLE, M., YURTMEN, S., DEMiR, T., BRIDGLAND, D., ROWBOTHAM, G. & MADDY, D. 2004. Pliocene and Quaternary regional uplift in western Turkey: the Gediz river terrace staircase and t h e volcanism at Kula. Tectonophysics, 391, 121-169. WESTAWAY,R., GUILLOU,H., YURTMEN, S., DEMiR, T. & ROWBOTHAM, G. 2005. Investigation of the conditions at the start of the present phase of crustal extension in western Turkey, from observations in and around the Denizli region. Geodinamica Acta, 18, 313-342.

WESTAWAY, R., DEM|R, T., SEYREK, A. & BECK, A. 2006. Kinematics of active left-lateral faulting in southeast Turkey from offset Pleistocene river gorges: improved constraint on the rate and history of relative motion between the Turkish and Arabian plates. Journal of the Geological Society, London, 163, 149-164. YALTIRAK, C., SAKINff, M. • OKTAY, F. Y. 2000. Comment on 'Westward propagation of the North Anatolian fault into the northern Aegean: timing and kinematics' by Armijo, R., Meyer, B., Hubert, A. & Barka, A. Geology, 28, 187-188. YILMAZ, Y., GENff, S. C., G~RER, F., et al. 2000. When did the Aegean grabens begin to develop? 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, 353-384. ZAGORCHEV, I. 1992a. Neotectonic development of the Struma (Kraigtid) lineament, southwest Bulgaria and northern Greece. Geological Magazine, 129, 197-222. ZAGORCHEV, I. 1992b. Neotectonics of the central parts of the Balkan Peninsula: basic features and concepts. Geologische Rundschau, 81, 635-654. ZAGORCHEV, I. 1994. Comment on 'Late Cenozoic extension in northeastern Greece; Strymon Valley detachment system and Rhodope metamorphic core complex' by Dinter, D.A. and Royden, L. Geology, 22, 283. ZAGORCHEV, I. 1995. Pirin; Geological Guidebook. Professor Martin Drinov Academic Publishing House, Sofia. ZAGORCHEV, I. 1998a. Rhodope controversies. Episodes, 21, 159-166. ZAGORCHEV, I. 1998b. Pre-Priabonian Palaeogene formations in southwestern Bulgaria and northern Greece: stratigraphy and tectonic implications. Geological Magazine, 135, 101-119. ZAGORCHEV, I. 2001a. Introduction to the geology of SW Bulgaria. Geologica Balcanica, 31, 3-52. ZAGORCHEV, I. 200lb. Low-angle normal faults and detachment hoaxes in SW Bulgaria. Geologica Balcanica, 31, 142-143. ZAGORCHEV, I. & DINKOVA, J. 1990. Geological Map

of the People's Republic of Bulgaria, 1:100 000 series, Petrich sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. ZAGORCHEV, I. & MOORBATH, S. 1986. Rb-Sr dating of the granitoid magmatism in Sahtinska Sredna Gora Mountains. Reviews of the Bulgarian Geological Society, 47(3), 62-68 (in Bulgarian with English abstract). ZAGORCHEV, I., LILOV, P. & MOORBATH, S. 1989a. Results of the rubidium-strontium and potassiumargon radiogeochronological studies of the metamorphic and igneous rocks of Southern Bulgaria. Geologica Balcanica, 19, 41-54. ZAGORCHEV, I., PoPOV, N. & RUSEVA, M. 1989b. Stratigrafiya Paleogena v chasti yugo-zapadnoi Bulgarii. Geologica Balcanica, 19(6), 41-69. ZAGORCHEV, I., GORANOV, A., VULKOV, V. & BOYANOV, I. 1999. Palaeogene sediments in the Padala graben, northwestern Rila Mountain, Bulgaria. Geologica Balcanica, 29, 59-69.

Neotectonic development of the (~ameli Basin, southwestern Anatolia, Turkey MEHMET

C I H A T A L ~ I ( ~ E K 1, J O H A N H . T E N V E E N 2'3 & M E H M E T

OZKUL 1

1Department of Geological Engineering, Pamukkale University, 20070 Denizli, Turkey (e-mail. alcicek@pamukkale, edu. tr) 2Faculty of Earth and Life Sciences, Free University, de Boelelaan 1085, 1081 H V Amsterdam, Netherlands 3Institute for Geology, Mineralogy and Geophysics, Universitdtsstrasse 190, D-44-801, Bochum, Germany This study of the ~ameli Basin presents a detailed basin evolution combined with structural analysis and provides the first detailed time-stratigraphic framework for the neotectonic development of Neogene grabens along the Fethiye-Burdur Fault Zone in southwestern Anatolia. During the Early Tortonian, the ~ameli Basin was established as a broad fault-bounded fluviolacustrine basin that experienced NW-SE extension. By MidPliocene time, continued NW-SE extension resulted in the formation of a new intrabasinal fault zone that split the basin longitudinally into two compartments. The development of a new generation of normal faults further split the basin into four narrow half-graben compartments at the end of the Late Pliocene. Structural analysis of basin-bounding and intrabasinal faults related to this three-stage basin development shows that NW-SE extension apparently persisted from Late Miocene to early Quaternary time. The youngest (i.e. Holocene), deformation is characterized by dextral shear along NE-SW-trending strikeslip faults and continuing NW-SE extension. The Late Miocene foundering of the basin was related to extension in the northerly hinterland zone of the still-emplacing Lycian nappes, whereas outward growth of the Hellenic Arc in response to the westward Anatolian extrusion is the main cause for NW-SE extension from the Pliocene onward. Dextral strike-slip faulting is localized and is associated with the activity of NW-SE-trending faults that accommodated NE-SW extension. The simultaneous activity of these faults suggests the existence of biaxial extensional tectonics, as initially proposed for the Burdur-Dinar area. Sinistral strike-slip faulting, continuing along the eastern Hellenic Arc, penetrated the southernmost part of Turkey but has not yet reached the Cameli Basin area. Our biostratigraphically well-constrained tectonosedimentary model for the evolution of the Cameli Basin provides a reliable time-stratigraphic framework for NE-SW extension in the 'Fethiye-Burdur Fault Zone' of SW Anatolia. We believe that this fault zone represents a broad zone of isolated or interconnected NE-SW-trending basins that formed under prevailing NW-SE extension, rather than being a significant strike-slip fault zone. Abstract:

Regional-scale tectonic extension has influenced the development of numerous fault-bounded intramontane basins in southwestern Anatolia. This extension follows the final stages of the Late Cretaceous-Miocene Tethys ocean closure and formation of the Tauride orogen ($eng6r & Yllmaz 1981; Robertson & Dixon 1984; Seng6r et al. 1985; Zanchi et al. 1993). Three tectonic provinces can be distinguished in SW Anatolia (Fig. 1): (1) the eastern Aegean extensional province; (2) the Isparta Angle; (3) the FethiyeBurdur Fault Zone that geographically connects the former two (Fig. 1). In the eastern Aegean extensional province (EAEP) extension is characterized by basins with a general N E - S W and east-west orientation, which are commonly

referred to as cross-grabens (~eng6r 1987). The cause(s) and timing of the crustal extension are subjects of continuing debate and, until now, remain controversial (Yllmaz et al. 2000; Bozkurt 2001, 2003). For several basins in the EAEP, Purvis & Robertson (2004, 2005a,b) have presented new field-based evidence and A r - A r dating that support a three-phase 'pulsed extension' model. A presumably Late Oligocene phase of extensional unroofing of the Menderes Metamorphic Massif created approximately N E - S W scoop-shaped depressions. The major east-west-trending grabens foundered during an Early-Late Miocene phase of north-south extension related to rollback of the Aegean subduction zone. This interpretation concurs with

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 591-611. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

592

M.C. AL(~I(~EKE T A L .

Fig. 1. Geodynamic framework of the eastern Mediterranean showing main structural features in the Hellenic Arc and southwestern Turkey with its three tectonic provinces: EAEP, eastern Aegean extensional province; FBFZ, Fethiye-Burdur Fault Zone; IA, Isparta Angle. Box shows location of Figure 2.

that of the evolution of other Miocene basins in the Aegean region such as Crete (ten Veen & Postma 1999) and Rhodes (ten Veen & Kleinspehn 2002). A young Pliocene-Quaternary phase of north-south extension in the EAEP is related to westward tectonic escape of Anatolia. Recent studies by Flecker et al. (1995, 2005) and Glover & Robertson (1998) have revealed the complex Miocene-Recent tectonic evolution of the Isparta Angle (Figs 1 and 2). Fault orientations in the Isparta Angle are NE-SW, NW-SE to north-south, and slicken-fibre patterns indicate multiple fault reactivations. Reverse faulting took place under compression during the Late Miocene Aksu phase and right-lateral strike-slip faulting occurred during latest Miocene-earliest Pliocene transtension. In the Late Pliocene-early Pleistocene, approximate east-west extension formed the present Aksu Basin as a north-south half-graben in the core of the Isparta Angle. The onset of this extension is thought to be related to a regional change in stress direction in the Aegean region (Glover & Robertson 1998), plausibly related to the onset of westward tectonic escape of Anatolia.

The Fethiye-Burdur Fault Zone (FBFZ) is characterized by the dominance of Late Miocene-Quaternary NE-SW-trending faults and basins. These occur in a roughly linear arrangement between Fethiye and Afyon and include the Cameli, Burdur, Aclg61, Sandlkll, t~ivril and E~en (~ay basins and their bounding faults (Fig. 2). To the north the FBFZ merges with a series of WNW-ESE grabens, including the Dinar, Bey~ehir, Ak~ehir-Afyon and Dombayova grabens and their bounding faults. The latter are interpreted as the easternmost expression of the east-west basins of the Aegean extensional province (Westaway 1990), or as a westernmost part of a reactivated Aksu thrust fault (Temiz et al. 1997). Many earthquakes originate from both of these WNW- and NE-trending structures, including the 3 October 1914 Burdur (M=7.1), 7 August 1925 Dinar (M=5.8), 19 July 1933 ~ivril (M=5.8), 12 May 1971 Burdur (M=6.2), 1 October 1995 Dinar (M=6.1) and 15 December 2000 Ak~ehir (M=5.8) earthquakes. Some workers (e.g. Dumont et al. 1979; Eyido~an & Barka 1996; Barka et al. 1997) have suggested that the FBFZ

NEOTECTONIC ~AMELI BASIN, SW TURKEY

593

Fig. 2. General geological map of southwestern Turkey, including the FBFZ and the Isparta Angle (based on ~enel 1997a-f), showing major lineaments detectable in satellite imagery (ASTER) and digital terrain models (GTOPO30). Main structural featm'es in the Isparta Angle are after Glover & Robertson (1998).

represents a regionally important sinistral, transtensional fault. However, sinistral strike-slip motions are not evident from earthquake focal mechanisms (Taymaz et al. 1991; Taymaz & Price, 1992), and Ko~yi~it et al. (2000) regarded it as a normal fault zone. The interpretation that the FBFZ is a continuation of the sinistral Pliny fault zone (Barka et al. 1997; A19igek et al. 2002) has been put in doubt by ten Veen et al. (2004), who showed that the Pliny 'trench' in fact continues in offshore southern Turkey (Fig. 1). As shown in Figure 2, NE-SW-trending faults occur not only along a zone from Fethiye to Burdur, but are numerous throughout southwestern Turkey. Although these faults are the most pronounced features, the actual geometries of the basins in this area appear to be related to a combination of N E - S W and north-south faults. This en echelon basin configuration becomes apparent on satellite images, such as for the Esen Gay Basin (see ten Veen 2004; Fig. 2), and in multibeam bathymetry images of the Anaximander Mountains, offshore southern Turkey (ten Veen et al. 2004; Fig. 1). This fault pattern continues westwards into the eastern Hellenic Arc, where deformation occurs as a result of a transtensional setting (e.g. ten Veen

& Kleinspehn 2002, 2003). A limited number of structural analyses in SW Turkey (e.g. Dumont et al. 1979; Temiz et al. 1997, 2001) indicate the presence of normal, oblique and strike-slip faults, and several conflicting regional interpretations have emerged from fault kinematic analyses of these faults. For the E~en Gay Basin (Fig. 2), structural and sedimentological data indicate that the Plio-Pleistocene period was marked by east-west to W N W - E S E extension, but the Holocene-Recent period was characterized by a complex combination of faults of which sinistral strike-slip faults trending 070 ~ are the most important. Fault-slip analysis suggests that deformation occurred in a transtensional setting involving the time-transgressive addition of a sinistral shear component, which was possibly produced by northeastward propagating transcurrent motion of the Hellenic forearc (ten Veen 2004). Thus, it appears that the FBFZ is situated between a zone of north-south neotectonic extension in the EAEP, a zone of transtension along the eastern Hellenic Arc, and a zone of east-west extension in the Isparta Angle. To what extent these geodynamic driving forces play a role in the neotectonic evolution of the study area is still

594

M.C. AL(~I(~EKET AL.

unclear, as is the role of any sinistral motion along the hypothetical FBFZ. The present study documents the tectonosedimentary evolution of the Late Miocene-Late Pliocene, intramontane (~ameli Basin in the central part of the FBFZ (Fig. 1), based on sedimentary facies analysis, biostratigraphic dating and structural analysis. We use the basin fill for temporal and palaeogeographical control in order to document internal basin deformation and adjacent basement kinematics that are related to regional driving mechanisms.

(~amefi Basin The (~ameli Basin (Fig. 3), c. 40 km wide and 60 km long, consists of a series of NNE-SSWtrending interconnected tilt-block compartments within the Lycian nappes ( de Graciansky 1972). Locally, these ophiolite and limestone thrustsheets are unconformably overlain by Lower Miocene deposits that were first interpreted by Altmh (1955) as marine-fossiliferous unit. These deposits comprise alluvial red beds overlain by shallow-marine sandstones, marls and fossiliferous limestones. Similar basal sediments, elsewhere in the Lycian nappes, were interpreted as syn-nappe emplacement units by Collins & Robertson (2003). This supra-allochthonous sedimentary cover is here regarded as part of the basement succession (Figs 3, 4; Alqiqek et al. 2005). Along the SE and NW margins, the Dirmil and Bozda~ faults (Fig. 3), respectively, are the main basin-bounding normal faults that delimit the extent of the ~ameli Formation. Northwestdipping secondary normal faults divide the basin into four approximately equal-sized compartments (Fig. 5). Although the ~ameli Basin is part of a larger area of NNE-SSW-trending basins that constitute the hypothetical FBFZ, the individual basin-bounding faults do not extend beyond the basin's northern and southern limits. Instead, NW-SE-trending faults delimit the basin and there is no evidence of cross-cutting fault relationships. Counterparts of the ENEWSW lineaments, which are clear from satellite imagery and terrain models (Fig. 2) are not observed as basin-scale faults in the (~ameli area. The deposits of the (~ameli Formation exhibit a general southeastward dip towards the NWdipping faults (Fig. 3) and are unconformably overlain by non-tilted Quaternary alluvial deposits, which are generally <20 m in thickness. The ~ameli Formation was originally mapped as the 'Neogene cover' of the Lycian nappes and assumed to be Pliocene in age (Becker-Platen 1970; Erakman et al. 1982; Me~hur & Akpmar

1984), although it was neither dated nor sedimentologically studied. More recently, the age of the ~ameli Formation has been determined as Late Miocene (Tortonian) to Late Pliocene (Gelasian) based on terrestrial mammal macro- and microfossils and molluscan remains (Al~i~ek et al. 2005). The succession has been grouped into three lithostratigraphic subunits referred to as the Derindere, Kumaf~an and Defgne Members, which consist of alluvial-fan, fluvial and lacustrine deposits, respectively. In the central part of the basin, these members overlie each other in a 500 m thick sequence, but they are laterally equivalent along the basin margins (Fig. 3; Table 1; A19iqek 2001; Alqiqek et al. 2004, 2005). The Derindere Member is composed of coarse-grained alluvial deposits and typically occurs in the lowermost and uppermost parts of the basin fill, especially along the basin margins. This member is 40-60 m thick, dark red in colour and composed of matrix-supported conglomerates and mudstones. The unit passes laterally and vertically into the fluvial Kumafsan member and the lacustrine Defgne Members, and is often found in faulted contact with basement rocks. The Kumaf~an Member is widespread in the northern part of the basin and in the middle stratigraphic level of the basin fill. The unit consists of up to 146 m of stacked fluvial deposits characterized by a light yellow colour. The unit passes laterally and vertically into the alluvial Derindere and lacustrine De~ne members. The De~ne Member is represented by a sequence of lacustrine deposits that varies in thickness between 75 and 300 m. The unit is most common in the southern parts of the basin and mainly constitutes the upper part of the basin succession. The member passes laterally and vertically into the fluvial Kumaf~arl Member and the alluvial De~ne Member, but is also found resting directly on the basement rocks.

Basin evolution Stage 1." opening o f the (2ameli Basin (Late Miocene) The mammal remains of Per&sodactyla found in the lowermost part of the basin-fill succession near Elmahyurt (Figs 3 and 5) indicate that sedimentation in the (~ameli Basin commenced by Vallesian time (Early Tortonian 10.8-8.5 Ma; Alqi~ek et al. 2005). At that time, an ephemerallake environment existed along the basin axis, represented by the lacustrine deposits of the De~ne Member. Axial-fluvial systems of the Kumaf~an Member supplied sediment

NEOTECTONIC (TAMELI BASIN, SW TURKEY

Fig. 3. Geological map and cross-section of the (~ameliBasin (modified from A19iCeket

predominantly from the NE and SW and alluvial fans prograded from the basin-margin fault escarpments, represented by the deposits of the Derindere Member (Fig. 6a). These Upper Miocene alluvial-fan deposits show an overall

595

al.

2005).

upward fining trend, but thicken and coarsen towards the now faulted basin margins, represented by the Bozda~ and Dirmil Faults to the N W and SE of the basin, respectively. Locally, the alluvial-fan deposits rest in small patches on

596

M.C. ALCICEK E T AL.

Fig. 4. Simplified statigraphy of the Cameli Basin (not to scale). Jaw symbols refer to levels with abundant mammal fossils.

the footwall block west of the Bozdafg Fault. A wedge shape of the stage 1 alluvial fans is particularly well seen at the western basin margin, but is obscured along the southeastern margin where later, ESE tilting has caused the burial of these deposits. Along the southwestern and northeastern basin margins, the Cameli Formation unconformably overlies the basement and locally Lower Miocene deposits (Fig. 3). All stage 1 deposits are characterized by dense, outcrop-scale growth faults (Fig. 7a), expressing synsedimentary extension throughout the Late Miocene. Interpretation. The thickening of alluvial fan deposits towards the Bozda~ and Dirmil Faults suggests that these faults formed the basin margins by Late Miocene times. Because the depth to the base of the basin-fill succession is unknown directly adjacent to the basin-margin faults, we

can only speculate about the amount of basin subsidence on either side of the basin. The general eastward tilt of the basin floor and asymmetrical distribution of sedimentary facies that is evident for subsequent stages of basin evolution cannot be affirmed for this stage. The presence of alluvial deposits on the footwall of the Bozdaf~ Fault indicates the potential of the alluvial fans to backlap across the basin margin, possibly controlled by continued basin deepening. The overall upward fining of the alluvial succession supports this basin deepening during stage 1. Palaeontological dates at the base of stage 2 (3.8-3.2 Ma; see next section) provide an estimate of c. 5-8 Ma for the duration of stage 1. The maximum basin depth is c. 210 m, based on the thickness of the stage 1 succession, not corrected for compaction, or the limited palaeo-water depths of the ephemerallake environment. These values suggest slow

NEOTECTONIC ~AMELI BASIN, SW TURKEY

597

Fig. 5. Lateral correlation of measured logs in the ffameli Basin. (For location of logs see Figs 3 and 6.)

basin subsidence of the order of 3-4 cm ka -1. However, erosion of sediment associated with block tilting during the later stages of basin development cannot be ruled out and could influence the basin subsidence rate.

Stage 2: intrabasinal faulting (Early-Mid-Pliocene transition) Whereas the Late Miocene-Early Pliocene development of the ~ameli Basin was characterized by slow overall subsidence, accommodated at the basin-margin faults, the following stage shows a distinctly different style as a result of the development of a NW-dipping normal fault in the centre of the graben. This Sankavak-Kumaf~an Fault Zone (SKFZ, Fig. 5) split the basin into two longitudinal compartments (Figs 3 and 6b). Close to the SKFZ, near Sankavak, the lowermost tufa unit has a thickness of 55 m, whereas

at Ericek 5 km to west, a thickness of c. 10 m is observed (A19igek & Ozkul 2005), indicating a wedge shape of the tufa unit. Laterally (to the NW), the tufa interfingers with ephemeral-lake sediment and is overlain by fluvial and peat-mire sediments (Fig. 5). The latter sediments were sampled near Ericek and appear to abound in mammal micro- and macro-fossils. These include Rodentia teeth and bones, which indicate a Late Ruscinian (Zanclean-Piacenzian) age (3.83.2 Ma, Al~igek et al. 2005). Also to the NE along the SKFZ, the tufa is transitional to the fluvial deposits of the Kumafsan Member. Peat-mire deposits at the lateral transition between the fluvial Kumaf~arl Member and the lacustrine De~ne Member near (~amhbel (Fig. 3) contain micro-remains of the mammal Rodentia, which indicate a Late Ruscinian-Early Villanian age (Piacenzian-Gelasian, 3.5-2.5 Ma, A19igek et al. 2005).

598

M.C. ALr

ET AL.

Table 1. Lithofacies characteristics of the Cameli Formation in the ~ameli Basin (based on Alfifek 2001; Alfifek et al. 2004, 2005) Member

Facies associations

Lithology

De~ne (DMb)

Open-lake (OL)

Laminated marl Clayey limestone Limestone Clayey limestone Massive-pebbly sandstone Stratified sandstone Planar cross-bedded sandstone Ripple cross-laminated sandstone Massive-pebbly sandstone Stratified sandstone Planar cross-stratified sandstone Clast-supported conglomerate Stratified conglomerate Laminated siltstone-mudstone Travertine, micritic limestone Laminated siltstone-mudstone Laminated marl Clayey limestone Coal Laminated siltstone-mudstone Planar cross-bedded sandstone Epsilon cross-bedded sandstone Stratified sandstone Laminated siltstone-mudstone Massive mudstone Ripple laminated sandstone Stratified conglomerate Massive-pebbly sandstone Planar cross-bedded sandstone Clast-supported conglomerate Matrix-supported conglomerate Clast-supported conglomerate Stratified conglomerate Massive-pebbly sandstone Massive mudstone

Ephemeral-lake (EL) Deltaic (D)

Fan-deltaic (FD)

Kumaf~an (KMb)

Tufa (T) (spring-outflow)

Peat-mire (PM) Meandering-river (MR)

Braided-river (BR) Derindere (DrMb)

Alluvial-fan (AF)

Basinwide, the fluvial and peat-mire deposits are overlain by up to 220 m of monotonous open-lake deposits with local indications of wave reworking. In the Kavalcllar area (Fig. 3), the open-lake deposits directly overlie the basement ridge that was elevated in the footwall of the SKFZ (Fig. 7c). The relatively thick open-lake succession passes upwards into ephemeral-lake deposits in the basin centre. Towards the SW and NE, this facies passes into Gilbert-type river deltas, whereas near the basin-margin faults the lake deposits pass into Gilbert-type fan deltas (Fig. 6c). These deltas are characterized by steep foresets with abundant slump structures. Based on the irregular occurrence of these delta deposits, it is inferred that individual delta bodies have limited aerial extent. Interpretation. The wedge shape of the tufa unit close to the SKFZ suggests a gradual, progressive

Depositional environment Lacustrine

Fluvial

Alluvial

tilt of the basin floor accommodated by fault displacement along the SKFZ. This caused a southeastward tilting of the basin floor and older (stage 1) basin deposits. The wedge shape implies a basin floor tilt of less than 1~ during the tufa deposition, which explains why the contact with the overlying deposits shows no recognizable angular unconformity. The tufa is interpreted to originate from spring waters along the newly created intrabasinal SKFZ from which calcium carbonate was precipitated in the confined area of the local fault segment (see Guo & Riding 1998). Tufa or travertine (hotspring) deposits are generally recognized to be related to faults (Heimann & Sass 1989), especially in extensional tectonic settings (Altunel & Hancock 1993; Ozkul et al. 2002). The general eastward dip of the basin floor and deepening towards the SKFZ is further interpreted from lateral facies transitions as

NEOTECTONIC ffAMELI BASIN, SW T U R K E Y

599

Fig. 6. Palaeogeographical reconstructions of successive stages of basin evolution in both map and cross-sectional view. (a) Opening of the basin during Late Miocene stage 1. (b) Development of intrabasinal Sankavak-Kumafs, an Fault Zone (SKFZ) and fault-related tufa wedge during the Early-Mid-Pliocene stage 2. (e) After overall deepening of the basin and basinwide occurrence of open-lake facies during the continuation of stage 2. The closure of stage 2 is marked by the re-occurrence of coarse clastic sedimentation. (d) Renewed intrabasinal faulting through the development of the Alcl-Kelek~i (AKFZ) and Uzunoluk-~ameli (UffFZ) faults zones during the Late Pliocene.

600

M.C. AL~I(~EK E T AL.

Fig. 7. Structural features related to the development of the (~ameli Basin. (a) Growth fault associated with the stage 1 basin opening. (b) Fault plane of Uzunoluk-Cameli Fault with groove marks indicative of dip-slip normal faulting. (c) Down-section, to the east of (d), steeply tilted open-lake deposits (OL) overlie metamorphic basement rocks that were uplifted along the Sankavak-Kumaf~an Fault. The open-lake deposits correspond to the lowermost part of the Kavalcllar log in Fig. 5. (d) Typical eastward-tilted basin fill in the hanging-wall block of the Uzunoluk-~ameli fault (arrows+) SE of B19ak~l (Fig. 3), corresponding to the ephemeral-lake (EL) fan-delta (FD)-alluvial fan (AF) transition in the upper part of the Kavalcllar log in Figure 5. Bedding planes show faint indications of up-section decrease in dip angle as a result of syndepositional basin-floor tilting. Quat, Quaternary alluvial fan.

depicted in Figure 6b. Subtle increases in base level, accommodated by displacement along the SKFZ, may have caused inundation of the alluvial floodplains by the expanding lake, which locally led to peat-mire development. The deepening into deep-lake conditions and the complete submergence of the SKFZ intrabasinal ridge substantiates the continuation of basin-floor subsidence. Although the deep-lake deposits locally overlie the Dirmil and Bozda~ faults, there is no evidence that the lake extended far beyond these faults. Therefore, subsidence along the primary basin-margin faults is held responsible for the basin deepening. The renewed input of coarse clastic material into the basin indicates either that basin subsidence slowed near the end of stage 2, or that

sediment supply drastically increased. From the steep foresets with abundant slump structures and the limited aerial extent of the fluvial topset beds we deduce that only limited progradation of the fan deltas toward the deep lake occurred initially (see Postma 1990). Therefore, the shallowing in the basin centre towards ephemerallake conditions is not likely to be much influenced by an increased sediment input, but rather by slowing of basin subsidence. However, slowing of basin floor subsidence in combination with continued input of coarse-clastic sediment could have forced fan deltas to prograde towards the basin centre, as deduced from the upward transition of open-lake to ephemeral-lake and to deltaic deposits (Fig. 6).

NEOTECTONIC ~AMELI BASIN, SW TURKEY

Stage 3." renewed intrabasinal faulting (latest Pliocene) After shallowing of the lake (Fig. 6c), a new phase of basin differentiation is represented by the generation of the Alcl-Kelek~i (AKFZ) and Uzunoluk-(;;ameli (UCFZ) fault zones (Fig. 3). These faults caused the basin to be divided into narrower half-graben compartments (Fig. 6d) and are best recognizable where they cut the stage 2 open-lake sedimentary succession. The strikes and northwesterly dips of these new faults are similar to those of the pre-existing SKFZ. South of (j;ameli along the UCFZ (Fig. 6d) the upper tufa unit exhibits a wedge shape with a thickness of 6 m near Kavalcalar (Fig. 5; Kavalcllar log) that rapidly increases towards the fault and is overlain by fluvial and alluvial-fan deposits. Along the AKFZ, south of Kelekgi, a thick fan-delta sequence progrades westwards. Basinwards, the delta foresets are transitional to ephemeral-lake deposits, which are considered as delta front or prodelta equivalents of the fan delta. The fan-delta sequence is cut by a series of west-dipping normal faults that represent fault-parallel splays of the AKFZ (Fig. 3). One of these faults delimits an eastward-thickening 'upper' tufa wedge that includes intercalated slump deposits. This third stage of basin development occurred in Late Villanian time (Gelasian, 2.6-1.8 Ma, Al~i~ek et al. 2005), as indicated by the remains of Rodentia fossils in the lower or distal part of alluvial-fan deposits that prograded onto the tufa deposits (Fig. 5). Towards the end of the Villanian (Late Gelasian) stage, the Cameli Basin was predominantly filled by fluvial and alluvial-fan depositional systems. Deep-lake conditions have not returned since that time. After the late Gelasian, basin-floor tilting continued, as inferred from the angular unconformity between the tilted stage 3 fluvial succession and the subhorizontal Pleistocene alluvium. Many northsouth-trending outcrop-scale normal faults that cut and displace the stage 3 sediments are not present in the Pleistocene cover (Fig. 8a).

Interpretation. The stage 3 faulting episode caused further syndepositional southeastward tilting of the segmented basin-fill succession and its onset was accompanied by the deposition of the upper tufa unit. Our observations strongly suggest that eastward basin-floor tilting ceased at some ill-defined time during the PlioceneQuaternary transition. In addition, the fact that deep-lake conditions did not return after the Villanian suggests an important lowering of the local base level at the Pliocene-Quaternary transition. In this scenario, the contemporaneous

601

local base-level lowering can be explained by the cessation of basin-floor subsidence and by far more important Plio-Quaternary isostacy-driven regional surface uplift (see Westaway et al. 2003).

Stage 4." Quaternary During the subsequent phase of Quaternary sedimentation several rivers incised the older basin fill and small alluvial fans developed at the dormant fault escarpments, which largely smoothed out the pre-existing tilted-block relief of the basin. Older Quaternary (?Pleistocene) river terrace deposits near Evciler unconformably overlie the tilted Neogene basin fill (Fig. 8a) and are, in turn, incised by more recent Quaternary fluvial systems that parallel the older basin axis. Along the southeastern basin margin, near Dirmil, an alluvial fan sequence is found that truncates the older lacustrine marls (Fig. 3). Close to the margin, this sequence consists of chaotic conglomerates with angular clasts representing scree and debris-flow deposits. To the west these chaotic deposits gradually pass into red and green clays rich in ostracodes and plant remains. Along the western basin margin, large scree sheets cover the Bozda~ Fault (Fig. 3). All of the Quaternary fluvial and alluvial deposits display subhorizontal bedding and are relatively poor in intraformational faults compared with the older deposits of the ~ameli Basin. The 020~ Cibyra Fault Zone (CbFZ) is, however, one of the few faults that indicates post-Pliocene fault activity in the ~ameli Basin. This fault zone is characterized by numerous anastomosing subvertical fault splays. These show vertical offsets ranging from 0 to 50 cm, although some offsets of several metres were observed. The fault cuts a pre-existing basement contact of onlapping Pliocene (stage 3) fluvial deposits west of GSlhisar (fault locality b in Fig. 3). Consequently, the fault either juxtaposes peridotite basement and Neogene conglomerates, or cuts only the latter unit. The ancient city of Cibyra was built on the Pliocene clastic sedimentary rocks and its damaged stadium is located at the trace of one of the fault segments of the CbFZ (Fig. 3). According to Akyiiz & Altunel (2001), the seating rows of the stadium were displaced sinistrally where they are crossed by the fault. There is ample archaeoseismolgical evidence that the stadium suffered serious earthquake damage (Fig. 8b), including a domino-style arrangement of fallen column blocks, the rotation and dilation of seating row blocks, and the almost complete collapse of the eastern side of the stadium. Akyiiz & Altunel (2001) observed slickenside lineations on exposed fault surfaces to the south and north

602

M.C. AL(~It~EK ET AL.

Fig. 8. Photographs showing characteristics of Quaternary strike-slip faulting: (a) older Quaternary (?Pleistocene) river terrace deposits unconformably overlie the tilted and faulted Neogene basin fill; (b) disturbed seating rows of the stadium at Cibyra; view to the south; (c) calcrete-filled, steep faults (see arrows) cutting tilted Pliocene conglomerates in the Kumklsl~l quarry (see Fig. 3 for location); (d) detail of (c), showing near-horizontal slickensides indicative of dextral strike-slip motion on the fault plane. of the stadium, which indicate sinistral oblique displacement of the CbFZ. We reinvestigated these localities and based on our fault kinematic data it appears that the fault zone comprises many differently oriented splays that exhibit either dextral, sinistral or normal displacement. The northernmost extent of the CbFZ is found in the Kumklsl/gl quarry (Fig. 3), where it cuts southeasterly tilted Pliocene conglomerates (Fig. 8c and d). There are also numerous faults with different orientations and with different styles of displacement, suggesting a complex deformation history. The analysis of these data is discussed in the next section. Interpretation. Despite the base-level lowering at

the Pliocene-Quaternary transition, during the Quaternary the basin was still fed by alluvial systems at both margins. The fact that these systems also cover the basin-margin faults suggests that fault displacement had ceased by Quaternary times. Active faults, such as the CbFZ, show negligible vertical offsets and we suppose that the thickness distribution of the Quaternary fan systems was not controlled by these faults.

Rather, the pre-existing basin was filled by these young alluvial systems. The archaeoseismological evidence at Cibyra and the lithostratigraphic units affected limit the age of fault activity to post-Pliocene to subrecent, (i.e. the Quaternary). This young age is in accordance with the relatively small fault offsets.

Structural analysis Fault measurements were made at several sites throughout the (~ameli Basin (Fig. 3). Generally, the master faults cut through basement limestone and display slickensides, striae, or groove marks (tails and scratches produced by asperity ploughing). Smaller-scale faults that cut the clastic or lacustrine basin fill exhibit only striated surfaces. In a single case, a fractured and displaced clast could be used to determine the sense of slip. The estimated age of fault activity in most cases is based on fault-controlled sedimentological events (see previous section). Alternatively, the maxim u m age of fault activity is indicated by the age of the youngest affected stratigraphic unit. A

NEOTECTONIC ~AMELI BASIN, SW TURKEY combination of both provides a time frame for fault activity during the last part of this period. In this study, fault data provide kinematic support for our model of basin evolution. The number of useful fault data are limited by available age constraints. Only 77 fault-kinematic indicators have therefore been considered for stress inversion, to obtain the orientations of the principal stress axes (see Table 2). Fault-slip analysis was performed with FaultkinWin 1.1 (a computer program for analysis of fault slip data by R. A. Almendinger et al. 2004). Faults that are suggested to be active during stages 1-3 of the basin evolution have northsouth to N E - S W orientations and exhibit kinematic evidence indicative of normal to normaloblique-slip faults (Fig. 8a-c). A few N W - S E trending faults represent dextral oblique-slip faults that developed at the same time as the normal faults. Fault-plane solutions, often used as a first approximation of the stresses (Marrett & Almendinger, 1990), indicate that this fault pattern was produced under N W - S E extension (Fig. 9a-c). Fault data for stage 4 of the basin evolution collected in the complex Cibyra Fault Zone and the Evciler area show many kinematic inconsistencies. This is demonstrated by the P/T scatter plot (Fig. 9g), which shows overlaps between the orientations of P and T axes of individual faults. Such overlap between P and T axes is often produced by a combination of fault kinematic indicators that were generated under different stress regimes. Representing the fault data in one single fault-plane solution gives an incorrect result that does not account for the stresses exerted by individual tectonic regimes. To overcome this problem we sorted the fault data according to the orientations of the P and T axes to produce kinematically compatible groups of fault data. Performing this exercise on the Cibyra and Evciler fault sets produces two groups of compatible fault data, which are referred to as stage 4a and stage 4b (Fig. 9). The first set is represented by normal and normal-oblique faults with north-south to N E - S W orientations in combination with a few north-south-trending sinistral strike-slip faults (Fig. 9d). The faultplane solutions of these fault sets suggest that the faults were produced under N W - S E extension, and are thus kinematically comparable with the stage 1-3 faults. The second subset of faults dominantly consists of NE-SW-trending dextral strike-slip faults and NW-SE-trending normal faults (Fig. 9e). At both the Cibyra and Evciler sites, a few NNW-SSE-trending sinistral strikeslip faults occur as well. Fault-plane solutions for both sites (Fig. 9e), show the different orientation

603

of the T axes and the near-horizontal plunge of the P axes compared with the stage 1-4a faults, suggesting that this fault set was produced in a strike-slip tectonic regime. In conclusion, two sets of fault-kinematic data are recognized that relate to different types of deformation and that were probably generated under different tectonic regimes. The translation from observed deformation to stresses exerted by regional tectonic processes is facilitated by stress inversion techniques. Following the method described by Marrett & Almendinger (1990), the orientations of stress axes are based on Bingham distribution statistics providing directional maxima of the shortening and extension axes of a fault array. The stress tensor aspect ratio defined a s (YZ--O'3/(Yl--O'3,was determined to indicate the tectonic regime. A best-fit stress inversion for the stage 1-4a faults yields a horizontal o3 oriented towards 311 ~ and a vertical maximum compression ol. In combination with the stress axes, the stressellipsoid shape parameter ~ = 0 . 6 5 indicates plane stress under N W - S E extension (D1 in Fig. 9f). Stress inversion for the stage 4b faults shows a slightly inclined % plunging toward 195 ~ and a % axis moderately inclined toward 088 ~ With ~=0.32, the orientations of the inclined stress axes indicate deformation in transtension (D2 in Fig. 9g). This transtensional regime is resolved by dextral shear along NE-SW-trending synthetic strike-slip faults, sinistral shear along NNW-SSE-trending antithetic strike-slip faults and tension at NW-SE-trending normal faults. The onset of D1 extension is well constrained by the Late Miocene initiation of the ~ameli Basin. Continued extension caused segmentation of the tilted basin floor, resulting in local uplift or deepening that characterizes the Late MioceneLate Pliocene sedimentation patterns of the basin. Although no tectonosedimentary evidence was found, the D~ extension obviously continued into the early Quaternary (stage 4a), as is evidenced by fault data from the Cibyra and Evciler localities. The transtensional D2 deformation does not clearly overprint the former D~ faults, but the late Quaternary (Holocene) timing is well constrained by the historical ruptures that destroyed Cibyra's stadium. However, we did not find that the recent colluvium was affected by these D2 faults.

Discussion The present study of the (~ameli Basin, involving basin analysis, and structural analysis, provides the first detailed time-stratigraphic framework

M.C. AL~I~EK E T AL.

604

Table 2. Fault-slip data collected in the study area providing the basis for kinematic interpretations (stress inversion) presented in Figure 9 Site

Fault

Fault strike (RHR)

Fault dip

Rake (0-180 ~

a

CbFZ

b

CbFZ

c

N of Evciler

d

Yaprakh darn

e

U~FZ

f

SKFZ

g

SKFZ

024 004 240 230 224 226 227 225 170 046 272 246 249 224 200 116 180 235 122 247 260 232 261 123 231 060 180 160 175 150 208 330 355 002 353 012 018 002 192 189 193 200 194 178 170 237 229 223 253 226 045 038 065 270 255 234 229 054 230

83 90 74 75 86 63 68 71 70 73 58 58 81 78 58 70 65 61 41 83 80 75 81 78 73 68 80 65 64 48 87 85 89 89 73 84 70 87 60 54 52 48 50 41 42 54 40 40 51 52 42 68 65 65 53 64 43 40 29

158 180 174 30 160 120 108 70 90 95 0 75 50 70 75 110 80 155 90 150 165 150 53 88 160 110 135 150 123 165 80 100 105 90 80 120 68 100 80 88 127 120 90 108 115 170 180 125 180 90 90 100 70 90 90 72 95 63 90

Indicator type

Upper block Lower block

str. str. str. str. str. str. str. o.c. o.c. s.s. str. s.s. str. g.m. g.m. s.s. g.m. str. str. str. str. str. str. str. str. str. str. str. str. str. str. str. str. str. g.m. g.m. g.m. g.m. g.m. g.m. str. str. g.m. s.s. str. str. str. s.s. str. str. str. g.m. g.m. g.m. g.m. g.m. g.m. str. str.

KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb DMb DMb DMb DMb DMb Basm Basm Basm Basm Basm Basm Basm Basm Basm Basm Basm Basm DMb DMb Basm Basm Basm Basrn DMb DMb

KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb KMb DMb DMb DMb DMb DMb DrMb DrMb DrMb DrMb DrMb DrMb Basm Basm DrMb DrMb DMb Basm KMb KMb Basm Basm Basm Basra DMb DMb

NEOTECTONIC ~AMELI BASIN, SW TURKEY

605

Table 2. Continued Site

Fault

Fault strike (RHR)

h

CbFZ

228 350 035 000 181 178 204 200 060 235 222 021 010 135 220 090 020 020

SKFZ AKFZ AKFZ BFZ

Fault dip Rake (0-180 ~ Indicator type Upper block Lower block 88 80 87 90 79 71 64 52 71 73 42 60 70 83 34 65 43 77

90 55 105 30 100 95 110 70 45 85 95 100 110 120 108 105 90 100

str. str. str. str. s.s. s.s. g.m. s.s. g.m. g.m. str. g.m. s.s. s.s. s.s. g.m. s.s.

KMb KMb KMb KMb Basm Basm Basm KMb KMb DMb DMb LMio LMio LMio LMio LMio LMio LMio

KMb KMb KMb KMb Basm Basm Basm KMb KMb DMb DMb DrMb LMio LMio LMio LMio LMio LMio

Localities (a-l) are indicated in Figure 3. Fault strikes according to right-hand rule (RHR). Rakes of kinematic indicators range form 0~ (pure sinistral slip), through 90~ (pure dip slip) to 180~ (pure dextral slip). Type of kinematic indicator: str., striae; g.m., groove mark; o.c., offset clast(s); s.s., slickensides. Lithologies: Basm, basement; DMb, De~ne Member; KMb, Kumaf~an Member; DrMb, Derindere Member; LMio, Lower Miocene deposits (basement); CbFZ, Cibyra Fault Zone; U(~FZ, Uzunoluk-~ameli Fault Zone; SKFZ, Sankavak-~ameli Fault Zone; AKFZ, Alci-Kelekqi Fault Zone; BFZ, Bozda~ Fault Zone.

for the neotectonic development of Neogene grabens in the Fethiye Burdur Fault Zone. However, it is not exactly clear what are the tectonic driving forces that trigger the formation of these basins. Collins & Robertson (1998, 1999) documented the southeastward emplacement of the Lycian allochthon during three episodes starting in the Late Cretaceous. The final emplacement over the most proximal foredeep (Kas Basin; Fig. 1) occurred in Late Miocene (Tortonian) time. Biostratigraphic evidence presented here indicates that sedimentation in the (j;ameli Basin had commenced by Vallesian time (Early Tortonian; 10.8-8.5 Ma), suggesting that the extensional basin opening coincided with this final phase of thrusting. Extension thus occurred in the hinterland zone of the emplacing nappes and coeval contraction occurred in the Lycian foreland zone (Ka~ Basin). Closer to the nappe front, i.e. to the SE, this extension probably did not develop during the Late Miocene. This is exemplified by the fact that the Esen Gay Basin (Fig. 1) originated as a NE-SW-trending fluviolacustrine basin in a ramp-fold close to the Lycian thrust front (ten Veen 2004). Here, N W - S E extension is reported only for the Late Pliocene-Pleistocene period. Such a combination of extension in the hinterland coeval with contraction in the foreland is interpreted to result from 'orogenic collapse'

(Seyito~lu et al. 1996; Collins & Robertson 1999, and references therein). The structural style of the N E - S W grabens in southwestern Anatolia was thus possibly established during the final stage of nappe emplacement, which also explains the parallelism of the Lycian thrust front and extensional basins. Price & Scott (1991) also mentioned the presence of a NE-SW-trending fabric within the pre-Neogene basement of the Burdur region, which might have influenced the orientation of faults formed during the subsequent Neogene and Quaternary extension. The Pliocene-Quaternary westward escape of Anatolia, which generated north-south extension in the EAEP (Purvis & Robertson 2004), thus postdates the foundering of NE-SW-trending basins in the FBFZ. However, westward escape of Anatolia can be still considered as one of the possible driving mechanism for extension in SW Anatolia during the younger, PlioceneQuaternary period of extension. Based on fault kinematic data for basinbounding faults of the Quaternary Burdur, Aclg61 and ~ivril basins, Price & Scott (1994) identified NW-SE-directed extension. Kinematic data of numerous mesoscale faults that cut the Late Miocene-Late Pliocene Cameli basin fill reveal similar orientations of palaeostress axes. The earthquake source mechanisms of the

606

M . C . A L C I C E K E T AL.

Fig. 9. (a-e) Fault data for the Cameli faults on lower hemisphere equal area projections together with fault-plane solutions used as a first approximation of the stress field (see text for explanation). (f) Stress inversion results of Late Miocene-early Quaternary (D~) deformation, characterized by NW-SE extension with (Yl > ~2 > (Y3and ~3 horizontal. (g) Stress inversion results of late Quaternary (D2) deformation, characterized by inclinations of all stress axes, which in this case is indicative of a dextral-transtensional tectonics regime, resolved by dextral shear along NE-SW-trending synthetic strike-slip faults, sinistral shear along NNW-SSE-trending antithetic strike-slip faults and tension at NW-SE-trending normal faults.

NEOTECTONIC ~AMELI BASIN, SW TURKEY May 12 1971 Burdur earthquake sequence also confirm the NW-SE orientation of extension (Taymaz & Price 1992). Price & Scott (1994) suggested that the discrepancy between northsouth extension in Western Anatolia and the NW-SE extension could be explained in terms of differences in crustal thickness between the rapidly extending Aegean region and the stable unstretched central Anatolia plateau to the east. Block rotations about vertical axes are accommodated by a hypothetical north-south dextral shear zone, which accounts for NE-SW faults with sinistral-normal senses of slip. This concept is somewhat outdated, because of the underestimated role of the major NW-SE-trending Dinar Fault. The 1 October 1995 Dinar earthquake, with a 10 km long surface rupture indicating normal faulting under NE-SW extension, changed existing ideas concerning the regional tectonic setting. Temiz et al. (1997) reintroduced the preliminary cross-fault model of Westaway (1990) for the Burdur-Dinar area and proposed a twostage evolution. The first stage of NE-SW extension caused initiation of the south-facing Dinar breakaway fault and led to the formation of the Burdur, Aclg61 and Givril 'hanging-wall' area. Temiz et al. claimed that at this stage sinistral oblique slip along the Burdur and Aclg61 faults accommodated differential stretching of the hanging wall. During the second stage, extension occurred in both NE-SW (Dinar Fault) and NW-SE (Burdur, Aclg61 and Givril faults) directions. The east-west to WNW-ESE extension of the Esen Gay Basin, forming the southernmost basin in the FBFZ of southwestern Turkey, is suggested to be related to the kinematic effects of outward growth of the Hellenic forearc (ten Veen 2004). The present-day GPS velocity field adequately demonstrates this east-west extension (e.g. McClusky et al. 2000) with effects that are also observed on Rhodes and Crete (ten Veen & Kleinspehn 2002, 2003) and possibly in the Isparta Angle, although interpreted differently by Glover & Robertson (1998). The GPS velocity field is thought to be determined by the westward extrusion of Anatolia in combination with effects of subduction rollback and backarc extension. Ten Veen (2004) suggested that a time-transgressive addition of a sinistral shear component was produced by the northeastward propagating transcurrent motions of forearc slivers that sheared from the expanding forearc as continental collision with the African promontory started. This process began on Crete and Rhodes in the Late Pliocene and is as young as the Holocene in the E~en Gay Basin. Based on these earlier finding and models, a geologically young (even historical) transtensional

607

deformation in the Gameli Basin is thus expected. However, throughout the entire eastern Hellenic forearc, transtension was always combined with NE-SW extension and sinistral shear resolved on NE- to ENE-trending faults. The 'young' transtension in the Gameli basin is, instead, explained by a combination of NE-SW extension and dextral shear resolved on NE-trending faults. The Gameli Basin area can, thus, be regarded as an interference zone of at least three different tectonic regimes: (1) outward growth of the eastern Hellenic forearc, resulting in NE-SW extension; (2) tectonic escape represented by strike-slip faulting approximately along the strike of the eastern forearc; (3) NE-SW extension towards the north of the FBFZ (Dinar and Denizli grabens), which can be interpreted as internal deformation related to Anatolian extrusion (Eyido~an & Barka 1996). The development of a particular basin probably involves the effects of more than one regime, although one stress field might dominate in some cases or could prevail during a certain time period. The model presented in Figure 10 illustrates how the southern part of the FBFZ is influenced by the sinistral transcurrent motions along, and expansion, of the eastern Hellenic Arc (1 and 2), while at the same time the northern part is influenced by the northsouth extrusion-related extension and Hellenic Arc expansion (3 and 1). The NW-SE-trending Gavdlr Fault (Fig. 3), forming the northern limit of the Gameli Basin, might be another important fault breakaway fault on a smaller scale, comparable with the Dinar fault. Differential motion of its hanging-wall block could have induced the dextral motion at the Cibyra Fault Zone, which is supported by the fact the latter fault does not continue into the footwall block of the Gavd~r Fault. Additionally, the increasing importance of NE-SW extension northward along the FBFZ is likely to limit the effect of sinistral transcurrent motions occurring in the southern part. Although local strike-slip faults are present in the FBFZ, from south to north they are related to different tectonic processes. Thus, it appears that this feature is better described as a broad zone of isolated or interconnected NE-SW-trending basins that formed under prevailing NW-SE extension in combination with local tectonic regimes that produce(d) the strike-slip faulting.

Conclusions (1) In the Late Miocene the Gameli Basin was established as a broad fault-bounded fluviolacustrine basin that experienced N W SE extension in the northerly hinterland zone of the still emplacing Lycian nappes.

608

M.C. AL~I(~EK ET AL.

Fig. 10. Sketch diagram showing the inferred Holocene-Recent geodynamic mechanisms that contribute to the deformation of the (~ameli Basin (indicated with bold C). Mechanisms include: (1) outward growth of the eastern Hellenic forearc, resulting in NE-SW extension; (2) tectonic escape represented by strike-slip faulting approximately along the strike of the eastern forearc; (3) NE-SW extension in central western Anatolia related to the Anatolian extrusion. Whereas sinistral strike-slip faulting penetrated only the southernmost part of Turkey, the dextral strike-slip faults in the q2ameli Basin are interpreted to accommodate NW-SE extension on general NE-SW-trending faults.

(2) Outward growth of the Hellenic Arc, as a result of westward Anatolian escape, accompanied a Pliocene period of intensified N W SE extension that resulted in the formation of several new intrabasinal fault zones. By the end of the Pliocene, these faults split the basin longitudinally into four narrow half-graben compartments. (3) Holocene deformation is characterized by continued N W - S E extension and dextral shear resolved on NE-SW-trending strikeslip faults. Dextral strike-slip faulting is localized and accommodates the activity of NW-SE-trending normal faults. (4) Sinistral strike-slip faulting, continuing along the eastern Hellenic Arc and penetrating the southernmost part of Turkey, has not yet reached the (~ameli Basin area. (5) The entire F B F Z is an expression of several regional tectonic processes including: (a) outward growth of the eastern Hellenic forearc, resulting in N E - S W extension; (b) tectonic escape represented by strike-slip

faulting approximately along the strike of the eastern Hellenic forearc; (c) N E - S W extension towards the north of the FBFZ, related to westward Anatolian escape. This study was supported by the Scientific and Technical Research Council of Turkey (Tf0BiTAK) research grant YDAB(~AG 100Y004 to M.C. Al~i~ek and Dutch Science Foundation, NWO grant 831.48.009 to J.H.T.Veen. Some figures were created using the GMT (Generic Mapping Tools) software, and we are indebted to its authors P. Wessel and W. Smith for making it available. This manuscript benefited from constructive reviews by A. H. F. Robertson and S. J. Boulton.

