Correlation of the early Paleogene in northwest Europe

Correlation of the Early Paleogene in Northwest Europe

Geological Society Special Publications Series Editor

A. J. FLEET

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 101

Correlation of the Early Paleogene in Northwest Europe EDITED BY

R. W. O'B. KNOX British Geological Survey Nottingham, UK

R. M. CORFIELD University of Oxford Oxford, UK and

R. E. DUNAY Mobil, North Sea Ltd London, UK

Stratigraphy Commission Petroleum Group

1996 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Society was founded in 1807 as the Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of 8000. It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists, SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London W1V 0JU, UK. The Society is a Registered Charity, No. 210161. Published by the Geological Society from: The Geological Society Publishing House Unit 7 Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel 01225 445046 Fax 01225 442836)

Distributors USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders: Tel (918) 584-2555 Fax (918) 548-0469)

First published 1996

Australia Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5065 Australia (Orders: Tel (08) 379-0444 Fax (08) 379-4634)

The publisher makes 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 1996. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/96/$07.00.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN 1-897799-47-0 Typeset by EJS Chemical Composition, Midsomer Norton, Bath BA3 4BQ, UK Printed by The Alden Press, Osney Mead, Oxford, UK

India Affiliated East-West Press PVT Ltd G-I/16 Ansari Road New Delhi 110 002 India (Orders: Tel (11) 327-9113 Fax (11) 326-0538) Japan Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surugadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel (03) 3255-3497 Fax (03) 3255-3495)

Contents PREFACE KNOX, R. W. O'B. Correlation of the early Paleogene in northwest Europe: an overview

vii

1

Regional studies: stratigraphy, tectonics and volcanism NEAL, J. E. A summary of Paleogene sequence stratigraphy in northwest Europe and the North Sea

15

NADIN, P. A. & KUSZNm,N. J. Forward and reverse stratigraphic modelling of CretaceousTertiary post-rift subsidence and Paleogene uplift in the Outer Moray Firth Basin, central North Sea

43

RITC~E, J. D. & HITCHEN,K. Early Paleogene offshore igneous activity to the northwest of the UK margin and its relationship to the North Atlantic Igneous Province

63

JoY, A. M. Controls on Eocene sedimentation in the central North Sea Basin: results of a basinwide correlation study

79

MUDGE, D. C. & BUJAK,J. P. An integrated stratigraphy for the Paleocene and Eocene of the North Sea

91

THOMAS, J. E. The occurrence of the dinoflagellate cyst Apectodinium (Costa & Downie 1976) Lentin & Williams (1977) in the Moray and Montrose Groups (Danian to Thanetian) of the UK central North Sea

115

WOOD, S. E. & TYSON, R. V. An integrated palynological-palynofacies approach to the zonation of the Paleogene in the Forties-Montrose Ridge area, central North Sea

121

ALl, J. R. & JOLLEY, D. W. Chronostratigraphic framework for the Thanetian and lower Ypresian deposits of southern England

129

POWELL, A. J., BRINKHUIS,H. & BUJAK,J. P. Upper Paleocene-Lower Eocene dinoflagellate cyst sequence biostratigraphy of southeast England

145

ELLISON, R. A., Au, J. R., H1NE,N. M. & JOLLEY,D. W. Recognition of Chron C25n in the upper Paleocene Upnor Formation of the London Basin, UK

185

ALI, J. R., HAtLWOOD,E. A. & KING, C. The 'Oldhaven magnetozone' in East Anglia: a revised interpretation

195

HOOKER, J. J. Mammalian biostratigraphy ascross the Paleocene-Eocene boundary in the Paris, London and Belgian basins

205

JOLLEY, D. W. The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation

219

MITLEHNER, A. G. Palaeoenvironmensts in the North Sea Basin around the Paleocene-Eocene boundary: evidence from diatoms and other siliceous microfossils

255

ScnMrrz, B., HEILMANN-CLAUSEN,C., KING, C., STEURBAUT,E., ANDREASSON,F. P., CORFIELD,R. M. t~ CARTLIDGE,J. E. Stable isotope and biotic evolution in the North Sea during the early Eocene: the Alb~ek Hoved section, Denmark

275

Global perspective: geochronology and the oceanic record BERGGREN,W. A. & AUBRY,M.-P. A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network

309

vi

CONTENTS

AUBRY,M.-R, BERGGREN,W. A., STOTr, L. & SINHA,A. The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology

353

STOTr, L. D., SINHA,A., THIRY,M., AUBRY,M.-R & BERGGREN,W. A. Global 813C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine correlations

381

THOMAS, E. & SHAC~ETON, N. J. The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies

401

CORFIELD,R. M. & NORRIS,R. D. Deep water circulation in the Paleocene Ocean

443

CHAmSI, S. D. & SCHMITZ,B. Early Eocene palaeoceanography and palaeoclimatology of the eastern North Atlantic: stable isotope results for DSDP Hole 550

457

INDEX

473

Preface The early Paleogene of northwest Europe has been the subject of intense investigation over the last quarter century, with important stimulus being provided by the search for oil and gas in the offshore basins and by lUGS-sponsored investigations of the onshore historical stage and system stratotype sections. The Paleogene has long been an exploration target offshore northwest Europe. Giant accumulations, such as the Forties oilfield (UK) and the Ekofisk oilfield and Frigg gasfield (Norway), were discovered in the early days of exploration in the central and northern North Sea. Exploration of the North Sea Paleogene is continuing, with total discoveries now exceeding 12 billion barrels of oil equivalent (BBOE). Paleogene exploration plays are also being actively pursued West of Shetlands, where discoveries reputed to be in excess of 1 BBOE have been made in recent years. The early Paleogene has thus been the focus of major industry interest, and continues to be an attractive exploration target. The onshore sections that fringe the southern margin of the North Sea Basin are home to the historical stratotype successions for most of the Paleogene system and stage boundaries. With the drive towards development of a global standard for the subdivision of Paleogene time, these successions have been the subject of detailed investigation in recent years. Attention is currently focused on the early Paleogene, with international collaboration taking place under the aegis of the lUGS 'Paleocene-Eocene Boundary' and 'Paleocene Stages' working groups and IGCP Project 308 'Paleocene-Eocene Boundary Events in Space and Time'. The igneous province of northwest Britain has also received much attention,with a better understanding of the timing and nature of the volcanism arising from the application of improved analytical techniques and from the acquisition of new information from shallow and deep drilling in the offshore areas. DSDP drilling in the eastern Atlantic has also played a significant part in recent advances in northwest European early Paleogene stratigraphy. Drilling in the Bay of Biscay and the Goban Spur (Legs 48, 80) has provided information on the oceanic succession nearest to northwest Europe, while drilling in the Rockall area (Legs 12, 48, 81) and adjacent parts of the North Atlantic has increased our knowledge of the crustal evolution of the region during the early Paleogene, leading to a better understanding of the history of tectonism and volcanism in northwest Europe. As illustrated by the papers in this volume, the wide range of activities listed above has led to the acquisition of a remarkably diverse dataset, which provides a unique opportunity for the development of a truly comprehensive regional stratigraphy, encompassing terrestrial, epicontinental marine and oceanic successions, and linking these to the tectonic and volcanic events associated with the onset of seafloor spreading between Greenland and Europe. A key element in realizing this potential is the integration of data derived from onshore studies, offshore hydrocarbon exploration activities and ocean drilling programmes. It is hoped that publication of this volume will add further stimulus to the necessary interchange of data and ideas between researchers in these different fields. R. W. O'B. Knox R. M. Corfield R. E. Dunay

Correlation of the early Paleogene in northwest Europe: an overview R. W. O ' B . K N O X British Geological Survey, Keyworth, Nottingham NG12 5GG, UK

The last two decades have seen a major resurgence of interest, both commercial and scientific, in the early Paleogene stratigraphy of northwest Europe. The commercial interest has arisen primarily as a result of major oil and gas finds in the central and Northern North Sea, mostly in deep-water sandstones of late Paleocene to mid Eocene age. Increased interest in the onshore sections has been stimulated partly in response to the offshore hydrocarbon exploration, but largely through the activities of international (IUGS/IGCP) working groups, whose primary concern is the establishment of a globally standardized system of series and stages. The onshore sections of the southern North Sea Basin area are of particular importance in these investigations, because they include the historical stratotypes for the Paleocene and Eocene series and for their constituent stages. Unfortunately, these historical stratotypes are inappropriate as global stratotypes because of their stratigraphic incompleteness, and their limited representation of the standard Paleogene biozones. Only through the fullest understanding of these historical stratotype sections, however, can we ensure that the standard stages are defined in a way that ensures the maximum compatibility with traditional assignments in NW Europe (Knox 1994; Schmitz 1994). For a long time the commercially driven and scientifically driven lines of investigation proceeded more or less independently, partly because of the confidential nature of the offshore investigations and partly because of difficulties in correlating widely separated sections of strongly contrasting lithofacies and biofacies. For these reasons, the earlier stratigraphic compilations for the Paleogene of northwest E u r o p e were concerned almost exclusively with the onshore areas (e.g. Curry et al. 1978; Cavelier & Pomerol 1986; Pomerol 1989). A notable exception is the compilation of data collected in relation to IGCP Project 124 (Vinken et al. 1988), which represents a remarkable achievement in the field of multidisciplinary and multinational stratigraphic collaboration. Other areas that have been the subject of detailed analysis are the oceanic successions of the eastern Atlantic (Fig.l), encountered during DSDP drilling in the Goban Spur area (Legs 48, 80) and in the

Rockall area (Legs 12, 48, 81), and the successions of British Tertiary Igneous (BTIP) and the North Atlantic Igneous (NAIP). Five, largely independent, areas can thus be identified:

volcanic Province Province of study

(1) the onshore sections of the southern margin of the North Sea Basin, restricted to inner shelf, littoral and terrestrial facies; (2) the offshore sections of the North Sea Basin and West of Shetlands area, dominated by outer shelf, slope and basinal facies, but including inner shelf to terrestrial facies around the Scottish landmass; (3) the offshore sections of the Goban Spur area, restricted to bathyal facies (largely calcareous nannofossil oozes); (4) the offshore sections of the Rockall area, representing inner shelf to bathyal facies; (5) the onshore and offshore stratified sections of the BTIP and GFIP, dominated by lavas and tufts, but with intercalations of nonvolcanogenic sediments.

_ !tI Fig.1. Distribution of early Paleogene sedimentary and igneous successions in NW Europe, with locations of the Bay of Biscay, Goban Spur and Rockall DSDP sites (Legs 12, 48, 80, and 81) and central North Sea well 22/10a-4.

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 1-11.

2

R.W. O'B. KNOX

A broad stratigraphic framework has now been established for each of these areas. For example, extensive studies have now been carried out on all aspects of biostratigraphy, such that regional zonation schemes are firmly established for the more important fossil groups. Similarly, comprehensive magnetic polarity zonations have been established for the onshore areas of the BTIP and the southern North Sea Basin, providing not only a means of correlation, but also a direct link to the geological timescale. It is only now that these comprehensive stratigraphic frameworks have been established that correlation betweeen the individual areas (referred to as 'interregional correlation' in this account) can be realistically attempted. Problems in interregional correlation arise primarily from differences in the type and nature of the available stratigraphic data. Examples of limitation due to data type are seismic data, which are available only for offshore sections, and magnetic polarity data, which can be obtained only from cored sections or outcrops. Even within a single discipline, interpretation can be hampered by different methods of data acquisition. Thus biostratigraphic data for the offshore hydrocarbons basins are largely based on cuttings, and thus dependent on first downhole occurrences (FDOs), whereas for the onshore sections they are based on the standard criteria of first and last appearance datums (FADs, LADs). A more serious limitation on interregional correlation is the effect of facies on the nature and diversity of fossil assemblages. Thus while the Goban Spur sections in the eastern Atlantic possess rich calcareous microfaunas and nannofaunas, allowing assignment to the standard Paleogene biozones, equivalent strata in central parts of the North Sea Basin are commonly devoid of calcareous fossils. Conversely, whereas palynomorphs are ubiquitous in the North Sea Basin, they are reported to be absent from the Goban Spur sections. Under such circumstances, correlation between the two successions must rely on a combination of techniques, with particular emphasis on those that are less facies dependent (e.g. magnetostratigraphy and tephrostratigraphy). Fortunately, not all of the correlation problems are so severe. For example, the early Paleogene succession of Denmark, accessible through outcrop sections and cored boreholes, is of deep-water facies, and provides a valuable insight into the succession of the central North Sea. It is therefore possible to apply techniques such as magnetostratigraphy to a basinal North Sea succession, and thereby assess the relationship between the zonally based biostratigraphic schemes established for onshore sections and the FDO-based schemes established for the offshore hydrocarbons boreholes. Additionally, the Paleocene sections of

SE England provide a more or less continuous transect from sublittoral facies in the historical Thanetian stratotype sections of Kent to outer shelf 'North Sea'facies in northern parts of East Anglia. The major challenges for the future are (i) interregional correlation within northwest Europe, and (ii) calibration of the northwest European sections in terms of the standard biozones and the geological timescale. Because it is clear that no one discipline can solve all these problems, a multidisciplinary approach is paramount. The purpose of this overview is to provide a brief assessment of how the different disciplines have contributed towards interregional correlations, and how the ultimate aim of a chronologically calibrated, integrated stratigraphy for the entire northwest European region might be achieved.

Essential elements of an interregional correlation The geological timescale The publication of a new Paleogene timescale (Cande & Kent 1995) is a welcome development in view of the significant discrepancies between earlier timescales, especially over the late Paleocene to early Eocene interval. The incorporation of a radiometric date obtained from a tephra layer of early NP10 age puts the new timescale on a much firmer footing. This is especially important for assessing the influence of 'Atlantic' tectonism on stratigraphic events, as it allows an improved correlation between the biostratigraphically dated sedimentary successions and the radiometrically dated lava successions of the North Atlantic borderlands. However, as cautioned by Berggren & Aubry (1996), timescales are constructed on a series of assumptions that will be subject to continuous reassessment, and the new Paleogene timescale is no exception. The Cande & Kent (1995) timescale is nevertheless considered to provide a much improved overall chronological calibration of the standard biozones and magnetic polarity zones (see Berggren et al. in press). Its adoption by the lUGS Subcommission on Paleogene Stratigraphy should ensure that it becomes the common standard until such time as sufficient data are accrued to warrant further refinement. The use of such a common standard will greatly facilitate comparison of stratigraphic data of all types and from all parts of the world.

Biostratigraphy Calcareous nannofossils. A more or less complete record of the early Paleogene standard calcareous nannofossil zones is present in the Goban

CORRELATION OF THE EARLY PALEOGENEIN NORTHWEST EUROPE" AN OVERVIEW Spur DSDP Sites 549 and 550, which together may be regarded as providing a composite oceanic reference section for NW Europe. This composite section, which has recently been the subject of a detailed reassessment (Aubry et al. 1996), provides a unique opportunity for linking the epicontinental successions with the global oceanic record. Direct correlation is admittedly hampered by the patchy distribution and in part endemic nature of calcareous nannofossil assemblages in NW Europe, but, as early studies on the onshore sections of the North Sea Basin (e.g. Aubry 1986; Siesser et al. 1987) are augmented by new data (e.g. Steurbaut 1990; Hine 1994; Ellison et al. 1996), a significant part of the NW European succession can now be at least broadly linked to the oceanic calcareous nannofossil record. The Goban Spur sites also provide an important record of the standard planktonic foraminiferal zonation for the eastern Atlantic, which has recently been the subject of a comprehensive review (Berggren & Aubry 1996). However, planktonic foraminifera are of limited value in the epicontinental successions, where faunas are characteristically dominated by endemic assemblages. Only in the early Paleocene and part of the early Eocene is there potential for direct assignment of the standard planktonic zonations. For the remainder of the succession, microfossil zonal schemes have been established largely for the deep-water, basinal areas of NW Europe, either on whole faunas (e.g. King 1989; Mudge & Copestake 1992) or selected elements (e.g. Mitlehner 1996). While these zonal schemes have potential for detailed correlation between the deep-water facies of the North Sea Basin, West of Shetlands basins and the offshore Norwegian basins, they cannot be applied to the shallow-water onshore successions of the southern North Sea Basin margins because of a fundamental change in biofacies. Zonal schemes that apply to the entire region (e.g. Gradstein et al. 1994) are therefore necessarily less detailed than the correlations developed specifically for the basinal areas. However, they have the advantage that they can be directly linked with the calcareous nannoplankton zonation and magnetostratigraphy established for the onshore sections. Microfossils.

Palynomorphs. Following their early use in the correlation of the marginal successions of the southern North Sea Basin, dinoflagellate cysts have proved to be particularly valuable in basinal areas (e.g. Heilmann-Clausen 1985; Powell 1988; Mudge & Bujak 1996). They are now established as perhaps the most effective means of correlating across the broad spectrum of facies encountered in

3

the onshore and offshore basins of NW Europe (see Powell 1992; Heilmann-Clausen 1994). Terrestrial palynomorphs have an even greater potential for interregional correlation, as they occur in both marine and continental facies, including sedimentary intercalations within the lava successions of the Hebridean Province and the Faeroes. Regional stratigraphic analysis of the terrestrial palynomorph assemblages is, however, less straightforward than for marine palynomorphs, because of greater variability in assemblages arising from terrestrial climatic and physiographical controls, and because of the strong geographical control exerted by fluvial catchment, transport and deposition. Nevertheless, the potential for local and regional high-resolution stratigraphy has been demonstrated in studies on the Thanet Formation of southern England (Jolley 1992) and on the Harwich Formation and its Central North Sea equivalents (Jolley 1996). Calibration of the palynomorph zonations against the standard zonal schemes is possible in those parts of the succession where the appropriate calcareous nannoplankton are present, including the early Paleocene, part of the late Paleocene (late NP6 to early NP9), and much of the early Eocene. Calibration over the interval spanning late NP9 to NP10 cannot, however, be achieved within the North Sea Basin, and application of calibrations developed for other areas (e.g. Rockall area: Morton et al. 1983) may be unsafe because of possible diachroneity of dinoflagellate cyst influxes resulting from isolation of the North Sea Basin at this time. Many of the limitations apparent from existing palynological studies will be overcome from the combined use of terrestrial and marine groups in establishing correlations between terrestrial, shallow marine and deep marine facies. Published accounts in which such an approach have been used (e.g. Jolley 1996) are relatively few, but the method is being increasingly used in the study of the offshore hydrocarbons basins. Though the mammal faunas of NW Europe are of very restricted occurrence compared with other zonally significant fossil groups, they provide important information on the evolution of the regional palaeogeography Hooker 1996). The succession of NW European mammal faunas indicates that interchange with North American faunas took place twice during the Paleocene and Eocene, firstly in the early late Paleocene and secondly in the Paleocene/Eocene 'boundary interval' (see below). Both events correspond to periods of maximum lowstand (middle Maureen Formation and basal Sele Formation lowstands of the central North Sea), which presumably resulted in exposure

Mammals.

4

R.W. O'B. KNOX

of the Greenland-Scotland landbridge. The initial highly cosmopolitan nature of the North American and European faunas within the Paleocene-Eocene boundary interval suggests that migration was exceptionally rapid, with favourable palaeogeographical conditions probably being enhanced by short-term climatic amelioration (Hooker 1996, and see below). Integrated biostratigraphic framework. As discussed above, there is a distinct dichotomy between the marine biostratigraphic record in the oceanic Atlantic succession and that in the epicontinental successions of the North Sea Basin and West of Shetlands area. Even within the epicontinental basins, biofacies changes from the centre to the margins of the basins puts severe constraints on long-distance correlations. No one fossil group can be relied on to correlate between all sections, and the key to developing a biostratigraphic correlation framework for the entire NW European region thus lies in determining the interrelationships between bioevents and biozonal schemes established for all fossil groups. Isotope stratigraphy

The standard techniques of carbon and oxygen isotope analysis are carried out on whole-rock carbonate or, for more informative results, on selected species of calcareous benthic and planktonic foraminifera. Analysis is therefore usually restricted to carbonate-rich facies. Knowledge of the Paleogene isotopic record has therefore been built up primarily from oceanic sections encountered in DSDP and ODP drilling (e.g. Charisi & Schmitz 1996; Corfield & Norris 1996; Stott et al. 1996). The long-term trends within the early Paleogene are now well established, and attention is currently focusing on specific aspects, such as the occurrence and age of one or more short-term negative carbon isotope excursions in the late Paleocene (e.g. Corfield & Norris 1996; Stott et al. 1996) and their relationship to other major events, such as the widespread oceanic 'benthic extinction' event (Thomas & Shackleton 1996). Appropriate facies for carbon and oxygen isotope analysis within the lower Paleogene are largely restricted in NW Europe to the lower Eocene. A detailed study on the Danish lower Eocene has revealed both long-term trends and short-term events, as well as demonstrating the influence of fresh water input on isotopic values (Schmitz et al. 1996). No such comprehensive study has been carried out on the NW European Paleocene, because the facies are largely unsuitable. However, Stott et al. (1996) have identified a distinct negative carbon isotope excursion within

early diagenetic (pedogenic) carbonate from the Argile Plastique Bariolre (equivalent to the Reading Formation) in the Paris Basin. This may well correlate with the major negative short-term excursion recorded from oceanic sections in late NP9, in which case it will corroborate the earlier findings of Koch et al. (1992) and constitute an important breakthrough in linking the Paleocene/Eocene succession of NW Europe with the oceanic record. R a d i o m e t r i c dating

Radiometric dating has been carried out on both high temperature minerals and glauconites. With two exceptions, the high temperature dates have been obtained from volcanic rocks in the BTIP and GFIE These dates have contributed to knowledge of the relative timing of volcanic events within and between individual igneous centres, and also to the gross timing of volcanism in the region (e.g. Mussett et al. 1988; Noble et al. 1988; Ritchie & Hitchen 1996). They are also of major significance in assessing the relationship between phases of volcanism and the tectonic and sea-level history of the region as inferred from coeval sedimentary successions. Dating of the sediments themselves has been based on K-At analysis of glauconite, and has played a significant part in the construction of some timescales (e.g. Harland et al. 1989). Single-crystal (sanidine) Ar/Ar dates from early Eocene tephras in Denmark have played a more specific role in the construction of the Cande & Kent (1995) timescale, with the age of ash-layer-17 being used as a tiepoint for the lower part of calcareous nannofossil zone NP10. Magnetostratigraphy

Magnetostratigraphy has also played an important role in the establishment of the igneous history of the region, supplementing the data obtained from superpositional/crosscutting relationships and radiometric dating (Mussett et al. 1988; Ritchie & Hitcben this volume). In the sedimentary successions of the onshore southern North Sea area, a detailed knowledge of the reversal history has now been established for many areas (see Ali et al. 1993; Ali & Jolley 1996), though reassessment and refinement are still possible through the application of improved techniques or through the study of new sections (e.g. Ali et al. 1996). As a means of providing precise chronological correlations, magnetostratigraphic data have unique potential in correlating between igneous and sedimentary successions and in assessing synchroneity of sealevel change both within NW Europe and beyond.

CORRELATION OF THE EARLY PALEOGENEIN NORTHWESTEUROPE: AN OVERVIEW Tephrostratigraphy

Though a detailed tephrostratigraphy had been established in the early Eocene ash-series of Denmark around the turn of the century, the significance of these tephras as regional or interregional correlation tools was not appreciated until their wide geographical extent was revealed by offshore drilling in the North Sea. Equivalents of the Danish ash-series are now known to extend beyond the North Sea Basin, into the West of Shetlands area and into the Goban Spur area, providing a valuable interregional marker for the lower NP10 interval. Several phases of ash deposition have now been recognized, with compositional changes reflecting progressive stages in the volcanic history of the northeast Atlantic rift zone (Morton & Knox 1990). S e q u e n c e stratigraphy a n d crustal history

As with the other methods of stratigraphic analysis, sequence stratigraphy has to a large extent developed separately in the offshore areas (e.g. Armentrout et al. 1993; Den Hartog Jager et al. 1993; Galloway et al. 1993; Mitchell et al. 1993; Jones & Milton 1994) and the onshore areas (e.g. Plint 1988; Grly & Lorenz 1991; Jolley 1992; Hardenbol 1994; Knox et al. 1994; Powell, et al. 1996; Vandenberghe in press). It is only recently that studies have been published that combine the two. Studies of this kind range from broad overviews (e.g. Neal 1996), in which depositional systems are linked to both long-term and short-term sea-level change, to detailed, often biostratigraphically driven, analysis of restricted stratigraphic intervals (e.g. Jolley 1996; Powell et al. 1996). Sequence stratigraphy provides a useful vehicle for the compilation of diverse, multidisciplinary stratigraphic data. However, it is clear from the NW European record that the sequence stratigraphy of the region cannot be assessed independently of its crustal history, which during the early Paleogene as a whole was strongly influenced by both Atlantic and Alpine processes. While the overall aim must be to develop a sequence stratigraphic scheme for the whole of NW Europe, it is clear that proper appreciation must be given to the effect of local tectonics on the relative sea-level curve for different parts of the region. The influence of tectonism during the Paleocene has been amply demonstrated by studies on the Paleogene uplift history of the central and northern North Sea, with an uplift of over 400 m proposed for the Outer Moray Firth area (Nadin & Kusznir 1996). Tectonic control is also proposed as the underlying mechanism for the generation of large-scale sequences in the Eocene (Joy 1996). Such a strong tectonic signal is hardly

5

surprising, considering the complex crustal history that is recorded in the British and North Atlantic igneous provinces, culminating in the opening of the North Atlantic in early Eocene times (Ritchie & Hitchen 1996). Under such circumstances, the erection of an interregional sequence statigraphic scheme will inevitably be hampered by geographical variation in the amount of uplift, and by the interplay between uplift and eustatic sea-level change.

Paleocene/Eoeene boundary events The Paleocene/Eocene boundary has yet to be formally defined, but historical considerations require that it will eventually be placed within the NP9 to early NP10 interval, which, in the meantime, is often referred to as the Paleocene/Eocene boundary 'interval' or 'transition'. This interval includes several events of global significance, of which the following may be considered the most important: (1) a pronounced short-term negative shift in carbon isotope values (Corfield & Norris 1996; Stott et al. 1996); (2) an oceanic 'benthic extinction' event, involving a dramatic reduction in both numbers and diversity of benthic foraminifera in the oceans (Thomas & Shackleton 1996); (3) an influx of kaolinite into both the oceans and shelf seas (Thomas & Shackleton 1996); (4) interchange of mammals between North America and Europe, leading to cosmopolitan faunas with an exceptional level of species in common (Hooker 1996); (5) extensive uplift and volcanism associated with the lead up to opening of the North Atlantic between Greenland and Rockall (Ritchie & Hitchen 1996). The relative timing of these events has yet to be fully established, and is the focus of IGCP Project 308. Until this timing is properly established, the ultimate cause of all these changes will not be known. Explanations for the individual features include: (1) a change in the pattern of oceanic circulation, with warming and/or increased salinity of bottom waters causing mass mortality among the benthic communities; (2) rapid changes in productivity in the oceans; (3) climatic warming, with enhanced chemical weathering leading to an increased production of kaolinite; (4) the development of a land-bridge between North America and Europe as a result of regional, probably plumerelated, uplift, perhaps enhanced by eustatic sealevel fall. The NW European succession, with its detailed, multi-component stratigraphic, tectonic and

6

R.W. O'B. KNOX

volcanic record, provides a unique opportunity to unravel many of these temporal relationships. However, the coverage of this critical interval is somewhat fragmented, and little attempt has been made to piece it together other than in the general context of regional stratigraphic compilations. Attempts at identifying the influence of global events on the NW European stratigraphic record have been largely concerned with the role of eustasy in the history of relative sea-level change. To assess the influence of specific global events, attention has focused largely on the early part of the Paleocene-Eocene boundary interval, during which major changes are known to have taken place in the oceanic environment and in the distribution of terrestrial faunas. Such studies include assessment of the history and mechanisms of mammal migration (Hooker 1996) and the recent search for isotopic signatures in the Paleocene of the Paris Basin (Stott et al. 1996). Unfortunately, both studies have of necessity been concerned primarily with the terrestrial sections of the southern North Sea margin, which suffer from the limitations imposed by incompleteness of the stratigraphic record. There is a real possiblity that some of the events identified in the oceans will be represented in these sections by hiatuses. The offshore sections of the central North Sea (and equivalent deep-water

sections in onshore Denmark) similarly have their limitations. In particular, the absence of calcareous facies precludes both the identification of the standard planktonic biozones and the application of standard techniques of carbon and oxygen isotope analysis. However, in their favour, these sections do provide a more or less continuous record across the Paleocene/Eocene boundary interval and must surely contain some reflection of any major change in the global environment.

The basinal record of the central North Sea In the context of a relatively deep-water, noncalcareous, clastic facies, and bearing in mind the events recorded from oceanic sections, reflections of any change in oceanic circulation or climate might be expected in (i) the benthic community (dominantly agglutinated foraminifera), (ii) the nature and relative dominance of terrestrial palynomorph assemblages,~ and (iii) the composition of the clay mineral assemblages. Evidence from published and unpublished sources indicates that significant environmental changes are recorded by all three of these components. In terms of central North Sea lithostratigraphy, the 'Paleocene/Eocene boundary interval' com-

Fig.2. Stratigraphic events in the early part of the Paleocene-Eocene boundary interval in well 22/10a-4, and possible correlation with the onshore succession of southern England.

CORRELATION OF THE EARLY PALEOGENE IN NORTHWEST EUROPE: AN OVERVIEW prises the uppermost Lista Formation, the Sele Formation, and the Balder Formation. Tephrostratigraphic correlation indicates that much of the Balder Formation and the uppermost Sele Formation (unit $3) is of NP10 age, with the NP9/10 boundary probably occurring in the upper part of unit $2. The remainder of the Sele Formation is probably of NP9 age. A pronounced upward reduction in the abundance and diversity of benthic foraminiferal assemblages has long been recognized to occur near the base of the Paleocene/Eocene boundary intervals offshore, corresponding to Sele/Lista formation boundary as defined by Knox & Holloway (1992). The existence of cores through this interval in a relatively expanded Central North Sea section (well 22/10a-4) has allowed other features associated with this 'benthic extinction' event to be recorded in detail for the first time (O'Connor & Walker 1993). Most significantly, the reduction in the benthic assemblages takes place at the level where pale greyish green, waxy, unbedded claystones (Lista Formation) are replaced upwards by medium to dark grey, crudely laminated mudstone (Sele Formation, base unit S la) (see O'Connor & Walker 1993, fig. 21). The boundary probably marks an increase in sedimentation rate, with greater retention of organic matter. However, the reduction in numbers of the agglutinating foraminifera cannot be ascribed simply to a dilution effect, as it is accompanied by a significant reduction in diversity. Further up the section, a second lithological change takes place, marked by a progressive increase in the number and thickness of turbidite sandstone layers (c.859Y6" in fig. 1 of O'Connor & Walker 1993). An upward increase in fine-scale lamination, and decrease in bioturbation, is observed within this unit, indicating the onset of bottom-water anoxia. The top of the turbidite sandstone unit is marked by a rapid upward transition to delicately laminated mudstone. A marked high gamma-ray wireline-log spike occurs at the top of this transition (here placed at c. 8567' 6" in fig. 1 of O'Connor & Walker 1993). This gamma-ray spike is accompanied by a sharp and sustained increase in uranium content. Benthic foraminifera disappear altogether in the uppermost part of the the turbidite sandstone, and are of only rare occurrence throughout the remainder of the Sele Formation. The changes in microfaunal abundance and diversity are accompanied by changes in the palynomorph assemblages (Thomas 1996), with the upward transition from greenish claystone to grey mudstone being marked by a dramatic increase in the ratio of terrestrial to marine forms. This feature persists to the top of the lower division of the Sele Formation (unit Sla of Knox & Holloway 1992), above which the turbidite sandstone unit is

7

characterized by an increase in the proportion of marine forms and by the incoming of the genus Apectodinium. A peak abundance of Apectodinium occurs in the upper part of the turbidite sandstone unit, immediate below the high-gamma spike. Clay mineral assemblages also display marked changes over this interval. The Lista Formation is characterized by kaolinite-free assemblages, dominated by smectite or chlorite (both probably derived from alteration of fine-grained volcanic material). Similar assemblages initially persist into the lower part of unit S l a, but kaolinite appears about half way up the unit, reaching a low peak within the turbidite sandstone unit. After a slight fall-off in the lower part of unit Slb, kaolinite increases rapidly, paralleling an overall increase in grain size. It is clear that important environmental changes took place in the central North Sea in the early part of the Paleocene/Eocene boundary interval. Two of the features described above, the reduction in benthic faunas and the influx in kaolinite, parallel events described from oceanic sections. The marked increase in the proportion of terrestrial palynomorphs can be interpreted in terms of increased terrestrial run-off, resulting from either sea-level fall or climatic change. According to some interpretations, the increase in relative abundance of Apectodinium might be an indicator of warming, but other factors, such as sea-level change, may also be involved (Thomas 1996). However, an increase in temperature around the S 1a/S l b boundary has previously been inferred by Schrrder (1992) from the composition of the palynomorph assemblages, and fluctuations in the composition of the palynomorph assemblages reported from the remainder of the Slb section (Forties Sandstone) may also be of climatic origin (Wood & Tyson 1996). The occurrence of kaolinite in the lower part of the Sele Formation is almost certainly an indicator of increased humidity in the source areas, as it is virtually absent from the underlying Lista Formation, even in sandier facies lower in the section. However, variations in the relative abundance of kaolinite within the Forties Sandstone are clearly related to the overall grain-size, and cannot be taken as a direct climatic indicator. A combination of palynological data plus clay mineral data thus points to at least a general climatic warming around the Sla/Slb boundary, close to the high gamma-ray wireline-log spike, and it is tempting to think that various events recorded over this interval may in some way be related to the broadly coeval events recorded from ocanic successions. However, many other factors need to be taken into account in view of the progressive restriction and evental isolation of the North Sea

8

R.W. O'B. KNOX

Basin over this period. For example, there is good evidence for substantial sea-level fall at the Lista/Sele formation boundary, leading to an increased land area and a sharp influx of terrestrial palynomorphs. A second sea-level fall may be represented by the base of the turbidite sandstone unit, but several features, such as the slight decrease in the proportion of terrestrial palynomorphs, the abrupt incoming of Apectodinium, and the increase in fine-scale lamination point more to a change in basin/hinterland configuration than to a simple sea-level fall. The succeeding laminated mudstones, which are associated with a reduction in kaolinite percentage, a reduction in Apectodinium abundance, and a sustained reduction in the proportion of terrestrial palynomorphs, are interpreted as representing transgression. The maximum-flooding may be represented by the high-gamma peak. Alternatively, it may occur close to the kaolinite and terrestrial palynomorph minima, in which case the highgamma peak could be regarded as a purely basinal phenomenon, perhaps related to the onset of full anoxia. The fall in sea-level represented by the Lista-Sele facies transition is reflected throughout the UK North Sea area, with an overall fall in sealevel of at least 100 m being inferred for southern England. In the Bradwell section (see Knox et al. 1994) fine-grained mudstones of Rhabdammina biofacies (Thanet Formation, Lista equivalent) are overlain by pedogenically altered lagoonal or shallow marine sandy mudstones (upper division of the Upnor Formation, Sele equivalent). It is difficult to explain a sea-level fall of such magnitude and rapidity other than by tectonic uplift (see also Neal 1996). Precise correlation of the onshore and offshore sections is uncertain. However, it seems likely that the hiatus at the base of the lower division of the Upnor Formation (Ellison et al. 1996), during which the Thanet Formation was subjected to decalcification and other effects of meteoric leaching, equates with the major sea-level fall recorded at the base of the Sele Formation (unit S 1a). In both cases, the overlying sediments appear to have been deposited at a time of continued, if somewhat restricted, connection with oceanic waters (allowing influx of NP9 calcareous nannofloras to southern England). Further restriction of free connection between the North Sea and the eastern Atlantic appears to have occurred prior to deposition of the upper division of the Upnor Formation, which is of restricted marine facies. The onset of continental deposition, represented by the base of the lower leaf of the Reading Beds and by the base of the Argiles plastiques bariolres marks the complete closure of the southwestern oceanic connection.

From these observations, it is plausible to suggest that the Upnor Formation and immediately overlying continental beds are represented in the central North Sea by the transitional, upwardcoarsening turbidite sandstone unit. On this correlation, the turbidite sandstone unit would approximate to the horizon of mammal migration (Hooker 1996) and the negative carbon isotope excursion recorded by Stott et al. (1996). The principal reduction in benthic faunas would, however, not be related to the oceanic 'benthic extinction', however, since it occurs lower in the succession. In the absence of any reliable onshoreoffshore corelation tool, such correlations must be considered speculative, as must any connection between the events recorded from the central North Sea and the global changes that were occurring at about the same time. It is quite possible to explain the central North Sea events purely in terms of regional tectonism, with the progressive elimination of benthic faunas and the onset of basin anoxia resulting from basin isolation and the influx of kaolinite resulting from increased humidity and precipitation in response to changes in palaeogeography. We are thus left with the tantalizing situation that some of the oceanic events recorded from the Paleocene/Eocene boundary interval are paralleled in the broadly coeval sediments of the central North Sea, but that on present evidence it is not possible to say whether some or all of the North Sea events are caused by local (North Atlantic) tectonism or by global climatic change. Of course, it may be wrong to think in terms of these two extremes, as changes in oceanic circulation pattern, global climatic change, and mammal migration could themselves all be explained in terms of plate reorganization prior to the opening of the North Atlantic (Eldholm & Thomas 1993). In conclusion, one of the most intriguing aspects of early Paleogene stratigraphy is the evidence in sediments of late NP9 age for a major but short-term interruption to the long-term trends of oceanic warming and climate change. It must be expected that such a profound change in the world's oceans would have left its mark on the terrestrial and epicontinental marine record, in which case the onshore and offshore successions of northwest Europe must be prime candidates for study. Such study would not be of purely academic interest, since an understanding of the relative roles of regional uplift, eustasy, and global climatic change would not only answer some long-standing scientific questions, but also throw new light on the origin of one of northwest Europe's most productive oil reservoirs, the Forties Sandstone. The elucidation of event stratigraphy within the Paleocene/Eocene

CORRELATION OF THE EARLY PALEOGENE IN NORTHWEST EUROPE: AN OVERVIEW boundary interval in northwest Europe thus provides a prime example of the benefits of interchange of data and ideas across the academic/commercial divide.

9

I am grateful to Richard Corfield, David Jolley and Andy Morton for helpful comments on the manuscript. Publication is with the approval of the Director, British Geological Survey.

References ALl, J. R. & JOLLEY,D. W. 1996. Chronostratigraphic framework for the Thanetian and lower Ypresian deposits of southern England. This volume. , HAILWOOD,E. A. & KING, C. 1996. The 'Oldhaven magnetozone' in East Anglia: a revised interpretation. This volume. , KING, C. & HAILWOOD, E. A. 1993. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAmWOOD, E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. ARMENTROUT, J. M, MALACEK, S. J., FEARN, L. B. SHEPPARD, C. E., NAYLOR, P. H., MILES, A. W., DESMARAIS,R. J. & DUNGY,R. E. 1993. Log-motif analysis of Paleogene depositional systems tracts, central and northern North Sea: defined by sequence stratigraphic analysis. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 45-58. AUBRY, M. E 1986. Paleogene calcareous nannoplankton biostratigraphy of Northwestern Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 55, 267-334. , BERGGREN,W. A., STOTF,L. & SINHA,A. 1996. The upper Paleocene - lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology. This volume. BERGGREN, W. A. & AUBRY, M. -E 1996. A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network. This volume. , KENT, D. V., SWISHER, C. C. I I I & AUBRY,M.-R 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A, KENT, D. V., AUBRY, M.-R & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society of Economic Geologists and Paleontologists, Special Publication, 54, Tulsa. 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 (B4), 6093-6095. CAVELIER, C. & POMEROL, C. 1986. Stratigraphy of the Paleogene. Bulletin de la Socidtd Gdologique de France, 8 (II,2), 255-265. CHARISI, S. D. & SCHMrrZ,B. 1996. Early Eocene palaeoceanography and palaeoclimatology of the eastern North Atlantic: stable isotope results for DSDP Hole 550. This volume. CORFrELD, R. M. & NORRIS, R. D. 1996. Deep water circulation in the Paleocene Ocean. This volume. COSTA, L. I., DENISON, C. & DOWNIE, C. 1978. The

Paleocene/Eocene boundary in the Anglo-Paris Basin. Journal of the Geologial Society, London, 135, 261-264. CURRY, D., ADAMS,C. G., BOULTER,M. C., DILLEY,F. C., EAMES, E E., FUNNELL, B. M., & WELLS, M. K. 1978. A Correlation of Tertiary Rocks in the British Isles. Geological Society, London, Special Report, 12, 1-72. DEN HARTOGJAGER,D., GILES, M. R. & GRIFFITHS,G. R. 1993. Evolution of Paleogene submarine fans of the North Sea in space and time. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 59-72. ELDHOLM,O. & THOMAS,E. 1993. Environmental impact of volcanic margin formation. Earth and Planetary Science Letters, 117, 319-329. ELLISON, R. A., ALI, J. R., HINE, N. M. & JOLLEY,D. W. 1996. Recognition of Chron 25n in the upper Paleocene Upnor Formation of the London Basin, UK. This volume. GALLOWAY, W. E., GARBER, J. L., XIJIN, LIU & SLOAN, B. J. 1993. Sequence stratigraphic and depositional framework of the Cenozoic fill, central and northern North Sea Basin. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 33-44. GELY, J. -P. & LORENZ,C. 1991. Analyse s6quentielle de l'Eoc~ne et de l'Oligocbne du bassin Parisien (France). Revue de l'lnstitut Franfais du Pdtrole, 46 (6), 713-747. GRADSTEIN,E M., KAMINSKI,M. A. & BERGGREN,W. A. 1994. Cenozoic biostratigraphy and paleoceanography. In: The North Sea and Labrador Shelf. Micropaleontology, 40, Supplement. HARDENBOL, J. 1994. Sequence stratigraphic calibration of Paleocene and Lower Eocene continental margin deposits in NW Europe and the US Gulf Coast with the oceanic chronostratigraphic record. GFF, 116, 49-51. HARLAND,W. B., ARMSTRONG,R. L., COX, A. V., CRAIG, L. E., SMITH, A. G. & SMITH, D. G. 1989. A Geologic Time Scale. Cambridge University Press. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks Geologiske UndersCgelse, Serie A, 7. 1994. Review of Paleocene dinoflagellates from the North Sea region. GFF, 116, 51-53. HINE, N. M. 1994. Calcareous nannoplankton assemblages from the Thanet Formation in Bradwell Borehole, Essex, England. GFF, 116, 54-55. HOOKER, J. J. 1996. Mammalian biostratigraphy across the Paleocene-Eocene boundary in the Paris, London and Belgian Basins. This volume.