References AKYI)Z, H. S. & ALTUNEL,E. 2001. Geological and archaeological evidence for post-Roman earthquake surface faulting at Cibyra, SW Turkey. Geodinamica Acta, 14, 95-101. ALClCEK, M. C. 2001. Sedimentological investigation of Cameli Basin (Late Miocene-Late Pliocene,

NEOTECTONIC ~AMELI BASIN, SW TURKEY Denizli, S W Anatolia). PhD thesis, Ankara University. ALflCEK, M. C. & OZKUL, M. 2005. Extensional faulting induced tufa precipitation in the Neogene ~ameli Basin of southwestern Anatolia, Turkey. In: OZKUL, M., YAQIZ, S. & JONES, B. (eds) Proceedings of the First International Symposium on Travertines and Technologies Exhibition, Proceedings, 21-25 September, Denizli, Turkey, 120-127. AL(~I~EK,M. C., KAZANCI,N., ~EMEN, i. & [~)ZKUL,M. 2002. Strike-slip faulting in the ~ameli basin, southwestern Turkey: implications for inland transform prolongation of the Hellenic subduction zone. Annual Meeting of the Geological Society of America, 27-30 October 2002, Denver. Abstracts with Programs, 34, 111-114, 250. AL~I~EK, M. C., KAZANCI, N., OZKUL,M. & SEN, $. 2004. Sedimentary infill and geological evolution of the ~ameli Basin, Denizli, SW Turkey. Bulletin of Mineral Research and Exploration of Turkey, 128, 99-123. AL~I(~EK, M. C., KAZANCI, N. & {~ZKUL, M. 2005. Multiple rifting pulses and sedimentation pattern in the ~ameli Basin, southwestern Anatolia, Turkey. In: KELLING, G., ROBERTSON, m. H. F. & VAN BUCHEM, F. H. P (eds), Cenozoic Sedimentary Basins of South Central Turkey. Sedimentary Geology, Special Issue, 173(1-4), 409-431. ALTINLI, I. E. 1955. The geology of southern Denizli. Revue de la Facultd des Sciences de l'UniversitO d'IstanbuL Sbrie B: Sciences Naturelles, 20(1-2), 1-47. ALTUNZL, E. & HANCOCK, P. L. 1993. Morphology and structural setting of Quaternary travertines at Pamukkale, Turkey. Geological Journal, 28, 335-346. BARKA,A., REILINGER,R., SAROGLU,F. & SENGI)R,A. M. C. 1997. The Isparta Angle: its importance in the neotectonics of the Eastern Mediterranean region. In: PI,SKIN, (~., ERGO/q, M., SAVA,SGIN,M. Y. & TARCAN, G. (eds) International Earth Science Colloquium on the Aegean Region, 9-14 October 1995, Proceedings, 1, 3-17, Izmir, Turkey. BECKER-PLATEN, J. D. 1970. Lithostratigraphische Unterschungen im Kanozoikum Siidwest Anatoliens ( Tiirkei) ( Kanozoikum und Braunkohlen der Turkei). Beihefte zum Geologischen Jahrbuch, 97. BOZKURT, E. 2001. Neotectonics of Turkey--a synthesis. Geodinamica Acta, 14, 3-30. BOZKURT, E. 2003. Origin of NE-trending basins in western Turkey. Geodinamica Acta, 16, 61-81. COLLINS, A. & ROBERTSON, A. H. F. 1998. Processes of Late Cretaceous to Late Miocene episodic thrust-sheet translations in the Lycian Taurides, SW Turkey. Journal of the Geological Society, London, 155, 759-772. COLLINS, A. & ROBERTSON, A. H. F. 1999. Evolution of the Lycian Allochthon, western Turkey, as a north-facing Late Palaeozoic to Mesozoic rift and passive continental margin. Geological Journal, 34, 107-138. COLLINS, A. & ROBERTSON, A. H. F. 2003. Kinematic evidence for Late Mesozoic-Miocene emplacement of the Lycian Allochthon over the Western

609

Anatolide Belt, SW Turkey. Geological Journal, 38, 295-310. DE GRACIANSKY, P. C. 1972. Rkcherches gkologiques dans le Taurus Lycien Occidental. DSc thesis, Universit6 de Paris-Sud. DUMONT, J. F., POISSON, A. & SAHINCI, A. 1979. Sur l'existence de coulissement s6nestre r6cent a l'extr6mit6 orientale de l'arc 6g6en (sud-ouest de la Turquie). Comptes Rendus de l'Acadkmie des Sciences, 289, 261-264. ERAKMAN, B., ME~HUR, M., GUL, M. A., ALKAN, H., OZTA~, Y. & AKPINAR, M. 1982. Toros projesine ba~,h Kalkan-K6yce~iz-~ameli-Tefenni arasmda kalan alanm jeolojisi ve hidrokarbon olanaklarz raporu. Natural Oil and Natural Gas Company of Turkey (TPAO), Ankara, Technical Report, 1732. EYIDOGAN,H. & BARKA,A. 1996. The 1 October 1995 Dinar earthquake, SW Turkey. Terra Nova, 8, 479-485. FLECKER, R., ROBERTSON, A. H. F., POISSON, m. & MOLLER, C. 1995. Facies and tectonic significance of two contrasting Miocene basins in south coastal Turkey. Terra Nova, 7, 221-232. FLECKER, R., POISSON, A. & ROBERTSON, A. H. F. 2005. Facies and palaeogeographic evidence for the Miocene evolution of the Isparta Angle in its regional Eastern Mediterranean context. In: KELLING, G., ROBERTSON, A. H. F. & VAN BUCHEM, F. H. P. (eds) Cenozoic Sedimentary Basins of South Central Turkey. Sedimentary Geology, Special Issue, 173(1-4), 277-314. GLOVER, C. & ROBERTSON,A. H. F. 1998. Neotectonic intersection of the Aegean and Cyprus tectonic arcs: extensional and strike-slip faulting in the Isparta Angle, SW Turkey. Tectonophysics, 298, 103-132. Guo, L. & RIDING, R. 1998. Hot-spring travertine facies and sequences, Late Pleistocene Rapolano Terme, Italy. Sedimentology, 45, 163-180. HEIMANN, A. & SASS, E. 1989. Travertines in the northern Hula Valley, Israel. Sedimentology, 36, 95-108. KOfYI6IT, A., ONAY, E. & SARAf, G. 2000. Episodic graben formation and extensional neotectonic regime in west Central Anatolia and the Isparta Angle: a case study in the Ak~ehir-Afyon Graben, Turkey. 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, 405421. MARRETT, R. A. & ALMEND1NGER, R. W. 1990. Kinematic analysis of fault-slip data. Journal of Structural Geology, 12, 973-986. MCCLUSKY, S., BALASSAANIAN,S., BARKA, A., et al. 2000. Global Positioning System constraints on plate kinematics and dynamics in the eastern Mediterranean and Caucusus. Journal of Geophysical Research, 105(B3), 5695-5719. ME~HUR, M. & AKPINAR, M. 1984. Yata~,an-MilasBodrum-Karacasu-Kale-A ctpayam- Tavas civarlarmm jeolojisi ve petrol olanaklart. Natural Oil and Natural Gas Company of Turkey (TPAO), Ankara, Technical Report, 1963.

610

M.C. AL(~It~EK ETAL.

OZKUL, M., VAROL, B. & AL~I~EK, M. C. 2002. Depositional environments and petrography of Denizli travertines). Bulletin of Mineral Research and Exploration of Turkey, 125, 13-29. POSTMA, G. 1990. Depositional architecture and facies of river and fan deltas: a synthesis. In: COLELLA,A. & PRIOR, D. B. (eds) Coarse Grained Deltas. International Association of Sedimen to Logists. Special Publications, 10, 13-27. PRICE, S. P. & SCOTT, B. C. 1991. Pliocene Burdur basin, SW Turkey: tectonics, seismicity and sedimentation. Journal of the Geological Society, London, 148, 345-354. PRICE, S. P. & SCOTT, B. C. 1994. Fault-block rotations at the edge of a zone of continental extension, south-west Turkey. Journal of Structural Geology, 16, 381-392. PURVlS, M. & ROBERTSON, A. H. F. 2004. A pulsed extension model for the Neogene-Recent E-Wtrending Ala~ehir Graben and the NE-SWtrending Selendi and G6rdes Basins, western Turkey. In: TAYMAZ, T., WESTAWAY, R. & REILINGER, R. (eds) Active Faulting and Crustal Deformation in the Eastern Mediterranean Region. Tec tonophysics, 391(1-4), 171-201. PURVIS, M. & ROBERTSON, A. H. F. 2005a. Sedimentation of the Neogene-Recent Ala~ehir (Gediz) continental graben system used to test alternative tectonic models for western (Aegean) Turkey. In: KELLING, G., ROBERTSON, A. H. F. & VAN BUCHEM, F. H. P (eds) Cenozoic Sedimentary Basins of South Central Turkey. Sedimentary Geology, Special Issue, 173(1-4), 373408. PURVlS, M. & ROBERTSON, A. H. F. 2005b. Miocene sedimentary evolution of the NE-SW-trending Selendi and G6rdes Basins, W Turkey: implications for extensional processes. Sedimentary Geology, 174, 31-62. ROBERTSON, A. H. F. & DIXON, J. E. 1984. Introduction: aspects of the geological evolution of the eastern Mediterranean. In: DXXON,J. E. & ROBERTSON, A. H. F. (eds) The Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 1-74. ~ENEL, M. 1997a. Geological maps of Turkey in 1:100 000 scale. Fethiye L8 sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara. ~ENEL, M. 1997b. Geological maps of Turkey in 1:100000 scale: Fethiye M8 sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara. ~ENEL, M. 1997c. Geological maps of Turkey in 1:100 000 scale: Denizli K9 sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara. ~ENEL, M. 1997d. Geological maps of Turkey in 1:250 000 scale: Isparta sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara. ~ENEL, M. 1997e. Geological maps of Turkey in 1:250 000 scale: Antalya sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara.

~ENEL, M. 1997f. Geological maps of Turkey in 1.'250 000 scale: Fethiye sheet. Mineral Research and Exploration Directorate of Turkey (MTA), Ankara. ~ENGOR, A. M. C. 1987. Cross-faults and differential stretching of hanging walls in regions of low angle normal faulting; examples from western Turkey. In: COWARD,M. P., DEWEY, J. F. & HANCOCK, P. L. (eds) Continental Extensional Tectonics. Geological Society, London, Special Publications, 28, 575589. ~ENG6R, A. M. C. • Y1LMAZ, Y. 1981. Tethyan evolution of Turkey: a plate tectonic approach. Tectonophysics, 75, 181-241. ~ENGt)R, A. M. C., Goal]R, N. & ~AROGLU, F. 1985. Strike-slip faulting and related basin formation in zones of tectonic escape: Turkey as a case study. In: BIDDLE, K. T. & CHRISTIE BLICK, N. (eds) Strike-Slip and Basin Formation. SEPM (Society for Sedimentary Geology) Special Publications, 37, 227-264. SEYITO6LU, G. & SCOTT, B. C. & RUNDLE, C. C. 1996. Timing of Cenozoic extensional tectonics in west Turkey. Journal of the Geological Society, London, 149, 533-538. TAYMAZ, T. & PRICE, S. P. 1992. The 12.05.1971 Burdur earthquake sequence, a synthesis of seismological and geological observations. Geophysical Journal International, 108, 589-603. TAYMAZ, T., JACKSON, J. & MCKENZIE, D. 1991. Active tectonics of the north and central Aegean Sea. Geophysical Journal International, 106, 433490. TEMIZ, H., POISSON, m., ANDRIEUX, J. & BARKA, A. 1997. Kinematics of the Plio-Quaternary BurdurDinar cross-fault system in SW Anatolia (Turkey). Annales Tectonicae, 11, 102-113. TEMIZ, H., POISSON, A. & ANDRIEUX, J. 2001. The Plio-Quaternary extensional system of the western side of the Isparta angle in SW Turkey. In: 4th International Symposium on Eastern Mediterranean Geology, 1sparta, Turkey, 125-129. TEN VEEN, J. H. 2004. Extension of Hellenic forearc shear zones in SW Turkey: the PlioceneQuaternary deformation of the E~en (~ay Basin. Journal of Geodynamics, 37, 181-204. TEN VEEN, J. H. & KLEINSPEHN, K. L. 2002. Geodynamics along an increasingly curved convergent plate margin: Late Miocene-Pleistocene Rhodes (Greece). Tectonics, 21, 10.1029/ 2001TC001287. TEN VEEN, J. H. & KLEINSPEHN, K. L. 2003. Incipient continental collision and plate-boundary curvature: Late Pliocene-Holocene transtensional Hellenic forearc, Crete, Greece. Journal of the Geological Society, London, 160, 161-181. TEN VEEN, J. H. & POSTMA, G. 1999. Roll-back controlled vertical movements of outer-arc basins of the Hellenic subduction zone (Crete, Greece). Basin Research, 11,243-266. TEN VEEN, J. H., WOODSIDE, J. M., ZITTER, T. A. C., DUMONT, J. F., MASCLE, J. & VOLKONSKAIA,A.

NEOTECTONIC ~AMELI BASIN, SW TURKEY 2004. Neotectonic evolution of the Anaximander Mountains at the junction of the Hellenic and Cyprus arcs. Tectonophysics, 391, 35-65. WESTAWAV, R. 1990. Block rotations in Western Turkey, 1. Observational evidence. Journal of Geophysical Research, B, 95, 1985%19884. WESTAWAY, R., PRINGLE, M., YURTMEN, S., DEMIR, T., BRIDGLAND, D., ROWBOTHAM, G. & MADDY, D. 2003. Pliocene and Quaternary surface uplift of western Turkey revealed by long-term river terrace sequences. Current Science, 84(8), 1090-1101.

611

YILMAZ, Y., GENff, ~. C., GORER, F., et al. 2000. When did the western Anatolian grabens begin to develop? 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, 353-384. ZANCHI, A., KISSEL, C. 8z TAPIRDAMAZ, C. 1993. Late Cenozoic and Quaternary brittle continental deformation in western Turkey. Bulletin de la Socidtd Gdologique de France, 164, 50%517.

Tectonic and sedimentary evolution of the Cenozoic Hatay Graben, Southern Turkey: a two-phase model for graben formation S A R A H J. B O U L T O N 1, A L A S T A I R H . F. R O B E R T S O N

1 & U L V I C. U N L U G E N ( ~ 2

1School o f GeoSciences, Grant Institute, Edinburgh University, Edinburgh E H 9 3JW, UK (e-maik sarah, boulton@glg, ed. ac. uk) 2Department o f Geological Engineering, ~ukurova University, Balcali, 01330 Adana, Turkey New structural and sedimentary studies form the basis of a new interpretation for the Neogene Hatay Graben. Fault analysis reveals three contemporaneous trends of fault orientation (000~ ~ 045~ ~ and 150~ ~ suggesting that the graben is transtensional in nature. Sedimentary studies show that, following shallow-marine deposition from the Late Cretaceous to the Eocene, a hiatus ensued until Early Miocene fluvial sedimentation. After a Mid-Miocene marine transgression, water depths increased until the Messinian salinity crisis, followed by a regression from the Pliocene to the present day. The basin initially developed as the distal margin of a foreland basin of the Tauride allochthon to the north, developing a classic sedimentary sequence during Mid-Late Miocene. Stresses caused by loading of the crust created a flexural forebulge to the south that supplied sediment mainly northwards. During the Plio-Quaternary, transtensional graben development took place, primarily influenced by the westward tectonic escape of Anatolia along the East Anatolia Fault Zone and left-lateral offset along the northward extension of the Dead Sea Transform Fault. This area is, thus, an excellent example of a foreland basin reactivated in a strike-slip setting. Our new two-phase model: foreland basin, then transtensional basin for the Hatay Graben, is in contrast to previous models, in which it was generally assumed that the Plio-Quaternary Hatay Graben represents a direct extension of the Dead Sea Fault Zone or the East Anatolian Fault Zone. Abstract:

The H a t a y region is located in the Eastern Mediterranean, near the border between Turkey and Syria (Fig. 1). This is an area of active neotectonics where three major structural lineaments intersect: the southeastern end of the East Anatolian Fault Zone (EAFZ), the northern end of the Dead Sea Fault Zone (DSFZ) and the Cyprus Arc (Peringek & Cemen 1990; Robertson 1998b). This study focuses on a prominent graben that trends N E - S W from near the town of Antakya to the Mediterranean Sea. This area was previously considered as an extension of a graben, variously known as the H a t a y Graben (Perin~ek & Cemen 1990), the Amanos Fault Zone (Lyberis et al. 1992), or the Karasu Rift (Lovelock 1984; Westaway 1994; Rojay et al. 2001; Over et al. 2002), and appears to link the E A F Z with the D S F Z through the Amik Plain (Fig. 1). Here, we apply the term Karasu Rift to the graben that trends northwards from the Amik Plain, whereas we will refer to our study area to the SW as the H a t a y Graben. The structural complexity of this area has led to a number of different tectonic interpretations. Tinkler et al. (1981) and Parlak et al. (1998)

considered the Karasu Rift as a northward continuation of the DSFZ, whereas Arpat & S,aro~lu (1972) saw the E A F Z as a whole as a continuation of the DSFZ. Hempton (1987), Perin~ek & (~emen (1990) and Westaway (1994) argued that the E A F Z runs from Karhova to a position to the north of Cyprus and that it is not connected to the DSFZ. Others considered the Karasu Rift to be an extension of the EAFZ, which continues through the H a t a y Graben and then offshore ($eng6r et al. 1985; Lyberis et al. 1992; Saro~lu et al. 1992), and thus separate from the DSFZ. In addition, Yfiriir & Chorowicz (1998) viewed the Karasu Rift as a separate structure from both the E A F Z and DSFZ. Each of these tectonic interpretations was proposed following work mainly on the E A F Z and the DSFZ. To date, research on the H a t a y Graben (Delaloye et al. 1980; Tinkler et al. 1981; Pipkin et al. 1986; Robertson 1986; Pirazzoli et al. 1991; Safak 1993a, b; Over et al. 2002) and on the Karasu Rift (Parlak et al. 1998; Rojay et al. 2001; Yurtmen et al. 2002; C)ver et al. 2004) has mainly focused on the regional tectonic setting and significance, whereas the sedimentary and

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 613-634. 0305-8719106l$15.00 9 The Geological Society of London 2006.

614

S.J. BOULTON E T A L .

'

EAFZ~

Ceyh~in

"---k

ye/

Lineamer

Iskenderun

9

ii km , " r - r AmikPlait

ANATOLIAN MICROPLATE

EcemiwFZ ,//. EAFZ

Misis-AndmnLineament

RiverAsi I ~ . ~ . . .

Taurides

DSFZ

N- ,qe~',~L fL~Crete _A

~ r r a ~

/{t,X/r',,ny

Cyprus 'P

Ridge

AreaT(

~ ' ~ "~4;k""~ """ AFRICANPLATE 0 ~ km ~ ~ ~ 24~'

281

/x~i[' wr" ~ rn

~ aSFZ /l

I

36]

Fig. 1. Regional tectonic setting of the Eastern Mediterranean; the box shows the location of the inset map. Inset: outline map of the Hatay region, southern Turkey, showing major structures and some locations referred to in the text; box shows the location of Figure 2.

structural processes involved and the overall tectonosedimentary evolution have received little attention. In this study we have investigated the evolution of the Hatay Graben, utilizing new sedimentological and structural studies to test existing models, and if appropriate develop a new one. We will show that the existing models are inadequate. The formation of the Hatay Graben predates the formation of the EAFZ, N A F Z and the northern DSFZ. The area was first part of the foreland basin to the Bitlis suture zone and normal faulting began during the Mid-Miocene, we suggest, as a result of isostatic loading. Following the cessation of continental collision in SE Turkey, strike-slip faulting and transtension become important in the Mid-Miocene and led to the formation of the present Hatay Graben. The Hatay Graben is an excellent example of a basin that has been dissected by normal faults in a foreland basin setting and later involved in strike-slip tectonics related to regional tectonic

escape, in this case Anatolia westwards towards the Aegean Arc.

Regional geological setting The post-Cretaceous tectonics of the Eastern Mediterranean region has resulted from the relative motions of the African, Arabian and Eurasian plates (McKenzie 1978). The northward motion of Africa and Arabia towards Eurasia has caused convergence and the emplacement of allochthonous units since the Late Cretaceous along what is now the Bitlis Suture Zone (McKenzie 1978; ~eng6r et al. 1985; Robertson 1998a). A foreland basin developed in SE Turkey along the Arabian margin parallel to the convergence front in Early-Mid Miocene time (Ydmaz 1993; Derman et al. 1996; Robertson 1998a, 2002). Continuing north-south-directed compression, as result of the motion of the Arabian plate, led to the uplift and crustal thickening

CENOZOIC HATAY GRABEN, S TURKEY of Anatolia (~eng6r & Kidd 1979) during Late Miocene-Quaternary time. By the Early Pliocene compressional tectonics were replaced by the westward extrusion of the Anatolian microplate along the dextral NAFZ, and the sinistral EAFZ (Fig. 1) (McKenzie 1978). These two strike-slip faults meet at a triple junction near the town of Karhova in Eastern Turkey ($eng6r et al. 1985; Bozkurt 2001). The N A F Z developed during the Mid-Miocene (McKenzie 1978; Seng6r et al. 1985, 2005) to Late Miocene-earliest Pliocene (Barka et al. 2000) and is moving at around 15-25 mm a -1 (Oral et al. 1995; Reilinger et al. 1997). The initiation of the EAFZ is variously dated as Late Miocene to Early Pliocene ($engSr et al. 1985; Dewey et al. 1986; Hempton 1987), Late Pliocene ($aro~lu et al. 1992), 3 Ma (Early Pliocene; Westaway & Arger 1996), or 1.8 Ma (Late Pliocene; Yiirtir & Chorowicz 1998). Global positioning system (GPS) measurements indicate that the current rate of slip along the EAFZ is 11 + 2 mm a -1 (Reilinger et al. 1997). The relative motion between the African and Arabian plates is accommodated by sinistral motion along the DSFZ (Mart & Rabinowitz 1986; Chaimov et al. 1990; Beydoun 1999). The DSFZ links the rifted plate margin in the Red Sea to the convergent Bitlis suture zone in Southern Turkey (Hempton 1987). The southern DSFZ is considered to have formed as a result of the rifting of Arabia from Africa at c. 18-20 Ma (Garfunkel 1981; Hempton 1987; Lyberis 1988; Rojay et al. 2001), possibly in two stages, with the second phase and formation of the northern strands of the DSFZ beginning at c. 4.5 Ma: Early Pliocene (Freund et al. 1968; Girdler & Styles 1978; Brew et al. 2001). A further regionally important strike-slip lineament, the Ecemis Fault Zone, to the NE of the study area (Fig. 1), is considered to show a left-lateral offset of c. 60 km mainly since Mid-Miocene time (Yeti~ 1978; Kogyl~lt & Beyhan 1998; Jaffey & Robertson 2001). Much early displacement of Anatolia was accommodated along this lineament, prior to the Plio-Quaternary activity of the EAFZ.

Sedimentary evolution of the Hatay Graben The topographic basin representing the Hatay Graben runs NNE-SSW from the present-day coast to the north of Antakya city (Figs 1 and 2). The basin has two well-defined topographic margins, which we refer to as the NW and SE margins, respectively. The SE margin exhibits a high topographic relief and is deeply incised by Quaternary streams, compared with the NW margin, which has a lower topographic relief.

615

Near the coast, Quaternary river channel incision has resulted in exposures of Neogene sediments, especially as river terraces. Sedimentary units are exposed on both of the margins and in the axis of the Hatay Graben, ranging in age from the Late Cretaceous to Quaternary (Safak 1993a,b; Pipkin 1986; Fig. 2). L a t e Cretaceous and Eocene ( K i s l a k and O k f u l a r Formations, Table 1.1)

In the SE, the Hatay ophiolite is overlain by a sequence of latest Cretaceous (Maastrichtian) to Mid-Eocene (Lutetian) carbonates. Latest Cretaceous facies are generally grey, medium-bedded (beds generally 20-50 cm thick) wavy and convolute laminated shallow-marine microbial limestones with common secondary chert. In the south, the Eocene carbonates are composed of fine-grained, pale grey to pink limestones that contain abundant large benthic Foraminifera (e.g. Nummulites sp.), suggesting a water depth of 20-100 m (Saller et al. 1993). Further north, Eocene limestones include graded calcarenites with parallel and cross-lamination. The bases of some individual beds are erosional with scour marks; these limestones are interpreted as mainly relatively high-density calcareous turbidites. Latest Cretaceous and Eocene sediments accumulated regionally on the northern margin of the Arabian platform (Rigo de Righi & Cortesini 1964; Yllmaz 1993). During the latest Cretaceous, shallow-marine conditions persisted (0-50 m water depth) and microbial mats formed creating wavy laminated stromatolites that accumulated in the inner intertidal zone. During the Eocene, shelf carbonates rich in benthic Foraminifera accumulated in the SE; water depths increased to an estimated > 100 m further NW, where hemipelagic carbonates and carbonate turbidites accumulated on a slope setting. The Oligocene was a period of non-deposition during which Eocene and older strata were folded and uplifted. Sedimentation resumed during Early Miocene (Aquitanian) time above an angular unconformity. Early M i o c e n e ( B a l y a t a ~ i Formation; Fig. 3, Table 1.1)

Lower Miocene sediments crop out along both of the present margins of the basin but only in the NE of the area (Fig. 3). Along the SE margin, matrix- and clast-supported, polymict conglomerates unconformably overlie Eocene limestones. The contact is occasionally marked by a carbonate interval (c. 2 m thick), rich in shallow-marine bivalves and bored pebbles (e.g. Harbiye area;

616

S.J. BOULTON E T AL.

Fig. 2. Geological map of the Hatay Graben, southern Turkey. Lines indicate positions of cross-sections (Fig. 8) and box indicates location of Figure 5 (palaeocurrent map). Stars indicate the location and present altitude of Messinian evaporite deposits; letters refer to locations described in the text. Numbers in circles refer to localities in the text where evidence for synsedimentary faulting was observed. Based on data collected during this study, and on the work of Pipkin (1986) and Turkish masters degree students (especially A. Kop, T. Mistik and N. Temizhan). Fig. 2). The conglomerates in the lower part of the formation are thick bedded, whereas higher in the sequence coarse conglomerates occur as lenses (tens of metres wide) within mediumgrained sandstone, to conglomerate. Clastic sediments at the top of the formation exhibit pebble imbrication, large-scale cross-bedding and well-developed palaeosols (Fig. 4a). Figure 5 shows palaeocurrent measurements, based on clast imbrication and cross-bedding, as recorded along the SE basin margin. The current flow was generally to the north to NE (00-050 ~ in westerly locations and to the north to NW (00-270 ~ in easterly locations. However, at two locations, measurements indicate south to SE (90~ ~ current directions (Fig. 5). Along the NW margin of the basin very similar conglomerates unconformably overlie the Hatay ophiolite. At some localities the basal sediments are composed of pale, fine-grained, serpentinite-derived, indurated mudstone with scattered serpentinite clasts. Clast abundance increases upwards, passing into matrix-supported conglomerate, in turn overlain by a palaeosol (up

to 25 m thick). This contrasts with the succession on the SE margin, where lenticular conglomerates are instead prominent. The thickness of the Lower Miocene succession on the NW margin varies from 0 to c. 175 m; in contrast, the succession reaches a maximum thickness of c. 300 m along the SE margin. On both margins the Lower Miocene succession thins and disappears towards the present coast. The Lower Miocene succession is interpreted as a braided-river environment, in which sheetflow processes dominated, forming the laterally continuous stratigraphically lower conglomerates. Higher in the formation, as observed in the SE, conglomerates are confined to channels, forming lenticular bodies; soils formed within inter-channel areas. A decrease in clast size upwards and the presence of foresets at the very top of the formation may indicate a decreasing palaeoslope with time. Palaeocurrent data, although variable, show that flow was generally to the north or west. The differences in palaeocurrent directions between the easterly and westerly exposures partly reflect a combination of

CENOZOIC HATAY GRABEN, S TURKEY

617

Table 1. Age, lithology and microfossil data for the sediments of the Hatay Graben

Formation Name

Age

Lithology

Selected Microfossils

Samanda~

Pliocene

Marl and sandstone

Vakxfll Member Nurzeytin

Messinian SerravalianTortonian Langhian

Marl, limestone and sandstone Limestone

K1~lak

AquitanianBurdigalian Lutetian

Conglomerates and palaeosols Limestone and marl

Okgular

Lutetian

Limestone

Kalebo~am

Late Cretaceous

Limestone and sandstone

Globerinoidesruber, Globorotalia scitula, Globigerinoides trilobus, sacculifer* Globoquadrina altispira Orbulinauniversa, Hastigerina rom. sp., Orbulina suturalis* Praeorbulina gloerosa curva, Orbulina suturalis* Globergerinoidestriobus, Globergerinoides ruber* Acarinina bullbrooki, Morozovella spirulosa* Morozovella aragonensis, Globigerina ineqispira* Globotruncana arca, Globotruncana gansseri, Globotruncana mayaroensist

Sofular Balyata~l

*$afak 1993a. -~Pi~kinet al. 1986.

[ lOm

I Quaternary Alluvium

~

Pliocene: Samandac~ Fm.

100m

~ t

Messinian' Vahfh Mb.

5-20m

Uooe Miocene:

300m

Nurzeytin Fm, m

O-300m I I

I I

I

I

Middle Miocene ' Sofular Fm. Lower Miocene: Balyata~l Fm.

O-300m ?,s% /-, /

Hatay ophiolite

Fig. 3. Generalized stratigraphic column of the stratigraphic succession, not to scale. regional and local palaeotopographic controls. The Hatay Graben was not clearly present in its current form, as braided rivers flowed northwards over an undissected topography, in contrast to the present flow through a dissected and

faulted topography westwards into the Mediterranean Sea. The anomalous southerly flow data were recorded near the top of the formation and might record bypass of the basin and the inception of flow towards the present coast. The dominantly ophiolitic composition of the conglomerates suggests a source from ophiolites that were regionally emplaced southwards over the Arabian Platform during latest Cretaceous time (Yflmaz 1993; Robertson 2002). As the flow was dominantly northwards and westwards the main source was probably ophiolites located to the south or east of the Hatay Graben, as now exposed in the K m l d a ~ Massif to the west and in the Ba~r-Bassit Massif to the SE, in Syria. M i d d l e M i o c e n e ( S o f u l a r Formation; Fig. 3, Table 1) The base of the Middle Miocene succession overlies Lower Miocene conglomerates in the north of the area on both sides of the graben. The basal contact is sharp with small, localized angular unconformities ( < 5~ Additionally, on the N W margin in the SW of the area these sediments overlie an eroded surface of the ophiolite (Fig. 4d). Along the N W margin the thickness of the succession increases from the NE to SW from 1-2 m to a maximum of c. 150 m. The basal Middle Miocene sediments are bioclastic limestone, generally wackestones, characterized by shallow-marine fauna including abundant bivalves, gastropods, corals and echinoids, together with oncolites, indicating low-energy conditions. In addition to bioclastic carbonates,

618

S.J. BOULTON E T A L .

Fig. 4. (a) View to the east of the type section of the Balyata[gl Formation (Lower Miocene) near the village of Enek; (b) thick turbiditic sand beds in marl (Nurzeytin Formation); car for scale; (c) dewatering pipe in alternating beds of hard and soft marly limestone with chert nodules, observed in the upper part of the Middle Miocene sequence at ~evlik; (d) Middle Miocene sediments overlying serpentinite along an irregular erosion surface. (Note the large gastropod in the top right corner; pen for scale, bottom right.)

Fig. 5. Location map and rose diagrams showing the directions of palaeocurrent flow of the Lower Miocene conglomerates (Balyat~l Formation shaded in grey) along the SE margin of the basin. Box corresponds to the area shown in Figure 2.

sandstones with low-angle cross-bedding (indicating southward current flow), conglomerates and palaeosols are present near the base of the succession in the NW. The sediments exposed in coastal exposures on the N W margin are typically bioturbated, bioclastic calcirudites. Sedimentary structures are uncommon but parallel laminations and a dewatering structure (Fig. 4c) were observed locally. Along the SE margin of the basin, the succession thickens towards the south, to a maximum of c. 300 m and is similar to that seen on the N W margin. In the south, at Kozkalesi (Fig. 2), the lower part of the formation includes repeated palaeosol horizons with sharp bases, grading into bioclastic limestones (Fig. 6). The total thickness of these cycles exceeds 100m. Upwards, palaeosol horizons gradually disappear and bioclastic calcirudites dominate the succession, rich in oncolites and reworked bioclastic material. The Middle Miocene succession thins to the NE, as a result of greater subsidence in this area. The sharp base of this formation, without

CENOZOIC HATAY GRABEN, S TURKEY

< 200 m for the upper part of the formation in this area. The thick succession on the N W basin margin shows evidence of sediment instability and gravity reworking (i.e. dewatering structures; slumps).

Q 5m ;;0 2~11=!--

"O

I1[~1 m

m m

_

r

m

m

5" i

tQ e"

m

m =

b

R

0m

•

Marly Limestone Limestone Palaeosol

~ .

Conglomerate Fossils

619

Q Oncolite

Fig. 6. Representative log showing the sedimentary

cycles observed at Kozkalesi (Fig. 2). erosional features, is suggestive of a rapid marine transgression. Local unconformities at the base of the formation suggest that the palaeotopography of the basal contact was irregular. Initially, water depths were very shallow (c. 0-10 m) along both basin margins, as there is evidence of coastal and non-marine processes (i.e. low-angle cross-bedding and soil formation), together with a coral build-up at one locality. The carbonates observed at Kozkalesi are interpreted as peritidal cyclothems on a carbonate platform. Sedimentation kept place with subsidence initially but the rate of subsidence apparently increased with time, resulting in deeper-water conditions possibly caused by tectonic subsidence; also there was an eustatic sea-level rise during this time (Haq et al. 1987). Planktic:benthic foraminiferal ratios (Meschede et al. 2002) suggest a water depth of

Late Miocene (Nurzeytin Formation; Fig. 3, Table 1) The contact between the Middle and the Late Miocene successions is gradational, marked by 5-20 m of interbedded bioclastic limestone and marl. This contact is defined as the level at which marl exceeds bioclastic limestone. A thick (c. 300 m) marl sequence dominates the Upper Miocene succession and is exposed both within the present topographic graben and to the SE of the basin margin. The marl is a relatively uniform medium grey, very fine-grained, well sorted and contains a rich fauna of benthic Foraminifera (e.g. Uvigerina peregrina) and planktic Foraminifera (e.g. Orbulina universa), together with occasional bivalves. Planktic:benthic foraminiferal ratios suggest a water depth of up to 700 m (Meschede et al. 2002). White mica (muscovite) is not present in the lower part of the marl sequence; however, in more northerly exposures outside the Hatay Graben, near Belen (Fig. 1), and higher in the basin sequence as a whole the sediments become markedly micaceous. Numerous beds composed of mixed calcareous and terrigenous material are interbedded with the marl; these beds vary from 1-2 cm to > 2 m thick. In the SW of the basin, calcarenite beds exhibit erosive bases with flute and groove casts, together with parallel and cross-lamination. Other calcarenites are massive, with bed thicknesses ranging from 0.1 to 2 m. In the same area a matrix-supported conglomerate horizon > 5 m thick was observed. The matrix of this conglomerate is composed of grey marl with clasts of calcarenite and marl, up to 2-3 m in size. Further NE, interbeds are composed of medium- to coarse-grained litharenite, in beds 0.05-2 m thick (Fig. 4b). These beds are mostly massive, although parallel- and cross-lamination, ripples, flutes and mud rip-up clasts were locally observed. These structures yielded rare palaeocurrent directions; these are very variable (090o-300 ~) but generally are orientated towards the axis of the graben. On the SE flank of the basin (outside the modern topographic graben), there is a similar thickness of marl, but litharenite interbeds are absent. The marl sequence is interpreted as background sedimentation in a relatively deep-water setting. The interbedded coarse sediments were probably reworked downslope as turbidites,

620

S.J. BOULTON ET AL.

grain-flows and low-density debris-flows, reflecting instability of the basin margins. The lack of reworked material within the Upper Miocene succession outside the present basin suggests that at least some basin topography had developed by this time, causing sediment flows to bypass the relatively higher basin flanks and be deposited on the basin floor. The presence of muscovite in the stratigraphically higher sediments is interesting as there is little or no mica present in the basement rocks of the Hatay region; this suggests that this material is extrabasinal and was probably derived from the Tauride Mountains to the north.

Messinian ( Vaklfh Member; Fig. 3, Table 1) During the Messinian, the Mediterranean as a whole was affected by the Messinian salinity crisis (Hsfi et al. 1978; Krijgsman et al. 1999). However, perhaps reflecting its marginal setting, only four evaporite localities are known in the Hatay Graben; three of these are near the axis of the modern graben and one on the SE margin (Fig. 2). The present altitude of gypsum ranges from 130m above sea level (a.s.l.) near the graben axis to 320 m a.s.1. (Fig. 2) in the SE. The thickest gypsum deposit (25 m) is mainly composed of fine-grained alabastrine gypsum (location a; Fig. 2). The exposure includes large angular blocks ( > 2 m) of laminated alabastrine gypsum set in a gypsiferous marl matrix. In places, the alabastrine gypsum has undergone diagenetic alternation to coarse selenitic gypsum. Other sequences (5-10 m thick) comprise coarsegrained selenitic gypsum (locations b, c, d; Fig. 2). Exposures b and c consist of massive selenitic gypsum. It is not clear if this is primary or diagenetic. Location d, by contrast, consists of several exposures where a succession can be measured. The basal gypsum is made up of bandedstacked selenite (e.g. as reported from Cyprus; Robertson et al. 1995), with repeated layers of selenite crystals, 1-5 cm in size. The upper part of the sequence is composed of thick (> 1 m), massive, fragmented selenite crystals, 5 cm or more in size, interpreted as debris-flow deposits. The gypsum formed when the basin became semi-isolated from the Mediterranean Sea as a result of a falling sea level (Hsfi et al. 1988). The fine-grained albastrine gypsum probably precipitated at the sediment-water and air-water interfaces (e.g. Schreiber et al. 1976). After precipitation, the gypsum was probably reworked into local depocentres, forming the banding seen at location d. The selenitic gypsum formed in a very shallow sub-aqueous marginal environment.

We interpret the alabastrine gypsum (location a; Fig. 2) as material that was reworked towards local depocentres, in line with its present position near the axis of the modem graben. By contrast, the selenitic gypsum at locality d (Fig. 2) probably formed near the margin of the basin in a very shallow-water environment. The broken selenite crystals at the top of this succession possibly represent gypsum debris flows that were triggered by sea-level change or tectonic activity.

Pliocene (Samanda~ Formation; Fig. 3, Table 1) Pliocene sediments crop out only near the modern basin axis. Following the Messinian, a Pliocene transgression resulted in a return to marl deposition (Hsfi et al. 1978). The marls resemble those of the underlying Upper Miocene succession but now contain a diverse shallow-marine fauna and common plant material. Within this marl sequence coarse siliciclastic horizons can be observed near the axis of the graben, for example, as thin ( < 5 cm), lenticular uncemented sand horizons ( < 1 m) with parallel lamination and rip-up clasts. There are also matrix-supported lenticular conglomerates, with subangular to rounded clasts (up to 1 m in size). Graded sandmud horizons with erosive bases and tops can also be observed within the marl. Elsewhere, in the SW, Lower Pliocene sediments are composed of coarse-grained litharenite, rich in bivalves, often as discrete horizons, lacking sedimentary structures. Low-angle cross-bedding, ripples, parallel lamination and conglomerate lenses are present higher in the succession. Palaeocurrent directions are variable and show a range of directions from 060 ~ to 225 ~. In contrast, the Upper Pliocene succession is generally composed of poorly cemented, coarse-grained, orangeweathering litharenites that are massive bedded and contain no bioclastic debris. Following the Early Pliocene transgression, normal marine sedimentation resumed. The ratios of planktic:benthic Foraminifera (c. 0.9) suggest a water depth of < 200 m. Background marl sedimentation was interrupted by the input of siliciclastic material by gravity processes. Exposures in the SW of the basin are typical of coastal deposits. Low-angle cross-bedding is associated with small gravel channel structures that are composed of rounded and sorted pebbles; these features are typical of beach processes. The succession probably represents sequence shallowing upwards from lower shoreface (sandy marl with bivalve lags) to upper shore-face and beach. By the end of the Pliocene, relative sea level had fallen further and the Hatay Graben became non-marine.

CENOZOIC HATAY GRABEN, S TURKEY

Quaternary alluvium (Fig. 3) Quaternary sediments are composed of coarse sands, gravels and conglomerates. Sediments are preserved in four main river terraces that formed progressively as the Asi Nehir (Fig. 1) cut progressively downwards into the underlying strata, towards the present coastline. The coarse-grained sediments of these terraces are similar in composition to those of the modern river (mainly carbonate clasts and serpentinite), and are generally formed of subrounded to rounded clasts with little or no matrix. The sediments are usually massive bedded but large (2.5m) high-angle cross-beds and erosional features are present in some exposures. In addition, Quaternary fault talus composed of poorly sorted, angular clasts and palaeosols was observed adjacent to major fault planes, especially to the south of Antakya. Also, around the town of Harbiye, tufa (cool-water carbonate) > 50 m thick was locally precipitated from streams flowing down a high-angle fault scarp. The Quaternary fluvial facies accumulated from a river system, characterized by meandering to braided channels, much as today. The progressively lower position of the terraces was caused by a relative sea-level fall, causing incision. Raised beaches, marine erosion notches and benches, and remnants of bioconstructed rims are found along, or near, the present coastline and these have been used to document two phases of rapid late Quaternary uplift (Pirazzoli et al. 1991).

Synsedimentary deformation Synsedimentary structures can be used to determine the relative timing of faulting. Key features are growth faults, sediment packages thickening into normal faults (i.e. sediment fanning), intraformational faults and phases of fault motion, as inferred from fault-derived talus. Synsedimentary features are absent from the Lower Miocene succession. However, three growth faults were identified within the Middle Miocene succession on the SE basin margin, near Kozkalesi (Fig. 2, location 1). Limestones are displaced by normal faults that dip northwestwards towards the axis of the basin and strike NE-SW. In two of these exposures there is a greater sediment thickness on the hanging-wall block compared with the footwall; undeformed strata overlie these faults (Fig. 7). In addition, within an upper Middle Miocene coastal exposure on the N W basin margin, beds at the base dip more steeply (35 ~ than those at the top (25 ~ as a result of sediment fanning (Fig. 2, location 2).

621

Sediment fanning was observed within Upper Miocene sediments at two localities elsewhere: one on the basin axis (Fig. 2, location 3) and one on the SE margin between two basin-bounding faults (Fig. 2, location 4). At the basin axis locality, the fanning (observed in a valley) was revealed by the difference in the angle of dip between the upper and lower beds (c. 10~ At location 4 (Figs 2 and 7), along the River Asi, the lower beds are subvertical, with the dip gradually decreasing upwards to c. 30 ~ These fanning sediments thicken towards the SE graben margin and it is possible that they thicken towards a graben-bounding normal fault. Pliocene sandstones are highly deformed adjacent to major basin-bounding faults, with dips as high as to 90 ~ locally (Fig. 2, location 5), implying that these faults are Pliocene or older. Microfaulting is commonly developed within axial Pliocene sediments; also, a growth fault and evidence of slumping were observed at location 6 (Fig. 2). Several angular discordances were observed within a Quaternary talus cone adjacent to a major fault bounding the SE margin of the graben (near the village of Dursunlu; Fig. 2). It is inferred that the talus was derived from the adjacent exposed fault scarp; this fault then moved, rotating the pre-existing talus and producing more material that was deposited on top of the original sediment along a discontinuity (Fig. 2, location 7). This process was repeated, creating multiple small-scale discontinuities within the talus fan. Only occasional faults were observed within the Quaternary deposits, probably because of the difficulty of recognition in such coarse and poorly consolidated sediments. However, rare faults were identified and locally the boundary between Pliocene sandstone and Quaternary conglomerate is faulted, confirming that fault motion has taken place during the Quaternary (Fig. 2, location 8). The synsedimentary features described above suggest that growth faulting began in the MidMiocene. During the late Mid-Miocene and Late Miocene fault motion resulted in the tilting of bedding and the creation of accommodation space, creating the local sediment fanning. The present elevation of the Messinian evaporites (up to 320 m; Fig. 8) suggests that there has been significant post-Messinian fault uplift; evaporites on the basin margin are now 190 m higher in altitude than similar deposits near the basin axis. This, in turn, suggests that after the Messinian the basin underwent a phase of subvertical fault movement. Faulting additionally deformed the

622

S.J. BOULTON E T AL.

Fig. 7. (a) Photograph of a growth fault observed near the village of Kozkalesl in Middle Miocene limestone. (b) Sketch of the geometry of the fault. It should be noted that the lower intervals (a and a') are displaced by the fault but are of the same thickness. Upwards, interval b has a greater thickness than b'. The throw on the fault increases downwards. The upper package of strata is not cut by the fault. Therefore, the fault began moving after time A, and was moving during the deposition of b but had ceased moving when the upper layer was deposited. (c) Photograph showing the Middle Miocene fanning sediments exposed along the River Asi. (d) Sketch of the fanning sediments; it should be noted that the dip of the bedding decreases up section. In the middle of the section is a bored intraformation unconformity; below are fine limestones (mudstone) and above are dominantly bioclastic calcirudites with some conglomerate horizons.

Fig. 8. Structural cross-sections of the Hatay Graben (see Fig. 2 for locations of sections and key).

Pliocene sediments and Quaternary talus throughout the basin. There is also the evidence of palaeoseismic to recent seismic activity from small earthquakes (US Geological Survey National Earthquake Information Centre).

Within the past two centuries the city of Antakya has been devastated by large earthquakes, two notable events occurring on 13 August 1822, M-~ 7.4 and 13 April 1872, M-- 7.2 (Over et al. 2002).

CENOZOIC HATAY GRABEN, S TURKEY

Structure of the Hatay Graben The Hatay Graben is an asymmetrical structure (Fig. 8) trending 030~ ~ The SE margin is characterized by normal faults. A number of en echelon fault segments step away from the axis of the graben to the east, forming two arrays of subparallel faults (Fig. 2). The outer array comprises three main segments, whereas the inner array is shorter with two main fault segments. The greatest throw ( > 200 m) is on the innermost of the major faults. Small ( < 10 km 2) sub-basins have formed on the margins of the graben as a result of the back-rotation of fault blocks. On the N W margin of the graben, mapscale faults (c. 100-200 m of displacement) dip into the graben; however, it appears that these are not as large as the faults bounding the SE margin. In total, over 850 measurements were made of fault planes in the field area. When these data are considered together, the majority of the faults strike between 060 ~ and 320 ~. Three main trends in the strike direction of large faults are recognized (Fig. 9): (1) N E - S W (c. 0350-060 ~ to 215~176 parallel or subparallel to the basin margins; (2) N W - S E (c. 140~ ~ to 320 ~ 340~ orthogonal to the basin margins; (3) north-south (c. 350~ ~ to 170~176 oblique to the basin margins. The majority of the faults trend either parallel to the graben or at a high angle to it (i.e. NW-SE). Normal, oblique, sinistral and dextral strike-slip faults are common, plus rare reverse faults. The direction of dip is variable, with the majority of faults being high angle. To aid interpretation, the data were then divided into subgroups, first by structural domain (Fig. 9a) and then by the maximum age. The maximum age was determined from crosscutting relationships, syndepositional structures and the age of displaced units in which the structures occur. Cross-cutting relationships did not reveal any consistent trend, suggesting that there was only one identifiable phase of deformation. Two sets of slickenlines were observed on a few fault planes. Of these, one set of lineations on a fault plane is commonly oriented at a high angle (dip-slip) and the other at a low angle (oblique or strike-slip). However, it was not possible to confirm if one set of the two slickenlines was the younger, based on only a small number of measurements (n = 13). When the faults are considered by geographical area (Fig. 9) the graben margins exhibit predominantly basin-parallel normal faults.

623

However, there is also a significant number of faults trending at a high angle to the graben. For example, in zones 4 and 6 numerous faults trend east-west. The faults in zones 2 and 3, covering the axial zone of the graben, are less influenced by basin-parallel faults, although there is still a significant number of normal faults oriented in this direction. The main trends are north-south or N N W - S S E . These faults are predominantly extensional but there is also a significant number of strike-slip (sinistral and dextral) faults. When fault patterns are considered according to the age of the formation in which they are observed, the patterns for each of the age categories, from Eocene to Pliocene, are very similar, with three main trends distinguishable (Fig. 10). Basin-parallel faults are not well represented within the Pliocene sediments, probably because Pliocene sediments are exposed only near the graben axis. The Pliocene also has greater numbers of strike-slip faults compared with other time periods. Faults within the Upper Cretaceous rocks exhibit a more north-south trend and may reflect a pre-existing stress regime. As there is little evidence that the normal and strike-slip faults in the Hatay Graben represent separate stages of faulting (of different age) we consider it likely that these variably trending faults coexisted in a transtensional setting.

Kinematics o f faulting A number of recent studies have investigated the process of oblique extension (transtension), both experimentally (Withjack & Jamieson 1986; Clifton et al. 2000; Tron & Brun 1991; McClay & White 1995) and using field evidence (Umhoefer & Stone 1996; ten Veen & Kleinspehn 2002). Transtension represents a range between two end-members: pure extension and strike-slip (where the trend of the basin is oblique to the extension direction). The acute angle, ~, between the rift trend and the direction of displacement on the plate edge is inversely related to obliquity; thus, a largely oblique regime (i.e. strike-slip basin) exhibits a low value of 0~. In areas of pure extension (~ = 90 ~ the majority of the faults are normal and strike parallel or subparallel to the graben, with only small numbers of strike-slip faults accommodating changes in the amount of extension along strike; these faults strike at a high angle to the boundary faults. In contrast, in pure strike-slip regimes where ~ = 0 ~ two dominant directions of faults occur c. 45 ~ apart, and normal, reverse and strike-slip faults develop within the fault zone. Neither of these scenarios is applicable to the Hatay Graben, where three main directions of faulting were determined. One

624

S.J. BOULTON E T AL.

\

/

'\ \

/

N=857 / largest petal / =25 values

i

b.

a.

~/fzone

/,

5 \'\

2.5 ~ classes o

o

i '

i N=42

zone 1

~.

N=94

zone 2 o

o

N=89

zone3

N=108

zone 4 0

0

N=118 zone 5

,

....

N=200 zone 6

C.

Fig. 9. (a) Rose diagram showing the strike of all the faults measured; (b) sketch map showing the sub-areas used for data analysis; (e) breakdown of fault data by area. Rose diagrams are divided into 5~ classes.

possible explanation is that this area represents an intermediate value of ~. A n a l o g u e experiments have shown that there is a change in the style of faulting between 0~=45 ~ and ~ = 3 0 ~ (Withjack & Jamieson 1986; Clifton et al. 2000).

W h e n ~ > 45 ~ all of the faults are of dip-slip type. Faults near a graben margin will strike slightly obliquely to the main trend, whereas near the axis of the graben faults strike near to the displacement-normal direction (Fig. 11). However, when

CENOZOIC HATAY GRABEN S TURKEY .

.

.

.........

/J/ ,i

.

......

.

.................. ~.......

0

,.

N=85

N=95

)

625

34

~,.....

, ,.

(a)

""

(b)

o

jJJ

(c)

" ...........

o

..,

.................. ~....... ........

.... .<_~.0

.........

. . . ........... iii.. i !IN

.....

Fig. 10. Rose diagrams showing all of the structural data measured for the following periods: (a) Late Cretaceous; (b) Eocene; (c) Early Miocene; (d) Mid-Miocene; (e) Late Miocene; (f) Pliocene. Roses are divided into 5~ classes.

_<30 ~ dip-slip, oblique-slip and strike-slip faults can develop in three populations. One population forms subparalM to the graben trend and two trend at a high angle to the rift and displacement direction. However, when ~ = 15~ the majority of the faults are strike-slip in nature. We, therefore, infer that the deformation in the Hatay Graben reflects oblique extension at the present, where is c. 30 ~ (Fig. 11). A lack of strike-slip faults in older sediments suggests that the angle of extension has decreased over time and prior to the Plio-Quaternary ~ was perhaps > 30 ~ To test this hypothesis and to gain a better understanding of the kinematics of faulting in the Hatay Graben, a stress inversion study was undertaken. This used the Angelier method (Angelier 1984), applied using the computer program Daisy 2.4 (Salvini 2001). The data were analysed using the same groups as before (i.e. age of formation and area). Figure 12 shows that zones 2 and 5 do not have a principal stress axis in the vertical and thus represent a transtensional stress regime. Zones 1, 3 and 6 have cy~positioned in the vertical; this is the maximum stress axis and relates to normal faulting. In these cases the minimum principal axis of stress, cy3, is oriented N E - S W (188 ~ 033 ~ and 210 ~ respectively) and normal faults are expected to trend NW-SE. In contrast, zone 4 has cy2 in the vertical and therefore represents a strike-slip regime. When the fault data are considered by age (Fig. 13),

the stress analysis results in or1 being vertical, thus corresponding to a normal palaeostress regime. The orientation of or3 is within the N E - S W quadrant of the stereonet. This apparently anomalous result is likely to reflect the large number of NW-SE-oriented small faults from which slip data were obtained compared with the relatively small number of large bounding faults along which most of the fault motion took place; these are commonly weathered or slip data are rarely available owing in inaccessibility. Therefore, if more fault data could be collected a clearer result would be expected.

Evolution of the Hatay Graben During latest Cretaceous-Eocene time, shallowmarine carbonates were deposited regionally across the Arabian platform. Sedimentary structures from the area studied indicate the presence of a slope dipping northwards towards the Neotethys Ocean. No sediments are known from Oligocene time, suggesting the area was then emergent and undergoing erosion. Eocene and older strata are folded indicating that fold and thrust tectonics caused uplift during this time; this was probably due to the southward migration of a flexural bulge related to continental collision along the Bitlis Suture to the north (Robertson e t a l . 2004).

626

S.J. BOULTON E T AL.

R

R

I' H1

ot = u

a -

R

30 ~

R I"

-H1

45 ~

or=90 ~

Fig. 11. Rose diagrams of fault trends predicted by a transtensional model (combined extension and sinistral shear). R is the rift trend; the large arrow indicates the direction of extension between the rift margins; EH~is the direction maximum of extensional strain; modified from Withjack & Jamieson (1986).

At the beginning of the Early Miocene the Hatay area remained above sea level and large braided streams fed sediment across the eastern part of the basin, mainly from the south, depositing coarse conglomeratic units. These sediments were probably largely derived from the BaerBassit ophiolite massif, suggesting that uplift and unroofing of this area had occurred by this time. Some ophiolite-related sediment could also have been sourced from exposures now entirely eroded. The energy of the streams appears to have waned over time, possibly as the relief of the source area decreased. It is likely that the inferred flexural high remained during Early Miocene time, perhaps controlled by the inversion of the Neotethyan-age extensional faults in the basement.

The Mid-Miocene witnessed a marine transgression and shallow-marine bioclastic limestone deposition in a variety of subenvironments. Greatest thicknesses in the SW imply more accommodation space in this area, in turn suggesting that extension increased in this direction. The inferred growth fault and sediment fanning confirm that normal faults were active during Mid-Miocene time. These were oriented NE-SW, implying that r was vertical and c~3 oriented c. NW-SE. By the Late Miocene, a rise in relative sea level led to moderately deep-marine conditions, and sufficient bathymetry to promote basin margin instability. The Mediterranean-wide salinity crisis during the Messinian resulted in the precipitation of evaporites, now exposed in only a few areas near the present basin axis and on its southern flank. Rapid uplift of the graben flanks took place during the Early Pliocene, displacing the evaporite horizon and confining Pliocene sediment deposition to within the present graben. The probable cause was a strong pulse of rifting with corresponding flexural uplift of the basin margins and relative deepening of the basin axis. Although marine sedimentation resumed in the Early Pliocene, relative sea-level fall culminated in continental conditions by the Late Pliocene, during a time when active faulting appears to have continued. During the Quaternary, continuing uplift has accentuated channel incision and formed wind-gaps, as seen on the SE margin. A series of river terraces formed inland and marine terraces developed (Pirazzoli et al. 1991) near the coast, reflecting continuing uplift during a time of global sea-level change. Historical seismicity (Over et al. 2002), and present-day seismicity, confirm that the area remains tectonically active at present, with normal and oblique faulting taking place (Over et al. 2002).

Regional comparisons and implications The Hatay Graben can usefully be compared with the tectonic evolution of the Tauride Mountains to the NE and Cyprus to the west. During the Cretaceous to Neogene, collisional processes affected the Taurus Mountains to the north of the Hatay Graben. The segment of the Bitlis suture directly north of the study area is the Misis-Andlrm lineament (Robertson et al. 2004; Fig. 1). Ophiolites were emplaced southwards onto the Arabian foreland during the partial closure of the southern Neotethys Ocean in latest Cretaceous time. This ocean remained partly open during the Early Tertiary until final closure by Mid-Miocene time along the Bitlis suture, which runs through southern Turkey to

CENOZOIC HATAY GRABEN, S TURKEY

627 30" 60"

t20"

ZoneI

~r

~-q--~,-, 150"

,-,'1~.

Zone 2

=1o;~

180"

3

180~

36~

330"

N=21

N=37

300"

270

Zone3

30~ m~

9

z~o. , ~ ~

"210~

Zone4

.... , ~

120~

180" _

I

330=

N=11

3

24O~ ~ Zone

5

0

30~

330~

~

30"

300~

60~

N=50o ~

120~

12o-

Zone6

,,o . . . . "r,-,- .~r~" ' ,r~o

,,0.~

180 ~

1W ~

/ x% ~" gg ~i

Key FaultPlanes Slickenlines Sigma 1 Sigma2 Sigma 3

Fig. 12. Equal-area stereonets showing the results of the Angelier-method stress inversion where the data are divided by sub-area (see Fig. 9 for location of the zones).

628

S.J. BOULTON E T A L . I 9

~~

31m*

~ 10~

,

30.

6o-

"

70

~

. I"~

Pliocene

N=

"

2'10 ~

" "~

~

1110~

'0~

300 ~

Oe ~

. o . ~

1,~.

"

180"

Early Miocene

~o

.

N 1

cene

Late Miocene 300 ~

60"

210 ~

~ r " ~

r~P ~''

klSO'

tiN) ~

Mid-Miocene

300 ~

80"

~

0

"

~

~

~ ~ 1 5 D ' " 180o

Late Cretaceous

Fig. 13. Lower hemisphere equal-area projection of the results of the Angelier-method stress analysis when the data are divided by the minimum age of the fault (see Fig. 12 for key).

Iran (Dewey et al. 1986; Yllmaz 1993; Robertson 1998a). After ophiolite obduction in latest Cretaceous time, passive margin conditions were restored along the northern margin of the Arabian plate, including the study area. By contrast, the northern margin of the southern Neotethyan ocean was characterized by northward subduction and active margin processes (Ydmaz 1993; Robertson 1998a). After a period of extension on the

northern active margin during Mid-Eocene time, remaining Neotethyan ocean crust was subducted until diachronous continental collision was initiated. As the northerly Tauride active margin impinged on the Arabian passive margin, this was flexurally upwarped during the Oligocene and then collapsed to form a foreland basin during the Early to Mid-Miocene (e.g. Lice Formation in SE Turkey). Collision was complete by the Mid-Miocene, the earliest time when

CENOZOIC HATAY GRABEN, S TURKEY transgression of the suture is observed. During the Mid-Late Miocene suture tightening took place until little further orthogonal shortening could take place (Robertson 2004). During MidLate Miocene time large volumes of clastic material were shed from the uplifting Taurus Mountains to the north, coupled with a much smaller amount of sediment sourced from the erosion of the inferred flexural forebulge to the south (Robertson et al. 2004). During the Early Pliocene there was then a switch to left-lateral strike-slip as the EAFZ became established ($eng6r et al. 1985; Dewey et al. 1986; Hempton 1987; Westaway & Arger 1996). As a result, compressional strain was released across the foreland as a whole. Miocene sediments directly to the north of the Hatay Graben are buried beneath the Iskenderun Basin. In general, exploration oil wells in this basin indicate that Upper Oligocene to Lower Miocene turbidites overlie ophiolitic rocks (Kempler & Garfunkel 1991), which were then covered by Messinian evaporites and PlioQuaternary clastic sediments, prograding from the Misis-Andxrm thrust front to the north (Aksu et al. 1992). On land, the Iskenderun Basin is marked by extensional faulting, especially around the northern margin of the Hatay ophiolite (unpublished data), opening a basin that was partially infilled with shallow-marine clastic sediments. Further north, thrust Miocene sediments structurally underlie the Misis Lineament, the southern front of the Tauride allochthon, and include Lower-Middle Miocene deep-water turbidites and Upper Miocene-Pliocene shallowwater clastic deposits (G6kgen et al. 1988). Cyprus is situated to the SW of the Hatay Graben. The Troodos Ophiolite formed in an above-subduction zone setting in Late Cretaceous time, in common with the chain of ophiolites that stretches from Cyprus, to Hatay, through SE Turkey and Iran to Oman (e.g. Robertson 2004). During the Miocene the present subduction zone, which dips NE under Cyprus (Pilidou et al. 2004), is likely to have been active. Subsequent rollback of this subduction zone resulted in extensional faulting during the Miocene, forming a number of basins in the area (e.g. Polis Graben, western Cyprus; Payne & Robertson 1995). The collision of the Eratosthenes Seamount with the Cyprus trench triggered the uplift of the Troodos Massif and the Kyrenia Range of northern Cyprus (e.g. Robertson 1998b) mainly during Late Pliocene to Pleistocene time. The Kyrenia Range continues offshore to connect with the Misis-Andlrm lineament (Fig. 1) and can be considered as part of the original northern continental margin of the

629

Southern Neotethys Ocean. Further south, the Latakia Ridge runs from southern Cyprus to northern Syria, coming onshore to the north of Latakia (Kempler & Garfunkel 1991; BenAvraham et al. 1995). Between these lineaments is the Latakia Basin, which may extend into the Mesaoria Basin on Cyprus. As such, the Hatay Graben is located at the interface between an area further to the east where continental collision took place by Mid-Miocene time followed by left-lateral tectonic escape, and an area to the west that is in a syncollisional setting with a deepwater basin still present. We consider that this particular setting has contributed to the tectonic development of the Hatay Graben.

Discussion Three models can be considered to explain the tectonic evolution of the Hatay Graben; of these, model 1 is the only one previously considered in the literature. M o d e l 1: s t r i k e - s l i p r e l a t e d m o d e l

Basin formation took place in response to stresses caused by the propagation of the EAFZ, the DSFZ, or both. Some workers see the Hatay Graben as a direct extension of the EAFZ ($eng6r et al. 1985; Lyberis et al. 1992; Saro~lu et al. 1992), whereas others (Arpat & Saro~lu 1972; Parlak et al. 1998) trace the DSFZ further north through the Amanos Mountains to the Turkish coast. Evidence from syndepositional faulting within the Hatay Graben indicates that normal faulting parallel to the present graben was taking place during the Mid-Miocene, around 13-15 Ma. However, the inception of the EAFZ is dated as Late Miocene to Pliocene ($eng6r et al. 1985; Dewey et al. 1986; Hempton 1987; Saro~lu et al. 1992; Westaway & Arger 1996; Yfiriir & Chorowicz 1998). In addition, the DSFZ formed synchronously with the opening of the Red Sea around 18-20 Ma (Garfunkel 1981; Hempton 1987; Lyberis 1988; Rojay et al. 2001). Studies of the northerly DSFZ indicate that the fault did not propagate northwards into northern Syria and southern Turkey for another 15 Ma, until a second phase of fault motion around 4.5 Ma (Freund et al. 1968; Girdler & Styles 1978; Zanchi et al. 2002). Therefore, the initiation of faulting within the Hatay Graben cannot be linked directly with the inception of either the EAFZ or the DSFZ, as the initiation of normal faulting predates the presence of the EAFZ and DSFZ in southern Turkey by c. 10 Ma. However, the inferred Plio-Quaternary

630

S.J. BOULTON E T A L . Reactivated

Flexural ,

faults .... b=oi.

forebulge

I

normal

in the

foreland ..,......

,-'.~_ , 7 :

, . . _ _ j r , .-.',,,

/X

+

+

+\+

+

+

O P b ~ a i t ~ ' ~ ' 7 ] + ~ ' ~ .

++

+

~=l'-'~~\.

+ + + +

+ Arabian +

4

+

+

+'++

P l a. . t. .f. o. . .r. m -t-

+

+

+

~

+

.

-

++

'

'

'-

,

,-

,--

+ - ' < 'I &/k . . ' '~- ' . ./ _ I- ,

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

§

+

4-

t-

I

+ P ' ~ ' ,~ . ~ . . - ' 4-

~---'

-

Th.,,=~,~=,i.="

-I-

+~"%="

I

\

- _, " "

,-

"

. .~," - /'

//

..

' -

, -.. ,

,

'

-1 /

. ""

-"

.."

I

/

+ "t', +

/

Fig. 14. Schematic diagram illustrating the model of basin formation with normal faulting occurring in the foreland basin. Box indicates location of the Hatay Graben.

transtension and related strike-slip faulting could be related to the EAFZ or DSFZ (see below).

like the Hatay Graben, which is less likely to be affected by extension-related rollback.

Model 2." extension related to subduction roll-back

Model 3: foreland basin followed by 'tectonic escape' model (Figs 14 and 15)

During the Early Miocene the Southern Neotethys to the north of the study area was in its final stages of subduction and incipient collision with the Arabian plate. The initiation of the graben, situated on the subducting Arabian slab, might relate to stresses associated with subduction, e.g. trench rollback caused by the negative buoyancy of subducting oceanic lithospheric crust. There is evidence that the overriding Tauride plate was strongly extended during the Early-MidMiocene, with the formation of thrust-top (piggyback) basins within the Misis-Andlrm lineament to the north (Robertson et al. 2004). This extension could be related to rollback of the downgoing slab during the final stages of subduction when old, dense oceanic crust was being subducted. This model was proposed for several basins in the eastern Mediterranean (i.e. Polis Graben, Cyprus, Payne & Robertson 1995; Mesaoria Basin, Cyprus, McCallum & Robertson 1990; Robertson et al. 1991). However, these basins are situated on the overriding place of the subduction zone, not the foreland

In this model the basin was initially part of the fill of a foreland basin developed ahead of the advancing Tauride allochthon. Peripheral foreland basins develop in response to loading of the lithosphere of the downgoing plate during continent-continent collision (Stockmal et al. 1986; Sinclair 1997). We believe that there was no distinct topographic graben present until Late Miocene-Pliocene time and that before this time the Hatay Graben area formed part of the foreland basin to the Bitlis Suture Zone. Models of continental collision suggest that the position of a flexural forebulge can be c. 400 km from the thrust front, making this feasible. Normal faulting, as seen in the area studied, has been documented in many foreland basins; for example, the Taconic foreland basin (Utica Shale) of New York (Bradley & Kidd 1991), the Timor Trough (Veevers et al. 1978), the French Alps (Sinclair 1997) and the central Mediterranean, Pelagian Shelf (Argnani & Torelli 2001). Such normal faulting can be related to two factors. First, the underlying passive margin is likely

CENOZOIC HATAY GRABEN, S TURKEY

631

Fig. 15. Tectonic sketches illustrating the evolution of the Eastern Mediterranean from the Early-Mid-Miocene to the present day. The Early-Mid-Miocene is a time of subduction near the coast of Anatolia, subsequently migrating to the south. Thrusting occurs along the Arabian margin, forming a foreland basin. Faulting is initiated in the Hatay Graben. By the Late Miocene, 'rollback' of the subduction zone has given rise to extensional basins on the overriding plate and thrust emplacement is complete in southern Turkey. The DSFZ is propagating northwards. During Pliocene-Recent time, collision results in westward escape of Anatolia, initiating the EAFZ, and the DSFZ has propagated northwards into Southern Turkey. The Eratosthenes Seamount is being thrust under Cyprus, and the Troodos Massif is uplifting in response. The vector diagram illustrates the orientation and motion of the DSFZ and EAFZ, with the orientation of the Hatay Graben shown between them. to have normal faults present that formed during continental break-up, forming deep-seated weaknesses in the lithosphere that could be reactivated later. Second, active faulting in the foreland may be due to extensional stresses generated on the outer arc of the fixed lithosphere (Bradley & Kidd 1991). However, it is more likely to have affected more proximal areas of the foreland basin than is represented by the H a t a y Graben area. We, therefore, we suggest that the initiation of the H a t a y Graben was due to extensional stresses caused by the flexure of the foreland that reactivated deep-seated normal faults in the Arabian passive margin. A thick upper turbiditic 'flysch' sequence is not present, as in a fully developed foreland basin (e.g. Sinclair 1997), probably because by the Mid-Miocene southward convergence of the Eurasian and African (Arabian) plates in SE Turkey was replaced by left-lateral strike-slip and tectonic escape of Anatolia towards the Aegean.

As a result, loading of the foreland diminished and thus allowed isostatic rebound and the observed regional regression to take place. We suggest the Plio-Quaternary deformation was transtensional because the H a t a y Graben is located at the interface between a zone of continent-continent collision in the east and a syncollisional zone in the west into which tectonic escape could take place.

Conclusions The first evidence of a discrete basin in the H a t a y area is from the Mid-Miocene time. Initial normal faulting was coincident with the basin becoming shallow marine in character. The basin progressively deepened during the M i d - L a t e Miocene, when deeper marine carbonate deposition dominated. During the Early Pliocene strong flank uplift and relative sea-level fall resulted in marginal-marine conditions followed by

632

S.J. BOULTON ET AL.

non-marine conditions from Late Pliocene time onwards. There are three main trends of faulting (NW-SE, north-south, NE-SW). The fault patterns are dissimilar to those known from pure strike-slip or extensional settings but are similar to those recorded in oblique extensional (transtensional) settings (e.g. Gulf of California, Gulf of Aden). Fault stress analysis indicates that ~ = 30o-45 ~ within a left-lateral (sinistral) shear zone (where ~ is the acute angle between the rift trend and the direction of displacement on the plate edge). The orientation of the graben was probably influenced by pre-existing zones of crustal weakness, probably including those of the Early Mesozoic rifting of Neotethys. Sediments of Late Miocene and younger age reflect the location of the developing basin near the distal edge of a foreland basin. During the Plio-Quaternary, when the present topographic graben developed, the main regional tectonic controls were the east-west left-lateral tectonic escape of Anatolia westwards along the East Anatolia Fault Zone and the northwards propagation of the DSFZ. SJB acknowledges receipt of an NERC PhD Studentship (NER/S/A/2002110361) and additional funding by the American Association of Petroleum Geologists. AHFR thanks the Carnegie Trust for financial assistance with fieldwork. We thank M. C. Al~igek, N. Jaffey and M. Purvis, whose reviews improved this paper.

References AKSU, A. E, ULUG, A. PIPER, D. J. W., KONUK, Y. T. & TURGUT, S. 1992. Quaternary sedimentary history of Adana, Cilicia and |skenderun Basins: northeast Mediterranean Sea. Marine Geology, 104, 55-71. ANGELIER, J. 1984. Tectonic analysis of fault slip data sets. Journal of Geophysical Research, 89, 5835-5848. ARGANI, A. & TORELLI, L. 2001. The Pelagian Shelf and its graben system (Italy/Tunisia). In: ZIEGLER, P. A., CAVAZZA, W., ROBERTSON, A. H. F. & CRASQUI-SOLEAU. S. (eds) Peri-Tethys Memoir 6." Peri-Tethys Rift~Wrench Basins and Passive Margins. MGmoires du MusGum National d'Histoire Naturelle, 186, 529-544. ARPAT, E. & SARO(3LU, F. 1972. The East Anatolian Fault System; thoughts on its development. Bulletin of the Mineral Research and Exploration Institute of Turkey, 78, 33-39. BARKA, A. A., AKYOZ, S. H., COHEN, H. A. & WATCHORN, F. 2000. Tectonic evolution of the Niksar and Tasova-Erba pull-apart basins, North Anatolian fault zone: their significance for the motion of the Anatolian block. Tectonophysics, 322, 243-264. BEN-AVRAHAM, Z., TIBOR, G., LIMINOV, A. F., LEYBOV, M. B., IVANOV,M. K., TOKAREV,M. Y. & WOODSIDE, J. M. 1995. Structure and tectonics of

the Eastern Cyprean Arc. Marine Petrology and Geology, 12, 263-271. BEYDOUN, Z. 1999. Evolution and development of the Levant (Dead Sea Rift) Transform system: a historical-chronological view of structural controversy. In: MACNIOCAILL, C. & RYAN, P. D. (eds) Continental Tectonics. Geological Society, London, Special Publications, 164, 239-255 BOZKURT, E. 2001. Neotectonics of Turkey--a synthesis. Geodinamica Acta, 14, 3-30. BRADLEY, D. C. & KIDD, W. S. F. 1991. Flexural extension of the upper continental crust in collisional foredeeps. Geological Society of America Bulletin, 103, 1416-1438. BREW, G., LUPA, J., BRAZANGI, M., SAWAF, T., ALIMAM, A. & ZAZA, T. 2001. Structure and tectonic development of the Ghab basin and the Dead Sea fault system, Syria. Journal of the Geological Society, London, 158, 665-674. CHAIMOV, T., BARAZANGI,O., AL-SAAD, O., SAWAF, T. & GEBREN, A. 1990. Crustal shortening in the Palmyride Fold Belt, Syria, and implications for movement along the Dead Sea Fault System. Tectonics, 9(6), 1369-1386. CLIFTON, A. E., SCHLISCHE,R. W., WITHJACK,M. O. & ACKERMANN,R. V. 2000. Influence of rift obliquity on fault-population systematics: results of experimental clay models. Journal of Structural Geology, 22, 1491-1509. DELALOYE, M., PISKIN, 6 , SEL(~UK,H., VUAGNAT,M. & WAGNER, J.-J. 1980. Geological section through the Hatay ophiolite along the Mediterranean coast, southern Turkey. Ofiliti, 5(213), 205-216. DERMAN, A. S., AKDA~, K., GOL, M. A. & YENIAY,G. 1996. Relationship between sedimentation and tectonics in the Mara~ Miocene basin. Turkish Association of Petroleum Geologists, llth Petroleum Congress, Turkey, 91-102. DEWEY, J. F., HEMPTON, M. R., KIDD, W. S. F., SAROC3LU,F. & SENGOR, A. M. C. 1986. Shortening of continental lithosphere: the neotectonics of eastern Anatolia--a young collision zone. In: COWARD, M. O. & RElS, A. C. (eds) Collisional Tectonics. Geological Society, Londons, Special Publications, 19, 3-36. FREUND, R., ZAK, I. & GARFUNKEL, Z. 1968. On the age and rate of sinistral movement along the Dead Sea rift. Nature, 220, 253-255. GARFUNKEL, Z. 1981. Internal structure of the Dead Sea leaky transform (rift) in relation to plate kinematics. Tectonophysics, 80, 81-108. GIRDLER, R. W. & STYLES,P. 1978. Seafloor spreading in the western Gulf of Aden. Nature, 271,615-617. GOKCEN, S. L., KEELING,G., GOKCEN,N. & FLOYD, P. A. 1988. Sedimentology of a Late Cenozoic collisional sequence: the Misis Complex, Adana, southern Turkey. Sedimentary Geology, 59, 205-235. HAQ, B. U., HARDENBOL,J. & VAIL, P. R. 1987. Chronology of fluctuating sea-levels since the Triassic. Science, 4793, 1156-1167. HEMPTON, M. R. 1987. Constraints on Arabian Plate motion and extensional history of the Red Sea. Tectonics, 6, 687-705. HsO, K. J., MONDADERT, L., BERNOULLI, D., et al. 1978. History of the Mediterranean salinity crisis. In: HsO, K., MONTADERT, L. et al. (eds) Initial Reports of the Deep Sea Drilling Project, 42. US

CENOZOIC HATAY GRABEN, S TURKEY Government Printing Office, Washington, DC, 1053-1078. JAFFEY, N. & ROBERTSON, A. H. F. 2001. New sedimentological and structural data from the Ecemi~ Fault Zone, southern Turkey: implications for its timing and offset and the Cenozoic tectonic escape of Anatolia. Journal of the Geological Society, London, 158, 367-378. KEMPLER, D. & GARFUNKEL, Z. 1991. The northeast Mediterranean triple junction from a plate kinematic point of view. Bulletin, Technical University oflstanbul, 44(3-4), 425-454. KOfYI~IT, A. & BEYHAN, A. 1998. A new intracontinental transcurrent structure: the Central Anatolian Fault Zone, Turkey. Tectonophysics, 284, 317-336. KRIJGSMAN, W., WILSON, D. S., HILGEN, F. J., RAFFI, I. & SIERRO, F. J. 1999. Chronology, causes and progression of the Messinian salinity crisis. Nature, 400(6745), 652-655. LOVELOCK, P. E. R. 1984. A review of the tectonics of the northern Middle East region. Geological Magazine, 121, 577-587. LYBERIS, N. 1988. Tectonic evolution of the Gulf of Suez and the Gulf of Aqaba. Tectonophysies, 153 (1-4), 209-220. LYBERIS, N., YfJROR, T., CHOROWITCZ, J., KASAPO~LU, E. & GONDO~U, N. 1992. The East Anatolian fault: an oblique collisional belt. Tectonophysics, 204, 1-15. MART, Y. & RABINOWITZ, P. D. 1986. The northern Red Sea and Dead Sea rift. Tectonophysics, 124, 85-113. MCCALLUM, J. E. & ROBERTSON,A. H. F. 1990. Pulsed uplift of the Troodos Massif--evidence from the Plio-Pleistocene Mesaoria Basin. In: MALPAS, J., MOORES, E. M., PANAYIOTOU,A. & XENPHONTOS, C. (eds) Ophiolites, Ocean Crustal Analogues.

Proceedings of the Symposium 'Trodoos 1987". Geological Survey Department, Nicosia, 217-229. MCCLAY, K. R. & WHITE, M. J. 1995. Analogue modelling of orthogonal and oblique rifting. Marine and Petroleum Geology, 12, 147-151. MCKENZIE, D. P. 1978. Active tectonism in the Alpine-Himalayan belt: the Aegean Sea and the surrounding regions (tectonics of the Aegean region). Geophysical Journal of the Royal Astronomical Society, 55, 217-254. MESCHEDE, M., WEISS, R., SCHMIEDL, G. & HEMLEBEN, C. 2002. Benthic foraminifera distribution and sedimentary structures suggest tectonic erosion at the Costa Rica convergent plate margin. Terra Nova, 14(5), 388-396. ORAL, M. B., REILINGER, R. E., TOKSOZ, M. N., KONG, R. W., BARKA, A. A., K1N1K,I. & KENK, O. 1995. Global positioning system offers evidence of plate motions in eastern Mediterranean. EOS Transactions, American Geophysical Union, 76 (9), 9-11. OVER, S., KAVAK, K. $., BELLIER, O. & OZDEN, S. 2004. Is the Amik Basin SE Turkey a triple junction area? Analyses of SPOT XS imagery and seismicity. International Journal of Remote Sensing, 25(19), 3857-3872. OVER, S., (JNLOGEN(~,U. C. & BELLIER, 0. 2002. Quaternary stress regime change in the Hatay region

633

(SE Turkey). Geophysics Journal International, 148, 649-662. PARLAK, O., KOP, A., (~INLf3GEN~,U. C. & DEMIRKOL, C. 1998. Geochronology and geochemistry of basaltic rocks in the Karasu graben around K m k h a n (Hatay), S. Turkey. Turkish Journal of Earth Sciences, 7, 53-61. PAYNE, A. S. & ROBERTSON, A. H. F. 1995. Neogene supra-subduction zone extension in the Polis graben system, west Cyprus. Journal of the Geological Society, London, 152(4), 613-628. PERIN~EK, D. & ~EMEN, i. 1990. The structural relationship between the East Anatolian and Dead Sea fault zones in southeastern Turkey. Tectonophysics, 172, 331-340. PILIDOU, S., PRIESTLEY, K., JACKSON,J. & MAGGI, A. 2004. The 1996 Cyprus earthquake: a large, deep event in the Cyprean Arc. Geophysical Journal International, 158(1), 85-97. PIRAZZOLI, P. A., LABOREL, J., SALIEGE, J. F., EROL, O., KAYAN, I. & PERSON, A. 1991. Holocene raised shorelines on the Hatay coasts (Turkey): palaeoecological and tectonic implications. Marine Geology, 96, 295-311. PIPKIN, O., DELALOYE,M., SELCUK,H. & WAGNER, J. 1986. Guide to Hatay Geology (SE Turkey). Ofioliti, 11, 8%104. REILINGER, R. E., MCCLUSKY, S. C., ORAL, M. B., KING, W. & TOKSOZ, M. N. 1997. Global positioning system measurement of present-day crustal movements in the Arabian-African-Eurasia plate collision zone. Journal of Geophysics Research, 102, 9983-9999. RIGO DE RIGHI, M. & CORTESINI,A. 1964. Gravity tectonics in foothills structure belt of SE Turkey.

American Association of Petroleum Geologists Bulletin, 48, 1911-1937. ROBERTSON, A. H. F. 1986. The Hatay Ophiolite (southern Turkey) in its eastern Mediterranean tectonic context; a report on some aspects of the field excursion. Ofioliti, 11(2), 105-119. ROBERTSON, A. H. F. 1998a. 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

Occean Drilling Program, Scientific Results, 160. Ocean Drilling Program, Collage Station, TX, 723-782. ROBERTSON, A. H. F. 1998b. Tectonic significance of the Eratosthenes Seamount: a continental fragment in the process of collision with a subduction zone in the eastern Mediterranean (Ocean Drilling Program Leg 160). Tectonophysics, 298, 63-82. ROBERTSON, A. H. F. 2002. Overview of the genesis and emplacement of Mesozoic ophiolites in the Eastern Mediterranean Tethyan region. Lithos, 65 (1-2), 1-67. ROBERTSON, A. H. F. 2004. Development of concepts concerning the genesis and emplacement of Tethyan ophiolites in the Eastern Mediterranean and Oman regions. Earth-Science Reviews, 66(3-4), 331-387. ROBERTSON, A. H. F., EATON, S., FOLLOWS, E. J. & PAYNE, A. S. 1995. Depositional processes and

634

S.J. BOULTON ET AL.

basin analysis of Messinian evaporites in Cyprus. Terra Nova, 7, 233-253. ROBERTSON, A. H. F., EATON, S., FOLLOWS, E. J. & MCCALLUM, J. E. 1991. The role of local tectonics versus global sea-level change in the Neogene evolution of the Cyprus active margin. In: MACDONALD, D. I. M. (ed.) Sedimentation, Tectonics and Eustasy: Sea-level Changes at Active Margins. International Association of Sedimentologists, Special Publications, 12, 331-369. ROBERTSON, A. H. F., UNLOGEN~, U. C., INAN, N. ~r TASL], K. 2004. The Misis-Andmn Complex: a Mid-Tertiary m61ange related to late-stage subduction of the Southern Neotethys in S Turkey. Journal of Asian Earth Sciences, 22, 413-453. ROJAY, B., HEIMANN, A. & TOPRAK, V. 2001. Neotectonic and volcanic characteristics of the Karasu fault zone (Anatolia, Turkey): the transition zone between the Dead Sea transform and the East Anatolian fault zone. Geodinamica Acta, 14, 197-212. ~AFAK, ()-. 1993a. Antakya Havzasl planktonic foraminifer biyostratigrafisi. (Planktonic Foraminifera biostratigraphy of Antakya Basin.) Ankara Suat Erk jeoloji simpozyumu, Ankara University, Ankara, 143-156. SAFAK, U. 1993b. Antakya Havzasl ostrakod foraminifer biyostratigrafisi. (The ostracode biostratigraphy of the Antakya Basin). Geological Bulletin of Turkey, 36, 115-137 (in Turkish, with English abstract). SALLER, A., ARMIN, R., ICHRAM, L. O. & SULLIVEN,C. 1993. Sequence stratigraphy of aggrading and backstepping carbonate shelves, Oligocene, Central Kalmantan, Indonesia. In: Carbonate Sequence Stratigraphy: Recent Developments and Applications. In: LOUCKS, R. G. & SARG, J. F. (eds) Carbonate Sequence Stratigraphy--Recenr Events and Applications. AAPG Memoirs, 57, 267-290. SALVINI, F. 2001. Daisy 2.4." the structural data integrated system analyser. University 'Roma Tre', Rome. ~AROGLU, F., EMRE, 6. & KUSCU, I. 1992. The East Anatolian fault zone of Turkey. Annales Tectonicae, 6, 99-125. SCHREIBER, B. C., FRIEDMAN, G. M., DEC1MA, A. & SCREIBER, E. 1976. Depositional environments of upper Miocene (Messinian) evaporite deposition of the Sicilian basin. Sedimentology, 23, 729-760. ~ENG()R, A. M. C. ~r KIDD, W. S. F. 1979. Postcollisional tectonics of the Turkish-Iranian plateau and a comparison with Tibet. Tectonophysics, 55, 361-376. SENGOR, A. M. C., GOROR, N. & SAROGLU, F. 1985. Strike-slip faulting and related basin formation in zones of tectonic escape: Turkey as a case study. In: BIDOLE, K. T. & CHRISTIE, BLICK, N. (eds) Strike-slip Deformation, Basin Formation and Sedimentation. Society of Economic Paleontologists and Mineralogists, Special Publications, 37, 227-264. SENGOR, A. M. C., TOYsCrz, O., IMREN, C., et aL 2005. The North Anatolian Fault: a new look. Annual Review of Earth and Planetary Sciences, 33, 37-112.

SINCLAIR, H. D. 1997. Tectonstratigraphic model for underfilled peripheral foreland basins: an Alpine perspective. Geological Society of America Bulletin, 109, 324--346. STOCKMAL, G. S., BEAUMONT, C. 8r BOUTILIER, R. 1986. Geodynamic models of convergent margin tectonics: transition from rifted margin to overthrust belt and consequences for foreland-basin development. American Association of Petroleum Geologists Bulletin, 70(2), 181-190. TEN VEEN, J. H. ~r KLEINSPEHN, K. L. 2002. Geodynamics along an increasingly curved convergent plate margin: Late Miocene-Pleistocene Rhodes, Greece. Tectonics, 21(3), 1017-1038. TINKLER, C., WAGNER,J. J., DELALOYE,M. & SEL~UK, H. 198 I. Tectonic history of the Hatay Ophiolites (South Turkey) and their relationship with the Dead Sea rift. Tectonophysics, 72, 23-41. TRON, V. ~r BRUN, J.-P. 1991. Experiments on oblique rifting in brittle~luctile systems. Tectonophysics, 188, 71-84. UMHOEFER, P. J. • STONE, K. A. 1996. Description and kinematics of the SE Loreto basin fault array, Baja California Sur, Mexico: a positive field test of oblique-rift models. Journal of Structural Geology, 18, 595-614. VEEVERS, J. J., FALVEY, D. A. & ROBINS, S. 1978. Timor Trough and Australia: facies show topographic wave migrated 80 km during the past 3 myr. Tectonophysics, 45, 217-227. WESTAWAV, R. 1994. Present day kinematics of the Middle East and Eastern Mediterranean. Journal of Geophysical Research, 99, 12071-12090. WESTAWAY, R. & ARGER, J. 1996. The G61ba~l basin, southeastern Turkey: a complex discontinuity in a major strike-slip zone. Journal of the Geological Society, London, 153, 729-743. WITHJACK, M. O. d~r JAMIESON, W. R. 1986. Deformation produced by oblique rifting. Tectonophysics, 126, 99-124. YETIS, C. 1978. Geology of the Amardi (Ni~de) region and the characteristics of the Ecemi~ Fault Zone between Madan Bo~azl and Kami~li. Istanbul University Fen Fakultesi Mecanik Series B, 43, 41-61. Y~LMAZ, Y. 1993. New evidence and model on the evolution of the southeast Anatolian orogen. Geological Society of America Bulletin, 105(2), 251-271. YURTMEN, S., HERVI~, G., WESTAWAY, R., ROWBOTHAM, G. & TATAR, O. 2002. Rate of strikeslip motion on the Amanos Fault (Karasu Valley, southern Turkey) constrained by K-Ar dating and geochemical analysis of Quaternary basalts. Tectonophysics, 344, 207-246. YOROR, T. & CHOROW]CZ, J. 1998. Recent volcanism, tectonics and plate kinematics near the junction of the African, Arabian and Anatolian plates in the Eastern Mediterranean. Journal of Volcanology and Geothermal Research, 85, 1-15. ZANCHI, A. CROSTA, G. B. & DARKAL, A. N. 2002. Palaeostress analysis in NW Syria: constraints on the Cenozoic evolution of the northwestern margin of the Arabian plate. Teetonophysics, 357, 255-278.

Evidence for late Holocene activity along the seismogenic fault of the 1999 Izmit earthquake, NW Turkey S. B. P A V L I D E S 1, A. C H A T Z I P E T R O S l, Z. S. T U T K U N 2, V. I D Z A K S O Y 3 & B. D O ( ] A N 4

1Department of Geology, Aristotle University, 54124, Thessaloniki, Greece (e-mail." pavlides@geo, auth. gr) 2Department of Geological Engineering, (2anakkale Onsekiz Mart University, 17020, Canakkale, Turkey 3General Directorate of Mineral Research and Exploration (MTA ), Ankara, Turkey 4Department of Geological Engineering, Istanbul Technical University, Maslak 34469, Istanbul, Turkey During the strong 1999 Izmit (Kocaeli) earthquake, 100 km long east-weststriking (N80~176 right-lateral, fault traces were formed. In the epicentral area the seismic ruptures did not follow any known or mapped fault traces, but morphology and tectonostratigraphic evidence from trenches reveal pre-existing earthquake-related features, e.g. elongated valleys, shutter ridges, high-angle slopes, scarplets and stream offsets. In the G61cuk Peninsula a characteristic NW-SE-trending extensional fault segment emerged at the surface with a 1.5-2 m maximum vertical displacement and a 0.30 m right-lateral component. The resulting coseismic fault scarp was mapped in detail, and two trenches were excavated at the DEniz Evler site. The 1999 displacement at this site was 1.50 m, whereas the penultimate event displaced the same sediments by 0.70 m, and a previous event by 0.20 m. Displacement is not characteristic, as fault-associated soft (recent) deltaic deposits, and the fault itself, are typically not coseismic, but rather a secondary accommodation structure in geometric consistency with the right-lateral main displacement zone. The data were compared with similar results from the A~ggt Yuvacik, Kular Yaylacik and Acisu sites between Izmit and Sapanca Lake. The same fault segment seems to have been activated and produced surface ruptures including during the earthquakes of AD 1509, AD 989 and AD 554, plus two prehistoric events. The palaeoseismological results provide clear evidence for repeated reactivation of the same fault or fault segments during historical seismic events.

Abstract:

On 17 August 1999 an Mw=7.4 earthquake occurred along the western part of the North Anatolian Fault Zone (NAFZ) (Fig. 1). Its epicentre was located in the Gulf of Izmit. This was a multiple event, which ruptured a sequence of segments with typical right-lateral displacements (0.2-5 m). During this strong earthquake, a series of east-west-striking (N80~176 rightlateral, fault traces appeared, their length being 100 km in total. At a small scale they show typical strike-slip displacements (2-5 m), pop-ups and pressure ridges (N 40-70~ P (N80~ R (N100110 ~ and R' (c. N350 ~ Riedel shears, extensional cracks (Nl15~176 restraining and releasing bends, and small pull-apart structures (Tutkun & Pavlides 2001). Although the earthquake ruptures are generally linear (Fig. 2), and east-west-striking (N80~176 they can be divided into branches (or segments) on the basis of geometry. The Karadere 'segment' (or branch) is an E N E WSW-striking (N75-85~ 20-25 km long fault

from Eften Lake to Akyazi town. This fault overlapped the 12 November, 1999 Duzce earthquake ruptures of the Aksu segment (e.g. Tutkun & Pavlides 2001; Emre et al. 2003a); the maximum observed displacement is about 2 m, and the average is 1 m. The Arifiye segment is a 25 km long linear and continuous seismic trace from Arifiye village to the east shore of Sapanca Lake. Between Arifiye and the town of Akyazi, there is a 5 km gap in the surface seismic traces. The maximum right-lateral displacement (4.8 m) was observed close to Arifiye village. The displacement gradually decreases toward the east (0.3 m), and the average displacement on this segment is about 2-2.5 m. The Karadere and Arifiye fault branches are close to the western edge of the 1967 Mudurnu seismic fault traces (Ambraseys & Zatopec 1969; Barka & Kadinsky-Cade 1988). This segment is separated from the Tepetarla one by the Sapanca Lake stepover and by a NW-SE-trending normal fault (G61cfik Peninsula-Kavakli area).

From:ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the EasternMediterranean Region. Geological Society, London, Special Publications, 260, 635-647. 0305-8719106l$15.00 9 The Geological Society of London 2006.

636

S.B. PAVLIDES E T AL.

Fig. 1. Regional structural map, after Emre et al. (2003a). The North Anatolian Fault Zone is divided into two main strands, a northern strand, activated during the 1999 Izmit earthquake, and a southern strand. Inset: the study area in relation to the broader geodynamic setting of the area. The Tepetarla segment is east-west-striking (90~ 35 km long, and extends from the northwestern shore of Lake Sapanca to the G61ctik Peninsula; for about 4 km the fault runs beneath the Izmit Gulf near its southern shoreline. The maximum displacement of this fault segment is 4 m, and the average is 2.5+ 0.5 m. The western edge of the segment in the Kavakli-DenizevlerYenik6y area is a typical normal fault 4 km long, striking N W - S E (120-130~ with a maximum vertical offset of 2.10m and a right-lateral component of 0.2-0.3 m. The westernmost Hersek and G61c/ik segments are two probable submarine east-west structures that extend from G61ctik to Yalova; they are about 50 km long, based on the distribution of aftershocks. The seismicity pattern defines two major segments that extend about 80 km east and 75 km west of the main shock epicentre (40.8~ 29.9~ G61ctik-Izmit, Fig. 1) (Barka 1999, 2000; Awata et al. 2000; Lettis et al. 2002; Emre et al. 2003a).

Palaeoseismology of the westernmost edge of the North Anatolian Fault The broader Aegean region, including western Anatolia, is mainly the result of Late Mesozoic and Early Tertiary deformation, with collisional shortening and an inherited fabric of folds, reverse faults and thrusts, nappes, sutures and strike-slip structures. It is also the result of postorogenic and lithospheric extension during the last 25 Ma, and especially during the youngest,

Mid-Late Quaternary extensional deformation, which is locally associated with important strikeslip motion. From the neotectonic point of view, Anatolia is a post-Alpine collisional intracontinental convergence zone affected by tectonic escape, and where typical active normal faults and important strike-slip faults dominate the region, e.g. the right-lateral North Anatolian Fault Zone (e.g. Bozkurt 2001). Indeed, it was only during the Quaternary that the North Anatolian Fault propagated westwards to merge with the North Aegean Trough ($eng6r 1979; Saro[glu 1988; Jackson 1994). The North Anatolia Fault Zone is the northern boundary of the Anatolian 'microplate' or 'megablock', which accommodates the westward motion of Anatolia relative to Eurasia ($eng6r et al. 1985; Taymaz et al. 1991; Gautier et al. 1999). This fault is an intracontinental strike-slip structure, with a slip rate of about 20-25 mm a-', a well-developed topographic expression and remarkable seismic activity. As a result of continuing tectonic activity, the broader Aegean region has been affected by many active faults (e.g. Jackson 1994; Bozkurt 2001). Many segments were reactivated during recent and historical earthquakes, associated with surface ruptures and other ground deformation (Pavlides & Caputo 2004). In geological terms, active faults are theoretically in constant motion, with episodic events or creep, for relatively long periods of time. The western segments of the North Anatolian Fault have been investigated for a better understanding of their seismogenic behaviour (e.g. Hancock & Barka

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY 1987; Barka & Kadinsky-Cade 1988; Bozkurt 2001). By analysing the geometry, kinematics and seismotectonic characteristics, several common features of these faults can be emphasized: they are commonly active since the early Pleistocene, are characterized by strong earthquakes (M = 7), have maximum vertical displacements of some tens to a few hundreds of centimetres and exhibit return periods of hundreds to thousands of years. Strike-slip faults associated with the North Anatolia-North Aegean Trough system are associated with very strong, devastating earthquakes, with horizontal displacements up to 5 m (Barka & Kadinsky-Cade 1988). Palaeoseismological studies based on investigation of fault colluvium in trenches extends the chronology of historical seismological information and can thus provide crucial data regarding the occurrence of destructive prehistoric earthquakes (e.g. Sieh 1978; McCalpin 1996; Pavlides 1996; Yeats et al. 1997; Pavlides et al. 1999). Knowledge of the spatial and temporal complexity of earthquake recurrence is essential for reliable seismic hazard evaluation and the development of important new concepts concerning earthquake generation (Yeats et al. 1997). In an early attempt to study the palaeoseimicity of the Iznik-Mekece and IzmitSapanca fault segments, Ikeda et al. (1991) determined the eastern extent of the IznikMekece fault. Two seismic events that have taken place between 200 and 500 years BP and the 1967 Mudurnu Valley earthquake ruptures have been

637

identified. Demirtas (1996) excavated and interpreted four trenches across the surface ruptures of the 1957 and 1967 earthquakes in the Mudurnu Valley segment, and suggested that at least three or four previous large earthquake events have produced surface faulting on the same fault prior to the last two earthquakes of 1957 and 1967. Radiocarbon dates show that all the faulting events in the Mudurnu Valley, prior to the last two earthquakes, occurred at some time between 4335 BC and 2500 years BP. The excavation demonstrated an average recurrence interval of this segment of 150 years. Palaeoseismological research by Rockwell et al. (2001) along the Ganos segment, which was last activated during 1912, has shown that recent sediments affected by the fault activity range in age from less than a few hundred years to about 6000 years. They were also affected by repeated coseismic fault behaviour, at least four times during the historical period and even more during the last 4000 years. The August 1999 Izmit-Kocaeli seismogenic fault comprises two major segments, east and west of Sapanca Lake; these in turn can be divided into six smaller segments. The western segment, according to Sugai et al. (2000) and Emre et al. (2003b), produced surface ruptures at least three times during the last millennium. There is palaeoseismological evidence for at least three or possibly four surface-rupturing events since the 17th century, according to Toda et al. (2003). The historical record suggests that the Izmit-Sapanza area experienced strong

Fig. 2. Map of the 1999 Izmit earthquake ruptures (modified from Emre et al. 2003). e, epicentre of the main shock. Ruptures have been divided into the following segments: HS, Hersek segment; GS, G61ciik segment; TS, Tepetarla segment; ARS, Arifiye segment; KS, Karadere segment; AKS, Aksu segment. Trenching sites discussed in this paper are indicated by arrows: DE, Dfiniz Evler; AY, Asa~l Yuvaclk; MA, Mahmutpa~agiftli~i; KY, Kullar-Yaylacik; AC, Acisu.

638

S.B. PAVLIDES E T A L .

e a r t h q u a k e s during 1509, 1719, 1754 and 1894 (Ambraseys & Finkel 1995; A m b r a s e y s 2002). Three seismic events were also defined during the last 1000 years in the eastern segment and the D u z c e fault. E m r e et al.'s (2003b) d a t a indicate

that the eastern Izmit segment exhibits a synchron o u s recurrence interval with the Duzce fault, unlike the western part o f the Izmit segment. D a t i n g results show that the penultimate event o c c u r r e d some time between 200 a n d 300 years

Fig. 3. (a) Log of the SE wall of DE-1 trench at DEniz Evler site. The detailed stratigraphy of Holocene sediments is shown in Figure 4 and described in the text. e, sampling points for OSL dating (results in italics); A, sampling points for ~4C dating. Results are shown in Tables 3 and 4. (b) Photograph showing part of the southeastern wall of DE- 1 trench and details of the main fault zone (faults and layer boundaries are shown by a series of colour-coded flags; grid is 1 m • 1 m). (c) Log of the NW wall of DE-1 trench. Although there are some differences in the logs of the two walls, the main stratigraphic and structural features are consistent.

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY BP, and possibly correlates with the well-known 1719 earthquake. Palaeoseismological results from both segments indicate that the 1719 event had a similar triggering process to the 1999 events (19 August, Izmit; 14 November, Duzce). Klinger et al. (2003) studied the NW-SEtrending G61ciik normal fault, part of the western segment, known as D~niz Evler, which also produced surface ruptures during the 1999 Izmit earthquake. The same fault segment was studied in the present work. The results of Klinger et al. were based on anisotropy of magnetic susceptibility (AMS) dating of colluvial wedges in two trenches, as well as on the total vertical displacement of two palaeoscarps, and they concluded that two possible earthquakes (1509 and 1719) were associated with the G61ciik fault.

639

D 6 n i z E v l e r site

This site is located east of G61cuk city, close to Kavakli and Yenik6y villages (Fig. 2). The eastwest-trending G61cuk fault segment extends eastwards marked by a NW-SE normal fault about 3 km long on land, where seismic ruptures emerge at the surface, with a maximum vertical displacement of 1.5-2 m and a right-lateral component of 0.30 m. This fault affects recent deltaic deposits, i.e. sandy clays, sands, gravel, conglomerates and soil. It has a typical extensional structure perpendicular to the ~3 main principal

Historical seismicity Prior to the 1999 earthquake the well-known strong 1894 earthquake seismic event probably ruptured the segments that were listed activated by the 1999 event (Ambraseys 2001). Several other large earthquakes (e.g. those of 1719 and 1754), as documented by Ambraseys & Finkel (1987) and Ambraseys (2002), could potentially be associated with the western branch of the NAFZ, especially in the Izmit region. Other earthquakes of similar size include the events Of AD 1766, 1754, 1719, 1509 and 989 (Ambraseys & Finkel 1991), and the less well-documented events of AD 554, 447(?), 362(?), 358(?), 268(?) and 268(?) and 68 (Ambraseys 2002). Thus, there is a strong possibility that some of these earthquakes, or possibly other unknown ones, have ruptured segments of the 1999 seismogenic fault.

Trenching sites In the present study, five palaeoseismological trenches were excavated during September 2001 at four sites along the 1999 fault ruptures (Fig. 2): (1) D~niz Evler (G61cuk-Kavakli area); (2) A~a~l Yuvaclk (Kocaeli, south of Izmit); (3) KullarYaylaclk (Kocaeli, SE of Izmit); (4) Acisu (Kocaeli, west of Sapanza Lake). Previous work included preliminary palaeoseismological trenching at Mahmutpa~a~iftli~i (in May 2000, Tutkun & Pavlides 2001; Bulu~-Kankkaya et al. 2006). All these sites were selected on the basis of the existence of surface ruptures, as well as on the basis of surface morphology. The objective during the data selection was to examine the 1999 surface ruptures, and also to try to establish a link between current and past faulting.