10

R. W. O'B. KNOX

JOLLEY, D. W. 1992. Palynofloral association sequence stratigraphy of the Palaeocene Thanet Beds and equivalent sediments in eastern England. Review of Palaeobotany and Palynology, 74, 207-237. 1996. The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation. This volume. JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: Palaeogene stratigraphy and relative sea-level history of the Outer Moray Firth, UK North Sea. Marine & Petroleum Geology, 11, 157-165. JoY, A. M. 1996. Controls on Eocene sedimentation in the central North Sea Basin: results of a basinwide correlation study. This volume. KING, C. 1989. Cenozoic of the North Sea. In: JENKINS,D. G. & MURRAY, J.W. (eds) Stratigraphical Atlas of Fossil Foraminifera (2nd edition). Ellis Horwood, Chichester, 294-298. KNOX, R. W. O'B. 1994. From regional stage to standard stage: implications for the historical Paleogene stratotypes of NW Europe. GFF, 116, 55-56. - & HOLLOWAY,S. 1992. Paleogene of the Central and Northern North Sea. In: KNox, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic nomenclature of the UK North Sea. British Geological Survey, Nottingham. , HINE, N. &ALI, J. 1994. New information on the age and sequence stratigraphy of the type Thanetian of Southeast England. Newsletters on Stratigraphy, 30 (1), 45--60. KOCH, P. L., ZACHOS, J. C. & GINGERICH, P. D. 1992. Coupled isotopic change in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary. Nature, 358, 319-322. MILTON, N. J., BERTRAM,G. T. & MANN,I. R. 1990. Early Palaeogene tectonics and sedimentation in the Central North Sea. In: HARDMAN, R. F. P. BROOKS, J. (eds) Tectonic Events Responsible for Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. MITCHELL, S. M., BEAMISH, G. W. A., WOOD, M. V., MALACEK, S. J., ARMENTROUT,J. A., DAMUTH,J. E. & OLSON, H. C. 1993. Paleogene sequence stratigraphic framework of the Faeroe Basin. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 1011-1023. MITLEHNER, A. 1996. Palaeoenvironments in the North Sea Basin around the Paleocene-Eocene boundary: evidence from diatoms and other siliceous microfossils. This volume. MORTON, A. C. & KNox, R. W. O'B. 1990. Geochemistry of late Palaeocene and early Eocene tephras from the North Sea Basin. Journal of the Geological Society, London, 147, 425-437. --, BACVdVIAN,J. & HARLAND,R. 1983. A reassesment of the stratigraphy of DSDP Hole 117A, Rockall Plateau: Implications for the Paleocene-Eocene boundary in N.W. Europe. Newsletters on Stratigraphy, 12(2), 104-111. MUDGE, D.C. & BUJAK, J.P. 1996. An integrated stratigraphy for the Paleocene and Eocene of the North Sea. This volume.

• COPESTAKE,P. 1992. Revised Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine & Petroleum Geology, 9, 53-69. MUSSETT, A. E., DAGLEY, P. & SKELHORN, R. R. 1988. Time and duration of activity in the British Tertiary igneous province. In: MORTON, A. C. 8,: PARSON,L. M. (eds) Early Tertiary Volcanism and the Opening of the North East Atlantic. Geological Society, London, Special Publication, 39, 337-348. NADIN, P. A. & KUSZNIR,N. J. 1996. Forward and reverse stratigraphic modelling of Cretaceous-Tertiary post-rift subsidence and Paleogene uplift in the Outer Moray Firth Basin, central North Sea. This volume. NEAL, J. E. 1996. A summary of Paleogene sequence stratigraphy in northwest Europe and the North Sea. This volume. NOBLE, R. H., MCINTYRE, R. M. & BROWN, P. E. 1988. Age constraints on Atlantic evolution: timining of magmatic activity along the E Greenland continental margin. In: MORTON, A. C. & PARSON, n. M. (eds) Early Tertiary Volcanism and the Opening of the North East Atlantic. Geological Society, London, Special Publication, 39, 201-214. O'CONNOR, S. J. & WALKER, D. 1993. Paleocene reservoirs of the Everest Trend. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 145-160. PLINT, A. G. 1988. Global eustasy and the Eocene sequence in the Hampshire Basin, England. Basin Research, 1, 11-22. POMEROL, C. 1989. Stratigraphy of the Palaeogene: hiatuses and transitions. Proceedings of the Geologists' Association, 100, 313-324. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Palaeocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327-344. - 1992. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. Chapman & Hall, London, 155-251. , BRINKHUIS, H. ~:; BUJAK, J. P. 1996. Upper Paleocene - Lower Eocene dinoflagellate cyst sequence biostratigraphy of southeast England. This volume. RITCHIE, J. D. & HITCHEN, K. 1996. Early Paleogene offshore igneous activity to the northwest of the UK margin and its relationship to the North Atlantic Igneous Province. This volume. SCHMITZ, B. 1994. The Paleocene Epoch - stratigraphy, global change and events. GFF, 116, 39-67. - - - , HEILMANN-CLAUSEN,C., KING, C., STEURBAUT,E., COR~ELD, R. M. & CARTLIDGE,J. E. 1996. Stable isotope and biotic evolution in the North Sea during the early Eocene: the Alb~ek Hoved section, Denmark. This volume. SCrmODER, T. 1992. A palynological zonation for the Paleocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-126. SIESSER, W., WARD,D. J. & LORD A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian stage

--

CORRELATION OF THE EARLY PALEOGENE IN NORTHWEST EUROPE: AN OVERVIEW (Palaeocene) in the type area. Journal of Micropalaeontology, 6, 85-102. STEURBAUT,E. 1990. Ypresian calcareous nannoplankton biostratigraphy and paleogeography of the Belgian Basin. In: DuPuIS, C., DE CONNINCK, J. & STEURSAUT,E. (eds) Bulletin de la Societd Belge de Gdologie, 97 (3-4), 251-285. STEWART,I. J. 1987. A revised stratigraphic interpretation of the early Palaeogene of the central North Sea. In: BROOKS, J. & GLENNIE,K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 557-576. STOTF, L. D., SINHA, A., THIRY, M., AUBRY, M. -P. & BERGGREN,W. A. 1996. Global 513C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine correlations. This volume. THOMAS, E. & SHACKLETON, N. J. 1996. The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies. This volume. THOMAS, J. E. 1996. The occurrence of the dinoflagellate

11

cyst Apectodinium (Costa & Downie 1976) Lentin & Williams (1977) in the Moray and Montrose Groups (Danian to Thanetian) of the UK central North Sea. This volume. VANDENBERGHE, N., LAGA, P., STEURBAUT, E., HARDENBOL, J. & VAIn, P. R. In press. Sequence stratigraphy of the Tertiary at the southern border of the North Sea Basin in Belgium. In: HARDENSOL,J., VAIL, P., DE GRACIANSKY,P.C. & JACQUIN,T. (eds) Sequence Stratigraphy of Mesozoic and Cenozoic European Basins. CNRS-IFP, Paris. VINKEN, R., VON DANIELS,C. H., GRAMANN,E, KOTHE, A., KNOX,R. W. O'B., KOCKEL,E, MEYER, K. J. & WEISS, W. (eds) 1988. The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100. WOOD, S. E. & TYSON, R. V. 1996. An integrated palynological-palynofacies approach to the zonation of the Paleogene in the Forties-Montrose Ridge area, central North Sea. This volume.

A summary of Paleogene sequence stratigraphy in northwest Europe and the North Sea J. E. N E A L

Rice University, Department of Geology & Geophysics, PO Box 1892, Houston, TX 77251, USA; present address: Exxon Production Research Co., PO Box 2189, Houston, TX 77252, USA Abstract: Sequence stratigraphic analysis of Paleogene central North Sea well-log, seismic and biostratigraphic data recognizes patterns of cyclic sedimentation seen in the physical stratigraphy and biostratigraphy. Numerous authors have documented cyclic sedimentation resulting from relative changes in sea level in northwest Europe, but interregional integration of these observations with North Sea subsurface data is lacking in the literature. Presented here is a chronostratigraphic correlation framework for the Paleogene of northwest Europe, built by integrating subsurface and outcrop data using sequence stratigraphic first principles. Biostratigraphic data from many sources is ordered with the composite standard method. Graphic correlation of this data documents certain correlations and helps suggest previously unrecognized ties. Paleogene North Sea sediments record five major regressions and their intervening major transgressions. Overprinting this low frequency signal are 19 higher frequency sequence cycles that control lithofacies distribution. In northwest Europe, western basins (London-Hampshire, Paris and Belgian) have shallow marine to non-marine settings which reveal basinward and landward facies shifts that indicate sea level changes. The biostratigraphy of these shallow water deposits is linked to deep water central North Sea biostratigraphy by correlating through deeper water deposits outcropping in Denmark that have been tied to western basin stratigraphy. Using this biostratigraphic framework, key bounding surfaces are correlated between basins using sequence stratigraphic principles. Depositional sequences are recognized onshore that are completely sediment starved in the North Sea. The mixing of low and high frequency sea level signals requires that all of northwest Europe be studied to recognize the 'true' signal. Final correlations resolve 30 depositional sequences with five long-term sea-level changes that can vary from one sub-basin to another.

Geologists have recognized transgressions and regressions in northwest European Tertiary sediments since the late 18th century, as Lavoisier noted in 1766 the 'flux and reflux' of the sea represented by map units in the Paris Basin (Rappaport 1969). These sediments challenge us to unravel their stratigraphic equivalence, from shelf to basin and region to region. Sequence stratigraphy is an infant technique compared with traditional correlation procedures of lithostratigraphy, biostratigraphy and even magnetostratigraphy. A geologist using sequence stratigraphy can synthesize results from more traditional correlation methods and suggest ways to solve stratigrapbic contradictions. A sequence stratigrapher uses regional and local stratigraphic and sedimentological observations to reconstruct the relative sea-level history represented within the rock record. Chronostratigraphic charts (Wheeler 1958) represent deposition and lacuna along a given profile through time, graphically identifying depositional sequences. A basin's chronostratigrahic chart will show a record of transgressions and regressions through time.

The northwest European Paleogene basin had many sub-basins, each with its own stratigraphy. Belgium, northern France and southeast England have outcrops of Paleogene shallow marine and estuarine sediments. Boreholes and outcrops from Denmark and northern Germany encounter Paleogene deeper shelf and basinal sedimentary deposits. The central North Sea was a depocentre for sandy submarine fans and prograding deltas identified by interpretation of seismic data and well information. Because all these sub-basins were connected (Ziegler 1990), the rock record of each should show similar relative sea-level histories unless local tectonic uplift obscures the signal. Under the auspices of the CNRS-IFP Sequence Stratigraphy of Mesozoic and Cenozoic European Basins Project, I present below a stratigraphic framework interpreted from a seismic data grid of 7000 line kilometres (Fig. 1) and approximately 250 well-logs (150 of which had detailed biostratigraphic reports). The framework developed from working with experts from European countries that have Paleogene deposits and using

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlationof the Early Paleogene in NorthwestEurope, Geological Society Special Publication No. 101, pp. 15-42.

15

16

J.E. NEAL

the ongoing and published biostratigraphic, sedimentological and physical stratigraphic research to correlate a central North Sea sequence stratigraphy with onshore sections. Graphic correlation of biostratigraphic data carried out by J. Stein and J. Gamber of Amoco (Neal et al. 1994) augmented key marker stratigraphy and suggested different and more precise correlations than have been previously published. Sequence stratigraphy can be used in two

different ways depending on the goals of the researcher. The more economically-minded approach focuses mainly on detailed description of the rock record, identifying key stratigraphic bounding surfaces and sedimentary bodies for their fluid flow properties. Regional correlation is less important at this scale. The second approach is more academic, using every available piece of information to construct an internally consistent, documented chronostratigraphic framework. The

Fig. 1. Map of northwestern Europe with the outline of Eocene deposition from Zeigler (1990). Also shown are the locations of North Sea seismic lines and key outcrop and borehole sections. Labelled North Sea Pre-Tertiary structural elements are: E.S.P., East Shetland Platform; W.G.G., Witch Ground Graben; S.V.G., South Viking Graben; EM.H., Forties-Montrose High.

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE second approach is used in this paper. Biostratigraphic data resolution is the limiting factor in the precision, accuracy and detail of a documented chronostratigraphic chart. Each subsurface depositional sequence carried in the framework below has a unique biostratigraphic signature, but a sequence stratigraphic approach integrating sequence stacking patterns within major regression/ transgression facies (R/TF) cycles was necessary when sequences were identified in outcrop section at a higher resolution than individual biostratigraphic schemes could resolve (i.e. two Belgian sequences in N P l l =two English sequences in NP11; an interval largely represented by sediment starvation in the central North Sea). Depositional sequences occur at many timescales (e.g. Mitchum & Van Wagoner 1991; Posamentier et al. 1992) and aliasing sequences with biostratigraphy is a constant pitfall. Employing sequence stacking patterns is one way to enhance the correlation. An important advantage this study has over previous correlation frameworks, as mentioned above, is the composite standard biostratigraphic method and graphic correlation (Shaw 1964; Miller 1977; Stein et al. 1995). Composite standard biostratigraphy synthesizes a detailed ideal section from the most complete individual section available, which is then enhanced with additional data from multiple, stratigraphically overlapping individual sections. Graphic correlation is a technique to graphically represent the completeness of an individual test section relative to the composite standard (Fig. 2). Graphic correlation helps interpret the rock record by identifying and quantifying stratigraphic gaps, which appear as terraces in a line of correlation that relates the completeness of a test section to the composite standard. The line of correlation (LOC) is interpreted from the scatter of individual fossil markers as they appear in the test section and the composite standard. Recognition of data terraces is then used to construct a framework of chronostratigraphic units based on overlapping graphic correlation terraces. If terraces in two or more wells share intervals of time [expressed in terms of composite standard units (CSU)], then the time of overlap can be correlated as a single or composite event (i.e. a lacuna). Terrace-bounded units bracket depositional sequences for correlation purposes, but not all depositional sequences are associated with a terrace (Neal 1994).

Previous work and correlation problems Numerous studies of northwest European Paleogene stratigraphy exist in the literature. The last decade witnessed a revision of Paleogene

17

stratigraphy onshore with the publication of detailed biostratigraphic (Aubry 1986), magnetostratigraphic (Ali et al. 1993) and sedimentological (Plint 1988) frameworks that employ sequence stratigraphic correlation techniques. Other regional papers will arrive with the publication of the Proceedings from the CNRS-IFP Meeting of Dijon 1992 (i.e. Michelsen et al. 1995; Vandenberghe et al. 1995). The central North Sea alone is on a second or third generation of stratigraphy papers, each new generation becomes more complex as data resolution increases. Parker (1975) published the first regional study of Paleogene North Sea deposition that inspired works through to the most recent round of papers presented at the 4th Conference on Petroleum Geology of Northwest Europe (Parker 1993). New aspects of North Sea stratigraphy are continually emerging, creating a need to standardize present published frameworks and document the sea-level signal for the North Sea and northwest Europe. The International Geologic Correlation Project (IGCP) Project 124 (Vinken et al. 1988) synthesized and standardized vast amount of raw stratigraphic information available in northwest Europe, but focused mainly on pure biostratigraphic correlations and emphasized lithostratigraphy. The IGCP Project 124 results provide a control on depositional sequence correlations and allow crosscorrelations of different fossil taxa types. Sequence stratigraphic analysis uses this data source as a framework, but provides additional information not covered the IGCP Project 124 report. The conclusions presented below differ from those of IGCP Project 124 mainly because the depositional sequence-based framework contains more subdivisions and is largely independent of strict lithostratigraphic correlation. This study attempts to standardize the present state of depositional sequence correlations similar to the way IGCP Project 124 ordered biostratigraphic and lithostratigraphic correlations. Standardizing a relative sea level history from subsurface data and published studies has, however, proved difficult. Various groups have created separate biostratigraphic zonation schemes and used different key correlation markers. This study used composite standard biostratigraphy to unify different zonal schemes without losing the ability to recognize individual key correlation markers. This advantage allowed comparisons with the published stratigraphic frameworks and led to a tie with outcrop and borehole sections of northwest Europe. In standardizing relative sea-level histories for northwest Europe, the first step was to evaluate the applicability of the (Haq et al. 1988) eustatic curve. Since northwest Europe was one of the key areas for the development of the curve, its documentation

18

J.E. NEAL

~T~E

.~=~ o,u

do ,..cZ

.~

r

"~

0

09

~~ ,,..a

,~,_qo

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE in this study would seem crucial. The fact that many key biostratigraphic markers in northwest Europe are endemic to the area made correlations to the eustatic curve difficult, but a correlation is made

19

by linking the subsurface framework to outcrop sections containing nannofossil zonations. A possible correlation between the eustatic curve and several North Sea frameworks is shown in

Fig. 3. Comparison of several sequence stratigraphic frameworks published in the literature, corrected to the timescale of Berggren et al. (1995). This figure highlights the different resolutions and internal inconsistencies of various frameworks for the same stratigraphic interval.

20

J.E. NEAL

Fig. 3. This correlation highlights problems of different resolutions and diachroneity in the various published works. The portion of the Haq curve shown is modified to be consistent with the nannofossil stratigraphy of Aubry et al. (1988), tied to the new geomagnetic timescale of Cande & Kent (1992) and Berggren et al. (1995). Armentrout et al. (1993) compared their sequence stratigraphic framework to the Haq curve and recognized fewer events than predicted by the global curve. They based their framework on a series of key fossil markers, which become tie points with the framework presented in this paper. Next is the more detailed framework of Den Hartog Jager et al. (1993), showing a higher resolution than both the Mobil (Armentrout et al. 1993) and Exxon (Haq et al. 1988) frameworks. Den Hartog Jager et al. (1993) based their correlations on the Shell North Sea Paleogene biozones (Schrfder 1992). Stewart (1987) published the first major lower Paleogene sequence stratigraphic framework in the North Sea, displayed at the far right in Fig. 3. This framework is diachronous in places when comparing the type wells and maps of Stewart (1987) to a biostratigraphy-based framework. All these different frameworks, constructed from similar data sets, highlight the problem of finding an agreed relative sea-level signal. The Paleogene North Sea relative sea level signal is easily divided into cycles of two categories: sequence cycles (ranging from 0.3 to 5 million years when tied to the present geomagnetic timescale) and major regressive/transgressive facies (R/TF) cycles (ranging from 3 to c. 13 million years). Sequence cycles of relative sea level form depositional sequences ( s e n s u Mitchum et al. 1977). They, along with sediment supply and the depositional profile, control distribution of lithofacies and seismic stratal patterns. These cycles are correlative within biostratigraphic resolution as stratigraphic events throughout the basin and may represent a eustatic signal. Every location does not record every sequence cycle, though, as composite unconformities or marine hiatal intervals can account for time represented by multiple sequences elsewhere in the basin (Neal et al. 1994). Major R/TF cycles consist of multiple sequence cycles. They reflect a reorganization of the depositional system in a basin or sub-basin and have been related to tectonic mechanisms. These cycles are not always correlative and can be as much as half a cycle out of phase when comparing their effects around northwest Europe. Major R/TF cycles control the facies characteristics of their component sequence cycles. For example, sequence cycles that occur on the falling sea-level limb of a major R/TF cycle will have well developed lowstands and are marked by fan-depositional pulses into distal parts

of the basin. Sequence cycles that occur during the turn-around and rising sea-level portions of a R/TF cycle are characterized by aggradational facies (thick coastal plain deposits). Sequence cycles occurring during the 'highstand' interval of a R/TF cycle may be completely starved in the central North Sea and are recognized only by correlation to onshore sections. Five major R/TF cycles, grouping 18 regionally identifiable sequence cycles, are recognized for the Paleogene central North Sea (Fig. 4). Sequence cycles are named for lithostratigraphic units that typically fall within the sequence (Neal et al. 1994). Various publications have noted major tectonic events in the North Sea area (e.g. Cavelier & Pomerol 1979; Galloway et al. 1993; Japsen 1993; Vandenberghe et al. 1995). Major R/TF cycles are linked to these events, which are recognizable throughout the central North Sea, but their magnitude and precise timing is not correlative throughout northwest Europe. For example, the lower Oligocene is regressive in the central North Sea and Denmark (Michelsen et al. 1995) as a result of the uplift of Norway and possibly the Shetland Platform (Japsen 1993; Galloway et al. 1993). This same time interval corresponds to a major transgression (RupelianStampian) in Belgium (Steurbaut 1992) and France (Cavelier & Pomerol 1979; Gfly & Lorenz 1991). A final complication is that sequence cycles have different apparent magnitudes around northwest Europe and their expression is enhanced or subdued by local tectonic events (Vail et al. 1991). A summary chronostratigraphic chart (Fig. 5) synthesizes all this stratigraphic information to produce a Paleogene North Sea relative sea-level history. Specific formations and fossils are noted below as they document changes in relative sea level from the top Cretaceous to the middle Oligocene.

Danian-Selandian deposition Everywhere in northwest Europe, the Paleogene sequence stratigraphy commences with a major regression that started at various times in the Upper Cretaceous, depending on sub-basin location. In the central North Sea, the Maastrichtian Tor Formation and Danian Ekofisk Formation arc separated by 'a change in sedimentary facies (that) is generally an omission or erosion surface, with the basal Ekofisk a condensed black shale' (Kennedy 1987, p. 477). The duration of this marine hiatus is difficult to determine. The central North Sea biostratigraphic framework relics on ditch cutting samples, and therefore is based on last appearance datums (LAD), which correspond to first downhole

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

Ma. Central North Sea R\TF Cycles

30 Paleogene (Lower 5 Oligocene)

Major North Sea Subsurface Tectonic Events Sequence Cycles Land ~ " - ~ > S e a Lower L/ark II Uplift oft~

L~

35 ~

Lark l

Brioc

Paleogene 4 (Upper40 Middle Eocene)

21

Skaqerrat:U Katteqat M

Platform]N~ _ Alpine,~ uompressiqn.

~i Alba Undiff.-

~

~

jr ~ "

e'i'umw

Nauchlan

ffaran

45

Upp;r Tay Phases of 1~'

Paleogene 3 SO 55

(Lower Eocene) Paleogene 2

.. (,Upper I~aleocene~ Paleogene I (Lower Paleocene)

~igg Middle Frigg LowerFriggj Upper

Badl"ff-er'uooer Sele,-~

- Middle ~ele

'-'-

.UDDer B~mgPa~ r Forties

Uplift in the Shetland Platform

J

Jt

voc.ncsh Thulea~n ! 1

Lower Balmoral Andrew

/ Maureen /

Ekofisk

/

Selandian Clastics

Dan~ian

Fig. 4. Long-term relative sea-level curve and location on that curve of major regressive/transgressive facies (R/TF) cycles and depositional sequences for the central North Sea. This curve is related to major regional tectonic events from Zeigler (1990), Milton et al. (1990), Galloway et al. (1993) and Vandenberghe et al. (1995).

occurrences (FDO). The most commonly reported fossils in the Ekofisk are planktonic foraminifera such as Globorotalia compressa, Globorotalia pseudobulloides and Eoglobigerina trivialis, which range upwards into the Maureen sequence. The oldest nannofossil marker found in the Ekofisk chalks is the acme occurrence of Prinsius dimor-

phosus, which has been placed near the NP2-3 boundary (Thomsen & Heilmann-Clausen 1985; Thomsen 1992). The oldest correlatable FDO of dinoflagellate cysts in the central North Sea Paleogene is Senoniasphaera inornata, marking the top of dinocyst zone D1 that occurs within NP3 (Vinken et al. 1988).

22

J.E. NEAL

Danian deposits around northwest Europe have a complex and variable stratigraphy. Depositional sequences are difficult to identify within the Ekofisk Formation due to numerous reworking events and allocthonous units (Johnson 1987; Kennedy 1987), therefore the Ekofisk has been considered here as a single sequence. In Denmark, detailed stratigraphic mapping of chalk facies reveals four sequence cycles of transgression and regression (Thomsen 1990, 1992). The oldest of these cycles occurs within the NP1 zone. Whether one or two cycles can be confidently identified and correlated remains a possibility, but at least one cycle can be correlated to other regions. A thin deposit, known as the 'Pa-layer', assigned to NP1 zone, is also reported in Belgium (Verbeek, pers. comm.) and Bignot (1993) placed the Mont AimEVertus Formation within the lowermost Danian. This sequence cycle does not appear on the eustatic curve of Haq e t al. (1988) and may be recognizable only in northwest Europe. The Marnes de Meudon of the Paris Basin are placed in zones NP2-3 by Aubry (1985), but this is contradicted by Bignot (1993) who positioned the base of the Montian stage boundary in NP3, not NP2. Sediments of NP2 age have only been positively identified in northwest Europe in Denmark and possibly the central North Sea. In the London-Hampshire Basin, and parts of the Paris Basin, the lower Danian section is not present and is represented by an unconformity that erodes down to the Campanian and Santonian (Aubry 1985; Pomerol 1989). Deposits of NP3--4 age exist in the Belgium, Paris and Denmark basins and document another sequence cycle. The Mons Basin near the BelgiumFrance border is a local area of subsidence that preserves deposits of a relative sea level cycle

with the Malogne conglomerate, Vroenhoven and Houthem Formations, and the Calcaire de Ciply (Vandenberghe e t al. 1995). Similar-aged deposits are found south and west of Paris in the Calcaire de Meudon (Vinken e t al. 1988; Pomerol 1989). The '61 Ma' flooding event on the Haq curve could be responsible for these deposits in Belgium and France (Aubry 1985; Bignot 1993; Vandenberghe e t al. 1995) and for the most widespread Danian Limestone depositional event (Thomsen 1992). A sequence cycle not identified by Haq e t al. (1988) has been recently identified in the Gulf of Mexico Alabama section (Mancini & Tew 1991) between the planktonic foraminifera zones of Morozovella uncinata (P2) and M o r o z o v e l l a a n g u l a t a (P3A). Vandenberghe e t al. (1995) note evidence for this cycle in the Maasmechlen Formation (P3A - Hooyberghs 1983), which is a thin transgressive calcarenite above the Calcaire de Ciply. The IGCP Project 124 assigned this unit to the NPF2 and B 1 foraminifera zones (Hooyberghs 1988; Laga 1988). These two fossil zones cover broad time intervals, but overlap only during NP4 (Vinken e t al. 1988).

Paleogene 1 R/TF cycle (Danian chalk/ Thanetian clastics) Uplift of northwest Europe continued through the next sequence cycle, bringing major siliciclastic deposits to the basin due to erosional removal of the blanketing Upper Cretaceous chalks. The type Selandian-Danian stage boundary marks the transition from chalk to siliciclastics in Denmark with deposition of the Lellinge Green Sand and Kerteminde Marl (Rosenkrantz 1924; Thomsen & Heilmann-Clausen 1985; Berggren 1994;

Fig. 5. Sequence stratigraphic framework chart for the Paleogene of Northwest Europe. 1 Age scale comes from Berggren et al. (1995) with nannofossil correlations from Aubry et al. (1988). 2 Outcrop correlations compiled primarily from the work of Heilmann-Clausen (1985, 1994), Thomsen & Heilmann-Clausen (1985), Nielsen et al. (1986), Heilmann-Clausen & Costa (1989), Michelsen et al. 0995), Vandenberghe et al. (1992, 1995), De Coninck (1990), Steurbaut (1988), Plint (1983, 1988), Knox (1994), Knox et al. (1994), All et al. (1993), Eaton (1976), J. Riveline & M. Renard (pers. comm.), Cavalier (1988), Cavalier & Pomerol (1986), Aubry (1983), 1986, 1994), Bignot (1993), G61y & Lorenz (1991) and Steurbaut & Nolf (1986). 3 Seismic sequences in the eastern North Sea (Denmark) comes from Michelsen et al. (1995). 4 Hiatuses in the Paris Basin come from Pomerol (1989). I'~-New IO &SS)], position of a sequence boundary from outcrop and subsurface data not picked in Haq et al. (1988); I-~--New (O) I , position of a sequence boundary from outcrop data not picked in Haq et al. (1988). Correlation of the frameworks from Den Hartog Jager et al. (1993) and Armentrout et al. (1993) are based on biostratigraphic calibration points. The (Haq et al. 1988) eustatic curve is modified based on a correlation of nannofossil zone differences between Haq et al. (1988) and Aubry et al. (1988). Lithology and facies key: hemipelagic mud/sediment starved, r----q;tuff, ~ ; dominantly-silt turbidites, I ; dominantly-sand turbidites, I ~ ; highstand silt/sand, Illn ; coal, I ; transgressive silt/sand, ~ ; erosion or missing section, ~ ; pelagic chalk, ~ ; allochthonous chalk debris, ~,m ; graphic correlation data terrace, I ; sequence boundary (sb), ~ ; incisive s b , " ' N J ; transgressive surface,- . . . . ; lowstand prograding, ~ (sand-silt/mud); R/TF cycle regressive phase, 9R/TF aggradational phase, ~ ; R/TF transgressive phase, , f f ~ 9

I

Ag( )-

i~1 st

uence ana Pormau(

I-'UDIisnea ~equence Prame~

uequence ~ii

|

~1~ ,.F J-,,----R/TF for CNS Central North Sea ,,u~,su,acei___._~., ~r eN-~

!il

0 Key B ~ . Darnel C h r o n o s t r a t i g r a p h i c (;;hart " R-'zI ~- = LAD (firstdownhole) S h e l f > Basin 9 -- = ac.me oc.curre~,ce (Moray Firth).. (Central Graben, . J~ ~ - - - - ~ : m l c r o x o S S l l $ ~oum viKinq ~raoen] IZ~J. palynomorphs = - .

n..~ ~" :~eq.

C N S Seq. F r a m e w o r k of S h e l l Mod. C N S Seq. of M o b i l Modified from f r o m Den H a r t o g J a g e r e t al., (1993) ~ ~J ou'~cro~)- - = F~ ~J . . . . Seq. in outcrop o~ ly Armentrout o ~J<---High sea leve! et al., (1993 ) (:1~1 Low sea leve~

Eustatic Curve Modified from

Haq

et aL,

(1988)

Danish Sub-basin

I SE England

Belgium

~ Formations J.~

Paris Basin

Formations I Formations

Formations

--28

A.arcuatum~--29

%%%%%%%%% %%%%%%%%%%%% %%%%%%%%%%%% Boom

--31

',~Ygie. " / / ,

~ernlel Double Boom

S~cookson/ae~-

_22_.

Kerkom

beds

Stratigraphy

k 4ktymokus~" 19-20

beds

__. i:i:i:

Rise

all

(3?sequences) sand

%%

R. porosum

i!iiiii

H. porosa r"

D. colligerume

Bdoc

~37 IIiI1~1 I I i i i ] i l l l l l I I I I I I t I I i i i i I I i i i i i I I I I I 1 ~ I I i I i i I I I I l i I I I I I~1 I I

~39

I I

I I

I I

I I I I i i I

I I I I I I

m40

I

I I I I I

I

I I I I I

,

''

~ |

,

| SablesdeMames/ SablescrEzanville

hcliff Sand

I-I Ii I

I I I~1 I I I ~1 I I 1 I 1 ~ I I 1 ~ I I 1 ~ I I I

Fisher XVl - Barton Clay

Sevind Marl Fm.

,,

N. prestwichlanus ,.

!!!~!'

,. ,.

:Cyc,e ;';4

~--41

--42

!iiiii

Nauchlan

9 S. placacantha-.

~4~

I W9 up

r-'--" 9

O. flcusoides~" ~44

Fisher Beds X- XV N. variolarlus

Mames et Callaisses su~,urs (Falun de Foulanguem)

Eaton Beds 11 & 12 (Fisher Beds VIIi & IX)

Mames et Callaisses infddeurs

~C. ~ p l ~ n s (common)

Caran

~-4~

23.3 3.3

D. pachydermumr"

i --46

PT23 ~4~

Upper Tay

I

__.

Lillebmlt clay

23.2

E. ursulae --'

! ~-4.

N. laevigatus abundant

to

Fisher Vl

9 E. ursulae -. H. tubiferum r"

~4~

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 i i i i

Upper Frigg

!

ipT

13

Vllerzel~-'Aalter

12

23.1

E~/.Eg~Stone%

O.condylos

(o)

u.M. Frigg'i

0. vn~ongitudum D. politum

---S2

D. simile r

~--S~

?

'~.~er. =C. warrlenense =P~ros~a A. augustum

I

z~nini~y~ FoP tmpove~shedAg; )

m56

(~op

gg.l

Portsmouth

Aalbeke

PT22122.3 22.2 22.1

M. Frigg

L Frigg

. . . . . .

W6

j

BC-12

Upper Sele Middle Sele Upper , Forties

Vmrum mbrJ Fur Fm. (pos. ash series) aslund.member eg. asn ser,es)

PT 20119.4 119.2 I19.1

Upper Balmoral

,~*orchi;,

(o)

PT 21

Balder

LondonCl/Nursling Christchurch

Roubaix

7

Lower Balmoral (incl. Glamk, member)

Cenodiscus

LI

Andrew

Cenodiscus 9

--59

9

~

PT 15

n

Sables de Mons en Pd~de---

B

Walton/London A2

UpOn.oFm_: r

BoisGilles

c~.de Meudon __Brar,beux Cgl. de Cernay

tOD B r a d ~ l l Borehole

~ h e i Clav Tuffeau d p,,~.-In_.r "lb~

,aleqgene 1 ',R/I'F/// 5CyCle ,5,,5/

o

:21~ckh~th

Holmehus Fm.

P. ~58

d

He e'-~i~_.

ilborg Sl

(o)

n

Continental Landen NooIwich v Reading 2 (Knokke ' en ~/Reading 1

D5

BC-IO BC-6

o

D5b

Granr

P. Dvro~horuJ

~

Knokke Member'X Tilehurst/Oldhaven

I 8

~n~s r

D7 ~

roD. oebisfeidensis-~

--55

\

Resn~es clay

9 R tubfferun~ 9 Inaperturopollenites spp.-~ -v~oscinodiscus so. 1&2 Bartenstetn et a/[1962) W. as~ra r"

~sil'lV Kortemerk

BC-141

22.4

S; seem talus (!t, sept#a)

S. linaperta gr.

10

.~em

W. meckelfe~,.sis r"

~Lsummissum r. ~ $. navarroana

~--J~-~Pk~rre ~ U a r d l - Glauconie Grossi~re w7

23i

C edwardsii e 9 PC lenuispinosum -

~50

---51

Fisher VII

Brussels

R~-~u~ulv~rSilt Mar ~ d t h ~;ay Mbr. Base Bed_

3l Gelinde~'--~-nMarl

ables de Chalons-~ CNWons-su Sables

Dark grey non-calc, clay

5 Hainin-Eisden nd

--60

G. compressa IG. ~ . .d.obuloides |E. triviali$ A r~iculata D. calffomica X. lubricurr

--61

Maureen

PT 13 (o)

Maas____mec___hle_nn Houthem & Vroenhov

--62

Calcaire de Meudon

- Shelf I~1-

--63

Ekofisk

~- 9 dimorphosus(NP2.3)

Basinal Shales

- Submarine Fan ITITI-I] - Hiatus - Maximum Flooding Surface - Sequence Boundary - - - Higher Order Sequence Boundary \ N

\ ~65

1

Pa4ayer

~,

Mont Alm~Verlus F

PALEOGENE SEQUENCE STRATIGRAPHYIN NORTHWESTEUROPE Michelsen et al. 1995). This boundary is a transgressive surface. In Belgium, the transgression is marked by deposition of the non-marine Hainin Formation unconformably overlying the Danian carbonates (Vandenberghe et al. 1995). The lowstand deposits of this sequence occur as siliciclastic submarine fans and a prograding deltaic wedge in UK Quadrants 13, 14 and 16. Because the fan deposits in Maureen field (UK well 16/29-4: Neal et al. 1994) fall within this sequence, it is called the Maureen sequence. The FDO of dinocysts Xenicodinium lubricum, Danea californica, and Alisocysta reticulata is noted at the base of the Maureen sequence, biostratigraphically correlating the lowstand deposits with the Selandian-Danian boundary in Denmark (Thomsen & HeilmannClausen 1985). Hardenbol & Berggren (1979) note that Paleogene stage boundaries in northwest Europe are unconformities and the SelandianDanian boundary is no exception (Thomsen & Heilmann-Clausen 1985). In a sequence stratigraphic framework, lowstand fans of the central North Sea subsurface correlate to the time represented by a subaerial unconformity in updip (onshore) sections. By this reasoning, the Maureen fan deposits are placed within the Danian (Fig. 5). Peak Paleocene uplift of northern Britain is placed at 6 0 M a (Green 1989), but shoreline regression continued in the North Sea as subsidence began to slowly increase (Fig. 4). The upper part of the Selandian contains another major North Sea submarine fan system, depositing thick sands in UK Quads 14, 15, 20 and 21 and a welldeveloped progradational shelf that is deeply eroded in Quad 14 (Jones & Milton 1994; Neal 1994). This unit is designated the Andrew sequence because it closely corresponds to the lithostratigraphic unit of the same name (Deegan & Scull 1977; Mudge & Copestake 1992a). It contains the biostratigraphic markers of the FDO of Isabelidinium? viborgense, the acme and FDO of Paleocystodinium australinum (bulliforme) and the acme of the radiolarian Cenodiscus (Fig. 5). The position of the so-called 'Cenodiscus claystone' (Stewart 1987; Mudge & Copestake 1992; O'Connor & Walker 1993) within this sequence bears some explaining as it is a likely source of confusion between sequence stratigraphic and lithostratigraphic frameworks. Present lithostratigraphic schemes place the Cenodiscus (Cenosphaera lenticularis) shale as the boundary between the Andrew and Maureen Formations (Knox & Holloway 1992; Mudge & Copestake 1992a). Published sequence stratigraphic frameworks have identified this marker as a highstand condensed section separating two sequences (top Sequence 2, Stewart 1987, p. 562; top Upper Maureen sequence, O'Connor & Walker 1993,

23

p. 151). While this claystone may, indeed, occur at a sequence boundary at certain locations, i.e. at 9024ft (2750.5 m) in well 16/18-1 of Knox & Holloway (1992, p. 67), it does not correlate to a consistently mappable seismic reflector in the data set used by the present study. Also, the Cenodiscus claystone itself has been found to be diachronous within the composite standard biostratigraphy used for this study (J. Stein, pers. comm.; Neal 1994). An example where the Cenodiscus claystone occurs above the Maureen-Andrew sequence boundary of this study is in well 14/25-1, where the MaureenAndrew lithostratigraphic boundary occurs at 7088 ft (2160.5 m) depth (Knox & Holloway 1992, p. 67), but the seismically-defined sequence boundary occurs closer to 7260 ft (2213 m) (Neal 1994; Neal et al. 1995). A tie of this well to the seismic line is shown by Armentrout et al. (1993, p. 49), who pick the sequence boundary higher in the section. Neal et al. (1995) demonstrate with additional wells, tied to the seismic line with synthetic traces, how the Maureen-Andrew sequence boundary is correlated in this study. Additional well ties show the sequence boundary occurring at 1885m (6184ft) in well 14/30-1 [lithostratigraphic boundary at 1788 m (5866 ft), Knox & Holloway 1992, p. 105] and at 1810 m (5938ft) in well 14/29-1 [lithostratigraphic boundary at 1760 m (5774 ft), Knox & Holloway 1992, p. 103]. The occurrence of L? viborgense within the Andrew sequence is a key tie to the stratigraphic succession in the Viborg borehole of Denmark (Heilmann-Clausen 1985). This borehole provides a link between North Sea sequences and the nannofossil stratigraphy of northwest Europe, as Viborg dinoflagellate zones have been indirectly correlated to the standard NP nannofossil zone (Powell 1992; Michelsen et al. 1995). Viborg zones 2 and 3 (Cerodinium speciosum and Palaeoperidinium pyrophorum zones of Powell 1992) are correlated to NP5-7 (Powell 1992; Michelsen et al. 1995). Deposits of NP6-7 age mark the onset of the Thanetian transgression in the Belgium, Paris and London-Hampshire basins (Aubry 1985). Some of these deposits stratigraphically overlapping in time with type Selandian deposits in Denmark (Thomsen & Heilmann-Clausen 1985; Berggren 1994; Knox 1994a). Graphic correlation of the Viborg borehole (Fig. 6) displays data terraces that overlap with similar graphs of North Sea wells. The overlap indicates a shared time of non-deposition (Neal 1994), which can be used to correlate lowstanddominated sequences Of the central North Sea to transgressive and highstand deposits in onshore sections. Since the Danish Paleogene section was deposited in a relatively deep water setting, graphs of this section closely match those of the North

24

J.E. NEAL

Fig. 6. Graphic correlation plot of dinoflagellate stratigraphy from the Viborg borehole (Heilmann-Clausen 1985). Nannofossil correlations come from the stratigraphic framework of Michelsen et al. (1995) and Powell (1992), correlated to Viborg dinoflagellate zones. Consistent times of overlapping data terraces from the central North Sea are plotted at the top for comparison and link to CNS sequences. Sea, and subsequent sequences can be correlated through this tie point. Using this method, the Andrew sequence of the central North Sea is correlated to deposits in northwest Europe that occur in the overlap between type Thanetian and type Selandian (Fig. 5).

Thanetian deposition The early Thanetian transgression is an indication of increased subsidence around northwest Europe and the transgressive phase of the first R/TF cycle of the Paleogene. This transgression commenced in Denmark with the Lellinge Greensand, which is

correlated with the Maureen sequence of the central North Sea (Fig. 5b) in agreement with Knox (1994b). Sediment supply in the central North Sea overwhelms the background sea level rise through the Andrew sequence and into the next sequence cycle, the Lower Balmoral (after Knox & Holloway 1992; O'Connor & Walker 1993; equivalent to the lower part of the Balmoral Formation of Mudge & Copestake 1992a). This sequence contains the sandiest and most widespread submarine fan-apron complex of the Paleogene ('Andrew' - Den Hartog Jager e t al. 1993; 'MT 4'-Morton e t al. 1993). The Lower Balmoral sequence also contains Thanetian volcaniclastics known as the Balmoral Tuffite

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE (Knox & Holloway 1992) or Glamis Member (Mudge & Copestake 1992a). This sequence is identified biostratigraphically with an acme occurrence of the dinocyst Palaeoperidinium pyrophorum at its base and the FDO of the acme of Areoligera gippingensis at its top. This upper boundary is important since it has been identified as part of a significant graphic correlation terrace in North Sea wells (Neal et al. 1994). When correlated into Danish sections, this terrace falls within the middle of the Holmehus Formation and spans most of the NP8 zone (Fig. 6). The NP8 zone has been reported to contain one, and possibly two, sequences (Knox et al. 1994; Vandenberghe et al. 1995). In Belgium, the Gelinden Marl has a bored upper boundary that is overlain by the glauconitic Tuffeau de Lincent and the Waterschei Clay, indicating a highstand sealevel cycle (Vandenberghe et al. in press). The Granglise sand appears locally above the Waterschei and may suggest a third 'Thanetian' sequence cycle (Vandenberghe et al. in press). Evidence for three lower Thanetian relative sea level cycles was found in the Bradwell borehole of southeast England, with glauconitic sands being found at the formation contact between the Pegwell Marls and the Reculver Silt and near the top of the Reculver Silt (Knox et al. 1994). The acme of A. gippingensis has been reported in the Pegwell Marls and is correlated with the Lista Formation of the North Sea (Powell et al. 1993). A single unit, the Sables de Tillet, has been identified as NP8 in the Paris Basin (Aubry 1994), but this does not exclude the possibility of two NP8 cycles. The two NP8 cycles are difficult to correlate with the central North Sea as biostratigraphic resolution in this study is insufficient to distinguish two separate sequences. This study can demonstrate the correlation of one sequence and a graphic correlation terrace into Thanetian sections underlying the 9Lambeth Group (Ellison et al. 1994; Knox 1994a) onshore (Fig. 5b). Some published frameworks have identified of two fan units for this time interval based on physical stratigraphy (Milton et al. 1990; Den Hartog Jager et al. 1993). Haq et al. (1988) considered the that the Thanet Formation represents a single transgressive event ranging from '58.5 Ma' to '55 Ma', therefore, two additional sequence cycles are present in northwest Europe that have not been globally correlated.