Fig 4. Synthetic stratigraphic column of the layers observed in DE-1 trench (both walls). Numbers refer to sampling points for OSL dating.

640

S.B. PAVLIDES E T AL.

stress axis. Extensional features, i.e. m o d e I cracks (see Scholz 1990) form in strike-slip systems, in response to simple shear at a b o u t 45 ~ to the master fault, e.g. the G61ciik-Kavakli normal fault. This fault is believed to have been activated during, or just after, the m a i n shock of 1999 in the f o r m o f a gravitational slip, following a

pre-existing structure, rather than as a typical coseismic rupture. This n o r m a l fault could possibly be a visible strand of a pull-apart structure, located close to the epicentre o f the main shock where the rupture p r o p a g a t i o n began. The seismic ruptures did not follow any k n o w n or m a p p e d fault traces. The round,

Table 1. Events and displacements observed at the DE-1 trench at the D~niz Evler site Event

Timing

1

1999; associated with activity of the main fault zone; cuts through the surface, causing significant ground distortion Before deposition of layer 2 and after deposition of layer 4; associated with activity on the main fault zone Before deposition of layers 6 and 7, and after deposition of layers 8 and 9; associated with an erosional surface and activity on a secondary fault Before the deposition of layers 8 and 9; not associated with any faults, but assumed to be due to the abrupt change of depositional facies

2 3 4?

Displacement (m) 1.5-1.6 0.7 At least 0.2 Unknown

Fig 5. Log of part of the western wall of AY-1 trench. A, sampling points for J4C dating (Table 6). Grid is 1 m x 1 m. Number indicate layers described in the text. Table 2. Events and displacements observed at the A Y-1 trench at the A~a~,l Yuvaczk site

Event

Timing

1 2

1999; associated with the activity of the main fault zone and cuts the surface After deposition of layer 5 and immediately before deposition of layer 4; associated with a secondary fault that stops at the base of layer 4

Displacement (m) 0.25-30 0.7

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY eroded pre-seismic topographic scarp height is similar to or slightly higher than the 1999 one. The 1999 fault rupture formed a N W - S E trending, NE-dipping, typical normal fault scarp with up to 2 m of normal displacement. The Dfiniz Evler excavation site is located in the middle of the G61ctik normal fault, east of Kavakli village. At this site the fault scarp had a vertical throw of 1.5-1.6m. However, the morphology of the area indicates a pre-existing fault scarp, indicating previous reactivation. Such reactivations were traced during the process of trench logging, as described below. Surface effects along this scarp were associated with marsh subsidence, tilting of trees, etc. Trenching at this site (DE-1 trench) revealed that the stratigraphy consists of fluvial sediments at the top, changing to lake deposits at the bottom of the trench (Fig. 3). This was the largest and deepest trench (12 m x 8 m, 4-5 m deep). The stratigraphy includes the following layers from the top to the bottom (Fig. 4): (1) Modern red-brown fine-grained soil, with few pebbles. (2) Alluvium with multicoloured bedrock pebbles, of up to 10-12 cm diameter.

641

(3) Coarse-grained sand with weak lamination towards the north and a very few pebbles (up to 2 cm in diameter). (4) As (2), but with a larger proportion of coarse-grained sand. The lower part of this layer is slightly Fe-oxidized (red-brown coloured). (5) Pebbles with black Mn-oxides; these are arranged roughly parallel to the general bedding. (6) Red-brown conglomerate with pebbles of various sizes (up to 15 cm in diameter). (7) Grey colluvium with a few large pebbles (up to 12 cm) and many smaller ones (2-5 cm), without bedding. (8) Red-brown conglomerate (large pebbles, up to 15 cm). (9) Brown-green fine-grained sand with a very few small pebbles, mainly near its southern edge. (10) Olive green clay with lenses of fine-grained sand. (11) Coarse-grained green sand. According to an analysis of the microstratigraphy and fault geometry, the events listed in Table 1 may be distinguished.

Fig 6. (a) View of the KY- 1 trenching site. Dextral displacement, as documented in displaced walls, is of the order of 3 m. (b) Detail of the 1999 shear zone. Slickenlines indicate a pure strike-slip movement at this site. (c) Liquefaction structures are associated with a previous event in KY-1 trench. Dating of this event was not possible because of lack of organic material.

642

S.B. PAVLIDES ET AL.

Evidence for three prior events is, therefore, present in the trench, each displacing layers by less than the 1999 event. Most of the displacement took place along the main fault zone, which is the same one that produced the 1999 event, but there is also another fault that dies out upwards, indicating that it was activated only once. Analysis of the microstratigraphy shows that there is an abrupt change of facies in the footwall (lacustrine to fluvial), possibly connected with a single earthquake or several events closely spaced in time on the main fault.

A4a~z Yuvactk site Seismic fault fractures at this site appeared in two separate roughly east-west-trending parallel strands, about 40 m apart; the displacement was predominantly normal, while the strike-slip vector was very small. Two north-south-trending

trenches were excavated at this site (AY-1 and AY-2) in both fault strands. AY-1 trench, located on the southern strand, was 12 m long and 3 m deep. Its stratigraphy consists of the following layers from the top to the bottom (Fig. 5): (1) Well-rounded pebbles (5-15 cm) containing abundant gravels and sand lenses. Towards the surface this changes to poorly developed reddish brown soil. (2) Coarse sand laminations with occasional fine gravels; greyish brown. (3) Clay with occasional small to medium-sized pebbles; reddish brown. (4) Dark grey organic silty material, probably associated with a periodic sag pond or swamp bed. (5) Coarse sand with occasional gravels. Analysis of the microstratigraphy shows that there was one event prior to the 1999 earthquake (Table 2).

Fig 7. (a) General view of the trenching site at Acisu. A sag pond is formed parallel to the main fault line. (b) The 1999 surface ruptures 100 m west of the trenching site. The sag pond is formed in a relief inversion, in which the trench AC-1 was excavated (arrow). (c) Detail of the eastern wall of AC-1 trench. Three successive palaeo-sag ponds are visible in the trench; Ottoman ceramic shards are visible in the middle one (arrows). (For dating see Table 7.)

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY AY-2 trench was excavated on the northern strand, where it crossed a sag pond, indicative of active strike-slip displacement. However, no evidence of previous events was found in this trench. K u l l a r - Y a y l a c t k site

This trench (KY-1) was excavated in the Yaylacik area, c. 1 km west of Kullar. A block of flats was demolished during the earthquake, which produced a right-lateral displacement of c. 3 m (Fig. 6a). Two main fault zones were detected in the trench, in contrast to the 1999 rupture, which was confined to just one. The sedimentary structure is chaotic; there are also multiple shear zones and older liquefaction phenomena (Fig. 6b). Liquefaction at this site directly relates to older earthquakes; however, dating of these earthquakes is difficult, because of the lack of suitable material. It seems that the intensity of at least one of the palaeoearthquakes was large enough to cause ejection of underlying material (Fig. 6c). A c i s u site

This trenching site is located east of Acisu village. The site was selected because of the morphology of the area, as well as its proximity to 1999 earthquake fractures. This was the only site where the trench did not cut exposed fault ruptures. The site is located in an elongated sag pond that forms a relief inversion, as it is located on the slope of a 'host' hill, not at its base (Fig. 7a and b). The sag pond is about 80 m long by 20 m wide, and strikes east-west, parallel to the N A F Z at this point. The trench was excavated perpendicular to the major axis of the sag pond, to detect variations in deposits that could be associated with subsequent faulting events. Trench AC-1 was about 10 m long and 2.5 m deep; it was excavated entirely in dark-coloured organic material, indicating that it operated as a sag pond for a long period of time. Two older sag ponds were detected, associated with earlier development phases, the uppermost one being rich in ceramic fragments of the Ottoman period (Fig. 7c). A log of this trench shows that the sag ponds did not form along the same axis, but that their axes are migrating towards the south, thus indicating possible tilting.

for each case was defined mainly by the amount of organic material present in samples. Samples dated by 14C were analysed at the Center for Applied Isotope Studies, University of Georgia, USA (CAIS-UGA), and the OSL dating was performed by E. Bulu~-I~nkkaya at the University from Kocaeli (Turkey). The following paragraphs describe the dating results from the trenches. D 6 n i z E v l e r site

Several samples were dated in this trench by both 14C and OSL methods. Table 3 summarizes the corrected ~4C dates, and Table 4 the OSL ones. Based on these results, as well as the description in Table 1, Table 5 summarizes the palaeoseismic history observed in trench DE-1.

Table 3. Corrected ~4C ages of the samples from the DE-1 trench at the D6niz Evler site UGA no.

Sample no.

Date (years Corrected BP___1 ) date 13C (~

10113 Sand 7C 5580• 10115 Sand 9C 1080•

5 5 9 0 -24.51 1 1 5 0 -20.75

+8 +69

Table 4. The palaeodose (De), annual dose and OSL ages obtained from six samples from the DE-1 trench (Bulus-Ktrtkkaya et al. 2006)

Sample no. De (Gy) 8T 3T 5T 6T 6T1 7T

25.293 15.211 11.939 3.120 2.938 1.303

Dose rate (Gy ka-1)

Additive-dose age (years)

2.469 2.505 2.525 2.534 2.470 2.590

10244 + 764 6072_+ 453 4728 +_353 1231 -t-92 1189 • 89 503 • 37

Table 5. The timing of the events observed at DE-1 trench at D~niz Evler site

Event

Date of the event

3 4?

1999 Prior to AD 800, younger than 2778 Bc; because of the geometry of the fault zone and the stratigraphic relations, probably the uppermost time limit rather than the lowermost one Between 3640 Bc and 4122 Bc Between 4122 Bc and 8294 BC

Dating Dating of samples from the trenches was performed by 14C with anisotropy of magnetic susceptibility (AMS) and by optically stimulated luminescence (OSL). The most suitable method

643

644

S.B. PAVLIDES E T AL.

Table 6. Corrected 14C ages of the samples from the A I1-1 trench at the A~a~t Yuvactk site

UGA no.

Date (years Corrected ~3C Years Be + 1 ) date (~

Sample no.

11472 Soil AY-1/1 11473 SoilAY-1/2

590+70 370___50

590 380

-25.18 -3 -24.12 +14

Asa~t Yuvactk site T w o samples f r o m characteristic spots were d a t e d in this trench (Fig. 5), as s h o w n in Table 6. Based on these results, as well as the interpretation of Table 2, we c o n c l u d e that, apart from the 1999 event, the previous large event occurred between AD 1360 a n d 1570. A cisu s i t e

Table 7. Corrected HC ages of the samples from the .4 C-1 trench at the Acisu site

UGA no.

Date (years Corrected '3C BP_+ 1) date (~

Sample no.

11471 SoilAC-1

750-t-80

770

Years

-23.80 +19

10o0 _c.es :ault tca~

The trench at this site did n o t cut the 1999 rupture; therefore, this e a r t h q u a k e is not docum e n t e d in the trench. Based on the stratigraphical relationships observed in the trench, we can relate the evolution of older sag p o n d s to previous earthquakes, which caused tilting of the surface at the site. One sample was taken from the b o t t o m o f the penultimate sag p o n d (Table 7). In urf

s a length (m) N Ik

W

102 ~

r

110 ~

E

Pressure ridges

80_90 ~

r

~.j 70 ~

~ .,,. . ~ l t ~ : ~ . . ~ . t i l

"--.-.....~ ~

F::\.\ \ . //"1 .v

b

Extensional cracks

~

.

~

100 ~

130 ~

~

Fig 8. (a) Stratigraphy of the Mahmutpa~a?iftlifgi trench. Distinct layers are shown (a, soil; b, sandy mud; c, soil, possibly of palaeo-sag -pond origin). The 1999 fault (contractional) is marked by the small-scale flower structure (I), liquefaction (II and III), seismites (IV) and cracks (V). Numbers 1-3 indicate layers described in the text. (b) Sketch of a series of pressure ridges along the right-stepping linear fault traces at Mahmutpa~a?iflli~i in the Kocaeli district (A~a~l Yuvaclk-Yailacik) where the trenches were excavated. In general, the orientation and movement along stepping small fault branches resulted in local compression; this caused shortening and uplift of c. 5-10 m long (grey areas), accompanied by extensional fractures (120-130~ Inset: the structure of a typical pressure ridge, where the main displacement zone is oriented at 100~ and the contractional structures at 70 ~.

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY

645

Dating of events Events

DE-l/1 2000

AY-1/1

AC-1/1

AY-1/2

DE-l/2

DE-l/3

DE-l/4

| --1999

'

--1999

'

~JP~89

' 920

1290 1120

4) 554

-2OO0

-3131 1-3616 -4000

T292

-6000

1-753o

-8000

-10O00

Fig 9. Synthetic graph of the events detected in palaeoseismological trenches in this study. Most of the events dated in the AD period are associated with known historical events.

the absence of other obvious factors, the creation of the sag ponds can be attributed to alterations in the morphology associated with earthquake events. In this case, the last event that produced enough ground distortion to modify the sag pond would be around AD 1180. M a h m u t p a 4 a f i f t l i ~ i site

Our interpretation of a previously excavated trench was also used. Although results from this trench (Fig. 8a) were already published (Tutkun & Pavlides 2001) based on two samples of 14C dating, new results show that there were subsequent activations of the same fault zone throughout the Holocene (Table 8). The Mahmutpasa~iftli~i area is located in the middle of a shallow linear valley with a flat floor.

Table 8. De: the annual dose rate and OSL ages obtained from the three samples from the Mahmutpasar trench (Bulu4,-Ktrtkkaya et al. 2006) Dose rate Sample no De (Gy) (Gy ka -1) c b a

28.607 23.055 22.650

2.529 2.489 2.756

OSL age (years) 11311 • 843 9263 • 8218 •

Two trenches, 8 m long and 4 m deep, were excavated across the 1999 fault traces on a small pressure ridge (Fig. 8b), to look for fossilized palaeoseismic structures or palaeoearthquakes (Fig. 8a). The trench walls display a record of Holocene (historical) sedimentation. These sediments can be broadly divided into at least three layers plus superficial recent soil; (1) 2 m dark soil; (2) 1 m yellow sandy mud, possibly of palaeo-sag p o n d deposits; (3) 1 m dark palaeosoil with a few pebbles. The pebbles are typically in horizontal position, but disseminated. Structural features, i.e. joints (cracks), seismites and

Table 9. Summary of palaeoseismic events in chronological order traced at the excavated sites Event DE-l/1 AY-1/1 AY-1/2 AC-1/1 DE-l/2 DE-l/3 DE-l/4

Minimum age Maximum age Probable age

AD 1290 AD 1120 3131 BC 4575 BC 9058 BC

1630 AD 1280 AD 920 AD 3616 BC 7530 BC

AD 1999 AD 1999 AD 1509 AD 1180(?) AD 554 4096 BC 8294 BC

DE, D~niz Evler site; AY, A~a~l Yuvaclk site; AC, Acisu site.

646

S.B. PAVLIDES ET AL.

Table 10. Summary table of palaeoseismic events traced in various locations along the NAFZ within the wider study area Events AD Location D~niz Evler D~niz Evler A~a~l Yuvaclk Acisu Eastern segment

1999

1719

1509

X X X

x

x x

X

x

BC 989

554

4096

8294

x

x

x

x

References This paper Klinger et al. (2003) This paper This paper Emre et al. (2003)

palaeo-liquefaction, are associated with the fault. These features are consistent with morphotectonic observations showing that a large-scale dextral movement is documented by 'blind' streams and truncated elongated hills (Tutkun & Pavlides 2001).

The authors are grateful to A. Pinar and A. H. F. Robertson for their useful reviews. Partial support for this research was provided by the General Secretariat for Research and Technology of Greece, the University of Kocaeli (Izmit) and the municipality of Izmit. Thanks are also due to E. Bulu~-Klnkkaya for OSL dating.

Discussion and conclusion

References

The trenches excavated at various sites along the 1999 surface rupture lines show that the same faults were reactivated several times during the late Holocene. Indeed, as shown in Table 9, a number of events were identified during this period. The results in Table 9, as well as their projection as a graph (Fig. 9), are associated with known historical earthquake activity, with the exception of the Ao 1180 event, detected in trench AC-1, which does not coincide with any known historical earthquake. The closest event is that of AD 989. A possible explanation for this discrepancy is that it is associated with the lower part of the middle sag pond at this site, which could have taken some time before it is formed. It is, therefore, assumed that the c. 100 year gap was associated with geomorphological processes, and that the palaeoseismologically derived age is linked to this event. Several of the above-mentioned events are comparable with findings in other trenches, as discussed above. Table 10 presents the recent seismic history of several sites along the N A F Z in the broader study area, as derived from palaeoseismological research. It is, therefore, concluded that the 1999 earthquake fault was indeed reactivated during the past, during several historical earthquakes. Of those, the 1719 event seems to have extended across both major segments of the 1999 fracture, whereas the 1509, 989 and 554 events were most probably confined to its western segment.

AMBRASEYS,N. N. 2001. The earthquake of 10th July 1894 in the Gulf of Izmit (Turkey) and its relation to the earthquake of 17 August 1999. Journal of Seismology, 5, 17-128. AMBRASEYS,N. N. 2002. The seismic activity of the Marmara Sea region over the last 2000 years. Bulletin of the Seismological Society of America, 92, 1-18. AMBRASEYS,N. N. & FINKEL, C. F. 1987. The SarosMarmara earthquake of 9 August 1912. Earthquake Engineering Structured Dynamics, 15, 189-211. AMBRASEYS,N. N. & FINKEL, C. F. 1991. Long-term seismicity of Istanbul and of the Marmara Sea region. Terra Nova, 3, 527-239. AMBRASEYS,N. N. & FINKEL, C. F. 1995. The Seismicity of Turkey and Adjacent Areas--a Historical Review, 1500-1800. EREN, Istanbul. AMBRASEYS,N. & ZATOPEC,A. 1969. The Mudurnu Valley, Western Anatolia, Turkey, Earthquake of 22 July 1967. Bulletin of the Seismological Society of America, 59, 521-589. AWATA, Y., YOSHIOKA,T., EMRE, O., DUMAN,T. Y., DO'AN, A. & TSUKUDA, E. 2000. Segment structures of the surface ruptures associated with the August 17, 1999 Izmit earthquake, Turkey. XXVII General Assembly ESC, Abstracts and Papers, Lisbon, Portugal, 10-15 September 2000. BARKA, A. 1999. The August 17th, 1999 Izmit earthquake. Science, 285, 1858-1859. BARKA, A. 2000. The August 17 and November 12 1999 earthquakes in the eastern Marmara Sea region. In: OKUMURA,K. et al (eds) Active Fault Research for the New Millennium, Proceedings, HOKUDAN-Japan International Symposium, 17-26 January, 2000, S 1-$6.

LATE HOLOCENE SEISMOGENIC FAULT, N TURKEY BARKA, A. A. & KADINSKY-CADE,K. 1988. Strike-slip fault geometry in Turkey and its influence on earthquake activity. Tectonics, 7, 663-684. BOZKURT, E. 2001. Neotectonics of Turkey--a synthesis. Geodinamica Acta, 14, 3-30. BULU,S-KIRIKKAYA, E., I~)ZER, A. M., OSKAY, Z., PAVLIDES, S. t% CHATZIPETROS, A. 2006. Optically stimulated luminescence (OSL) dating of sediment samples of the fault of the Kocaeli (Turkey) earthquake. Quaternary Research (in press). DEMIRTA,S, R. 1996. Paleoseismology of the North Anatolian fault: a case study of the Mudurnu Valley segment, Turkey, International Journal of Rock Mechanics and Mining Sciences and Geomechanics Abstracts, 33(5), 195A-195A(1). EMRE, O., AWATA, Y. & DUMAN, T. Y. (eds) 2003a. Surface rupture associated with the 17 August 1999 Izmit earthquake. General Directorate of Mineral Research and Exploration, Ankara, Special Publication. EMRE, O., TODA, S., DUMAN, T. Y., et al. 2003b. Paleoseismicity of eastern part of 1999 Izmit rupture and Dtizce fault, North Anatolian Fault Zone, Turkey. International Workshop on the North Anatolian, East Anatolian and Dead Sea Fault Systems, Ankara, Turkey, 31 August-12 September 2003, Abstracts Volume, 129. GAUTIER, P., BRUN, J.-P., MORICEAU, R., SOKOUTIS, D., MARTINOD, J. & JOLIVET, L. 1999. Timing, kinematics and cause of Aegean extension: a scenario based on a comparison with simple analogue experiments. Tectonophysics, 315, 31-72. HANCOCK, P. L. & BARKA,A. A. 1987. Kinematic indicators on active normal faults in western Turkey. Journal of Structural Geology, 9, 415-430. IKEDA, Y., SUZUKI, Y., HERECE, E., SARO~LU, F., ISIKARA, A. M. & HONKURA, Y. 1991. Geological evidence for the last two faulting events on the North Anatolian fault zone in the Mudurnu Valley, western Turkey. Tectonophysics, 193, 335-345. JACKSON, J. 1994. Active tectonics of the Aegean region. Annual Review of Earth and Planetary Sciences, 22, 239--271. KLINGER, Y., SIEH, K., ALTUNEL, E., et al. 2003. Paleoseismic evidence of characteristic slip on the western segment of the North Anatolia Fault, Turkey. Bulletin of the Seismological Society of America, 93, 231%2332. LETTIS, W., BACHHUBER, J., WITTER, R., et al. 2002. Influence of releasing step-overs on surface fault ruptures and fault segmentation: examples from the 17 August 199 Izmit earthquake on the North Anatolian Fault, Turkey. Bulletin of the Seismological Society of America, 92(1), 19-42. MCCALVIN, J. P. 1996. Paleoseismology. Academic Press, San Diego, CA. PAVLIDES, S. 1996. First palaeoseismological results from Greece. Annali di Geofisica, 34, 545-555.

647

PAVLIDES, S. tfr CAPUTO, R. 2004. Magnitude versus faults' surface parameters: quantitative relationships from the Aegean. Tectonophysics, 380, 59-188. PAVLIDES, S., ZHANG, P. & PANTOSTI, D. 1999. Earthquakes, active faulting, and paleoseismological studies for the reconstruction of the seismic history of faults. Tectonophysics, 308. ROCKWELL, T., BARKA, A. A., THORUP, K., DAWSON, T. & AKYUZ, S. 2001. Paleoseismology of the Gazik6y-Saros segment of the North Anatolian fault, northwestern Turkey: comparison between historical and trench records. Journal of Seismology, 5, 433-448. SARO~LU, F. & YILMAZ, Y. 1987. Geological evolution and basin models during neotectonic episodes in eastern Anatolia. Bulletin of Mineral Research and Exploration, 107, 74-94. SCHOLZ, CH. 1990. The Mechanics of Earthquakes and Faulting. Cambridge University Press, Cambridge. ~ENGOR, A. M. C. 1979. Mid-Mesozoic closure of Permo-Triassic Tethys and its implications. Nature, 279, 590-593. ~ENGOR, A. M. C., G6ROR, N. & SAROrLU, F. 1985. Strike-slip faulting and related basin formation in zones of tectonic escape: Turkey as a case study. In: BIDDLE, K. T. & CHRISTIE-BLICK,N. (eds) Strikeslip Deformation, Basin Formation and Sedimentation. Society of Economic Paleontolgists and Mineralgists, Special Publications, 37, 227-264. SIEH, K. E. 1978. Prehistoric large earthquakes produced by slip on the San Andreas Fault at Pallet Creek, California. Journal of Geophysical Research, 83, 3907-3939. SUGAI, T., TODA, S., EMRE, l~)., et al. 2000. A preliminary result of trenching study across the 1999 August Izmit earthquake rupture, Turkey. EOS Transactions American Geophysical Union, 81(48), Fall Meeting Supplement, Abstract $51A-02. TAYMAZ, T., JACKSON, J. & MCKENZIE, D. 1991. Active tectonics of the north and central Aegean Sea. Geophyscis Journal International, 106, 433-490. TODA, S., EMRE, 0., DUMAN, T. Y., et al. 2003. Behavior of the 1999 rupture zones, western North Anatolian Fault during the past 500-2000 years. International Workshop on the North Anatolian, East Anatolian and Dead Sea Fault Systems, Ankara, Turkey, 31 August-12 September 2003, Abstracts Volume, 21. TUTKUN, Z. & PAVLIDES, S. 2001. Small scale contractional-extensional structure and morphotectonics along the fault traces of Izmit (Turkey). Bulletin of the Geological Society of Greece, 34(1), 345-532. YEATS, R. S., SIEH, K. & ALLEN, C. R. 1997. The Geology of Earthquakes. Oxford University Press, Oxford.

Neotectonic and seismological data concerning major active faults, and the stress regimes of Northern Greece D. M O U N T R A K I S

1, M. T R A N O S l, C. P A P A Z A C H O S 2, E. T H O M A I D O U

E. K A R A G I A N N I

2 & D. V A M V A K A R I S

1,

2

~Department of Geology, School o f Geology, Aristotle University o f Thessaloniki, GR-54124 Thessaloniki, Greece (e-mail: dmountra@geo, auth.gr) 2Geophysical Department, School o f Geology, Aristotle University o f Thessaloniki, GR-54124 Thessaloniki, Greece Abstract: Northern Greece is an intracontinental region behind the Hellenic subduction zone, with widespread seismic activity (ranging from low to high), with strong destructive earthquakes of M >_6.0 in historical to recent times. Geological and seismological data indicate that recent seismic activity is mainly localized along large, inherited, fault zones, which have transected Northern Greece since Oligocene-Miocene times. The main active fault zones in Thrace, and Eastern and Central Macedonia strike approximately east-west, with lengths of 40-120 km. Fault segments strike WNW-ESE to ENE-WSW and range from 10 to 30 km in length. In Western Macedonia the main active fault zones strike NE-SW to ENE-WSW with lengths of 40-60 km and consist of 10-30 km segments. The region's strong earthquakes are usually associated with reactivation of these fault segments and are estimated at M = 5.6-6.5. Focal mechanisms and fault-slip data from the fault zones indicate a change in the trend of extension axes from NNE-SSW in Eastern Macedonia-Thrace to NNW-SSE in Western Macedonia. Thus, neotectonic and seismological data suggest that variations in fault patterns, as determined from the large inherited fault zones transecting Northern Greece, are the major factor governing this change in the trend of maximum extension. This interpretation is consistent with the long-lived arcuate shape of the Hellenic subduction zone.

Brittle deformation within intracontinental regions is usually associated with distributed faulting and seismic activity. However, within such broad deforming regions some parts, although transected by large faults, are characterized by relatively less intense seismic activity and slip rates. Northern Greece lies in the inner part of the continental collision region of N W Greece and Albania, close to the intensely n o r t h south Aegean extended Sea. This deformation is driven by the Hellenic subduction zone and the westward extrusion of Anatolia (Fig. 1a). In the older Greek seismic code, some large areas of Macedonia (e.g. Western Macedonia) are classified as almost 'aseismic' and other adjacent ones (e.g. Central Macedonia) as highly seismic. Recent instrumental data obtained from the regional seismological network of the Aristotle University of Thessaloniki, after its installation in 1980, have confirmed the existence of significant seismicity in Northern Greece (Fig. lb), particularly along narrow rupture zones (Papazachos et al. 2001). Furthermore, large earthquakes have been reported in historical times (Papazachos & Papazachou 2003) and several strong and destructive ones have been recorded instrumentally during the last century.

Of these, the best known are the 1978 Thessaloniki earthquake of M = 6.5 (Papazachos et al. 1979) and the 1995 Kozani-Grevena earthquake (Pavlides et al. 1995; Mountrakis et al. 1996c, 1998; Papazachos et al. 1998a), with the latter occurring in the 'low seismicity' Western Macedonia region. The rupture zones in Northern Greece need a thorough re-examination in the light of recent advances in our understanding of active faulting and the new data provided by the Aristotle University seismological network. Our aim is to identify the large active fault zones in Northern Greece from their geometric, kinematic and seismotectonic features, to define controlling stresses, and to elucidate the continuity or extent of these zones. The main aim is to better understand the active deformation of these areas and to determine whether fault deformation in Northern Greece is concentrated into relatively narrow bands or zones. As one of the major issues in continental mechanics is how fault slips in the upper crust reflect more distributed flow in the lower lithosphere (e.g. McKenzie & Jackson 1983; Molnar & Gipson 1994; Bourne et al. 1998), information about active faulting and driving stresses in such areas is very useful.

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopment of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 649470. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

650

D. MOUNTRAKIS E T A L .

Fig. 1. (a) Schematic map showing the position of Northern Greece in the broader Eastern Mediterranean geotectonic regime (modified from Papazachos & Papazachou 2003). (b) Spatial distribution of instrumentally recorded earthquakes (1911-2004) with magnitudes M > 3.0 for the broader study area. CTF, Cephalonia Transform Fault; RTF, Rhodes Transform Fault; PTF, Paphos Transform Fault.

Geology and seismotectonics of Northern Greece Northern Greece lies in the inner part of the Hellenic orogen and comprises rocks belonging

to the Internal Hellenide zones and the Hellenic hinterland. The Mesozoic to Tertiary Alpine orogeny began, on a more regional scale, with the convergence of the Eurasian plate and the Cimmerian and Apulian continental

ACTIVE FAULTS AND STRESS, N GREECE fragments (Mountrakis et al. 1983; Mountrakis 1986; Robertson et al. 1996). The rocks of these zones form the pre-Alpine and Alpine basement of Northern Greece, on which large Neogene and Quaternary basins developed. The late collisional processes dating from Late OligoceneEarly Miocene times were associated with large strike-slip faults that are recognized in Thrace (Karfakis & Doutsos 1995), Central Macedonia (Tranos 1998; Tranos et al. 1999) and Western Macedonia (Mountrakis 1983; Zelilidis et al. 2002; Vamvaka et al. 2004). From the Late Miocene onwards, the subduction of the African plate beneath Eurasia along the section of the Hellenic arc from the Ionian Islands southwards to Crete and further east to Rhodes has dominated Greece and created the Hellenic volcanic arc. Northern Greece is characterized by intracontinental brittle deformation; it lies within the internal part of the Hellenic subduction zone and reveals considerable extensional deformation, orthogonal to the subduction zone. Other geotectonic processes include the continuing collision of Eurasia and the Adriatic microplate and the lateral extrusion of the Anatolia microplate towards the Aegean Sea (McKenzie 1978; Taymaz et al. 1991; Papazachos 1999). The influence of the latter processes on Northern Greece is being considered although recent papers suggest that the rightlateral strike-slip deformation of the North Aegean Trough, activated by the lateral extrusion of Anatolia westwards, also encompasses faults in Central Macedonia and Eastern Thrace (Pavlides et al. 1990; Koukouvelas & Aydin 2002). In addition, Koukouvelas & Aydin (2002) have attributed the exposure of large basins in Central Macedonia and Thrace to the contemporaneous activation of faults that strike ENEWSW and function as right-lateral strike-slip faults, and NW-SE-striking normal faults. Since the Late Miocene, the neotectonic deformation of Northern Greece has been dominated by an extensional stress regime, with the least principal stress axis (CY3)oriented NE-SW during the Late Miocene-Pliocene and north-south during the Early Pleistocene-present (Mercier et al. 1989). The NE-SW extensional stress field mainly activated NNW-SSE- to NW-SEstriking normal faults and led to the formation of many fault-bounded basins (e.g. Drama, Strymonas, Axios-Thessaloniki and Ptolemais), whereas the north-south extensional stress field has mostly activated east-west trending normal faults, thus reshaping the already developed fault-bounded basins. However, the least principal stress axis (%) of the contemporary stress field, as determined

651

from neotectonic observations, reveals a distinct change from NNE-SSW in Eastern Macedonia and Thrace to NNW-SSE in Western Macedonia (Mercier 1981; Le Pichon et al. 1982; Mercier et al. 1987; Tranos & Mountrakis 1998). A similar spatial variation is also observed in the T-axis of the available earthquake fault-plane solutions and the corresponding strain-rate tensor extensional eigenvalues (Papazachos et al. 1992; Papazachos & Kiratzi 1996). This has been attributed either to a spatial change of the lithospheric loading as a result of contemporary lithospheric processes (Mercier et al. 1987) or to the pre-existing fault pattern (Tranos & Mountrakis 1998), and the fact that the faults behave not only in an Andersonian mode, but also obey the 3D deformational strain (Tranos 1998; Tranos & Mountrakis 1998).

Seismic activity and fault-plane solutions in Northern Greece The most recent seismic activity in Northern Greece, as shown by the M > 3 . 0 earthquakes of 1982-2001 (Fig. 2), strongly reflects the aftershock sequences related to the latest strong events, such as the 1978 Thessaloniki, 1990 Griva and 1995 Kozani-Grevena earthquakes. The seismic information clearly defines the rupture zones associated with these strong earthquakes and indicates areas of high seismicity; it also indicates the strike of some fault zones. For this reason, a recently developed database of fault-plane solutions derived from the seismological network of the Aristotle University of Thessaloniki was used to define the seismotectonic characteristics of the active faults and to complement available geological information (i.e. neotectonic criteria; also geometric and kinematic characteristics) used to recognize such faults. The seismological and neotectonic criteria that have been used for the characterization of faults as being active are those already adopted during the neotectonic mapping of Greece by the Greek Earthquake Planning and Protection Organization. Seismically active faults are defined as being directly associated with welldefined historical earthquakes. Using only stratigraphic criteria, active faults are defined as those activated since the late Pleistocene. Additionally, several other features of faults are used, such as: (1) Geomorphological features, i.e. the linear trend of a mountain front along which successive Quaternary fan or colluvial deposits, triangular facets, fault scarps etc. are distributed; (2) Tectonic features, i.e. correlation of fault-slip data

652

D. MOUNTRAKIS E T A L . .

.

42 ~

.

.

.

.

.

.

.

.

.

Fault Plane Solutions W~fogm

?, .......................................................................

T.axis

m~l~g

"~,.

Firm motions

....." .

U$~d , . ,h,$ $,ud,

~R..~.

....,_j ......

.............

j,

..9

).

\

i?

x.,,,

..........

, 40 ~

i :I

............................ ~ ~ ......................... i......................... " [ II

............................................................

22 ~

;

/

:

"

24"

26 ~

Fig. 2. Distribution of available extension (T) axes, corresponding to the most recent database of earthquake fault-plane solutions for the study area (Papazachos et al. 2004). Information corresponding to fault-plane solutions was determined using different methods, as well as those used in the present study for specific fault zones, as are denoted by different arrows (see key).

with those of verified, well-known seismic faults with similar orientations; (3) linear alignment of springs or spring deposits. In addition to data from the Thessaloniki network, the database includes fault-plane solutions of large and intermediate magnitude events, as determined by waveform modelling, from international centres (Harvard, ETH, INGV, etc.) or elsewhere, together with several intermediate and smaller-magnitude events as determined from first motions. The procedure for finding firstmotion fault-plane solutions was calibrated using the available common solutions from waveform modelling (Papazachos et al. 2004). The spatial distribution of extension axes, as determined from earthquake fault-plane solutions, is presented in Figure 2, where black vectors denote the 0-3-axes used in the detailed analysis of the selected active faults presented below. The faultplane solutions were correlated with field observations along the faults to help define their activity. Faulting of the area

To identify the active rupture zones exposed in Northern Greece, we tried to combine information from the latest seismic activity with that from geological observations along the large fault zones, or as reported in previous work. The most important issues are the spatial distribution

and focal mechanisms of both small and large magnitude earthquakes. This information has been used to locate the strain mainly along large fault zones, which are the most likely to produce strong earthquakes. To compare easily the principal strain axes derived from focal mechanisms with fault-slip data recorded along the faults studied, these faults were analysed with a simple graphical method that constructs the kinematic axes of the faults, i.e. the principal incremental shortening (P) and extension (T) axes using the program 'FaultKin' (Allmendinger 2001). Each pair of axes lies in the movement plane of the fault (i.e. a plane perpendicular to the fault plane that contains the unit vector parallel to the direction of accumulated slip, and the normal vector to the fault plane). Each pair of axes makes an angle of 45 ~ with both vectors. To distinguish the shortening and extension axes, information on the relative sense of slip is needed. Also, the principal stress axes (0-1, 0-2, 0-3) of the rupture zones were defined using a program Duyster (1999). This calculates the stress directions from the recorded fault-slip data with the PT method after Turner (1953). The method is a very simple way to determine palaeostress directions assuming that fractures generate parallel to 0-2 with the angle | between the 0-~ and the fault plane being 30 ~ Although this is valid according to the M o h r Coulomb criterion applied to a homogeneous rock mass, experimentally obtained values of

ACTIVE FAULTS AND STRESS, N GREECE | range between 17~ and 40 ~ (e.g. Hubbert 1951; Byerlee 1968; Jaeger & Cook 1979) and imply that an angle of O = 30 ~ is a reasonable approximation. Using this approach we subdivide the large area 1 Northern Greece into three areas, which, from east to west, are Eastern Macedonia and Thrace, Central Macedonia and Western Macedonia. The fault pattern, as defined by the larger fault zones, can be briefly described as follows. (1) Eastern Macedonia and Thrace are dominated by NE-SW- and east-west-striking faults. (2) In Central Macedonia large basins strike N N W - S S E to NW-SE; however, eastwest-striking faults dominate the recent fault pattern. (3) In Western Macedonia, NE-SW- to ENE-WSW-striking faults predominate, with subordinate NW-SE- and east-west-striking faults. On this basis the following large rupture zones have been established (Fig. 3).

Eastern Macedonia and Thrace The mountainous Eastern MacedonianThrace region includes several large east-weststriking fault-bounded basins, namely the Alexandroupolis, Drama and Kavala-XanthiKomotini basins. This area has low seismicity, with very few historically reported earthquakes; of these, Drama in 1829 (M=7.3) and 1867 (M = 6.0), Komotini (M=6.7) in 1784 and Didimoticho (M = 7.5) in 1752 were the strongest (Papazachos & Papazachou 2003). This low activity is puzzling, as Eastern Macedonia and Thrace are close to the seismically active North Aegean Trough and contain kilometres-long fault zones of similar strike (i.e. large eastwest- to ENE-WSW-striking fault zones), which dominate the fault pattern of the area. The east-west-striking fault zones include two different geometric types: (1) large rectilinear fault zones with constant east-west strike; (2) fault zones related to faults whose strike varies from N E - S W to W N W - E S E and that coalesced during the neotectonic period, e.g. the KavalaXanthi-Komotini fault. In the second case, the NE-SW- and W N W ESE-striking faults bounded the EoceneOligocene molasse-type sediments and controlled exposures of Oligocene volcanic rocks (see also Karfakis & Doutsos 1995). The main fault zones exposed in Eastern Macedonia and Thrace are KavalaXanthi-Komotini, Maronia-Alexandroupolis, Drama-Prosotsani, Serres-Nea Zichni and Ofrinio-Galipsos, which all strike more or less east-west.

653

Kavala-Xanthi-Komotini fault zone The Kavala-Xanthi-Komotini fault zone, the most important in Eastern Macedonia-Thrace, is an east-west master fault zone > 120 km long (Lyberis 1984) that runs very close to the cities of Kavala, Xanthi and Komotini (Figs 3 and 4a). This clearly demonstrates the importance of assessing the related seismic hazard. Although the fault zone generally strikes east-west, it comprises four fault segments that vary in strike from N E - S W to W N W - E S E and reveal different geological features (Mountrakis & Tranos 2004). The four segments are as follows:

Chrisoupolis-Xanthi fault segment. This 35 km long NE-SW-striking (c. 55 ~ segment runs along the SE slopes of Mt Lekani, between the coast east of the city of Kavala and the city of Xanthi. The fault separates the marbles of the Pangeon Unit in the footwall from the tectonically overlying migmatites and gneisses of the Sidironero Unit that constitute the hanging wall, along with overlying post-Alpine Tertiary molasse-type sediments (Kilias & Mountrakis 1998). The segment is characterized by remarkable triangular facets and fault scarps. The fault surface strikes N E - S W and dips at moderate angles towards the SE, and is well exposed at Paradisos village. A thin, dark brown oxidized carapace covers this fault surface, and records three generations of slickenlines (Fig. 4a). The older slickenlines indicate strike-slip movement, whereas the younger indicate a N N W - S S E extension axis (T). Xanthi-Iasmos fault segment. This 27 km long segment strikes W N W - E S E to east-west. It runs from the city of Xanthi to the east of Iasmos village, until it ends against a N N W - S S E rectilinear fault trace along Xiropotamos stream. This segment contains several subparallel fault branches that form successive fault scarps, with the most basinward ones being the most impressive. The fault affects the metamorphic rocks of the Sidironero Unit and the Tertiary molassetype sediments; it also affects the Late Oligocene Xanthi granitoid (c. 28 Ma, Kyriakopoulos 1987), forming well-exposed fault surfaces dipping steeply southward. The slickenlines along this segment indicate successive strike-slip movements, with an oblique left-lateral normal movement and finally a right-lateral oblique movement (Fig. 4a).

Iasmos-Komotini fault segment. This 16km long segment strikes E N E - W S W and comprises two parallel left-stepping and overlapping fault strands: the Polyanthos-Mega Piston and Mega Piston-Agiasma fault strands, which are about 6 km and 13 km long, respectively (Fig. 4a).

654

D. M O U N T R A K I S

E T AL.

~< ..~=~ ~< , ,,,,~

"~ O r~

0

0

0

z~

9~ o ~

~o ;;o "~ m~" N

"~Nrm

r~.~ 9 ~ 0 ~

~

m ,-.M . . - .

~.~

.~0

AI ~ ~ ~-i

ACTIVE FAULTS AND STRESS, N GREECE These fault segments juxtapose basement with Plio-Quaternary fanglomerates and Holocene deposits that entirely cover the underlying Neogene sediments of the Komotini basin (Diamantis 1985). However, the Mega Piston-Agiasma strand can be traced further northeastwards. Along this fault Tertiary molasse-type sediments were juxtaposed with the metamorphic rocks of the Rhodope massif, suggesting that the N E SW-striking faults are older, perhaps reactivated pre-Neogene structures. The Iasmos-Komotini fault segment is characterized by multiple reactivations revealing slickenlines of right-lateral strike-slip movement overprinted by younger ones that indicate normal-sense reactivations.

Komotini-Sapes fault segment. This segment, over 30 km long, reveals a complicated geometry, as it resembles several W N W - E S E and east-west synthetic fault strands < 8 km long that dip SSW to south at medium to high angles (Fig. 4a). Predominant are the Tichiro, Gratini, Dokos and Fillira-Skaloma faults, which gradually lower the hilly landscape towards the south. This is a boundary fault, striking WNW-ESE, which controlled the deposition of molasse-type and especially Neogene sediments (Karfakis & Doutsos 1995). Our mapping indicates that WNW-ESE-striking faults are truncated by east-west trending faults, e.g. the east-west Gratini fault truncates the W N W - E S E Tichiro fault, whereas other smaller and non-continuous east-west-striking faults have been observed to continue westwards to the city of Komotini. Right- and left-lateral oblique normal movements here give rise to N N E - S S W and northsouth extension, respectively, along this fault segment. The youngest activity can be traced east of Polyanthos village, where a rectilinear fault line, a few tens of metres long, with a vertical offset of less than a metre, has been found within the Plio-Pleistocene fanglomerate sediments of the hanging wall. Concerning the latest kinematics of the Kavala-Xanthi-Komotini fault zone, the latest slickenlines recorded along the variously oriented fault segments are related to normal reactivations and define a stress ellipsoid whose least principal stress axis (%) is oriented almost north-south (Fig. 4b). The most recent seismic activity along the fault zone is around Komotini city, where a destructive M = 6.7 earthquake was reported in 1784 (Papazachos & Papazachou 2003). It seems reasonable to associate this earthquake with the Komotini-Sapes fault segment, as the latter faces the epicentre (Fig. 4a) and a young fault scarp has been found cutting the Plio-Pleistocene fanglomerate sediments of its hanging wall.

655

Maronia-Alexandroupolis fault zone This 35 km long east-west-striking fault zone localizes the coast from Maronia village to the city of Alexandroupolis (Figs 3 and 4a) and is very important for a seismic hazard assessment of the city, which is built on its extension. However, this zone does not exhibit a single traceable fault surface, but rather seems to comprise fault segments of W N W - E S E and E N E - W S W strike. Hence, another fault strand could be present in this fault zone; i.e. the 7 km long, WNW-ESE-striking Avantas fault, which bounds the Alexandroupolis basin to the north (Fig. 4a). Close to Avantas village this is clearly observed to modify the contact between overlying Eocene-Oligocene clastic marls and clays and underlying Upper Lutetian nummulitic limestones, which dip southwards at moderate angles, forming a rectilinear, steep fault scarp along which fault slickensides exhibit right-lateral oblique slickenlines and indicate a N E - S W extension axis (T) (Fig. 4c). The WNW-ESE-striking fault segment controls the coastline east of Maronia village, forming steep corrugated slickensides, which dip SSW at about 60 ~ It also affects the Mesozoic and Tertiary rocks and forms a composite cataclastic zone with corrugated slickenside surfaces and a 1 m thick cohesive cataclasite. The slickensides exhibit dip-slip slickenlines and shearing microstructures that indicate normal reactivation and a N N E - S S W extension axis (T) (Fig. 4c). Horizontal continental-type Pleistocene deposits abut these fault slickensides. The small antithetic faults in those sediments suggest Quaternary reactivation of the fault. The closest earthquake (M=4.6, 25.60~ 40.69~ which occurred in the hanging wall of this fault on 5 March 2002, was located in the Thracian Sea, between the coast and the island of Samothraki. Its focal mechanism exhibits similar geometry and kinematics to the western segment of the fault zone (Fig. 4c and d) suggesting that it is active. A similar conclusion can be drawn from a fault-plane solution of the M = 5.1 earthquake that occurred close to this fault zone on 27 June 2004 (26.04~ 40.78~ This exhibits a similar strike and extension axis (T) to the Avantas fault, although it seems to have a more significant right-lateral strike-slip component.

Drama-Prosotsani fault zone The 30km long east-west-striking D r a m a Prosotsani fault zone is located on the southern slopes of Mt Falakron and places the NeogeneQuaternary sediments of the Drama basin against the marbles of the pre-Miocene basement (Figs3 and 5a). The zone has a rather

656

D. MOUNTRAKIS E T A L .

Fig. 4. (a) Generalized geological-tectonic map of the Kavala-Xanthi-Komotini and MaroniaAlexandroupolis fault zones (modified from Mountrakis & Tranos 2004). Stereographic projections on the map show the main movements of the fault zone segments, with 1, 2, 3, being the order from the oldest to the youngest movements. Pa, Paradisos; Si, Simantra; Ko, Koptero; Ias, Iasmos; Po, Polyanthos; MP, Mega Piston; Mi, Mischos; Ag, Agiasma; Ti, Tichiro; Gr, Gratini; Do, Dokos; Fi, Fillira; Sk, Skaloma; Ni, Nikites; Ki, Kinira. (b) The contemporary stress regime, as defined by the latest normal movement along the KavalaXanthi-Komotini fault zone, using the program by Duyster (1999). (c) Stereographic projection of the latest movement of the Maronia-Alexandroupolis fault zone and Avantas fault. (d) Focal mechanisms of the earthquakes closest to the fault zone. The latest movement of the fault zone corresponds well to that defined by the focal mechanisms and indicates a NNE-SSW to NE-SW extension axis (T). complicated geometry, with a main east-weststriking boundary fault, the Prosotsani fault and, basinwards, several smaller, subparallel and interrupted fault strands that dip steeply southward. The latter faults delineate another eastwest-striking fault branch, the D r a m a fault; along with the Prosotsani fault this forms a

right-stepping fault geometry covered by an extensive alluvial plain, which is also affected by steep to vertical east-west- to E N E - W S W striking mesoscale joints and faults. This alluvial plain obscures the trace of the main fault, suggesting that the slip rate of the fault zone, particularly the boundary fault, is rather small.

ACTIVE FAULTS AND STRESS, N GREECE The latest observed movement along the Drama fault has a normal offset. Near Kaliphytos village (Fig. 5a), we observed that the fault rock, Quaternary reddish cemented brecciated fault gouge, is transected by younger faults, indicating normal reactivation, and that Quaternary screes and fanglomerates rest with a buttress unconformity on Neogene sediments. In addition, the presence of travertine deposits associated with perennial springs (e.g. in the centre of the city of Drama and in Mylopotamos village) along the Drama fault branch also suggests recent reactivation. The kinematics of the Drama-Prosotsani fault zone, as defined by the latest slickenlines (Fig. 5b), corresponds to a north-south extensional strain field. A similarly oriented extension axis (T= 162-12 ~ is defined by the focal mechanism of the nearby M = 5.5 Volakas earthquake

657

that occurred on 9 November 1985 (23.9~ 41.3~

Serres-Nea Zichni fault zone The east-west-striking Serres-Nea Zichni fault zone lies east of the NW-SE-striking Neogene Strymon basin and defines the basin boundary east of the city of Serres (Figs. 3 and 6). This fault is about 30 km long and controls the deposition of the Quaternary sediments along the southern slopes of Mt Menikion, from the city of Serres to Nea Zichni village. It exhibits a complicated geometry, as it includes several fault segments of ENE-WSW, east-west and N W - S E strike. These are: (1) the Serres segment; (2) the Eptamili-Ag. Pnevma segment; (3) the Ag. Pnevma-Metalla segment; (4) the Dafnoudi-Nea Zichni segment.

Fig. 5. (a) Generalized geological-tectonic map and (b) schematic cross-section and stereographic projection of the latest movement and the defined P, T kinematic axes of the Drama-Prosotsani fault zone.

658

D. MOUNTRAKIS ET AL.

The Serres fault segment, exposed between Lefkonas and Eptamili villages, runs through the city of Serres. It includes ENE-WSW- and eastwest-trending faults that differentiate a hilly area to the north, made up of Neogene sediments, from the Quaternary floodplain to the south for a length of about 6.5 km. The east-west-striking Eptamili-Ag. Pnevma and WNW-ESE-striking Ag. Pnevma-Metalla fault segments are the main boundary faults that separate the crystalline basement from the Neogene sediments of the Strymon basin. These are both c. 10 km long and consist of subparallel faults towards the basin that dip southwards at high to very high angles. The fault surfaces are dominated by normal slickenlines that overprint older strike-slip slickenlines (Tranos & Mountrakis 2004). The strike-slip slickenlines exhibit similar kinematics to that of the similarly oriented faults that were observed within Neogene sediments, north of the city of Serres (Karistineos 1984), but not within the Quaternary sediments of this area. The strong modification of this boundary related to Quaternary normal reactivation of

fault segments has resulted in the juxtaposition of upper Pleistocene fan deposits with the basement. Although no data are available for any historical earthquakes in the Serres-Nea Zichni fault zone, considering its possible future reactivation, it seems that the potentially most active sections are the Eptamili-Ag. Pnevma and Ag. Pnevma-Metalla segments, which both define a N N W - S S E extension and which both affect the later Quaternary sediments.