Paleogene 2 (Upper Paleocene) R/TF cycle deposition The next North Sea R/TF cycle, the Paleogene 2 (Fig. 4), has its lower boundary at the A. gippingensis or Upper Lista graphic correlation data terrace. The first central North Sea sequence cycle

25

in this package is called the Upper Balmoral (from the upper part of the Balmoral Member of Mudge & Copestake 1992a). Further explanation is required to clarify the position of this sequence in relation other published work. O'Connor & Walker (1993) examined in great detail the relationship of depositional sequences in the 'Forties' interval to lithostratigraphic divisions of the Sele and Lista Formations. The Upper Balmoral sequence of the present study is equivalent to the Lower Forties sequence of O'Connor & Walker (1993). Confusion over lithostratigraphic nomenclature arises when sands of this sequence overlap the List-Sele Formation boundary and a biostratigraphic or seismic tie is necessary to identify the sequence boundary. Wells around Forties Field provide type examples for the Upper Balmoral (Lower Forties) sequence. The Balmoral Member of Mudge & Copestake (1992a, p. 58) occurs above 2370 m, a depth coincident with the 'Upper Lista' graphic correlation terrace of Stein et al. (1995). In this well, the Balmoral Member falls within the Paleogene 2 R/TF cycle, whereas in well 14/25-1 the Upper Lista terrace occurs at 1609 m in the well and the Balmoral Member is placed within the Paleogene 1 RfrF cycle. An additional example comes from the discovery well of Nelson field, 22/11-1, where a thick sandy package occurs from 2438 to 2298 m. This sandy deposit occurs above the Upper Lista terrace but below the high gamma-ray spike noted by O'Connor & Walker (1993), which marks the base of the Apectodinium augustum-rich Upper Forties sequence. The Upper Balmoral sequence commonly contains the FDO of Alisocysta margarita and marks the start of a bottom-water oxidation crisis that limits, then excludes, agglutinated foraminifera from the sediment. This crisis is also responsible for the change from grey-green bioturbated mudstones of the Lista Formation to dark grey laminated mudstones of the Sele Formation (Mudge & Copestake 1992a). This boundary is noted in Denmark where grey Osterrenden clay is found above the green Holmehus Formation (Heilmann-Clausen et al. 1985). North Sea agglutinated foraminifera markers (M4 and M5 Mudge & Copestake 1992a) are somewhat controlled by sedimentary facies and can range into the next younger sequence, a relationship thoroughly discussed in O'Connor and Walker (1993). The precise upper sequence boundary is usually picked at a gamma-ray log spike interpreted to indicate sediment starvation. Above this spike the Forties sandstones appear, containing abundant dinocysts, especially Apectodinium augustum. Graphic correlation of the dinoflagellates in the D.G.I. 83101 borehole (Nielsen et al. 1986) at Osterrenden (Fig. 7) demonstrates the tie

26

J.E. NEAL

o

. ,,...

9 ,,-.z, c.o

2; E o

c

,..., o

o ,.c

o

=~ -L

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

of North Sea sequences to a lithological column with additional formations not seen at Viborg. Outcrop evidence for sequences in the Paleogene 2 R/TF cycle is most complete in the LondonHampshire Basin. The Upnor Formation starts deposition of this cycle above the 'Chron C25n hiatus' (Aubry et al. 1986; All 1994). Although Chron C25n has now been identified in a London area borehole (404T-Ali 1994; Ellison et al. 1994), most of the Upnor formation is considered as transgressive deposits for the North Sea Upper Balmoral sequence (Fig. 5). The Chron C25n sediments could be a lowstand remnant due to their restricted distribution (Ali 1994) and should correlate within the Upper Balmoral lowstand (sometimes called 'Lower Forties' sandstone - see lithostratigraphic note above), which would agree with the prediction of Knox (1990, p. 60). The Upnor Formation has been identified as NP9 in age and correlated to the Bois Gilles sand of Belgium and the Sables de Bracheux in France (Aubry 1986, 1994). The Woolwich and Reading Formations are nonmarine to marginal marine and have been divided into upper and lower parts based on the transgression of the Woolwich shell beds (King 1991; Ellison et al. 1994). This shift to non-marine deposition coincides with the peak regression of the Upper Paleocene cycle in the central North Sea. Palynological investigations of the Lambeth Group reveal the presence of abundant Apectodinium sp. (Powell et al. 1993; Ellison et al. 1994). This marker provides a biostratigraphic tie to the North Sea Upper Forties sequence with correlation through the Osterrenden borehole of Denmark (Fig. 7), where the base Apectodinium sp. acme zone occurs with the top A. augustum (Nielsen et al. 1986), a correlation observed by Powell (1992) as corresponding to nannofossil zone NP9 (Powell 1988; Ellison et al. 1994). As the major R/TF cycle transforms from regressive to transgressive, another sequence cycle can be identified on seismic and logs (Neal 1994), which is designated as the Middle Sele and contains acme occurrences of the acritarch Pterospermella and the dinocyst Cerodinium wardenense. This sequence also contains a palynofacies of large leiospheres that has its FDO in a younger sequence, the Upper Sele. The sequences occurring at the peak regression and aggradational phase of the Paleogene 2 R/TF cycle can be correlated from the North Sea to the upper and lower Reading with full transgression continuing up to the London Clay B Division of King (1981). Absolute magnitude of high frequency (sequence cycle) sea-level rise and fall is difficult to determine and even more difficult to correlate, especially within the Paleogene 2 R/TF cycle. The sequence boundary at the base of the Upper Forties

27

sequence in the North Sea is demonstrated on seismic data to record a minimum 150 m sea level (Neal 1994; Neal et al. 1994). This major fall could not have the same magnitude around northwest Europe (as illustrated by the lack of evidence for such a fall in the outer shelf deposits of Denmark Heilmann-Clausen et al. 1985) and is thought to be enhanced (Vail et al. 1991) in the North Sea by tectonic uplift of the Shetland Platform (Milton et al. 1990). This sea level fall is, none the less, a major event evidenced by the fact that the Upper Forties sequence has no equivalent in the Paris Basin (HP4 - Pomerol 1989) and possibly Belgium, where the Continental Landen deposits cannot be uniquely placed sequence stratigraphically (Vandenberghe et al. 1995). The magnitude of relative sea-level falls can also appear greater onshore than in the North Sea. This is the case for the North Sea Upper Sele sequence that correlates with the deeply incisive Blackheath-Hales Clay sequence of the London-Hampshire Basin. The Upper Sele sequence occurs as the R/TF cycle transgression is increasing, while in western Europe there is evidence that the London-Brabant Massif is being uplifted (Cavelier & Pomerol 1979; Vandenberghe et al. 1995). Such local factors prohibit the designation of absolute magnitudes, but do not hinder correlations based on biostratigraphy and major R/TF cycles.

Upper Paleocene-lower Eocene deposition As the 'Upper Paleocene' transgression begins onshore, it is appropriate to discuss the PaleoceneEocene boundary as it relates to the sequence stratigraphy. Clearly this transgression continues into the Eocene, but where does one pick the boundary? Recent work on 813C isotopes suggests that an excursion towards light values around the Paleocene-Eocene boundary might be a good marker since it has been seen in many depositional environments and even the non-marine Sparnacian deposits of the Paris Basin (Sinha & Stott 1993; Stott et al. 1995; Thomas & Shackleton 1995). Another marker that has been recognized around northwest Europe and the North Atlantic is the ashfall tufts of the Balder and Olst Formations (Knox & Harland 1979; Heilmann-Clausen et al. 1985; Knox & Morton 1988). A detailed discussion of the problems with the Paleocene-Eocene boundary is covered in the meeting proceedings of the Paleocene Working Group, introduced by Schmitz (1994). Four sequences are observed in the central North Sea during Upper Paleocene regression and aggradation until the basin becomes starved of sediment in the early Ypresian. The Upper Forties, Middle and Upper Sele sequences have already

28

J.E. NEAL

been discussed briefly with only the Balder sequence remaining. The Balder sequence encompasses most of the Balder Formation and part of the Beauly Member of Knox & Holloway (1992). The Balder sequence has its base above the FDO of the large leiospheres palynofacies and the FDO of the acme of C. wardenense. Sediments containing these two markers have been placed within the Upper Sele sequence. The top of the Balder sequence is a major graphic correlation terrace containing acme occurrences of Inaperturopollenites spp. and Hystrichosphaeridium tubiferum. The sequence itself contains the FDO of the acme of Deflandrea oebisfeldensis and the FDO of Wetzeliella astra. Using graphic correlation, a tie of the Balder sequence to the V~erum Member of the Olst Formation is demonstrated (Figs 6 & 7). This tie is extended to the volcaniclastic Member X of King (1990) in the Knokke well of Belgium and the Oldhaven/Harwich (London Clay Division A1 King 1981) sequence in the London-Hampshire Basin. The 513C spike is found in type Sparnacian deposits of the Paris Basin (Sinha & Stott 1993; Stott et al. 1995), which are traditionally correlated with the Reading Beds of England. The Ypresian (i.e. lower Eocene boundary) traditionally commences with the start of the London Clay transgression (Knox 1994a), which is one, or possibly two, sequences younger than the 613C spike. The ~513Cspike should not be confused with the gammaray spike at the top of the central North Sea's Upper Balmoral sequence. The gamma-ray log spike is correlated as being a sequence older than the fi13C excursion. This study places the PaleoceneEocene boundary at its traditional location, the transgressive surface of the Balder sequence (Fig. 5), which is different from Powell (1992) and Knox & Holloway (1992), who place the boundary at the LAD of A. augustum. If this convention is followed, the Upper and Middle Sele sequences will be placed within the Eocene. The final placement of this important boundary remains a topic of much discussion.

Ypresian deposition Lower Eocene deposition resumed in the central North Sea within the area covered by this study after a significant graphic correlation terrace and marine hiatus at the top of the Balder sequence. Graphic correlation of the dinoflagellates in the Wursterheide borehole of northern Germany (Heilmann-Clausen & Costa 1989) displays an expansion of the Balder terrace into 100 m of section. Lithostratigraphically, the Balder terrace is associated with a condensed section of red mudstone equivalent to the RCsna~s clay of

Denmark (Heilmann-Clausen et al. 1985), but the expanded 'Lower Eocene 3' Formation of Wursterheide represents significant deposition occurring at the same time sediment starvation is prevalent in Denmark and the central North Sea. Knox & Holloway (1992) also note that 'Balder Sands' (Odin Member of Mudge & Copestake 1992b) represent significant deposition during this time interval. Although the northern North Sea section was not interpreted for depositional sequences within the scope of the present study, the fact that additional sequences are observed in onshore sections of the London and leper Clays opens up additional possibilities for improvement to the present framework. The expanded section of the Wursterheide borehole was logged with a gamma-ray tool and that log can be interpreted for depositional sequences (Fig. 8). Similar log patterns in the leper Clays are interpreted as fining- and coarsening-upward trends caused by relative sea-level change (Vandenberghe et al. 1988; 1995). Key dinoflagellate FDO from the North Sea correlate with LAD in Belgium (De Coninck 1990), Denmark (Nielsen et al. 1986) and north Germany (Heilmann-Clausen & Costa 1989) to provide the biostratigraphic framework for early Eocene depositional sequences. The Lower Frigg sequence commences Eocene North Sea deposition above the Balder terrace. This sequence is best developed in the Frigg Field area of the South Viking Graben where thick, sandy submarine fans occur (Heritier et al. 1980). The FDO of D. oebisfeldensis, Dracodinium solidum and Dracodinium simile commonly occur in this sequence. HeilmannClausen & Costa (1989) find these markers in the Wursterheide well, and place the sediments containing these fossils within the D 7 zone. These fossils are graphically correlated from the North Sea to Denmark, positioning the Lower Frigg sequence within the lower ROsn~es Clay (Fig. 7), above the hiatus between the RCsn~es and Olst formations (Nielsen et al. 1986). These markers also locate the Lower Frigg sequence within the Roubaix Clay in Belgium (De Coninck 1990), correlated to sequence 3 of Vandenberghe et al. (1995), which contains the N P l l - 1 2 boundary (Steurbaut & Nolf 1986). With the dinoflagellate and nannofossil information, the Lower Frigg sequence is placed within the same depositional sequence as London Clay Division C (King 1981) and the Sables de Mons-en-P6v~le (Steurbaut & Nolf 1986; De Coninck 1990) in the LondonHampshire and Paris basins (Fig. 5b). This means that Ieper sequences 2 and 1 of Vandenberghe et al. (1995), which fall within the NPll-upper NP10 range, are contained within the Balder terrace of the North Sea (Fig. 5b). This observation is confirmed

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

29

Fig. 8. Graphic correlation of dinoflagellates in the North German Wursterheide borehole from Heilmann-Clausen & Costa (1989). In this section, the 'Balder' terrace observed in all central North Sea wells corresponds to an interval of deposition. Interpretation of the gamma-ray log identifies depositional sequences that are correlated to sequences observed elsewhere in northwest Europe. A terrace at 571 m well depth corresponds to a possible condensed section or sediment starvation hiatus.

30

J.E. NEAL

in the Wursterheide well and can be extended to the London-Hampshire Basin where King (1981) notes relative sea-level cycles for the Bognor and Walton Members of the London Clay. The central North Sea Paleogene 3 R/TF cycle starts with the Lower Frigg sequence. This R/TF cycle is smaller in magnitude than the preceding cycles and is less well coupled to onshore sections. Three major sequence cycles can be confidently identified within the Lower Eocene, with local evidence for possibly two additional sequences. The Middle Frigg sequence succeeds the Lower Frigg and contains the FDO of many key markers. The Middle Frigg sequence is capped by the Intra-Frigg graphic correlation terrace. This sequence and its upper boundary terrace often contain the FDO of dinocysts Dracodinium politum, Dracodinium varielongitudum, Spiniferites septatus, Dracodinium condylos, Apectodinium summissum and Wetzeliella meckelfeldensis. Also found are the FDO of foraminifera Spiroplectammina navarroana and the Subbotina linaperta group, although these markers have been observed in deposits with Upper Frigg dinoflagellates as well. The Upper Frigg sequence commonly contains the FDO of Charlesdownia (Kisselovia) edwardsii, Hystrichosphaeridium tubiferum and of acme Homotryblium tenuispinosum. The Upper Frigg sequence is capped by a major graphic correlation terrace that contains the FDO of Eatonicysta ursulae. Correlation to onshore sections note the consistency of this marker, but onshore data suggests that two better markers could be used to identify Upper Frigg-equivalent deposits, namely the first appearance datum (FAD) of Dracodinium pachydermum and an 'apparent' LAD of Kisselovia coleothrypta. These markers are difficult to use for the central North Sea composite standard. Fossil FAD (or bases) are unreliable data points in wells using only ditch-cutting sampling, making the base D. pachydermum an ineffective marker for correlations from the central North Sea to onshore section that have core or outcrop sampling. The true LAD of K. coleothrypta occurs in the Lower Oligocene (D 13 - Vinken et al. 1988), and it is carried there in the North Sea composite standard. In Denmark and northern Germany (Heilmann-Clausen & Costa 1989; Michelsen et al. 1995) the K. coleothrypta zone (based on the FAD) occurs in D 8 and a disappearance of the marker occurs in the early-mid Eocene D 9 zone. These inconsistencies illustrate the fact that certain key local markers can be problematic or unreliable for correlation of subsurface to onshore sections. The lower Eocene sections of the LondonHampshire and Belgium Basins have been studied for depositional sequences (e.g. Plint 1988;

Vandenberghe et al. 1995), producing stratigraphic frameworks that correlate to eustatic cycles on the Haq et al. (1987) curve. Correlation problems exist for this interval since three to five sequences are identified in the central North Sea subsurface, but similar-aged sections onshore can arguably display seven depositional sequences. Eaton (1976) published dinocyst data from the Bracklesham Beds on the Isle of Wight at Whitecliff Bay. Those data are graphed with the North Sea composite standard (Fig. 9). This interpretation of the same data differs slightly from that published by Neal et al. (1994). Depositional sequences were interpreted in this section by Plint (1988), based on a correlation to the Haq et al. (1987) eustatic curve and on sedimentological investigations previously published (Plint 1983). Neal et al. (1994) interpreted depositional sequences within this section, based on Plint's work and graphic correlation of Eaton's dinoflagellate succession (Fig. 9). The interpretation of Whitecliff Bay's sequence stratigraphy shown in Fig. 9 differs from that presented by Neal et al. (1994) as a result of discussions about the Belgian section (N. Vandenberghe, pets. comm.) and the sequence stratigraphic implications for the central North Sea correlation. Fisher (1862) Beds IV and V are now interpreted to fall within the 'Intra-Ffigg' hiatus in the subsurface, honoufing the LAD of W. meckelfeldensis. Revised correlation of Fisher Beds VIII-XIX comes from a more precise tie to the framework of Newton & Flanagan (1993). The most difficult task in correlating North Sea submarine fans to Isle of Wight exposure surfaces is also the most meaningful addition to a total stratigraphic analysis. Graphic correlation supplements traditional correlation methods and suggests alternative correlations. Figure l0 is a summary chronostratigraphic chart that focuses only on the correlation of the Lower Eocene R/TF cycle and its constituent sequence cycles onshore and in the central North Sea. This chart is a synthesis of the sequence stratigraphy, zonal and key marker biostratigraphy, and graphic correlation interpretations from Whitecliff Bay (Isle of Wight - Eaton 1976), the Wursterheide borehole (N Germany Heilmann-Clausen & Costa 1989) and various publications from Belgium. This time interval is lumped into one sequence in Denmark (Michelsen et al. 1995), and is mostly hiatal in the Paris Basin (Pomerol 1989).

Paleogene 3 R/TF cycle and sequence cycle summary Correlation of the Paleogene 3 R/TF cycle begins with the Balder terrace, which expands into

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

31

o o "~,

e~0 ~.~ ,.~

~ o o

32

J.E. NEAL

multiple depositional sequences onshore and possibly in the northern North Sea (Fig. 8 discussed above). The o~-sequence represents the youngest cycle that is recognized onshore, but remains starved in the central North Sea. This sequence corresponds to the regressive pulse responsible for deposition of the Bognor Member of the London Clay (King 1981, 1991) as well as the Orchies sequence of Belgium between Ypresian sequence boundaries 2 and 3 of Vandenberghe et al. (1995). This sequence is fully within the NP11 zone (Steurbaut & Nolf 1986) and should record a normal magnetic polarity zone, the Ieper-1 of Ali et al. (1993) or C24n.3n of Cande and Kent (1992). The [3-sequence marks the renewal of central North Sea deposition with the Lower Frigg sequence. As noted above, this sequence correlates with the Roubaix sequence of Belgium (Vandenberghe et al. 1995) that spans the NPll-12 boundary. The Christchurch Member (Division C; King 1981) of the London Clay falls between solid nannofossil dates of N P l l for Division B1 and NP12 for Fisher (1862) Bed IV (Aubry 1986). The Nursling Member (King 1981, 1991) is included in this sequence as well, though an additional sequence may exist here based on King's summary figure (King 1981, p. 22). A correlation of these deposits in the London-Hampshire Basin to the Roubaix Clay of Belgium is supported by dinoflagellate correlations, although the LAD of D. simile appears to be stratigraphically higher at Whitecliff than expected from correlation to Belgium (De Coninck 1990). Correlations from onshore sections to the North Sea become very complicated with the next sequences. The Middle Frigg sequence in the central North Sea contains many key biostratigraphic markers, listed above, that make correlation into the Wursterheide well relatively straightforward. However, many of these key markers do not have the same stratigraphic range at Wursterheide as they do in the central North Sea. This observation is demonstrated by the LOC, which passes through relatively few tops in the Portsmouth-Christchurch equivalent interval (c. 625-575 m). The LOC does, however, define a data terrace, associated with the occurrence of the FAD of Samlandia chlamydophora with the LAD of A. summissum. According to the dinoflagellate correlation chart of Vinken et al. (1988), this terrace would represent the upper part of zone D 8. The z-sequence combines Middle Frigg deposition with sediments that contain typical Middle Frigg dinoflagellates in the Wursterheide well, but occur below the graphic correlation terrace. The z-sequence is correlated to England and Belgium as the sequence occurring immediately above the 13-sequence. The graphic correlation terrace above

the z-sequence at Wursterheide is also associated with the FAD of Areosphaeridium diktyoplokus (although this marker actually occurs 1 m above the terrace sample). The FAD of A. diktyoplokus is correlated to NP13 at Whitecliff Bay (Eaton 1976; Aubry 1986) and in the IGCP Project 124 correlation chart (Vinken et al. 1988). Sequence analysis of the stratigraphic successions in Belgium (Vandenberghe et al. 1995) and at Whitecliff Bay (Plint 1988) suggest that three depositional sequences are recognizable from the ]]-sequence to NP13-aged sediments. In Belgium, Vandenberghe et al. (1995) have recognized sequences encompassing the Aalbeke Clay (z-sequence), Kortemark Silt (A-sequence) and Egem Sand (e-sequence) before NP13 is encountered in the Merelbeke Clay. Likewise, Plint (1988) recognizes sequence boundaries at the base of the Portsmouth Member (z-sequence), Whitecliff Member (A-sequence) and Fisher (1862) Bed II (e-sequence) before Aubry (1986) can confidently identify NP13 at the top of Fisher Bed VI. The two youngest of these three sequences are correlated within the graphic correlation terrace at 571 m well depth at Wursterheide (labelled 'Condensed Sequences?' in Fig. 8). The relationship to central North Sea Middle Frigg deposition is less clear as the IntraFrigg terrace can vary in length (see discussion in Neal et al. 1994). I make an attempt at this correlation in Fig. 10, but fully realize that additional work is required to validate or contradict this tie. The Cuisian(Ypresian)-Lutetian boundary is another stratigraphic problem that is not as simple as it might appear on the surface. Aubry (1991) examined this boundary with high resolution nannofossil stratigraphy around the world, highlighting the complexity of this international stage boundary. In northwest Europe, the Lutetian transgression occurs with the Glauconie grossibre in the Paris Basin, the Brussels Sands in Belgium, and Fisher Beds VI and VII in England, all of which occur in NP14 (Aubry 1986; Steurbaut & Nolf 1986). In Belgium, the deeply incisive Vlierzele sequence (De Baptiste 1989) underlies the Brussels Sands and on the Isle of Wight, Fisher Bed V at Whitecliff Bay occurs beneath Beds VI and VII. The equivalence of these units to the CuisianLutetian unconformity in the Paris Basin is explored below. A well documented hiatus occurs between Fisher Beds IV and V at Whitecliff Bay, where Aubry et al. (1986) demonstrate a missing normal polarity magnetostratigraphic event at the YpresianLutetian stage boundary using nannofossil stratigraphy. Ali et al. (1993) point out that this hiatus occurs in a data-poor section and the actual missing normal polarity event could correlate through more than one location in Fisher Bed V. The sequence

PALEOGENE SEQUENCE STRATIGRAPHYIN NORTHWEST EUROPE

33

Fig. 10. Detailed correlation of depositional sequences in the lower to middle Eocene of northwest Europe. 1 Based on data from Heilmann-Clausen & Costa (1989). 2 Based on data from Aubry et al. (1986), Eaton (1976) and Plint (1983, 1988). 3 Based on the framework of Vandenberghe et al. (1995). I--, FAD; - q , LAD.

stratigraphic interpretation I make of the missing section at the base of Fisher Bed VI is done with the aid of graphic correlation and is different from that of Aubry et al. (1986) and Plint (1988), but allowable within data control points from the most recent magnetostratigraphy of Ali et al. (1993). Sequence stratigraphic interpretation of the Wursterheide well above 571 m recognizes three depositional sequences. The oldest sequence occurring above the graphic correlation terrace ties precisely with the consistent Intra-Frigg terrace of the central North Sea (Fig. 8). With graphic correlation, Fisher Bed V appears to also overlaps with the Intra-Frigg terrace of the North Sea. Correlative deposition in Belgium is difficult to document since no well or borehole was graphed with the composite standard. In the Egem (Ampe) clay and sand pit a limestone layer is found at the top of the Egem Sand. This 'Egem Stone Ban' may represent another sequence boundary, which was noted, but not correlated by Vandenberghe et al. (1995). Fisher Bed V is correlated here with the Egem Stone Band-Merelbeke sequence (0-sequence) below the incisive Vlierzele. The incisive nature of the Vlierzele testifies to the magnitude of this sequence cycle sea level fall.

In the central North Sea, the Upper Frigg sequence (y-sequence) is correlated to this sea level fall, as is the middle Wursterheide sequence above 571 m and below the major unconformity at the base of the Rupelian (Fig. 8). This tie is corroborated by the LAD of C. e d w a r d s i i in both the central North Sea and Wursterheide sections. The magnitude of this sea level fall is so great that sea level apparently does not subsequently onlap far enough inland above the sequence boundary to create the accommodation space necessary for a new sequence. The result is a composite unconformity in the Whitecliff Bay section at the base of Fisher Bed VI, and nondeposition of the q~-sequence. An interesting aspect of the incisive Vlierzele Sand is its estuarine, not fluvial, nature (Houthuys & Gullentops 1988). A major graphic correlation terrace occurs in the central North Sea above the Upper Frigg sequence. This terrace is interpreted to represent sediment starvation in the deep basin due to an overall transgression in northwest Europe (Fig. 4). The estuarine fill of Vlierzele channels is an indication of long-term increasing shelfal accommodation space (Posamentier & Vail 1988). Above the Vlierzele, the Brussels Sands transgress far inland, leaving remnants that are onlapped by the next

34

J.E. NEAL

(Lede) transgression (Vandenberghe et al. 1995). The transgressive phase of the Lower Eocene R/TF cycle provides for deposition onshore (q-sequence) that is starved in the central North Sea.

Paleogene 4 R/TF cycle The mid to late Eocene was a time of regression following the peak Lutetian transgression. The upper and middle Eocene is noted especially for submarine fan deposition in the Gannet (Armstrong et al. 1987) and Alba (Harding et al. 1990; Newton & Flanagan 1993) areas. The Tay sand in the Gannet area (southwest comer of UK Quad 21) is divided into an upper and lower part with the boundary occurring at a mudstone unit containing the E. ursulae marker (Knox & Holloway 1992). The Lower Tay correlates with the Upper Frigg sequence as the FAD of A. diktyoplokus is found in a cored section near the base of the Lower Tay Sand in the type section well UK 21/18-1A (Robertson Research 1987, unpublished Central Graben Report; Knox & Holloway 1992). Graphic correlation of this well places the Top Frigg terrace with the E. ursulae marker and the Upper Tay deposition falls below another terrace associated with the FDO of Diphyesficusoides and the FDO of common Cyclammina? amplectens (Base Nauchlan). In the Alba area (southwest comer of UK Quad 16, northwest corner of UK Quad 22), PaleoServices has constructed a detailed biostratigraphy that Newton & Flanagan (1993) correlated to lithostratigraphic units. This framework is adopted here, with the units published by Newton & Flanagan (1993) being namesakes for depositional sequences of similar age (Fig. 5a). The Caran Sandstone is equivalent to the Upper Tay of Knox & Holloway (1992). However, well-logs from the Outer Moray Firth area studied with graphic correlation and sequence analysis suggest that at least two depositional sequences exist between the Top Frigg and Base Nauchlan terraces. The FDO of Dracodinium pachydermum serves as a provisional marker between these two sequences, here called the Upper Tay and Caran (Fig. 5a). Heilmann-Clausen & Costa (1989) place the LAD of D. pachydermum within D 10? zone, which is consistent with the PaleoServices scheme that locates D. pachydermum at near the base of D 10 in NP15 (B. Braham, pers. comm.). IGCP Project 124 agrees with the placement of the LAD of D. pachydermum in D 10, but correlates the top of this zone to NP17 (Vinken et al. 1988). Above the Caran sequence, a consistent graphic correlation terrace occurs below what is termed the Nauchlan sequence (Fig. 5a). This terrace is

associated with the FDO of D. ficusoides and common C.? amplectens. Within the sequence, high abundances of Systematophora placacantha and the FDO of Spiroplectammina aft. spectabilis sensu Knox & Holloway (1992) occur. Above the Nauchlan sequence, the Brioc sequence is identified as the youngest package of the Paleogene 4 central North Sea R/TF cycle (Fig. 5a). The Brioc sequence contains discrete sandy fan deposits in the Outer Moray Firth and South Viking Graben areas, but also develops sharp-based transgressive sands with a blocky well gamma-log signature in western parts of the basin. The base of the Brioc sequence is often marked by the FDO of Heteraulacacysta porosa. Other biostratigraphic markers common to this sequence include the FDO of Diphyes colligerum and Rhombodinium porosum, with the top of the sequence marked by a major graphic correlation terrace that includes the FDO of A. diktyoplokus. The middle and upper Eocene central North Sea succession is difficult to tie with onshore sections. Graphic correlation of data from the Isle of Wight (Eaton 1976) and Gartow borehole (K/3the 1990) yields mixed results, due to the complications involved in graphic correlations of sections from different depositional environments (Neal 1994). Therefore, the sequence stratigraphy of northwest Europe for the middle and upper Eocene relies heavily on the shallow water sections studied in Belgium (Vandenberghe et at. 1995), the Isle of Wight (Plint 1988) and Paris Basin. Information in the Paris Basin comes from ongoing stable isotope work in the Marnes et Callaisses Formation (J. Riveline, pers. comm.) and a study of hiatuses throughout the succession (Pomerol 1989). All these studies have correlated observed sequence boundaries and hiatuses to eustatic events on the Haq et al. (1988, 1987) chart. Correlation of the Bartonian and Lutetian sections for all three basins show a match in number of events and timing of transgressions and regressions. An indirect tie to Denmark using nannofossils places the entire middle and upper Eocene within one depositional sequence that correlates to the upper Lilleba~lt Clay and S~vind Marl onshore (Michelsen et al. 1995). Central North Sea sequences of the middle Eocene (Fig. 5a) are tied to onshore sections with a combination of graphic correlation (Fig. 9), nannofossil stratigraphy from the PaleoServices biostratigraphic framework (Newton & Flanagan 1993; B. Braham, pers. comm.), positioning within the major R/TF cycle, key marker ties with the Haq et al. (1988) curve and any additional stratigraphic information published from the IGCP Project 124 (Vinken et al. 1988). Based on these sources of information, the Upper Tay sequence occurs above the Top Frigg terrace

PALEOGENE SEQUENCE STRATIGRAPHYIN NORTHWEST EUROPE and below the FDO of D. p a c h y d e r m u m , placing it near the NP14-15 boundary. It correlates with a sequence in the Whitecliff Bay section encompassing Fisher Beds VIII and IX (Fig. 9). In Belgium, the Lede Sands are deposited during the transgression of this sequence cycle. Regression in mid Eocene northwest Europe began after a peak transgression in the Brussels-Calcaire Grossi~re sequence and the Lede is slightly less transgressive than the Brussels (Vandenberghe et al. 1995). The type Lutetian sections in the Pads Basin shallow up to the Marnes et Callaisses, which have been divided into two sequences based on strontium chemo-stratigraphy (J. Riveline & M. Renard, pers. comm.). Some debate exists about this interpretation as it differs from that given by Haq et al. (1988, p. 89). G61y & Lorenz (1991) discuss the differences between these two possible interpretations, highlighting the differences in the Lower Lutetian and timing of peak transgression. Presently, the Brussels Sand and Calcaire Grossibre are placed within the same sequence and correlated to the TA3.2 ('48.5-46.5 Ma') sequence of Haq et al. (1988). This sequence is biostratigraphically represented by the Top Frigg terrace in the central North Sea. The Upper Tay and correlative formations are placed within the TA3.3 ('46.5-44 Ma') sequence of Haq et al. (1988). The Caran sequence graphically correlates with a sequence in Whitecliff Bay that contains Fisher Beds IX-XV (Fig. 9). The FDO of D. ficusoides and common C. ? amplectens occur near the NP15/16 boundary (B. Braham, pers. comm.), allowing a correlation of the Base Nauchlan terrace to the transgressive and highstand deposits of the Wemmel and Asse Formations in Belgium (Fig. 5a) that straddle the NP15-16 boundary (Vandenberghe et al. 1995). This sequence can be correlated to the TA3.4 ('4442.5 Ma') cycle of Haq et al. (1988). This cycle is correlated to the Marnes et Callaisses sup6rieurs as the hiatus above this formation is tied to the 42.5 Ma sequence boundary (Pomerol 1989). The Nauchlan sequence in the central North Sea occurs at the peak regression of the Paleogene 4 R/TF cycle. Its base can be tied to the '42.5 Ma' sequence boundary of Haq et al. (1988) based on its superposition relation with the Caran sequence. The top of this sequence is more difficult to place as the key marker for the base of the Brioc sequence, FDO of H. porosa, is reported on the biochronostratigraphy chart of Haq et al. (1988) at the base of the Priabonian. K6the (1990) reports this marker above the LAD of A. diktyoplokus, a widely recognized marker for the Eocene/ Oligocene limit (Vinken et al. 1988; Brinkhuis & Biffi 1993). Based on this very tenuous tie to the Haq curve, the base of the Brioc sequence is correlated to the '39.5 Ma' event on the Haq chart.

35

Between '39.5 and 42.5 Ma', a 40.5 Ma event appears on the Haq curve. Its expression has been noted in the Isle of Wight, Belgium and the Pads Basin, but cannot be reliably identified in the central North Sea at present. The Paleogene 4 R/TF cycle has peak regression with the Nauchlan sequence and is fully transgressive above the Brioc sequence. This observation is based on recognition of the Top Eocene graphic correlation terrace, which is reported as an unconformity by Harding et al. (1990) and Newton & Flanagan (1993). This gap in sedimentation is interpreted as a period of sediment starvation. In Belgium, the upper Eocene Bassevelde Sand transgresses over a major unconformity that erodes Paleocene deposits in northeastern Belgium (Vandenberghe et al. 1995). Nannofossils from the Bassevelde span from NP18-21 (Schuler et al. 1992), accounting for the entire Priabonian stage. Priabonian deposits in the Paris and LondonHampshire basins are non-marine or marginal marine (Cavalier 1988; King 1988) and are interpreted to indicate the peak of the mid and late Eocene regression (cycle 4 - Kockel 1988). This study interprets these deposits as the transgressive phase of the Paleogene 4 R/TF cycle, correlative with the Bassevelde Sand of Belgium and Top Eocene terrace in the central North Sea. Keen (1993) identifies four depositional sequences in the marginal marine upper Eocene of the London-Hampshire Basin based on ostracod stratigraphy within coarsening-upward cycles. Likewise, G61y & Lorenz (1991) identify cycles of transgression and regression within the lagoonal and evaporitic Ludien deposits in the Pads Basin. The thick succession of upper Eocene gypsum in the Paris Basin is attributed to increased subsidence by Cavalier & Pomerol (1979), which is consistent with observed transgressions from more marine settings in Belgium and the North Sea. The thick marginal marine depocentres record multiple depositional sequences that become condensed in the North Sea and are difficult to distinguish in Belgium. These depositional sequences are correlated with the Priabonian events on the Haq curve (Keen 1993; G61y & Lorenz 1991).

Paleogene 5 R/TF cycle (Lower Oligocene) During the early Oligocene, stratigraphic correlations become increasingly more difficult. The Paleogene 5 R/TF cycle in the central North Sea (Fig. 3) can only be correlated to Denmark. Southern North Sea estuary basins in Belgium, France and England record a major transgression (Vail & Hardenbol 1979; Kockel 1988). Japsen (1993) discusses uplift of the Skagerrak-Kattegat

36

J.E. NEAL

Platform, located at the southern end of the Baltic Shield, which occurs from the Oligocene into the Neogene. Michelsen et al. (1995) note major regressions during the Oligocene that may result from this uplift, with possible subaerial exposure of the onshore succession above the uppermost Eocene Viborg Formation (Heilmann-Clausen, pers. comm.). The central North Sea records a thick succession (up to 1000 m) of distal shelf mudstones called the Lark Formation (Knox & Holloway 1992). Two depositional sequences are identified on seismic data and gamma-ray well-logs. The Lower Lark 1 sequence occurs above the Top Eocene terrace, which contains the FDO of A. diktyoplokus. The top of the Lower Lark 2 sequence is marked by the FDO of A r e o s p h a e r i d i u m arcuatum. Between these two biostratigraphic markers, the FDO of Svalbardella cooksoniae appears. The biochronostratigraphic chart of Haq et al. (1988) locates the LAD of A. arcuatum just below the major mid-Oligocene sea level fall at '30' Ma. The inter-regional range chart of dinoflagellates from IGCP Project 124 places the LAD of A. arcuaturn within the Middle Eocene (Dll) and the LAD of S. cooksoniae within the Early Oligocene (Vinken et al. 1988). The positioning ofA. arcuatum by Haq et al. (1988) is consistent with observations in the central North Sea, as is the position of S. cooksoniae by IGCP Project 124 (Vinken et al. 1988). The two eustatic sea level cycles on the Haq curve are tentatively correlated to the Lower Lark 1 and 2 depositional sequences (Fig. 5a). These depositional sequences are also correlated with Sequences 4.1 and 4.2 in the Danish North Sea (Michelsen et al. 1995) based on dinoflagellate zones and an indirect nannofossil correlation through the biochronostratigraphic chart of Haq et al. (1988). In Belgium, a complex transgression is represented by Tongeren then Rupelian sediments (Steurbaut 1992). The sequence stratigraphy of the Rupelian and the Eocene-Oligocene transition is discussed by Vandenberghe et al. (1995). They recognize two complete depositional sequences through this interval, correlated to the two early Oligocene eustatic events of Haq et al. (1988). The Paris Basin has a more complex stratigraphic interpretation. Pomerol (1989) recognizes four hiatuses within the Stampian transgressive deposits, two of which (HP11 and HP12) are correlated with the eustatic events of Haq et al. (1987). One of the two remaining hiatuses in the Upper Stampian (HP14 or HPI5) could correlate to Sequence 4.3 in the Danish North Sea based on a nannofossil correlation (Michelsen et al. 1995). The Lower Oligocene of the London-Hampshire Basin has not been studied for depositional sequences.

Discussion and conclusions The IGCP Project 124 published the synthesized stratigraphic results of many researchers working in all parts of northwest Europe and the North Sea. The correlations presented by the IGCP Project 124 link local stratigraphic frameworks by biostratigraphic and lithostratigraphic means to create a northwest European stratigraphy. Only one figure (Kockel 1988) was devoted to erecting an 'assumed' relative sea-level curve responsible for creating the wonderful stratigraphic framework IGCP Project 124 had just deciphered. The results of IGCP Project 124 formed a foundation that could be examined in light of sequence stratigraphic first principles and a stratigraphic framework constructed with well-logs, biostratigraphy and seismic data from the central North Sea. Figure 11 compares the present sequence stratigraphic interpretation of northwest European Paleogene deposition with Kockel's (1988) interpretation of assumed relative sea-level changes in northwest Europe as part of IGCP Project 124. Two fundamental theoretical differences exist between these two curves.

The first general difference concerns resolution and the recognition of long and short term sea-level signals. This study recognizes 30 sequence cycles and five R/TF cycles, while Kockel has eight relative sea level cycles. Certain cycles correlate between frameworks (e.g. the Thanetian transgression, Brussels transgression), other intervals are more difficult to link as the high and low frequency signals become mixed (e.g. between the Balder and Thanet, lower Oligocene). This study combines the sequence stratigraphy of depocentres around northwest Europe to produce a sequence cycle curve for all of northwest Europe, not surprisingly containing more events than were correlated by Kockel (1988). He identified cycles thought to be recognizable throughout northwest Europe, whereas the present study wanted to identify all biostratigraphically resolvable sea-level events, whether they were present, eroded or condensed in other sub-basins. The second major difference is the interpretation of non-marine deposits in terms of relative changes in sea level. As stated above in the discussion of the upper Eocene, thick non-marine deposits are taken as indicators of transgression and increased accommodation space for sediment (Posamentier & Vail 1988; Duval et al. 1992). Kockel interprets these deposits as peak regressions. This philosophical difference leads to sea-level curves that are out of phase. For example, the mid and late Eocene Aalterbrugge Formation was interpreted by Kockel as representing a peak regression, whereas it is here interpreted as representing an early transgressive

37

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

IGCP # 124 (Kockel. 1988) =

Relative SeaLevel changes ~ .~ in the NW e~~ European Z Tertiary Basins

i ~N

23 ~

~

Reference formations for transgressive and regressive pulses

~_

~

19

Ji .

-~ N'W Europe Depositional ~ Sequence Cycles ~e . Relative Sea L e v e l . 9l ' l q ~ K 9

- - K e r k o m I O. Biezen Latdorf / Grimmertingen

Reference formations for Refer ~gressive/Regressive Facies Facies Cycles of the Central North Sea

"Jp~l~

]

M. and v.e.L.

' Rupel

/ ~ ~

22 21

This Study Major Central North Sea

i

Lower Lark

'- i

5. Cycle t ~. Oligo -cene ,

i

-- Pomerze ~ U .

Barton / L. Headot

~~

18

~ N

~ 16-15 / 14 13 ' ~ " ;~ 12

-Horda

Alba Brussel / Calcalregrmsier '

Aalterbrugge i

- - Frigg

11 10 9 8

.l 6

9

~

Balder/London A Woolwich

~

-- -- -- Balder Balr

,Sele / Erquelines Thanet t Landen / Lista

~ i *~

- Sele

~---Forties

,, ~ t~

Gelinde / Lellinge

--

C

~

Balmoral Tuffite

-- Andrew -Lista

[

~H~n

!

~'~---<~ohlingen (~assenheide

i

~- Manreen i

3 m

2

DanI Ekoflsk

.~.

Ekofisk i

Fig. 11. Summary figure comparing the relative sea-level curves of this study and IGCP Project 124 (Kockel 1988). Note differences in resolution of events and phase of curves related to differences in the sequence stratigraphic interpretation of thick non-marine deposits (see text for further discussion).

phase. Evidence in support of this position is that thick non-marine deposits onshore correlate to sediment starvation in the central North Sea, contrary to what would be expected in a peak regression. Peak regressions are represented by unconformities onshore and by submarine fan deposits in the North Sea. Detailed stratigraphic research in the future will certainly recognize additional sequence cycles, but the synthesis presented here places currently available stratigraphic data into terms of relative changes in sea-level as recognized in the central

North Sea subsurface and onshore northwest Europe sections. This study recognizes a hierarchy of sea-level cycles with a low frequency signal controlling the characteristics of higher frequency events. Graphic correlation makes possible a tie of marine hiatuses and submarine fan depositional pulses from the North Sea to shallow marine deposits of continental Europe. This interregional perspective allows a correlation of detailed subbasin frameworks and resolves the underlying relative sea level history of the Paleogene North Sea.

38

J.E. NEAL

The author would like to thank Exxon Production Research Co. for permission to publish this manuscript. Many thanks are due to J. Stein and J. Gamber of Amoco Production Co. for their help in the graphic correlation of North Sea wells and outcrop sections, and their great contribution to this work. The author wishes to thank: P. Vail, J. Hardenbol, R. W. O'B. Knox, N. Vandenberghe, C. Heilmann-Clausen, E. Thomsen, B. Braham, C. King,

J. Riveline, M. Renard, C. Pomerol and E Copestake for sharing their expertise on Paleogene stratigraphy in NW Europe. This research was supported by Rice University, Amoco Production Co., Amoco (UK) Exploration Co. and Ecole des Mines de Paris, where many thanks are due to E Ch. de Graciansky. Helpful reviews by R. W. O'B. Knox, M. Farley and anonymous reviewers significantly improved the final manuscript.