Ofrinio-Galipsos fault zone Along the southern slopes of Mt Pangeon, east of the River Strymon (Fig. 3), that is a 10 km long fault array composes of three synthetic subparallel east-west-striking faults that dip steeply southwards, as mapped between the villages of Ofrinio and Galipsos. The northernmost of these faults bounds Neogene sediments and delineates the east-west front of the mountain. In places, it forms triangular facets and fault scarps that dip steeply southward. The southern faults affect the Neogene sediments, causing tilting of up to 40 ~

Fig. 6. Geological-tectonic map of the Serres fault zone (modified from Tranos & Mountrakis 2004). Inset: stereographic projection (equal area, lower hemisphere) of the fault-slip data along the fault zone. @, rye;&, cr2;IB, ~3-

ACTIVE FAULTS AND STRESS, N GREECE to either the south or the north. In addition, near Galipsos village small steeply- dipping normal fault surfaces located along the trace of the boundary fault have affected Quaternary alluvial fan deposits (Tranos 1998), suggesting Quaternary reactivation. The Ofrinio-Galipsos fault zone extends ENE as far as the south-dipping normal faults of the Kavala-Eleftheroupoli fault zone, although it seems to truncate the latter. The fault-slip data along the fault zone indicate a N N W - S S E extension axis (T) and a subvertical shortening axis (P) (Fig. 7).

Central Macedonia Central Macedonia possesses several NeogeneQuaternary basins, namely the Thessaloniki, Yanitsa, Kilkis, Mygdonia and Strymon basins, which strike N W - S E and east-west, forming large plains between the mountainous terrain of the pre-Alpine and Alpine basement (Fig. 1). The fault pattern of Central Macedonia is similar to that of Eastern Macedonia and Thrace, i.e. east-west, varying from W N W - E S E to ENE-WSW, and includes faults that strike N W SE and NE-SW. More precisely, the prevalent east-west-striking faults form large fault systems that bound Neogene and Quaternary basins. The NW-SE-striking faults follow the orogenic fabric and form large N W - S E Neogene basins. These

N

659

faults are nowadays less well defined, as they have been cut or truncated by east-west-striking faults (Tranos et al. 2003). The most significant east-west-striking faults exposed in Central Macedonia (Fig. 1) are the South Mygdonia fault system, the Stratoni fault, the Sochos-Mavrouda fault zone, the Vourvourou fault, the Northern Almopias fault zone, the Kerkini fault zone, the Anthemountas fault zone and the Northern Pieria fault zone. After the 1978 Thessaloniki earthquake, several of these were described in detail. Here, we will summarize those faults for which published information exists; i.e. the Southern Mygdonia fault system, the Stratoni fault, the Sochos-Mavrouda fault zone, the Vourvourou fault and the Northern Almopia fault zone. Southern M y g d o n i a f a u l t system

The Southern Mygdonia fault system (Fig. 3), the most intensely studied in Northern Greece (Papazachos et al. 1979; Mercier et al. 1983; Mountrakis et al. 1983, 1996a,b; Pavlides & Kilias 1987; Tranos 1998; Tranos et al. 2003), delineates the stretched Mygdonia graben to the south for 60 km. Its complex geometry has resulted from the coalescence of pre-existing 2 5 k m long WNW-ESE-striking faults and 10km long NE-SW- to ENE-WSW-striking faults that dip steeply to the north and which were reactivated in Quaternary to Recent times as active normal fault segments defining northsouth extension. Fault segments of this fault zone were reactivated, causing the 1978 Thessaloniki earthquake; because of this reactivation, impressive seismic fissures have been observed for 20 km along the fault zone (Papazachos et al. 1979). Stratoni fault

The Stratoni fault is also an east-west-striking active normal fault with an observed length of 15-20 km (the sea obscures its eastward extension) and defines north-south extension (Pavlides & Tranos 1991). According to Pavlides & Tranos, the 1932 Ierissos earthquake of M =7.1 was possibly due to reactivation of the Stratoni fault. S o c h o s - M a v r o u d a f a u l t zone

Fig. 7. Stereographic projection (lower hemisphere, equal area) of the latest movement along the boundary fault of the Ofrinio-Galipsos fault zone and the P, T. kinematic axes.

The c. 30km long Sochos-Mavrouda fault zone strikes east-west and dips southwards (Mountrakis et al. 1996a). The most active segments of this zone are the Sochos and Mavrouda faults, which have similar strikes and lengths of about 8-10km, forming a right-stepping

660

D. MOUNTRAKIS ET AL.

This is a 25 km long ENE-WSW-striking fault zone that transects the Internal Hellenide zones of Almopia and Paikon close to and parallel the Greek-FYROM border (Fig. 3). Its geometry and kinematics have been described by Pavlides et al. (1990). It is noteworthy because it abruptly ends the prolongation of Mt Voras to the south and forms the large inter-mountain Almopia or Aridea basin. The zone consists of three segments; from west to east, these are the Loutraki, Promachi and Theriopetra faults, which are about 8-10 km long and dip steeply south, producing an asymmetric or half-graben development. The Northern Almopia fault zone originated as an old strike-slip fault in Tertiary times, and was reactivated as a normal fault in neotectonic times (Mountrakis 1976; Pavlides et al. 1990). Thermal springs and Quaternary travertine outcrops have been mapped along the fault. The fact that these travertines were later affected by ENE-WSW-striking, south-dipping normal faults, with similarly orientation to the Loutraki fault segment of the Northern Almopia fault zone, suggests that neotectonic reactivation has taken place. The latest reactivation of this fault zone is normal, and defines a NNW-SSE extension axis (T).

east-west-elongated Mt Kerkini and forms an elongated narrow valley filled with Quaternary fan deposits. It has been characterized as an active fault using stratigraphic and geomorphological data (Psilovikos & Papaphilipou 1990) (Figs 3 and 8a). This is the dominant morphotectonic feature of northernmost Central Macedonia, and consists of two main fault segments named the Kastanousa and Poroia-Petritsi segments, respectively; those both abruptly downthrow the southern slopes of Mt Kerkini (Fig. 8b) forming a right-stepping geometry. The 18 km long Kastanousa segment dips steeply southwards, and, together with smaller antithetic faults, bounds a narrow valley filled with Quaternary proluvial and alluvial sediments dipping gently southwards. To the west the fault joins an ENE-WSW-striking fault that extends towards Lake Doirani. The Kastanousa fault segment contains at least two more subparallel strands towards the centre of the valley, as defined by geophysical surveying along crosssection 1 (Fig. 8a) (G. Vargemezis, unpubli. data). The southernmost of these strands might be considered as the westward extension of the Poroia-Petritsi segment. The small fault scarps, and small parallel exposures of Holocene alluvial sediments that form the east-west strike, imply recent reactivation of the fault. The eastern fault segment, the Poroia-Petritsi segment, is about 24 km long and juxtaposes the Strymon basin sediments against the metamorphic rocks of the Serbo-Macedonian massif that makes up Mt Kerkini. This segment clearly truncates the NE-SW- to ENE-WSW-striking faults that obliquely cut the mountain chain north of Petritsi village. In addition, the River Strymon possibly represents a cross-cutting feature, and consequently a possible barrier to the fault. This fault seems to have been bypassed by the fault reactivation as the fault trace continues eastwards without any deflection or change. The kinematics of the fault zone are well defined along both segments, as indicated in Figure 8c, and define a NNE-SSW extensional stress field. It should be noted that there is an almost complete lack of recent seismic activity along most of this fault. Small-magnitude seismic activity in the wider region is concentrated further south towards the fault edges, close to Lakes Doirani and Kerkini, showing a more or less north-south extension (T) (see Fig. 3). However, few seismological data are available for this fault zone.

Kerkini f a u l t zone

A n t h e m o u n t a s f a u l t zone

The 45 km long east-west-striking Kerkini fault zone runs along the southern slopes of the

This 40 km long zone is one of the most spectacular in Central Macedonia; it bounds the narrow

geometry. The faults define rectilinear mountain slopes, along which Quaternary scree and fan sediments were deposited in the hanging wall, whereas triangular facets characterize the mountain escarpments in the footwall. These faults typically undergo normal reactivation that defines north-south extension (Mountrakis et al. 1996a). The 1932 Sochos earthquake of M = 6 . 2 in this area (Papazachos & Papazachou 2003) was possibly due to reactivation of the Sochos fault zone. Vourvourou f a u l t The Vourvourou fault was described by Tranos (1998) as a c. 15 km long WNW-ESE-striking normal fault that dips steeply N N E and downthrows the mountainous terrain of the Sithonia Peninsula of Chalkidiki at its northern end. The fault exhibits more than one reactivation; the latest one defines a subhorizontal extension axis oriented NNE-SSW. N o r t h e r n A l m o p i a or Aridea f a u l t zone

ACTIVE FAULTS AND STRESS, N GREECE

661

Fig. 8. (a) Detailed mapping of the Kerkini fault zone. Continuous bold line traces the boundary fault; dashed lines indicate the covered or less well-defined faults. (b) Schematic cross-sections of various parts of the fault zone. The Quaternary sediments are coarse-grained fanglomerates, fine-grained fanglomerates and floodplain deposits prograding from the mountain slope towards the basin. (e) Stereographic projection of the latest movement of the fault zone as recorded at the different fault segments and the calculated stress axes, as defined using the program by Duyster (1999). " , cry;A, oh; II, or3.

east-west-striking Anthemountas basin to the south (Mountrakis et al. 1996b) (Fig. 3). The zone can be divided into two fault segments based on the geomorphological features and different lithology of the rocks that it separates. In particular, the segment running from the Thermaikos Gulf eastwards for about 30 km reveals an almost rectilinear strike and separates Neogene sediments of the footwall from Holocene alluvial and coastal deposits of the hanging wall. The other fault segment is about 20 km long and curves concavely northwards. The basement rocks of the mountainous terrain are exposed along this segment, and a Late Pleistocene terrace system, consisting of the older Pleistocene sediments, has formed in the hanging wall. The two segments form a right-step overlapping geometry, with the western one providing the most evidence of recent reactivation, as indicated by the distribution of the small recorded earthquakes. The fault zone possibly extends westwards into the Thermaikos Gulf, and further west may

join the Northern Pieria fault zone (see below). The 1759 earthquake (M ,-, 6.5), reported to have destroyed a large part of the city of Thessaloniki (Papazachos & Papazachou 2003), could be related to reactivation of this part of the Thermaikos Gulf fault zone. However, because information is limited, this is uncertain. The database of the seismological network does not indicate any significant seismic activity along this fault. The fault-slip data of the fault zone define an extensional stress regime with the least principal stress axis (cr3) oriented N N W - S S E (Fig. 9a). This fits with the extension axes as defined by the focal mechanisms of small earthquakes along the zone (Fig. 9b).

N o r t h e r n Pieria f a u l t zone The ENE-WSW-striking Northern Pieria fault zone (Figs 3 and 10a) lies in the northernmost Pieria region and downthrows low-mountainous to hilly terrain over about 20 km. It is a wide zone

662

D. MOUNTRAKIS E T A L .

of faults several kilometres long and dipping steeply northward, thus forming the Aliakmonas basin. Among these, the 10 km long VerginaPalatitsa, Neokastro and Kolindros faults are the most prominent. The faults of this zone affect Mesozoic rocks and Neogene sediments and are marked by continuous or discontinuous linear fault scarps. In addition, several possible fault lines representing concealed faults have been mapped basinwards, e.g. N W of Meliki village and these control the course of the River Aliakmonas. Small landslides have occurred along the faults affecting the Neogene sediments (e.g. Agathia fault). Dip-slip slickenlines along the faults described above indicate that normal reactivation characterizes the fault zone, defining N N W - S S E extension axes (T) (Fig. 10b). These kinematics fit well with the focal mechanisms of the small earthquakes along this zone (Fig. 10c). The Northern Pieria fault zone exhibits morphotectonic similarities to the western part of the Anthemountas zone, suggesting a link, or at least simultaneous evolution.

Western Macedonia Western Macedonia (Fig. 3) lies west of the Voras, Vermio and Pieria mountain chains, and

N

is separated from Thessaly to the south by the River Aliakmonas. This is a mountainous terrain interrupted by the Grevena, Florina, Ptolemais, Kozani-Ag. Dimitrios and Serbia basins. The Grevena basin is filled with molasse-type sediments of the Mesohellenic Trough, whereas all of the other basins contain Neogene and Quaternary sediments. The fault pattern of Western Macedonia differs from that of Central Macedonia and Thrace, because the most prevalent faults of several kilometres length strike not east-west but rather N E - S W to ENE-WSW, cutting the NW-SE orogenic fabric of the Hellenides at high oblique to orthogonal angles. These Late Tertiary strikeslip faults were reactivated as normal faults in the Quaternary (Mountrakis 1983). The fault pattern of Western Macedonia includes NNW-SSEstriking faults of several kilometres that follow the orogenic fabric and bound the Grevena, Kozani-Ag. Dimitrios and Florina basins, without, however, affecting the Quaternary sediments. These faults exhibit a normal reactivation that defines a N E - S W extension axis (T), suggesting that they were mainly reactivated during the Pliocene within the previous N E - S W extensional stress field. The east-west-striking faults exposed in the pre-Neogene basement appear, in conjunction with the NE-SW- to

N

Fig. 9. (a) Stereographic (equal area, lower hemisphere) projection of the fault-slip data of the latest movement of the Anthemountas fault zone and the calculated stress axes oh, ~2, ~3, using the program by Duyster (1999). ~ (0)= 126-82~ cy2(A) =263-05 ~ and c~3( I ) =353-04 ~ (b) Stereographic projection (equal area, lower hemisphere) of the focal mechanisms of the small earthquakes that occurred along the western segment of the Anthemountas fault zone.

ACTIVE FAULTS AND STRESS, N GREECE

663

Fig. 10. (a) Fault map of the Northern Pieria fault zone, (b, c) Stereographic projections (equal area, lower hemisphere) of the latest movement of the zone (b) and the focal mechanisms that occurred along the fault zone (c). ~, shortening (P) axis; D, extension (T) axis.

ENE-WSW-striking faults, to indicate oblique left-lateral normal movement. They also affect the Quaternary sediments forming isolated small faults between the NE-SW- and NNW-SSEstriking faults. In the latter case, they are steepdipping to vertical, indicating normal movement, and possibly originated along similarly striking neotectonic joints (Tranos & Mountrakis 1998). In general, the most numerous faults exposed in Western Macedonia are: (1) the ENE-WSW-striking Aliakmonas fault zone and the nearby east-west-striking ChromioVari, Pontini-Pilori and Feli faults; (2) those of the Vegoritis-Ptolemais fault system; (3) the east-west-striking Ag. Dimitrios (or KoiladaKremasti-Kapnochori) fault. The faults in the Aliakmonas zone were investigated after the 1995 Kozani-Grevena earthquake (see Pavlides et al. 1995; Mountrakis et al. 1996c, 1998; Chatzipetros et al. 1998). The VegoritisPtolemais fault system was described by Pavlides (1985) and by Pavlides & Mountrakis (1987).

Here, we focus mainly on new structural and seismological data. Aliakmonas fault zone

The 7 0 k m long ENE-WSW-striking Aliakmonas zone consists of several subparallel faults that strike E N E - W S W to NE-SW, parallel to the River Aliakmonas (Fig. 3). These faults cut Mts Vourinos and Vermio and extend into Central Macedonia, where they join the ENE-WSWstriking Northern Pieria fault zone. The great length of the Aliakmonas zone and the linking of the recent 1995 Kozani-Grevena M =6.6 earthquake with a reactivation of the fault suggests that this is the most significant fault zone in Western Macedonia. The most important segments are the Rimnio-Kentro, the Serbia-Velventos and the Polifitos-Polidendri faults. Rimnio-Kentro fault segment. This 30 km long ENE-WSW-striking fault dips N N W and affects

664

D. MOUNTRAKIS ET AL.

the Lower-Middle Miocene molasse sediments and the ophiolitic complex of Mt Vourinos. The 1995 Kozani-Grevena earthquake was related to this fault segment. Just after the earthquake many seismic fractures along the fault were observed to cut much younger formations such as the Upper Pliocene-Pleistocene sediments that rest unconformably on the molasse. Thus, the fault has been considered 'seismic' (Mountrakis et al. 1998). It is characterized by rectilinear fault scarps, most clearly seen between Paleochori and Sarakina villages, and intense liquefaction phenomena were also observed in a broader area near Rimnio village (Pavlides et al. 1995; Mountrakis et al. 1996c, 1998; Chatzipetros et al. 1998). The latest reactivation of the fault, as defined by the slickenlines and the sense-of-slip indexes measured along its surface, suggest a normal movement defining a N N W SSE extension axis (T) similar to that defined by the fault-plane solution of the 1995 KozaniGrevena earthquake (Mountrakis et al. 1998). Towards its western end the fault splays into several smaller subparallel faults that mainly dip NNW. Serbia-Velventos fault segment. This is 24 km long, strikes E N E - W S W and dips steeply (c. 6080 ~ NNW, running from Rimnio to Servia village, and also probably extends eastwards as far as Velventos village. The fault forms a 10 km long rectilinear steep mountain slope about 200 m high, along which the Triassic-Jurassic marbles of the Pelagonian zone are separated from the Neogene lacustrine sediments. Because of the abruptness of the mountain slope, talus formations were also deposited, indicating recent vertical movement. Eastwards, the fault is shifted southwards beside Serbia village, where it again forms a rectilinear mountain slope. Here, the fault forms an analogous abrupt fault scarp, along which a cemented tectonic breccia is cut by corrugated slickensides that exhibit striations indicating normal reactivation. However, this fault was not reactivated during the 1995 Kozani-Grevena earthquake (Mountrakis et al. 1998). Polyphytos-Polydendri fault. This 20 km long fault, which forms the easternmost segment of the Aliakmonas fault zone, cuts across the Vermio-Pieria mountain chain and reaches the Pieria fault zone to the east. On a larger scale, the fault west of Polydendri village seems to be subdivided into branches with slightly different orientations. Closely related to the main Aliakmonas zone are three smaller faults, Chromio-Vari, PontiniPilori and Feli, which are found in its hanging

wall. These faults strike east-west and cut Mt Vourinos, forming narrow valleys filled with Plio-Pleistocene sediments (Mountrakis et al. 1996c, 1998). Chromio-Vari is a 16 km long fault zone consisting of two parallel faults that form a right-overlapping geometry. The Pontini-Pilori fault runs close to Pilori and Pontini villages and to the east cuts across Mt Vourinos, forming a tongue filled with molasse-type sediments of Mid-Miocene age. This indicates that the fault is a pre-existing structure, which was reactivated in the Plio-Quaternary. The seismic ruptures along it during the 1995 Kozani-Grevena earthquake confirm a recent reactivation of the fault (Mountrakis et al. 1998). The kinematics of these faults were described in detail by Mountrakis et al. (1998) as normal faults defining a NNW-SSE extensional stress regime similar to that defined by the focal mechanisms of the 1995 Kozani-Grevena earthquake sequence. Vegoritis-Ptolemais fault system

This is 40 km long and consists of an array of NE-SW-striking (40-60 ~ faults that affect the pre-Alpine and Alpine basement, forming the large Neogene and Quaternary VegoritisPtolemais basin (Figs 3 and l la) (Pavlides 1985; Pavlides & Mountrakis 1987). The larger faults of this system are the SE-dipping Nimfeo-Xino Nero-Petra fault and the NW-dipping ProastioKomnina-Mesovouni fault, believed to be the main boundary faults of the depression. Within the depression, other subparallel synthetic or antithetic faults have been mapped (e.g. the Emporio-Perdika, Chimaditis, Peraia-Maniaki and Vegora faults). The Nimfeo-Xino Nero-Petra and ProastioKomnina-Mesovouni boundary faults are the main faults of the system and are described below. Nimfeo-Xino Nero-Petra fault. This is a 30 km long, NE-SW-striking fault that dips steeply SE and bounds the Ptolemais-Vegoritis depression, as it delimits the Neogene and Quaternary sediments and the pre-Alpine and Alpine basement (Mountrakis 1983). It is best exposed between Nymfeo and Aetos villages, where it forms a rectilinear mountain slope up to 400 m high, to the NE it has affected the Neogene sediments of Xino Nero village, depressing them by about 100 m (Mountrakis 1983). Proastio-Komnina-Mesovouni fault. This 30 km long fault zone forms the southeastern borders of the Ptolemais basin and consists of two subparallel fault segments: the 10 km long Proastio

ACTIVE FAULTS AND STRESS, N GREECE

665

Fig. 11. (a) Generalized tectonic map of the Vegoritis-Ptolemais fault system, (b, c) Stereographic projections (equal area, lower hemisphere) indicating (b) the latest movement of the large faults of the fault system and the contemporary stress axes (I~, or1;Ik, cr2;II, or3)and (c) the focal mechanisms that occurred within or near the basin. ~, shortening (P) axis; [B, extension (T) axis. Both projections indicate the same extensional stress field with the least principal stress axis oriented NW-SE. fault and the 20 km long Komnina-Mesovouni fault (Fig. 1l a), both of which dip steeply NW. The Proastio fault mainly affects the Upper Villafranghian Proastion Formation, which consists of conglomerates, forming a rectilinear fault scarp within these sediments. The K o m n i n a Mesovouni fault cutting the Triassic-Jurassic Pelagonian marbles forms a narrow valley filled with Quaternary sediments. The latest formed slickenlines and microstructures, widely accepted as sense-of-shear indicators along these faults, define an extension stress field with a N W - S E least principal stress axis (Fig. l lb). This extension is also indicated by the trend of the extension axes (T), as defined by the focal mechanisms of earthquakes that have occurred within, or close to, the

Vegoritis-Ptolemais fault system (Fig. 11c), e.g. the M = 5.4 earthquake of 9 July 1984.

Ag. Dimitrios or Koilada-KremastiKapnochori fault This is an east-west-striking fault that to the south bounds the Neogene-Quaternary KozaniAg. Dimitrios basin; this constitute the southern part of the larger Ptolemais basin. Steep slope escarpments and a concave mountain slope characterize the fault towards the north, for a length of about 12 km. Recently formed slickenlines along the fault define a normal reactivation with a left-lateral component that exhibits a N N W - S S E to north-south extensional stress field (Fig. 12).

666

D. MOUNTRAKIS E T A L .

Discussion

Both the geological and seismological data indicate that seismic activity in Northern Greece is concentrated along normal faults of several kilometres that stand either alone or as segments of larger fault zones. In Eastern Macedonia and Thrace, the main faults are the east-west-striking Kavala-Xanthi-Komotini, Maronia-Alexandroupolis, Drama-Prosotsani, Serres-Nea Zichni and Ofrinio-Galipsos fault zones. They are typically about 30 km long, except for the 120 km Kavala-Xanthi-Komotini and the 10km Ofrinio-Galipsos fault zones. However, the Kavala-Xanthi-Komotini fault zone is a composite of four fault segments that strike either W N W - E S E or ENE-WSW with lengths of about 30 km. In Central Macedonia, intense recent seismicity has promoted study of the majority of the east-west-striking active faults. However, in the present work, we found that there are some other very important east-west-striking fault zones, such as the Kerkini, Anthemountas and Northern Pieria faults, that might contribute to the recent seismic activity. Several microearthquakes recorded along the Northern Pieria and Anthemountas zones indicate geometry and kinematics similar to the latest movement of these faults, indicating that both fault zones

G

Fig. 12. Stereographic projections (equal area, lower hemisphere) of the latest movement observed along the Ag. Dimitrios or Koilada-Kremasti-Kapnochori fault. The stress field having cr1=165-81~ cr2=264-01 ~ ~3=355-08 ~ has been calculated using the program by Duyster (1999).

should be considered in the seismic hazard assessment of the city of Thessaloniki. In Western Macedonia, the main active faults are the Aliakmonas fault zone and the VegoritisPtolemais fault system, which strike NE-SW to ENE-WSW. The east-west-striking faults that prevail in Central Macedonia and further east are fewer and shorter in Western Macedonia. These are the Feli, Chromio-Vari, Pontini-Pilori and Ag. Dimitrios faults. However, they could also contribute to the seismic activity of the area, as indicated by the 1995 Kozani-Grevena earthquake sequence (Mountrakis et al. 1996c, 1998; Papazachos et al. 1998a). The rupture zones in Northern Greece require re-examination in the light of new data concerning the active faults and the surprisingly low recent seismicity. Thus, we try to estimate the expected magnitude of an earthquake using the scaling law suggested by Papazachos (1989) for the region of Greece. The fault length (L) can be related to the magnitude (M) of the earthquake by the equation. log L = 0 . 5 1 M - 1.85.

(1)

Thus, taking into account the 30 km length of the ruptures in most fault zones, we estimate a maximum probable earthquake magnitude of 6.5. The few strong earthquakes that have been reported in the region of Eastern MacedoniaThrace, for example the 1784 Komotini earthquake (M=6.7), correspond to this rupture length. However, several of the 30km long fault zones of Serres-Nea Zichni and D r a m a Prosotsani consist of fault segments of c. 10 km length, similar to the length of Ofrinio-Galipsos fault zone. In this case, the maximum expected earthquake magnitude is 5.6. Recent results (Papazachos et al. 2006) show that for events in the range M = 6.0-7.5 (including the faults studied here) the surface length is on average 30-50% smaller than the true subsurface length. This indicates that the maximum probable magnitude can be up to 0.4 units larger that estimated from the observed fault length using equation (1). However, the fault structures studied here are well developed and in most cases correspond to faults that have been reactivated several times, thus revealing the total subsurface length on the surface. As a result, application of equation (1), where L represents the observed fault length, should not lead to a systematic underestimation of the maximum probable earthquake magnitude for these faults. The earthquake magnitudes to be expected from reactivation of the Kerkini and Anthemountas fault zones, which consist of segments

ACTIVE FAULTS AND STRESS, N GREECE varying from 18 to 24 km and from 20 to 30 km, respectively, are 6.0-6.3 for the former and 6.2-6.5 for the latter. If the entire 45 km long Kerkini fault zone ruptured, the fault magnitude there could reach 6.8. Such magnitudes are in good agreement with those of several strong historical earthquakes in Central Macedonia (Papazachos & Papazachou 2003), particularly along the Serbo-Macedonian massif during the instrumental era, with the latest being the 1978 Thessaloniki earthquake of M=6.5. The total rupture of the 20 km long Northern Pieria fault zone could produce an earthquake of up to M = 6 . 2 , but ruptures of its 10 km long fault segments are more likely, which would produce earthquakes of magnitude not exceeding 5.6. In Western Macedonia, although the Aliakmonas fault zone and the Vegoritis fault system have been more or less modified by the recent stress regime, these are the main structures along which the recent seismic energy is concentrated. The 1995 Kozani-Grevena earthquake of M = 6.5 was caused by the reactivation and rupture of the 30 km long part of the total 70 km long Aliakmonas fault zone. Taking into account the 30 km length and using equation (1), we estimate a magnitude of 6.5, similar to that of the 1995 earthquake. The Vegoritis-Ptolemais fault system appears to differ from the others described above, in which the strain is localized longitudinally. All the faults in this system exhibit recent activity, similar geometric or kinematic characteristics, and similar morphotectonic features. It thus seems that the strain is not concentrated along the NE-SW-striking boundary faults, but is distributed among most of the faults in the system. The earthquake magnitude to be expected from reactivation of the Vegoritis-Ptolemais faults, whose lengths vary from 10 to 30 km, could be 5.6-6.5.

Conclusions The large fault zones of Northern Greece are the 70 km long Aliakmonas zone in Western Macedonia, the Southern Mygdonia and Kerkini fault zones in Central Macedonia, and the Kavala-Xanthi-Komotini fault zone in Eastern Macedonia and Thrace. These faults bounded and influenced the late Tertiary basins of Northern Greece and show clear evidence of Neogene tectonic movements and, therefore, could be considered as pre-existing large structures. These inherited structures were reactivated as normal faults in Quaternary-Recent times, producing the historical and recent earthquakes.

667

On a very large scale, the fault zones are interlinked and delineate major tectonic structures. One such major tectonic structure is that formed by the linkage of the NE-SW-striking Aliakmonas and the east-west-striking KavalaXanthi-Komotini fault zones, which form en echelon bridge structures through the east-westto WNW-ESE-striking Northern Pieria, Anthemountas, Southern Mygdonia and Sochos fault zones. Another characteristic, although interrupted, structure of the same type is that formed by the Kerkini, Northern Almopia and Vegoritis fault zones. These inherited major structures, along which strike-slip slickenlines have been recorded, could initially represent indent-linked strike-slip faults, similar to those that characterize the brittle deformation of intracontinental regions (Woodcock 1986; Woodcock & Schubert 1994). This hypothesis is supported by the following: (1) late- to post-orogenic brittle deformation related to large strike-slip faults has been reported in Western Macedonia, e.g. in the Mesohellenic Trough (Doutsos 1994; Zelilidis et al. 2002; Vamvaka et al. 2004), in Thessaly (Mountrakis et al. 1993), and in Central Macedonia and Eastern Macedonia-Thrace (Pavlides & Kilias 1987; Tranos 1998; Tranos et al. 1999); (2) most of these fault zones exhibit strike-slip movements that precede normal reactivations (Pavlides & Kilias 1987; Tranos 1998; Tranos et al. 1999; Mountrakis & Tranos 2004; Tranos & Mountrakis 2004); (3) most of the fault zones bound older late Tertiary sediments; (4) they do not represent neotectonic faults related to the contemporary stress regime, as they do not reveal a single uniform reactivation, but are instead inherited structures related to the general fault pattern of the area; (5) the maximum extension defined by the focal mechanisms and the fault-slip data varies even in adjacent areas, and seems to be related to the orientation of the inherited older structures. The stress regime in Northern Greece is extensional with the least principal stress axis (cy3) oriented north-south during the Quaternary. However, a significant variation around this orientation is well known from both earthquake focal mechanisms (Figs 2 and 3) and fault-slip data (Fig. 3). In Eastern Macedonia and Thrace focal mechanisms are too few to precisely define the trend of the extension axes. However, such fault-slip data as have been obtained from the active faults show that the extension in the area trends from NNW-SSE (340 ~ to N E - S W (40~ In Central Macedonia focal mechanisms are more numerous and indicate that the commonest trend of extension is N N W - S S E (c. 355~ although the extension axes (T) reveal a

668

D. MOUNTRAKIS ET AL.

significant variation around this trend even in adjacent areas. This change of trend of extension, as mentioned by Le Pichon et al. (1982) and others (Pavlides 1985; Mercier et al. 1987, 1989; Papazachos et al. 1992, 1998b; Papazachos & Kiratzi 1996; Tranos & Mountrakis 1998), has been confirmed. This swing of the extension in Central Macedonia and Eastern M a c e d o n i a Thrace is also shown by the present fault-slip data, although with a much smaller variation than that obtained using earthquake focal mechanisms. In Western Macedonia, the trends of the extension, as defined by the focal mechanisms and the fault-slip data, are concentrated along N W - S E to N N W - S S E orientations. As a result, a change in the trend from Eastern Macedonia-Thrace to Western Macedonia is well established, but this change concerns the near-stress field imposed by the geometry of the large inherited faults. Although the origin of this change in trend could be attributed to the tectonic stresses related to the arc-shape of the Hellenic subduction zone, we suggest that fault pattern variations as determined by the large inherited fault zones that transect the region of Northern Greece are in fact the most influential factor in this change. A similar swing of the trend of the least principal stress axis (or3) has also been defined in Northern Greece from the neotectonic joints, and has also been attributed to the pattern of fault differentiations (Tranos & Mountrakis 1998; Tranos 1998). This work is part of scientific project 20321 funded by the Earthquake Planning and Protection Organization (Greece). The manuscript was revised in the light of comments by two reviewers.

References ALLMENDINGER, R. W. 2001. FaultKin program. http://www.geo.cornell.edu/geology/RWA. BOURNE, S. J., ARNADOTTIR, T., BEAVAN, J., et al. 1998. Crustal deformation of the Marlborough fault zone in the South Island of New Zealand: geodetic constraints over the interval 1982-1994. Journal of Geophysical Research, 103, 3014730165. BYERLEE,J. D. 1968. Brittle-ductile transition in rocks. Journal of Geophysical Research, 73, 47414750. CHATZIPETROS,A. m., PAVLIDES,S. B. & MOUNTRAKIS, D. M. 1998. Understanding the 13 May 1995 western Macedonia earthquake: a palaeoseismological approach. Journal of Geodynamics, 26(2-4), 327-339. DIAMANTIS, I. B. 1985. Hydrogeological study of the basin of the Vistonida lake. PhD thesis, Demokritos University of Thrace, Xanthi (in Greek with English summary).

DOUTSOS, T. 1994. Late orogenic uplift of the Hellenides. Bulletin of the Geological Society of Greece, 30, 37-44. DtrVSTER, J. D. 1999. StereoNett, version. 2.4. http://homepage.Ruhr-uni-bochum.de/Johannes. P.Duyster/Stereo/Stereo 1.htm. HUBBERT, M. K. 1957. Mechanical basis for certain familiar geologic structures. Geological Society of America Bulletin, 48, 1459-1519. JAEGER, J. C. & COOK, N. G. W. 1979. Fundamentals of Rock Mechanics. Chapman & Hall, London. KARFAKIS, I. & DOUTSOS, T. 1995. Late orogenic evolution of the Circum Rhodope Belt, Greece. Neues Jahrbuch fiir Geologic und Paldiontologie, HS, 305-319. KARISTINEOS, N. K. 1984. Palaeogeographical evolution of the basin of Serres. PhD thesis, University of Thessaloniki. KILIAS, A. A. & MOUNTRAKIS, D. M. 1998. Tertiary extension of the Rhodope massif associated with granite emplacement (Northern Greece). Acta Volcanologica, 10(2), 331-337. KOUKOUVELAS,I. K. & AYDIN, A. 2002. Fault structure and related basins of the North Aegean Sea and its surroundings. Tectonics, 21(5), 10, doi:1029/ 2001TC901037. KYRIAKOPOULOS, K. 1987. A geochronological, geochemical and mineralogical study of some Tertiary plutonic rocks of the Rhodope massif and their isotopic characteristics. PhD thesis, University of Athens (in Greek). LE PICHON, X., ANGELIER, J. & SIBUET, J. C. 1982. Plate boundaries and extensional tectonics. Tectonophysics, 81, 239-256. LYBERlS, N. 1984. Tectonic evolution of the North Aegean trough. In: DIXON, J. E. & ROBERTSON, A. H. F. (eds) The Geological Evolution of the Eastern Mediterranean. Geological Society, London, Special Publications, 17, 709-725. MCKENZIE, D. P. (1978). Active tectonics of the Alpine-Himalayan belt: the Aegean Sea and surrounding regions. Geophysical Journal of the Royal Astronomical Society, 55, 2 t 7-254. MCKENZlE, D. & JACKSON, J. A. 1983. The relationship between strain rates, crustal thickening, palaeomagnetism, finite strain and fault movements within a deforming zone. Earth and Planetary Science Letters, 65, 182-202. MERCIER, J.-L. 1981. Extensional-compressional tectonics associated with the Aegean arc: comparison with the Andean Cordillera of south Perunorth Bolivia. Philosophical Transactions of the Royal Society of London, Series A, 300, 337-355. MERCIER, J.-L., CAREY-GAILHARDIS,E., MOUYARIS, N., SIMEAKIS, K., ROUNDOYANNIS, TH. & ANGHELIDHIS, CH. 1983. Structural analysis of recent and active faults and regional state of stress in the epicentral area of the t978 Thessaloniki earthquakes (Northern Greece). Tectonics, 2(6), 577-600. MERCTER, J.-L., SOREL, D. & SIMEAKIS, K. 1987. Changes in the state of stress in the overriding plate of a subduction zonei the Aegean Arc from

ACTIVE FAULTS AND STRESS, N GREECE the Pliocene to the Present. Annales Tectonicae, 1, 20-39. MERCIER, J.-L., SIMEAKIS,K., SOREL,D. & VERGI~LY, P. 1989. Extensional tectonic regimes in the Aegean basins during the Cenozoic. Basin Research, 2, 49-71. MOLNAR, P. & GIPSON, J. M. 1994. Very long baseline interferometry and active rotations of crustal blocks in the Western Transverse Ranges, California. Geological Society of America Bulletin, 106, 594-606. MOUNTRAKIS, D. 1976. Contribution in the knowledge of the geology of the north boundary of the Axios and Pelagonian zones in the K. Loutraki-Orma area (Almopia). PhD thesis, University of Thessaloniki (in Greek with English summary). MOU~TRAKIS, D. 1983. Structural geology of the North Pelagonian zone s.l. and geotectonic evolution of the internal Hellenides. 'Habilitafion' thesis, University of Thessaloniki (in Greek with English summary). MOUNTRAKIS, D. M. & TRANOS, M. D. 2004. The Kavala-Xanthi-Komotini fault (KXKF): a complicated active fault zone in Eastern MacedoniaThrace (Northern Greece). In: CHATZIPETROS, A. A. & PAVL]DES, S. B. (Eds), 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 2, 857-860. MOUNTRAKIS, D., PSILOVIKOS,A. & PAPAZACHOS,B. 1983. The geotectonic regime of the Thessaloniki earthquakes. In: PAPAZACHOS, B. C. & CARYDIS, P. G. (eds) The Thessaloniki, Northern Greece, Earthquake of June 20, 1978 and its Seismic Sequence. Technical Chamber of Greece, Thessaloniki, 11-27. MOUNTRAKIS, D., KILIAS, A., PAVLIDES, S., ZOUROS, N., SPYROPOULOS, N., TRANOS, M. & SOULAKELLIS,N. 1993. Field study of the Southern Thessaly highly active fault zone. Proceedings of the 2nd Congress of the Hellenic Geophysical Union, Florina, 2, 603-614. MOUNTRAKIS, D., KILIAS, A., PAVLIDES, S., et al. 1996a. Neotectonic map of Greece, Langadhas sheet, scale 1:100 000. Earthquake Planning and Protection Organization and European Centre on Prevention and Forecasting of Earthquakes. MOUNTRAKIS, D., KILIAS, A., PAVLIDES, S., et al. 1996b. Special publication of neotectonic map of Greece, Thessaloniki sheet. Earthquake Planning and Protection Organization and European Centre on Prevention and Forecasting of Earthquakes, Athens. MOWNTRAKIS, D., PAVLIDES, S., ZOUROS, N., CHATZIPETROS, A. & KOSTOPOULOS, D. 1996c. The 13 May 1995 Western Macedonia (Greece) earthquake. Preliminary results on the seismic fault geometry and kinematics. Proceedings of the X V Congress of the Carpatho-Balkan Geological Association, Seismicity of the Balkan Region, 112-121. MOUNTRAKIS,D., PAVE]DES,S., ZOUROS,N., ASTARAS, TH. & CHATZIPETROS, A. 1998. Seismic fault geometry of the 13 May 1995 Western Macedonia (Greece) earthquake. Journal of Geodynamics, 26(2~), 175-196.

669

PAPAZACHOS, B. C. 1989. Measures of earthquake size in Greece and surrounding areas. Proceedings of the 1st Scientific Conference of Geophysics, 19-21 April 1989. Geophysical Society of Greece, Athens, 438--447. PAPAZACHOS, C. B. 1999. Seismological and GPS evidence for the Aegean-Anatolia interaction. Geophysical Research Letters, 26, 2653-2656. PAPAZACHOS,C. B. & KIRATZI,A. A. 1996. A detailed study of the active crustal deformation in the Aegean and surrounding area. Tectonophysics, 253, 129-153. PAPAZACHOS, B. & PAPAZACHOU,C. 2003. The Earthquakes of Greece. Ziti, Thessaloniki. PAPAZACHOS, B., MOUNTRAKIS, D., PSILOVIKOS, A. & LEVENTAKIS,G. 1979. Surface fault traces and fault plane solutions of the May-June 1978 major shocks in the Thessaloniki area, Greece. Tectonophysics, 53, 171-183. PAPAZACHOS,C. B., KIRATZI,A. A. & PAPAZACHOS,B. C. 1992. Rates of active crustal deformation in the Aegean and the surrounding area. Journal of Geodynamics, 16, 147-179. PAPAZACHOS, B. C., KARAKOSTAS, B. G., KIRATZI, A. A., PAPADIMITRIOU,E. E. & PAPAZACHOS, C. B. 1998a. Basic properties of the faulting which caused the 1995 Kozani-Grevena seismic sequence. Journal of Geodynamics, 26, 217-231. PAPAZACHOS, B. C., PAPADIMITRIOU,E. E., KIRATZI, A. A., PAPAZACHOS,C. B. & LOUVARI,E. K. 1998b. Fault plane solutions in the Aegean Sea and the surrounding area and their tectonic implication. Bolletino di Geofisica Teorica ed AppIicata, 39, 199-218. PAPAZACHOS, B., MOUNTRAKIS,D., PAPAZACHOS,C., TRANOS, M., KARAKAISIS, G. & SAWAIDIS, A. 2001. The faults which have caused the known major earthquakes in Greece and surrounding region between the 5th century BC and today. Proceedings of the 2nd Panhellenic Congress of Earthquake Engineering and Engineering Seismology, 28-30 September. Technical Chamber of Greece, Thessaloniki, 1, 17-26. PAPAZACHOS, C., MOUNTRAKIS,D., KARAGIANNI,E., TRANOS, M. & VAMVAKARIS, D. 2004. Stressfield and active tectonics in northern Greece using seismological and neotectonic information, lOth Congress of Geological Society of Greece, 15-17 April, Thessaloniki, Greece, 541. PAPAZACHOS, B. C., KARAKAISIS,G. F., PAPAZACHOS, C. B. t~ SCORDILIS, E. M. 2006. Perspectives for earthquake prediction in the Mediterranean and contribution of geological observations. In: ROBERTSON, A. H. F. & MOUNTRAKIS, D. (eds) Tectonic Development of the Easton Mediterranean Region. Geological Society, London, Special Publications, 260, 689-707. PAVLIDES, S. B. 1985. Neotectonic evolution of the Florina- Vegoritis-Ptolemais basin ( W. Macedonia, Greece). PhD thesis, University of Thessaloniki (in Greek with English summary). PAVE]DES, S. B. & KILIAS, A. A. 1987. Neotectonic and active faults along the Serbomacedonian zone (Chalkidiki, N. Greece). Annales Tectonicae, 1, 97-104.

370

D. MOUNTRAKIS ET AL.

PAVLIDES, S. B. & MOUNTRAKIS, D. M. 1987. Extensional tectonics of northwestern Macedonia, Greece, since the late Miocene. Journal of Structural Geology, 9(4): 385-392. PAVLIDES, S. B. & TRANOS, M. D. 1991. Structural characteristics of two strong earthquakes in the North Aegean: Ierissos (1932) and Agios Efstratios (1968). Journal of Structural Geology, 13, 205-214. PAVLIDES, S., MOUNTRAKIS,D., KILIAS,A. & TRANOS, M. 1990. The role of strike-slip movements in the extensional area of the northern Aegean (Greece). Annales Tectonicae, 4, 196-211. PAVLIDES, S., ZOUROS, N., CHATZIPETROS, A., KOSTOPOULOS, D. & MOUNTRAKIS, D. 1995. The 13 May 1995 western Macedonia, Greece (Kozani Grevena) earthquake; preliminary results. Terra Nova, 7(5), 544-549. PSlLOVIKOS, A. & PAPAPHILIPOU,E. 1990. Pediments, alluvial fans and neotectonic movements of the Mt Kerkini/Belassitsa. Geologica Rhodopica, 2, 95-103. Aristotle University Press, Thessaloniki. ROBERTSON, A. H. F., DIXON, J. E. & BROWN, S., et al. 1996. Alternative models for the Late PalaeozoicEarly Tertiary development of Tethys in the eastern Mediterranean region. In: MORRIS, A. & TARLING, D. H. (eds) Palaeomagnetism and Tectonics of the Mediterranean Region. Geological Society, London, Special Publications, 105, 239-263. TAYMAZ, T., JACKSON, J. & MCKENZIE, D. 1991. Active tectonics of the north and central Aegean Sea. Geophysical Journal International, 106, 433-490. TRANOS, M. D. 1998. Contribution to the study of the neotectonic deformation in the region of Central Macedonia and North Aegean. PhD thesis, University of Thessaloniki (in Greek with extended English summary). TRANOS, M. D. & MOUNTRAK1S, D. M. 1998. Neotectonic joints of Northern Greece: their

significance on the understanding of the active deformation. Bulletin of the Geological Society of Greece, 32(1), 209-219. TRANOS, M. D. & MOUNTRAKIS, D. M. 2004. The Serres Fault Zone (SFZ): an active fault zone in Eastern Macedonia (Northern Greece). In: CHATZIPETROS, A. A. & PAVLIDES, S. B. (eds) 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 2, 892-895. TRANOS, M. D., KILIAS, A. A. & MOUNTRAKIS,D. M. 1999. Geometry and kinematics of the Tertiary post-metamorphic Circum Rhodope Belt Thrust System (CRBTS), Northern Greece. Bulletin of the Geological Society of Greece, 33, 5-16. TRANOS, M. D., PAPADIMITRIOU, E. E. & KILIAS, A. A. 2003. The Thessaloniki-Gerakarou fault zone (TGFZ): the western extension of the 1978 Thessaloniki earthquake (Northern Greece). Journal of Structural Geology, 25, 2109-2123. TURNER, F. J. 1953. Nature and dynamic interpretation of deformation lamellae in calcite of three marbles. American Journal of Science, 251, 276-298. VAMVAKA, A. KILIAS, A. & MOUNTRAKIS, D. 2004. Geometry and structural evolution of the Mesohellenic Trough. A new approach. In: CHATZIPETROS, A. A. & PAVLIDES, S. B. (eds) 5th International Symposium on Eastern Mediterranean Geology, Thessaloniki, Greece, 1,209-212. WOODCOCK, N. H. 1986. The role of strike-slip faults at plate boundaries. Philosophical Transactions of the Royal Society of London, Series A, 317, 13-29. WOODCOCK, N. H. & SCHUBERT, C. 1994. Continental strike-slip tectonics. In: Hancock, P. L. (ed.), Continental Deformation. Pergamon, Oxford, 251-263. ZELILIDIS, A., PIPER, D. & KONTOPOULOS, N. 2002. Sedimentation and basin evolution of the Oligocene-Miocene Mesohellenic basin, Greece. American Association of Petroleum Geologists Bulletin, 86(1), 161-182.

Major active faults of SW Bulgaria: implications of their geometry, kinematics and the regional active stress regime M A R K O S D. T R A N O S 1, V A S S I L I S G. K A R A K O S T A S 2, E L E F T H E R I A E. P A P A D I M I T R I O U 2, V L A D I S L A V N. K A C H E V 1, B O Y K O K. R A N G U E L O V 3 & D R A G O M I R K. G O S P O D I N O V 3

1Department of Geology, School of Geology, Aristotle University of Thessaloniki, GR54124 Thessaloniki, Greece (e-mail: [email protected]) 2Geophysics Department, School of Geology, Aristotle University of Thessaloniki, GR54124 Thessaloniki, Greece 3Seismological Department, Geophysical Institute, Bulgarian Academy of Sciences, Sofia 1113, Bulgaria Southwest Bulgaria is an intracontinental region between the Dinaro-Hellenic and Balkan mountain ranges that has experienced infrequent, but strong and destructive earthquakes. The general geometric and kinematic characteristics of the major faults, mainly the active ones, are investigated, as the seismic activity is insufficient to describe thoroughly the active crustal deformation associated with the faulting. The results suggest a major rupture zone with a length of more than 50 km. The east-west-striking Kochani-KroupnikBansko 'rupture zone' was potentially associated with the large 1904 Kroupnik earthquakes, and has been found to transect the region joining the Kochani, Kroupnik and Bansko faults. In addition, a long-term slip rate ranging from 0.14 to 0.7 mm a-1 has been estimated for some large faults in the region using morphotectonic features. The most active faults are normal ones striking WNW-ESE to ENE-WSW, whereas the NNW-SSE- to NW-SEstriking faults tended to act as barriers to the growth of the former faults, as they do not exhibit much indication of recent reactivation. The stress regime determined is extensional with the least principal stress axis (cy3)subhorizontal and oriented north-south. The fact that the active faults show geometric and kinematic characteristics, as well as estimated long-term slip rates, similar to those of the active faults of central and eastern Macedonia and Thrace (Northern Greece) suggests that both of these regions share a single contemporary stress field. Abstract:

Intense seismic activity in SW Bulgaria is revealed by both historical information and instrumental records. The area has suffered severe damage caused by the 1904 Kroupnik earthquake (M = 7.8), which was preceded by a strong foreshockjust 23 minutes before (Shebalin et al. 1974). For the main shock, which is considered one of the largest events in the South Balkan Peninsula, a magnitude up to 7.8 has been estimated (Christoskov & Grigorova 1968). However, a recently re-estimated magnitude yields a much smaller value of 7.0-7.2 (Pacheco & Sykes 1992; Dineva et al. 2002). Although the seismic properties of the area have been investigated (Dineva et al. 1998; Rizhikova et al. 2000), the seismicity during the instrumental era is as yet insufficient to define the properties of the major rupture zones of the region. This is because the strongest events are infrequent and smaller magnitude seismicity is diffuse. The most reliable seismological information concerning the study

area includes the focal mechanisms of small earthquakes along the Strouma River (Van Eck & Stoyanov 1996), and isoseismals of the large events of the Kroupnik earthquake sequence (Grigorova & Palieva 1968; Shebalin et al. 1974; Papazachos et al. 1997). In general, the former determine a north-south extension, and the latter the existence of large rupture zones striking eastwest. Studies based on global positioning system (GPS) measurements during the period 19961998 (Kotzev et al. 2001) indicate that SW Bulgaria, i.e. mainly the area along the Strouma River, is dominated by N N W - S S E extension. This supports the view that the North Aegean extensional regime extends to Bulgaria, as suggested by Burchfiel et al. (2000). Geological information about the faulting in SW Bulgaria mainly refers to the fault network of the region (Zagorchev 1992a,b) and the N E - S W striking Kroupnik fault, which is considered to be associated with the 1904 Kroupnik earthquake

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. TectonicDevelopmentof the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 671-687. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

672

M.D. TRANOS E T AL.

(Zagorchev 1992a; Shanov 1997; Shanov & Dobrev 2000; Meyer et al. 2002). According to Zagorchev (1992a,b), the NNW-SSE-striking faults forming the Strouma fault system and smaller kilometres-long faults orthogonal or oblique to the Strouma River dominate the fault network of the region. However, more recent investigation of the faulting in the region indicates that the faults that trend orthogonally or obliquely to the Strouma River, i.e. those striking NE-SW, ENE-WSW, east-west and W N W ESE, are more prevalent than NNW-SSEstriking ones (Tranos 2004). As a result, several differently oriented faults could be considered as rupture zones activated under the contemporary stress regime in the region, as well as for the Kroupnik earthquake. In addition, although the contemporary stress regime is extensional, arguments exist about the orientation of the least principal stress axis (or3) as derived from fault-slip data analysis (Shanov 1997; Shanov & Dobrev 2000; Shanov et al. 2001; Tranos 2004). The purpose of this paper is to examine in detail the major faults of the region, mainly those that bound Quaternary basins in order to identify their geometry and kinematics, and obtain a better understanding of the fault activity as well as the active stress regime of the study area.

Geological and seismotectonic setting of the region Southwest Bulgaria is a region where the N N W SSE-trending Dinarides-Hellenides mountain belt and the east-west-striking Rhodope and Balkan orogen join (Fig. 1, inset map). In this region, several Alpine tectonic zones forming the inner part of the mountain chain are cut by the Strouma Lineament, as inferred by Bonchev (1958, 1971) and Jaranoff (1960), and described in detail by Zagorchev (1992b). An additional and more complete description of the geology of the area was given by Zagorchev (2001). The study area extends from Kresna and Bansko in the south to Kyustendil and Doupnitsa in the north (Fig. 1). The geology of the area comprises a pre-Alpine and Alpine basement consisting of: (1) Pre-Palaeozoic and Palaeozoic high-grade metamorphic rocks that belong to the Rhodopian Supergroup and the Ograzhdenian Supergroup, or to the Serbomacedonian massif; (2) a few relatively small outcrops of Mesozoic rocks; (3) Palaeogene sediments that include Middle Eocene to Lower Oligocene and Upper Oligocene continental elastic sequences.