References ALl, J. R. 1994. Magnetostratigraphy of the upper Paleocene and lowermost Eocene of SE England. GFF, 116, 41-42. , KING, C. & HAILWOOD, E.A. 1993. Magnetostratigraphic calibration of the Early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD, E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. ARMENTROUT,J. M, MALACEK,S .J., FEARN, L. B. ET AL. 1993. Log-motif analysis of Paleogene depositional systems tracts, central and northern North Sea: defined by sequence stratigraphic analysis. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 45-58. ARMSTRONG,L. A., TEN HAVE,A. & JOHNSON,H. D. 1987. The geology of the Gannet Fields, central North Sea, UK sector. In: BROOKS,J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 533-548. AUBRY, M.-E 1985. Northwestern European Paleogene magnetostratigraphy: calcareous nannofossil evidence. Geology, 13, 198-202. 1986. Paleogene calcareous nannoplankton biostratigraphy of Northwestern Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 267-334. 1991. Sequence stratigraphy: eustasy or tectonic imprint? Journal of Geophysical Research, 96 (B4), 6641-6679. 1994. The Thanetian Stage in NW Europe and its significance in terms of global events. GFF, 116, 43-44. 9 HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous-nannofossil stratigraphy of the Lower Palaeogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735. --, BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD, K. D., OBRADOVICH,J. D. t~ PROTHERO, D. R., 1988. Paleogene geochronology: an integrated approach. Paleoceanography, 3 (6), 707-742. BERGGREN, W. A. 1994. In defense of the Selandian Age/Stage. GFF, 116, 44 A6. --, KENT, D. V., SWISHER, C. C. I I I & AUBRY, M.-E 1995. A revised Paleogene geochronology and chronostratigraphy. In: BERGGREN, W. A, KENT, D. V., AUBRY, M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical 55,

-

-

Geology, Society of Economic Geologists and Paleontologists, Special Publication, 54, Tulsa. BIGNOT, G. 1993. The position of the Montian Stage and related facies within the stratigraphic-palaeogeographic framework of NW Europe during the Danian. Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie, 29 (3-4), 47-59. BRINKHUIS, H. & BIFFt, U. 1993. Dinoflagellate cyst stratigraphy of the Eocene/Oligocene transition in central Italy. Marine Micropaleontology, 22, 131-183. CANDE, S. C. 8s KENT, D. V. 1992. A new geomagnetic polarity time scale for the late Cretaceous and Cenozoic. Journal of Geophysical Research, 97 (B10), 13 917-13 951. CAVALIER9C. 1988. Lithostratigraphy, lithology and environment of the sediments of the Paris Basin. In: VINKEN, R., VON DANIELS, C. H., GRAMANN,E, ETAL. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, fig. 8. t~ POMEROL,C. 1979. Chronologie et interpr6tation des 6v~nements tectoniques c6nozo'/ques dans le Bassin de Paris. Bulletin de la Socidtg Ggologique de France, (7), 21, (1), 33-48. & 1986. Stratigraphy of the Paleogene. Bulletin de la Socidte G~ologique de France, (8), t. 11, (2), 255-265. DE BATIST, M. 1989. Seismistratigrafie en Struktuur van her Paleogeen in de Zuidelijke Noordzee. Doctoraatsthesis, Rijks Universitet Gent. DE CONINCK, J. 1990. Ypresian organic-walled phytoplankton in the Belgian Basin and adjacent areas. Bulletin de la Socidt~ Belge de Gdologie, 97 (3/4) (1988), 287-319. DEEGAN, C. E. • SCULL, B. J. 1977. A Standard Lithostratigraphic Nomenclature for the Central and Northern North Sea. Institute of Geological Sciences Report 77/25. DEN HARTOGJAGER, D., GILES, M. R. & GRIFF1THS,G. R. 1993. Evolution of Paleogene submarine fans of the North Sea in space and time. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 59-72. DUVAL, B., CRAMEZ,C. & VAIL, P. 1992. Types and hierarchy of stratigraphic cycles (abstract). In: Abstract Volume, Sequence Stratigraphy of European Basins Meeting, Dijon, France, May 18-20, 1992, CNRS-IFP. EATON, G. 1976. Dinoflagellate cysts from the Bracklesham Beds (Eocene) of the Isle of Wight, -

-

-

-

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE Southern England. Bulletin of the British Museum of Natural History (Geology), 26, 225-332. ELLISON, R. W., KNOX, R. W. O'B., JOLLEY, O. W. & IZdNG, C. 1994. A revision of the lithostratigraphic classification of the early Palaeogene strata of the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. FISHER, O. 1862. On the Bracklesham Beds of the Isle of Wight Basin. Quarterly Journal of the Geological Society of London, 18, 65-93. GALLOWAY,W. E., GARBE,J. L., XIJIN,LIU & SLOAN,B. J. 1993. Sequence stratigraphic and depositional framework of the Cenozoic fill, central and northern North Sea Basin. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 33-44. GELY, J.-P. & LORENZ, C. 1991. Analyse s6quentielle de l'Eoc~ne et de l'Oligoc~ne du bassin Parisien (France). Revue de l'Institut Franfais du P(trole, 46 (6), 713-747. GREEN, P. E 1989. Thermal and tectonic history of the East Midlands shelf (onshore U.K.) and surrounding regions assessed by fission tract analysis. Journal of the Geological Society, London, 146, 755-773. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, / 156-1167. , HARDENBOL,J. & VAIL, P. R. 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of relative sea level change. In: WILGUS, C. K., HASTINGS, B. S., KENDALL,G .C. ST. C., POSAMEYnER,H. W., Ross, C. A. & VAN WAGONER,J. C. (eds) Sea-level Changes: An Integrated Approach. SEPM, Special Publication, 42, 71-108. HARDENBOL, J. & BERGGREN, W. A. 1979. A new Paleogene numerical time scale. Studies in Geology, 6, 213-234. HARDING, A. W., HUMPHREY, T. J., LATHAM, A., LUNDSFORD, M. K. & STRIDER, M. H. 1990. Controls on Eocene Submarine Fan Deposition in the Witch Ground Graben. In: HARDMAN,R. F. P. & BROOKS, J. (eds) Tectonic Events' Responsible for Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 353367. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the Uppermost Danian to Ypresian in the Viborg I borehole, Central Jylland, Denmark. Danmarks Geologiske UndersCgelse, Series A, 7. 1994. Review of Paleocene dinoflagellates from the North Sea region. GFF, 116, 51-53. & COSTA, L. 1989. Dinoflagellate zonation of the Uppermost Paleocene? to Lower Miocene in the Wursterheide Research Well, NW Germany. Geologisches Jarbuch, A 111, 431-521. - - , NIELSEN,O. • GERSNER, [7. 1985. Lithostratigraphy and depositional environments in the Upper Paleocene and Eocene of Denmark. Bulletin of the Geological Society of Denmark, 33, 287-323. HERITIER, E E., LOSSEL, P. & WATHE, E. 1980. Frigg Field: Large submarine-fan trap in Lower Eocene rocks of the Viking Graben, North Sea. In: HALIBOUTY, M. (ed.) Giant Oil and Gas Fields of -

-

-

-

39

the Decade 1968-1978. American Association of Petroleum Geologists, Memoir 30, 59-80. HOOYBERGHS, H. 1983. Contribution to the study of planktonic foraminifera in the Belgian Tertiary. Aarkundige Mededelingen, 2, Leuven University Press. 1988. The distribution of planktonic foraminifera in sediments of the Kempen-northern Limburg area (Belgium). In: VINKEN, R., VON DANIELS, C. H., GRAMANN, F. Er AL. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, fig. 94. HOUTHUYS, R. & GULLEYrOPS, E 1988. The Vlierzele Sands (Eocene, Belgium): a tidal ridge system. In: DE BOER, P. L., VAN GELDER, A. & NIO, S. D. (eds) Tide-Influenced Sedimentary Environments and Facies. Reidel, 139-152. JAPSEN,P. 1993. Influence of lithology and Neogene uplift on seismic velocities in Denmark: implications for depth conversion of maps. American Association of Petroleum Geologists Bulletin, 77 (2), 194211. JOHNSON, H. 1987. Seismic expression of major chalk reworking events in the Palaeocene of the Central North Sea. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 591-598. JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: the Palaeogene stratigraphy and relative sea level history of the Outer Moray Firth, UK North Sea. Marine and Petroleum Geology, 11, 157-164. KEEN, M. 1993. Ostracods and sequence stratigraphy (abs.). In: Programme with Abstracts from the British Micropalaeontological Society Spring Meeting on Biostratigraphy in Sequence Stratigraphy at the University of Aberdeen, 22-24 March. KENNEDY, W. J. 1987. Sedimentology of Late CretaceousPalaeocene Chalk reservoirs, North Sea Central Graben. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 469-481. KaNG, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6. 1988. Lithostratigraphy, lithology and environment of the sediments of the Hampshire Basin (Great Britain). In: VINKEN, R., VON DANIELS, C. H., GRAMANN, E, ET AL. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, fig. 7. 1990. Eocene stratigraphy of the Knokke Borehole (Belgium). In: LAGA, P. & VANDENBERGHE,N. (eds) The Knokke Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M6moires pour servir ~t rExplication des Cartes G6ologiques et Mini~res de la Belgique, 29, 67-102. -1991. Stratigraphy of the London Clay Formation (Early Eocene) in the Hampshire Basin. PhD Thesis, Kingston Polytechnic. KNOX, R. W. O'B. 1990. Thanetian and early Ypresian -

-

-

-

-

-

40

J.E. NEAL

chronostratigraphy in south-east England. Tertiary Research, 11 (2-4), 57-64. 1994a. The age and regional context of the Thanetian stratotype sections of SE England. GFF, 116, 55-56. 1994b. From regional stage to standard stage: implications for the historical Paleogene stratotypes of NW Europe. GFF, 116, 56--57. -& HARLAND, R. 1979. Stratigraphic relationships of the early Palaeogene ash-series of NW Europe. Journal of the geological Society of London, 136, 463-479. & HOLLOWAY,S. 1992. 1. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & Cordey, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. -& MORTON, A. C. 1988. The record of Tertiary N Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. --, HINE, N. & ALl, J. 1994. New information on the age and sequence stratigraphy of the type Thanetian of Southeast England. Newsletters on Stratigraphy, 30 (1), 45-60. KOCKEL, E 1988. Assumed basin-wide relative sea-level changes in the NW European Tertiary Basin. In: VINKEN, R., VON DANIELS, C. H., GRAMANN, E ETAL. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, fig. 35a. KOTHE, A. 1990. Paleogene dinoflagellates of northwest Germany biostratigraphy and paleoenvironment. Geologisches Jarbuch, A 118, 7-508. LAGA, P. 1988. The distribution of benthic foraminifera in sediments of the Kempen-northern Limburg area (Belgium). In: VINKEN, R., VON DANIELS, C. H., GRAMANN, F. er AL. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, fig.91. MANCINI, E. & TEW, B. 1991. Relationships of the Paleogene stage and planktonic foraminiferal zone boundaries to lithostratigraphic and allostratigraphic contracts in the eastern Gulf Coastal Plain. Journal of Foraminiferal Research, 7 (1), 48-66. MICHELSEN, O., DANIELSEN,M., HEILMANN-CLAUSEN,C., JORDT, H., LAURSEN, G. & THOMSEN, E. 1995. Occurrence of major sequence stratigraphic boundaries in relation to basin development in the Cenozoic deposits of the southeastern North Sea. Sequence Stratigraphy of North West Europe, Norwegian Petroleum Society, Stavanger, in press. MILLER, E X. 1977. The graphic correlation method in biostratigraphy. In: KAUFFMAN,E. G. & HAZEL,J. E. (eds) Concepts and Methods in Biostratigraphy. Dowden, Hutchinson & Ross, Stroudsburg, Pennsylvania, 165-186. MILTON, N. J., BERTRAM,G. T. & VANN,I. R. 1990. Early Palaeogene tectonics and sedimentation in the Central North Sea. In: HARDMAN, R. F. P. & BROOKS, J. (eds) Tectonic Events Responsible for -

-

-

-

Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. MITCHUM, R. M. JR & VAN WAGONER, J. C. 1991. Highfrequency sequences and their stacking patterns: sequence stratigraphic evidence of high-frequency eustatic cycles. Sedimentary Geology, 70, 131-160. --, VAIL, P. R. & THOMPSON, S. III 1977. Seismic stratigraphy and global changes of sea level, part 2, the depositional sequence as a basic unit for stratigraphic analysis. In: PAYTON, C. E. (ed.) Seismic Stratigraphy -Applications to Hydrocarbon Exploration. American Association of Petroleum Geologists, Special Publication, 26, 53-62. MORTON,A. C., HALLSWORTH,C. R. & WILKINSON,G. C. 1993. Stratigraphic evolution of sand provenance during Paleocene deposition in the North Sea area. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 73-84. MUDGE, D. • COPESTAKE, P. 1992a. Revised Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. & -1992b. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. NEAL, J. E. 1994. A stratigraphic analysis of Paleogene deposition in Northwest Europe and the role of graphic correlation in sequence stratigraphy. PhD Thesis, Rice University, Houston. --, STEIN, J. A. & GAMBER, J. H. 1994. Graphic Correlation and Sequence Stratigraphy in the Paleogene of NW Europe. Journal of Micropalaeontology, 13 (1), 55-80. , & -1995. Nested stratigraphic cycles and depositional facies of the central North Sea. In: DE GRACIANSKY,P. Ch., HARDENBOL,J., VAIL, P. R. & JACQUIN, T. (eds) Mesozoic-Cenozoic Sequence Stratigraphy of NW Europe. Society of Economic Geologists and Paleontologists, Tulsa, in press. NEWTON, S. K. & FLANAGAN,K. P. 1993. The Alba Field: evolution of a depositional model. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 161-173. NIELSEN, O., BAUMANN, J., DEYU, Z. & HEILMANNCLAUSEN, C. 1986. Tertiary deposits in the Store B~elt. The Tertiary section of borehole D.G.I. 83101, Osterrenden, Store B~elt, Denmark. Geoskrifter, 24, 235-253. O'CONNOR, S. J. & WALKER, D. 1993. Paleocene reservoirs of the Everest Trend. In: Parker, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 145-160. PARKER, J. R. 1975. Lower Tertiary sand development in the Central North Sea. In: WOODLAND,m. W. (ed.) Petroleum Geology and the Continental Shelf of Northwest Europe. Applied Science, Barking, 447-453. (ed.) 1993. Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society; London. PLtNT, A. G. 1983. Facies, environments and sedimentary -

-

PALEOGENE SEQUENCE STRATIGRAPHY IN NORTHWEST EUROPE

-

-

cycles in the Middle Eocene, Bracklesham Formation of the Hampshire Basin: evidence for global sea-level change? Sedimentology, 30, 625-653. 1988. Global eustasy and the Eocene sequence in the Hampshire Basin, England. Basin Research, 1, 1 1 - 2 2 .

POMERO1, C. 1989. Stratigraphy of the Palaeogene: hiatuses and transitions. Proceedings of the Geologists' Association, 100 (3), 313-324. POSAMENTIER,H. W. & VAIL,P. R. 1988. Eustatic controls on clastic deposition II -Sequence and systems tract models. In: WILGUS, C. K., HASTINGS, B. S., KENDALL,G. C. ST. C., POSAMENTIER,H. W., ROSS, C. A. & VAN WAGONER, J. C. (eds) Sea-level Changes: An Integrated Approach. SEPM, Tulsa, Special Publication, 42, 125-154. , ALLEN, G. & JAMES, D. P. 1992. High-resolution sequence stratigraphy - the East Coulee delta. Journal of Sedimentary Petrology, 62, 310-317. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Palaeocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327-344. 1992. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. Chapman & Hall, London, 155-251. , BRINKHUIS, H. & BUJAK, J. P. 1996. Upper Paleocene to Lower Eocene dinoflagellate cyst sequence biostratigraphy of southeast England. This volume. RAPPAPORT, R. 1969. The geological atlas of Guettard, Lavoisier, and Monnet; conflicting views of the nature of geology. In: SCHNEER,C. J. (ed.) Towards a History of Geology. M.I.T. Press, Cambridge, Massachusetts, 272-287. ROSENKRANTZ,A. 1924. De kObenhavnske GrCnsandslag og deres Placering i den danske Lagra~kke. Med et Skema over det danske Paleoc~en. Meddelelser fra Dansk Geologisk Forening, 6, 1-39. SCHMITZ, B. 1994. The Paleocene Epoch - stratigraphy, global change and events. GFF, 116, 39-41. SCHRODER, T. 1992. A palynological zonation for the Paleocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-126. SCHULER, M., CAVALIER,C., DUPUIS, C., STEURBAUT,E. & VANDENBERGHE,N. 1992. The Paleogene of the Paris and Belgian basins. Standard-stages and regional stratotypes. In: 8th International Palynological Congress, Aix-en-Provence, 13-16th September 1992, Excursion C. Cahiers de Micropa16ontologie N.S., 7 (1/2), 29-92. SHAW, A. B. 1964. Time in Stratigraphy. McGraw-Hill, New York. SINHA, A. & STOTT, L. D. 1993. Recognition of the Paleocene/Eocene boundary carbon isotope excursion in the Paris Basin, France (abs.). In: Symposium on the Correlation of the Early Paleogene in Northwestern Europe, 1-2 December 1993. Geological Society, London. STEIN,J. A., GAMBER,J. H., KREBS,W. N. & LA COL, M. 1995. A Composite Standard Approach to -

-

41

Biostratigraphic Evaluation of the North Sea Paleogene. Sequence Stratigraphy of North West Europe. Norwegian Petroleum Society, Stavanger, in press. STEURBAUT, E. 1988. New Early and Middle Eocene calcareous-nannoplankton events and correlations in the middle to high latitudes of the northern hemisphere. Newsletters on Stratigraphy, 18 (2), 99-115. 1992. Integrated stratigraphic analysis of Lower Rupelian (Oligocene) in the Belgian basin. Annales de la Soci6t6 G6olgique de Belgique, 115 (1), 99-115. & NOLF, D. 1986. Revision of Ypresian Stratigraphy of Belgium and Northwestern France. Mededelingen van de Werkgroeop voor Tertiaire Geologie, 23 (4), 115-172. STEWART,I. J. 1987. A revised stratigraphic interpretation of the early Palaeogene of the central North Sea. In: BROOKS, J. & GLENNIE,K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 557-576. STOTT, L. D., SINHA, A., TItlRY, M., AUBRY, M.-R & BERGGREN, W. A. 1996. Global 813C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine con'elations. This volume. THOMAS, E. & SHACKLETON, N. J. 1996. The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies. This volume. THOMSEN, E. 1990. Kalkaflejringer fra Kridt og Danian. In: NmLSEN,O. B. & SANDERSEN,R (eds) Danmarks Geologi - f r a Ovre Kridt til i Dag, Bind 1: Ovre Kridt/terticer og Geofysiske Metoder. Geologisk Institut, Aarhus, 1-42. i992. Sequence stratigraphy of the Danian (Lower Paleocene) limestone deposits in Denmark (abs). In: Abstract volume from the Sequence Stratigraphy of European Basins Meeting in Dijon, France, May 18-20, 1992. CNRS-IFP. & HEILMANN-CLAUSEN, C. 1985. The DanianSelandian boundary at Svejstrup with remarks on the biostratigraphy of the boundary in western Denmark. Bulletin of the Geological Society of Denmark, 33, 341-362. TOWNSEND, H. A. & HAILWOOO, E. A. 1985. Magnetostratigraphic correlation of Paleogene sediments in the Hampshire and London Basins, southern UK. Journal of the Geological Society, London, 142, 957-982. VAin, R R. & HARDENBOL, J. 1979. Sea-level changes during the Tertiary. Oceanus, 22, 71-79. - - - , AUDEMARD,E, BOWMAN,S., EISNER, P. & PEREZCRUZ, G. 1991. The stratigraphic signatures of tectonics, eustasy, and sedimentation. In: SEILACHER,D. & EISNER,G. (eds) Cycles and Events in Stratigraphy. Springer-Verlag, Bedim 617-659. VANDENBERGHE, N., LAGA, P., STEURBAUT, E., HARDENBOL, J. & VAIL, R R. 1995. Sequence stratigraphy of the Tertiary at the southern border of the North Sea Basin in Belgium. In: HARDENBOL, J., VAIL, R, DE GRACIANSKY,R C. & JACQUIN, T. (eds) Sequence Stratigraphy of Mesozoic and Cenozoic European Basins. CNRS-IFP, Paris, in press. -

-

42

J.E. NEAL , VANDORMAEL,C. & ELEWAUT,E. 1988. The

geophysical log correlations in the leper Clay sections in Belgium. In: Dupuls, C., DE CONINCK,J. & STEURBAUT, E. (eds) The Ypresian StratoOpe. Bulletin de la Socidt6 Beige de G6ologie, 97 (3-4), 437-440. V1NKEN, R., VON DANIELS, C. H., GRAMANN, E ET AL. (eds) 1988. The Northwest European Tertia;y Basin.

Results of the IGCP Project No. 124. Geologisches Jarbuch, A 100, 7-508. WHEELER, H. E. 1958. Time-Stratigraphy. American Association of Petroleum Geologists Bulletin, 42, 1047-1063. ZIEGLER, P. A. 1990. Geological Atlas of Western and Central Europe, 2nd Edition. Shell Internationale Petroleum Maatschappij B.V., The Hague.

Forward and reverse stratigraphic modelling of CretaceousTertiary post-rift subsidence and Paleogene uplift in the Outer Moray Firth Basin, central North Sea R A. N A D I N & N. J. K U S Z N I R Department o f Earth Sciences, University o f Liverpool, PO Box 147, Liverpool L69 3BX, UK Abstract: The Paleogene stratigraphy in the Outer Moray Firth is part of the North Sea post-rift sedimentary sequence that was deposited in a thermally subsiding basin following Late Jurassic rifting. Previous studies have drawn attention to anomalous uplift and departure from the McKenzie (1978, Earth Planetary Science Letters, 40, 25-32) post-rift subsidence trend in the early Paleocene followed by an accelerated phase of basin subsidence in the early Eocene before returning to the normal post-rift subsidence. 2D forward and reverse stratigraphic modelling incorporating first-order global sea-level variations have been used to determine the timing and magnitude of departures from McKenzie post-rift thermal subsidence in the early Paleogene for the Outer Moray Firth. Reverse post-rift modelling, consisting of flexural backstripping, decompaction and reverse thermal subsidence, has been used to produce restored post-rift sections from present-day stratigraphy, which have been constrained by palaeobathymetric data. Forward syn-rift and post-rift structural and stratigraphic modelling uses the flexural cantilever model of rift basin formation and has been constrained by syn-rift structural data, present-day stratigraphy and palaeobathymetric markers. Forward and reverse modelling show that observed Cretaceous and Tertiary stratigraphy of the Outer Moray Firth was generated by the combined effects of inherited Late Jurassic syn-rift accommodation space, post-Jurassic rift thermal subsidence, sediment supply, long-term eustasy and an additional transient uplift event in the Paleocene. The much thicker Tertiary compared with Cretaceous can be explained by sediment infilling of starved Cretaceous palaeobathymetry. Reverse and forward post-rift modelling predict regional Paleocene uplift in the Outer Moray Firth of the order of 375-390 m (375 m reverse modelling, 390 m forward modelling), followed by c. 160 m of rapid Eocene subsidence, both superimposed on post-rift thermal subsidence following Late Jurassic rifting. Regional Paleocene uplift in the Outer Moray Firth is attributed to dynamic uplift associated with the development of the Iceland plume.

The area of this study covers the Outer Moray Firth Basin (Fig. la), an intracontinental rift basin which forms part of the North Sea trilete rift system. The Outer Moray Firth Basin has experienced two phases of Mesozoic rifting, the first in the Triassic and the second in the Late Jurassic. N-S orientated Late Jurassic extension ended in Portlandian to mid-Ryazanian times (Rattey & Hayward 1993), and post-rift thermal subsidence commenced with the development of a thick and conformable Cretaceous and Tertiary succession. Well and depth-converted seismic data show that the Cretaceous succession in the Outer Moray Firth Basin axis is up to 2 km thick and conformably overlain by up to 2.5 km of Tertiary, of which c. 1 km are early Paleogene sediments (Andrews et al. 1990). The larger thickness of the Tertiary succession compared to the thinner Cretaceous is not as expected from the application of the McKenzie (1978) post-rift subsidence model, which predicts more early Cretaceous post-rift subsidence (Fig. lb). This much thicker Tertiary

can largely be explained by infilling of earlier starved Cretaceous accommodation space (Bertram & Milton 1989; Barr 1991). However, previous workers in the Outer Moray Firth Basin (Milton et al. 1990; Jones & Milton 1994) have demonstrated that uplift and departure from the first-order M c K e n z i e (1978) post-rift subsidence trend occurred during the early Paleogene (Fig. lb). Regional Paleogene uplift is consistent with the northwestwards uptilting and rejuvenation of the clastic source areas (the Orkney-Shetland Platform and the Scottish Highlands), which resulted in a major clastic influx into a basin that had previously been accumulating deep-water carbonates. Sediment supply from the west was large enough to produce rapid shelf progradation (Fig. l a) late in the Paleocene (Knox et al. 1981; Rochow 1981; Stewart 1987; Andrews et al. 1990; Milton et al. 1990; Mudge & Copestake 1992; Den Hartog Jager et al. 1993). The sediment input and deposition of the Paleocene succession ended with a regional transgression and flooding of the Shetland Platform

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 43-62.

43

44

P. A. NADIN & N. J. KUSZNIR

Fig. 1. (a) A location map of the Outer Moray Firth Basin showing the distribution of early Paleogene sediments (adapted from Andrews et al. 1990). (b) A schematic Cretaceous and Tertiary subsidence history curve for the Outer Moray Firth Basin, in which Paleogene uplift is superimposed on the main exponential McKenzie (1978) post-rift trend (adapted from Bertram & Milton 1989).

in the early Eocene (Jones & Milton 1994). This regional transgression coincides with an accelerated phase of basin subsidence (Bertram & Milton 1989; Milton et al. 1990; Jones & Milton 1994), before returning to a trend of normal postrift subsidence (Fig. lb). Total Cretaceous and Tertiary post-rift subsidence, shown schematically in Fig. lb, can be accounted for by Triassic and Jurassic rifting (Giltner 1987; Marsden e t al. 1990). In this study, 2D reverse post-rift stratigraphic modelling (consisting of flexural backstripping, decompaction and reverse thermal subsidence modelling), and 2D forward structural and stratigraphic basin modelling have been used to analyse Cretaceous and Tertiary post-rift subsidence in the Outer Moray Firth Basin. In particular, stratigraphic basin modelling has been used to investigate early Paleogene departures from the first-order McKenzie (1978) post-rift thermal subsidence curve.

1D Airy backstripping studies In order to determine the post-rift subsidence history of the North Sea Basin, many workers (e.g. Sclater & Christie 1980; Joy 1992; White & Latin 1993; Hall & White 1994) have used ID Airy backstripping according to the method of Sclater & Christie (1980). Airy backstripping studies have been used to suggest a rapid increase in basement

subsidence during the early Paleogene (e.g. Joy 1992; White & Latin 1993; Hall & White 1994), rather than Paleocene uplift as suggested by Milton e t al. (1990) and Jones & Milton (1994). However, determination of post-rift subsidence history by Airy backstripping has several important sources of error which may lead to unreliable and misleading results. (1) The validity of previously published backstripping studies on the North Sea Basin depends fundamentally on the quality and quantity of palaeobathymetric data. Airy backstripping is critically dependent on precise and accurate water depth estimates through time. While Late Jurassic erosion surfaces and late Paleocene coals provide reliable estimates of palaeobathymetry, there is no good palaeobathymetric control through the Cretaceous, and the use of fossil assemblages to estimate palaeo-water depths may carry very large errors (Bertram & Milton 1989). Many Airy backstripping studies have assumed, without justification, that Cretaceous palaeobathymetry was negligible and that errors in Cretaceous palaeobathymetry were < 200 m (e.g White & Latin 1993; Hall & White 1994). Both of these assumptions are probably wrong, with errors in Cretaceous palaeobathymetry estimates being >> 200 m. In Airy backstripping, basement subsidence is estimated at a set of points in time (see Sclater & Christie 1980) by: (a) sequentially removing layers from top to bottom; (b) decompacting remaining

2D STRATIGRAPHICMODELLING IN THE OMF BASIN units; (c) using palaeobathymetric estimates to set the depth to the top of the remaining units; (d) subtracting the loading effects of remaining sediments using Airy isostasy. This procedure generates a curve whose form reflects the basement subsidence through time. Each estimate of basement subsidence at each point in time should carry errors arising from uncertainty in the palaeobathymetry and from the decompaction process. The consequence of using inaccurate estimates of palaeobathymetry in Airy backstripping is illustrated in Fig. 2. Basement subsidence history (Fig. 2a) is usually estimated by fitting a curve through the basement subsidence determinations (e.g. Sclater & Christie 1980). Figure 2a shows a typical subsidence history with an apparently complex history of basement subsidence and uplift. Furthermore,

(a)

~, ,'~, r

.'2.

Base subsidence .1_ determination with error bars

r./3

Time BP (Ma)

: ',T | ~

[ ,/ .L / / /

~,1 -J, ~"" ]" ~ _L"~

Apparent basement | subsidence history .L.

,--

"~ "~ = r.~

(d)

Time BP ( M a ) - -

while some palaeobathymetry estimates carry realistic estimates of bathymetry (solid error bars in Fig. 2a), many Airy backstripping studies use values in which the error in palaeobathymetry is underestimated (dashed error bars in Fig. 2a). It is important that errors in palaeobathymetry are not underestimated, otherwise a meaningless subsidence history is produced by the Airy backstripping process. An alternative approach to estimating basement subsidence is to use only palaeobathymetry estimates with very small errors (Fig. 2b), such as may be obtained from the presence of coals, erosion surfaces and carbonate reefs. These wellconstrained estimates of palaeobathymetry may then be used to calibrate the post-rift subsidence as predicted by the McKenzie (1978) model of

(b)

Time BP (Ma)

45

,

I

Calibrated McKenzie post-dft thermal subsidence

Time BP (Ma)

iii

i

Z i I

i

Departure from McKenzie post-rift thermal subsidence

Fig. 2. Schematic plots of basement subsidence obtained from Airy backstripping. Error bars for individual basement subsidence determinations are shown. Solid lines represent realistic estimates of errors, while dotted lines represent underestimates. (a) Fitting a trend curve (dashed line) through basement subsidence determinations gives an apparently complex history of basement subsidence and uplift. (b) An alternative approach is to use the most accurate palaeobathymetric determinations to calibrate the McKenzie post-rift thermal subsidence model (solid exponential curve) giving a [~-stretching factor. (e) Many of the basement subsidence determinations with realistic errors also fit with the calibrated McKenzie subsidence curve. (d) Those that do not may be used to identify departures (dotted line) from McKenzie post-rift thermal subsidence.

46

P. A. NADIN t~ N. J. KUSZNIR

extensional basin formation and subsidence; calibration of the McKenzie (1978) model taking the form of an estimate of the 13-stretching factor. This approach was used by Bertram & Milton (1989) and Barr (1991). Fundamental to this approach is the assumption that post-rift subsidence between times of reliable palaeobathymetric estimates is due to McKenzie post-rift lithosphere cooling. For some points in time estimates of basement subsidence based on accurate palaeobathymetry may conflict with the calibrated McKenzie post-rift subsidence trend (Fig. 2b). If this occurs, assumptions must be made as to which of the reliable basement subsidence points are due to McKenzie post-rift subsidence alone, and which are influenced by some additional uplift or subsidence process, or further rifting events. Assuming that McKenzie (1978) post-rift subsidence may be adequately calibrated, the basement subsidence estimates based on realistic palaeobathymetry errors should be compatible with the calibrated McKenzie post-rift subsidence curve (Fig. 2c). If palaeobathymetry estimates with realistic low errors do not lie on the calibrated McKenzie subsidence curve, then this implies anomalous subsidence (Fig. 2d). (2) Airy backstripping ignores the flexural strength of the lithosphere; the effective elastic thickness of the lithosphere, Te, for Airy isostasy is zero. The assumption that an Airy response can approximate a flexural response with low effective elastic thickness (Te < 5 km) is invalidated where isostatic loads have a short wavelength component. For example, the subsidence on fault-block highs where many wells are located is influenced by sediment loading in the deeper adjacent halfgrabens and vice versa. This is especially important in the Outer Moray Firth Basin where half-grabens are typically 10-30 km wide, i.e. the Witch Ground and Buchan half-grabens (Fig. 4a), and where the bathymetry at the end of rifting was of the order of 1000 m in the graben axis (Bertram & Milton 1989; Barr 1991). Backstripping stratigraphic wells without allowing for the effects of lateral loading (i.e. Airy isostasy) overestimates the post-rift load due to the cooling lithosphere, and hence overestimates the l-stretching factor (Roberts et al. 1993; Kusznir et al. 1995).

Reverse post-rift stratigraphic modelling: flexural backstripping, decompaction and reverse thermal subsidence modelling In order to overcome the limitations of Airy backstripping, reverse post-rift modelling (Fig. 3) consisting of flexural isostatic backstripping, decompaction and reverse thermal subsidence

modelling has been carried out. The technique has been applied to geological depth sections to produce a sequence of restored sections with time, whose predicted palaeobathymetry and/or emergence can be tested against the observations of well-constrained palaeobathymetric markers such as sea-level erosion surfaces and coals. A full description of the reverse modelling technique can be found in Roberts et al. (1993) and Kusznir et al. (1995). The starting point of the model is a depth-converted present-day crosssection. The procedure involves reverse modelling (i.e. backwards in time) the time-dependent postrift thermal subsidence (defined by 13in 2D) of the extensional sedimentary basin (Fig. 3). Lithosphere cooling, sediment loading and compaction are tracked backwards in time to the end of the syn-rift succession, i.e. the base of the post-rift. Thermal subsidence is defined by a 2D [3-stretching factor according to McKenzie (1978). Isostatic response to thermal, sediment and water-loads are calculated using flexural isostasy. 2D reverse post-rift modelling produces a series of stratigraphic restorations with time, showing predicted palaeobathymetry and/or emergence (Fig. 3). The history of palaeobathymetry and emergence are strongly controlled by the magnitude of 13 used to define the thermal subsidence component of the reverse model. The [3-factor may therefore be calibrated using reverse post-rift modelling constrained by palaeobathymetric estimates. 2D forward syn-rift and post-rift modelling, using the flexural cantilever model of continental lithosphere extension, has also been carried out to constrain estimates of [3-stretching factors and lithosphere flexural strength used in the reverse post-rift modelling.

Quantitative 2D analysis of post-rift subsidence in the Outer Moray Firth Basin using reverse post-rift modelling A regional geological cross-section based on seismic reflection and well data across the Outer Moray Firth Basin has been investigated using reverse post-rift modelling. The profile, shown in Fig. 4a, runs N-S across the Outer Moray Firth from the East Shetland Platform to the Peterhead Ridge and crosses the Halibut Horst, Witch Ground Graben and Buchan Graben which became prominent structures during Mesozoic rifting.

Reverse post-rift stratigraphic modelling (flexural bac kstrippin g ) The present-day depth cross-section (Fig. 4a) has been reverse modelled to the base of the post-rift

2D STRATIGRAPHIC MODELLING IN THE OMF BASIN

47

Fig. 3. A schematic illustration of 2D reverse post-rift modelling consisting of flexural backstripping, decompaction and reverse post-rift thermal subsidence modelling. Starting with a present-day depth section, sediment layers are successively removed down to the base of the post-rift succession. A sequence of restored sections is produced which can be tested against observed palaeobathymetric markers (adapted from Roberts et al. 1993).

succession, i.e. the end of Late Jurassic rifting at 157 Ma BP. First-order eustatic sea-level variations from Haq e t al. (1987) have been incorporated in the modelling (Fig. 40. An important test and constraint of the reverse modelling process is that stratigraphic surfaces once at sea level, i.e. eroded fault-block crests, wave cut platforms and coal bearing sequences, should be restored to sea level at the appropriate point in time. Trial estimates of a constant Late Jurassic [3-stretching factor were used to drive the reverse thermal subsidence model. In addition a 15 of 1.10 was used to describe post-rift thermal subsidence from the earlier Triassic rift event (at 250 Ma he) which also makes a minor contribution to Cretaceous and Tertiary post-rift thermal subsidence. A value of effective elastic thickness, Te = 4 km derived from forward

modelling, was used to describe the flexural strength of the lithosphere in the reverse model. A model using a constant 15= 1.12 for Late Jurassic rifting across the profile was found to restore end-of-rift erosion surfaces across the Shetland Platform margin and Halibut Horst back to or close to sea level at the base of the post-rift sequence (Fig. 4b). A lower value of [5 generated too much end-Jurassic bathymetry, while a greater value of J3 generated too much emergence. The reverse model with a 13 of 1.12 may also be used to interpolate McKenzie (1978) post-rift subsidence and produce restored cross-sections within the early Paleogene. The restored section at the end of the Paleocene (55 Ma) is shown in Fig. 4c. The top Paleocene horizon across this section contains a shallow water coal-bearing

Fig. 4. Reverse post-rift model of the Outer Moray Firth with eustatic sea-level variations but no Paleocene uplift. (a) Present-day depth section (original section from Andrews et al. 1990). (b) Restored cross-section to the endJurassic, modelled using a constant p-factor profile of 1.12. (c) Restored cross-section to the end-Paleocene, modelled using a constant ~-profile of 1.12. (d) Restored cross-section to the end-Jurassic, modelled using a variable [3-factor generated from the preferred forward syn-rift model (Fig. 7). (e) Restored cross-section to the end-Paleocene, modelled using the variable B-factor with ~av = 1.14. (f) First-order eustatic sea-level variations (Haq et al. 1987) used in the reverse model.

2D STRATIGRAPHICMODELLING IN THE OMF BASIN succession (Fig. la), corresponding to the Beauly Member of the Dornoch Formation (Knox & Holloway 1992). However, the restored Paleocene section produced by reverse modelling shows a substantial palaeobathymetry of c. 375 m (Fig. 4c). This restoration is therefore not consistent with the simple 2D McKenzie (1978) post-rift thermal subsidence curve. The discrepancy in water depth at the end of the Paleocene is interpreted as 375 m of additional regional uplift relative to the firstorder McKenzie (1978) post-rift subsidence curve. Increasing the Late Jurassic ~ to 4.0 correctly restored the end-Paleocene coal horizons to sea level, however a 13 of 4.0 (300% extension) is not supported by structural data and gave too much syn-rift thermal uplift and emergence at the end of the Jurassic. 2D forward syn-rift and post-rift modelling, using the flexural cantilever model of continental lithosphere extension, was carried out to constrain estimates of ]3-stretching factors and lithosphere flexural strength. The forward syn-rift model gave a profile of laterally varying 13 (Fig. 7b). This laterally varying ]3-profile, with a ~av-- 1.14, was also used in the reverse post-rift modelling. The reverse post-rift model, with laterally varying ]3, satisfactory restores eroded syn-rift surfaces to sea level (Fig. 4d). The restoration to the endPaleocene (Fig. 4e) using the variable [3-profile is little different from the constant [3-model (Fig. 4c), since both 13estimates are similar (1.12 cf. 1.14).

Reverse post-rift modelling using a regional uplift in the early Paleogene The discrepancy in the water depth of 375 m at the end of the Paleocene in the constant-]3 model (Fig. 4c) implies 375 m of regional Paleocene uplift across the Outer Moray Firth with respect to McKenzie (1978) post-rift thermal subsidence with = 1.12. The present-day section has been reverse post-rift modelled using ~ = 1.12 and the Haq sealevel curve, but also with an additional regional Paleocene uplift of 375 m (Fig. 5). The restored section to the end-Paleocene, as well as to the endJurassic, is in agreement with the observed palaeobathymetric indicators. This reverse post-rift model, with regional Paleocene uplift, may also be used to produce a restoration to the end of the Cretaceous (65 million years). The restored crosssection to the end-Cretaceous (Fig. 5b) predicts substantial palaeobathymetry of 500-800 m. The post-rift subsidence history for two locations in the Outer Moray Firth Basin is shown in Fig. 5d & e, for a reverse model incorporating the Haq sea-level curve augmented by regional Paleocene uplift. The reverse post-rift model (Fig. 5b, d & e) demon-

49

strates that the assumption of little or zero palaeobathymetry for the Late Cretaceous, frequently made in Airy backstripping studies, in the central North Sea is in error.

Reverse post-rift modelling with respect to an end-Paleocene datum The Beauly Member coals at the end of the Paleocene provide a very reliable estimate of zero palaeobathymetry for the Outer Moray Firth profile. The effect of post-Paleocene global eustasy and regional uplift/subsidence can be removed from the reverse modelling subsidence analysis by restoring the section to end-Paleocene time. In order to achieve this, Eocene and younger stratigraphic units must be removed, lower units decompacted and the end-Paleocene surface set to zero bathymetry. The resulting restored section to the end-Paleocene is shown in Fig. 6a. Flexural backstripping and reverse thermal subsidence modelling from the flattened end-Paleocene datum has then be continued back to the end of the Jurassic using the preferred constant ]3-profile of 1.12 and the Haq et al. (1987) sea-level curve (Fig. 4f). However, in this reverse model (Fig. 6b), the eroded fault-block of Halibut Horst and stratigraphy across the West Fladen High are restored too high at the end-Jurassic, with over 300 m of predicted palaeoemergence (Fig. 6b, e & f). Backstripping using a [3 of 1.0 was required to restore the eroded fault-block of Halibut Horst back to sea level (Fig. 6c). However, a [3 of 1.0 implies no stretching, which is clearly inconsistent with the structural evidence. Furthermore, if the section was backstripped from the present-day using a ~ of 1.0, i.e. no syn-rift thermal uplift and no post-rift thermal subsidence, the eroded crest of Halibut Horst was restored 400 m below sea-level (Fig. 6d & g). This is also an unsatisfactory restoration and the present-day stratigraphy clearly requires some post-rift thermal subsidence during the Cretaceous and Tertiary, as shown in Figs 4 & 5. The results of reverse modelling from a zero bathymetry at the end-Paleocene datum show that net basement subsidence at the end-Paleocene with respect to the end-Jurassic was small or zero. A similar result has also been reported by Joy (1992). Substantial extension clearly took place in the Late Jurassic, and should have been followed by post-rift thermal subsidence in the Cretaceous and Tertiary according to the McKenzie (1978) rift basin model. The evidence for deep-water conditions at the end of the Cretaceous (Hancock 1984; Lovell 1986; Bertram & Milton 1989) suggest that this thermal subsidence did take place. The most reasonable conclusion to be drawn from

50

19. A. NADIN ~ N. J. KUSZNIR

2D STRATIGRAPHICMODELLING IN THE OMF BASIN this is that McKenzie post-rift thermal subsidence took place through the Cretaceous and Paleocene, but was offset by regional Paleocene uplift. An alternative suggestion that post-rift thermal subsidence was delayed until after the Cretaceous (Joy 1992) is not compatible with the physics of the McKenzie (1978) model.

Subsidence analysis of the Outer Moray Firth Basin using structural and stratigraphic forward modelling Forward syn-rifl and post-tift structural and stratigraphic modelling using the 2D flexural cantilever model of continental lithosphere extension (Kusznir et al. 1991; Kusznir & Ziegler 1992) has been applied to the Outer Moray Firth profile. In this model extensional deformation is accommodated by extension along planar faults in the brittle upper crust and an equal amount of distributed plastic deformation in the lower crust and mantle. The flexural isostatic response to extensional faulting in the upper crust produces footwall uplift and hanging-wall subsidence. The distribution of plastic deformation in the lower crust and mantle is defined by a 2D pure-shear l-stretching factor. The flexural cantilever model takes into account the geothermal perturbation generated by lithosphere extension (as a function of ~ in 2D) and subsequent post-rift thermal subsidence. Isostatic loads generated by crustal thinning, geothermal perturbation and re-equilibration are compensated flexurally. The flexural isostatic consequences of instantaneous compacted sediment loading as well as erosional unloading are also incorporated in the model.

Best-fit forward syn-rift model Fault throws and positions observed on the seismic reflection profile across the Outer Moray Firth for Late Jurassic extension were input into the flexural cantilever model. The forward syn-rift model (Fig. 7) was allowed to thermally subside by 18 million years to the end of the Jurassic, to enable comparison with, and constraint by, the reverse modelled section to the base Cretaceous

51

(Fig. 4b). Reverse modelled stratigraphy in Fig. 7c is represented by [+]. The best-fit forward syn-rift model and resulting l-factor profile are shown in Fig. 7a-c. The dip of the rotated fault-blocks predicted by the flexural cantilever model is particularly sensitive to Te and gave a value of Te = 4 km. The l-profile produced by forward syn-rift modelling (Fig. 7b) was also used to refine the t-estimate used in reverse post-rift modelling.