The Neogene sediments are mainly exposed in basinal parts of the area and have been grouped into three depositional cycles (Nedjalkov et al. 1988): (1) the Late Badenian-Sarmatian cycle, which consists of red polymict conglomerates, siltstones and sandstones with clay interbeds; (2) the Meotian-earliest Pontian cycle, which consists of whitish or yellowish alluvial sand and clay, interbedded with pebble gravel lenses; (3) the Pontian-Pliocene cycle, characterized by well-sorted conglomerates and sandstones. Finally, Quaternary proluvial and alluvial sediments were formed along the NNW-SSE Strouma River and its tributaries, as well as along the ENE-WSW-striking fault bounded by the Doupnitsa and Kyustendil basins and the WNW-ESE Bansko-Razlog basin. Since the Late Miocene the region appears to have experienced crustal extension, implying that it might represent the northernmost part of the Aegean extended domain (e.g. Jackson & McKenzie 1988; Burchfiel et al. 2000; Tranos 2004). The neotectonic movements succeeded and destroyed the planation surfaces of the initial peneplain formed in the Early-Mid-Miocene (Zagorchev 1992b). Recent focal mechanisms of small earthquakes (Van Eck & Stoyanov 1996), geodetic measurements (Kotzev et al. 2001) and fault-slip data (Tranos 2004) indicate that the mean extension axis in SW Bulgaria is oriented north-south, although a WNW-ESE extension of the area was previously suggested (Shanov 1997; Shanov & Dobrev 2000; Shanov et al. 2001). The fault network of the area is complicated, comprising many inherited faults from the late orogenic stages (Zagorchev 1992a,b; Tranos 2004). Three main fault orientations can be distinguished: NNW-SSE-striking faults that follow the orogenic fabric of the Dinarides and Hellenides; NE-SW- to ENE-WSW-striking faults orthogonal to the previous fabric; W N W ESE-striking faults that follow the Balkan and Rhodope structural fabric. These orientations include faults that could have experienced recent reactivation. In particular, Van Eck & Stoyanov (1996) mentioned that the faults along the Strouma River could have undergone recent reactivation as their lengths correspond to the strong earthquakes that have occurred in the area. Meyer et al. (2002) suggested that the 1904 Kroupnik earthquakes were related to reactivation of the Kroupnik fault, which strikes NE-SW. Historical destructive earthquakes of M > 6.0 are known to be associated with the Kroupnik (1866, M 6.7) and Kyustendil (1641, M 6.7) faults (Papazachos & Papazachou 2003). In particular,

ACTIVE FAULTS OF SW BULGARIA

673

Fig. 1. Generalized tectonic map of SW Bulgaria and neighbouring Former Yugoslav Republic of Macedonia (F.Y.R.O.M.) indicating the large rupture zones and faults as compiled from Landsat satellite imagery, geological maps (Marinova & Zagorchev 1990a-d) and field observations. In the middle, the trace of the Kochani-Kroupnik-Bansko rupture zone is shown on the Landsat imagery with white arrowheads. The map at the lower right indicates the study area in the Balkan Peninsula with respect to the large alpine orogenic structures (HSZ, Hellenic subduction zone; NAT, North Anatolian Trough; SL, Struma Lineament). Ko. F, Kochani fault; Kr. F., Kroupnik fault; Krs. F., Kresna fault; G-P. F, Gradevo-Predela fault; B. F., Bansko fault; B1. F., Blagoevgrad fault; St. F., Stob fault; Do. F., Dobrovo fault; Sa. F., Saparevo fault; Ky. F., Kyustendil fault. the seismicity associated with the Kroupnik fault is characterized as the highest in Bulgaria (Ranguelov et al. 2001). Recent seismicity of

M _>4.0 since 1964 (Fig. 1) is mainly concentrated along an E N E - W S W elongated zone, extending from the area west of the B u l g a r i a n - F Y R O M

674

M.D. TRANOS E T AL.

border toward the villages of Simitli and Kroupnik and the area east of Blagoevgrad (Karakostas et al. 2004).

Large faults of SW Bulgaria Our investigation is based on the use and interpretation of Landsat imagery to define the dominant fault pattern of the area and consequently the large fault zones. This interpretation shows that the NE-SW- to ENE-WSW- and W N W ESE- to east-west-striking faults are the most prominent fault lineaments in the region. The former are curved or rectilinear lineaments forming a large-scale anastomosing fault network of varied width, similar to that in a mountainous area NW of the Kroupnik-Simitli basin (Fig. 1). On the other hand, the W N W ESE- to east-west-striking faults are commonly lineaments of larger length and spacing, i.e. Kochani and Bansko, that breach shorter and more diffuse ENE-WSW-striking faults, forming large rupture zones of arcuate shape with an east-west general strike. The well-defined eastwest-striking Kochani-Kroupnik-Bansko 'rupture zone' has a total length of > 100 km. This rupture zone joins faults that dip north, such as the Kochani, Kroupnik, Gradevo-Predela and Bansko faults. A concavity to the north and an abrupt bend between the NE-SW-striking Kroupnik fault and the WNW-ESE-striking Bansko fault also characterize this zone. This bend seems also to be common for the western part of the Rila Mt, as both WNW-ESE- to eastwest- and NE-SW-striking faults could be traced there. The recent seismic activity shown in Figure 1 is distributed along the Kochani-KroupnikBansko rupture zone and also delineates this bend. In addition, the isoseismals of the 1904 earthquake (Shebalin et al. 1974) fit better with the orientation and length of this zone than with the N E - S W Kroupnik fault alone. Northwards, several discontinuous lineaments of N E - S W to E N E - W S W strike have been traced from the eastern part of the Kochani basin towards Bulgaria to form a horsetail-type structure or fan, opening eastwards. The discontinuity of these lineaments can be ascribed to persistent NNW-SSE, inherited Alpine structures (Fig. 1). In addition, Landsat imagery has allowed us to recognize that large Neogene and Quaternary basins, such as the Doupnitsa, Kyustendil, Kocherinovo-Rila-Stob (Rilska-reka), Kroupnik and Bansko-Razlog basins, have been oriented along both NE-SW- to ENE-WSW- and WNW-ESE-striking faults. For this reason, we attempted to define the geometric and kinematic

characteristics of these boundary faults and their kinematics. For the description of fault kinematics, we use the shortening (P) and extension (T) axes, as defined by Marett & Allmendinger (1990), as these axes are analogous to the P-T axes of the fault-plane solutions of the earthquakes. Thus, a direct comparison between the geological and seismological data could be achieved. Additionally, in the cases where we used the Holocene period for the estimation of the long-term slip rate of the faults, we used its nominal age, i.e. 10 ka. The geometric and kinematic characteristics of the faults investigated are described as follows. Kroupnik fault

This ENE-WSW- to NE-SW-striking fault has brought into contact basement rocks and the Neogene and Quaternary sediments that filled the Simitli-Kroupnik basin (Figs 1, 2a and 3a), and has been studied by several workers, e.g. Zagorchev (1970, 1975), Vrablianski (1974) and Vrablianski & Milev (1993), and more recently by Shanov & Dobrev (2000) and Meyer et al. (2002). Zagorchev (1970) described it as a normal fault that strikes N025 ~ to N040 ~ and dips at 40-70 ~ to the WNW. However, this description reflects the geometry of the easternmost part of the Kroupnik fault, more than that of the entire fault system that bounds the basin. Our observations, taking into account the strike, suggest that the fault is subdivided into two segments (Fig. 2a): an eastern one with N E - S W (c. N045 ~ strike, which is exposed east of the Strouma River, throwing the Neogene sediments against basement rocks; and a western one that strikes ENE-WSW (N065-070 ~ and is mainly exposed west of the Strouma River, juxtaposing basement rocks and rocks as young as Holocene proluvial deposits. The latter fault, as traced by the juxtaposition of the sediments against the basement rocks, appears to be shifted c. 200 m towards the south, possibly as a result of NNW-SSE-striking faults that form the Strouma River (Fig. 2a, point A). The same shift is also evident on the map presented by Dobrev & Kostak (2000). East of the Strouma River, the ENE-WSW-striking fault continues for about 1 km, where it truncates a NE-SW-striking fault (Fig. 2a, point B). Although the trace of both fault segments is easily recognizable in the landscape, the strike deflections and the cross-cutting NNW-SSE-striking faults might act as barriers where high stress levels could be concentrated. The 1904 Kroupnik earthquake was associated with many rock fall and landslide phenomena (Dobrev & Tacheva 2000) and the formation

ACTIVE FAULTS OF SW BULGARIA of a fault scarp (Meyer et al. 2002). Meyer et al. also suggested that the Kroupnik fault is an active normal fault exhibiting uniform reactivation of the order of c. 0.1 mm a -1 slip rate during the last 6 Ma. Shanov (1997), Shanov & Dobrev (2000) and Shanov et al. (2001), from their faultslip data analysis carried out along the Kroupnik fault, suggested that the stress regime is extensional with the least principal stress axis (or3) subhorizontal and oriented W N W - E S E (c. 280290~ However, their conclusions are mainly based on the dip-slip slickenlines collected along the NNE-SSW to N E - S W fault slickensides that affect the basement rocks and belong to the eastern NE-SW-striking fault segment that separates the basement from the Neogene sediments (Shanov & Dobrev 2000, pp. 120-121, points 2, 3 and 4). The fact that the N N E - S S W to NE-SWstriking faults are inherited structures that were previously reactivated as normal faults by W N W - E S E extension during the Mid-Late Miocene (Tranos 2004), raises questions about whether the fault-slip data reflect the contemporary stress regime. In any case, the W N W - E S E extension determined by Shanov & Dobrev (2000) does not match the north-south extension determined by focal mechanisms (Van Eck & Stoyanov 1996) and GPS measurements (Kotzev et al. 2001). Therefore, we argue that the dip-slip slickenlines that indicate a normal reactivation of the ENE-WSW- to east-west-striking fault slickensides are more reliable indicators of the contemporary regional stress field, as their reactivation, owing to older deformation events, is that of oblique or strike-slip faults (Tranos 2004). Furthermore, their kinematics, which defines a subhorizontal extensional kinematic axis (T) and slip vectors along the N N W - S S E to north-south orientation (Fig. 4a), suggests an extension orientation similar to that defined from the GPS measurements and focal mechanisms.

675

alluvial sediments of the Rilska River, and is characterized by a relative asymmetry, as the youngest deposits are parallel and close to the SW edge of the depression. At its N W edge, it forms a rectilinear terrace > 40 m high, whereas at its SW edge it is bounded by the NE-SWstriking Stob fault, which affects pre-Alpine crystalline rocks and Neogene sediments. This fault has been considered as active by Zagorchev (1975), because it is very well defined in relief. Although the strike of the fault and the plateau front indicate the prevalence of a N E - S W strike, several physiographic features also indicate recent activity along faults that strike E N E WSW. In particular, c. 2 km SW of Rila village, along the Stob fault, we have observed a rectilinear shutter ridge that forms an apparent rightlateral shifting of the proluvial fan (Fig. 3c). In addition, rectilinear trench-rills and landslides have formed along the E N E - W S W orientation in the area upslope of Stob village, where erosion of Neogene sediments forms the 'Stobski Piramidi' geomorphological features. At the same location (Fig. 2b, site A), an ENE-WSW-striking fault dipping at high angles towards the N N W has been found to displace not only the Neogene sediments but also the overlying Pleistocene gravelly silty-sandy deposits of the Badino Formation by about 1 m, forming a colluvial wedge that is potentially Holocene in age (Fig. 3d). Assuming that this colluvial wedge resulted from a palaeo-earthquake event, then a vertical slip of about 1 m can be estimated for the Holocene period. Taking an average fault dip of 45 ~ for the Stob fault, a value that typifies active normal faults in the seismogenic crust (e.g. Jackson & White 1989) and the nominal age of the Holocene, i.e. 10 ka, a long-term slip rate of c. 0 . 1 4 m m a -1 since the Holocene can be suggested.

Saparevo f a u l t Stob f a u l t A 10 km long, NE-SW-striking narrow depression into which the Rilska River flows is the dominant topographic feature of the area, in which the towns of Kocherinovo, Stob and Rila are located (Figs 1 and 2b). This depression, named Rilska-reka, intersects a well-defined plateau of mainly Neogene sediments onto which Early Pleistocene talus-proluvial deposits of the Badino Formation (Zagorchev 1992b) and Pleistocene alluvial sediments have been deposited, forming well-extended pediments. The depression is filled with Pleistocene and Holocene

This ENE-WSW-striking fault, here named the Saparevo fault, bounds the graben east of Doupnitsa city (previously Stanke Dimitrov) until Sapareva Banya village and then aligns itself with the northern mountain front of the Rila mountain range (Figs 1 and 2c). The fault was first described as an active one by Jaranoff (1960), who mentioned that it strikes N065 ~ and is associated with thermo-mineral springs. Vrablianski (1977) named it the Klisoura fault, whereas Zagorchev (1969) referred to it as the Saparevo fault. This fault affects the large, anastomosing fault zones that dip at medium angles

676

M . D . TRANOS E T AL.

. ,,...~

,~2"~

~<~

~

o,-~

~a-k .az~

ACTIVE FAULTS OF SW BULGARIA

677

678

M . D . TRANOS E T AL.

Fig. 3. Field photographs of the large fault zones exposed in SW Bulgaria. (a) Kroupnik fault: a general view (towards the SE) of the fault east of Kroupnik village and the Strouma River. (b) Detail of the surface of the fault (site A, Fig. 2) showing the dip-slip striations that indicate the normal reactivation of the fault. (c) Stob fault: a general view (towards the south) of the fault, SW of Rila village, where the right-lateral offset of the Holocene alluvial fan deposits is clearly observed. (tl) Detailed view of the (?)palaeo-earthquake fault exposed in the right abutment of the fan shown in (c), arrow indicates the trace of the fault. (e) Saparevo fault: a general view (towards the south) of the fault, west of Resilovo village. (f) Saparevo fault: abrupt vertical surface offsets within the Holocene proluvial sediments along the Saparevo fault close to Ovchartsi village. (g) Kyustendil fault: view of the two branches of the Gurlyano-Skakavitsa fault south of Gurlyano village (view towards the ESE). The lower branch affects the Holocene proluvial deposits.

ACTIVE FAULTS OF SW BULGARIA

679

Fig. 3. (h) Dobrovo fault: a very well-formed fault scarp in the western edge of Dobrovo village (view towards the west). (i) Close-up view (towards the ESE) of the Dobrovo fault cutting the Frolosh Formation (Dobrovo-Skrino road). Dashed white lines indicate the fault zone and white arrows the normal sense of shear. (j) Predela-Bansko fault: view of the fault (towards the SSE) in the slopes south of Bansko town. White arrows indicate successive points on the fault trace.

680

M.D. TRANOS E T AL. N

N

N

Kroupnik

Saparevo

Kyustendil

N

N

N

Dobrovo

Bansko

Gradevo-Predela

a

N

Kresna

N

Blagoevgrad

Fig. 4. Stereographic projections (equal area, lower hemisphere) indicating the latest kinematics of the large faults exposed in SW Bulgaria: (a) Kroupnik fault; (b) Saparevo fault; (e) Kyustendil fault; (d) Dobrovo fault; (e) Bansko fault; (f) Gradeno-Predela fault; (g) Kresna fault; (h) Blagoevgrad fault. Filled squares and diamonds indicate the extension (T) and shortening (P) axes, respectively, from the fault-slip data. Slip vectors are shown by arrows.

towards the N W to west and that, according to Shipkova & Ivanov (2001), form part of the Djerman detachment fault. These fault zones are the dominant structures that affect the basement rocks of the Rila Mt in its W N W part (i.e. between Sapareva Banya and Rila towns) and are related to a W N W - E S E extensional stress regime during the Mid-Late Miocene (Tranos 2004).

The Early Pleistocene talus and proluvial sediments of the Badino Formation that have been deposited along the escarpments of this part of the mountain (SSW of Bistritsa village) were entirely cut by ENE-WSW-striking fault surfaces that dip at steeper angles towards the N N W and belong to the Saparevo boundary fault. As a result of this retreat, younger Pleistocene and

ACTIVE FAULTS OF SW BULGARIA Holocene proluvial and alluvial sediments were deposited along the mountain front (Fig. 3e). However, the most recent reactivation of the mountain front occurs along faults that strike E N E - W S W to east-west, resulting in further retreat of the NE-SW-striking mountain front. This is well observed west of Bistritsa village (Fig. 2c), where successive fault scarps striking N70-80 ~ within the Pleistocene and Holocene proluvial and alluvial sediments retreat to the preceding N E - S W boundary fault. A retreat of the mountain front as a result of an active stress regime has been also observed in an escarpment south of Ovchartsi village (Fig. 2c, site A). There, proluvial sediments, dated as young as Holocene (Marinova & Zagorchev 1990a), have been deposited against the low-dipping N N W boundary fault surface and form a very gentle, northdipping slope. In the hanging-wall part and close to the boundary fault, steeply inclined to almost vertical ENE-WSW- to east-west-striking faults cut these proluvial sediments, forming steeply inclined rectilinear fault scarps (Fig. 3f). The length of these fault scarps is a few tens of metres, whereas they cause a vertical offset of about 3.8M m to the very gentle, north-dipping slope formed by the Holocene sediments. Taking into account the fault dip and the Holocene age, as for the Stob fault, we estimated a long-term slip rate of c. 0.6 mm a -1 for the Saparevo fault since the Holocene. Moreover, the westward extension of these ENE-WSW- to east-west-striking faults is recognized as far as Resilovo village, where an abrupt fault scarp of east-west-strike is also well observed, and separates the Holocene proluvial deposits from the older Neogene and Palaeogene sediments (Fig. 2c, site B). Thus, the overall observations along the Saparevo boundary fault suggest that the most recently reactivated faults are ENE-WSW- to east-west-striking faults that lead to a retreat of the mountain front along the Saparevo boundary fault. The kinematics of the Saparevo fault was defined from fault surfaces exposed in basement rocks in several places. Observed slickenlines suggest normal to oblique right-lateral normal movements driven by a north-south extension axis and a subvertical shortening axis (P) (Fig. 4b).

681

(Figs 1 and 2d). Jaranoff (1960) mentioned that in this area faults striking E N E - W S W to eastwest, segments of the Kyustendil fault, continue to be active, and some of them are associated with the hottest thermo-mineral springs of Bulgaria. Jaranoff also stated that these faults are shorter than the longitudinal NW-SE-striking faults, i.e. the faults parallel to the orogenic fabric, and in several cases, the latter bound the former. The Kyustendil fault can be subdivided into two segments (Fig. 2d): the GurlyanoSkakavitsa fault segment in the west and the Gurlyano-Kyustendil fault segment in the east. The former dips at very high angles towards the NNW, forming a very steep rectilinear mountain front. Along this mountain front, Holocene proluvial and alluvial sediments (dating according to Marinova & Zagorchev 1990c) were deposited against the gneissic granite that constructs the mountain (Fig. 3g). The eastern fault is traceable from Gurlyano village to the city of Kyustendil and exhibits a complicated geometry, as it consists of several fault branches that strike either E N E - W S W or east-west. In particular, east of Gurlyano village the mountain front is aligned east-west, as a result of the presence of east-west-striking inherited vertical discontinuities that behave as short fault bridges. Further east, to the city of Kyustendil, two distinct ENE-WSW-striking right-stepping fault branches, partly overlapped, have been observed; these diminish gradually towards the northern slopes of the mountain. The kinematics of the Kyustendil fault zone is well defined along the very steep slopes east of Gurlyano village, where the ENE-WSW-striking slickensides of the fault are well exposed; dip-slip slickenlines indicate an extensional strain field with the extension axis (T) almost horizontal along the N N W - S S E orientation (Fig. 4c). In addition, in the abutment east of Gurlyano village (Fig. 2d, site A) two rectilinear successive and almost vertical faults displace Holocene proluvial sediments and form a steeply dipping fault scarp with a total vertical offset of about 4.7-5 m. As for the previous faults, we estimated the long-term slip rate of the Kyustendil fault as c. 0.7 mm a -1 since the Holocene.

Dobrovo fault Kyustendil fault The Kyustendil fault is located west of Kyustendil city along the northern front of Mt Osogovo, and forms a 50 km long, E N E - W S W to east-west narrow basin filled with Pleistocene and Holocene proluvial and alluvial sediments

This is a c. 10 km long WNW-ESE-striking fault that cuts the mountainous terrain of Mt Vlahina along the course of the Strouma River and southwards (Figs 1 and 2e). The fault is better observed from the villages of Skrino and Dobrovo towards the Strouma River, where for about 6 km it is

682

M.D. TRANOS ET AL.

quite rectilinear, dipping NNE at high angles (Fig. 3h and i). In particular, at the western edge of Dobrovo village, the fault throws Pleistocene reddish proluvium pebble to cobble sandy gravel against basement rocks (Fig. 3h). In addition, between Dobrovo and Skrino villages the fault cuts the Quaternary erosion mantle material at high angles, as well as NNW-SSE- and N N E SSW-striking faults that dominate the broader area and control the flow of the Strouma River. The exposed slickensides of the fault there exhibit dip-slip slickenlines that define a subhorizontal NNE-SSW extension axis (T) and subvertical shortening axis (P) (Fig. 4d).

and Predela villages, is an area where the Kochani-Kroupnik-Bansko rupture zone bends (Fig. 1). Here, the east-west-striking faults are the most prevalent in the fault network and have lengths from about 2 km to 10 km (Fig. 2g). In particular, three distinct fault strands tend to bridge the NE--SW-striking Kroupnik fault and the WNW-ESE-striking Bansko fault. The fault surfaces that dip towards the north and belong to the northern strand, namely the GradevoPredela fault, exhibit slickenlines and microstructures that suggest normal kinematics. This is similar to the kinematics of the Bansko fault and defines a NNE-SSW subhorizontal extensional kinematic axis (T) (Fig. 4t).

Bansko fault This is a 30 km long WNW-ESE-striking fault that dips at high angles to the NNE, diminishes towards the northern slopes of the Mt Pirin and forms the wedge-shaped Bansko-Razlog graben (Fig. 1). The fault consists of two segments that form a left-stepping geometry around Bansko town, possibly because of the existence of N N E SSW-striking faults that act as barriers. The western fault segment is rectilinear and forms very abrupt steep slopes and a mountain front (Fig. 2f). Beyond Bansko town the mountain front shifts c. 2 km to the NNE, where the eastern fault segment is defined by a similar WNW-ESE strike for c. 14 km. At the base of the mountain front of the western fault segment (Fig. 3i), a dense vegetation cover in conjuction with soft-weathering Holocene proluvial and drift fan deposits (sands and unsorted pebble to boulder gravels) inhibit the recognition of fresh fault surfaces. However, in the upslope direction we have found parallel fault slickensides within marbles of the Rhodopian Supergroup that exhibit dip-slip slickenlines. The latter define a NNE-SSW extensional kinematic axis (T) and a steeply inclined shortening kinematic axis (P) (Fig. 4e). Some doubtful observations to the SW of Bansko town (Fig. 2f, site A) concern a few small rectilinear north-facing scarps that set down the hills made of granite and superficial Holocene sandy proluvium. This feature may indicate recent fault activation. Along these scarps, small landslide phenomena affect the sandy proluvium and the soil developed on it.

Gradevo-Predela fault The hilly to low-mountainous terrain that separates the Simitli-Kroupnik and BanskoRazlog basinal areas, i.e. the area of Gradevo

Kresna fault In the area of Kresna, the regional topography is dominated by the faults striking NNW-SSE to NW-SE. Several of these faults, and particularly those exposed SW of Kresna, have been described by Zagorchev (1970, 1971) as the 'Gradeshka fault zone', whereas some other faults, such as those east of Kresna village, were described by Moskovski & Georgiev (!970) as the Kresna fault. These faults have been recorded on the geological map (Razlog sheet) of Bulgaria (Marinova & Zagorchev 1990d) and their lengths vary between 6 and 12 km. They dip at high angles, towards either the NE or SW, and form rectilinear or anastomosing zones that control the course of the Strouma River and its tributaries. The most well-traced one dips towards the SW and bounds the Pontian-Pliocene upper sandstone-conglomerate member of the Kalimanci Formation (Marinova & Zagorchev 1990d) in the area ENE of Kresna village. Along several fault slickensides striking NNW-SSE to NW-SE, we observed two generations of slickenlines, the chronological order of which has been well established using overprinting criteria. The older slickenlines correspond to a prevalent strike-slip movement, whereas the later ones correspond to a normal movement that is compatible with a NE-SW-trending extension axis (T) (Fig. 4g). The extension axis (T) deviates significantly from the north-south extension defined by the GPS measurements and the focal mechanisms, and fits well with the NE-SW extension that governed the area during the Late Miocene-Pliocene (Tranos 2004). This last feature suggests that these faults are not optimally oriented parallel to the contemporary stress regime and therefore their reactivation was not enhanced.

ACTIVE FAULTS OF SW BULGARIA Blagoevgrad fault

The approximately 12 km long ENE-WSWstriking Blagoevgrad fault (first reported by Vrablianski 1977) is a linear structure that affects the Ograzhdenian gneissic basement and dips at high angles towards the SSE. It forms a narrow valley that reaches the northern suburb of the city of Blagoevgrad, where it seems to control the deposition of Quaternary sediments relative to Neogene sediments (Fig. 1). Dense vegetation cover along the valley prevents the exposure of fresh fault surfaces; however, the few exposed ones indicate that the Blagoevgrad fault dips at about 60-70 ~ towards the SSE and its kinematics is characterized by an oblique right-lateral normal movement that defines a NNW-SSE extension axis (T) (Fig. 4h). Some smaller antithetic fault surfaces have been also found and indicate normal slickenlines.

Active stress regime Burchfiel et al. (2000) considered the region of Bulgaria as the northern end of the Aegean extensional province. They suggested that during latest Pliocene and Quaternary time Bulgaria and Northern Greece were affected by somewhat different extensional regimes, involving northsouth extension in Bulgaria and both northsouth and NE-SW extension in Northern Greece; a complex of strike-slip and extensional faults in SW Bulgaria could represent a diffuse boundary between the two regimes. The contemporary stress regime of SW Bulgaria, and consequently the present tectonic framework, is not as much investigated as that of Northern Greece. Controlling factors could include the Hellenic subduction zone, the southwestward motion of the Aegean microplate (McKenzie 1972; Jackson 1994; Papazachos 1999), the right-lateral strike-slip of the North Anatolian Fault and the large ~ - S S E striking Strouma (Kraigtid) Lineament formed in Tertiary times (Zagorchev 1992b). The Hellenic subduction zone has been well established to dominate the Greek region (Mercier et al. 1989), whereas the influence of the North Anatolian Fault, including in southern Bulgaria, was discussed by Pavlides et al. (1990), and more recently by Burchfiel et al. (2000) and Koukouvelas & Aydin (2002). Burchfiel et al. (2000) suggested that the north-south extension in central Bulgaria increased in intensity as a result of the propagation of the North Anatolian Fault system into the southern Balkan region during the last 4 Ma, and that the PlioceneQuaternary faults, which parallel the North

683

Anatolian Fault (i.e. the faults striking ENEWSW), may have both normal and dextral components. Burchfiel et al. also considered that that the NW-SE-striking faults and the NE-SW-striking faults bound mountains and adjacent valleys, implying recent reactivation. Koukouvelas & Aydin (2002) also considered the North Anatolian Fault as the first-order structure that produced a basin-and-range type morphology in the area north of the North Aegean Trough. The Strouma Lineament is an inherited structure with repeated reactivation and a length of > 800 km (Zagorchev 1992b). Its length is rather overestimated, and it is questionable whether it could act as a boundary element to the regional stress regime. Accordingly, Van Eck & Stoyanov (1996) suggested that the large magnitude earthquakes of the region could relate to the many fault segments of the Strouma Lineament, with lengths ranging between 15 and 50 km. A detailed knowledge of the kinematics of plate motions and interpolate deformations is necessary to help constrain forces responsible for deformation. In particular, in SW Bulgaria, where the available GPS measurements (Kotzev et al. 2001) and the seismological data (Van Eck & Stoyanov 1996) are insufficient to establish the active deformation pattern, a geological approach that provides geometric and kinematic information on crustal deformation is very helpful. In addition to the geometry and kinematics of the main faults exposed in SW Bulgaria, one of the most important issues of active deformation is the definition of the contemporary stress regime. For this purpose, we utilized the stress ellipsoid by applying the algorithm of Carey & Brunier (1974) to the youngest-formed slickenlines along the large active faults described above. This algorithm has been widely used in many neotectonic studies to determine the stress regime in the neighbouring Greek mainland (e.g. Mercier et al. 1989; Mercier & CareyGailhardis 1989) and also worldwide (e.g. CareyGailhardis & Mercier 1992; Bellier & Zoback 1995). From this analysis it is concluded that the normal movement on the main faults of the area, that is the youngest ones, has been controlled by an extensional stress field, which has a vertical greatest principal stress axis (or1) and a horizontal least principal stress axis (or3) trending northsouth (Fig. 5). This stress field is in accordance with the extension defined by focal mechanisms of small earthquakes in the region (Van Eck & Stoyanov 1996). It also fits well with the northsouth-striking extensional stress field recognized further south in central northern Greece

684

M.D. TRANOS E T AL.

o1:081-85 ~ 02:261-05 ~ 03:351-02 ~

10 8 N

R: 0,94

6

ii: :i :i84

~:i ~ :)i~ ::.

: .....

.

. . . . . . . . . . . .

10

20

30

40

50

(t^s)

a

b

Fig. 5. (a) Stereographic projection (equal area, lower hemisphere) indicating the latest kinematics of the faults and the principal stress axes oh, ~2, or3of the strain ellipsoid defined by the algorithm of Carey & Brunier (1974). (b) Bar diagram showing the angle between theoretical and measured slip vector (t^s) of the faults shown in (a) as calculated using the algorithm of Carey & Brunier (1974). N, number of fault-slip data; R, stress ratio.

from fault-slip data stress-inversion methods (Mercier et al. 1989; Tranos 1998), neotectonic joints (Tranos & Mountrakis 1998) and focal mechanisms (Mercier & Carey-Gailhardis 1989; Papazachos et al. 1998, 2001). It is also in good agreement with a north-south-trending stress regime, as defined by the similarly E N E - W S W - to east-west-striking active rupture zones of central Macedonia and Thrace, i.e. the ThessalonikiGerakarou fault zone (Tranos et al. 2003), the Serres fault zone (Tranos & Mountrakis 2004), and the K a v a l a - X a n t h i - K o m o t i n i fault zone (Mountrakis & Tranos 2004).

Conclusions Our investigations in SW Bulgaria suggest the presence of major normal faults striking E N E WSW (70-80 ~ and W N W - E S E (100~ both related to seismic activity of a broader area. The ENE-WSW-striking faults are more common along the Strouma River, in contrast to the WNW-ESE-striking faults, which are more prevalent in central and eastern Bulgaria. The former are kilometres-long normal faults that bound narrow basins such as the Doupnitsa, Kyustendil and Kroupnik basins; these are characterized by long-term slip rates that vary from 0.1 to 0.7 mm a -1. However, both W N W - E S E

and ENE-WSW-striking faults merge to form major kilometres-long arcuate rupture zones. Such a main rupture zone joining the Kochani, Kroupnik, Gradevo-Predela and Bansko faults is well defined, as confirmed by recent seismic activity distributed along it. This zone can be considered as the seismogenic fault for the strong Kroupnik earthquakes that struck the region in 1904. By contrast, the inherited NNW-SSE and N W - S E Alpine structures and faults (Fig. 1) do not exhibit any significant reactivation in the contemporary stress regime, and were apparently barriers to rupture propagation. The E N E WSW-striking faults are normal faults without a significant strike-slip component and demonstrate similar geometric features (e.g. strike and length) and similar kinematics to the active faults of central~eastern Macedonia and Thrace (Northern Greece). Therefore, they could be considered as active or possibly active faults. The driving stress regime as proposed here is a north-south extensional stress field, which corresponds to the well-recognized north-south extension of Northern Greece since the Quaternary. Therefore, we suggest that SW Bulgaria and central-eastern Macedonia and Thrace (Northern Greece) share seismotectonic properties at least during the Quaternary to the present, whereas the influence of the North Anatolian

ACTIVE FAULTS OF SW BULGARIA F a u l t in N o r t h e r n Greece and SW Bulgaria has been overestimated in previously published models. Critical reading of the manuscript by B. Papazachos is greatly appreciated. The authors also acknowledge constructive reviews by I. S. Zagorchev and R. Westaway. This study was supported by the bilateral research project between Greece and Bulgaria EPAN-M.4.3.6.1 and NZ-1209102. References

BELLIER, O. 8~; ZOBACK, M. L. 1995. Recent state of stress change in the Walker Lane zone, western Basin and Range province, United States. Tectonics, 14(3), 564-593. BONCHEV, E. 1958. l~lber die tektonische Synthese Westbulgariens. Geologija na Balkanite, 2, 5-48 (in Bulgarian with German abstract). BONCHEV,E. 1971. Problems of Bulgarian Geotectonics. Tehnika, Sofia. (in Bulgarian). BURCHFIEL, B. C., NAKOV, R. & TZANKOV, T. 2000. Cenozoic extension in Bulgaria and Northern Greece: the northern part of the Aegean extensional regime. 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, 325-352. CAREY-GAILHARDIS, E. & BRUNIER, B. 1974. Analyse th6orique et numerique d'un mod61e m6canique 61ementaire appliqu6 ~ l'6tude d'une population de failles. Comptes Rendus de l'AcadOmie des Sciences, 279, 891-894. CAREY-GAILHARDIS, E. 8~ MERCIER, J. L. 1992. Regional state of stress, fault kinematics and adjustments of blocks in a fractured body of rock: application to the microseismicity of the Rhine graben. Journal of Structural Geology, 14(8-9), 1007-1017. CHRISTOSKOV, L. 8~; GRIGOROVA, E. 1968. Energetic and space-time characteristics of the destructive earthquakes in Bulgaria after 1900. Bulletin of the Geophysical Institute, Bulgarian Academy of Science, 12, 79-107 (in Bulgarian with English summary). DINEVA, S., SOKEROVA, D. & MICHAILOV, D. 1998. Seismicity of south-western Bulgaria and border regions. Journal of Geodynamics, 26(2-4), 30%325. DINEVA, S., BATLLO, J., MIHAYLOV,D. & VAN ECK, T. 2002. Source parameters of four strong earthquakes in Bulgaria and Portugal at the beginning of the 20th century. Journal of Seismology, 6, 99-123. DOBREV, N. & KOSTAK,B. 2000. Fault dynamics in the Simitli graben (SW Bulgaria) and its monitoring. In: MILEV, G. (ed.) Reports on Geodesy, Warsaw University of Technology, 4(49), 123-135. DOBREV, N. & TACHEVA, E. 2000. A short review on the seismogenic terrain deformations in the Simitli

685

graben (SW Bulgaria). In: MILEV, G. (ed.) Reports on Geodesy, Warsaw University of Technology, 4(49), 145-147. GRIGOROVA, E. & PALIEVA, K. 1968. Macroseismic characteristics of the destructive earthquake from 04.04 1904. Bulletin of the Geophysical Institute, Bulgarian Academy of Science, 12, 109-112 (in Bulgarian with English summary). JACKSON, J. 1994. Active tectonics of the Aegean region. Annual Review of Earth and Planetary Sciences, 22, 239-271. JACKSON, J. A. & MCKENZIE, D. P. 1988. The relationship between plate motions and seismic tensors, and the rate of active deformation in the Mediterranean and Middle East. Geophysical Journal International, 93, 45-73. JACKSON, J. A. ~; WHITE, N. J. 1989. Normal faulting in the upper continental crust: observations from regions of active extension. Journal of Structural Geology, 11, 15-36. JARANOFF, D. 1960. La Tectonique de la Bulgarie. Tehnica, Sofia. (in Bulgarian with French summary). KARAKOSTAS, V. G., PAPADIMIDRIOU,E. E., TRANOS, M. D., et al. 2004. Seismotectonic properties of SW Bulgaria. In: Geophysics in Economic Activity," environment and cultural heritage investigations. Book of Abstracts, 4th National Geophysical Conference, Sofia, 4-5 October 2004, 79-80. KOTZEV, V., NAKOV, R., BURCHFIEL,B. C., KING, R. & REILINGER, R. 2001. GPS study of active tectonics in Bulgaria: results from 1996 to 1998. Journal of Geodynamics, 31, 189-200. KOUKOUVELAS, I. K. & AYDIN, A. 2002. Fault structure and related basins of the North Aegean Sea and its surroundings. Tectonics, 21(5), 10, TC 901037. MARINOVA, R. & ZAGORCHEV, I. 1990a. Geological map of the People's Republic of Bulgaria, 1:100 000 series, Blagoevgrad Sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MARINOVA, R. & ZAGORCHEV, I. 1990b. Geological map of the People's Republic of Bulgaria, 1:100 000 series, Delchevo Sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MARINOVA, R. & ZAGORCHEV, I. 1990C. Geological map of the People's Republic of Bulgaria, 1:100 000 series, Kyustendil Sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MARINOVA, R. & ZAGORCHEV, I. 1990d. Geological map of the People's Republic of Bulgaria, 1:100 000 series, Razlog Sheet. Geological Institute, Bulgarian Academy of Sciences, Sofia. MARRETT, R. & ALLMENDINGER, R. W. 1990. Kinematic analysis of fault-slip data. Journal of Structural Geology, 12(8), 973-986. MCKENZIE, D. P. 1972. Active tectonics of the Mediterranean region. Geophysical Journal of the Royal Astronomical Society, 30, 109-185. MERCIER, J. L. & CAREY-GAILHARDIS, E. 1989. Regional state of stress and characteristic fault kinematics instabilities shown by aftershock sequences: the aftershock sequences of the 1978

686

M . D . TRANOS ET AL.

Thessaloniki (Greece) and 1980 CampaniaRIZHIKOVA, S., TOTEVA, T. & RANGUELOV, B. 2000. Seismicity of the Kresna Source zone for eightyLucania (Italy) earthquakes as examples. Earth and year post active period (1909-1989). In: MILEV, G. Planetary Science Letters, 92, 247-264. (ed.) Reports on Geodesy, Warsaw University of MERCIER, J.-L., SIMEAKIS,K., SOREL, D. & VERGELY, Technology, 4(49), 56-60. P. 1989. Extensional tectonic regimes in the Aegean SHANOV, S. B. 1997. Contemporary and neotectonic basins during the Cenozoic. Basin Research, 2, stress field in the eastern part of Balkan Peninsula. 49-71. PhD thesis, Bulgarian Academy of Sciences, Sofia MEYER, B., ARMIJO, R. & DIMITROV, D. 2002. Active (in Bulgarian). faulting in SW Bulgaria: possible surface rupture SHANOV, S. ~r DOBREV, N. 2000. Tectonic stress field of the 1904 Strouma earthquakes. Geophysical in the epicentral area of 04.04.1904 Krupnik earthJournal International, 148, 246-255. quake from striae on slickensides. In: MILEV, G. MOSKOVSKI, S. & GEORGIEV,A. 1970. On the structure (ed.) Reports on Geodesy, Warsaw University of of Kresna Gorge region. Annuaire de l'Universitb Technology, 4(49), 117-122. de Sofia, 62, 95-111 (in Bulgarian with English SHANOV, S., KURTEV, K., NIKOTOV, G., BOYKOVA,A. summary). 8r RANGUELOV, B. 2001. Seismotectonic characMOUNTRAKIS, D. M. 8r TRANOS, M. D. 2004. The teristics of the western periphery of the Rhodope Kavala-Xanthi-Komotini fault (KXKF): a comMountain region. Geologica Balcanica, 31(1-2), plicated active fault zone in Eastern Macedonia53-66. Thrace (Northern Greece). In: CHATZIPETROS, SHEBALIN, N. V., KARNIK, V. 8r HADZIEVSKI,D. 1974. Catalogue of Earthquakes, Atlas of Isoseismal A. A. 8r PAVLIDES, S. I . (eds) 5th International Maps. UNDP-UNESCO Survey of the Seismicity Symposium on Eastern Mediterranean Geology, of the Balkan Region, Skopje. Thessaloniki, Greece, 2, 857-860. SHIPKOVA, K. t~ IVANOV, Z. 2001. Effects of Late NEDJALKOV,P., KOJUMDGIEVA,E. & BOZHKOV,I. 1988. Alpine extension in the northwestern foot of Rila Sedimentation cycle in the Neogene grabens along Mountain. Geologica Balcanica, 31, 138-139. Strouma valley. Geologica Balcanica, 18(2), 61-66. TRANOS, M. D. 1998. Contribution to the study of the PACHECO, J. F. 8r SYKES, L. R. 1992. Seismic moment neotectonics deformation in the region of Central catalog of large shallow earthquakes, 1900 to 1989. Macedonia and North Aegean. PhD thesis, UniBulletin of the Seismological Society of America, versity of Thessaloniki (in Greek with extended 82, 1306-1349. English abstract). PAPAZACHOS, C. B. 1999. Seismological and GPS TRANOS, M. D. 2004. Late Cenozoic faulting deforevidence for the Aegean-Anatolia interaction. mation of SW Bulgaria. In: CHATZIPETROS,A. A. 8r Geophysical Research Letters, 17, 2653-2656. PAVLIDES, S. B. (eds) 5th International Symposium PAPAZACHOS, B. C. & PAPAZACHOU, C. 2003. The on Eastern Mediterranean Geology, Thessaloniki, earthquakes of Greece. Ziti, Thessaloniki. Greece, 1,400-403. PAPAZACHOS, B. C., PAPAIOANNOU, CH. A., TRANOS, M. D. & MOUNTRAKIS, D. M. 1998. PAPAZACHOS, C. B. 8~; SAVVAIDIS,A. S. 1997. Atlas Neotectonic joints of northern Greece; their sigof isoseismal maps for strong earthquakes in Greece nificance on the understanding of the active and surrounding area (426Bc-1995). Publication of deformation. Bulletin of the Geological Society of Geophysical Laboratory, Thessaloniki University. Greece, 32, 209-219. PAPAZACHOS, B. C., PAPADIMITRIOU, E. E., KIRATZI, TRANOS, M. D. • MOUNTRAKIS, D. M. 2004. The Serres Fault Zone (SFZ): an active fault zone A. A., PAPAZACHOS,C. B. & LOUVARI, E. K. 1998. in Eastern Macedonia (Northern Greece). In: Fault plane solutions in the Aegean and the CHATZIPETROS, A. A. 8r PAVLIDES, S. B. (eds) 5th surrounding area and their tectonic implications. International Symposium on Eastern Mediterranean Bollettino di Geofisica Teoricaed Applicata, 39, Geology, Thessaloniki, Greece, 2, 892-895. 199-218. TRANOS, M. D., PAPADIMITRIOU,E. E. & KILIAS,A. A. PAPAZACHOS, I . K., MOUNTRAKIS, D. M., 2003. Thessaloniki-Gerakarou Fault Zone PAPAZACHOS, C. B., TRANOS, M. D., KARAKAISIS, (TGFZ): the western extension of the 1978 G. PH. 8/; SAWAIDIS, A. S. 2001. The faults caused Thessaloniki earthquake fault (Northern Greece) the known strong earthquakes in Greece and the and seismic hazard assessment. Journal of Strucaround area since 5~ B.c. In: 2nd Proceedings of tural Geology, 25, 2109 2123. Seismic Mechanics and Engineering Seismology, VAN ECK, T. & STOYANOV, T. 1996. Seismotectonics 28-30 November, Volume A, 17-26 (in Greek). and seismic hazard modelling for Southern Technical Chamber of Greece, Thessaloniki. Bulgaria. Tectonophysics, 262, 77-100. PAVLIDES, S., MOUNTRAKIS,D., KILIAS,A. & TRANOS, VRABLIANSKI, B. 1974. Neotectonic studies in Simitli M. 1990. The role of strike-slip movements in the Graben and its framework. Bulletin of the Geologiextensional area of the northern Aegean (Greece). cal Institute, Bulgarian Academy of Science, 23, In: BOCCALETTI, M. & NUR, A. (eds) Active and 195-220 (in Bulgarian with French summary). Recent Strike-slip Tectonics. Annales Tectonicae, 4, VRABLIANSKI, B. 1977. Neotectonic regime of Struma 196-211. fault zone. Geotectonics, Tectonophysics and RANGUELOV, B., RIZHIKOVA, S. & TOTEVA, T. 2001. Geodynamics, Sofia, 7, 18-41 (in Bulgarian). The earthquake (M 7.8) source zone (south-west VRABLIANSKI, B. & MILEV, G. 1993. Neotectonic Bulgaria). Academic Publication House 'M. features of the Struma fault zone. Acta Montana, Drinov', Sofia. 4(90), 111-132.

ACTIVE FAULTS OF SW BULGARIA ZAGORCHEV, I. S. 1969. The Struma deep fault during the Late Alpine orogenic stage. Acta Geologica Academiae Scientiarum Hungaricae, 13, 437-441. ZAGORCHEV, I. S. 1970. On the neotectonic movements in a part of South-West Bulgaria. Bulletin of the

Geological Institute, Bulgarian Academy of Science, 19, 141-152 (in Bulgarian with French summary). ZAGORCHEV, I. S. 1971. Certain features of the Young Alpine block structure in a part of South-West Bulgaria. Bulletin of the GeologicalInstitute, Bulgarian Academy of Science, 20, 17-27 (in Bulgarian with English summary).

68'/

ZAGORCHEV, I. S. 1975. Basement block structure of a part of the Krai~tid-Vardar Lineament. Geologica Balcanica, 5(2), 3-18. ZAGORCHEV, I. S. 1992a. Neotectonics of the central parts of the Balkan Peninsula: basic features and concepts. Geologische Rundschau, 81, 635-654. ZAGORCHEV, I. S. 1992b. Neotectonic development of the Strouma (Krai~tid) Lineament, southwest Bulgaria and northern Greece. Geological Magazine, 129(2), 197-222. ZAGORCHEV, I. S. 2001. Introduction to the geology of SW Bulgaria. Geologica Balcanica, 31(1-2), 3-52.

Perspectives for earthquake prediction in the Mediterranean and contribution of geological observations B. C. P A P A Z A C H O S ,

G. F. K A R A K A I S I S ,

C. B. P A P A Z A C H O S

&

E. M. S C O R D I L I S

Geophysical Laboratory, Aristotle University, P O B o x 352-1, 54124, Thessaloniki, Greece (e-mail." karakais@geo, auth. gr) Accelerating seismic strain caused by the generation of intermediate-magnitude preshocks in a broad (critical) region, accompanied by decelerating seismic strain caused by the generation of smaller preshocks in the seismogenic region are systematically observed before strong mainshocks. On the basis of this seismicity pattern a model has been developed that seems promising for intermediate-term earthquake prediction, called the 'Decelerating in-Accelerating out Seismic Strain Model'. Recent seismological data for the Mediterranean region are used here for backward and forward testing of this model. The selection of the broader Mediterranean region as a test area was motivated not only by the interest of timedependent seismic hazard assessment in a high-seismicity and highly populated region but also by the fact that the Mediterranean is a natural geophysical and geological laboratory where both complex multi-plate and continuum tectonics are found in a more or less convergent zone. Within this complex geotectonic setting several geological phenomena such as subduction, collision, orogen collapse and back-arc extension take place, leading to the generation of a broad spectrum of mainshocks, reaching Mw = 8.0 or greater for subductionrelated thrust events and a variety of corresponding seismicity levels and neotectonic activity ranging from very low (e.g. large parts of Iberian peninsula) to very high (broader Aegean area). The backward procedure shows that all six strong (M > 6.8) mainshocks that have occurred in the Mediterranean since 1980 had been preceded by preshock sequences that followed this seismicity pattern and satisfy all model constraints. Application of the model for future mainshocks has led to the identification of nine regions (in the Pyrenees, Calabria, NE Adriatic, Albania, Northern Greece, SE Aegean, NW Anatolia, western Anatolia, NE Anatolia) where current intermediate-magnitude seismicity satisfies the constraints of the model and corresponds to strong (M >_6.2) mainshocks. The magnitudes, epicentres and origin times of these probably ensuing mainshocks, as well as their corresponding uncertainties, are estimated, so that it is possible to evaluate the model potential during the next decade (2006-2015). Furthermore, it is shown that geological observations of surface fault traces can contribute to the accurate location of the foci of future strong mainshocks in the Mediterranean and to an estimation of their sizes. For this purpose, globally valid relations between fault parameters based on geological observations (surface fault length, Ls, and fault slip, Us) and measures of mainshock size (mainshock magnitude, subsurface fault length, L, and fault slip, u) are proposed. Abstract:

Recently, it has become more evident that antiseismic measures cannot be effective without knowledge on the location, size and time of future strong earthquakes, that is, without prediction of individual strong earthquakes. At present, however, only knowledge of the spatial distribution of strong earthquakes is of practical use because their time distribution is considered as random. This is because the prediction of all three basic parameters (space, time, magnitude) with reasonable uncertainties is a very difficult scientific task. Short-term earthquake prediction (time uncertainty of the order of days to weeks) is not feasible with the present state of knowledge (e.g. Wyss 1997). Long-term prediction (time uncertainty of the order of decades) of a future strong

earthquake (mainshock) requires accurate knowledge of the physical process of generation of the previous mainshock on the same fault, but such knowledge is not feasible at present (Jaum6 & Sykes 1999). It seems, however, that intermediate-term earthquake prediction (time uncertainty of the order of a few years) is possible, on the basis of precursor seismicity patterns (Evison 2001). Accelerating generation of intermediatemagnitude preshocks in broad regions (Tocher 1959; Papadopoulos 1986; Sykes & Jaum6 1990; Knopoff et al. 1996; Tzanis et al. 2000, among m a n y others) and decelerating generation (seismic quiescence) of preshocks in the narrower (seismogenic) region (Wyss & H a b e r m a n n 1988; Bufe et al. 1994; Hainzl et al. 2000; Z611er et al.

From: ROBERTSON,A. H. F. & MOUNTRAKIS,D. (eds) 2006. Tectonic Development of the Eastern Mediterranean Region. Geological Society, London, Special Publications, 260, 689-707. 0305-8719/06/$15.00 9 The Geological Society of London 2006.

690

B.C. PAPAZACHOS E T AL.