Best-fit forward post-rift model The forward syn-rift model (Fig. 7a-c) includes the geothermal perturbation generated during extension and the subsequent post-rift cooling using a 2D adaptation of McKenzie (1978), defined by a ~-stretching factor profile. Post-rift thermal subsidence arising from this l-profile, and resulting stratigraphy, has been forward modelled for the Outer Moray Firth profile from the end-Jurassic to the present-day. The best-fit present-day stratigraphy obtained using forward modelling is shown in Fig. 7d. First-order eustatic sea-level variations from Haq et al. (1987) were applied (Fig. 8a). The supply of sediment was iterated to provide a bestfit with observed present-day stratigraphic thicknesses, taking into account the effects of sediment loading and compaction, and the observation of zero palaeobathymetry at the end-Paleocene, indicated by the Beauly Member coals (Fig. la). Forward modelling of observed early Paleogene thicknesses across the profile required 390 m of regional uplift in the Paleocene, followed by sediment infill to sea level at the end-Paleocene, in order to be consistent with the zero bathymetry implied by the Beauly Member coals. This regional uplift is equivalent to a regional relative sea-level fall of 390 m (Fig. 8a; line 2). In order to generate the accommodation space for the observed thickness of Eocene sediments (c. 100 m in the north to c. 250 m in the south; Andrews et al. 1990), the forward model required a decrease in the regional uplift of the Paleocene by 160 m in the Eocene, equivalent to a regional 160 m relative sea-level rise (Fig. 8a; line 2). Through the remainder of the Oligocene-Recent the Paleocene regional uplift was decreased to zero at the present-day (Fig. 8a; line 2), and the model was infilled to the presentday basin bathymetry of c. 130 m. The forward

Fig. 5. Reverse post-rift model of the Outer Moray Firth including Paleocene uplift and eustatic sea-level variations. (a) Section reverse modelled to end of the Paleocene using a constant ~ of 1.12 and incorporating an additional 375 m of Paleocene uplift (water-loaded). (b) Section reverse modelled to end-Cretaceous using a constant [3 of 1.12. (e) Section reverse modelled to end-Jurassic using a constant [~of 1.12.(d) Subsidence history for a location in the Dutch Bank Basin. (e) Subsidence history for a location on the Halibut Horst. (f) Global eustatic and regional sealevel variations used for this model.

52

P. A. NADIN & N. J. KUSZNIR

Fig, 6. Reverse post-tift model of the Outer Moray Firth with respect to an end-Paleocene datum. (a) Decompacted and flattened section at the end of the Paleocene, consistent with the presence of end-Paleocene coals. (b) Section reverse modelled from the end-Paleocene to the end-Jurassic using a constant ~1of 1.12. (c) Section reverse modelled from the end-Paleocene to the end-Jurassic using a constant ~ of 1.00. (d) Section reverse modelled from the presentday to the end-Jurassic using a constant ~ of 1.00. (e) Subsidence history for (b) for a location on the West Fladen High. (f) Subsidence history for (b) for a location on the Halibut Horst. (g) Subsidence history for (d) for a location on the Halibut Horst. Reverse models incorporate first-order eustatic sea-level variations (Fig. 4f) but no Paleocene uplift.

2D STRATIGRAPHICMODELLING IN THE OMF BASIN post-rift model to the present-day (Fig. 7d) compares well to the observed present-day stratigraphic thicknesses (Fig. 7d; dotted lines).

53

Magnitude of Eocene subsidence (Fig. 7i) The combined thickness of Eocene-Recent stratigraphy which is observed (Fig. 7i) can be forward

Eustatic sea-level, sediment supply and loading (Fig. 7e & f) The forward model may be used to explore the consequences of changing sea level and sediment supply on stratigraphy. The model shown in Fig. 7e incorporated first-order eustatic sea-level variations (Fig. 8b; line 1) and assumed sediment infilling to sea level at all Cretaceous-Tertiary post-rift times. The stratigraphic thicknesses generated (Fig. 7e) showed little resemblance to those observed; the Cretaceous was too thick and the Tertiary was absent. This was expected since the Outer Moray Firth Basin has a semi-starved Cretaceous post-rift succession (Bertram & Milton 1989; Andrews et al. 1990). In the model shown in Fig. 7f, sediment supply through the Cretaceous was iterated to provide a best-fit with observed present-day stratigraphic thicknesses, taking into account the effects of sediment loading and compaction. The eustatic sea level variation is shown in Fig. 8b. The model was then infilled to sea-level at the end-Paleocene to be consistent with the Beauly Member coals. This model without Paleocene uplift generated too much Paleocene and no Eocene-Recent (Fig. 713. Post-rift thermal subsidence, eustasy and sediment supply alone are not able to fully account for observed early Tertiary stratigraphic thicknesses and observed zero end-Paleocene palaeobathymetries in the central North Sea Basin without additional regional uplift in the early Paleocene followed by rapid regional subsidence in the early Eocene.

Magnitude of Paleocene uplift (Fig. 7g & h) The forward post-rift model of present-day stratigraphy is very sensitive to the magnitude of Paleocene uplift. The model shown in Fig. 7g incorporated 550 m of regional Paleocene uplift (Fig. 8b; line 2) and generated insufficient Paleocene stratigraphic thicknesses and too much Eocene-Recent stratigraphy. The effect of reducing Paleocene uplift to 250 m (Fig.8b; line 3) is shown in Fig. 7h, it generated too much Paleocene and no Eocene-Recent. As discussed earlier in this section, the correct thickness of Paleocene sediments can be modelled by incorporating 390 m of regional Paleocene uplift.

modelled assuming a linear transgression from the end-Paleocene through to the present day (Fig. 8b; line 4). However, the linear transgression gives relatively slow regional subsidence through the Eocene (Fig. 8b; line 4), even when combined with post-rift thermal subsidence, and gives little accommodation space for Eocene sedimentation. Much less Eocene sediment (40-75 m) was generated in this model than is observed, while the Oligocene-Recent was too thick (Fig. 7i). A rapid reduction of uplift is required in the early Eocene of 160 m (Fig. 8a; line 2) in order to generate the observed thickness of Eocene (Fig. 7d).

Subsidence history plots Figure 9 shows the computed subsidence histories of the forward post-rift models for a point located in the Witch Ground Graben. The subsidence history shows the development of stratigraphic thicknesses (including the effects of sediment loading and compaction) and palaeobathymetry. The subsidence history for the preferred post-rift subsidence model is shown in Fig. 9a. This model predicts substantial Cretaceous palaeobathymetry followed by rapid shallowing during the Paleocene then deepening through the Eocene. Regional Paleocene uplift and Eocene subsidence is superimposed on post Jurassic-rift thermal subsidence. This 'kick' in early Paleogene basement subsidence is consistent with that reported for the Outer Moray Firth Basin by previous workers (Bertram & Milton 1989; Milton et al. 1990; Jones & Milton 1994). The subsidence histories shown in Fig. 9b-f of the discounted post-rift models (Fig. 7e-i) are not consistent with the observed present-day stratigraphic thicknesses.

Errors in the analysis of CretaceousTertiary post-rift subsidence Errors in the analysis of the post-rift subsidence arise through errors in interpreting and depth convetting stratigraphy from seismic data and from parameters used to define the forward and reverse modelling processes: compaction/decompaction, effective elastic thickness (lithosphere flexural strength) and the [3-stretching factor.

~stretching factor. Reverse post-riff modelling is dependent on the [3-factor profile used to define post-rift thermal subsidence. A [3 of 1.0 implies no extension and therefore no thermal subsidence,

54

F'. A. NADIN & N. J. KUSZNIR

2D STRATIGRAPHIC MODELLING IN THE OMF BASIN while increasing [3 increases extension and thermal subsidence. The preferred constant ~-factor profile across the Outer Moray Firth profile had a value of 1.12. The error on this relatively small amount of extension (12%) is estimated to be _+0.05, which gives an error in the estimation of regional Paleocene uplift of + 25 m.

Effective elastic thickness. Sensitivity tests using different values of Te have shown that determination of 13 by reverse modelling is relatively insensitive to variations in Te within the range 1-25 kin. The preferred Te constrained by forward syn-rift modelling was 4 km. The tilt of faultblocks calculated using forward syn-rift modelling (the flexural cantilever model) is particularly sensitive to Te. When a Te of 0 km (Airy isostasy) is used in backstripping, large errors can arise when isostatic loads have a short wavelength component, i.e. in the early post-rift phase. If Airy isostasy is used in reverse post-rift modelling, a [3 of 1.35 is required to restore the eroded fault-block crest of Halibut Horst back to sea-level. Such a large [3-factor can not be supported by structural data and forward syn-rift modelling. Sensitivity tests have shown that the estimate of Paleocene uplift using a very large Te of 25 km (360 m uplift) is more similar to results using a Te of 4 km (375 m uplift) than those obtained using a Te of 0 km (290 m uplift). Sediment compaction/decompaction. Forward and reverse post-rift modelling are also dependent on the parameter used to define sediment compaction (compaction length, surface porosity and matrix density). However, for the range of feasible values for each compaction parameter, tests have shown, neglecting overpressuring, that the estimates of end-Paleocene palaeobathymetry from reverse modelling was relatively insensitive to the variations in these parameters. Sensitivity tests for compaction parameters for the Outer Moray Firth profile gave small changes in predicted 13; however, the estimates for end-Paleocene palaeobathymetry

55

and inferred Paleocene uplift were relatively stable (+_ 35 m).

Triassic extension. While most of the Cretaceous and Tertiary post-rift thermal subsidence which must be reverse modelled is derived from Late Jurassic rifting, a small component of the Cretaceous and Tertiary post-rift thermal sub sidence was inherited from earlier Triassic rifting at c. 250 Ma. The line of section of profile 1 crosses an area where the Triassic sequence is thin and it is therefore assumed that the profile was relatively unaffected by Triassic rifting. Tests have shown that using a Triassic [3 of 1.00 (i.e. ignoring Triassic rifting), gave a 25% increase in the estimated Late Jurassic [3 from reverse post-rift modelling. Using a Triassic [3 of 1.50 gave a 40% decrease in Late Jurassic [3. However, for variation in Triassic [3 the palaeobathymetrical estimate for the end-Paleocene was stable and showed little change (_ 10 m). The 375 m of regional Paleocene uplift estimated by reverse post-rift modelling is significant compared with the observed thickness of the Cretaceous and Tertiary and cannot be attributed merely to errors in the parameterization of the reverse model, discussed above. The error on the end-Paleocene uplift is estimated to be c. + 60 m. The best estimate, from reverse post-rift modelling, of the magnitude of regional Paleocene uplift across the Outer Moray Firth profile is therefore 375 m +_60 m (water-loaded).

Discussion Long-term global eustasy Changes in sea level, superimposed on the combined effect of infill of accommodation space created by rifting, subsidence from lithosphere cooling, sediment loading and compaction, controlled the development of accommodation space and distribution of sediments in the Outer Moray Firth Basin. It is important to incorporate an

Fig. 7. Forward syn-rift and post-rift stratigraphic models of the Outer Moray Firth using the flexural cantilever model of rift basin formation. All models include long-term global sea level from Haq et al. (1987). (a) Late Jurassic syn-rift crustal structure before erosion. (b) Variable p-profile (solid) and crustal thinning factor profile after erosion (dashed) generated by forward model; ~av = 1.14. (c) End-Jurassic basin geometry after erosion and 18 million years early post-rift sediment infill; ++++, represents backstripped stratigraphy. (d) Preferred forward post-rift model following 390 m of regional Paleocene uplift and 160 m Eocene subsidence superimposed on global sea-level variation; . . . . represents present-day stratigraphy. (e) Model with no Paleocene uplift infilled to sea level at all CretaceousTertiary post-rift times gives too much Cretaceous and no Tertiary. (f) Model with no Paleocene uplift infilled to sea level at end-Paleocene gives too much Paleocene and no Eocene-Recent. (g) As (f) with 550 m Paleocene uplift, gives insufficient Paleocene. (h) As (f) with 250 m Paleocene uplift, gives too much Paleocene. (i) As (f) with 390 m Paleocene uplift and 50 m of Eocene subsidence gives observed Paleocene but insufficient Eocene. See Fig. 8 for sea-level variations applied to models (d)-(i).

(a)

D

0

0

mr-

m c

~o

o~i C~ 6o

f-

m

,,

m

~.~ o~

~

~ 6o

m

oo

o~o

o~

o

~

m

;;0 .-<

-- + 250 -- + 200 - - + 150

"''~176 "'-

- + 100 ~" ,~

.

+50

o

--50

K~"

-100 - -

~

K~

- 150

_ - 200 - - 250

1

Global

2

Regional sea-level (model 7d)

57

(b)

~ - ~~ _ --

e u s t a t i c s e a - l e v e l ( f r o m H a q et al. 1 9 8 7 )

Time BP (Ma)

mr-

m"a

~0

tJ

C

rn~~

~ --

C

~

~a

m

0

m

r-

0

~

~ ~

r_

0

0

rrl

.~

,.,,

31Q

-- § 250

-- + 200 - - + 150 - - - 100 ,=..., --~-50 ~. .

h

9

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

,l'. ..........

:,

,-"

,i

. . . -"

. I.. ',

1 2 3 4 .......

157

---50

."

---100

,-."

"' r K~"

150 ~

-"

-- - 200 _ - 250

...-"

9 ."' Global eustatic sea-level (model 7e & 7f) Regional sea-level (model 7g) Regional sea-level (model 7h) R e g i o n a l s e a - l e v e l ( m o d e l 7i)

T i m e BP (Ma)

c~

2 ;'~.~""

.''."

H

-- - 300 - - - 350

0

Fig. 8. Sea-level variations used in forward post-rift modelling. (a) Line l, global eustatic sea-level curve from Haq et al. (1987); line 2, combined global eustatic sea-level plus effect of regional Paleocene uplift applied to the preferred forward model Fig. 7d. (b) Line 1, global eustatic sea-level curve; lines 2-4, combined global eustatic sea-level curve plus effect of varying magnitudes of Paleocene uplift and Eocene subsidence.

57

2D STRATIGRAPHIC MODELLING IN THE OMF BASIN

~

e'~

"~

~g ,x:l

9

"~ O

O

8~

o

58

P. A. NADIN t~ N. J. KUSZNIR

estimate of first-order eustatic sea-level into the modelling process. While uncertainty may exist in the magnitude of the global eustatic sea-level curve (cf. Pitman 1978; Watts & Steckler 1979; Watts & Thorne 1984; Hallam 1984; Haq et al. 1987), the incorporation of an estimate of first-order sea-level variations into the forward and reverse modelling process is considered to be more accurate than not using a correction for sea-level change at all. The apparent discrepancy between reverse post-rift modelled (375 m) and observed (zero) palaeobathymetry at the end-Paleocene is interpreted as additional regional uplift relative to the McKenzie (1978) post-rift subsidence curve. This discrepancy of 375 m cannot be accounted for by errors in the Haq sea-level curve alone. Reverse post-rift modelling using a constant sea-level curve relative to present-day gave 175 m of regional Paleocene uplift, while forward post-rift modelling gave 190 m of uplift. The use of alternative long-term eustatic sea-level curves (e.g. Pitman 1978; Watts & Steckler 1979; Watts & Thorne 1984; Hallam 1984) would also give regional Paleocene uplift in the Outer Moray Firth Basin. First-order eustatic sea-level variations from Haq et al. (1987) incorporated into forward postrift models predict Lower and Upper Cretaceous stratigraphy consistent with that observed in the Outer Moray Firth Basin. During the early Tertiary, global eustasy accounts only for minor changes in sea level (Fig. 4f), and cannot explain observed early Tertiary stratigraphic thicknesses and zero palaeobathymetry at the end of the Paleocene in the Outer Moray Firth Basin. Forward and reverse modelling shows that an additional regional uplift event is required, superimposed on post-rift subsidence in the Paleocene followed by rapid decrease in uplift in the early Eocene, in order to generate the observed stratigraphy. Cretaceous palaeobathymetry

Studies by Bertram & Milton (1989), Barr (1991), Milton (1993) and Rattey & Hayward (1993) have

shown that rates of sediment input during Kimmeridgian rifting in the northern North Sea were insufficient to fill the rift bathymetry, and many rift sub-basins were starved of syn-rift fill. Consequently the deposition of the post-rift succession was strongly influenced by the rift topography. Deposition in the early post-rift, e.g. Late Jurassic (Portlandian to mid-Ryazanian; Rattey & Hayward 1993) primarily represents passive infill of syn-rift accommodation space. The pattern of sedimentation in the Lower Cretaceous shows marine onlap on to the major highs, again consistent of infill of inherited syn-rift bathymetry. There is also no evidence from faunal assemblages for shallow-water conditions in the grabens for the early Cretaceous. Early post-rift rates of subsidence outpaced sediment supply which led to increases in water depths to c. 1000+ m in the deeper parts of the basin (Bertram & Milton 1989). Substantial Cretaceous palaeobathymetry in the Outer Moray Firth Basin has also been demonstrated in this study by post-rift reverse modelling (flexural backstripping, decompaction and reverse thermal subsidence modelling), where [3 and Te are constrained by structural information from forward modelling. Reverse post-rift modelling predicts substantial end-of-rift (end-Jurassic/early Cretaceous) palaeobathymetries of the order of 1200 m in the deepest part of the Witch Ground Graben (Fig. 5c & Table 1). By the mid to late Cretaceous, transgression and continued thermal subsidence resulted in the submergence of surrounding clastic hinterlands which led to the starvation of clastic infill. Biogenic and pelagic sediments only partially infilled the early Cretaceous bathymetry created by rifting. Sediment starvation and chalk compaction together with postrift thermal subsidence resulted in substantial palaeobathymetries at the end of the Cretaceous. Reverse post-rift modelling predicted palaeo-water depths of c. 500-800 m at this time for the Outer Moray Firth (Fig. 5b & Table 1). Numerous authors argue for deep-water conditions at the end of the Cretaceous, e.g. Hancock (1984; 1000 m for

Table 1. Estimates of Cretaceous axial palaeobathymetry for the Outer Moray Firth Basin predicted from 2D forward and reverse post-rift stratigraphic modelling Outer Moray Firth Modelling methodology Sea-level variations End Jurassic/Early Cretaceous Mid Cretaceous Late Cretaceous

Cretaceous axial palaeobathymetry Reverse post-rift modelling Relative to Cretaceous sea level up to 1200 m up to 1200 m up to 800 m

Cretaceous axial palaeobathymetry Forward post-rift modelling Relative to Cretaceous sea level up to 1200 m up to 1100 m up to 900 m

2D STRATIGRAPHICMODELLING IN THE OMF BASIN

59

Table 2. Water and air-loaded estimates of regional Paleocene uplift and Eocene subsidence in the Outer Moray Firth Basin predicted from 2D forward and reverse post-rift stratigraphic modelling Outer Moray Firth Modelling methodology Sea-level variations Apparent 'water-loaded' uplift/subsidence Driving 'air-loaded' uplift/subsidence

Paleocene uplift Reverse post-rift modelling Relative to Early Paleocene sea level 375 m 255m

the Chalk), Parker (1975; 600-900m, early Paleocene) and Lovell (1986; 300-500 m, early Paleocene); however these water depth estimates come from fossil evidence and may be unreliable (Bertram & Milton 1989).

R e g i o n a l P a l e o c e n e uplift

The results of 2D reverse modelling of the observed Cretaceous and Tertiary stratigraphy estimate that the 'water-loaded' Paleocene uplift in the Outer Moray Firth had a magnitude of 375 m (Table 2). This 'water-loaded' uplift consists of an initial driving 'air-loaded' uplift of 255-265 m, together with an isostatic response to 'water unloading' of 120-125 m. Previous workers (Milton et al. 1990; Jones & Milton 1994), using sequence stratigraphic evidence, have also demonstrated that uplift in the Outer Moray Firth (e.g. 800 m; Jones & Milton 1994) accompanied sedimentation during the midPaleocene. While Paleocene uplift did generate widespread emergence and erosion in areas such as the East Shetland Platform and Inner Moray Firth, the regional profile in this study is located in the Outer Moray Firth Basin depocentre (Fig. l a) where the stratigraphic succession is conformable and end-Cretaceous palaeobathymetry (up to 900 m; Table l) exceeded Paleocene uplift. The magnitude of regional Paleocene uplift (c. 390 m) predicted for the Outer Moray Firth in this study by forward modelling reduced the water depths in the graben axes to c. 400-450 m. Shallower water depths of c. 100-150 m are predicted across Halibut Horst elevating Ekofisk Formation sediments of the Chalk Group (Knox & Holloway 1992) to within wave-base but not above sea-level.

Eocene-Recent subsidence

Forward and reverse modelling show that regional Paleocene uplift, superimposed on post-Jurassic rift

Paleocene uplift Forward post-rift modelling Relative to Early Paleocene sea level 390 m 265m

Eocene subsidence Forward post-rift modelling Relative to Early Paleocene sea level 160 m llOm

thermal subsidence (McKenzie 1978), decayed from the end-Paleocene to the present. This decay of uplift was rapid (160 m) in the Eocene. Accommodation space for Eocene and younger sediments was provided by starved end-Cretaceous palaeobathymetry inherited from Late Jurassic tiffing and enhanced by post-rift thermal subsidence and global sea-level rise. This endCretaceous accommodation space was greatly reduced in the Paleocene by regional uplift and sediment infill, but then partly restored in the Eocene by collapse of the Paleocene uplift. While Eocene to Recent accommodation space was enhanced by continuing post-rift thermal subsidence this was offset by the global sea-level fall of the Tertiary. By providing accommodation space for EoceneRecent sedimentation, the existence of substantial palaeobathymetry at the end of the Cretaceous precludes the need for a Tertiary rifting episode to generate early Paleogene accommodation space, as suggested by Hall & White (1994). They propose early Tertiary extension with a [3-stretching factor of 1.05-1.20 in order to explain early Tertiary subsidence obtained from Airy backstripping (Hall & White 1994). Such large amounts of stretching, equivalent in magnitude to those seen for the Late Jurassic on the East Shetland Terrace of the northern North Sea Basin (Roberts et al. 1993), are not seen within geological data. While early Tertiary rifting does occur on the Norwegian margin (Skogseid & Eldholm 1988) this extension does not extend as far south as the North Sea Basin. That extension which is seen on seismic reflection data for the early Tertiary is restricted to small faults associated with compaction and intrasediment slope failure, and is very minor. No other mechanism is required for the formation of Tertiary accommodation space other than inherited endCretaceous palaeobathymetry and post-Jurassic thermal subsidence modified by a transient Paleocene uplift event and global sea-level change.

60

P. A. NADIN & N. J. KUSZNIR

Paleocene uplift in the northern North Sea and the development of the Iceland plume Forward and reverse structural and stratigraphic modelling has also been carried out in the South Viking and North Viking Grabens (Nadin & Kusznir 1995) and shows similar regional Paleocene uplift and Eocene subsidence to that of the Outer Moray Firth, but with magnitude increasing towards the north. The timing of uplift and increase in its magnitude towards the north coincides and corresponds with the development of the Iceland plume, located 700-900 km NW of the northern North Sea in Paleocene times (White & McKenzie 1989). The initiation of the Iceland plume in the early Tertiary can be expected to have generated regional uplift. Plume uplift is generated by three distinct mechanisms: (1) dynamic uplift associated with the fluid-flow field of the rising plume material; (2) the isostatic response to hot asthenosphere and lithosphere heated by the plume; and (3) magmatic underplating of the crust (Courtney & White 1986; White & McKenzie 1989). While any or all of these mechanisms could be responsible for regional Paleocene uplift in the northern North Sea, the rapid Eocene subsidence suggests that heated asthenosphere or lithosphere, or magmatic underplating under the northern North Sea Basin are not responsible. Magmatic underplating would give a permanent uplift, while cooling of hot asthenosphere or lithosphere would occur too slowly over 10-100 Ma (Nadin & Kusznir 1995). It is therefore suggested that regional Paleocene uplift in the northern North Sea was generated by the dynamic uplift of the fluid-flow field of the plume. The northern North Sea may have been too distant from the initiating Iceland plume for the other two mechanisms to occur. Rapid Eocene subsidence may have been caused by a decrease in plume vigour and associated dynamic support. Eocene subsidence coincides with the onset of sea-floor spreading in the Norwegian-Greenland Sea which may have decreased the intensity, or modified the pattern of plume flow.

Summary and conclusions Quantitative 2D forward and reverse post-rift stratigraphic modelling has been applied to the Outer Moray Firth Basin in order to examine the post-rift subsidence history, and in particular the magnitude of departure from McKenzie (1978) post-rift thermal subsidence. Reverse post-rift modelling, comprising flexural backstripping, decompaction and reverse thermal subsidence modelling, has been constrained by palaeobathymetrical markers,

consisting of syn-rift erosion surfaces at the end of the Jurassic and coal horizons at the end of the Paleocene. Forward post-rift modelling has been constrained by syn-rift structural data, present-day stratigraphic thicknesses and the zero bathymetry indicated by coals at the end-Paleocene. Palaeobathymetric indicators have been used to calibrate McKenzie (1978) post-rift thermal subsidence and determine [3-factors. Poorly constrained palaeobathymetric markers derived from fossil assemblages have not been used in the subsidence analysis. During forward and reverse post-rift modelling each post-rift unit marks a correlatable step in the development of the Outer Moray Firth Basin. Inherited starved syn-rift accommodation space, as well as thermal subsidence, was important for controlling Lower and Upper Cretaceous sediment deposition. The mainly pelagic Cretaceous sequence lacks accurate water depth indicators. Forward and reverse post-rift modelling of the Outer Moray Firth Basin predict substantial early Cretaceous axial bathymetries of up to 1200 m. The assumption of little or zero palaeobathymetry for the Late Cretaceous, frequently made in Airy backstripping studies in the central North Sea, is inaccurate. The substantial axial bathymetry that existed, c. 500-900 m, during the Late Cretaceous provided much of the accommodation space for Tertiary sedimentation and precludes the need for an early Tertiary rift event. The observed Cretaceous and Tertiary stratigraphy of the Outer Moray Firth Basin was generated by a combination of the effects of inherited syn-rift accommodation space, postJurassic rift thermal subsidence, sediment supply, long-term eustasy and an additional transient regional uplift event of the order of 390 m in the Paleocene (c. 65-55 Ma). Regional Paleocene uplift was followed by rapid regional Eocene subsidence (c. 55-50 Ma) of the order of 160 m. The magnitude and timing of regional Paleocene uplift and rapid Eocene subsidence is consistent with the development of the Iceland plume, situated at a distance of c. 900 km NW of the Outer Moray Firth. Regional Paleocene uplift and Eocene subsidence were superimposed on postrift thermal subsidence (McKenzie 1978) derived from earlier Late Jurassic rifting. No additional mechanism is required to generate Eocene to Recent accommodation space.

The authors wish to thank Alan Roberts and Derek Hendrie for their comments and discussions during the preparation of this paper. Also, we acknowledge the two reviewers for very helpful comments. Phil Nadin was supported by a NERC research studentship.

2D STRATIGRAPHIC MODELLING IN THE OMF BASIN

61

References ANDREWS, I. J., LONG, D., RICHARDS, P. C., THOMSON, A. R., BROWN, S., CHESHER,J. a. & MCCORMAC,M. 1990. The Geology of the Moray Firth. United Kingdom Offshore Regional Report, HMSO, London, for the British Geological Survey. BARR, D. 1991. Subsidence and sedimentation in semistarved half-graben: a model based on North Sea data. In: ROBERTS,A. M., YIELDING,G. & FREEMAN, B. (eds) The Geometry of Normal Faults. Geological Society, London, Special Publication, 56, 17-28. BERTRAM, G. T. & MILTON, N. J. 1989. Reconstructing basin evolution from sedimentary thickness: the importance of palaeobathymetrical control, with reference to the North Sea. Basin Research, 1, 247-257. COURTNEY,R. C. & WHITE, R. S. 1986. Anomalous heat flow and geoid across the Cape Verde rise: evidence for dynamic support from a thermal plume in the mantle. Geophysical Journal of the Royal Astronomical Society, 87, 815-867. DEN HARTOGJAGER, D., GLEES,M. R. & GRIFFITHS, G. R. 1993. Evolution of Paleogene submarine fans of the North Sea in space and time. In: PARKZR,J. R. (ed.) Petroleum Geology of NW Europe: Proceedings of the 4th Conference. Geological Society, London, 59-71. GILTNER, J. R 1987. Application of extensional models to the Northern Viking Graben. Norsk Geologisk Tidsskrift, 67, 339-352. HALL, B. D. & WHITE, N. J. 1994. Origin of anomalous Tertiary subsidence adjacent to North Atlantic continental margins. Marine and Petroleum Geology, 11, 702-714. HALLAM, A. 1984. Pre-Quaternary sea-level changes. Annual Review of Earth and Planetary Sciences, 12, 205-243. HANCOCK, J. M. 1984. Cretaceous. In: GLENNIE, K. W (ed.) Introduction to the Petroleum Geology of the North Sea 1st Edition. Blackwell Scientific Publications, Oxford, 133-150. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea-levels since the Triassic. Science, 235, 1156-1167. JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: Palaeogene stratigraphy and relative sea-level history of the Outer Moray Firth, UK North Sea. Marine and Petroleum Geology, 11, 157-165. JoY, A. M. 1992. Right place, wrong time: anomalous post-rift subsidence in sedimentary basins around the North Atlantic Ocean. In: STOREY, B. C., ALABASTER, T. & PANKHURST, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication, 68, 387-393. KNox, R. W. O'B., MORTON,A. C. & HARLAND,R. 1981. Stratigraphical relationships of Palaeocene sands in the UK sector of the central North Sea. In: ILL1NG, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of NW Europe, Heyden & Son, London, 267-281.

--,

& HOLLOWAY, S. 1992. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey. KUSZNm, N. J. & ZIEGLER, P. A. 1992. The mechanism of continental extension and sedimentary basin formation: a simple-shear/pure-shear flexural cantilever model. Tectonophysics, 215, 117-131. --, MARSDEN, G. & EGAN, S. S. 1991. A flexural cantilever simple-shear/pure-shear model of continental lithosphere extension: application to the Jeanne d'Arc Basin and Viking Graben. In: ROBERTS, A. M., YIELDING,G. & FREEMAN,B. (eds) The Geometry of Normal Faults. Geological Society, London, Special Publication, 56, 41-60. --, ROBERTS, A. M. & MORLEY, C. 1995. Forward and reverse modelling of rift basin formation. In: LAMBIASE, J. (ed.) The Hydrocarbon Habitat of Rift Basins. Geological Society, London, Special Publication, 88, 3-56. LOVELL,J. P. B. 1986. Cenozoic. In: GLENNIE,K. W. (ed.), Introduction to the Petroleum Geology of the North Sea 2nd Edition. London Scientific Publications, 151-169. MARSDEN, G., YIELDING,G., ROBERTS, A. M. & KUSZNIR, N. J. 1990. Application of a flexural cantilever simple-shear/pure-shear model of continental lithosphere extension to the formation of the northern North Sea Basin. In: BLUNDELL,J. & GraBS, A. D. (eds) Tectonic Evolution of the North Sea Rifts. Oxford University Press, 241-261. MCKENZIE, D. P. 1978. Some remarks on the development of sedimentary basins. Earth Planetary Science Letters, 40, 25-32. MmTON, N. J. 1993. Evolving depositional geometries in the North Sea Jurassic rift. In: PARKER, J.R. (ed.), Petroleum Geology of NW Europe: Proceedings of the 4th Conference. Geological Society, London, 425--442. , BERTRAM, G. T. & VANN, 1. R. 1990. Early Palaeogene tectonics and sedimentation. In: HARDMAN, R. E P. & BROOKS, J. (eds) Tectonic Events Responsible for Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. MUDGE, D. C. & COPESTAKE, P. C. 1992. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. NADIN, R A. & KUSZNIR, N. J. 1995. Paleocene uplift and Eocene subsidence in the northern North Sea Basin from 2-D forward and reverse stratigraphic modelling. Journal of the Geological Society, London, 152, 833-848. PARKER, J. R. 1975. Lower Tertiary sand development in the central North Sea. In: WOODLAND, A. W. (ed.) Petroleum and the Continental Shelf of NW Europe, Volume I, Geology. Institute of Petroleum & Applied Science Publishers, Barking, 447-452. PITMAN, W. C. 1978. Relationship between eustasy and stratigraphic sequences of passive margins.

62

P. A. NADIN •

Geological Society' of America Bulletin, 89, 1389-1403. RA~EV, R. P. & HAYWARO,A. B. 1993. Sequence stratigraphy of a failed rift system: the Middle Jurassic to early Cretaceous basin evolution of the Central and Northern North Sea. In: PARKER, J. R. (ed.) Petroleum Geology of NW Europe: Proceedings of the 4th Conference. Geological Society, London, 215-249. ROBERTS,A. M., YIELDING,G., KUSZNIR,N. J., WALKER,I. & DORN-LOPEZ,D. 1993. Mesozoic extension in the North Sea: constraints from flexural backstripping, forward modelling and fault populations. In: PARKER, J. R. (ed.) on the Petroleum Geology of NW Europe: Proceedings of the 4th Conference. Geological Society, London, 1123-1136. ROCHOW, K. A. 1981. Seismic stratigraphy of the North Sea 'Palaeocene' deposits. In: ILLING, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of NW Europe. Heyden & Son, London, 255-266. SCLATER, J. G. & CHRISTm, P. A. E 1980. Continental stretching: an explanation of the post-midCretaceous subsidence in the North Sea Basin. Journal of Geophysical Research, 85, 3711-3739. SKOGSEID, J. & ELDHOLM, O. 1988. Early Cenozoic

N. J. KUSZNIR evolution of the Norwegian volcanic passive margin and the formation of marginal highs. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 49-56. STEWART,I. J. 1987. A revised stratigraphic interpretation of the early Palaeogene of the Central North Sea. In: BROOKS, J. & GLEYYlE, K.W. (eds) Petroleum Geology of NW Europe. Graham & Trotman, London, 557-576. WAWS, A. B. & STECKLER, S. 1979. Subsidence and eustasy at the continental margin of eastern North America. American Geophysical Union Series, 3, 218-234. & THORNE, J. 1984. Tectonics, global changes in sea-level and their relationship to stratigraphical sequences at the US Atlantic continental margin. Marine and Petroleum Geology, 1, 319-339. WHITE, N. J. & LATIN, D. M. 1993. Subsidence analyses from the North Sea 'triple junction'. Journal of the Geological Society, London, 150, 473-488. WHITE, R. S. & MCKENZIE, D. P. 1989. Magmatism at rift zones: The generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, 94, 7685-7729. -

-

Early Paleogene offshore igneous activity to the northwest of the UK and its relationship to the North Atlantic Igneous Province J. D. RITCHIE 1 & K. H I T C H E N 2

Petroleum Geology and Basin Analysis Group and 2Marine Geology and Operations Group, British Geological Survey, Murchison House, West Mains Road, Edinburgh EH9 3LA, UK

Abstract: Offshore occurrences of Early Tertiary igneous rocks to the northwest of Britain are allocated to sevenmajor categories using criteria such as geographical location, structure, age and genetic relationship. The categories are: (1) Faeroe Plateau Lava Group; (2) North Rockall Trough-Hebrides Lava Group; (3) central igneous complexes; (4) volcaniclastic rocks; (5) Faeroe-Shetland Intrusive Complex; (6) the Minch; and (7) Wyville-Thomson Ridge. An igneous chronology is presented based on radiometric age dating (mainly K-Ar whole rock), biostratigraphical data and seismic relationships. Most landward central complexes (i.e. Skye, Rhum, Ardnamurchan, Mull, Arran) pre-date those which are more oceanward (i.e. Erlend, Darwin, St Kilda, Rockall). Exceptions to this trend (i.e. Rosemary Bank, Anton Dohrn, Hebrides Terrace) are probably genetically linked to crustal thinning in the Rockall Trough. The age of the Faeroe Plateau Lava Group, and especially the contiguous Rockall Trough-Hebrides Lava Group, are not particularly well constrained. The main intrusive phase on, and to the southeast of, the Faeroe Islands (i.e. Faeroe-Shetland Intrusive Complex) was a relatively late stage magmatic event, coeval with the opening of the North Atlantic between Greenland and NW Europe.

In any large-scale correlation of the early Paleogene in NW Europe, within either a chronostratigraphic, lithostratigraphic or event-stratigraphic framework, important consideration should be given to the spatial and temporal development of Early Tertiary magmatism within the North Atlantic Igneous Province (NAIP) (Fig. 1). For example, in Scotland and Northern Ireland, plateau lavas, central intrusive complexes, dyke swarms and intrusive sheets comprise the vast majority of preserved Lower Tertiary rocks. Offshore, the area under consideration in this study extends along a broad tract bordering the North Rockall TroughFaeroe-Shetland Basin-M0re Basin axis and represents a volcanic province c. 1000km in length (Fig. 2). The area covered by Lower Tertiary igneous rocks offshore is much greater than that onshore. The most recent review of Tertiary igneous activity within the NE Atlantic borderlands as a whole was published in a special volume edited by Morton & Parson (1988). This followed an earlier review by Noe-Nygaard (1974). Hitchen & Ritchie (1993) have previously attempted to place some aspects of Early Tertiary igneous activity from offshore NW Britain within the broader chronostratigraphic framework of the NAIP. The development of a chronostratigraphic framework for igneous activity within the NAIP is dependent on absolute ages from radiometric age-

dating combined with geochemical data for rocksuite comparisons and magnetic polarity signatures. Biostratigraphical analyses can also be a useful constraint on the validity of radiometric ages, especially if tied into seismic reflection data. In the study area, ages are mainly based on K-Ar (whole rock) data. Such ages may be regarded as significant if they are repeatable, if they do not contradict other geological evidence and if they are derived from fresh (unweathered) rock samples which are not low in potassium and radiogenic argon. This paper is based on relevant information from most of the offshore hydrocarbon wells drilled northwest of Britain and all the British Geological Survey (BGS) shallow boreholes in the area. This assessment includes the interpretation of thousands of kilometres of commercial multichannel and BGS high-resolution seismic data.

Regional setting Following initiation, the Iceland mantle plume or hotspot continued to develop until it impinged on the base of the lithosphere at c. 62 Ma (Skogseid et al. 1992; White 1993). This occurred c. 8 million years before continental separation and the opening of the North Atlantic Ocean between Greenland and NW Europe (Fig. 3). The plume was originally centred close to the margin of East Greenland at

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlationof the Early Paleogene in NorthwestEurope, Geological Society Special Publication No. 101, pp. 63-78.

63

64

J. D. RITCHIE & K. HITCHEN

Fig.1. North Atlantic reconstruction during magnetic anomaly C23 time (Ypresian) showing the location of the North Atlantic Igneous Province (modified after White 1989). The inferred position of the Iceland Plume is indicated by a circle. The dark shading represents areas of seaward dipping reflectors; the lighter shading represents areas of other igneous features including flood basalts, central igneous complexes, sill and dyke complexes, etc. The area outlined by dashes represents the part of the province under consideration in this study (see Fig. 2).

Kangerlussuaq (White 1993) (Fig. 1) and was responsible for an extremely large outburst of magmatism (crustal underplating, intrusive activity and volcanism) which affected an area > 2000 km in diameter, concentrated along the conjugate rifted margins of East Greenland and West Rockall Plateau-Faeroes-North Norway (e.g. White 1988, 1989, 1993). Igneous activity associated with this mantle plume reached its acme within 3-4 million years of its inception (White 1989) corresponding with the peak of igneous activity in northwest Britain at c. 59 Ma (Mussett et al. 1988) and in West Greenland (Upton 1988; Larsen et al. 1992).

The presence of the plume had a big influence on the subsidence and rifling history of existing basins along the NW European continental margin. The effects on the Faeroe-Shetland and the Sea of Hebrides Basins have been partly addressed by Turner & Scrutton (1993) and England et al. (1993), respectively (Fig. 2). Both noted late Danian to early Thanetian uplift around the basins, followed by rapid subsidence. These events are related to the predicted effects of uplift in the initial stage of plume formation followed by subsequent extension (e.g. White 1988; Wilson 1993).

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT

Structure, distribution, nature and age Faeroe Plateau Lava Group (FPLG)

On the Faeroe Islands, geological mapping, seismic refraction investigations and the results of deep drilling the Lopra-1 and Vestmanna-1 boreholes (Fig. 2) have suggested that the generally northwesterly-derived basalts of the FPLG are c. 5.5 km thick (Berthelsen et al. 1984; Waagstein 1988). The FPLG comprises three distinct 'formations', originally defined on the basis of field-parameters and unconformities (Noe-Nygaard & Rasmussen 1968), and designated the lower (thickness > 3 kin), middle (thickness 1.4 km) and upper (thickness > 0.9 kin) 'series'. All the FPLG basalts are subaerially extruded tholeiites with most meaningful geochemical variation apparent within TiO 2, FeO and MgO ratios. Parts of the uppermost Middle Lava Formation and the Upper Lava Formation are midocean ridge type basalts (MORB) (Waagstein 1988). Offshore dredges close to the Faeroes recovered basalts of similar compositions to the onshore FPLG, although some mildly alkalic types have also been recorded (Waagstein 1988). The three formations of the FPLG are cut by a later intrusive phase of basalts and picrites (Hald & Waagstein 1991). These basalts are mainly tholeiitic with a MORB-type genesis and are geochemically very similar to those of the Upper Lava Formation (Hald & Waagstein 1991). The age of the FPLG is poorly constrained onshore because of the lack of marine sediments and fossils and the generally unsuitable nature of the volcanics for radiometric age-dating purposes. Tarling et al. (1988) recorded a wide range of K-Ar ages (66-50 Ma, latest Maastrichtian-Ypresian) from the freshest samples. A number of magnetostratigraphic correlations exist for the FPLG (e.g. Abrahamsen et al. 1984; Tarling et al. 1988; Waagstein 1988; Morton et al. 1988a; Hitchen & Ritchie 1993). The scheme preferred in Fig. 3 agrees mainly with that of Waagstein (1988) and Hitchen & Ritchie (1993), i.e. the Lower Lava Formation was extruded during C26r-C25n (although one or two flows at the top of the sequence are reversely magnetized) and the Middle and Upper Lava formations during the later part of C24r. The top of the Lower Lava Formation is unconformably overlain by a very thin, tuffaceous coal-bearing lacustrine sequence, biostratigraphically dated as late Paleocene in age (Lund 1983). Any basalts deposited during the earlier part of C24r have almost all been removed prior to this sedimentary sequence being deposited (Fig. 3). There is no direct age for the intrusive group of rocks other than that part of the complex appears synchronous with the Upper Lava Formation while

65

other parts clearly post-date it (Hald & Waagstein 1991). Offshore, to the southeast of the Faeroes, only well 205/9-1 has proved the mapped continuation of the FPLG (Fig. 2). Biostratigraphical data from above and below the thin lava interval confirm that this well penetrated lavas of the same age as the Faeroes Lower Series. The apparent bilobal featheredge of the volcanics drilled by wells 209/3-1, 209/4-1A and 209/9-1 (Fig. 2) is caused by the radial extrusion of lavas from two central igneous complexes, the Erlend and West Erlend centres (see below) (Gatliff et al. 1984). The Erlend lavas postdate the bulk of the FPLG Lower Lava Formation although seismic reflection data cannot resolve the nature of the contact between the two groups of volcanic rocks. Strong regional reflections from volcanic rocks in the intervening area between the WyvilleThomson Ridge and to the east of the Brendan Seamount (Fig. 2), thought to represent the Middle and Lower Lava formations (Smythe 1983; Smythe et al. 1983; Hitchen & Ritchie 1987), have been mapped using thousands of kilometres of commercial seismic reflection data. The Lower and Middle Lava formations on land are inferred to be represented offshore by seismic intervals which have transparent and layered characteristics respectively (Smythe et al. 1983). The FaeroeShetland Escarpment (FSE) (Fig. 2) represents a seismically-observable 'step' (Smythe 1983) which marks the southeastern extent of the Middle Lava Formation. The FSE probably represents a palaeoshoreline, where sub-aerially extruded basalts rapidly consolidated against the sea. The Lower Lava Formation extends much further to the southeast of the FSE where it gradually feathers out, possibly at an older paiaeo-shoreline (Fig. 2). This is reasonable evidence for rising sea level or increased tectonic subsidence during the time between Lower and Middle Lava formation extrusion. The age of the Lower Lava Formation offshore can be partly constrained by regional seismic mapping of the Balder Formation tied into commercial well data from the Faeroe-Shetland Basin. The Balder Formation onlaps the Lower Lava Formation hut the nature of the contact is uncertain and beyond the resolution of the seismic data. Smythe (1983) and Smythe et al. (1983) correlate the Balder Formation (latest Paleocene) 'ash-marker' or unit 2b of Knox & Morton (1988) with the thin, tuffaceous, coal-bearing, lacustrine sequence, of similar age (Lund 1983), which separates the Lower and Middle Lava formations on the Faeroes. If the contact offshore is a pinchout and not a very thin interval that can be traced back to the Faeroes, then the correlation of Smythe

66

J. D. RITCHIE & K. HITCHEN

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT

Fig. 2. Location maps of the study area to the (a) west and (b) north of the UK.