"o"~

~0.o=

r

~o o

.o ~ .~

g~o ' n 'E

~ ~ "~'~ 0 ~ d

.N

~.~ ~

~ ~

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN 2002) are two of the most distinct precursory patterns. The term 'preshock' does not refer to the traditional 'foreshock' term, which corresponds to earthquakes occurring in the vicinity of the fault region a few days or months before the earthquake generation, but to the intermediate earthquake activity that occurs on a much larger scale (up to almost 10times the fault length) within the critical region of the ensuing earthquake, which is preparing for its generation through stress alignment and long-range earthquake interactions. The simultaneous occurrence of the two patterns (seismic excitation and seismic quiescence) in a region has been called a 'doughnut pattern' by Mogi (1969). Accelerating generation of preshocks in the broad (critical) region is considered as a critical phenomenon culminating in a mainshock (critical earthquake), and is considered as a critical point (Sornette & Sornette 1990; Sornette & Sammis 1995), whereas decrease of preshock activity in the narrower (seismogenic) region has been attributed to stress relaxation as a result of preseismic sliding (Kato et al. 1997; Wyss et al. 1981). Recent detailed investigation of accelerating preshock sequences in broad critical regions (Papazachos & Papazachos 2000, 2001; Papazachos et al. 2005) and of decelerating preshock sequences in corresponding narrow seismogenic regions (Papazachos et al. 2004a,b) has revealed important predictive properties, observed almost simultaneously in both regions (critical and seismogenic) corresponding to the same mainshock. These properties have been formulated by analytical relations, which are the basis of a promising intermediate-term earthquake prediction model (Papazachos et al. 2006), which can be called the 'Decelerating in-Accelerating out Seismic Strain Model'. This model is applied in the present study by using recent data for strong earthquakes in the Mediterranean and surrounding region (35~ 45~ 5~176 Because the proposed model presents uncertainties that can be estimated and assessed only by studying a large number of events, tests performed on a smaller-scale area (e.g. a well-studied fault zone such as the North Anatolian Fault), where data for only a very few recent mainshocks can be used, may lead to misleading results. Furthermore, the establishment of the main model relations presented below requires a variety of seismotectonic environments where different magnitude mainshocks (large range of moment magnitude values) and different seismicity levels (large range of seismicity rate values) are found. The Mediterranean region is

691

a natural geophysical laboratory, which allows for an efficient backward and forward testing of the model, as it includes various seismotectonic regimes (subduction, collision, back-arc extension, etc.) with a variety of strong mainshocks (up to Mw= 8.0 or larger) and seismicity levels ranging from very low (e.g. large parts of Iberian peninsula) to very high (wider Aegean area). The main goal of this work is to test the above model on preshock sequences of strong mainshocks that have already occurred in the area (backward testing), as well as on probable preshock sequences of future strong mainshocks (forward testing). In addition, global relations are defined between geologically observed fault parameters (surface fault length and slip) and seismic quantities (moment magnitude, subsurface fault length and slip) (Papazachos et al. 2004c), to facilitate the contribution of geological observations to the estimation of the location and size of future mainshocks.

Study area The Mediterranean region (Fig. 1) represents the boundary between the Eurasian and African lithospheres, and shows a complex geotectonic setting. This setting comprises several mountain belts, as well as Neogene basins that were formed by back-arc extension and/or Alpine crust collapse (e.g. Dercourt et al. 1986; Dewey 1988), despite the fact that no generally accepted model for the extension mechanism exists. However, the most important factor in present-day tectonics is widely accepted to be the convergence between Eurasia and Africa in a more or less north-south direction at a rate of about 10 mm a -1. Several smaller plates also contribute significantly to this tectonic setting, such as the northward motion of the Arabian plate, the westward motion of the Anatolian plate, the SW motion of the Aegean plate, the N N W motion and anticlockwise rotation of the Adriatic plate, and the expansion towards both the east and west of the western Mediterranean lithosphere (e.g. McKenzie 1970, 1972). The clearest manifestation of these plate motions is the significant seismicity level, with events exceeding M = 8 . 0 (Papazachos 1990). These events, which are the result of this complex lithospheric interaction, occur along several types of faults, such as the strike-slip faults of northern Anatolia, dip-slip faults in the continental crust system (e.g. normal faults in the Aegean or Apennine region, lowangle thrust faults in the southern coasts of the Western Mediterranean) and dip-slip faults in the lithospheric subduction regions (e.g. convex side

692

B.C. PAPAZACHOS E T AL.

of the Hellenic arc). It is interesting to note that a large number of active faults (mainly of normal type) are activated as a result of strong back-arc extension (e.g. Aegean, Anatolia, Apennines), far from the compressive margins shown in Figure 1. Figure 1 shows significant spatial variations in the seismic activity throughout the Mediterranean. Relatively low seismicity levels are found along the more or less compressive tectonic boundaries of the Western Mediterranean, such as the Pyrenees, Betics, Alboran Sea and Atlas (e.g. Andrieux et al. 1971; Buforn et al. 1988a,b; Jimenez-Munt & Negredo 2003). The Central Mediterranean shows higher activity as a result of the Inner Apennine extension, the Tyrrhenian Sea subduction and the Adriatic collision with the Dinarides (e.g. McKenzie 1972; Gasparini et al. 1985; Malinverno & Ryan 1986; Anderson & Jackson 1987; Faccenna et al. 1996, 2001; Wortmann et al. 2001). The fastest plate motions and corresponding seismicity levels are found in the Eastern Mediterranean, with Arabia moving northwards along the Dead Sea Fault, resulting in a transpressional regime along the East Anatolian Fault. Deformation rates and seismicity increase in the Anatolia region as a result of its westward migration, with most of its westward motion taken along the North Anatolian Fault, whereas the Aegean region exhibits the highest deformation rates and seismicity, moving rapidly towards the SW, owing to the combined effect of Anatolia westward motion and subduction rollback (McKenzie 1970, 1972; Mercier et al. 1976, 1989; Dewey & Seng6r 1979; LePichon & Angelier 1979; Jackson & McKenzie 1988; Taymaz et al. 1991; Papazachos & Kiratzi 1996; Meijer & Wortel 1997; Kahle et al. 1998; Papazachos et al. 1998; McClusky et al. 2000). The seismicity pattern shown in Figure 1 should not be considered as either static or representative. Significant long-term temporal seismicity variations have been identified in this area, and several large events occur on major faults with long return periods (e.g. Dead Sea Fault). Furthermore, several destructive events often occur along sections of the Mediterranean tectonic boundaries, which are considered to be of relatively low seismic hazard potential, such as the recent Morocco and Algeria events. Moreover, the exact geometry and seismicity potential of several of the tectonic boundaries shown in Figure 1, especially at sea, are still poorly understood. Therefore, it is imperative to incorporate all available and future geological information to better understand and constrain the active tectonic setting of the Mediterranean and facilitate earthquake prediction efforts.

The model The model used in the present work has its observational basis in two precursory seismicity patterns, the 'accelerating seismic strain in a broader (critical) region' and the 'decelerating seismic strain in a narrower (seismogenic) region'. Two corresponding methods have been proposed for intermediate-term earthquake prediction. The first of these, called the 'timeto-failure method' (Bufe & Varnes 1993) is based on the accelerating generation of the Benioff strain (square root of seismic energy) released by intermediate-magnitude preshocks in the broader region. The second one, called the 'seismic quiescence method' (e.g. Wyss et al. 1981), is based on the decrease of the rate of generation of small preshocks in the narrower (seismogenic) region. Both methods have been recently developed further and combined to improve their efficiency for intermediate-term earthquake prediction. Regarding the accelerating strain method, additional predictive properties, which are expressed by empirical relations (Papazachos et al. 2005), have been recently proposed. Also, recent developments of the seismic quiescence method consider the decelerating strain instead of the decrease of the frequency of the small preshocks as precursory pattern and have proposed predictive properties, which are also expressed by empirical relations (Papazachos et al. 2004b). The most important relative progress, however, is the identification of simultaneous occurrence of the two patterns during the preshock period (Papazachos et al. 2004a,b), which suggests the formation of a model with predictive properties called the 'Decelerating in-Accelerating out Seismic Strain Model'. This model is applied in the present work and its basic characteristics are briefly described. Accelerating

seismic strain

The time variation of the accelerating preshock seismic strain, S (in Joulel/2), in the broader (critical) region is given by the relation S ( t ) = A - B(t~ - t) m

(1)

where t is the time to the mainshock, tc is the origin time of the mainshock and A, B and m are parameters that are determined by observations (Bufe & Varnes 1993). Bowman et al. (1998) quantified the degree of deviation of the time variation of S from linearity by proposing the minimization of a curvature parameter, C, which is defined as the ratio of the root mean square

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN error of the power-law fit (equation (1)) to the corresponding linear fit error. Papazachos & Papazachos (2000, 2001) suggested additional constraints to the critical earthquake model expressed by empirical formulae, which relate parameters of the accelerating preshock sequence to the mainshock magnitude and the long-term seismicity rate in the critical region. Very recently, Papazachos and colleagues (Papazachos et al. 2005, 2006) used global data (from the Mediterranean, Himalayas, California and Japan) to derive the following empirical relations: log R=0.42M-0.301og s,+ 1.25, cy=0.15 (2) log (tc--t~a)=4.60--0.571og s,, ~=0.10 M = M13 + 0.60, cy= 0.20

(3) (4)

where R is the radius (in km) of the equivalent circle of the elliptical critical region, sa (in Joule m per year and per 104 kin:) is the rate of the long-term seismic strain in the critical region, ts, (in years) is the start time of the accelerating sequence, M is the mainshock magnitude and M~3 is the mean magnitude of the three largest preshocks. The smallest preshock magnitude, M~,, of an accelerating preshock sequence for which the best solution is obtained is given by the relation M - Mmin= 0.54M- 1.91

(5)

where M is the mainshock magnitude (Papazachos et al. 2005). Thus, for mainshock magnitudes 6, 7 and 8, the Mminis 4.7, 5.1 and 5.6, respectively. The probability, P, that the calculated parameters for an examined region fit these relations is also estimated by the available data for this region. Furthermore, for each point of an investigated area a 'quality index', q~, has been defined (Papazachos et al. 2002) by the formula P qa

mC

(6)

Hence, qa increases with increasing probability, P, showing a similarity to previous preshock (critical) region behaviours, to the degree of deviation from linearity of the time variation of the strain (decrease of C), and to the degree of acceleration (decrease of m). The following cut-off values have been determined (Papazachos et al. 2005) for these four parameters: C<0.60, P>0.45, m<0.35, qa_>3.0.

(7)

693

From all grid points of the examined area that fulfil these relations, the one for which the quality index, qa, takes its largest value is considered as the geometric centre, Q, of the critical region and the corresponding solution (M, tsa, M13, s. . . . . ) as the best solution. Decelerating seismic strain

Decelerating seismic strain released by intermediate-magnitude preshocks in the seismogenic region during the critical period, when accelerating seismic strain occurs in the broader region, also follows a power law (equation (1)) but with m > l . 0 (Papazachos et al., 2006). Additional properties of the decelerating strain in the seismogenic region are expressed by the relations log a=0.23M--0.141og Sd+ 1.40, CY=0.10 (8) log (tc-tsa)=2.95-0.311og Sd, CY=0.12 (9) where a (in km) is the large axis of the elliptical seismogenic region (typically with ellipticity e =0.70), M is the magnitude of the mainshock, tsd (in years) is the start time of the preshock decelerating strain and sa (in Joule 1/2per year and per 104 km 2) is the long-term seismic Benioff-strain rate of the seismogenic region. The smallest magnitude, Mmin, of the decelerating preshocks for which the best solution is obtained is given by the relation M - M ~ n = 0 . 7 1 M - 2 . 3 5 , cy=0.1

(10)

where M is the mainshock magnitude (Papazachos et al. 2006). Thus, for mainshock magnitudes 6, 7 and 8, the values of Mm~nare 4.1, 4.4 and 4.7, respectively. A quality index can be also defined by the relation qa -

P.m C

(11)

where P is the probability that decelerating preshock observations in a seismogenic region are compatible with equations (8) and (9). The following cut-off values have been determined for these parameters by the use of global data (Papazachos et al. 2006): C<0.60, P_>0.45, 2.5_<m_<3.5, qa>3.0. (12) The geographical point of the narrower (seismogenic) region that fulfils equations (12) and corresponds to the largest qd value (best solution) is considered as the geometric centre, F, of the seismogenic region.

694

B.C. PAPAZACHOS ET AL.

Estimation o f parameters o f ensuing mainshocks

present work, as an indication of the accuracy of the proposed predictions.

For the estimation (prediction) of the parameters of ensuing mainshocks we make use of properties of both the accelerating pattern and decelerating pattern of the seismic strain. Thus, the estimated origin time, to*, of the ensuing mainshock is the average of the origin time corresponding to the best solution of the accelerating seismic strain (equation (3)) and of the origin time corresponding to the best solution of the decelerating seismic strain (equation (9)). Similarly, the estimated magnitude, M*, of the ensuing mainshock is the average of the value calculated by equations (2) and (4) and equation (8). For an estimation of the geographical coordinates of the epicentre, E*(% k), of the ensuing mainshock we make use of the locations of: the geometric centre, F, of the seismogenic region; the mean epicentre, Pf, of the decelerating preshocks, which is considered as the physical centre of the seismogenic region; the geometric centre, Q, of the critical region; and the mean epicentre, Pq, of the accelerating preshocks, which is considered as the physical centre of the critical region. The two centres F and Pf are usually close, and for this reason we make use of the middle point, D, of the line segment F-Pf. For the same reason, we make use of the middle point, A, of Q-Pq. From a large sample of previous mainshocks (Papazachos et al. 2006) it has been shown that the mean distance between D and the mainshock epicentre is DE= 100 km with a standard deviation of 40 km, and the distance between A and E is AE = 180 km with a standard deviation of 80 km. Then, the mainshock epicentre, E, is defined by the circle (D, 100 km) with centre D and radius 100 km and by the circle (A, 180 km). In the cases when the two circles intersect at two close points or these circles do not intersect there is a unique solution. In these cases, the estimated epicentre is considered to be the intersection of circle (D, 100 km) with the line DA that is closer to the circle (A, 180 kin). The mainshock epicentre, E, lies between D and A at a mean distance DE = 0.4 DA. From comparison of the estimated parameters (to*, M*, E*) using this approach of previous mainshocks with the known parameters (to, M, E) of these mainshocks, it can be concluded (Papazachos et al. 2006) that the two standard deviation model uncertainties for the estimated parameters are _+2.5years for the origin time of the mainshock, _+0.4 for the magnitude and 80_+ 70 km for its epicentre. These uncertainties can be also adopted for the nine probably ensuing mainshocks considered in the

Application of the model for earthquakes of the Mediterranean As it is not possible to test the validity of an earthquake prediction model by producing data at will (for instance, by laboratory experiments), retrospective predictions of previous earthquakes (postdictions) or predictions of future earthquakes are usually used instead. Both these procedures are examined in the present work. For this reason, three data samples are necessary for the study area: (1) a sample that includes the examined mainshocks; (2) a sample containing the preshocks of each mainshock; (3) all shocks that are used to define the long-term mean strain rate release both in each critical region, sa, and in the seismogenic region that engulfs each fault, sd. For the purposes of the present work, data for the broader Mediterranean region have been taken from a recently compiled catalogue for this region (Papazachos et al. 2005). The standard catalogue uncertainties involved are typically less than 30 km for the epicentre and 0.3 for the moment magnitude. It should be noted that the data necessary to compute the long-term mean strain release (M>5.2) are complete since 1911 for the Mediterranean region (Western Mediterranean, Aegean, Anatolia). Furthermore, all magnitudes reported in the catalogue are either originally reported moment magnitudes or equivalent to moment magnitudes; that is, magnitudes that have been transformed to moment magnitudes from any other available scale (usually Ms or mb published by the International Seismological Center, ISC, and/or National Earthquake Information Center, NEIC) by using appropriate formulae (Scordilis 2006); this ensures the proper computation of Benioff strain from the earthquake catalogue.

B a c k w a r d testing Such testing, to be reliable, must be applied to a complete sample of mainshocks. All shallow mainshocks with M > 6 . 8 that occurred in the Mediterranean and surrounding area (35~ 45~ 5~176 since 1980 have been considered as such a sample. The minimum magnitude, 6.8, and the minimum time, 1980, have been defined to have high accuracy and completeness of the available data used for the study area. In the first line of each of the six cases presented in Table 1, the date, the geographical coordinates of the epicentre and the magnitude of these six mainshocks are given. In the second line,

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN

695

Table 1. Parameters of the critical regions (second line in each case) where accelerating preshock seismic deformation has been observed and of the seismogenic regions (third line in each case) where decelerating preshock seismic deformation has been observed before the generation of the six large ( M > 6. 8) mainshocks (first line in each case) that occurred in the Mediterranean region between 1980 and 2003 Event 1 2 3 4 5

6

Date 10.10.1980 Algeria 23.11.1980 Italy 19.12.1981 N Aegean 17.1.1983 W Greece 9.10.1996 Cyprus 17.8.1999 Anatolia

9, X

M

36.2, 01.4 36.4, 05.7 36.2, 01.6 40.8, 15.3 43.5, 14.3 40.6, 15.1 39.0, 25.3 40.0, 24.3 39.6, 24.5 38.1, 20.2 38.6, 17.8 38.2, 20.1 34.5, 32.1 35.5, 30.6 34.3, 32.3 40.8, 30.0 38.8, 28.4 40.8, 29.8

7.1 7.3 6.9 6.9 7.2 7.2 7.0 7.1 6.8 6.7 7.4 7.6

a

z

1314 100 138 100

e

C

m

q

Mm~n n

ts

Log sr

0.95 0.70

0.51 0.58

0.3 2.8

3.4 4.5

5.5 4.4

29 60

1911 4.5 1944 5.2

443 123

140 130

0.70 0.70

0.45 0.56

0.3 2.8

4.9 2.5

5.1 4.3

81 49

1951 5.3 1958 5.7

362 89

40 40

0.70 0.70

0.47 0.42

0.3 3.4

4.6 4.4

5.2 4.5

44 83

1969 5.9 1964 6.0

391 36

0 50

0.70 0.70

0.32 0.40

0.3 3.0

6.4 3.3

5.1 4.5

61 58

1963 5.6 1970 6.2

324 71

150 40

0.70 0.70

0.49 0.42

0.3 2.5

3.3 2.5

5.0 4.2

40 16

1981 5.6 1963 5.3

1089 129

90 90

0.90 0.70

0.30 0.32

0.3 3.2

5.5 8.7

5.1 4.6

181 17

1988 5.8 1980 5.8

the parameters for the broader (critical) region are listed: q0, X are the geographical coordinates of the centre, Q, of the critical region, M is the predicted magnitude, a (in km) is the length of the large axis of the elliptical critical region, z is the azimuth of this axis, e is the ellipticity of the region, C is the curvature parameter, m is the parameter from equation (1), q is the quality index, Mrs. is the minimum preshock magnitude, n is the number ofpreshocks, ts is the start year of the preshock sequence and Sr (in Joule 1~2per years and per 104km 2) is the long-term strain rate. In the third line of each of the six cases presented in Table 1 the corresponding parameters for the narrow (seismogenic) elliptical regions are presented, such as the geographical coordinates of the geometric centre of this region, and the length, a, of the large axis of the seismogenic

region. Figure 2 shows the broader (critical) and the narrower (seismogenic) elliptical regions for the six strong mainshocks in the Mediterranean. In the same figure the time variation of the cumulative strain, S(t), in the critical regions (accelerating strain, upper part of Fig. 2) and in the seismogenic regions (decelerating strain, lower part of Fig. 2) are also shown. Open circles and dots in this figure show epicentres of accelerating and decelerating preshocks, respectively. Numbers in Figure 2 correspond to code numbers of mainshocks in Table 1, and stars denote the epicentres of the six mainshocks.

The procedure followed in the present work considers the generation (in the broader critical region) of accelerating strain that satisfies equations (7) and the generation (in the narrower seismogenic region) of decelerating strain that satisfies equations (12). Preshock activity associated with all six strong (M > 6.8) mainshocks that occurred in the Mediterranean region during the period 1980-2003 fulfils constraints imposed by the model, as parameters C, m, P and q calculated for each preshock sequence (see Table 1) satisfy these relations. Therefore, the backward test can be considered as successful. F o r w a r d testing

For a forward testing of the model of Decelerating in-Accelerating out Seismic Strain, all of the Mediterranean and surrounding region (35~ 45~ 5~176 has been separated into a grid of points (0.2~ x 0.2~ and each point has been considered as the centre of an elliptical critical region corresponding to mainshocks with M > 6.2. Nine groups of points that fulfil equations (7) have been identified (Pyrenees, Calabria, N E Adriatic, Albania, Northern Greece, SE Aegean, N W Anatolia, Western Anatolia, NE Anatolia). The point for each group, which corresponds to the best solution (largest qa value) is considered as the geometric centre, Q, of the corresponding elliptical critical region. Information

696

B.C. PAPAZACHOS E T AL.

,.o

-~ ~.~

"~

~

~ 9 o ~:~

~ "~-~ ~

~,'~

, ~

~.~

~

" ~ ~ 9~ ' n

=

~

0,..0

0

.~

~ c ~

~~

o~

No ~

0 .= 0 "~ ,-~ r.~

9~

~

~

.~

m "~

~

=.,~ ~

~:~ome

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN

697

Table 2. Parameters of the critical regions (first line in each case) where accelerating seismic deformation currently occurs and parameters of the seismogenic regions (second line in each case) where decelerating seismic deformation currently occurs in the Mediterranean region

Event 1 2 3 4 5 6 7 8 9

tc

%L

M

a

z

e

C

m

q

Mr.in

n

ts

Log sr

2009.0 2007.2 2008.3 2009.6 2008.5 2008.7 2009.0 2008.6 2009.2 2007.1 2009.0 2010.2 2009.0 2008.0 2008.4 2006.6 2009.0 2008.6

43.2, 09.2 42.5, 01.5 38.2, 17.6 38.6, 15.6 43.5, 14.0 44.6, 16.1 44.1, 22.4 40.7, 20.5 41.0, 25.5 40.9, 23.4 36.3, 26.0 37.8, 26.8 42.6, 26.8 41.1, 27.8 39.4, 29.6 38.8, 29.5 39.6, 38.3 40.8, 38.7

6.4 6.5 6.8 6.6 6.5 6.5 6.6 6.6 6.3 6.7 7.0 6.7 6.3 6.3 6.2 6.2 7.1 6.9

397 248 292 174 297 170 349 155 209 154 533 160 250 133 183 118 525 201

90 80 100 60 20 130 10 20 50 130 30 10 50 60 100 80 130

0.90 0.70 0.60 0.70 0.70 0.70 0.60 0.70 0.60 0.70 0.70 0.70 0.00 0.70 0.70 0.70 0.70 0.70

0.42 0.21 0.50 0.34 0.42 0.28 0.37 0.36 0.41 0.38 0.32 0.31 0.33 0.45 0.46 0.30 0.29 0.46

0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0 0.3 3.0

6.4 13.5 6.4 7.5 7.4 10.4 7.0 7.5 5.9 6.7 7.0 6.1 9.5 5.6 5.8 5.7 10.2 6.2

4.8 4.2 4.9 4.3 4.9 4.3 4.9 4.3 4.7 4.3 5.0 4.2 4.7 4.2 4.7 4.1 5.1 4.2

59 38 38 37 52 44 42 141 35 48 141 497 35 45 86 55 82 43

1964 1963 1984 1991 1972 1988 1944 1994 1987 1993 1988 1996 1957 1993 1986 1992 1977 1990

5.17 5.27 5.64 5.42 5.33 5.27 4.89 5.76 5.71 5.81 5.75 5.80 5.06 5.72 5.70 5.76 5.43 5.42

on the values of the parameters of the best solution for the nine broad critical regions are listed in the first line for each of the nine cases in Table 2. Each grid point defined in the previous step has been considered as the centre of the elliptical seismogenic region, and the point for which equations (12) hold and the quality index, qd, has the largest value is considered as the geometric centre, F, of the seismogenic region. The nine ellipses in Figure 3 show the probable seismogenic regions defined by this method for the nine probably ensuing mainshocks. The time variation of the cumulative strain, S, in the corresponding nine assumed seismogenic regions is also shown. In the second line of each of the nine cases listed in Table 2, the parameters of the decelerating seismic strain are given. The estimated (predicted) by this method parameters (to*, E*, M*) of the nine probably ensuing mainshocks are listed in Table 3.

Contribution from geological observations Prediction of an individual earthquake requires knowledge of the source location, its size and its origin time before its generation. Because earthquakes are generated by slip on seismic faults and generation of strong shallow earthquakes on land is associated with observable surface fault traces, geological observations can contribute to the estimation of the location of the foci of strong earthquakes. The surface fault length, L , and surface

fault slip, us, can be also estimated by geological observations and as these fault parameters are related to the magnitude of the maximum earthquake on the fault, the expected mainshock magnitude, M, can also be assessed by geological observations. Active faults where no strong earthquakes have occurred during the instrumental period, and that have considerable probability of generating such earthquakes in the near future, can be reliably located by geological observations, whereas only faults that have been recently activated can be identified by seismological methods (e.g. spatial distribution of aftershocks). However, geological observations must be very carefully handled if they are to be reliably used for the estimation of seismic parameters (e.g. earthquake magnitude), because two important issues must be considered, as follows. (1) It is well known (e.g. Wells & Coppersmith 1994; Ambraseys & Jackson 1998) that the surface fault length, Ls, and surface fault slip, u , are usually only a part of the real (subsurface) fault length, L, and fault slip, u. Observed values of Ls and us vary significantly for the same mainshock magnitude. This holds particularly for the estimation of the mean surface slip for which even secondary ground effects have sometimes been considered for the calculation of the mean fault slip. Relative errors can be considerably reduced by excluding outliers through valid statistical procedures, which take into

698

B.C. PAPAZACHOS E T AL.

O O

~a

o

o~ ~az

~o

o~

o

o

~

o

~

.~ 9 m

~

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN Table 3. The estimated origin time, to*, epicentre coordinates, E*(~p, Z), and magnitude, M*, for each of the nine probably ensuing mainshocks in the Mediterranean and surrounding regions Event Area 1 2 3 4 5 6 7 8 9

to*

Pyrenees 2008.1 Calabria 2008.9 NE Adriatic 2008.6 Albania 2008.8 N Greece 2008.1 SE Aegean 2009.6 NW Anatolia 2008.5 W Anatolia 2007.5 NE Anatolia 2008.8

E*(~p, L) 42.8~ 38.5~ 44.1~ 40.1~ 40.3~ 37.2~ 40.9~ 38.8~ 40.3~

02.1~ 16.8~ 15.5~ 20.5~ 24.7~ 26.7~ 27.7~ 29.5~ 39.2~

M* 6.4 6.7 6.5 6.6 6.7 6.6 6.3 6.2 7.0

Model uncertainties are: _+2.5years for the origin time, < 150 km for the epicentre and _+0.4 for the magnitude of each expected mainshock.

consideration relations between geological observations and independently estimated fault and seismic parameters. In the present work, geological observations (Ls, us) are compared with both real fault parameters (L, u) and moment magnitude, M. (2) Relations between fault parameters and earthquake size depend strongly on the type of faulting, as well as on the seismotectonic environment in which these faults are found. For instance, thrust faults from continental regions have different properties from those in subduction regions, where shallow earthquakes occur along low-angle megathrust faults. Recently, Papazachos et al. (2004c) have shown that different scaling relations apply for strike-slip faults, dip-slip faults in continental regions and dip-slip faults in regions of lithospheric subduction. Furthermore, dip-slip faults in continental regions have the same scaling-law behaviour whether they are normal or thrust. It is therefore interesting to investigate the corresponding behaviour of the fault length and surface displacement based on geological observations. In the present work, information collected from field geological observations for the fault length and surface displacement (L~, u0 is correlated with corresponding fault parameters (L, u), as these are defined from independent information (aftershock activity, fault parameters derived from waveform inversion), which can evaluate the 'true' subsurface fault parameters. Their dependence on the moment magnitude, M, is also examined. Following Papazachos

699

et al (2004c), we examine separately the corresponding relations for dip-slip faults (normal or thrust) in continental regions and for strikeslip faults. Unfortunately, no direct geological observations (Ls, u0 are available for faults in lithospheric subduction regions. Geological data for Ls and u~ have been reported by many researchers world-wide (e.g. Wells & Coppersmith 1994; Stock & Smith 2000; Papazachos & Papazachou 2003; Pavlides & Caputo 2004). The values of surface fault length, Ls, and surface fault displacement, us, reported in these papers and for which reliable moment magnitudes were available are presented in Table 4 and have been used here. It should be noted that one could potentially examine the scaling-law behaviour of reverse and normal dip-slip faults separately. However, because of the limited amount of geological field information available for reverse faults in continental areas, we adopted the approach of Papazachos et al. (2004c), who showed that these two faulting types have almost indistinguishable scaling relations for their total length and displacement as these have been derived from independent information. Relation o f f a u l t length to m a g n i t u d e

Figure 4a shows a plot of the surface fault length (dots) as a function of moment magnitude for strike-slip faults. The continuous line corresponds to the relation between the real (subsurface) fault length, L, and moment magnitude (Papazachos et al. 2004c): log L = 0.59M--2.30

(13)

which also appears to fit the data of surface fault lengths for M ~ >7.5. The dashed line is a fit to the data for M < 7.5 using the slope of equation (13), corresponding to the relation log L~=0.59M-2.50, 5.8 _<M <7.5, n=28, ~=0.18

(14)

where n is the number of observations and cr is the standard deviation. Figure 4b shows the surface fault length against moment magnitude for dip-slip faults in continental regions. The continuous line is the relation between the length, L, of the real (subsurface) dip-slip fault and magnitude (Papazachos et al. 2004c): log L = 0 . 5 0 M - 1.86

(15)

which also appears to fit the data of surface fault lengths for M ~ > 7.3, whereas the dashed line is

B . C . P A P A Z A C H O S ET AL.

700

Table 4. Date, epicentre coordinates, moment magnitude, M, fault parameters and region of the 48 dip-slip

earthquakes and 47 strike-slip earthquakes for which data are used in the present study Event Year 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

1904 1906 1920 1928 1928 1929 1931 1932 1939 1941 1941 1943 1944 1944 1947 1949 1951 1952 1953 1954 1954 1954 1954 1954 1955 1956 1957 1957 1958 1959 1962 1964 1966 1966 1966 1966 1966 1967 1967 1967 1968 1968 1968 1969 1970 1970 1971 1971 1971 1973 1973 1975 1976 1976 1977 1978 1978 1978

Date-Time 0404102600.0 0418131200.0 1216120543.0 0414093000.0 0418192248.0 0501153722.0 0810211840.0 0926192042.0 1226235721.0 0216163859.0 0301035247.0 1126222036.0 0201032236.0 0625041619.0 0923122810.0 0822040111.0 1118093543.0 0721115211.0 0318190613.0 0430130236.0 0706111318.0 0824055130.0 1216110712.0 1216111133.0 0716070710.0 0209143239.0 0526063334.0 1204033748.0 0710061556.0 0818063715.0 0901150058.0 1006143123.0 0628042616.3 0628042616.3 0819122210.5 0901142257.0 0912164101.5 0105001440.1 0722165658.0 1130072350.0 0409022900.2 0831104741.3 1014025851.8 0328014829.5 0104170039.4 0328210223.5 0209140040.6 0512062515.0 0522164359.3 0206103707.2 0714045120.0 0906092012.0 0204090143.9 1124122215.6 1219233433.3 0523233411.4 0620200321.5 0916153556.7

Lat. 41.90 38.00 35.79 42.15 42.10 37.70 47.00 40.45 39.50 33.40 39.67 41.00 41.50 39.15 33.70 53.75 31.00 35.10 40.00 39.28 39.50 39.50 39.20 39.20 37.55 31.90 40.67 45.21 58.36 44.67 25.60 40.30 35.88 35.88 39.17 37.39 39.40 48.22 40.67 41.39 33.22 34.15 -31.54 38.55 24.12 39.21 34.40 37.70 38.85 31.33 35.16 38.51 15.28 39.05 30.93 40.73 40.78 33.37

Long.

M

Ls

Us

KF

Region

Reference

23.00 -123.00 105.74 25.28 25.00 57.80 90.00 23.86 38.50 58.90 22.54 34.00 32.50 29.46 58.70 -133.25 90.50 -118.90 27.30 22.29 -118.50 -118.30 -118.00 -118.00 27.15 -115.80 30.86 99.24 -136.34 -110.83 65.22 28.23 -120.42 -120.42 41.56 22.14 -120.16 102.90 30.69 20.46 -116.19 59.01 117.00 28.46 102.49 29.51 -118.43 30.00 40.52 100.49 86.40 40.77 -89.19 44.04 56.48 23.25 23.24 57.44

7.3 7.9 8.0 6.8 7.0 7.3 7.9 7.0 7.8 6.1 6.3 7.6 7.6 6.1 6.8 8.1 7.7 7.4 7.4 7.0 6.2 6.9 7.2 6.9 6.9 6.6 7.0 8.1 7.8 7.3 7.4 6.9 6.4 6.3 6.9 6.0 6.0 7.0 7.4 6.3 6.6 7.2 6.6 6.6 7.3 7.1 6.6 6.2 6.6 7.5 7.0 6.6 7.6 7.2 5.9 5.8 6.5 7.4

25 430 220 38 50 70 180 20 360 12 7 280 180 18 20 440 140 57 18 34 57 45 35 22 40 236 95 27 99 40 38 39 30 2 10 34 54 16 31 80 36 32 48 45 16 38 89 27 26 235 55 12 85

10 330 720 210 740 30 180 10 128 800 60 280 100 25 45 280 210 50 55 650 214 163 18 230 90 40 210 180 150 14 130 50 260 205 12 10 30 150

N S S N N T S N S S N S S N S S S R S N N N S S N S S S S N R S R S N N S S S N S S R N S N S N S S N N S S S N N R

Bulgaria California China Bulgaria Bulgaria Iran China Greece Turke y Iran Greece Turke y Turke y Turkey Iran Queen Charlotte Is. China California Turke y Greece Nevada Nevada Nevada Nevada Turke y Mexico Turkey Mongolia Alaska Montana Iran Turke y California California Turkey Greece California Mongolia Turke y Albania California Iran Australia Turkey China Turke y California Turkey Turke y China China Turkey Guatemala Turkey Iran Greece Greece Iran

7 1 1 7 7 8 1 7 1 8 7 1 1 7 8 4 1,3 1,3 6 6 1,3 1 1 1 7 1 1 1,9 1,3 1 1 6 1 1 1 7 1,3 1,3 1,3 7 1 1 1 1,6 1,3 3,6 1 7 1 1 3 1 1 1,3 1 6 6 1

E A R T H Q U A K E P R E D I C T I O N IN THE M E D I T E R R A N E A N Table 4.

Continued

Event

Year

Date-Time

Lat.

59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95

1979 1979 1979 1979 1979 1980 1980 1980 1980 1981 1981 1981 1981 1981 1981 1982 1983 1983 1983 1985 1986 1986 1987 1987 1987 1988 1988 1988 1988 1988 1988 1989 1990 1992 1995 1995 1999

0602094758.7 0806170522.8 1015231657.9 1114022118.2 1127171033.0 0124190009.2 0709021157.4 1010122522.1 1123183452.2 0123211352.0 0224205337.0 0225023553.5 0304215807.2 0611072425.1 0728172223.0 1213091250.9 1028140607.4 1030041228.1 1222041129.3 1027193457.2 0330085352.0 0913172434.3 0302014234.7 1124015416.7 1124131556.6 0122003558.1 0122035725.6 0122120458.0 1106130319.9 1106131542.0 1207074124.3 1225142432.8 0716072634.7 0628115735.9 0615001549.0 1001155712.6 0817000138.6

-30.73 37.13 32.86 34.03 34.08 37.77 39.29 36.16 40.86 30.89 38.23 38.17 38.24 29.90 29.99 14.67 44.10 40.35 11.85 36.43 -26.30 37.08 -37.93 33.23 33.12 -19.82 -19.78 -19.80 22.80 23.20 40.96 60.07 15.70 34.25 38.27 38.06 40.76

Long. 117.21 -121.51 -115.46 59.81 59.79 -121.70 22.91 1.40 15.33 101.15 22.97 23.12 23.26 57.72 57.77 44.23 -113.81 42.18 -13.51 6.78 132.77 22.15 176.78 -115.65 -116.02 133.86 133.92 133.95 99.59 99.46 44.16 -73.48 121.22 -116.48 22.15 30.15 29.96

M

Ls

Us

6.1 5.8 6.5 6.6 7.2 5.8 6.5 7.1 6.9 6.6 6.7 6.4 6.3 6.6 7.1 6.3 6.9 6.7 6.2 6.0 5.8 6.0 6.5 6.2 6.6 6.3 6.4 6.6 7.1 6.8 6.8 6.0 7.7 7.3 6.4 6.2 7.5

15 14 31 17 65 6 8 31 38 44 19 15 65 15 34 12 9 13 15 18 10 27 10 7 16 35 16 25 10 120 71 7 11 100

50 18 120 21 154 64 60 60 80 45 10 12 54 63 60 93 70 60 295 1 12 300

KF

Region

R S S S R S N R N S N N N R R N N S S S S N N S S R R R S S R R S S N N S

Australia California California Iran Iran California Greece Algeria Italy China Greece Greece Greece Iran Iran N Yemen Idaho Turkey W Africa Algeria Australia Greece New Zealand California California Australia Australia Australia China China Armenia Canada Philippines California Greece Turkey Turkey

701

Reference 1 1,2 1,2,4 1,3 1,3 1 8 1 1 1 6 1 6 1 1 1 1 1 1 1 1 6 1,5 1,4 1,5 1 1 1 1 1 1 1 1 1,2 7 7 6

References: 1, Wells & Coppersmith (1994); 2, Pegler & Das (1996); 3, Stock & Smith (2000); 4, Fujii & Matsuura (2000); 5, Henry & Das (2001); 6, Papazachos & Papazachou (2003); 7, Pavlides & Caputo (2004); 8, Ambraseys & Jackson (1998); 9, present study. K F is the kind of faulting (N for normal fault, R for reverse, S for strike-slip), Ls is the length (in km) of the surface trace of the fault, and us is the mean slip (in cm) on the surface fault trace.

a fit to t h e d a t a (dots) f o r M _< 7.3, c o r r e s p o n d i n g to log L s = 0 . 5 9 M - - 2 . 0 1 , 6 . 0 < M < 7 . 3 , n - - 3 7 , cy = 0.22.

(16)

Relation of fault slip to magnitude F i g u r e 5a s h o w s the d i s t r i b u t i o n o f s u r f a c e f a u l t slip v. m a g n i t u d e f o r strike-slip faults. T h e c o n t i n u o u s line s h o w s s u c h v a r i a t i o n f o r real ( s u b s u r f a c e ) f a u l t slip, u, ( P a p a z a c h o s et al. 2004c):

log u = 0 . 6 8 M - 2 . 5 9

(17)

w h e r e a s t h e d a s h e d line is a fit (using t h e slope o f 0.68) to t h e o b s e r v e d s u r f a c e slips, us, f o r M < 7.6, c o r r e s p o n d i n g to log u s = 0 . 6 8 M - 2 . 7 6 , 5.8 < M _<7.6, n = 2 3 , cy = 0.25.

(18)

F i g u r e 5b s h o w s t h e c o r r e s p o n d i n g v a r i a t i o n o f the s u r f a c e f a u l t slip w i t h t h e m a g n i t u d e f o r dip-slip faults in c o n t i n e n t a l regions. T h e c o n t i n u o u s line is t h e r e l a t i o n b e t w e e n t h e

702

B.C. PAPAZACHOS E T AL. 6.0

6.5

7.0

7.5

8.0

6.0

8.5

(a) Strike slip

6.5

' ' I ....

' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' ' 1 ' ' ' 6 1 0 ' ' '

7.0

7.5

I ' ' ' ' I ....

8.0

I ....

I '

1000

1000

:

(a) Stdke slip

~."

9 9

9 eeS 9

9

100

100

E o 100 1"

1"

100

..I

10

4,

10

-

9

J J

e 9

a

a i , , I ', : : ', I ', ', ', ', I ', ', ', ', I : ', ', ', I : '

- t000

1000

(b) Dip slip

(b) Dip slip

100

100

9

m" 10

1"

lOO

9 i

,,~~

100

9

ej

y

9149

;'-. j

9

a I I

I

b J , l , , , , I , , t ~ l ,

6.0

6.5

7.0 M~

7.5

8.0

8.5

Fig. 4. Variation of observed length (in km) of surface fault traces with the moment magnitude (dots) for strike-slip faults (a) and dip-slip faults (b). Continuous lines show the variation with magnitude of the real (subsurface) fault length, L; dashed lines show the corresponding variation of the surface fault length, L,, for the low-magnitude part of the data (see text for explanation).

real (subsurface) fault slip and the magnitude (Papazachos et al. 2004c), which is log u = 0 . 7 2 M - 2 . 8 2 .

(19)

The relation fitting the surface fault slip data is log u s = 0 . 7 2 M - 3 . 1 8 , 5.8 < M <7.4, n = 29, cr = 0.42.

9

1

. . . . . . . . . . .

(20)

All previous results suggest that the observed fault length and slip both in strike-slip and dip-slip faults in continental areas bear a rather complex relation to the true fault length and width. This is mainly due to the fact that for smaller earthquakes the probability for a surface

I

t I

6.0

i

b ....

6.5

I ....

7.0 M~

I , , , , I , ,

7.5

8.0

Fig. 5. Variation of observed surface fault slip (in cm) with the moment magnitude (dots) for strike-slip faults (a) and dip-slip faults (b). Continuous lines show the variation with magnitude of the real (subsurface) fault slip, u; dashed lines show the corresponding variation of the surface fault slip, u, for the low-magnitude part of the data (see text for explanation).

fault trace becomes increasingly smaller (e.g. Wells & Coppersmith 1994) and when the rupture does reach the surface, the observed fault trace length and slip is a fraction of the true length and slip, which also depends on the mainshock magnitude. To further examine this dependence, in Figure 6a and b the ratio of the observed fault length and slip to the length calculated by equations (13) and (15) for length, and calculated by equations (17) and (19) for slip are presented for both types of faults examined here (strike-slip and dip-slip) as a function of magnitude. As the logarithm of these quantities is typically considered to be related to magnitude, a log-scale is used in the plots. Despite the data scatter, there appears to be a general agreement

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN + '

'

I

I

I

I

I

I

I

I

9

9

O

o,

4

4

_1_ +~ g

.o

4

4

=o

O O

9

*4 4

9

*o*

9

4

$9 ~O * 4 0 ~

9o 4

9

:

9

9

_*_0 4

9

+

~

4 ~O

4

4 9

+

++

~o

9

~

~

40~4 ~

gO

*

9 9

9

4

S 4

Dip-slip

~-

0.1

d _

0.2

__

__

0.2

0

0

,,,,.. --

9 4

9 Strike.slip 1

*

c

n

9

9

~

+

9

!

9

8

_-

_._

= +

go o

_J

0.1

I

b

a

n

I

703

',....,

,,-,,,

--.;

__

...

A

& 8'

-0.2

-0.2 -,

- -

- ~ - - -

- - - -

-0.4

-0.4

-0.6

.... 6.0

6.5

7.0

7.5

8.0

M-->

8,5

6.0

6.5

7.0

7.5

8,0

0,6 8.5

M ---->

Fig. 6. Variation of the observed/true ratio of fault length (a) and fault-slip (b) v. magnitude for strike-slip and dip-slip faults. The corresponding five-point weighted moving average log-curves and their errors are shown in (c) and (d).

between strike-slip and dip-slip events, although the slip observed for dip-slip events shows somewhat smaller values than that for strike-slip events (Fig. 6b). Furthermore, a general increase of the observed/true length and slip ratios with magnitude is observed. To further quantify this magnitude dependence, a five-point weighted moving average window of the observed ratio and its corresponding error has been estimated and is presented in Figure 6c and d. The zero-logarithm ratio line, corresponding to Ls = L and Us= u, is shown in both plots. In Figure 6c, the L+/L ratio is practically constant for 5.8 < M < 7.3, with values increasing from 66% to 7 1 ~ with an average of 69%. For larger magnitudes this LJL ratio increases rapidly and reaches values of c. 1.0 (100%) for magnitudes M > 7 . 6 , which means

that the observed fault length is practically equal to the total fault length, within statistical accuracy. On the other hand, the observed/true slip ratio exhibits a slightly different behaviour, with relatively low values for M < 6 . 4 , increasing (more or less linearly) from c. 40% (M = 5.8) to c. 63% (M=6.4). For magnitudes between 6.3 and 7.5 the Us/Uratio remains practically constant at the value of c. 63% and then rapidly increases to almost 100% at M ~ >7.8-7.9. Therefore, it appears that the observed/true slip ratio exhibits a similar rapid increase to almost 100% with a 'hysteresis' of c. 0.3 magnitude units. This 'hysteresis', together with the much larger dispersion of log (Us~U) values compared with log (L+/L) values and the lack of dip-slip data for M_> 7.5, suggests that the use of observed slips for prediction of expected earthquake magnitude should

704

B.C. PAPAZACHOS E T AL.

be performed with care and additional constraints should be used by field geologists for such estimations. Global relations presented here allow the use of geological observations (L~, us) to reliably estimate measures of earthquake size (M, L, u), necessary for earthquake prediction research. Thus, the location of active faults in the Mediterranean region, and an estimation of the seismic parameters of these faults (L, u, M) by a combined use of geological and seismological information, can be of primary importance for the application of the intermediate-term earthquake prediction method, such as that presented here. A continuous monitoring for the identification of decelerating seismic strain in each of these faults, accompanied by accelerating seismic strain in the broader region, could lead to the prediction of the origin time of the mainshock about to be generated on this fault, and its expected magnitude can be estimated by the fault dimensions.

Discussion and conclusions Decelerating seismic strain caused by the generation of intermediate-magnitude preshocks in a seismogenic region, accompanied by accelerating seismic strain caused by the generation of larger intermediate-magnitude preshocks in the broader (critical) region, is a distinct premonitory seismicity pattern that has led to the formulation of a promising intermediate-term earthquake prediction model. This model, which can be called the 'Decelerating in-Accelerating out Seismic Strain Model', is also supported by theoretical work. This model has been applied retrospectively to a complete set of six strong (M > 6.8) shallow mainshocks, which occurred in the Mediterranean and surrounding region (35~176 5~176 between 1980 and 2003. All of these six mainshocks have been preceded by such patterns, which obey quantitative constraints imposed by the model. This indicates that this premonitory pattern is a general one or at least that the probability for a future large mainshock in the Mediterranean region to be preceded by such a pattern is high. This result, however, does not exclude the possibility for the occurrence of such patterns that are not followed by mainshocks (false alarms). The model has been also applied to identify such currently occurring patterns in the Mediterranean and surrounding region, which indicate the probable generation of such strong (M >_6.2) mainshocks in this area. Nine such patterns have been observed and the estimated (predicted) parameters (origin time, epicentre coordinates,

magnitude) of the corresponding, probably ensuing mainshocks are listed in Table 3. The two standard deviation model uncertainties are _+2.5 years for the origin time, _+0.4 for the magnitude and less than 150 km for the epicentre. There is a probability confidence of about 75% for the occurrence of each of these mainshocks, indicated by tests on synthetic catalogues (Papazachos et al. 2002, 2004b), whereas the probability for random occurrence is of the order of 10%. It should be noted that the approach used here follows that adopted in our previous studies (e.g. Papazachos & Papazachos 2000, 2001; Papazachos et al. 2004a,b, 2005, 2006), which is based on the critical earthquake model that is widely used in related work (e.g. Bufe & Varnes 1993; Bowman et al. 1998; Jaume & Sykes 1999). However, the results obtained in the present work represent our first attempt at a forward test of the proposed model, where observations during the next decade will show the degree of validity of this forward test and of the applied method. Globally valid relations have been derived between the surface fault length, L~, derived from geological observations and real (subsurface) fault length, L, and the mainshock magnitude, M, separately for strike-slip faults and for dipslip faults in continental regions. Similar relations have been also derived for the surface fault slip, Us. From the results of the prediction model previously presented, it is clear that the associated uncertainties and false-alarm rates can often lead to ambiguous results concerning the candidate neotectonic fault that will produce the expected earthquake. For this reason, it is necessary to approach the proposed method as a tool for intermediate-term time-dependent seismic hazard estimation, rather than as an accurate individual earthquake prediction method. However, the previously defined scaling relations may even help to narrow the number or spatial extent of neotectonic faults that can be associated with expected earthquakes. Thus geological observations can be used to locate active faults in the Mediterranean and through such relations to define the dimension of each such fault and the magnitude of the next oncoming mainshock in the fault. Continuous seismological monitoring of all known active faults in the Mediterranean and the identification of seismogenic regions where decelerating strain occurs (when accelerating strain occurs in the broad critical region) can lead to prediction of the next mainshock in each of these faults. The advantage of this method is that it requires data of intermediate-magnitude shocks (e.g. M>4.0), which are easily obtained by the existing seismological networks.

EARTHQUAKE PREDICTION IN THE MEDITERRANEAN Because the potential o f the m o d e l examined for predicting mainshocks is still u n d e r investigation, the present w o r k must be considered as part o f the model testing. Independently, however, of the model capacity for predicting mainshocks, it is of importance for dealing with the p r o b l e m of t i m e - d e p e n d e n t seismic hazard assessment, as this m o d e l reliably defines critical regions, which are at a seismic excitation for a predicted time interval as a result of the generation of interm e d i a t e - m a g n i t u d e shocks. The largest of these shocks (with M,,~6) often cause considerable d a m a g e w h e n their epicentres are on land (e.g. the Athens 1999 M = 5.9 shock killed 143 people and caused extensive destruction). The results of the present work, which concern expected mainshocks, will be u p d a t e d regularly as new observations are expected to accumulate over the next few years. Eventually, any significant change o f the currently observed seismicity behaviour will be noticed and taken into consideration during the final tests. We would like to thank Wessel & Smith (1995) for freely distributing the GMT software that was used to produce the maps of the present study. This research was partially supported by the Earthquake Protection and Planning Organization of Greece (project 20242, Res. Comm. AUTH). This paper benefited from the comments of two reviewers.

References AMBRASEYS, N. N. & JACKSON, J. A. 1998. Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region. Geophysical Journal International, 133, 390406. ANDERSON, H. & JACKSON,J. 1987. Active tectonics of the Adriatic Region. Geophysical Journal of the Royal Astronomical Society, 91, 937-983. ANDRIEUX, J., FONBOTE,J. M. & MATTAUER,M. 1971. Sur un mod6 le explicatif de l'Arc de Gibraltar. Earth and Planetary Science Letters, 12, 191-198. BOWMAN, D. D., QUILLON, G., SAMMIS, C. G., SORNETTE, A. & SORNETTE, D. 1998. An observational test of the critical earthquake concept, Journal of Geophysical Research, 103, 24359-24372. BUFE, C. G. & VARNES,D. J. 1993. Predictive modeling of seismic cycle of the Great San Francisco Bay Region. Journal of Geophysical Research, 98, 9871-9883. BUFE, C. D., NISHENKO, S. P. & VARNES, D. J. 1994. Seismicity trends and potential for large earthquakes in Alaska-Aleutian region. Pure and Applied Geophysics, 142, 83-99. BUFORN, E., UDIAS,A. & MEZCUA,J. 1988a. Seismicity and focal mechanisms in South Spain. Bulletin of the Seismological Society of America, 78, 2008-2024.

705

BUFORN, E., UDIAS, A. & COLOMBAS, M. A. 1988b. Seismicity, source mechanisms and tectonics of the Azores-Gibraltar plate boundary. Tectonophysics, 152, 89-118. 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. 1988. Extensional collapse of orogens. Tectonics, 7, 1123-1139. DEWEY, J. F. & ~ENGOR, A. M. C. 1979. Aegean and surrounding regions: complex multiplate and continuum tectonics in a convergent zone. Geological Society of America Bulletin, 90, 84-92. EVISON, F. F. 2001. Long-range synoptic earthquake forecasting: an aim for the millennium. Tectonophysics, 333, 207-215. FACCENNA, C., DAVY, P., BRUN, J. P., FUNICIELLO,R., GIARDINI, D., MATTEI, M. & NALPAS, T. 1996. The dynamics of back-arc extension: an experimental approach to the opening of the Tyrrhenian Sea. Geophysical Journal International, 126, 781-795. FACCENNA, C., BECKER, T. W., LUCENTE, F. P., JOLIVET, L. & ROSSETTI, F. 2001. History of subduction and back-arc extension in the Central Mediterranean. Geophysical Journal InternationM, 145, 809-820. FuJII, Y. & MATSUURA, M. 2000. Regional difference in scaling laws for large earthquakes and its tectonic implication. Pure and Applied Geophysics, 157, 2283-2302. GASPARINI, G., IANNACCONE,G. & SCARPA, R. 1985. Fault-plane solutions and seismicity of the Italian peninsula. Tectonophysics, 117, 59-78. HAINZL, S., ZOLLER, G., KURTHS, J. & ZSCHAU, J. 2000. Seismic quiescence as an indicator for large earthquakes in a system of self-organized criticality. Geophysical Research Letters, 27, 59%600. HENRY, C. & DAS, S. 2001. Aftershock zones of large shallow earthquakes: fault dimensions, aftershock area expansion and scaling relation. Geophysical Journal International, 147, 272-293. JACKSON, J. & MCKENZIE, D. 1988. The relationship between plate motions and seismic moment tensors and the rates of active deformation in the Mediterranean and Middle East. Geophysical Journal of the Royal Astronomical Society, 93, 45-73. JAUMI~, S. C. & SYKES, L. R. 1999. Evolving towards a critical point: a review of accelerating seismic moment/energy release rate prior to large and great earthquakes. Pure and Applied Geophysics, 153, 279-306. JIMENEZ-MUNT, I. & NEGREDO, A. 2003. Neotectonic modeling of the western part of the Africa-Eurasia plate boundary: from the mid-Atlantic ridge to Algeria. Earth and Planetary Science Letters, 205, 257-271. JOLIVET, L. & FACCENNA, C. 2000. Mediterranean extension and the Africa-Eurasia collision. Tectonics, 19, 1095-1107. KAHLE, H.-G., STRAUB,C., REILINGER, R., et al. 1998. The strain field in the eastern Mediterranean estimated by repeated GPS measurements. Tectonophysics, 294, 237-252.