67

68

J . D . RITCHIE 8/; K. HITCHEN

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT (1983) and Smythe et al. (1983) is incorrect and the tuffaceous lacustrine sediments on the Faeroes are younger. Waagstein (1988, p. 229) stated that the geochemistry of the tufts on the Faeroes and those of the Balder Formation are disparate, precluding any common source. However, in the same paper (p. 234) he supports the temporal correlation between them (Smythe et al. 1983) on the grounds that non-commonality of source does not necessarily preclude their ages being similar. Morton et al. (1988a) suggested that the Faeroes tufts and some Balder Formation tufts have similar geochemistries. Clearly, the relative source, timing of accumulation and geochemistry of these tufts remains unresolved (Fig. 3). There is no offshore evidence for the age of the Middle Lava Formation. It is possible that the MORB-type Faeroes Intrusive Group tholeiitic rocks, which cut all the basalt formations on the Faeroes, represent the northwestern margin of the Faeroe-Shetland Intrusive Complex (FSIC) (see below). At present, the northwestern extent of this complex is apparently masked by the Lower Lava Formation (Fig. 2). The geochemistries of the FSIC, the stratigraphically youngest part of the Middle Lava Formation, Upper Lava Formation and Faeroes Intrusive Group are similar (see below), and it is also likely that they are coeval (Fig. 3). North Rockall Trough - Hebrides Lavas Group (RTHG)

The area covered by volcanics within the RTHG lies to the southwest of the Wyville-Thomson Ridge (Fig. 2). The distribution of the lavas to the south of the Anton Dohrn and Hebrides Terrace seamounts (Fig. 2) is not well known and their general distribution within the North Rockall Trough is still the subject of some debate (e.g. Roberts et al. 1983; Wood et al. 1987; Abraham & Ritchie 1991; Boldreel & Andersen 1993). Arguably, the lavas are patchily developed, leaving some areas where they are completely absent (Fig. 2). The RTHG forms a heterogeneous, undifferentiated group of volcanic rocks. This is because, unlike the FPLG, there are numerous central igneous complexes within the North Rockall Trough (Fig. 2). Seismic reflection data cannot resolve the contacts between basalts derived from

69

fissures within the trough (the Hebridean Escarpment may mark one such site), and those extruded radially from ?temporally different central volcanoes such as Geikie, Darwin, Sigmundur, St Kilda, Hebrides Terrace Seamount, etc. (Fig. 2). An interesting observation of present subcrops of lavas derived from central complexes, on land and offshore, is that they generally extend 40-50 km from their source (Hitchen & Ritchie 1993). This should be regarded as a minimum figure since these are erosional limits. It suggests that some of the lavas, sampled in boreholes 85/5B, 90/7 and 90/10, might have been sourced from the Geikie and St Kilda central volcanoes (Fig. 2). This is somewhat simplistic since the original extent of the lavas is controlled by factors such as the scale of the complex, palaeo-topography, lava viscosity and whether they were extruded subaerially or into a submarine environment. The southeastern feather-edge of the RTHG has been drilled and samples dated from one commercial well (164/25-1), six BGS boreholes (85/5B, 85/7, 88/10, 90/4, 90/7 and 90/10) and a single dredge haul (J. 1986) (Fig. 2). Well 163/6-1A and borehole 90/18 recovered volcanic rocks from the Darwin Complex (Morton et al. 1988b; Abraham & Ritchie 1991) and Rosemary Bank Seamount (Hitchen & Ritchie 1993), respectively. Boreholes 85/5B, 90/7 and 90/10 are all located west of Lewis (Fig. 2) near the Geikie and St Kilda centres. Basalt samples from 90/7 and 90/10 are subalkaline and have within-plate type tholeiite compositions (A. C. Morton, pers. comm.). The sample from 85/5B is also subalkaline but plots in a transitional area between within-plate and N (normal)-MORB-type basalt (Stoker et al. 1988). Nearby dredge samples (Jones et al. 1986) (Fig. 2) comprise tholeiitic basalts with geochemistry suggestive of a within-plate genesis. Basaltic rocks on St Kilda exhibit some geochemical scattering but plot mainly within the ocean-floor field (Jones et al. 1986). Samples from north of Lewis (164/25-1, 85/7, 88/10 and 90/4) are generally subalkaine basalts of mainly N-MORB type, although the specimen selected for dating purposes in borehole 88/10 was clearly an exotic clast of alkali basalt within a conglomerate made up mainly of subalkaline basalt clasts. K-Ar (whole rock) ages have been obtained from samples in all the wells and boreholes described

Fig. 3. Comparison of existing age data for the North Atlantic Igneous Province. Geomagnetic polarity timescale based Cande & Kent (1995). Partly compiled from Hitchen & Ritchie (1993), Mussett et al. (1988 and refs therein), White & McKenzie (1989), Morton et al. (1988b), Waagstein (1988), Miles & Roberts (1981), Buckley & Bailey (1975), Omran (1990), Hald & Waagstein (1991), Parson et al. (1988), Macintyre & Hamilton (1984), Larsen et al. (1992), Thorpe et al. (1990), Hodgson et aL (1990) and Noble et al. (1988).

70

J. D. RITCHIE • K. HITCHEN

above (Jones et al. 1986; Stoker et al. 1988; Hitchen & Ritchie 1993). Apart from the Jones et al. (1986) dredge sample, ages were obtained from rocks which are petrographically fresh and show no obvious anomalies in the experimental details. Young ages from the dredge sample are dismissed on petrographic grounds due to severe alteration of both phenocryst and groundmass phases (Hitchen & Ritchie 1993). At first glance, the age of the RTHG appears to span a considerable range i.e. 63.1 +__6.4-46.1 _+2.7 Ma (Maastrichtian-Lutetian) (Fig. 3). However, borehole 90/4 ages of 46.4 + 2.8 and 46.1 + 2.7 Ma should be regarded with caution as the percentage of atmospheric contamination was high and K20 levels extremely low (Hitchen & Ritchie 1993). Ages obtained from 88/10 and 164/25-1 are compatible with the biostratigraphic data at these sites. At other sites, where Quaternary sediments immediately overlie the lavas, no such constraints are available. In the absence of any conflicting evidence, these ages should cautiously be regarded as valid. However, as in the Faeroes case, where the 66-50 million years K-Ar age range for the FPLG as a whole has been constrained to 60.5-53.5 Ma using other evidence, so the 63-50 Ma K-Ar age range for the RTHG lavas may yet be further constrained by future investigations.

Central igneous complexes

Sixteen central igneous complexes have been identified offshore within the study area (Fig. 2). Basalts have also been dredged from Bill Bailey Bank and Lousy Bank (Waagstein 1988) (Fig. 2), but these are probably basalt-covered continental fragments rather than central complexes. Most centres are Tertiary in age (either proven or presumed) but Anton Dohrn and Rosemary Bank seamounts are definitely late Cretaceous or older (Jones et al. 1974; Hitchen & Ritchie 1993). The age of the Brendan Seamount remains conjectural but is considered to be Turonian or older by Smythe et al. (1983). These pre-Tertiary complexes are not included in this review. Centres which have not been drilled or which do not crop out at the seabed have been defined purely on geophysical grounds, e.g. Geikie (Evans et al. 1989), Sigmundur (Andersen 1988, unpublished data), Faeroe Bank Centre (Dobinson 1970), Faeroe Channel Knoll (Roberts et al. 1983) and North and South Westray (Rumph et aL 1993). The North Westray centre was formerly called Judd (Hitchen & Ritchie 1987). Hebrides Terrace Seamount has been modelled geophysically by Buckley & Bailey (1975) and Omran (1990). It is considered to the largest

of the BTIP central complexes (Abraham & Ritchie 1991). K-Ar ages of 6 7 - 6 0 M a (late Maastrichtian-Danian) were recorded from reverse magnetized basic rocks (most likely attributed to magnetic anomaly chron C24r) from this seamount (Fig. 3) (Omran 1990). The Blackstones Centre (Durant et al. 1976) has been modelled by McQuillin et al. (1975). K-Ar analysis on a number of samples has yielded a spread of ages with a best estimate for the true age of the centre of 58.6 _ 0.9 Ma (Selandian-Thanetian) (Mitchell et al. 1976) (Fig. 3). Both these centres are probably closer in age to the onshore, rather than offshore, BTIP central igneous complexes (Fig. 3). St Kilda Complex. The detailed geology of the

deeply eroded St Kilda Igneous Complex is well known and summarized in Cockburn (1935) and Harding et al. (1984). The structure and extent of the centre has been defined geophysically by Himsworth (1973) and is considered to be one of largest of the BTIP central complexes (Abraham & Ritchie 1991). Early K-Ar (whole rock) radiometric age-dating of the complex was carried out by Miller & Mohr (1965). Re-evaluation of that work by Pankhurst (1982) reached the conclusion that an age of 60 _+3 Ma (Danian-Thanetian) for the Glen Bay granophyre represented the best general estimate for the overall emplacement for the complex. More recently, a Rb-Sr age of 55 _+0.5 Ma was obtained from the Conachair granite (Brook 1984), considered to be one of the youngest of all the intrusive units on St Kilda. The younger age is preferred by Mussett et al. (1988). Palaeomagnetic determinations by Morgan (1984) on all the major sequences of Hirta (the main Island of the St Kilda archipelago) revealed that they are reversely magnetized. This fact, together with the age determined by Brook (1984), suggests that the complex was emplaced during C24r (or older) and may therefore be younger than central igneous complexes on land, but similar in age to the offshore centres of Darwin, Rockall and Erlend (Fig. 3). Rockall Complex. Detailed rock descriptions from

Rockall Island were first published by Sabine (1960). The Rockall Complex, which is deeply eroded and includes Rockall Island, Hasselwood Rock and Helen's Reel was first defined using magnetic anomaly patterns by Roberts (1969). The earliest K-Ar determination on a sample of aegirine-granite from Rockall Island yielded an age of 60 _+ 10 Ma (Miller & Mohr 1965). Jones et al. (1972) recorded a K-Ar (whole rock) best-fit age of 53.6 _+ 1.2 Ma from samples of aegirine-

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT riebeckite granite [recalculated by Pankhurst (1982) as 55 _ 1 Ma] and 44.3 + 1.8 Ma from an alkali olivine basalt in dredge hauls, c. 14 km east of Rockall Island. Sanderson et al. (1975) reported ages of rocks from submarine outcrops around Rockall Island of 63 +_ 2 Ma (trachydolerite), 65 _ 2 Ma (olivine basalt) and 50 _+2 Ma (olivine basalt cutting an aegirine-granite). The granitic rocks of Rockall Island have yielded Rb-Sr (whole rock) ages of 57+_4, 54_+4 and 54+_4Ma (Danian-Ypresian) (Hawkes et al. 1975). These results, combined with the recalculated average age of Jones et al. (1972), suggest emplacement close to the Paleocene-Eocene boundary (Fig. 3). The reversed magnetic polarity of the Rockall Centre (Lewis et al. 1975) could mean crystallization during magnetic chron anomaly C24r. The Rockall Complex is younger than the BTIP onshore central complexes but similar in age to the offshore St Kilda, Darwin and Erlend centres. It should be noted that a late Cretaceous age (83 _+3 Ma) has been obtained from a troctolitic gabbro taken from Helen's Reef (3 km ENE of Rockall Island) (Roberts et al. 1974). E r l e n d a n d West E r l e n d C o m p l e x e s . The existence

of the Erlend Complex, comprising a layer of volcanic rocks overlying a plutonic body, was initially inferred from geophysical modelling by Chalmers & Western (1979). West Erlend was discovered by seismic mapping and gravity modelling by Gatliff et al. (1984), who described both complexes in detail. The geology of Erlend has also been proved by drilling (wells 209/3-1, 209/4-1A and 209/9-1) (Fig. 2). Seismic reflection profiles indicate that the upper structural levels of the complex are unusually well preserved, with lavas that can be mapped radially away from a central volcanic vent (Gatliff et al. 1984; Hitchen & Ritchie 1987). Drilling results and seismic mapping suggest that the volcanic sequence is up to 1000 m thick and comprises an upper basaltic sequence and a lower intermediate to acidic sequence. Furthermore, the basic lavas can be divided into two distinct groups by well-log response and seismic reflection characteristics (i.e. an upper seismically layered group and a lower seismically more transparent group). The underlying intermediate to acidic sequence is difficult to map but generally appears more restricted in distribution. Samples of the basaltic sequence, recovered from the three wells, have been described as quartz and olivine tholeiitic basalts, although all show a significant degree of alteration (Ridd 1983; Mitchell & Euwe 1988; Kanaris-Sotiriou et al. 1993). In well 209/3-1 the basalts are geochemically N- to T-MORB (transitional between

71

normal and enriched). They are comparable with the lower suite of basalts in well 163/6-1A drilled on the northwest flank of the Darwin Complex in the North Rockall Trough (Morton et al. 1988b) (Fig. 2) and basalts within the VCring Plateau Lava Upper series drilled during ODP Leg 104 (Viereck et al. 1988). Samples of the intermediate to acidic suite of volcanics, recovered from the wells drilled into Erlend, are described as rhyolites and cordierite dacites (often glassy) with subordinate pitchstones and obsidian (Ridd 1983; Mitchell & Euwe 1988; Kanaris-Sotiriou et al. 1993). The unusual peraluminous graphite and cordierite-bearing acidic volcanics of the Erlend Complex probably formed by anatexis of aluminous pelitic Mesozoic sediments (Kanaris-Sotiriou et al. 1993). Similar mineralogies have also been described from the Darwin Complex (Morton et al. 1988b) and the Vcring Plateau Lava Lower series (Viereck et al. 1988). This unusual petrographic similarity suggests a common geological setting for all these occurrences. Interpretation of the radiometric ages from the Erlend Complex is difficult. Samples from the basaltic succession in well 20913-1, analysed using the 4~ (whole rock) technique, yielded inconclusive (unpublished) results of 'broadly Paleocene' aspect. The basalts are too altered to provide a reliable extrusion age. The same well penetrated intermediate to acidic sills beneath the main basalt pile. These were intruded into shales from which Cretaceous microfossils have been recovered (R. E. Dunay, pers. comm.). 4~ (whole rock) age dating from a sample of rhyolite at the base of the intermediate to acidic sequence in well 209/9-1 yielded an age of 64.8 +_0.8 Ma (close to the CretaceousPaleocene boundary) (Hitchen & Ritchie 1987). Ridd (1983) noted that Campanian microfossils had been recovered from peperites (mixed sediment/volcanic rocks formed in a subaqueous environment by explosive interaction) within the same acidic succession as the rhyolite. Mitchell & Euwe (1988) recorded discordant K-Ar (whole rock) age dates from acidic volcanics from nearby well 209/4-1A but their model suggested a 'formation' age of between 58 + 3 and 55 _+4 Ma. They concluded that the heat generated from the overlying 'Lower Tertiary' basalts may have expelled pre-Tertiary noble gas isotopic evidence and reduced the age of the acidic succession to Paleocene. Significantly, late Cretaceous microfossils have been recorded from sediments intercalated with the acidic volcanics (R Connell, pers. comm.). Kanaris-Sotiriou et al. (1993) pointed out that the model of Eldholm et al. (1989) for the VCring

72

J. D. RITCHIE ~ K. HITCHEN

Plateau volcanic margin suggested the possibility of a time gap between the acidic and basaltic magma extrusion. However, this requires magmas to have been ponded in the upper crust for several million years. If the Erlend and Darwin Complexes are of the same age, as structural, geochemical, petrographic and petrogenetic evidence suggest, then a K-Ar (whole rock) age of c. 55 Ma for both the Darwin acidic and basaltic sequences as a whole (Morton et al. 1988b) (see later) would rule out a pre-Tertiary history for the Erlend Complex. Hence, a latest Paleocene age for the Erlend centres seems preferable, although a more thorough review of all the evidence is clearly required. Other magnetic and stratigraphic data help constrain the age of the Erlend Complex. Rumph et al. (1993) noted that the Erlend intrusive centres are reversely magnetized. The results of detailed seismic reflection mapping by Gatliff et al. (1984, fig. 4) show that the Balder Formation onlaps the basaltic volcanic succession. This clearly indicates that the centre is pre-Balder Formation in age (older than c. 54.5 Ma) (Fig. 3), and together with the reversed negative magnetic signature (?C24r), would suggest a minimum age of c. 55.554.5 Ma (Thanetian) for the complex (Fig. 3). The age of the underlying acidic suite of rocks has not yet been fully resolved (see above). To the north of the Erlend Complex, the Brendan Seamount (Fig. 2) is normally magnetized and therefore cannot be synchronous with Erlend. As mentioned above, Smythe et al. (1983) suggest the possibility that it might be Turonian or older. Darwin Complex. The existence of the Darwin

Igneous Complex was predicted from geophysical modelling (Himsworth 1973; Abraham & Ritchie 1991). The northwest flank of the complex was drilled by a stratigraphic test well 163/6-1A in 1980. The well terminated after penetrating 689 m of basaltic volcanics overlying at least 356 m of dacitic extrusives (Morton et al. 1988b). Seismic reflection mapping indicates unusual preservation of the upper structural levels of this volcanic centre (similar to Erlend) with lavas dipping radially away from a volcanic vent (Abraham & Ritchie 1991 ). The basaltic volcanics can be split into an upper succession of olivine-tholeiites, which have alkalic and picritic affinities, and a lower group of olivinetholeiites with N-MORB type compositions (Morton et at. 1988b). The underlying intermediate to acidic volcanics are peraluminous and are described as cordierite dacites (Morton et al. 1988b). K-Ar (whole rock) age determinations on the basalts and dacites by Morton et al. (1988b) show that most of the analysed samples fall within a restricted range between 58 and 51 Ma

(Paleocene-Eocene) with an average of c. 55 Ma. Some dacite samples yielded late Cretaceous ages; however, because of the likely presence of extraneous radiogenic argon, these have been dismissed as being too old (Morton et al. 1988b). The Erlend and Darwin centres show remarkable structural and geochemical similarities and it is possible that they were emplaced synchronously at 55.5-54.5 Ma (Thanetian) (Fig. 3). No palaeomagnetic investigations have been carried out on rocks from Darwin. Volcaniclastic deposits

Paleogene volcaniclastic sediments are commonly preserved in marine sediments in offshore boreholes and commercial wells to the northwest of Britain, especially within the Faeroe-Shetland Basin (Fig. 2). They represent wind-transported debris from sub-aerial volcanic eruptions, possibly including seamounts. The timing of episodes of sub-aerial lava extrusion within the North Atlantic Igneous Province is often difficult to obtain using radiometric age-dating techniques, due to alteration. Volcaniclastic deposits preserved in marine basins can often be biostratigraphically dated and, providing reasonable geochemical comparisons can be made, their study provides a useful tool to assist with the dating of major extrusive episodes. The distribution, age and geochemistry of Paleogene volcaniclastic rocks within the North Sea is fairly well understood (e.g. Knox & Morton 1988). To the northwest of Britain the situation is less clear and no detailed wide-ranging study has yet been undertaken. However, results from commercial wells suggest that unit 2b of Knox & Morton (1988), i.e. the lower tuff-rich part of the Balder Formation corresponding to the lower part of nannofossil zone NP10, close to the PaleoceneEocene boundary within magnetochron C24r (Fig. 3), is by far the most commonly preserved (or observed). The Balder Formation produces a good seismic reflector which can be mapped across, and tied into, many wells in the West Shetland and Faeroe-Shetland basins (Fig. 2). It represents a useful stratigraphic marker and helps constrain the age of the Erlend Complex, and the FPLG Lower Lava Formation (see above). Waagstein and Heilmann-Clausen (1995) have suggested that Balder Formation post-dates the FPLG Upper Lava Formation. This conclusion is highly speculative as it was based on poor quality single-channel seismic data calibrated by four dredge hauls. Only one dredge haul, from 7 km east of the basalt seabed outcrop, yielded dinoflagellate cysts, but these did not produce corroborative ages.

DATING BTIP WITHIN NORTH ATLANTICCONTEXT Volcaniclastic rocks from other stratigraphic levels have also been described. For example, the Selandian tuffaceous horizon in BGS borehole 82/12 (100 km west of southern Shetland) (Fig. 2) is thought to have been deposited during C26r (Morton et al. 1988a) broadly synchronous with the BTIP plateau lavas and central complexes and FPLG Lower Lava Formation (Fig. 3). The geochemistry of the tuff, especially the high Ti content, suggests it may be derived from a Faeroes source (Morton et al. 1988a). F a e r o e - S h e t l a n d Intrusive C o m p l e x ( F S I C )

The FSIC forms a mainly continuous, elongate, NE-SW trending belt of intrusive rocks over 400 km in length and at least 100 km wide (Fig. 2). It extends further to the northeast beyond the study area into the MOre Basin. Its northwest margin is obscured beneath the southeast margin of the FPLG Lower Lava Formation (Fig. 2). However, the FSIC may extend as far as the Faeroes (see above). On seismic reflection data, sills of the intrusive complex can be distinguished from subaerial volcanic intervals by their high amplitude, characteristic diffraction patterns, lateral discontinuity and general chaotic appearance. They are not necessarily concordant with the bedding and tend to mask reflections from deeper seismic events to a greater degree than lavas. The FSIC as a whole includes satellite groups of intrusives around the West Shetland Basin and along the Rona Ridge (Fig. 2), which were formerly assigned to the West Shetland Basin Intrusive Belt (Hitchen & Ritchie 1987). The southwest margin of the complex has not been rigorously defined and may extend into the North Rockall Trough. The FSIC has been proved by a number of wells within the FaeroeShetland Basin (Fig. 2). The volume of intrusive activity, as deduced from seismic mapping, decreases dramatically towards the southwest and southeast. A maximum thickness of intrusive material is difficult to estimate due to masking effects of uppermost intrusive units. However, well 208/15-1A, drilled off the southern flank of the Erlend Complex (Fig. 2), terminated within an apparently massive single intrusion at least 340 m thick (Hitchen & Ritchie 1987). Petrographic, mineralogical and geochemical analysis of intrusives from wells 208/21-1 and 219/20-1 (Fig. 2) suggest that they are dolerites of olivine-tholeiitic aspect with a T-MORB-type genesis (Gibbet al. 1986; Gibb & Kanaris-Sotiriou 1988). A similar type of investigation by Fitch et al. (1988) from 'upper' and 'lower' sills in well 219/28-2 (Fig. 2) reveals olivine-dolerites of disparate chemical composition, but of tholeiitic aspect, (similar to 219/20-1) and alkalic aspect

73

respectively. Major and trace element geochemistry of the upper, tholeiitic sill is similar to that of the Upper Lava Formation on the Faeroes (Fitch et al. 1988) and rare:earth element patterns from tholeiites in wells 219/20-1 and 208/21-1 also suggest geochemical similarities with the Upper Lava Formation (Gibb & Kanaris-Sotiriou 1988). Plausible K-Ar and Ar-Ar (whole rock) ages have been obtained from samples recovered in seven wells which penetrated the FSIC (206/13-1, 208/15-1A, 214/27-1,214/28-1,209/6-1,209/12-1 and 219/28-2) (Figs 2 & 3) (Hitchen & Ritchie 1987, 1993; Fitch et al. 1988). Results from wells 205/10-2B, 208/17-1, 208/21-1 and 219/20-1 (Fig. 2) were considered to be too old because the host sediments are biostratigraphically younger (Gibb & Kanaris-Sotiriou 1988; Hitchen & Ritchie 1993). This problem is most likely caused by the presence of excess radiogenic argon which can be introduced by circulating fluids (Macintyre & Hamilton 1984). The spread of radiometric ages indicates that the sill complex was intruded over a long period of time, i . e . c . 80-48.7 Ma (Campanian-early Lutetian) (Fig. 3). However, most of the activity was concentrated between 55 and 53 Ma (Thanetian-Ypresian), approximately synchronous with the start of the opening of the North Atlantic and intrusion of the Middle and Upper Lava Formations on the Faeroes (Fig. 3). This is in broad agreement with Gibb & KanarisSotiriou (1988), who assigned a maximum Eocene age for the complex mainly on structural grounds. The age of the sill dated in well 209/6-1 (Fig. 3) is now considered geologically implausible as the sediments into which the intrusion was emplaced have been re-dated biostratigraphically as Danian (E Connell pers. comm.). The alkalic, lower sill in well 219/28-2 yielded an age of c. 80 Ma (Campanian) (Fig. 3) and is also geochemically distinct from the tholeiitic Lower Tertiary intrusives. Volumetrically, this phase of intrusive activity is likely to represent a minor part of the FSIC and, speculatively, may have more in common with late Cretaceous igneous activity noted within the contiguous Rockall Trough, e.g. Helen's Reef (Roberts et al. 1974), Anton Dohrn Seamount (Jones et al. 1974), Rosemary Bank Seamount (Hitchen & Ritchie 1993) and the Barra Complex (Scrutton & Bentley 1988). The M i n c h

Samples of igneous rocks have been recovered from two locations within The Minch (Fig. 3). Borehole 88/5 recovered a subalkaline dolerite with geochemistry suggestive of intrusion in a within-plate setting (Stoker 1989). Borehole 88/12

74

J. D. RITCHIE • K. HITCHEN

recovered Paleocene conglomerates with clasts of subalkaline basalt material (Hitchen & Ritchie 1993). K-Ar (whole rock) ages of 57.2 + 0.8 and 57.6 + 0.8 Ma (Selandian-Thanetian) were obtained from the sill drilled by borehole 88/5. These are bracketed by Rb-Sr ages of 59.3 _ 0.7 and 53.5 _+0.4 Ma (latest Danian-Ypresian) from granitic intrusions within the Cuillins and Red Hills, respectively, on Skye (Dickin 1981). A single basaltic clast in 88/12 yielded K-Ar (whole rock) ages of 59.8_+ 1.1 and 61.9 + 1.3 Ma (DanianSelandian). The Minch Fault may have been active at this time with a syn-tectonic basin in the hanging wall collecting coarse-grained basaltic debris from the adjacent Skye Complex (Pankhurst 1982). Alternatively, as boreholes 90/7 and 85/5B (Fig. 2) recorded similar ages, it is possible that lavas of the RTHG extended much further southeast than the present erosional limit (Fig. 2) so that basaltic material could have been derived from the footwall of the Minch Fault.

Wyville-Thomson Ridge (WTR)

The WTR and adjacent Ymir Ridge form a N W SE-trending bathymetric and structural feature which apparently separates the FPLG from the RTHG (Fig. 2). In order to explain the observed gravity-high anomaly, Himsworth (1973) suggested that the ridge was formed by a thick pile of lavas, resting directly on oceanic crust, with a lowdensity root projecting 10 km down into the mantle. A later model by Roberts et al. (1983) suggested that the ridge comprises 12 km of lavas, erupted through NW-SE aligned parallel fissures, overlying sediments on downwarped Cretaceous oceanic crust. Bott (1984) surmised that the present shape of the WTR is the result of simple volcanic loading. More recent seismic data caused Boldreel & Andersen (1993) to suggest that both the WTR and Ymir Ridge are post-latest Thanetian ramp anticlines connected with a fault plane dipping to the north. It is unclear how this structural interpretation accounts for the observed free-air gravity anomalies over the WTR. Borehole 85/2B, located on the extreme southeast of the WTR (Fig. 2), penetrated tuffaceous sediments biostratigraphically dated as latest Paleocene-earliest Eocene age (Stoker et al. 1988) and overlying the main volcanic lava pile. Samples S1, $2 and $3, recovered by Jones & Ramsay (1982) from close to the northwest margin of the WTR, were biostratigraphically dated as Ypresian (earliest Eocene). Hence, although the WTR is overstepped by uppermost Paleocene-lowermost Eocene sediments, it may have been affected by

several compressional events in the Tertiary. Clearly, the ridge is pre-latest Paleocene in age.

Discussion and conclusions Chronostratigraphic correlations of igneous activity within the NAIP are often difficult to make due to the absence of good marine biostratigraphic control and the problems of obtaining reliable radiometric age data. Notwithstanding this, the scheme proposed in Fig. 3 is, we believe, the best attempt at placing igneous activity to the northwest of Britain within a broader NAIP correlation framework. Rosemary Bank and Anton Dohrn are Late Cretaceous in age or older and are associated with rifting in the North Rockall Trough. The initiation of massive Early Tertiary igneous activity on this margin (and also further afield in Greenland and the Faeroes) occurred in response to the Iceland Plume impinging on the base of the lithosphere at c. 62 Ma (White 1993) (Fig. 3). Maximum magmatic activity occurred at 62-58 Ma, after which the influence of the plume gradually waned. Initiation of the North Atlantic Ocean, and formation of the first new oceanic crust between Greenland and NW Europe, happened at c. 5655 Ma (Dewey & Windley 1988; White 1993) during magnetic chron C24r (Fig. 3). The extent to which the plume facilitated the process of crustal splitting and seafloor spreading between Greenland and Europe is still debatable. The dipping reflector series of basalts represent the initial terrestrial extrusions on the flanks of the developing new oceanic margin (Figs 1 & 2). The age of the reflector series varies (Fig. 3), with parts younger than the age of formation of the first oceanic stripe associated with the opening of the North Atlantic (Parson et al. 1988). Onshore, most of the igneous activity occurred during 63-58 Ma, prior to the opening of the North Atlantic. This activity included plateau lavas, central complexes and intrusive swarms, with a culmination c. 59 Ma (Mussett et al. 1988) (Fig. 3). Offshore, the FPLG Lower Lava Formation is probably broadly synchronous with the major part of the onshore BTIE extruded during magnetic chron anomaly C26r-C25n (Fig. 3). The best understood offshore central igneous complexes are Darwin and Erlend. These centres share many similarities in terms of structure, geochemistry and petrogenesis and were probably emplaced c. 55.5-54.5 Ma during magnetic chron anomaly C24r, just prior to the opening of the North Atlantic (Fig. 3). The St Kilda centre is similar in age to Rockall, whereas the more landward sited Hebrides Terrace Seamount and

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT Blackstones Complex are slightly older and closer in age to the onshore central igneous centres (Fig. 3). Mussett et al. (1988), and others before, have stated that there is no temporal pattern to the geographical location of igneous activity within the BTIE However, over the UK northwestern margin as a whole, it appears that the early Tertiary magmatism commenced in the southeast (onshore areas) and generally progressed northwestwards through time. Volcaniclastic deposits are often described from wells drilled offshore to the northwest of Britain. Phase 2b of Knox & Morton (1988), within the Balder Formation (latest Paleocene-earliest Eocene), is by far the most commonly described and forms a strong seismic reflection marker mappable over most of the offshore study area. The age is coincident with the onset of seafloor spreading between Greenland and NW Europe (Fig. 3). The FPLG Middle and Upper Lava formations were probably extruded at c. 54.3-53.5 Ma (latest Thanetian-earliest Ypresian) (Fig. 3), after the Balder Formation tufts and most onshore UK activity. The majority of the FSIC is geochemically similar to the uppermost Middle and Upper Lava formations and was emplaced at c. 55-53 Ma (Thanetian-Ypresian), probably synchronous with the Faeroes Intrusive Group (Fig. 3). An alkaline phase of intrusive activity is thought to have preceded the main part of the complex during late Cretaceous times. In broader NAIP terms, most of the initial pre-rift burst of plume-related magmatism (62-56 Ma) occurred in West Greenland, the Faeroes (Lower

75

Lava Formation) and onshore, followed by offshore, NW Britain. The syn- to post-North Atlantic opening phase of igneous activity was more restricted, occurring closer to the developing oceanic crust between in East Greenland (c. 5654 Ma for East Greenland) (Upton 1988; Noble et al. 1988), the Faeroes (Middle and Upper Lava Formations) and also Rockall and St Kilda. Mainly alkaline intrusive activity in East Greenland continued sporadically until c. 28 Ma ago (Noble et al. 1988). Further advancements in the understanding of offshore chronostratigraphy depend on the acquisition of high quality seismic reflection data and more material being recovered from commercial exploration wells or shallow boreholes. This depends to a great extent on the future hydrocarbon prospectivity of the Faeroe-Shetland Basin and Rockall Trough. Furthermore, more detailed studies should be undertaken regarding the Early Tertiary sedimentary evolution of North Atlantic marginal basins such as the Faeroe-Shetland Basin, Mere Basin and Rockall Trough. In particular attention should be given to the regional tectonic effects caused by the early Tertiary Iceland Plume development and subsequent crustal splitting, when the North Atlantic opened between Greenland and Europe. An understanding of these processes is important in developing the lithostratigraphic and chronostratigraphic relationships between early Tertiary magmatism and sedimentation. Comments from Andy Morton and Robert Knox greatly improved the original text. This paper is published with permission of the Director, British Geological Survey (NERC).

References ABRAHAM, D. A. & RITCHIZ, J. D. 1991. The Darwin complex, a Tertiary igneous centre in the Northern Rockall Trough. Scottish Journal of Geology, 27, 113-125. ABRAHAMSEN,N., SCHOENHARTING,G. & HEINESEN,M. 1984. Palaeomagnetism and magnetic age of the Vestmanna core and magnetic age and evolution of the Faeroe Islands. In: BERTHELSEN,O., NOENYGAARD, A. & RASMUSSEN, J. (eds) The Deep Drilling Project 1980-1981 in the Faeroe Islands. FCroya Frrdskaparfelag, T6rshavn, 93-108. ANDERSEN, M. S. 1988. Late Cretaceous and early Tertiary extension and volcanism around the Faeroe Islands. In: MORTON,A. C. & PARSON,L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 115-122. BERTHELSEN, O. NOE-NYGAARD,A. & RASMUSSEN, J. (eds) 1984. The Deep Drilling Project 1980-1981

in the Faeroe Islands. Annales Societatis Scientiarum Fa~roensis, Supplementum IX. FCroya Fr6dskaparfelag, T6rshavn. BOLDREEL, L. O. & ANDERSEN, M. S. 1993. Late Paleocene to Miocene Compression in the FaeroeRockall area. In: PARKER, J.R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 1025-1029. BOTT, M. H. P. 1984. Deep structure and origin of the Faeroe-Shetland Channel. In: SPENCER, A. M., JOHNSEN, S. 0., MOERK, A., NYSAETHER, E., SONGSTAD, E & SPINNANGR,A. (eds) Petroleum Geology of the North European Margin. Graham & Trotman, London, 341-347. BROOk, M. 1984. The age of the Conachair granite. In: HARDING,R. R., MERRIMAN,R. J. & NANCARROW, P. H. A. (eds) St Kilda: An Illustrated Account of the Geology. British Geological Survey Report, 16 (7).

76

J. D. RITCHIE t~ K. HITCHEN

BUCKLEY, J. S. & BAILEY, R. J. 1975. Geophysical evidence on the nature of the Hebrides Terrace Seamount. Scottish Journal of Geology, 11, 37-45. CANDE, S.C. & KENT, D.V. 1995. Revised calibration of the geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100, 6093-6095. CHALMERS, J. A. & WESTERN, P. G. 1979. A Tertiary igneous centre north of the Shetland Islands. Scottish Journal of Geology, 15, 333-341. COCKBURN, A. M. 1935. The geology of St Kilda. Transactions of the Royal Society of Edinburgh, Earth Sciences, 63, 511-547. DEWEY, J.E & WINDLEY, B.F. 1988. PalaeoceneOligocene tectonics of NW Europe. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 25-36. DICKIN, A. E 1981. Isotope geochemistry of Tertiary Igneous rocks from the Isle of Skye. Journal of Petrology, 22, 155-189. DOBINSON, A. 1970. 1. The development of a marine seismic recording system; 2. A magnetic survey of the Faeroe Bank. PhD Thesis, University of Durham. DURANT, G. E, DOBSON, M. R., KOKELAAR, B. P., MACtYTYRE, R. M. & REA, W. J. 1976. Preliminary report on the nature and age of the Blackstones Igneous Centre, western Scotland. Journal of the Geological Society, London, 132, 319-326. ELDHOLM,O., THIEDE,J. & TAYLOR,E. 1989. Evolution of the Voting margin. In: ELDHOLM, O., THIEDE, J., TAYLOR, E. ET AL. (eds) Proceedings of the Ocean Drilling Programme, Initial Reports, 104B, 1033-1065. ENGLAND, R. W., BUTLER,R. W. H. & HUTTON,D. H. W. 1993. The role of Paleocene magmatism in the Tertiary evolution of basins on the NW seaboard. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 97-105. EVANS, D., ABRAHAM,D. A. & HITCHEN, K. 1989. The Geikie igneous centre, west of Lewis: its structure and influence on Tertiary geology. Scottish Journal of Geology, 25, 339-352. FITCH, E J., HEARD, G. L. & MILLER, J. A. 1988. Basaltic magmatism of late Cretaceous and Palaeogene age recorded in wells NNE of the Shetlands. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 253-262. GATLIFF, R. W., HITCHEN, K., RITCHIE, J. D. & SMYTHE, D.K. 1984. Internal structure of the Erlend Tertiary volcanic complex, north of Shetland, revealed by seismic reflection. Journal of the Geological Society, London, 141, 555-562. GIBB, E G. E & KANAR1S-SOTIRIOU,R. 1988. The geochemistry and origin of the Faeroe-Shetland sill complex. In: MORTON,A. C. & PARSON,L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 241-252.

& NEVES, R. 1986. A new Tertiary sill complex of mid-ocean ridge basalt type NNE of the Shetland Islands: a preliminary report. Transactions of the Royal Society of Edinburgh, Earth Sciences, 7 7 , 223-230. HALD, N. & WAAGSTEIN,R. 1991. The dykes and sills of the Early Tertiary Faeroe Island basalt plateau. Transactions of the Royal Society of Edinburgh, Earth Sciences, 82, 373-388. HARDING,R. R., MERRIMAN,R. J. & NANCARROW,P. H. A. 1984. St Kilda: An Illustrated Account of the Geology. Report of the British Geological Survey, 1 6 (7).

HAWKES, J. R., MERRIMAN, R. J. HARDING, R. R. & DAR~YSHIRE, D. P. E 1975. Rockall Island: new geological, petrological, chemical and Rb-Sr age data. In: HARRISON, R.K. (ed.) Expeditions to Rockall 1971-1972, Institute of Geological Sciences Report, 75/1, 11-50. HIMSWORTH, E. M. 1973. Marine geophysical studies between Northwest Scotland and the Faeroe Plateau. PhD Thesis, University of Durham. HITCHEN, K. & RITCHIE, J. D. 1987. Geological review of the West Shetland area. In: BROOKS, J. & GLENNIE, K. W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 737-749. - & -1993. New K-Ar ages, and a provisional chronology, for the offshore part of the British Tertiary Igneous Province. Scottish Journal of Geology. 29, 73-85. HODGSON, B. D., DAGLEY, P. & MUSSETT, A. E. 1990. Magnetostratigraphy of the Tertiary igneous rocks of Arran. Scottish Journal of Geology, 26, 99-118. JONES, E. J. W. & RAMSAY, A. T. S. 1982. Volcanic ash deposits of early Eocene age from the Rockall Trough. Nature, 229, 342-344. --, MITCHELL. J. G. & PERRY, R. G. 1986. Early Tertiary igneous activity west of the Outer Hebrides, Scotland - evidence from magnetic anomalies and dredged basaltic rocks. Marine Geology, 73, 47-59. --, SHIDO, E & PHILLIPS, J. D. 1972. Igneous rocks dredged from the Rockall Plateau. Nature, Physical Sciences, 237, 118-120. --, RAMSAY,A. T. S., PRESTON,N. J. & SMITH, A. C. S. 1974. A Cretaceous guyot in the Rockall Trough. Nature, 251, 129-131. KANARIS-SOTIR1OU,R., MORTON, A. C. & TAYLOR, P. N. 1993. Palaeogene peraluminous magmatism, crustal melting and continental break-up: the Erlend complex, Faeroe-Shetland Basin, NE Atlantic. Journal of the Geological Society, London, 150, 903 -914. KNOX, R. W. O'B. & MORTON,A. C. 1988. The record of early Tertiary N Atlantic volcanism in sediments of the North Sea basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. LARSEN, L. M., PEDERSEN, A. K. & PIASECKI, S. 1992. Timing and duration of early Tertiary volcanism from West Greenland. In: STOREY, B. C., ALABASTER, Z. & PANKHURST, R. J. (eds) Magmatism and the Causes of Continental

DATING BTIP WITHIN NORTH ATLANTIC CONTEXT

Break-up, Geological Society, London, Special Publication, 68, 321-334. LEWIS, A. G., BECKMAN,G. E. J. & MERRIMAN,R. J. 1975. Notes on geomagnetic work and palaeomagnetic data. In: HARRISON, R. K. (ed.) Expeditions to Rockall 1971-1972. Institute of Geological Sciences Report, 75/1, 50-51. LUND, J. 1983. Biostratigraphy of interbasaltic coals from the Faeroe Islands. In: BOTT, M. H. P., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge: New Methods and Concepts. Plenum Press, New York, 417-423. MACINTYRE, R. M. & HAMILTON,P. J. 1984. Isotopic geochemistry of lavas from sites 553 and 555. In: ROBERTS, D. G., SCHNITKER,D. ET AL. (eds) Initial Reports of the Deep Sea Dilling Project. 48, 775-781. McQUILLIN, R., BACON, M. & BINNS, P. E. 1975. The Blackstones Tertiary igneous complex. Scottish Journal of Geology, 11, 179-192. MILES, P. R. & ROBERTS, D. G. 1981. The magnetisation of Rosemary Bank Seamount, Rockall Trough, Northeast Atlantic. Earth and Planetary Science Letters, 54, 442-448. MILLER, J. & MOHR, P. A. 1965. Potassium-Argon age determination on rocks from St Kilda and Rockall. Scottish Journal of Geology, 1, 93-99. MITCHELL, J. G. & EUWE, M. G. 1988. A model of singlestage concomitant potassium-argon exchange in acidic lavas from the Erlend volcanic complex, north of Shetland Isles. Chemical Geology (Isotope Geoscience Section), 72, 95-109. MITCHELL,J. G., JONES,E. J. W. & JONES, G. T. 1976. The composition and age of basalts dredged from the Blackstones igneous centre, western Scotland. Geological Magazine, 113, 525-533. MORGAN, G. E. 1984. Palaeomagnetism. In: HARDING, R. R., MERRIMAN, R. J. & NANCARROW, P. H. A. (eds) St Kilda: An Illustrated Account of the Geology. Report of the British Geological Survey, 16 (7).

MORTON, A. C. & PARSON, L.M. (eds) 1988. Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39. --, EVANS,D., HARLAND,R., KING, C. & RITCHIE, J.D. 1988a. Volcanic ash in a cored borehole W of the Shetland Islands: evidence for Selandian (late Palaeocene) volcanism in the Faeroes region. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 263-269. , DIXON, J. E., FITTON, J. G., MACINTYRE, R. M., SMYTHE, D. K. & TAYLOR, E N. 1988b. Early Tertiary volcanic rocks in well 163/6-1A, Rockall Trough. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 293-308. MUSSETT, A. E., DAGLEY, P. & SKELHORN, R. R. 1988. Time and duration of activity in the British Tertiary Igneous Province. In: MORTON, A. C. & PARSON,

77

L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 337-348. NOBLE, R. H., MACINTYRE, R. M. & BROWN, P. E. 1988. Age constraints on Atlantic evolution: timing of magmatic activity along the E Greenland continental margin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the ArE Atlantic. Geological Society, London, Special Publication, 39, 201-214. NOE-NYGAARD, A. 1974. Cenozoic to Recent volcanism in and around the North Atlantic Basin. In: NAIRN, A. E. M. & STEHLI, F. G. (eds) The Ocean Basins and Margins (Volume 2). Plenum Press, New York, 391-443. - & RASMUSSEN,J. 1968. Petrology of a 3000 metre sequence of basaltic lavas in the Faeroe Islands. Lithos, 1,286-304. OMRAN, M. A. 1990. Geophysical studies in the Hebrides Terrace Seamount Area. PhD Thesis, University College of Wales, Aberystwyth. PANKHURST, R. J. 1982. Geochronological tables for igneous rocks. In: SUTHERLANDD. S. (ed.) Igneous Rocks of the British Isles. John Wiley, London, 575-581. PARSON, L. M. & THE ODP LEG 104 SCIENTIFIC PARTY. 1988. Dipping reflector styles in the NE Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 57-68. RIDD, M. F. 1983. Aspects of the Tertiary geology of the Faeroe-Shetland Channel. In: BOTT, M. H. P., SAXOV, S., TALWANI,M. & THIEDE,J. (eds) Structure and Development of the Greenland-Scotland Ridge: New Methods and Concepts. Plenum Press, New York, 91-108. ROBERTS, D. G. 1969. A new Tertiary volcanic centre on the Rockall Bank. Nature, 223, 819-820. - - - , Bowr, M. H. P. & URUSXI, C. 1983. Structure and origin of the Wyville-Thomson Ridge. In: BOTT, M. H. P., SAXOV,S., TALWAN1,M. & THIEDE,J. (eds) Structure and Development of the GreenlandScotland Ridge: New Methods and Concepts. Plenum Press, New York, 133-158. - - - , FLEMMING,N. C., HARRISON,R. K., BINNS, P. E. & SNELLING, N. J. 1974. Helen's Reef: a microgabbroic intrusion in the Rockall intrusive centre, Rockall Bank. Marine Geology, 16, M21M30. RUMPH, B., REAVES, C. M., ORANGE, V. G. & ROBINSON, D. L. 1993. Structuring and transfer zones in the Faeroe Basin. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 999-1009. SABINE, P. A. 1960. The geology of Rockall, North Atlantic. Bulletin of the Geological Survey of Great Britain, 16, 156-178. SANDERSON,R. W., DARBYSHIRE,D. P. E, ROBERTS, D. G. & EDEN, R. A. 1975. The petrology and isotopic-age dating of in situ rocks samples from the D. E. Vickers Voyager and Pisces III cruise to Rockall

78

J. D. RITCHIE •

Bank, 1973. Journal of the Geological Society, London, 132, 334. SCRUTTON, R. A. & BENTLEY, P. A. D. 1988. Major Cretaceous volcanic province in southern Rockall Trough. Earth and Planetary Science Letters, 198-204. SKOGSEID,J., PEDERSEN,T. ELDHOLM,O. & LARSEN,B. T. 1992. Tectonism and magmatism during NE Atlantic continental break-up: the V0ring Margin. In: STOREY, B. C., ALABASTER, T. & PANKHURST, R. J. (eds) Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication, 68, 305-320. SMYTHE, D. K. 1983. Faeroe-Shetland Escarpment and continental margin north of the Faeroes. In: BOTT, M. H. P., SAXOV,S., TALWANI,M. & THIEDE,J. (eds) Structure and Development of the GreenlandScotland Ridge: New Methods and Concepts. Plenum Press, New York, 109-119 --, CHALMERS, J. A., SKUCE, A. G., DOBINSON, A. & MOULD, A. S. 1983. Early opening history of the North Atlantic - I. Structure and origin of the Faeroe-Shetland Escarpment. Geophysical Journal of the Royal Astronomical Society, 72, 373-399. STOKER, M. S. (compiler) 1989. British Geological Survey' shallow drilling programme 1988. Final report. BGS Technical Report WH/89/19C. --, MORTON, A. C., EVANS, D., HUGHES, M. J., HARLAND,R. & GRAHAM,D. K. 1988. Early Tertiary basalts and tuffaceous sandstones from the Hebrides Shelf and the Wyville-Thomson Ridge. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 271-282. TARLING, D. H., HAILWOOD, E. A. & LOVLIE, R. 1988. A palaeomagnetic study of the lower Tertiary lavas in E Greenland and comparison with other lower Tertiary observations in the northern Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the ArE Atlantic. Geological Society, London, Special Publication, 39, 215-224. THORPE, R. S., TINDLE, A. G. & GLEDHILL,A. 1990. The petrology and origin of the Tertiary Lundy granite (Bristol Channel, UK). Journal of Petrology, 31, 1379-1406. TURNER, J. D. & SCRUTTON, R. A. 1993. Subsidence patterns in western margin basins: evidence from the Faeroe-Shetland Basin. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 975-983. UPTON, B. G. J. 1988. History of Tertiary activity in the N Atlantic borderlands. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 429-453. VIERECK, L. G., TAYLOR,R N., PARSON, L. M., MORTON, A. C., HERTOGEN, J., GIBSON, I. L. & THE ODP 9 1 ,

K. HITCHEN LEG 104 SCIENTIFIC PARTY. 1988. Origin of the Palaeogene V0ring Plateau volcanic sequence. In: MORTON, A.C. & PARSON,L.M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 69-83. WAAGSTEIN, R. 1988. Structure, composition and age of the Faeroe basalt plateau. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 225-238. & HALD, N. 1984. Structure and petrography of a 660m lava sequence from the Vestmanna-1 drill hole, lower and middle basalt series, Faeroe Islands. In: BERTHELSEN, O., NOE-NYGAARD,A. & RASMUSSEN, J. (eds) The Deep Drilling Project 1980-1981 in the Faeroe Islands. FCroya FrGdskaparfelag, TGrshavn, 39-70. & HEILMANN-CLAUSEN,C. 1995. Petrography and biostratigraphy of Palaeogene volcaniclastic sediments dredged from the Faeroes Shelf. In: SCRUTTON, R. A., STOKER, M. S., SHIMMIELD, G. B. & TUDHOPE, A. W. (eds) The Tectonics, Sedimentation and Palaeoceanography of the North Atlantic Region. Geological Society, London, Special Publication, 90, 179-197. , MORTON, A. C., PRAEGEL,N. O. & TAYLOR,P. N. 1989. Highly alkaline lapilli tufts from Rosemary Bank. Terra Abstracts, 1, 31-32. WHITE, R. S. 1988. A hot-spot model for early Tertiary volcanism in the N. Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the PIE Atlantic. Geological Society, London, Special Publication, 39, 3-13. 1989. Initiation of the Iceland Plume and the opening of the North Atlantic Margins. In: TANKARD, A. J. & BALKW1LL, H. R. (eds) Extensional Tectonics and Stratigraphy of the North Atlantic Margins. American Association of Petroleum Geologists Memoir 46, 149-154. 1992. Crustal structure and magmatism of North Atlantic continental margins. Journal of the Geological Society, London, 149, 841-854. 1993. Development of the Rockall Trough and the northwest European continental margin. In: PARKER, J. R. (ed.) Petroleum Geology of Northeast Europe: Proceedings of the 4th Conference. Geological Society, London, 985. & MCKENZIE,D. 1989. Magmatism at rift zones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, 94 (BG), 7685-7729. WILSON, M. 1993. Geochemical signatures of oceanic and continental basalts: a key to mantle dynamics. Journal of the Geological Society, 150, 977-990. WOOD, M. V., HALL, J. & VAN HOORY, B. 1987. PostMesozoic differential subsidence in the north-east Rockall Trough related to volcanicity and sedimentation. In: BROOKS, J. & GLENNIE, K. W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 677~585. -

-

-

-

-

-

-

-

-

-

Controls on Eocene sedimentation in the central North Sea Basin: results of a basinwide correlation study AIDAN

M. J O Y

Fina Exploration Ltd, Fina House, Ashley Avenue, Epsom, Surrey, KT18 5AD, UK; present address: Kerr-McGee Oil (UK) plc, 75 Davies Street, London W1 Y 1FA, UK

Abstract: The Eocene of the central North Sea Basin consists of mudstones with locally developed deltaic and gravity-flow sandstones, some of which are important hydrocarbon reservoirs. The pronounced lateral variability in sandstone thickness, combined with the difficulties of seismic interpretation in the Eocene, obliges explorationists to develop a regional understanding of Eocene depositional systems in order to predict sandstone distribution. However the development of such a regional understanding is hampered by the need to correlate between widely spaced wells. The correlation of basinal sections with their marginal equivalents has proved to be a particularly intractable problem. This study utilizes a simple subdivision of the Eocene into two sequences based upon the recognition of a small number of areally extensive wireline-log and seismic markers. Though the sediments belonging to these two sequences are very similar, the nature and distribution of the large-scale depositional systems developed during the two intervals are quite different. During sequence I time (early to mid Eocene) sand supply was dominated by two deltas, one in UK Quadrant 9 and one in UK Quadrant 21. The late Paleocene Moray Delta formed a positive bathymetric feature in UK Quadrant 15 at this time, suppressing delta development in this region. By sequence II time (mid to late Eocene) a rise in relative sea level had created space in Quadrant 15, allowing a single major deltaic depocentre to develop here; this delta dominated sand supply during sequence II time. Both the location and the morphology of this delta suggest that it was fluvially dominated. This interpretation suggests that the availability of accommodation space for delta development was a fundamental controlling factor in the supply of sand to the basin. There is no evidence for a eustatic control on sand supply.

It has been recognized for some time that the Eocene section in the central North Sea Basin contains commercial hydrocarbons, and exploration interest in the Eocene was strongly increased by the Alba oil discovery in 1984 (Brennand et al. 1990; Bain 1993). Like the underlying Paleocene reservoirs, the Eocene reservoirs are gravity-flow sandstones. However, the Paleocene sandstones are laterally continuous, and Paleocene traps are therefore predominantly structural. By contrast, the Eocene sandstones, though widely distributed, are laterally highly discontinuous (Vining et al. 1993), and Eocene traps are generally stratigraphic. Because of the predominantly stratigraphic nature of E o c e n e prospects the presence of sand is a significant risk on any Eocene prospect, and therefore it is commercially important to understand Eocene depositional systems in order to accurately evaluate this risk. Through its participation in several Eocene plays in the central North Sea, Fina Exploration Limited has access to an extensive seismic and well database covering this economically important section.

In this study, well and regional seismic data have been systematically examined in order to produce widely applicable correlation criteria which may be used to understand Eocene depositional systems over the entire Central North Sea Basin. The depositional systems highlighted during this process have been interpreted in terms of the interplay between sedimentation, subsidence, eustasy and accommodation space.

Depositional environments during the Eocene The Eocene of the central North Sea Basin was deposited in a marine basin some 300 km across; sedimentological and subsidence history modelling studies suggest that water depths in the basin centre may have been as great as 1000 m (Joy 1993). Belts of progradational clinoforms are preserved on both sides of the basin, but while the western margin of the basin was dominated by deltas supplying both sand and mud derived from northern Britain, depositional systems on the eastern margin were exclusively mud-prone.

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlationof the Early Paleogene in NorthwestEurope, Geological Society Special Publication No. 101, pp. 79-90.

79

80

A.M. JOY biostratigraphically in order to facilitate correlation of lithostratigraphic units. However this is not always possible. It was therefore decided to adopt a pragmatic approach in order to subdivide the Eocene. Welllog and seismic markers were sought which are both widespread and easily recognizable and which conformed to the biostratigraphic subdivision of the Eocene illustrated in Fig. 1. For the purposes of this study three markers were identified across the entire area, enabling the Eocene to be subdivided into two sequences (sequences I and II). These markers are as follows. (1) The top of the Eocene; this is a variably developed downhole gamma ray log decrease, often overlain by a distinctive highgamma basal Oligocene section and accompanied by a commonly reported sample colour change from brown-grey to green-grey. The top of the Eocene is usually a strong seismic reflector except in areas of closely spaced intraformational normal faulting. It is taken as the top of sequence II. (Note: in many early wells this marker was originally

Eocene gravity-flow sandstones are laterally restricted. They are typically associated with the palaeoslope, rarely extending far across the basin floor. They are generally massive sandstones containing very little mud; though often structureless they may contain dish structures, low-angle laminations, syn-sedimentary folds, sand injection structures and/or reverse grading (Shanmugam et al. 1995). On seismic data they may present a mounded or hummocky appearance with abrupt lateral terminations, especially in the updip direction. These features suggest that these massive sandstones are not classical turbidites but the deposits of liquefied flows (Lowe 1979, 1982) or debris flows (Shanmugam et al. 1995).

Stratigraphic subdivision of the Eocene The Eocene in the study area is a marine sequence which may be subdivided using a variety of biostratigraphic markers (Fig. 1). Theoretically, therefore, it should be possible to date all available wells

MICRO GEOLOGICAL AGE

PLANKTON

~, ~ N

Ma

="

360--

i

J

THIS STUDY

~ ~) ON

.s . . . .

o

_

Q.

,,=,

-

m ~

~ , 'Z-_--_ ,'.

394--

IZ: m

N _

PE3d

--

:

-

~

"'

i

u

=. ,

0

.

. . . . . . ..

I1.,I

.

.

tl.I

~=

.=

~

?::)

UPPER

TOP EOCENE (TOP SEQUENCE II)

Z O

E II)

=E IE

~.

.~

i

0

=

ILl

_

_ _ -

i,i

iii

o

PE4a _

W

GEOLOGICAL AGE

~ LITHOLOGY

:2Z-Z-Z-Z- Z

~

3

] '

.

.

.

..........

-r

. . . . . . . . . . .

==

tu 0

0

Z

PE3b

z .

--

PE~ 49D--

,?,

PE2b

.

~-~

,

..

PE, b

--~

PE/Plc

. . .. .

.

.

.

LI~ U

,,=,

.

z

.

.

.

.

.

~

~0- -

.

.

.

.

.

O

~

r ttl

==

','~ a:..

ou

. . . . .

_-Z-S-

53.0 ~w

-- TOP SEQUENCE I

-..........,.

,

. . . .

-

_v -v-

v_

3B

BALD,FM.

--z

=

TOP BALDER FM. (BASE SEQUENCE I)

Fig. 1. Composite stratigraphy of the Eocene section in the study area. The positions of some of the more widespread sandstone-bearingintervals are shown. Microplankton zones were erected by Robertson Research International/ Simon Petroleum Technology.

EOCENE SEDIMENTATION IN THE CENTRAL NORTH SEA BASIN

81

!

~

:

~ "0

,.,,,,,,,,

o

<

i,.,., 9

i,,li'i'i,, IO

]!]

.,,,!>

9

.

~

9

0

z~

~J

tie

=z~

.,.c: ,

..........

~ .......

,, 1. . . . .

I,',',',',!~

9

> ~ > ~1 o z

g

u.:2

o

ELSE] t~

Lu

~ =z,,, _uJuJ/ -r.j~(n

(/) '1"

i

. . . . . . . . . . . . "Ll. btl.'.'.'," ". . . . . . . I' . ' . ' . 11' . . I. I ' . ' I . . . , . .. . , . ,. .I . 9. . ,

r

1 f I 1 ~ 9 .] F.. i p~l~,tl~ J,llJ* I I,i I/'," I~'"" III I Jll'~JJ]>>:>> ~ I' . ~, J , ' ' - ' l.P l.l i l.' . . . . 9 .I ' . .~ . ~ [. ' . . ' . ' .I ~. ~. ~. .' .~ .' . ' ' ' ' ' ' ' ' ] >

>

>1 >

I

<

I:=

o uJ

=

~

z~,,,

_

W.,-

-J

.~ llJ

,,,--

"

{~)

o,

:~

I:i:it::l :i:i: :i:i:i:::::::::::::::::::::::::: : : I t 'i

i

"'

.

.

.

.

.

II

IS

':: :'I~: :,;

iir"l~"'*ll

i r

n:3 ::3

t,,,,,,,,,,f ..... :~:

i

>.

I I I I [ ~

"0

I~ ,

. . . . . . . . . . . . . . . . - ..'.'.'.','.','." *

I I I I I I I I I I { I I rl i ill II IlJl [ J IIII

",'.'.'."

~,

.~

I I~> > > >1 I II> >1

0

0 Z

~-

u~o

9 '.'

.'.'

.'.','

.'.'

.1' |~ I II ' - ' - / q "

9'

.......... I ............

:J:J

l:t :1

~ r.,T-1 9>- .~

9 -,--, 0

82

A.M. JOY

dated as the top of the Lower Eocene.) (2) The midEocene high-gamma zone; this distinctive zone is typically 10-40 m thick. The top of the zone is dated as basal to intra-mid Eocene; it is taken to mark the boundary between sequences I and II. It correlates with the H2-H3 boundary of Knox & Holloway (1992), separating their informal Caran Sandstone (H2) and Brodie Sandstone (H3) units. This is also usually a strong seismic reflector, except in areas of closely spaced intraformational normal faulting. (3) The top of the earliest Eocene Balder Formation, a distinctive gamma ray and sonic log marker with an almost invariably strong, unfaulted seismic response. This is taken as the base of sequence I (so Balder Formation sands are not included in sequence I). Where the Balder Formation is absent (i.e. in the extreme west of the study area) the base of sequence I is taken to be the top of the Beauly Formation (see Mudge & Copestake 1992; Knox & Holloway 1992). These three horizons appear to fulfil the criteria for well-log and seismic markers mentioned above, and they may be equated with widespread stratigraphic markers identified in detailed recent studies (e.g. Knox & Holloway 1992). It is considered that no other Eocene marker is as widespread or as distinctive as these; however, over smaller areas it may be possible to subdivide the section more finely using more locally developed well-log and seismic markers backed up by biostratigraphic data.

Distribution and sandstone content of Eocene sequences This subject is addressed using regional maps of seismic isopach and of net sandstone thickness for the two Eocene sequences defined above. Reference is also made to a series of representative schematic well logs illustrating the Eocene stratigraphy of the study area (Fig. 2).

Sequence I isopach in time (Fig. 3) This sequence is thickest in two discrete areas: in the Viking Graben in central Quadrant 9 and on the western edge of the Western Central Graben in Quadrant 29. These two depocentres are both mudstone dominated, but whereas the southern depocentre comprises exclusively mudstone, the section in the northern depocentre has an important sandstone component in its lower part (see Fig. 4). Sequence I sediments in the northern depocentre were supplied from northern Britain; the southern depocentre is part of an arcuate belt of mudstone depocentres, mainly off the mapped area, which occur around the southern margin of the Central Graben (Joy 1993) and which may have been

Fig. 3. Time isopach of sequence I. Contours in ms twt (as a first-order approximation, 100 ms twt is c. 100 m).

sourced by a palaeo-Rhine drainage system from continental Europe. Sequence I thins markedly by onlap, and locally appears to pinch out completely, against the upper Paleocene Moray Delta. This delta therefore appears to have formed a palaeobathymetric high during the early to mid Eocene. The basal onlap relationships at the base of sequence I indicate that this sequence infilled pre-existing bathymetry. Over the basinal areas (i.e. Central Graben, South Viking Graben), away from concentrated sediment supply, sequence I is consistently thin [between 100 and 200 ms two-way time (twt)], reflecting its relatively distal, sandstone-poor nature. The sequence is particularly thin (< 100 ms twt) on the Jaeren High and over parts of the

EOCENE SEDIMENTATION IN THE CENTRAL NORTH SEA BASIN

Western Platform. This suggests that these areas, which were the shoulders of the Mesozoic rift basin, still formed positive features during the early to mid Eocene (Joy 1993).

Sequence I net sandstone isopach (Fig. 4) Sequence I sands include the Frigg Formation sandstones in Quadrant 9 (Deegan & Scull 1977) and the majority of the Tay Formation sandstones in Quadrants 21 and 22. However, they do not include the Gryphon reservoir sandstones and the lower Tay Formation sandstones which belong to the Balder Formation. During sequence I times the main sand supply route to the basinal Frigg Formation was situated in the Quadrant 9 area; this appears from its seismic character to have been a delta. A number of some-

83

what smaller sand fairways in the south of Quadrant 21 supplied sand to the Tay Formation. Two, possibly three, sand supply routes can be distinguished; these may have been associated with a delta visible on seismic data to the west. There is a prominent NW-SE component to the sand isopachs; this reflects deflection of sand-bearing gravity flows by newly rejuvenated fault-controlled structures. This is linked to a widespread, if mild, episode of tectonism of earliest Eocene age, possibly associated with the opening of the Norwegian-Greenland Sea. Sandstones are present throughout sequence I in the central part of the mapped area, but sand-supply routes cannot be mapped with confidence. A relatively thin early Eocene delta formed in Quadrant 14 and western Quadrant 15, but while this may have supplied sand during early sequence I time it cannot be correlated with later sequence I sandstones. The upper part of sequence I in Quadrants 14 and 15 appears to be a thin, condensed mudstone (Figs 2 & 3).

Sequence H isopach in time (Fig. 5) Sequence II is generally 50-200 ms twt thick. Typically, it is less prone to short wavelength fluctuations than sequence I. This is because sequence I was deposited on a rugose bathymetric surface which reflected the positions of the Moray Delta, of newly rejuvenated fault structures and of the underlying Mesozoic rift structures (e.g. the South Viking Graben; Fig. 3). By the end of sequence I time much of this short-wavelength bathymetry had been infilled, and sequence II was deposited upon a much smoother surface. The main fluctuations in sequence II thickness are associated with a major delta centred on the northern Witch Ground Graben and the western Fladen Ground Spur. Within this feature the time isopachs reflect the positions of the two long-lived southeastward-directed sand-supply fairways shown on Fig. 6 (see below).

Sequence H net sandstone isopach (Fig. 6)

Fig. 4. Sequence I net sand isopach. Contours in m.

Sequence II sands include the Nauchlan Member of the Alba Formation (Newton & Flanagan 1993). A major delta developed at this time in the western part of the study area, and two persistent sandsupply routes may be distinguished on the basis of the sandstone isopach. One of these fed a series of gravity flows across the Fladen Ground Spur and into an area centred upon central Quadrant 16, while the other fed the Alba sandstones in southern Quadrant 16. The edge of the major sequence II delta protruded into the basin in the centre of Quadrant 15, marking the progradation of the sand

84

A . M . JOY

Fig. 5. Time isopach of sequence II. Contours in ms twt (as a first-order approximation, 100 ms twt is c. 100 m).

supply fairways; this lobate morphology is typical of fluvially dominated deltas (Elliot 1986). The position of the main fluvial input to the depositional system does not appear to have been closely tectonically controlled. There are thick sequence II sands in the Viking Graben in Quadrant 9 and northern Quadrant 16; these appear to be gravity-flow sandstones. However, because of a scarcity of well data it is not clear how these are related to the body of sequence II progradational clinoforms identified on seismic data to the west. There are thin sequence II sands in southwestern Quadrant 21; from seismic and well data these appear to be gravity-flow sandstones that were deposited on or at the base of the palaeoslope. They do not extend on to the basin floor.

Fig. 6. Sequence II net sand isopach. Contours in m.

Patterns of Eocene sedimentation in the central North Sea Basin This subject is addressed using a series of schematic cross-sections across the northern, central and southern parts of the study area.

Northern area (Fig. 7) During sequence I time gravity-flow sandstones were supplied to the South Viking Graben from a series of deltas to the west. The morphology of the basin floor was strongly controlled by buried Mesozoic rift structures, which may have been rejuvenated during earliest Eocene tectonism. Consequently, the sequence I sandstone isopach is closely related to the positions of the Mesozoic structures.

EOCENE SEDIMENTATION IN THE CENTRAL NORTH SEA BASIN

Fig. 7. Schematic cross-section: northern area. Length of section c. 50 km. No vertical scale implied.

85

86

A.M. JOY

It is believed that the location of the sequence I delta in this area was controlled by the position of the upper Paleocene Moray Delta. There was probably relatively shallow water above the Moray Delta throughout sequence I time, and consequently a large sequence I delta could not develop in the central part of the study area. The main fluvial sand supplies were therefore diverted to the northern [and to the southern (see below)] parts of the study area. No sand was supplied to the basin during the later part of sequence I time; instead a thick sequence of mainly middle Eocene mudstones accumulated. The distribution of these mudstones was strongly controlled by the position of the South Viking Graben, which they partially infilled and the flanks of which they clearly onlap (Fig. 3). Sequence II is dominated by sandstone in the northern area. In numerous wells gravity-flow sandstones with 'box-car' log motifs are overlain by a thick deltaic coarsening-upward sequence, indicating that the delta was prograding over deeper-water sediments during sequence II time. However, the South Viking Graben was a pro-

nounced bathymetric feature at this time and the sequence II delta was unable to completely prograde across this area of deep water.

Southern area (Fig. 8) This area lay to the south of the Moray Delta at the end of the Paleocene, and during sequence I time a major mud-dominated delta developed to the south in Quadrants 28 and 29. It is argued that sand supply to the basin was confined between the bathymetric high formed by the Moray Delta to the north and the developing bathymetric high of the mud-dominated delta to the south. This effect was most pronounced during sequence I time; it may have persisted into sequence II time but by then very little sand was being supplied to this part of the basin. Sequence I sands were supplied from the west; some sand was ponded against newly rejuvenated NW-SE structures while some was able to reach the Central Graben in southern Quadrant 22. Seismic evidence suggests that the bathymetric surface had been smoothed by the end of sequence I

Fig. 8. Schematic cross-section: southern area. Length of section c. 70 km. No vertical scale implied.

EOCENE SEDIMENTATION IN THE CENTRAL NORTH SEA BASIN time. The distribution of the scarce sequence II sands in this area appears to have been controlled by the base of the palaeoslope.

Central area (Fig. 9) Deposition in this area during sequence I time was dominated by the presence of the Moray Delta. The position of this delta was originally controlled by the Moray Firth Basin; sand supplies were concentrated into this structure, which formed a major palaeobathymetric low at the end of the Cretaceous (Joy 1993). However, by the end of the Paleocene the Moray Delta formed a positive palaeobathymetric feature with somewhat deeper water areas both to the north, on the East Shetland Platform/Fladen Ground Spur, and to the south, on the Western Platform. As a result, fluvial sand supplies were diverted both north and south of the Moray Delta during sequence I time, and deltas

87

of this age built up in response to this supply of sediment. However, overlying the Moray Delta itself, sequence I consists of a thin condensed mudstone (Fig. 1), and eastward-thickening sequence I mudstones onlap the front of the Moray Delta (Fig. 9). The sequence I sandstones in the central part of the study area are isolated from their marginal equivalents by a sand bypass zone (Fig. 4), which in turn was controlled by the slope of the Moray Delta. Though sand-supply routes cannot be mapped with confidence in this area, the sand was probably supplied from the west and northwest. The gravity flows from which the sand was deposited appear to have been extremely sensitive to bathymetry. Sequence I sediments infilled a significant portion of the water column in the central part of the study area. The ultimate effect of this reduction in palaeo-water depths was to allow the sequence II

Fig. 9. Schematic cross-section: Central area. Length of section c. 100 km. No vertical scale implied. (a) End Paleocene. Maximum basinward extent of Moray Delta. (b) Early to Middle Eocene. A large delta cannot develop due to limited accommodation space. Sequence I is consequently mudstone-dominated. Thin sequence I gravity flow sands supplied from out of plane section. (c) Middle to late Eocene. Subsidence causes relative sea level rise, creating space for a sequence II delta. Delta progrades over 'platform' created by sequence I.

88

A.M. JOY

delta to prograde some 30 km basinwards of its Late Paleocene predecessor. Sand supply to the basin was dominated during sequence II time by this central area delta; the sand bypass zone was less well developed in this area than during sequence I time, and the sand-supply routes can generally be followed from the delta top to the base of the palaeoslope (Fig. 6).

Controls on Eocene sand deposition in the central North Sea Basin This study raises some interesting points about the importance of the various controls on Eocene sand supply and deposition in the central North Sea Basin. The main conclusion is that the pattern of sand supply was dominated by the locations of the deltas, and that the positions of the deltas were critically dependent upon accommodation space. During the late Paleocene, the position of the Moray Delta was controlled by the accommodation space provided by the Moray Firth Basin. By the start of sequence I time this accommodation space had been completely filled and fluvial sand supply was diverted around the northern and southern edges of the Moray Delta to areas where accommodation space was still available. By the start of sequence II time, a rise in relative sea level had created accommodation space over the Moray Delta, and this space was exploited as the main deltaic depocentre returned, broadly speaking, to its earlier location in the central part of the study area. The inferred rise in relative sea level may also have had the effect of reducing sand supply to the basin in the southern part (and to a certain extent in the northern part) of the study area. It seems likely that the rise in relative sea level referred to above was primarily tectonic rather than eustatic in origin. This is because: (1) it allowed the accumulation of a sequence II deltaic depocentre several hundred metres thick, and this is too great for a second-order eustatic fluctuation (Haq et al. 1987); and (2) according to the seismic data, the sequence II delta top was not eroded during the subsequent sea-level fall which would be expected in a eustatic cycle. Basin modelling studies have shown that the Eocene was a period of unusually rapid tectonic subsidence in the central North Sea Basin (Joy 1993). Palaeobathymetry also controlled the degree to which the deltas were able to prograde basinward. In the northern and the southern areas the deltas appear, from seismic-stratigraphic evidence (Milton et al. 1990; Joy 1993), to have been confined to the basin margins by deep water, which was in turn controlled by the structure of the Mesozoic rifts. The pattern of delta development in these areas was therefore dominated by aggrada-

tion. However, replacement of a significant proportion of the water column by sequence I sediment in the central area allowed the sequence II delta to subsequently prograde a significant distance (tens of kilometres) basinward of earlier deltas. On a smaller scale the distribution of gravity-flow sandstones was strongly controlled by fault-controlled palaeobathymetric features, such as the Crawford Ridge in the north of the study area and the L NW-SE structures on the Western Platform in the~ south. Since sand supply to deep-water depositional settings is sometimes used to infer relative sea-level change, attribution of a eustatic cause to stratigraphic phenomena in the central North Sea Basin might yield conflicting results. A study of southern Quadrant 21, for instance, might conclude that relative sea level was low during sequence I time, resulting in erosion in the marginal areas and sand supply to the basin, and that relatively high sea levels during sequence II time had resulted in the reduction of sand supply to the basin. By contrast, a study of southern Qua&ant 16 might reach the opposite conclusion with respect to relative sea level, since sequence II sandstones are more voluminous than sequence I sandstones. Needless to say, both interpretations of apparent sea level change cannot be attributed to eustasy. In fact it is difficult to discern any eustatic signature on this scale of investigation, since: (1) the pattern of sand supply through time and space was so varied; and (2) sand supply appears to have been so strongly controlled by delta location. Even during deposition of the mid-Eocene high-gamma zone, a widespread condensed sequence, which might be attributed to a eustatic highstand, sand supply continued uninterrupted in some parts of the South Viking Graben (Figs 2, 4 & 6).

Distribution o f hydrocarbon shows In the northern part of the study area shows and discoveries are abundant in sequence I sandstones while sequence II sandstones are almost invariably water wet. This indicates that the former are sealed by the thick upper sequence I mudstones. In the central area, by contrast, shows are almost exclusively confined to sequence II sandstones. Clearly, effective migration pathways must cross sequence I (Fig. 3), and it is not clear why no significant hydrocarbons have been discovered in this part of the section. It is hard to believe that this is exclusively due to infelicitous selection of well locations. It may be that the top seal to the sequence I sands has locally been breached by the numerous closely spaced intraformational normal faults which seismic data has revealed cutting the Eocene in this area (Higgs & McClay 1993; Joy 1993).

EOCENE SEDIMENTATION IN THE CENTRAL NORTH SEA BASIN

o

o

e-~

>..,

,..., ~

9

.o, ~

09

,.~

0"~ ~.~ ~ ~

m~ 9

,.~'~

9

=~

.H o

.= " ' ~

.,.~ ~

X4 __~

9

4 s~<

o

89

90

A.M. JOY

In the southern area shows may be encountered in sandstones in any part of the Eocene section. The pattern of discoveries indicates that the main risk on hydrocarbon charge in this area is associated with the presence of a migration pathway from a mature source in the western Central Graben into the reservoirs at the time of migration. Conclusions

The most important feature of this interpretation is that sand supply to the central North Sea Basin during the Eocene was primarily governed by the location of sand supply routes. These were laterally confined by pre-existing accommodation space, and their locations switched from place to place through time (Fig. 10). Consequently, different areas received sand at different times, while some areas received no sand at all during the Eocene. Another important control was the nature of the gravity flows which supplied sand to the basinal

areas. These were liquefied or debris flows rather than turbulent flows; the resulting deposits are laterally highly restricted, and they were preferentially deposited on and at the base o f the palaeoslope. It is a matter of conjecture as to why sand was supplied to the basin by liquefied flows and/or debris flows during the Eocene, in contrast to the more widespread classical turbidites deposited over the same area during the Paleocene. Finally, it is difficult to discern any eustatic control on sand supply to the central North Sea Basin during the Eocene. This may be a consequence of the rapid subsidence which the basin was undergoing at this time. This work was carried out at the Epsom office of Fina Exploration Limited and it is published with Fina's permission. The technical support and advice of Fina's geoscientists - in particular Mick Cope and Joe Staffurth - is gratefully acknowledged, as are the efforts of Fina's draughtspersons.

References

BAIN, J. S. 1993. Historical overview of exploration of Tertiary plays in the UK North Sea. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 5-13. BRENNAND, T. R, VAN HOORN, B. & JAMES, K. H. 1990. Historical review of North Sea exploration. In: GLENN1E, K. (ed.) Introduction to the Petroleum Geology of the North Sea 3rd Edition. Blackwell, Oxford, 1-33. DEEGAN, C. E. & SCULL, B. J. 1977. A Standard Lithostratigraphic Nomenclature for the Central and Northern North Sea. Institute of Geological Sciences Report 77/25. ELLIOT, T. 1986. Deltas. In: READING, H. G. (ed.) Sedimentary Environments and Facies. 2nd Edition. Blackwell, Oxford, 113-154. HAQ, B. U., HARDENBOL, J., & Vail, P. R. 1987. Chronology of fluctuating sea levels since the Triassic (250 million years ago to present). Science, 235, 1156-1166. HIGGS, W. G. & MCCLAY,K. R. 1993. Analogue sandbox modelling of Miocene extensional faulting in the Outer Moray Firth. In: WILLIAMS,G. D. & Dobb, A. (eds) Tectonics and Seismic Sequence Stratigraphy. Geological Society, London, Special Publication, 71, 141-162. JoY, A. M. 1993. An analysis of the subsidence history of the North Sea Basin. PhD Thesis, University of London. KNox, R. W. O'B. & HOLLOWAY, S. 1992. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY,W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham.

LOWE, D. R. 1979. Sediment gravity flows: their classification and some problems of application to natural flows and deposits. SEPM Special Publication, 27, 75-82. 1982. Sediment gravity flows II: depositional models with special reference to the deposits of high density turbidity currents. Journal of Sedimentary Petrology, 52, 279-297. MILTON, N. J., BERTRAM,G. T. & VAYN, I. R. 1990. Early Palaeogene tectonics and sedimentation in the Central North Sea. In: HARDMAN, R. F. P. & BROOKS, J. (eds) Tectonic Events Responsible for Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. MUDGE, D. C. & COPESTAKE, P. 1992. Revised Lower Paleogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. NEWTON, S. K. & FLANAGAN,K. R 1993. The Alba Field: evolution of the depositional model. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 161-171. SHANMUGAM,G., BLOCH, R. B., MITCHELL, S. M. ET AL. 1995. Basin-floor fans in the North Sea: sequence stratigraphic models vs. sedimentary facies. American Association of Petroleum Geologists Bulletin, 79, 477-512. VINING, B. A., IOANNIDES, N. S. & PICKERING, K. T. 1993. Stratigraphic relationships of some Tertiary lowstand depositional systems in the Central North Sea. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 17-29.

An integrated stratigraphy for the Paleocene and Eocene of the North Sea D A V I D C. M U D G E 1 & J O N A T H A N

P. B U J A K 2

1 6 Kilmardinny Crescent, Bearsden, Glasgow G61 3NR, UK 2 The Lexis Group, Albion House, 9 Albion Avenue, Blackpool, Lancashire FY3 8NA, UK Abstract: A high-resolution sequence stratigraphy has been developed for the Paleocene and

Eocene rocks of the North Sea Basin, using electric log data and biostratigraphy. A total of nine stratigraphic sequences has been defined, together with a further 11 subsequences. The sequences are bounded by high-gamma log markers representing maximum condensation surfaces that can be readily correlated over the whole basin. Biostratigraphic analysis has allowed the recognition of 37 last occurrence or top abundance dinocyst bioevents that have also been correlated with the microfaunal succession. Integration of the sequence stratigraphy and biostratigraphy demonstrates that a consistent relationship exists between individual sequence boundaries and bioevents, suggesting that these surfaces are time lines. The usefulness of developing an accurate stratigraphic framework is illustrated by a series of maps showing the lithofacies distribution for individual sequences on a regional scale. At the prospect or field level, it can be used in conjunction with seismic data to map the more subtle Tertiary hydrocarbon plays in the North Sea Basin.

This paper summarizes the results of detailed stratigraphic analysis of Paleocene and Eocene rocks in the North Sea from the Mid North Sea High northwards to the East Shetland Basin (latitudes 56000'-62000 ' N). Its aim is to develop a unified stratigraphic framework through the integration of published biostratigraphic, sequence and lithostratigraphic data. This framework has been used to produce a lithofacies distribution map for each sequence, based upon log interpretation of a database of more than 800 released wells. This study builds upon the work of Milton et al. (1990), Mudge & Copestake (1992a, b), Knox & Holloway (1992), Morton et al. (1993), Mudge & Bujak (1994), Bujak & Mudge (1994) and Neal et al. (1994). The definition of a high-resolution dinocyst zonation, based on analysis of biostratigraphic data from more than 120 wells, has enabled the succession to be divided into 14 zones and 34 subzones that are considered to have chronostratigraphic significance within the North Sea Basin. A total of 37 last occurrence or top abundance dinocyst bioevents have been recognized. The dinocyst zonation has also been integrated with published bioevents based on other microfossil groups. The zones have been correlated with onshore sections, permitting calibration of the North Sea biostratigraphy with the standard nannoplankton and planktonic foraminiferal zones. Nine stratigraphic sequences bounded by maximum condensation surfaces have been defined in the Paleocene to Eocene section. These surfaces are represented by high-gamma log peaks that record

the effect of greatly reduced clastic input into the North Sea Basin as a result of regional sediment starvation and changes in patterns of sediment supply. Variations in rate of tectonic uplift of the western hinterland and differential tilting and subsidence within the basin and along its margins are believed to have been the dominant control on sediment supply and deposition. However, some condensation surfaces may be related to episodes of marine transgression or climatic change. Integration of these stratigraphic sequences with the biostratigraphy has been critical to the development of a confident stratigraphic framework for the Paleocene and Eocene. It can be shown that individual sequence boundaries consistently fall within the same dinocyst subzone, giving a high degree of confidence that these surfaces represent time lines. The sequences also have regional significance and can be mapped over a large part of the North Sea Basin, retaining their internal character from Quadrant 211 in the north to Quadrants 29 and 30 in the south, a distance of over 500 km. Stratigraphic sequence tops for a selection of reference wells in the North Sea are listed in Table 1. The corresponding key dinocyst bioevents for these wells are shown in Table 2. The final phase of stratigraphic analysis has involved the integration of lithostratigraphy into the sequence framework. Recent revisions of Paleocene-Eocene lithostratigraphy by Mudge & Copestake (1992a, b) and Knox & Holloway (1992) closely tie rock units to the sequence stratigraphy and biostratigraphy, thereby reducing

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 91-113.

91

92

D . C . MUDGE 8s J. P. BUJAK

t"~t'r t"qC'qt'q~'q

t'~

< O ~ O

i

< i

-.7 '--"

Cq

E

.~ r~

N

g,

H

93

INTEGRATED NORTH SEA PALEOCENE AND EOCENE STRATIGRAPHY

t"qc'qt"q

,---~ ,---~ ,---~ t"q t"q t-q ~ q

c,,I t"q t"q

.,.-.~,-.~,---~t"qt"-lt"qt"qt-,-lt"qt-'q

t"q q"~

~

r162 "~

9--~ ,.--~

r

(',1('-,I

t",l~,-I

e4

t"xle-,It--,It"qt'-,It"q

~'fi 11

94

D. C. MUDGE • J. P. BUJAK

the potential for miscorrelation and avoiding the creation of diachronous units. The Paleocene and Eocene rocks of the North Sea consist of marine mudstones, sandstones and limestones laid down in a broad basin bounded to the west by a Scotland-Shetland landmass (Fig. 1). Large volumes of clastic material were shed into the basin during periodic uplift and tilting of this western hinterland. Thick sand-prone sediments infilled the relict topography of the earlier Mesozoic graben system, extending on to the platform areas and basinal highs. These deposits thin eastwards and southwards, suggesting that there was very little sediment input from a Norwegian land mass or from the Mid North Sea High. Uplift and rejuvenation of the ShetlandScotland area is believed to have been associated with the development of a hotspot during the Paleocene (Joppen & White 1990). Northwestward migration of this thermal dome coincided with major Hebridean volcanism and source-area subsidence. The cause of subsequent episodes of uplift during the Eocene are not clear; Galloway et al. (1993) relate them to intra-plate stresses developing during rifting between Greenland and Norway. The Paleocene-Eocene section of the North Sea has been the subject of a number of recent stratigraphic studies. Stewart (1987) defined ten depositional sequences in the Paleocene to lower Eocene of the Outer Moray Firth and central North Sea. These were correlated with a microfaunal and dinocyst zonation. The sequences were seismically defined, bounded by unconformities or disconformities marked by downlap, onlap or truncation of reflectors. He interpreted deposition of these sequences as being driven by eustatic sea-level changes with local subsidence important in controlling sediment distribution. Milton et al. (1990) redefined the Stewart (1987) sequences as ten depositional episodes showing that, within the limits of seismic resolution, they were bounded by maximum flooding surfaces rather than unconformities. They postulated that Paleocene to lower Eocene sediments were laid down during a single, second order tectonic cycle of uplift and subsidence, producing relative sea-level fall during Maureen to Forties deposition and sea-level rise during subsequent Dornoch to Frigg times. Morton et al. (1993), in a study of the Paleocene-lower Eocene section in the northern North Sea, tied these sequences to a succession of 30 bioevents. Jones & Milton (1994) recognized a second, lower amplitude tectonic cycle of mid to late Eocene age, terminated by a major late Eocene transgression. They quantify the amplitude and timing of both tectonic cycles in the Outer Moray Firth using seismic data and introduce 13 higher frequency seismo-stratigraphic units, where T20-

'210

NORTH

62~

1211 \ 33

34

~

TAMPEN

SHETLAND

/

PLATFORM

'\i

i

2

EAST ---SHETLAND

'

i 61~

" 2,5

i ' 60~

A

J

i

Bc.~'yl~ !

i 16 ('

i 15

--.

k ~

9

PLATFORM

r' 14

30

~

, / .15

~~,

Fladen Ground

L16

I 59~

UTSIRA

/ t

\

/ ~. ,'\

Ft"IR I~N

!. ~7*t~'P,

SALT PLATFORM

Tn' C i ksedrlnentsPaal eoceand ne Eocene J 1~

",,~

andD~ tPEoceneInFuflaeUnfcni gDeposPalaeocene otin I

0o00,

I~

I

A "''A' Lineol WellCorrelation I

2~

56~ 3~O0'E

Fig. 1. Structural elements of the North Sea basin. WGG, Witch Ground Graben; HI-I, Halibut Horst; NBG, North Buchan Graben; SBG, South Buchan Graben; FBB, Fisher Bank Basin; FMH, FortiesMontrose High.