706

B.C. PAPAZACHOS ET AL.

KATO, N., OTHAKE,M. & HIRASAWA,T. 1997. Possible mechanism of precursory seismic quiescence: regional stress relaxation due to preseismic sliding. Pure and Applied Geophysics, 150, 249-267. KNOPOFF, L., LEVSHINA,T., KEYLLIS-BOROK,V. J. & MATTONI, C. 1 9 9 6 . Increased long-range intermediate-magnitude earthquake activity prior to strong earthquakes in California. Journal of Geophysical Research, 101, 5779-5796. LEPICHON, X. & ANGELIER, J. 1979. The Hellenic arc and trench system: a key to the neotectonic evolution of the eastern Mediterranean area. Tectonophysics, 60, 1-42. MALINVERNO,A. & RYAN, W. B. E. 1986. Extension in the Tyrrhenian Sea and shortening in the Apennines as result of arc migration driven by sinking of the lithosphere. Tectonics, 5, 227-245. MCCLUSKY, S., BALASSANIAN, S., BARKA, A., et al. 2000. Global Positioning System constraints on plate kinematics and dynamics in the eastern Mediterranean and Caucasus. Journal of Geophysical Research, 105, 5695-5719. MCKENZIE, D. P. 1970. The plate tectonics of the Mediterranean region. Nature, 226, 239-243. MCKENZIE, D. P. 1972. Active tectonics of the Mediterranean region. Geophysical Journal of The Royal Astronomical Society, 30, 109-185. MEIJER, P. T. & WORTEL, M. J. R. 1997. Present-day dynamics of the Aegean region: a model analysis of the horizontal pattern of stress and deformation. Tectonics, 16, 879-895. MERCIER, J. L., CAREY, E., PHILIP, H. R. & SOREL, D. 1976. La n6otectonique plio-quaternaire de l'arc eg6en externe et de la Mer eg6en et ses relations avec seismicit6. Bulletin de la Soci~tk G~ologique de France, 18, 159-176. MERCIER, J. L., SOREL, D., VERGI~LY,P. 8~ SIMEAKIS, K. 1989. Extentional tectonic regimes in the Aegean basins during the Cenozoic. Basin Research, 2, 49-71. MOGI, K. 1969. Some features of the recent seismic activity in and near Japan II. Activity before and after great earthquakes. Bulletin of Earthquake Research Institute, University of Tokyo, 47, 395-417. PAPADOPOULOS, G. A. 1986. Long term earthquake prediction in western Hellenic arc. Earthquake Prediction Research, 4, 131-137. PAPAZACHOS,I . C. 1990. Seismicity of the Aegean and surrounding area. Tectonophysics, 178, 287-308. PAPAZACHOS, B. C. & PAPAZACHOS,C. B. 2000. Accelerated preshock deformation of broad regions in the Aegean area. Pure and Applied Geophysics, 157, 1663-1681. PAPAZACHOS, I . C. & PAPAZACHOU, C. B. 2003. The Earthquakes of Greece. Ziti, Thessaloniki. PAPAZACHOS, B. C., PAPADIMITR1OU, E. E., KIRATZI, A. A., PAPAZACHOS,C. B. & LOUVARI, E. K. 1998. Fault plane solutions in the Aegean Sea and the surrounding area and their tectonic implications. Bollettino di Geofisica Teorica ed Applicata, 39, 199-218. PAPAZACHOS, B. C., SCORDILIS, E. M., PANAGIOTOPOULOS, D. G., PAPAZACHOS, C. B. & KARAKAISIS,G. F. 2004c. Global relations between

seismic fault parameters and moment magnitude of earthquakes. Bulletin of the Geological Society of Greece, 25, 1-8. PAPAZACHOS, B. C., COMNINAKIS,P. E., SCORDILIS,E. M., KARAKAISIS,G. F. & PAPAZACHOS,C. B. 2005.

A catalogue of earthquakes in the Mediterranean and surrounding area for the period 1901-2004. Publication Geophysical Laboratory, University of Thessaloniki. http://lemnos.geo.auth.gr/the_ seisnet/medcatalog.txt PAPAZACHOS,C. B. & KIRATZI, A. A. 1996. A detailed study of the active crustal deformation in the Aegean and surrounding area. Tectonophysics, 253, 129-153. PAPAZACHOS,C. B. 8z PAPAZACHOS,B. C. 2001. Precursory accelerating Benioff strain in the Aegean area. Annali di Geofisica, 144, 461474. PAPAZACHOS,C. B., KARAKAISIS,G. F., SAVVAIDIS,A. S. & PAPAZACHOS,B. C. 2002. Accelerating seismic crustal deformation in the southern Aegean area. Bulletin of the Seismological Society of America, 92, 570-580. PAPAZACHOS, C. B., KARAKAISIS, G. F., SCORDILIS, E. M. & PAPAZACHOS,B. C. 2004a. Probabilities of activation of seismic faults in critical regions of the Aegean area. Geophysical Journal International, 159, 679-687. PAPAZACHOS,C. B., SCORDILIS,E. M. KARAKAISIS,G. F. and PAPAZACHOS, B. C. 2004b. Decelerating preshock seismic deformation in fault regions during critical periods, Bulletin of the Geological Society of Greece, 36, 1-9. PAPAZACHOS, C. B., KARAKAISIS,G. F., SCORDILIS,E. M. & PAPAZACHOS, B. C. 2005. Global observational properties of the critical earthquake model. Bulletin of the Seismological Society of America, 95, 1841-1855. PAPAZACHOS, C. B., KARAKAISIS,G. E., SCORDILIS,E. M. & PAPAZACHOS,B. C. 2006. New observational information on the precursory accelerating and decelerating strain energy release. Tectonophysics, doi: 10.1016/j .tecto.2006.03.004. PAVLIDES, S. 8l; CAPUTO, R. 2004. Magnitudes versus faults' surface parameters: quantitative relationships from the Aegean region. Tectonophysics, 380, 159-188. PEGLER, G. & DAS, S. 1996. Analysis of the relationship between seismic moment and fault length for large crustal strike-slip earthquakes between 197792. Geophysical Research Letters, 23, 905-908. SCORDILIS, E. M. 2006. Empirical global relations for MS, mb, ML, and moment magnitude. Journal of Seismology, doi: 10.1007/S10950-006-9012-4. SORNETTE, A. & SORNETTE, D. 1990. Earthquake rupture as a critical point. Consequences for telluric precursors. Tectonophysics, 179, 327-334. SORNETTE, D. • SAMMIS, C. G. 1995. Complex critical exponents from renormalization group theory of earthquakes: implications for earthquake predictions. Journal of Physics Series/., 5, 607-619. STOCK, C. & SMITH, E. G. V. 2000. Evidence for different scaling of earthquake source parameters for

E A R T H Q U A K E P R E D I C T I O N IN THE M E D I T E R R A N E A N large earthquakes depending on fault mechanism. Geophysical Journal International, 143, 157-162. SYKES, L. R. & JAUME, S. 1990. Seismic activity on neighboring faults as a long term precursor to large earthquakes in the San Francisco Bay area. Nature, 348, 595-599. TAYMAZ, T., JACKSON, J. & MCKEYZlE, D. 1991. Active tectonics of the north and central Aegean Sea. Geophysical Journal International, 106, 433-490. TOCHER, D. 1959. Seismic history of the San Francisco bay region. California Division of Mines Special Report, 57, 39-48. TZANIS, A., VALLIANATOS, F. & MAKROPOULOS, K. 2000. Seismic and electrical precursors to the 17-1-1983, M = 7 Kefallinia earthquake, Greece, signatures of a SOC system. Physics and Chemistry of the Earth, 25, 281-287. WELLS, D. L. t~ COPPERSMITH,K. J. 1994. New empirical relationships among magnitude, rupture length,

707

rupture width, rupture area and surface displacement. Bulletin of the Seismological Society of America, 84, 974-1002. WESSEL, P. & SMITH, W. 1995. New version of the Generic Mapping Tools. EOS Transactions, American Geophysical Union, 76, 329. WORTMANN, U. G., WEISSERT, H., FUNK, H. & HAUCK, J. 2001. Alpine plate kinematics revisited: the Adrian problem. Tectonics, 20, 134-147. WYSS, M. 1997. Cannot earthquakes be predicted? Science, 278, 487-488. WYsS, M. ~; HABERMANN, R. E. 1988. Precursory seismic quiescence. Pure and Applied Geophysics, 126, 319-332. WYss, M., KLEIN, F. & JOHNSTON, A. C. 1981. Precursors of the Kalapana M = 7.2 earthquake. Journal of Geophysical Research, 86, 3881-3900. ZOLLER, G., HAINZL, S., KURTHS, J. • ZSCHAU, J. 2002. A systematic test on precursory seismic quiescence in Armenia. Natural Hazards, 26, 245-263.

Index Page numbers in italic denote figures. Page numbers in bold denote tables. accretionary wedge doubly vergent model, Apulian margin 509,510, 512, 516 Lycian Nappes 460462 see also pro-wedge Adria 15, 16, 155, 159 Cenozoic 173 Cretaceous 171,172, 174 indentation 174 Jurassic motion 23-24, 170 Palaeozoic 167 Permo-Triassic reassembly 17-18,167 Triassic-Jurassic 167-I 69 Aegean, extensional province 557,558, 577,579, 591,592, 671, 683 Aegean, south tectonic models 93-148 Cenozoic setting 101,102 recognizing settings 95 tectonic evolution 101,144-148 Africa-Arabia plate motion 614, 615 Africa-Europe plate motion 11,25, 25-27, 27, 30, 309-310, 311, 691 and ophiolite creation and emplacement 26, 28, 29-31,29 reconstruction 15-21 Agios Dimitrios fault 665-666, 666 Albania geodynamic evolution 544 geology 539, 540, 541 544 ophiolites 267-296, 304 geology 268, 269-272 Albanides 540,541 fission-track thermochronology 544-554 Alboran fragment 14,15 Alborz 184, 185, 191 Alcx-Kelel~i Fault Zone 599, 601 Aliakmonas fault zone 663-664, 666, 667 Almopias Ocean 378 subduction 381,408 Almopias Zone 373-374, 375,376, 377, 406-408 foreland basin 403,404, 405 late Jurassic transgression 389,393-396, 397 metamorphism 388, 393, 395 oceanic crust genesis 380, 381,398-399 Palaeogene suturing 405406 passive margin subsidence 380, 398,401 sea-level rise 401 subduction 403 Triassic rifting 378, 408 amphibole blue, Pindos Flysch 502 K6mtirhan ophiolite 338, 341,343 Vourinos 246 amphibolite, Balkan Peninsula 169-170 Anatolia basin development 591-608 northwest, Palaeozoic terranes 53, 58-63 ophiolites 305,327-346 plate motion 309, 345

Anatolide-Tauride fragment 14, 15, 21 Ankara-Erzincan suture 305 see also Izmir-Ankara-Erzincan suture zone Anthemountas fault zone 661,662,666-667 apatite, fission-track ages 544-554 Apulia 15,507, 542 subduction 493,507-518 see also Adria Apulia-Pelagonia suture 493,507-518,508, 516 Arabia ophiolites 305 308 plate motion 309, 345, 614, 628,630, 691 Arakapas Fault Belt 353, 360 Armorican Terrane Assemblage 52, 6~63 Arna unit, metabasic igneous rock 135 Arnea granite 38, 39 Aspropotamus complex 242 243,244,249 Atlantic Ocean, north, continental break-up 11, 12-15, 21, 23 augengneiss, biotite 37, 39 Avalonian Terrane 52, 61-63 Avdella m~lange 240, 243,244, 384, 388 Axios Zone see Vardar Zone Ay~kayas~ Formation 426, 429, 428, 430 back-arc basins early Mesozoic, Black Sea-Caucasus 179-196 Jurassic 185-193 Triassic 181-185 Jurassic, Vardar Zone 381 Baar-Bassit ophiolite 306, 318,352, 354-355 age 366 palaeomagnetic studies 351,353,355, 357, 363-364 rotation 363-364 Balkan Peninsula evolution 155-175 Cambrian-Devonian 156, 159, 167-168 Carboniferous-Permian 156, 166-167,168 Cretaceous 158, 171-173 Maastrichtian-Cenozoic 158, 173-174 rotation 173-174 Triassic-Jurassic 157, 167-171,168,169 geology 157, 160, 159, 162, 159 oceanic crust 166 ophiolites 165 terranes 156--158, 157, 162, 159, 164, 167-173 Balkan Terrane 53 correlation with Istanbul Terrane 61 origin 62-63 Gondwanan 56, 57 palaeogeography 57 stratigraphy 55, 56-57 Baltica 52, 56, 61 Balyata~i Formation 615 617, 618 Bansko fault 674, 677, 679, 680, 682 basalt mid-ocean ridge see MORB Pindos thrust sheets, geochemistry 472474 Pontides, geochemistry 4204-23,424, 425, 431432, 432

710

INDEX

basement, pre-Alpine Crete 69-88 Levantine Basin 206-209 Serbo-Macedonian Massif35-48 basin-ophiolite relationship 309-319 Baskil arc magmatic unit 328,329, 331 biochronology, Mesozoic radiolarites, Lycian Mrlange 229-234 Bitlis Suture Zone 614, 615,626 Black Sea-Caucasus early Mesozoic back-arc basins, evolution 179-196 Blagoevgrad Basin 560-569, 578 Blagoevgrad fault 680, 683 blueschist 449-450 boninite 22, 23, 312 Borlu, Lycian Nappe klippe 453,455, 456 Bozda~ fault 594, 595, 596, 600 Bulgaria Palaeozoic terranes 53, 54-57, 61-63 southwest active faults 671-684 geology 672 late Cenozoic extension 557-586, 559, 672 regional kinematics 578-581 slip rates 577-578 vertical crustal motion 581 584 mammal fossils 560, 565 566, 569, 575 sedimentary correlation 575-576 tectonics 672, 673 Cadomian magmatism 59, 61 Calabria 14, 15 t~ameli Basin 595, 596 evolution 596-610 early-mid Pliocene 599-602 fault kinematics 604-605,606--607, 608 late Miocene 594-597 latest Pliocene 601 Quaternary 601-602 ~ameli Formation 594,595, 596, 598 Carboniferous-Jurassic, basement evolution, Crete 69-88 Carboniferous-lower Triassic succession western Crete 109-119 interpretation 117-119 sediment chemistry 111, 112 Carboniferous-Permian, Balkan Peninsula 166-167 carpholite, Fe-Mg 448,449, 452-453,452, 453, 454458, 459, 460-462 Caucasus, early Mesozoic back-arc basins 179-196 Cenozoic late, southwest Bulgaria, extension 557 586 South Aegean region 102 thrust belts 96 Central Bosnian Mountain terrane 159,164, 166-167 Chalkidiki Peninsula 36, 37 Chamezi crystalline complex 71, 81,83, 84, 87, 88, 119 garnet zonation 79-81 microfabrics 73, 79 Phyllite-Quartzite unit 120-123 structural evolution 77 chromite, Vourinos 246, 253 Cibyra Fault Zone 601-602 Cimmerian orogeny 21, 92, 105, 107, 119, 131,138, 139, 184

Circum-Rhodope belt 37,157 see also mrlange, Pirgadikia Unit ~ivril, Lycian Nappe klippe 453455, 457, 459, 460 clinopyroxene 272-273, 274, 278, 279, 283, 293-294 geothermobarometry 283-284, 286-287, 289-290, 291 K6miirhan ophiolite 337, 341,343 conglomerate Mana unit 115-117, 118-119 Tripokefala beds 123 Corsica 14, 15 MORB ophiolites 30 Crete central, Phyllite-Quartz unit 119-130 eastern, Phyllite-Quartz unit 119-132 geology 70-71 pre-Alpine basement 69-88 tectonic model 85-88 U-Pb dating 69-70, 71, 74-76 sedimentary studies 104-132 tectonostratigraphy 102-104 western, Phyllite-Quartz unit 109-119 Crimea, deformation 192 crust continental 157 oceanic Balkan Peninsula 166 transitional 13 see also Almopias Zone, oceanic crust generation cumulates, ultramafic Albanian ophiolites 269,270, 271,272-296 geochemistry 276, 279, 281,284 Cycladic Blueschist Complex 449~,50, 451

Cyprus geology 358 mid-Cretaceous ophiolites 20, 24-25, 30-31 see also Troodos ophiolite complex Cyprus Arc 613,614 Dalmatian-Herzegovina Composite terrane 159, 165 Dead Sea Fault Zone 355, 581, 615,616,617, 629 deformation and HP-LT metamorphism 458 kinematics, Mesohellenic Trough 524-529 synsedimentary, Hatay Graben 621-622 De~ne Member, ~ameli Formation 594, 598 Delchevo Formation 569, 571,573,578 Derindere Member, t~ameli Formation 594, 598 Devolli ophiolite 269, 270, 271 272, 273,287 Devonian, Balkan Peninsula, terranes 161-162 Dilek Peninsula 449-450, 451,452 Lycian Nappe klippen 452-453, 460 Dinaridic ocean basin 159,165, 169-170 closure 169-170 Dirmil fault 594, 595, 596,600 Dobrovo fault 677, 679, 680, 681-.682 Dotsikos strip ophiolite 249-250 Drama-Prosotsani fault zone 656-659, 666 Dramala complex 242 243,244 crustal section 248-249 fabric analysis 254 kinematic zones 251-252 mantle section 246 Drina-Ivanjica terrane 159, 164, 166-167, 170

INDEX dunite, Pindos-Vourinos ophiolite 246,247, 248,249 dykes Albanian ophiolite 269 Hatay ophiolite 354 K6miirhan ophiolite 331 Pindos-Vourinos ophiolite 246, 247,248 249, 259,260 Troodos ophiolite 353 Dzherman detachment 562, 584, 680 earthquakes Izmit 635-646 northern Greece 649, 651,666-667 prediction 689-705 geological observations 697, 699-704 preshocks 689, 691,692-693 seismic strain model 692-697 accelerating 692-693, 694 decelerating 693-694 southwest Bulgaria 671-672,674 East Anatolian Fault Zone 577, 580, 613,614, 615, 629 East Bosnian-Durmitor terrane 159,164, 169 East Coast Magnetic anomaly 12, 14 East Serbian Carpatho-Balkanides 166, 167 Eastem Hellenides Platform 159, 161 Ecemis Fault Zone 614, 615 Elam~ area geochemistry 333,334-338, 339-341 mineral chemistry 341-344 petrography 331-333 regional geology 328 331 tectonic model 345, 346 Elazl~ magmatic unit 328, 329, 331 Epirus area, convergent model 513 514 Eptahori Formation 512, 512,523, 524, 524-526,527, 528, 529, 530, 532, 534, 535 Eratosthenes Seamount 15, 203, 206,207,217,220, 222 E~en t~ay Basin 592, 605, 607 Eskikry Formation 414-415,416, 4 t 7 Eurasia 9142, 139, 160 active margin evolution 413~140 Black Sea-Caucasus terranes 179-181,184,194, 196 Cretaceous 171 Gondwana suture zone 92, 155, 159, 159 see also Izmir-Ankara-Erzincan suture zone Maastrichtian-Paleocene 173 Palaeozoic 161 evaporite, Messinian Hatay Graben 616, 620, 626 Levantine Basin 204-205, 206, 213,218~19 Evia 138, 139, 140, 141 exhumation, Vardar Zone 391-396, 406 extension late Cenozoic, southwest Bulgaria 557 586 Neogene, southwest Anatolia 591 608 Pernaian 19 Permo-Triassic 148 Faulting active northern Greece 649-668 southwest Bulgaria 671-684 earthquake prediction 697, 699~04 Hatay Graben 621,623-625

711

Mesohellenic Trough 524-529 seismogenic, Izmit 635-646 southwest Anatolia 592-593,597, 601,602-603 southwest Bulgaria 559-558,559, 56(~586 low-angle normal 565, 573,584-585 Fethiye-Burdur Fault Zone 591,592-594 flysch 24 Albania 542 Balkan terrane 56-57, 61 Epirus area 511 Krania unit 398, 510 Pindos 243,244, 510-511 Vai 127, 128, 131 Variscan 58 Fodele unit 105 fossils mammal late Cenozoic, Bulgaria 560, 565-566, 569, 575 Neogene, ~ameli Basin 594, 595,596, 597, 601 micro, Cenozoic, Hatay Graben 615,617, 617, 619, 620 gabbro Albanian ophiolites 268, 269,270, 271,273,274, 277, 278, 280, 282 geochemistry 281,283, 289, 290 MOR or SSZ 291-296 BaEr-Bassit ophiolite 355 Hatay ophiolite 354 K6mfirhan ophiolite 331,336-338, 341,343,344 Galicia Bank, peridotite ridge 13 garnet, chemistry, Cretan basement 73-74, 79-81,85 Gavrovo-Tripolitza carbonate platform 134-138,148, 471, 481-485 Gavrovo-Tripolitza zone 467, 468, 483, 493, 507,508, 509, 511, 514 deformation 484485 geochronology, Serbo-Macedonian Massif 37, 4~42, 44, 45 geothermobarometry, clinopyroxene 283-284,286-287, 289-290, 291 Global Palaeomagnetic Database 18 gneiss biotite 37 granitic 135, 137 G6kstm ophiolite 327, 344 Golitza fault 187 Gondwana Eurasia suture zone 92, 155, 159, 415 northern margin 48 in Palaeotethys tectonic models 91-93,144, 379 and pre-Alpine basement 48, 86-88 separation of Pelagonian microcontinent 91-92, 93,105, 139, 379 terranes see terranes, Gondwana-derived Gorno Spanchevo Fault 572, 573, 578, 584 Gotse Delchev basin 566, 574-575, 577 Gotse Delchev fault 575,577 Gradevo-Predela fault 674, 677, 680, 682 Grande Kabylie 14, 15 graptolites, Balkan Terrane 56-57 Greater Caucasus back-arc basin 179, 183, 189,190, 193,194, 195 deformation 191,192 Greece, northem fault stress directions 652-653,655-666

712 fault-plane solutions 651-652 geotectonics 65~651,654 seismicity 649,651 652 Greenland-western Europe margin 13, 14 Guevgueli back-arc basin 187, 381,391,408 gypsum Chamezi area 123 see also evaporite, Messinian, Hatay Graben harzburgite Central Pontides 425 SSZ ophiolites 268-269, 272, 273, 306, 307, 312-316 Vourinos 246, 303 see also ophiolites, harburgitic Hatay Graben 613,614, 616,623 comparison with Tauride Mountains 626, 628-629 fault kinematics 623-625, 626, 627, 628 sedimentary evolution 613-619, 623-624 synsedimentary deformation 621-622 tectonic models 629-631 Hatay ophiolite 352,354-355,615 age 366 palaeomagnetic studies 351,353,355,357, 361-363,364 rotation 363, 364 see also Klzdda~ ophiolite Hawasina basin 307,308 Heletz fault 215, 221 Hellenic orogen 48 Hellenic-Dinaric orogen, ophiolites 20, 22-24, 303-305, 315 Hellenides 522 External, orogenic mode1507 518 Internal 35, 36, 48 Hercynian orogeny 92, 100, 144-146 deformation and metamorphism 144-145 Herodotus Basin 15,19 hotspots 11, 15 Iceland 14 Hun Terrane 48, 92 Iberia 15 see also Newfoundland-Iberia margin i kigam Formation 415,416, 417, 418,420, 421,422,423 Ionian zone 507, 508, 509-512, 514, 542 Iraklion, Phyllite-Quartzite unit l 19,130 Iran late-Triassic back-arc rifting 185,188, 195 mid-Jurassic deformation 189 191 island arc Balkan Terrane 56, 157 tholeiite 22, 23, 344, 345, 393 Isparta Angle 591-592, 593 Istanbul terrane 53 correlation with Balkan Terrane 61 origin 62-63 palaeogeography 58 59 stratigraphy 58, 60 Izmir-Ankara suture zone 229, 233,234, 457 Izmir-Ankara-Erzincan suture zone 413-442, 414 lzmir-Ankara-Sevan basin 185, 186, 187,191,192, 193,194, 195-196 lzmit earthquake 635-646 fault segments 635 636, 637 trenching sites 639-643 14C dating 643-646

INDEX Jadar block terrane 159,164, 167 Jurassic Black Sea-Caucasus back-arc basins 185 193 early, geology 21 late, reconstruction 24 mid, ophiolites 21-24 Kadlklzl Formation 418,420 Kakopetria detachment fault 353,358 Kalavros crystalline complex 71,81 82, 84, 87, 88 garnet zonation 79 81 microfabrics 78, 79 structural evolution 77 Kalimantsi Formation 565-569, 571,572, 573,578, 584 Kalin granite pluton 560, 562 Karada[g Formation 425,426, 429, 428, 431 Karakaya accretionary complex 181,414 Karaova Formation 450, 452, 453,454, 456, 457-458, 459, 460 Karasu Rift 613,614 Karayaprak Mdlange 424, 430 Kataraktis Passage Member 471,494495, 496, 497,498,499, 500 palaeocurrents 470, 497, 501 palaeogeography 501-502 Kato-Loutraki Fault 378 Kavala-Xanthi-Komotini fault zone 655,656,666,667 Kayaaltl Formation 450 Kerdillion Unit 36, 37 Kerkini fault zone 660-661,666-667 Kirazba~i M61ange 418, 419-420, 423,433 Kir~ehir fragment 14, 15, 19, 21,184-185, 186 Klzdda(g ophiolite 306, 312, 315, 354 see also Hatay ophiolite Klzillrmak Ophiolite 415, 416, 417, 418,420, 421,422,423 klippen, Lycian Nappes 451,452-454 Klissochori unit 380, 393,395,396,397, 401,403 Kocaeli basin 193 K6miirhan ophiolite 327, 328,329 geochemistry 333, 334-335, 339 341,342 mineral chemistry 336-338, 341 petrography 331-333 Kopaonik block and ridge unit 159, 170 171 Korab-Western Macedonian terrane 159 Kraishte region 53, 57 Krania Formation 233,523, 524,525 Krania Unit 398 399,400 flysch 389, 510, 512, 533, 535 Kresna fault 680, 682 Kroupnik earthquake 565, 671 672, 674 Kroupnik normal fault 565, 566, 567, 568, 577, 671 673, 674 675, 676, 678, 680 Ku6aj terrane 157, 164 motion 163 164, 167 Kumaf~an Member, (2ameli Formation 594, 598 K/ire basin see Tauric back-arc basin Kyustendil normal fault 562, 577,677, 678, 680,681 Levantine basin 15, 19, 201-223, 202 crystalline basement 206-209 depositional supersequences 204-205,209-210, 211-214, 215-220 Neotethyan rifting 220 222 seismic stratigraphy 204-220

INDEX structure 203 Syrian Arc inversion 222-223 tectonic evolution 221~223 models 203-204 lherzolite 246, 268-269, 272, 273, 312, 316-317 see also ophiolites, lherzolitic Ligurian Sea 22, 23-24, 311 Liki-Margarita unit 393,397, 403 Limassol Forest Complex 353, 357, 359,360, 361,362 limestone, 'Bellerophon-type' 167 Liri unit 139,140 Livadia Unit 381 Lycian Nappes 450 accretionary wedge geometry 460-462 Cameli Formation 594 geology 450, 452 HP-LT metamorphism 447-462 mineral chemistry 455-457 klippen 451,452-454 Lycian Ophiolitic Mdlange 230~31,450, 452, 458 Mesozoic radiolarites 229-234 Lycian Thrust Sheets 230, 450, 452,454 peridotite 231,450, 458 Maastrichtian-Cenozoic, Balkan Peninsula 158, 173-174 Macedonia Central, fault stress directions 659-662 Eastern, fault stress directions 653,655-659 Western, fault stress directions 662~566 Maden unit 328,329 magma generation, at subduction zones 315-316, 317 magmatism alkaline, Crete 109, 112, 118 Cadomian 59, 61 magnesiocarpholite see carpholite, Fe-Mg magnetization 356-357 Malatya-Keban metamorphic unit 328-329, 344, 346 Malatya-Ovaclk Fault Zone 577, 578,580 Mamonia Complex 354 Mana unit 115-117, 118-119 Mani unit 134-135 mantle, Pindos-Vourinos ophiolite 246-248, 251 mantle wedge 315-316 marble Dobrostan 569, 571,574, 574 Mana unit 115-117, 118 Pirgadikia unit 39 Rhizarion 380-381 Vassilikon 130 Maronia-Alexandrouplis fault zone 655,656,666 Mavri Rakhi Fornlation 389, 391 Mavrolakkos Unit 397, 398 399 Mediterranean earthquake prediction 694-705,696, 697 seismicity 690, 691-692 Meglenitsa ophiolite 398, 399,405,407, 408 mdlange 24 Avdella 240, 243,244 Pelagonian Zone 383-384 Pirgadikia unit 37, 39 Vai area 125, 126, 127, 131 Vourinos 240, 384 Melnik Fault 571-572, 578

Menderes Massif 229-230, 233,448 geology 449 HP-LT metamorphism 447-462 Mesohellenic Trough 508, 509, 510, 521 536, 522,523 deformation kinematics 524-529 geology 523-524 ophiolites 235-261 structural evolution 521 536, 529 536 tectonic events 526-529 Mesovouni massif250 Mesozoic early back-arc basins, Black Sea-Caucasus 179 196 Tethyan tectonic models 93 148 radiolarites, Lycian Mdlange 229-234 Mesta River 557, 574, 574, 582 metabasites 112-113, 135 metamorphism Almopias Zone 388,403 Alpine, Crete 71, 112-113 east Arabian ophiolite 308 Hercynian 144-145 HP-LT Crete 102, 112, 148 Lycian Nappes 447-462 deformation 458 mineral chemistry 455-457 Vardar zone 403,406 Pelagonian Zone 388 pre-Alpine, Crete 69, 71,81 Serbo-Macedonian Massif 36-38 metaquartzite, mylonitic 38, 39, 42 metaserpentinite 135 microdiamonds, Rhodope I l-I 2 Mid-Atlantic Ridge 13, 23, 24 Mirdita ophiolites 268-269, 542 543 Mirdita-Pindos ophiolites 159,165,169 170, 238 Moesian microplate 155, 157, 171 motion 161,166, 173 Moesian Terrane 53, 54-56 correlation with Zonguldak Terrane 61 Gondwanan affinities 56 origin 61-63 palaeogeography 55-56 stratigraphy 54-55 Triassic folding 185 monazite, U-(Th)-Pb dating, Crete 69, 71, 76, 77, 81, 85 Morava ophiolite 269,270,271,287 MORB 20, 21, 22, 23, 30, 303 Mirdita ophiolite 268-269, 291~96 Pindos ophiolite 239 Voras Massif 378 Mouzaki area 514-515 Myrsini crystalline complex 71, 81 85, 84, 87, 88 garnet zonation 79-81 microfabrics 78 Paraspori orthogneiss 76 structural evolution 77 Nafpaktos area, convergence 513 514 neotectonics 2, 3, 4 Neotethys 105, 144, 170-171 biochronology, Mesozoic radiolarites 233-234 Cenozoic 102

713

714

INDEX

closure 102, 223 Cretaceous 172, 327, 353 definition 7-8 Mesohellenic Trough, evolution 237, 239 Mesozoic, Leventine Basin 201,203 Mesozoic subduction 230-231 origin of ophiolites 11,302 palaeomagnetic studies 351-368 rifting, early-Mesozoic, Levantine basin 220-222 spreading 92, 93, 96, 107, 109, 351-368 palaeomagnetic implications 364-366 Newfoundland-Iberia margin 13,14, 24, 29 Niliifer Plateau 21 plumes 15 Nission Fault 378, 399, 404 North Anatolian Fault Zone 558, 577, 578,579, 580-581,614, 615 Izmit earthquake 635 646,636 palaeoseismology 63~639 North Dobrogea basin 183-184, 185, 193 Northern Almopia fault zone 660 Northern Pieria fault zone 662, 663,666, 667 Nurzeytin Formation 618, 617-618 obduction 169, 317-319 north Arabian ophiolites 306 Ofrinio-Galipsos fault zone 658~659, 666 Ograzhden Fault 573 olivine Albanian ophiolite 27~273,274, 275, 293 K6miirhan ophiolite 332, 336, 341,343,344 Oman, basin margin 308 Ondria Formation 523, 524, 533 ophiolites Albania 267-296, 304 gabbro 269,270, 271,273, 274 geochemistry 276, 279, 281 geological setting 269-272 MOR vs. SSZ origin 291-296 tectonic setting 294-295 ultramafic cumulates 269-296 Anatolia 305, 327 346 Apennine-Ligurian-Alpine 11,21-22, 23, 28, 29 30 Arabian east 307-308 north 305-307 Balkan Peninsula 165 Cretaceous formation 30%312 models 31~317 internal structure 311 312 late, palaeomagnetic studies 351-368 mid 20, 24-5, 30 Tethyan, emplacement 11-31 developmental stages 309-319 distribution 20 harzburgitic 11,246, 268~69, 273, 303-304, 306, 307, 312-316 Hellenic-Dinaric 11, 22-24, 28, 30, 303-305, 315 IAT 303, 305, 306, 312 Jurassic, mid 21-24 lherzolitic 246, 268-269, 273,304, 312, 316-317 MOR 291-296, 303, 316-317

MORB 20, 21, 22, 23, 30, 239, 303, 312 Mirdita 268-269, 291-296 MORB and SSZ, Pindos 239 obduction 317 319 origin 302 303 Othris 303-304 Pelagonian Zone 383-392 Pindos-Vourinos 237 263, 303-304, 406 crustal section 248-249 fabric analysis 254, 257 kinematic zones 251-253 mantle section 246-248, 251 metalliferous zones 253-254 original geometry 257 259 slab heterogeneity 25%263 spreading characteristics 249 Pontides, Central 181 182 supra-subduction zone 12 Anatolia-Arabia 305-306, 307, 308 developmental stages 309-319,313 Hellenic-Dinaric 22-23, 24-25, 30, 303-305 Mirdita 268-269, 291-296 Vourinos 239-241 tectonization 317-319 orthogneiss mylonitic, Pirgadikia unit 37-39, 42, 48 radiogenic dating, Crete 69, 71,74-76 orthopyroxene 272, 273,274, 277 K6mfirhan ophiolite 337, 343,344 Othris ophiolites 239,242,246, 303-304 Padezh Basin 567, 585 Paikon ridge 187 Paikon Zone 373, 377, 378, 381,395,401,405 palaeocurrents, northwest Peloponnese 470, 471,485, 495,497, 501 palaeomagnetism 16-18 Black Sea-Caucasus 179 and tectonic rotation 318, 351-368 database 355-357 Palaeotethys 16 definition 7 8 evidence in Balkan terranes 167 evidence in Pelagonian zone 138-141,378-379 nomenclature 7, 93 Tauric basin subduction zone 181,182, 183,193 tectonic models 91-95, 94, 109, 130-142, 145, 144 convergence related 92, 93, 95, 109, 131, 139, 144 divergence related 91,92-93, 95, 102, 107, 109, 119, 130, 138, 144 counter-arguments 14~145 Palaeozoic late-early Mesozoic, Tethyan tectonic models 93 148,147 terranes 51 63 Balkan Peninsula 159,164, 161--163 Bulgaria 53, 54-57 palaeogeography 55-56, 57 northwest Turkey 53, 58-61 palaeogeography 58-59 palaeogeography 51, 52 palynomorphs, Balkan Peninsula 172 Pannonia 24 Panthalassa 91

INDEX Parnon window 516, 517, 519 Pechenega-Camena fault 183-184, 193 Pelagonia 14, 15, 21, 23 separation from Gondwana 91-92, 93 Triassic volcanism 19 see also Apulia-Pelagonia suture Pelagonian Massifterrane 159, 493 Pelagonian Zone 373-374, 376, 377, 406408 carbonate platform 138-141,380-381,382, 383, 384 evidence for Palaeotethys model 138 141,378-379 foreland basin 403,404, 405 late Jurassic transgression 389, 393 396 metamorphism 388, 392, 393,395 ophiolite emplacement 384, 388-389, 392, 406 ophiolitic mrlange 383-384, 385, 386,406 Palaeogene suturing 405-406, 510 passive margin subsidence 380, 399, 401 sea-level rise 401 subduction 24, 392, 403 Triassic firing 378-379 Peloponnese convergent model 514-517 northwest 468, 469 Pindos Flysch Formation 470, 471, 493 504 Pindos Ocean evolution 467487 succession 132-135 interpretation 136 tectonostratigraphy 132-134 Penrose pseudostratigraphy 241,306, 312,353 Pentalophos Formation 523, 524, 526, 528, 529, 531,532,534, 535 Peonais Zone 373,378, 381,395,405,408 peridotite see Lycian Thrust Sheets, peridofite Permian Balkan Peninsula 161-163 extension 19 Permo-Triassic extension 148 reassembly 17-19 rifting, Vardar Ocean 378 380 succession, western Sicily 9(~102, 97 Peshkopia tectonic window 542, 544 Petite Kabylie 14, 15 Phyllite-Quartzite Unit 509 Crete 69, 70, 71, 72, 86, 88, 103, 105,106 central 119,129, 130 eastern 106, 119-132, 146 western 106, 108, 109-119 sediment chemistry 111, 112,113 Peloponnese 132, 133, 134, 135, 137, 516-518 Pietra di Salomone block 97, 99, 100 Pindos Flysch Formation 493-504, 510-511 palaeocurrents 470, 471,497, 501 palaeogeography 499, 501-502, 503 petrology 502 stratigraphy 494-495 tectonics 50~504 Pindos Group sediments 468-469, 470, 471 Pindos Mountains 237,238, 467 Pindos Ocean 23, 109, 137, 139, 147, 148, 237, 507 accretionary processes 479, 481 continent-ocean transition zone 471-474, 485 evolution 467-487 foreland 481485

715

passive margin 468-469, 471,481 regional evidence 485487 thrust sheets 471 481 Pindos ophiolite 238, 239, 241-243,241,244, 245, 259-263, 303 304, 485 crustal section 248-249 fabric analysis 254, 257 mantle section 246-248,251 metalliferous zone 254 original geometry 258, 259 slow spreading 249 Pindos suture 468, 485 Pindos thrust sheets 471 481,482 basalt 472474 Central Imbricates 476, 477 Eastern Imbricates 476, 478, 479 Frontal Imbricates 475, 476, 4797 genesis and emplacement 474-48 l Pindos unit 70, 103, 104 Pindos zone 467,468,493, 507, 509, 514, 517 evidence for Palaeotethys model 138 Pirgadikia Unit geochemistry 40 geochronology 40-42, 43, 44, 45, 46 geology 37-39 metaquartzite 38 mylonitic orthogneiss 37-39, 48 shear 37 39 Sr-isotope ratios 45, 4748 Pirin massif 557,569, 573 plagioclase 273, 276, 282 Krm/irhan ophiolite 336, 341,343 plate margins, late Jurassic 24 platforms, carbonate 100, 148 Crete 103, 107 Pelagonian zone 38~381,382, 383, 384 see also Gavrovo-Tripolitza carbonate platform; Pelagonian zone, carbonate platform Plattenkalk Unit Crete 70, 72, 102, 104-107, 144, 148, 507 Peloponnese 132, 133, 134-135, 144, 507, 514-515 plug, uplifted 509, 510, 512, 514, 515-517 plumes 11,14 Nilfifer Plateau 15 Podgorie Fault 573, 577 Pontides 179, 181,183, 192, 195 Central 416422, 420 comparison with Eastern Pontides 433435 geochemistry 420-421,422, 423 424, 425 Eastern 423-431,428, 429 comparison with Central Pontides 433435 geochemistry 424, 425, 431-433, 432 IAESZ 413442, 414 structural vergence 433,433 tectonic evolution 433441 models 438-441,439, 440 Predela Fault 574, 577 see also Gradevo-Predela fault pro-wedge 509, 510~512, 510, 514, 517 Piitiirge metamorphic unit 328,329 pyroxene, Pindos-Vourinos ophiolite 248, 251 radiolarites Almopias Zone 398 Mesozoic, Lycian Ophiolitic Mrlange 229-234

716 Ranovac-Vlasina terrane 157 motion 166, 166 Ravdoucha unit 103, 148 Razlog basin 566, 574-575, 578 reconstruction Africa-Europe continental fragments 15-21 continental fragments, early Triassic 1~ 1 9 tectonic 7 Refahiye Complex 424, 425, 426, 429, 428, 429, 429 Rehove ophiolite 269, 270, 271 retro-wedge 509-510, 514 Rhodope 14, 15 microdiamonds 11-12 Rhodope Massif 36, 36, 48,157 back-arc basin 181,183, 185, 187 rift-settings counter-arguments 142-144 evidence from Crete 107, 108, 117-119, 128 130 evidence from Pelagonian zone 141 evidence from Peloponnese 136,137 rifting late-Palaeozoic, Crete 105, 107, 109, 118 119, 130 late-Triassic, Crete 108 Neotethyan, Levant basin 220-222 Permo-Triassic Pelagonian and Vardar Zones 378 380 Sicily 102 see also extension Rila massif 557, 560 Rila normal fault 562-563,564, 577 Rilska River 562, 675 gorge 563,564, 565 roll-back 23, 24, 315 rotation, tectonic Balkan Peninsula 173-174 palaeomagnetic studies 351 368 Troodos microplate 367-368 futile, U-PB dating, Crete 71, 75, 76, 84 Samanda~ Formation 620 Sana-Una-Banija-Kordun terrane 159, 167 Sandanski Basin 565, 566, 568, 569, 570, 571,577, 583-584 Sandanski Formation 571,572, 573, 578 Saparevo normal fault 560, 561,562, 577, 675,676,678, 680-681, 680 Sardinia 14, 15 Sankavak-Kunaaf~arl Fault Zone 597, 598-600, 599 Scythian platform 183,190 volcanism 188-189 sea-level rise, Vardar Zone 401 sediment Cenozoic, Hatay Graben 615-623 Permo-Triassic, western Sicily 96-102,146 seismicity, Mediterranean 690, 691-692 Sel~uk Formation 449450 Semail ophiolite 307, 308-309, 312, 315, 317 318 Serbian-Macedonian Composite terrane 157, 166,379 Serbo-Macedonian Massif 3548, 36, 187 geology 37-39 Serbo-Macedonian Zone, Triassic rifting 378 Serres-Nea Zichni fault zone 657-658,666 Shatsky rise 179, 181,183,190, 191,192 shear, Pirgadikia Unit 37-39 Sheeted Dyke Complex 353,358, 359, 360

INDEX Shipka, Palaeozoic succession 53, 57 Shpati ophiolite 269,270, 271 272, 273, 287, 289 Sicanian basin 96, 148 Sicily,western Permo-Triassic succession 96-102, 97 interpretation 100-102, 148 Simitli Basin 565-566, 567, 568-569 Sipik6r Formation 429,429, 428,433 Sisses unit 105, 107 Sochos-Mavrouda fault zone 660 Sofular Formation 617-619 sole, metamorphic 24, 317-318, 333, 338, 339, 342 Barr-Bassit ophiolite 354-355 Hellenic-Dinaric ophiolite 22-23,304 Zagros ophiolite 307-308 Solea Graben 357, 358, 359 Soulopoulo backthrust 512 South Troodos Transform Fault Zone 353,357,359-361,362 Southern Mygdonia fault system 661 spinel 272, 273,274, 276, 280, 281,293,424, 425 spreading Pindos-Vourinos ophiolite 249 sea-floor 309, 310, 311-312 Atlantic Ocean 11, 12-15, 21, 24 Stara Planina-Pore6 terrane 157,164 motion 166, 163 Stob fault 675,676,678 Stobski Piramidi 560, 562,564, 675 strain, seismic, model 691,692~594 Strandzhides, Palaeozoic succession 53, 57 Stratoni fault 659 Strouma Lineament 683 Strouma River 557, 562, 565, 568, 571-572, 573,582 active faults 671-672, 674, 676, 677, 682 subduction 313-317 Almopias Zone 403 Alpine 71,85-86 at convergent plate margins 509-518,510 Mesozoic Tauric basin 181,182, 185, 191,192, 193 pre-AIpine 86, 87, 88 southward late-Palaeozoic 86--.88, 92-93, 95, 144, 168,379 Triassic 137, 138 Vardar Ocean 167, 168, 170, 172, 187 see also ophiolites, supra-subduction zone superplume, mid-Cretaceous 11 supersequences, depositional, Levantine basin 204-205, 209-210, 211-214, 215-220 Siitpmar Formation 425,429, 429, 428 sutures 3, 155 Gondwana-Eurasia 92, 155, 159 Vardar zone 405406 Svanetia basin 183 Svoula Flysch 378, 381 Svoula Schist Formation 38, 39 Syria, mid-Cretaceous ophiolites 20, 30-31 Syrian Arc inversion, Levantine basin 222-223 Talea Ori unit 102, 104-107, 144 Tauric back-arc basin 179, 181 185, 187 188, 190, 191,192, 193-196 Taxiarchis, metaquartzite 38, 39 zircon dating 4 0 ~ 1, 42

INDEX Taygetos window 514, 515, 517 terrane accretion, Balkan Peninsula 155 terranes Balkan Peninsula 156-158, 157, 159, 162 see also Balkan Terrane Eurasian margin, Black Sea-Caucasus 180-181 Gondwana-derived 48, 51-63, 56, 167, 181,184, 185, 379-380, 493 Tethys closure 157, 591 models 1,2 nomenclature 7-8, 93 remnants in Balkan Peninsula 155, 167 see also Palaeotethys thermochronology, fission-track, Albania 544-550 tholeiite, island arc 22, 23,239, 304, 339, 344, 345, 383 Thrace, fault stress directions 653,655-659 Transcaucasian massif 181, 183,193 Triassic Black Sea-Caucasus back-arc basins 181-185 early, reassembly 17-19 late, geology 21 western Crete inverted succession 113-114 right-way up succession 115-117 Tripali unit 102,106, 107-109 Tripolitza Unit 70, 72, 102, 103-104, 128, 148 troctolite 269,273 geochemistry 281,293 Troodos ophiolite complex 12, 306-307, 312, 315, 317 318,352 age 366 geology 353-354 palaeomagnetic studies 351,355, 357 361,363, 366-368 rotation 357 361 South Troodos Transform Fault Zone 353,357, 359~61,362 Tsotyli Formation 523, 524-526, 528, 529, 530, 531,532, 533 tufa, (~ameli Basin 597-598,599, 601 Turkey Cenozoic, Hatay Graben evolution 613-632 late-Cretaceous ophiolite 327-346 mid-Cretaceous ophiolite 20, 24, 30-31,305 Palaeozoic terranes 53, 58-63 Permo-Triassic reassembly 18 19 western, Lycian Belt HP-LT rocks 447-462 Tyrnyauz-Pshekish fault 189,190 Tyros unit 103, 134, 136-137, 148 uplift, southwest Bulgaria 581-584 Uzunoluk-~ameli Fault Zone 599, 600, 601 Vai crystalline complex 81, 84, 85, 87, 88, 119 orthogneiss 69, 71, 74-76 Phyllite-Quartzite unit 123-128 structural evolution 78 Valais Ocean 29 Vallamara ophiolite 2 70, 271-272 Vardar Ocean 159, 165, 167, 374-375, 378,485-486 closure 48, 170-171,408

717

Main Vardar ophiolitic Belt 159,165 subduction 167, 168, 170, 172, 187 western margin, evolution 373-408,407 Vardar Zone 37, 159, 162, 373,375, 378-379 exhumation 391-396, 406 Jurassic subduction 381 late-Jurassic-early Cretaceous transgression 393-396 oceanic crust 380-381,398 399 passive margin subsidence 380-381,399, 401 Permo-Tfiassic riftilag 378 sea-level rise 401 Western oceanic basin 165, 167, 171,172, 173 Variscan Orogenic Belt 51 52, 63 Vatolakkos section 250 Vegoritis-Ptolemais fault system 664-665,666, 667 Vertiskos Unit 36, 37, 38-39 augengneiss, biotite 37, 39-40, 48 Sr-isotope ratios 45, 47-48 zircon dating 42, 43, 44, 45, 46, 47 volcanism late-Jurassic, Almopias unit 398 mid-Jurassic, Sicily 102 Triassic 19, 121,122, 136-137, 169 Scythian Platform 188-189 Voras Massif 374, 376,378, 381,403 Voskopoja ophiolite 269,270, 271,287 Vourinos SSZ ophiolite 237,238, 239-241,241,242,259-263, 303 304, 383, 388 crustal section 248 fabric analysis 254-257 fast spreading 249 mantle section 246, 248, 251 metalliferous zone 253-254 original geometry 257~59 Vourvourou fault 660 West Crimea fault 183-184 West Pirin Fault 571,578 Yaprakh Formation 415, 416, 418,419 Yaylagayt Formation 415, 416, 418-419, 420, 421,422, 423 Yiiksekova complex 328,329 Zagros ophiolite belt 30~308 zircon fission-track dating Albanides 544-545, 547, 549, 550, 553, 554 Crete 69, 71, 76, 85, 86 Pb-Pb dating, Serbo-Macedonian Massif40 42, 43, 44, 45, 46 U-Pb dating, Vai orthogneiss, Crete 69-70, 71,74-76, 81 Zonguldak Terrane 53 correlation with Moesian Terrane 61 origin 61 63 palaeogeography 59, 61 stratigraphy 59, 60 Zygosti ophiolite 250-251

Tectonic Development of the Eastern Mediterranean Region Edited by A. H. F. Robertson and D. Mountrakis

The Eastern Mediterranean region is a classic area for the study of tectonic processes and settings related to the development of the Tethyan orogenic belt. The present set of research and synthesis papers by Earth scientists from countries in this region and others provides an up-to-date, interdisciplinary overview of the tectonic development of the Eastern Mediterranean region from Precambrian to Recent. Key topics include continental rifting, ophiolite genesis and emplacement, continental collision, extensional tectonics, crustal exhumation and intraplate deformation (e.g. active faulting). Alternative tectonic reconstructions of the Tethyan orogen are presented and discussed, with important implications for other regions of the world. The book will be an essential source of information and interpretation for academic researchers (geologists and geophysicists), advanced undergraduates and also for industry professionals, including those concerned with hydrocarbons, minerals and geological hazards (e.g. earthquakes). Visit our online bookshop: http://www.geolsoc.org.uk/bookshop Geological Society web site: http://www.geolsoc.org.uk

Cover illustration: View acrossFeneosValleyto Mt Dourdouvana,from near Mosia village, NW Peloponnese,Greece. Photograph by A. H. F. Robertson