T82 approximate to the Montrose, Moray and lower Stronsay Groups, and T84-T98 to deposits of the upper Stronsay Group. Galloway et al. (1993) outlined 11 stratigraphic sequences bounded by condensed mudstones within the post-Ekofisk

INTEGRATED NORTH SEA PALEOGENEAND EOCENE STRATIGRAPHY Paleocene-Eocene succession of the North Sea. They also recognized two cycles of relative sealevel rise and fall, but placed the boundary between the two at the maximum condensation surface marking the top of the Balder Formation. They also attributed the lower cycle to an episode of tectonic uplift and subsidence but interpreted the second cycle as being of eustatic origin associated with sea-level changes during major plate reorganization. Mudge & Copestake (1992a, b) revised the lithostratigraphy of the Paleocene-lower Eocene in the Outer Moray Firth and northern North Sea using the maximum condensation surfaces as reference lines. They argued the importance of using biostratigraphy and tied the lithostratigraphic units to ten bioevents. The papers include a set of reference wells displaying gamma, sonic and resistivity log traces, lithostratigraphy and bioevents. Much of this lithostratigraphy was adopted by Knox & Holloway (1992) in their lithostratigraphic revision of the Paleogene in the central and northern North Sea. They also published a large number of graphic reference well sections, including bioevents in some wells. More recently, Mudge & Bujak (1994) have produced a stratigraphic framework for the Eocene of the North Sea, integrating the stratigraphic sequences with bioevents and lithostratigraphic units. A revised dinocyst zonation for the Eocene of the North Sea has also been published by Bujak & Mudge (1994). These papers include a set of graphic reference wells. A Paleocene-Eocene sequence stratigraphy integrated with bioevents has also been produced by Neal et a1.(1994), who identify 14 depositional sequences defined by high-gamma markers in basinal wells.

Biostratigraphic analysis Biostratigraphic subdivision of the Paleocene and Eocene succession in the North Sea Basin is based upon a variety of microfossil groups that occur in wells in the area. These groups can be broadly catagorized as: (1) palynomorphs, mostly comprising dinoflagellate cysts (dinocysts), miscellaneous algae, and the pollen and spores of plants; (2) microfaunas, mostly comprising benthic and planktonic foraminifera, diatoms and radiolaria; and (3) calcareous nannoplankton. Different microfossil groups have been utilized in the various biostratigraphic schemes proposed for the Paleocene and Eocene of the North Sea Basin. Bujak & Mudge (1994) erected a zonation for the Eocene based on 29 dinocyst lastoccurrence and top-abundance bioevents observed in more than 150 wells examined from the North Sea, with recognition of many zones being aided by

95

last occurrence events of additional taxa. The high resolution of this zonal scheme is possible due to the high diversity of dinocysts through most of the section, the planktonic nature of cyst-forming dinoflagellates and the broad geographical ranges of taxa, with many being known outside the North Sea. Morton et al. (1993) identified 15 dinocyst, four miospore and 11 microfaunal bioevents in the Paleocene-lower Eocene section in the northern North Sea, whilst Neal et al. (1994) used 42 bioevents in their study of the Paleocene-lower Eocene in the central North Sea. Many of these last occurrence events had been previously tied to well-log data by Mudge & Copestake (1992a, b) and Knox & Holloway (1992). Examination of sidewall and conventional cores from many wells has also established the first stratigraphic occurrences of dinocyst taxa in the North Sea (Fig. 2). Comparisons of the stratigraphic ranges of taxa in offshore North Sea strata with those in onshore northwest European stratotypes and reference sections, and to a lesser extent worldwide, has permitted correlation with the standard planktonic foraminiferal (P) and calcareous nannoplankton (NP) zones (Bujak & Mudge 1994; Neal et al. 1994). This is necessary because offshore North Sea successions contain few index species characterizing the NP and P zones. It is thus possible to place the North Sea succession within a framework of global stratigraphy and geological processes using dinocyst biostratigraphy (Fig. 3). The dinocyst zonation can also be correlated with the microfaunal succession documented for the North Sea by various authors including King (1983, 1989), Gradstein et al. (1992), Armentrout et al. (1993) and Harland et al. (1992). The microfossils included in these schemes represent a variety of organisms with differing habitats, including benthic calcareous and non-calcareous foraminifera, planktonic foraminifera, diatoms and radiolaria. Although King (1989) erected a relatively highresolution microfaunal zonation for the North Sea, he recognized that correlations based on benthic taxa were subject to diachronism. This view was shared by other authors who erected more conservative low-resolution microfaunal schemes, with Gradstein et al. (1992) only recognizing three zones in the entire post-Balder Eocene section. Most microfaunal bioevents occur in the latest Danian to Ypresian section due to rapidly changing environments that affected most of the semienclosed and environmentally restricted basin at the same time. Early Eocene transgression at the base of the Frigg sequence led to development of increased biotic niches as the basin deepened, so that most benthic microfaunal events became

D.C. MUDGE & J. P. BUJAK

96

slue^eo!8 lsX~ou?a Jo!el~

~.,. ~ c

c~ ~

,

~III

.

~

I

o .

I II

I

I

JO|elf eJ~ "d

elnwo~ '0 ~ l e W O U ~ IS

sno!J!u6euJ"S eleo!lep~o "1 ........................................

I

r

~

"r

,

Lun~oqdoJ~d "d s!sue6u!dd!6 V

i

[ .

[

i . i . :: .

Ig|LIIg6JR LU "~ wmsnene "V

.

.

.

.

.

.

.

.

:: .

:,

:,

i . E . :: .

i

i

!

!

!

i

8$USUepJgM "~)

i

i

!

i

wnp!Ios "O

i . i.

i

i

i

i

s!sueple~s!qao O

I

i . i.

SOl/,puo:)a

i

::

=:

i

i

i

::

i

',

::

wnpni!6uoleue^

'0

. .

. .

c~

~i

=,,

: I ~ i

gSsgJduJo0 "~

! i

,o,,,o~oo.o

wnu~epAqogd "C]

uJme,!qnl H

i

: i

I i ~ i

!ii

: i

,

:

,

,

i

i1:

!li

i

~"~""~ II ii

i ,

i---i ............................

i,

ii ::

i! :i

sep!osno!,~a

..... iiiiill ---, i!!

. . . . . . . .

!

, ,

-='''mmm

--

i ,, ,,

,

snxeld s!wJol!~lesnpeu~ V

9

I i

'

,

0

,

,

; i

i

,

,

, .

sep!osno!lopnesd 'O

:,

wnlou!~s!p a

',

i

I

wnep!oqLuoUJ "~

i

i

I:

,

,

,

i

;"

!

uJrljelR!llOOa

~J

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

eso~od H

!

i

I

II

E

~

-

1i,.ii]i IIILII I

,

!!pnoqo!w "V

I + ............................

~,

l

ewotnel "V

~"

I

0

sn~IOldO,~pi!p"V

wns~

'H

z

Ii .==

eLIiUeoe3eldS Z

_...ou. ees 4 ~ O N 3

se!JeS i

3N3OO3 I:13MO7 1

[

r

=l

I

I

i

I

I

1

l

I

I

3N::IOO3 7Vd kl3dclA I

I

I

1

l

I

3N::IOO3"IVd I:13MO7 I

I

I

I

I

;

I

.e

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY

& .~_ k= ; E

O w z L~

' 0~ o ,9~ o i ,r"- , North SeaZonation ~ N IO N I _ ~ o i~:o J Sub9~ i~ ~ a. i5 Zone zone i : ~NP22 ~ P1'8 not assigned NP21

,o,

.,

Z --

N--'-~" P17

~i

NP19' P16

-

Eeb (E8)

n~:

40

INP18

~_

P15

NP17

m i uJ Z uJ 0 0 uJ u.I

E8a

P14

H.porosa (E7)

q'~ A.tauloma

PI~ ~

"--'1NP16

E6c

P121 D.colligerum (E6) Z,,r LU NP15

] S.placacantha , ! Abundance (E5) , _~ .

,-I

~ D.ficusoides P10

(E4)

ILl Z

50

0 O LLI tr" LU

"

~,

LU Z ILl . s _

rr 13. >-

NP11

Z W-

rr LU

~ I--

NP7

-I ~

~L E3 i"

-

H.tubiferum Acrne ( E l )

P6A , P5

Aaugustum (P6)

'~ ' W.articulata brewcomuta (Web) D.pachydetmum (Dp) E.ursulae (Eu), D.pachydermum 9 ~' E.ursulae (Eua) 9 M, glaDra ~ ' C.columna (Cc) M.compressa (Mc), H.tenuispinosum 9 1 4(Hot), 9 M.g/abra 9 M furensis, A.medusettiformis 9 A,medusettitormis IB I I (Am), O.varielongitudum

'"'

A.margarita

D.sohdum

(Dv)

(Drc), Dpolitum O , D,slm#e, D, vanelongitudinum 0

E2b E2a

{Dso), D.stm#e 9

P6 ". ,

P4

P. pyrophorum . ----'-- I {P4) I NP5 P 3 B ' I. ?viborgense -------- i (P3) ' : P3A! S.magnilicus P2 ..... (P2)_ NP4 '

Z <~

65 65-

P6B'

(Dc), R.thomboideurn (Rr)

R d/sbnctum (Pd), P,powellii, S, placacantha (Spc) O

ElC: H.tubifemm 9 (Hi), T.hiatus 9 (Th') ' L E l c - ~ I b "~1 Elb: D.oebisfe/dens~s 9 (Do), G.ordinata 0 C.wardenense 9 (Cw) ~ L: Leiosphaendia spp.'* 9 C.dattmoonum, C.~oeciosum, R magnifJcum Pterosperme//a spp.'" 9 A.augustum (Aa) I i

(PS)

LU

'-,

E2c

' P7 ' A.medusetti. i (E2) 1 ,

NP9 ~ -NP8

Z LJJ L) O LLI " ~.1

(At), A.sentosa

D.coltigenJm

D.condylos

NP10

0. ::3 60-,___,._.~ "

E.ursulae (E3)

'

E3a

NP12

LU

=_J

P9

.,

O,pseudoficusoides (Dpf) EsbE6a iL'~' S.placacantha 9 (Spa) P clithndium (Pc), P. regaiis ESa '". . , ' O.ficusoides (D~), C.depressum, G.exuberans, T,delicata E4d .qm S.ptacacantha 9 9 (Sps), ,it=t C.magna (Cm), A.polypetellum A. Wttatum G,ordinata E4c ,,r= H.tubifetum, H, oostae

E4b E4a E3d E3c

L P8

~

O ..J

55-

:

NP131

frequent abundant superabundant pollen taxon algal taxon

w'~ H,dau,senii

,

NP14 9

common

E6b

I Pll

o

E7a

I

45-

, H.porosa (Hp), R.potosum R.botussca

E7b

key consistent top

9 9 9 9 9 I * 9*

(Ami)

, A miclloudii

'

,~ 9

,=

A.~k~yOptokUS (Ad)

A.diktyoplokus

_

13_ ira

Dinocyst Zonal Markers (major bioevents s h o w n in brackets)

A,marganta

(Aim), A,gippingens~s (Ag), L.obscurum, S.momboideus

P5 A.gippingensis 9

(Age)

P4

'* , R pyropho~m (Pp), P australinum ~ I~ pyrophorum 9 (Ppa) ' - ' L ?viDorgense (Iv), C.saeptum, E annetorpense

P3

~

P2

. , S. magniflcus (Sin) m A.reticulata (Ar), S.magnilicus 0 ,, Sinomata (Si)

Tcl. deticata

(TO)

P1C NP3,

' '

S.inornata (P1)

NP2

P1B

NP1

P1A

i

i

C.comuta

Fig. 3. North Sea Paleocene-Eocene dinocyst zonation and zonal markers. Absolute time scale after Haq et al. (1987).

97

98

D. C. MUDGE t~L J. P. BUJAK

diachronous across the basin. This is reflected by the low number of microfaunal bioevents available for basinwide correlation in the Middle and Upper Eocene. Sixteen microfaunal bioevents with regional North Sea chronostratigraphic significance can thus be delineated in the Paleocene-Eocene, based on observations from numerous wells, and are integrated with the dinocyst bioevents to establish a basinwide chronostratigraphy (Fig. 4). The correla-

I= 35 _ , ~

to~ o~'

WE

"= = o

o ~

~ not r,.- assigned z_~

'

Major Micro/aunal Events

Ad

key

E8b

" ~ _~

ESa

~

E7b

~ O 9 9 9

F Ami Hp

4C

s

E7a E6c

~Z

~ "~

z

u~ ..~

E6a

Dpf Spa Pc Of

50

,,,~
E4t~ E4a E3d E3c E3b E3a 9

E2c E2b

o_ >"

55

Spiroplectammina

aff.spectabilis (Sas)

Cenodiscus spp. 9 (Cda)

ESb E5a E4d E4c

common frequent abundant pollen bioevent

At Dc Rr ~. Pd Spc ~

E6b 45

consistent top

~ Sps - Cm Wab Dp Eu Eua Cc ~ Hot MC

Cenosphaeraspp. O (Cs) Spiroplectammlna navarroana (Sn)

Am Dv -,.-, Subbotina finaperta (SI) ~r~ Subbotina linaperta 9 (Sla)

Drc

DSo

E2a E1 E c ~ Ib .~ HtTh'Do~ Coscinodiscus spp.1 & 2 (C1) L~'#la Cw ~r~ Coscinodiscus spp. 2, 4 & 7 9 (D) ' Aa Impoverished agglutinate assemblage (IA) P6

0=,


.".5, ,,=, ~

AImAg ,,~ Diverse agglutinate assemDlage (DA)

'

P5

P3

P2

-Ag~, ,:,p Iv - Td

~ Cenodiscussp.t (Cdl) q'~ Cenosphaera lenticularis (CI)

Sm - Ar Si

~

,r~ Globtgerma simplicissimum (Gs)

~9 uJ

O,

Globigerina pseudobulloides (Gp) Globigerina cf.compressa (Gc) Globlgerina trivialis (Gt)

- Cac

Fig. 4. North Sea Paleocene-Eocene microfaunal correlation.

tion with standard NP and P zones, and hence the absolute timescale shown on Figs 2-4 is primarily based on dinocyst biostratigraphy for the postMaureen section, and planktonic foraminifera in the Ekofisk and lower part of the Maureen sequence.

Sequence stratigraphic analysis The stratigraphic framework for the Paleocene and Eocene rocks of the North Sea Basin is based upon the identification of nine sequences from electric log data and their calibration with biostratigraphic data from key wells across the region. Individual stratigraphic sequences display characteristic gamma, sonic and resistivity log patterns that reflect waxing and waning of sedimentation in the basin and on the basin margins. Each sequence is bounded by a high-gamma log peak that is often accompanied by a low velocity sonic trough, or occasionally a sonic spike, the latter indicative of limestone. The sequence boundaries are interpreted as surfaces of clastic starvation and are marked by a thin condensed mudstone section recording the effect of greatly reduced sediment input to the basin. Mudge & Bujak (1994) defined five Eocene stratigraphic sequences, naming them after the major rock or reservoir units that they contain. Milton et al. (1990) recognized seven Paleocene sequences, defining them as depositional units bounded by maximum flooding surfaces. Morton et al. (1993) tied these sequences to a succession of 23 bioevents in a study of this section in the northern North Sea. These Paleocene sequences and their definition have been adopted here, but renamed Ekofisk, Maureen, Lista and Forties, with three subsequences recognized in the Lista (Fig. 5). Redefinition of the lithostratigraphy by Mudge & Copestake (1992a, b) and Mudge & Bujak (1994) within this sequence stratigraphic framework gives confidence to the use of rock unit names, rather than alpha-numeric schemes of stratigraphic nomenclature. Typical log patterns are illustrated in Fig. 6. In sand-poor, more distal, well sections, e.g. in the Central Graben, individual sequences display bowshaped gamma log patterns, often mirrored by the sonic and resistivity responses. Basinal sands present in the section interrupt but do not destroy these log patterns, as seen in the Forties sequence. In sand-rich, more proximal wells, e.g. in the Moray Firth Basin, these bow-shaped log patterns bounded by high-gamma log peaks can still be discerned, representing basinal sand deposition. However, the Dornoch, Balder and Grid sequences contain asymmetrical gamma log patterns corn-

99

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY ~"

~~ =

I=

:~m

~E o> _=o ~,i~

o

=o

~ =~

Stratigraphic Sequence MCS! UC Mudge& Bujak(1994) & ThisPaper ~'~

Ad ~

~

u~ ~

Ami

E

E8a

~

E7b

= 40

..~ m

E8b

- -

Hp

E7a uJ z ILl

I

At Dc Rr Pd

E6c

GRID

- 3'

NW

\Spc_sas

-T

,

R

-Cda Z

E6a

~

ESa E4d

Pc Df Sps

,~ z O

O,

55

tu z ~

Dp III Eu - Cs Eua Sn Cc - y4 Hot Mc II

Ez= -

-3'--t

I

O~o

E2a H1EC_~Ib .. H t T h - C 1 - Do L~,Eia CWAa - D -IA P6

P5

FRIGG

Am D v - S l Drc - S l a

< ~

Sonic

II

E2b >"

GR

Wab

~.

,,:5, a. -

Sonic

"~ - - ~ 1

Cm

E4b ~" E4a E3d E3c E3b E3a

~z

GR

CENTRAL GRABEN

ALBA

'sP

E4C

50

III

Dpl

SE MORAY FIRTH BASIN

E6b tu

sequences can be divided into subsequences separated by subordinate high-gamma log peaks. These stratigraphic sequences broadly correspond to the genetic stratigraphic sequence of Galloway (1989), which was defined as a package of sediments recording a significant episode of basin margin outbuilding and basin filling bounded by periods of widespread basin margin flooding. The bow-shaped log patterns reflect cycles of sediment progradation, aggradation and retrogradation, both within the basin and on its margins. However, the Galloway sequences were identified in Tertiary

-T

BALDER 9 -T-'Y~-DORNOCH - Z 8 Z -T-F O R T I E S ____7__

- AImAg-DA -Y III _ Aga -'r-~

<

II

Pp pa

g

,v

~-Cdl "'Y-~ I - CI

P3

- Td

P2

Gp Sm --GtGc - Ar Si

tu z LU L)

6 LISTA

4+5 3

-T.

MAUREEN

2

EKOFISK

1

- Gs

g~ ~<

65 Q

m

-

Cac

Fig. 5. Integration of Paleocene-Eocene stratigraphic sequences and bioevents. tion

prising a series of upward-decreasing and upwardincreasing cycles. These represent parasequences deposited in a shelf environment (Van Wagoner et al. 1990). Several of the Paleocene and Eocene

Fig. 6. Identification of Paleocene and Eocene stratigraphic sequences from electric log patterns.

100

D.C.

M U D G E & J. P. B U J A K

basin margin sections on the US Gulf Coast, where they were defined as being eustatic in origin, caused by episodes of rapid marine transgression that cut off fluvially-derived clastic input into the basin by raising base level. In the North Sea region, where periodic uplift of the Scotland-Shetland hinterland during the Paleogene contributed large volumes of coarse clastics to the basin, tectonism and sediment supply were equally important influences on sequence development. Thus, the definitions of the stratigraphic sequence and its bounding surfaces are based simply upon empirical interpretation of well and biostratigraphic data and carry no genetic connotation. The bounding surfaces are defined as surfaces of maximum condensation and may have a number of origins including marine transgression, changes in sediment supply to the basin and intrabasinal variations in tectonic uplift and subsidence. The fundamental criteria for their use in mapping and correlation are their regional extent and synchroneity within the North Sea Basin. Biostratigraphy demonstrates that individual sequence boundaries are consistently bracketed by the same dinocyst bioevents, giving a high degree of confidence that these surfaces represent time lines. The sequences can be mapped over a large part of the North Sea, persisting in both shelf and basinal settings. Figure 5 integrates the stratigraphic sequences with dinocyst and microfaunal bioevents, demonstrating the degree of precision available for stratigraphic analysis. Sequence and bioevent tops for a selection of North Sea wells are listed in Tables 1 & 2. There is no evidence to indicate that the highgamma mudstones coincide with increased abundances or diversity of microfossils. Nor does there appear to be any relationship between microfossil last occurrences or last acme occurrences and the majority of these maximum condensation surfaces. The important exceptions to this latter observation are the surfaces that bound the shortduration sequences spanning the PaleoceneEocene boundary. In this part of the section, highgamma mudstones are associated with assemblage changes, designated as bioevents, that reflect water circulation and temperature changes affecting the North Sea Basin. Lithostratigraphic

analysis

The final phase of stratigraphic analysis has involved integration of the sequence stratigraphic framework and bioevents with the lithostratigraphy (Fig. 7). The Paleocene-Eocene succession is represented by the Montrose, Moray and Stronsay Groups, together with the uppermost part of the Chalk Group. The lithostratigraphy of the

~

Lithostratigraphy

.-c8~ Stratigraphic "~ Sequence

CNS

OMF

NNS

35 _j~-Gs -Ad

=.

,8,

,

~ -Ami

UPPER MOUSA FM

GRID

-Hp

UPPER MOUSA FM

-At I .DcRr - Pd XSpc III .Dpf -Spa -Pc T

HORDA FM

ALBA

03 Z 0 I--

MIDDLE MOUSA FM

-Crn

HORDA

-War:

-Eua

:~*I" FRJGG

ray Sst Mbr

[--

FM

--

Frigg, Sst M br

LOWER MOUSA FM

-~,t~ -- Drc -Dso I

-~

:DORNOCH. __LCM_L~ELE.DORNOCHFM_I. D(~tNOCHFM N FORTIES Forliesa IFM I F0rl~ M SELEFM U' LISTA FM

LISTA 9 FM

Ill -AIm~ LISTA -Aga _p~,p, iT -

3eJmom=Mbrl I A.d,e. Mb, jI

BalmoralMbr

a.

Heimdal, ibr

- -

(5 ~n O z

- rd ~-

0

MAUREEN MAUREEN - ~" FM - si

MAUREEN FM

MAUREEN FM

EKOFISK FM

EKOFISK FM

EKOFISK FM

n EKOFISK -Caci

<"1O

Fig. 7. North Sea Paleocene-Eocene lithostratigraphy integrated with stratigraphic sequences and bioevents.

Paleocene-lower Eocene Montrose and Moray Groups has been revised by Mudge & Copestake (1992a, b), who defined six regionally mappable units; the Maureen, Lista, Sele, Balder, Dornoch and Beauly formations. These are argillaceous units containing a number of basinal sandstone members of less widespread distribution. The Dornoch and Beauly Formations are genetically distinct shelf sandstone units that pass basinwards into the age-equivalent Sele and Balder Formations, respectively. Their scheme retains much of the original Deegan & Scull (1977) nomenclature, but substantially modifies the hierarchical structure of the constituent units. The revision used bioevents and high-gamma log markers to produce a framework

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY within which the lithostratigraphic units could be confidently defined. Much of this revision has been incorporated into the BGS/UKOOA publication on PaleoceneEocene lithostratigraphic nomenclature by Knox & Holloway (1992). However, Knox & Holloway (1992) have redefined the Beauly as a :member of the Dornoch Formation, in order to overcome the problem of lithostratigraphic assignment at formation level in wells where the Beauly lignites are poorly represented or absent. They thus include all Moray Group deltaic/shelf sediments in the Dornoch Formation. Mudge & Bujak (1994) assigned the Dornoch and Beauly Formations to different sequences, showing that each had its own distinctive pattern of sediment distribution. In many wells a lignite bed occurs at the base of the Beauly Formation, allowing separation of these units. Where this lignite is missing, a persistent high-gamma mudstone is present. Gamma log patterns may also help in distinguishing between these two units; the Dornoch section is dominated by upward-coarsening log traces, whilst those in the Beauly are much more variable, and include upward-fining patterns. Knox & Holloway (1992) also only recognized two formal members in the Lista Formation, the Heimdal Sandstone and the Mey Sandstone. The latter unit comprises the Andrew, Glamis and Balmoral Members of Mudge & Copestake (1992a). The Eocene Stronsay Group was defined by Knox & Holloway (1992) to include the sandrich Mousa Formation and argillaceous Horda Formation. They restricted the Mousa Formation to the western part of the basin, where a series of shelf sandstones are present. This formation passes eastwards and southwards into the Horda Formation, which comprises basinal mudstones with subordinate sandstones, including the Skroo Sandstone, Frigg Sandstone and Tay Sandstone Members and the informal Caran Sandstone (Knox & Holloway 1992). Mudge & Bujak (1994), in their paper on Eocene sequence stratigraphy, recognized three distinct sandstone units within the Mousa Formation, which they separated out as the lower, middle and upper Mousa Formation. The upper Mousa Formation can be correlated with the Grid Formation of Isaksen & Tonstad (1989) and includes the upper Grid Sandstone Member (Brodie Sandstone) of Knox & Holloway (1992).

101

and is of Lower Paleocene (Danian) age. The contact with the Upper Cretaceous is believed to be disconformable (Knox & Holloway 1992). Seismic data indicate that this surface forms a regional reflector traceable over much of the North Sea Basin. A fundamental shift in microfossil assemblages is also seen across the Cretaceous-Tertiary boundary. Seismic data show that the upper boundary commonly forms a pronounced unconformity against which the overlying Maureen sequence downlaps (Stewart 1987). Erosional channelling of this surface is seen in the Moray Firth Basin. Units of reworked chalk are common in the sequence in southern parts of the Central Graben (Kennedy 1987). The sequence comprises chalky limestone with a low-gamma log signature, interbedded locally with argillaceous limestones. The sequence contains poorly preserved planktonic foraminifera and calcareous nannoplankton, with dinocysts only rarely present. Some nannoplankton bioevents can be used for local correlation within the sequence, but do not appear to be regionally extensive across the North Sea Basin. In a number of wells (16/8-1, 16/17-8A and 21/11-2), the sequence is characterized by the dinocyst S. inornata. The top occurrence of this species (bioevent Si) is associated with the planktonic foraminiferal bioevent Gs in the upper part of the sequence where it may be related to the upward change in facies from chalk to clastic sediments. The Ekofisk sequence is present over most of the central North Sea, Moray Firth Basin and South Viking Graben. In these areas it typically has a thickness of 50-60 m, thinning northwestwards on to the East Shetland Platform and southwestwards on to the Forties Salt Platform. It is absent over these platform areas as well as on the Halibut Horst and parts of the Fladen Ground Spur and Jaeren High. Its distribution shows evidence of deep structural control with depocentres present in the South Viking Graben and Central Graben, where it thickens to > 100 m. The sequence is absent north of latitude 5 9 ~ due to non-deposition. However, thin marls at the base of the Maureen sequence in the East Shetland Basin may represent a thin Ekofisk section (Knox & Holloway 1992).

Maureen Sequence The Maureen sequence (sequence 2 of Morton

et al. 1993) corresponds closely to the Maureen

Paleocene-Eocene sequence stratigraphy (Figs 8 & 9) Ekofisk Sequence The Ekofisk sequence (sequence 1 of Morton et al. 1993) corresponds closely to the Ekofisk Formation

Formation and is of early Paleocene (Danian) to late Paleocene (earliest Thanetian) age. Where the Maureen and Lista sequences are sandy, the top Maureen maximum condensation surface occurs near the base of a relatively high gamma mudstone section which typically has a thickness of 10-20 m. This mudstone separates limestones, chalks and

102

D.C. MUDGE & J. P. BUJAK

Fig. 8. Eocene well correlation, Beryl Embayment to South Viking Graben.

Fig. 9. Paleocene well correlation, Moray Firth Basin to Fisher Bank Basin.

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY sandstones of the Maureen Formation from Andrew sandstones in the Moray Firth Basin and parts of the Central Graben, and from Heimdal sandstones in the South Viking Graben. The Maureen sequence comprises a heterogeneous mix of mudstone, marl, sandstone and reworked chalk, which gives rise to very irregular log patterns. Maureen sediment distribution clearly shows the pervasive influence of deep structural control with the development of a series of depocentres within the relict Mesozoic graben system (Fig. 10). The sequence is absent across the East Shetland Platform, Halibut Horst, Jaeren High and Forties Salt Platform. Away from these areas, two sandrich depositional fairways can be recognized, one spreading southeastwards through the Moray Firth

103

Basin and the other occupying the South Viking Graben and Beryl Embayment. These sediment systems are separated by a relatively thin zone of reworked chalk deposits extending southwards from the Fladen Ground Spur to the FortiesMontrose High. Within the Witch Ground Graben in Quadrant 15, the sequence reaches a thickness of 240 m, thinning northwards to zero over the East Shetland Platform. In the South Viking Graben up to 150 m of sediment is present in depocentres close to the faulted East Shetland Platform margin. However, the thickest development of Maureen sediments is seen in the western Beryl Embayment in Quadrant 9, where up to 600 m of thickly bedded sandstones were deposited close to the platform margin fault.

Fig. 10. Facies distribution Lista and Maureen stratigraphic sequences. Light stipple, sand; heavy stipple, sand thicks; grey shading, mudstone; limetone shading, Maureen chalk facies.

104

D.C. MUDGE • J. P. BUJAK

These sandstones thin out rapidly westwards on to the platform and pass eastwards into a sand sheet 100-150 m thick that occupies the Viking Graben system. Local mounded sandstones occur within this sand sheet; in the Frigg field area these reach 330 m in thickness. The limits of the Maureen sand systems in the North Viking Graben and Central Graben are marked by passage into mudstones 30-50 m in thickness. Thin sands continue as far south as Quadrant 30 in the Central Graben. Increased clastic input and decreased carbonate production within the Maureen sequence resulted in a major reduction in numbers of planktonic foraminifera and nannoplankton, although the former provide some biomarkers (bioevents Gc, Gt and Gp) in the lower, transitional part of the sequence. Non-calcareous agglutinated benthic foraminifera become abundant but are chronostratigraphically undiagnostic, whereas dinocysts may be common and provide several biomarkers (bioevents Ar, Sm and Td). Of these markers, the top occurrence of S. magnificus is the most consistently identified, this species characterizing the lower part of the Maureen sequence. The dinocyst L ?viborgense is another typical Maureen species. Although its top occurrence (bioevent Iv) is usually logged within the upper part of the Maureen sequence, its true extinction datum is believed to lie within the lowest Lista section (wells 15/28b-4 and 22/5a-lA). This species is very small in size and relatively uncommon, making its top occurrence difficult to locate. Radiolaria become common in the upper part of the Maureen sequence and provide a biomarker (bioevent C1) near the top of the sequence.

Lista Sequence The limits of the Lista sequence (sequences 3-6 of Morton et al. 1993) are more or less coincident with those of the Lista Formation. The sequence is of late Paleocene (Thanetian) age. Sandstones are widely developed within the Moray Firth Basin, East Shetland Basin, Viking Graben and Central Graben. These are the Andrew, Balmoral and Heimdal Members (Mudge & Copestake 1992a, b). In the Moray Firth Basin, where the sequence reaches its maximum development, the Andrew and Balmoral sandstones are separated by tuffaceous sandstones of the Glamis Member. In this area a number of subordinate maximum condensation surfaces can be identified, enabling the Lista to be divided into as many as five subsequences. These are referred to as Lista Ia, Lista Ib, Lista II, Lista Ilia and Lista IIIb, following Knox & Holloway (1992), who subdivided the Lista Formation into three informal units based on

recognition of a number of dinocyst biomarkers. Lista I corresponds to sequence 3 of Morton et al. (1993), Lista II to sequences 4 and 5, and Lista III to sequence 6 (Fig. 5). Where tuffaceous sediments are absent, bioevents indicate that the equivalent sandstone section should be included within the upper part of the Andrew Member. In the Central Graben, biostratigraphy indicates that Balmoral sandstones are more widely distributed than those of the Andrew. Only the upper part of the Andrew is present and this includes the lateral equivalent of the Glamis Member. Within the Viking Graben system, Heimdal sandstones extend throughout Lista sequence. The lower and middle parts of the Lista sequence contain abundant dinocysts, with a succession of biomarkers delimiting each of the Lista subsequences (bioevents Iv, Ppa, Pp, Aga, Aim and Ag - see Tables 1 & 2). These bioevents are also observed in Denmark (Heilmann-Clausen 1985) and southeast England (Powell et al. 1995), so that detailed biostratigraphic correlations can be made between these areas. Most bioevents also occur on the Scotian Shelf and Grand Banks of Canada, indicating marine connection with the Atlantic (Williams & Bujak 1977). However, foraminiferal assemblages continued to be dominated by agglutinated benthic forms, which became restricted to a few species in the uppermost part of the Lista above bioevent DA. This section also contains a low diversity dinocyst assemblage of Spiniferites, Areoligera and Hystrichosphaeridium species. Paleocene-Eocene sand deposition reached its maximum during Lista times. Sediment distribution patterns closely resemble those of the older Maureen sequence and were intimately influenced by the deep structural elements (Fig. 10). The two major zones of sand development through the Moray Firth Basin and Viking Graben are separated by the structural high of the Fladen Ground Spur over which sands are thin or absent. The E-W Halibut Horst also split the Moray Firth sand system into two parts, a northern fairway extending through the Witch Ground Graben and a southern one through the Buchan Graben. The Lista sequence reaches thicknesses of > 600 m in parts of the relict graben system. It typically thins to < 200 m over their margins, where it is represented mainly by mudstones. An isolated depocentre is sited in the East Shetland Basin in Quadrant 2, where up to 975 m of sandy sediments occur in the hanging wall of the East Shetland Platform margin fault. The sequence thins eastwards into the Norwegian sector and southwards into the central North Sea. In these areas it is typically < 100 m thick, thinning progressively on to the Mid North Sea High.

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY

Forties sequence The Forties sequence (sequence 7 of Morton et al. 1993) is of late Paleocene (latest Thanetian) age. It only has significant thickness in areas where it contains sandstones. It reaches its maximum thickness where these are well developed, as in the Buchan Graben and northern part of the Central Graben (Fig. 11). These sandstones are referred to the Forties Member of Mudge & Copestake (1992a) and occur within the lower part of the Sele Formation. Elsewhere in the North Sea Basin sandstones are only locally present, e.g. in the South Viking Graben (Skadan Sandstone Member of Knox & Holloway 1992) and in the North Viking

Fig. 11. Facies distribution Forties stratigraphic sequence. Light stipple, sand; heavy stipple, sand thicks; grey shading, condensed mudstone.

105

Graben (Teal Sandstone Member of Knox & Holloway 1992). Where sandstones are absent the Forties sequence often persists as a condensed unit of mudstone, calcareous sandstone or limestone, typically < 10 m in thickness. This displays a characteristic gamma-sonic log bow that separates the lower and upper high-gamma peaks. The boundary between the Lista and Forties sequences coincides with the last occurrence of North Sea Paleocene agglutinated foraminifera (bioevent IA) and the first occurrence of Apectodinium augustum. These events may be related to 'rapid global warming and oceanographic changes that caused one of the largest deepsea benthic extinctions of the past 90 million years' (Kennett & Stott 1991). This warming event occurred c. 0.5 million years before the PaleoceneEocene boundary and is associated with increasingly abundant kaolinite, suggesting high weathering rates that resulted from an unusually warm climate with high precipitation, plus a large influx of organic nutrients to sediments (Thomas & Shackleton 1996). This would explain the predominance of pyritized sapropelic kerogen and diatoms, and the increase in subtropical pollen. Although the North Sea Basin was largely enclosed during this period, warm-water Apectodinium dinocysts were able to migrate into the basin from the Tethys, probably via an eastern seaway, to predominate during Forties deposition. The consistent termination of these assemblages at the top of the Forties sequence (bioevent Aa see Tables 1 & 2) probably resulted from cooling of the surface waters. Several subzones based on pollen have also been suggested by Schr0der (1992), reflecting changing plant communities that were partially controlled by climatic variations. The Forties sequence has a thickness of > 250 m in the West Central Graben. The sand fairway is constrained to the west and east by rapid thinning on to the margins of the Forties Salt Platform and Jaeren High. The sandstones also thin southeastwards, pinching out in Quadrants 29 and 30. Thick sandstones are present in the Gannet field area on the western flank of the Central Graben in southern Quadrant 21. In the Moray Firth Basin the Forties sandstones are located in a linear zone along the structural trend of the Witch Ground Graben. Similar zones in Quadrant 20 parallel the SW-NE trend of the Buchan Graben and Buchan Ridge. To the north of the main Forties sand fairway, local depocentres are present in the Viking Graben and Beryl Embayment.

Dornoch Sequence The Dornoch sequence (sequence 8 of Morton et al. 1993) contains the Dornoch Formation as defined

106

D . C . MUDGE & J. P. BUJAK

by Mudge & Copestake (1992a, b). It also contains the upper part of the Sele Formation with its locally distributed, sand-rich Hermod, Flugga Sandstone and Cromarty Members. The sequence is of early Eocene (early Ypresian) age. In the Moray Firth Basin and East Shetland Basin, the Dornoch sequence displays a number of upward-coarsening sand cycles separated by high-gamma mudstones. These cycles are often strongly asymmetrical with the upward-fining part of the log pattern thin or missing. These cycles can be tied to prograding clinoforms on seismic data (Milton et al. 1990) and equate to the parasequences of Van Wagoner et al. (1990). Dornoch sandstones may rest unconformably on Lista mudstones or pass down gradationally into siltstone and silty mudstone. The uppermost sandstone unit is succeeded abruptly by the high-gamma mudstone marking the boundary with the Balder sequence. East of the pinchout of Dornoch sandstones, the sequence comprises dark grey laminated silty mudstones with thin limestones, sandstones and tufts. Hermod, Flugga and Cromarty sandstones, with typical blocky log patterns and sharp tops and bases, are locally developed within the sequence. The Dornoch sequence reaches its maximum thickness across the Moray Firth Basin, East Shetland Platform, Beryl Embayment and East Shetland Basin, where it is represented by sandstones (Fig. 12). In these areas it is generally 90300 m thick, reaching 400 m in a series of local depocentres in the Beryl Embayment. The western erosional limit of the sequence can be mapped seismically across the East Shetland Platform southwards into Quadrant 14 (Milton et al. 1990). The sequence thins eastwards and southwards to < 50 m in the South Viking Graben and Central Graben, where it becomes mud prone. In the northern North Sea, the sequence generally has a thickness of up to 30 m and is absent or highly condensed on the Tampen Spur. However, a major sand-rich depocentre is present in the Frigg field area in the North Viking Graben, where thick Balder and Frigg sandstones are also present. The Dornoch sequence represents a very short depositional episode, possibly as little as 0.2 million years in duration, during which widespread, shallow marine shelf sand deposition took place along the western margin of the North Sea Basin over a distance of > 400 km. The top Dornoch surface is deeply incised on regional seismic data in the Beryl Embayment and on the East Shetland Platform, indicating extensive erosion. Dornoch sand deposition was preceded by the easterly outbuilding of a wedge of muds and silts, which downlaps the maximum condensation surface marking the base of the Dornoch sequence.

Considerable thicknesses of this sediment filled bathymetric lows in the Beryl Embayment and Moray Firth Basin. To the east of the pinchout of the Dornoch shelf sands, a series of muddy basins developed along the axis of the old Mesozoic graben system. In places, the Dornoch sand limit may have been localized by a slope break above the graben-bounding faults. Sand build-ups can often be mapped close to this limit. The basins contain areas of thick sands, whose location can be related to deep faulting and the position of the Dornoch sand limit. The sands in the South Viking Graben and Buchan Graben are believed to have originated by catastrophic submarine slope failure in areas of rapid sand build up, with the sands sliding down into bathymetric lows controlled by palaeotopography. These processes also occurred on the East Shetland Platform, where no deep basement control can be invoked, indicating very rapid deposition and dispersal of Dornoch sands. The complex distribution and chaotic nature of Dornoch sands in the Beryl Embayment also indicates some degree of slope failure and gravity transport to the east of the platform margin fault. Here however, the sands have not become detached from their source sediments. The Dornoch sequence contains a few dinocyst taxa that were able to tolerate the restricted marine environments that characterized the North Sea Basin at this time. These are dominated by Cerodinium wardenense, whose acme is believed to be associated with the prominent high-gamma mudstone marking the top of the sequence (bioevent Cw). This species continues in declining numbers into the overlying Balder sequence (see Tables 1 & 2). Its final extinction may have been caused by decreased water temperature. Proximal marine to paralic environments were characterized by the development of the algal Leiosphaeridia biofacies (bioevent L). The microfaunal assemblages continued to be dominated by pyritized diatoms, which were reduced in diversity within the Dornoch sequence (bioevent D). The basin probably became landlocked at times, giving rise to anoxic bottom conditions that persisted until the end of the Balder, when the re-establishment of fully marine conditions took place. Pulses of the early Ypresian marine transgression may have temporarily introduced more marine waters into the basin and led to the development of subordinate flooding surfaces in the shelf areas where Dornoch sands were being deposited. Balder Sequence

The Balder sequence (Mudge & Bujak 1994; sequence 9 of Morton et al. 1993) contains the

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY

107

Fig. 12. Facies distribution Balder and Dornoch stratigraphic sequences. Light stipple, sand; heavy stipple, sand thicks; grey shading, tuffaceous or silty mudstones.

argillaceous Balder Formation and the sand-rich Beauly Formation as defined by Mudge & Copestake (1992a, b). The sequence is of early Eocene (early Ypresian) age. Thick sandstones belonging to the Odin Member are locally developed within the Balder Formation in parts of the Viking Graben and Beryl Embayment. The lower sequence boundary lies within the topmost part of the Sele Formation. The Balder sequence is dominated by tuffaceous mudstones and typically displays an asymmetrical gamma-sonic log profile. This reflects the abrupt incoming of tufts in the basal part of the sequence followed by a slow

decrease in the number of ash layers upwards. The upper part of the sequence comprises high-gamma mudstones terminated at the sequence boundary by a maximum condensation surface. The sonic and resistivity log profiles often closely follow that of the gamma log, giving rise to a characteristic log bow. Where Odin sandstones are present in the sequence, these log patterns are lost; however, the bounding high-gamma log peaks are still preserved, enabling separation of Odin sandstones from those of the Frigg or Dornoch sequences. In the Moray Firth Basin, and on the East Shetland Platform, the sequence contains Beauly

108

D . C . MUDGE ~: J. P. BUJAK

sandstones, which show a contrasting log pattern of upward-fining and -coarsening cycles. Upwardfining patterns are dominant and lignites may also be present at the top and base of the Beauly. The top Beauly sandstone may show an upward gradation into tuffaceous mudstones terminated by the top Balder high-gamma log peak. The dinocyst and microfaunal assemblages in the Balder sequence represent low-diversity assemblages within a restricted North Sea Basin. The dinocyst populations were terminated at the end of Balder times at the last abundance of Deflandrea oebisfeldensis (bioevent Do), with nearshore marine environments being dominated by Hystrichosphaeridium tubiferum (Subzone Elc, bioevent Ht). The last Taxodiaceaepollenites pollen also occur in abundance at the end of the Balder (bioevent Th). Diatoms continued to dominate the microfaunas until the end of the Balder (bioevent C 1), so that the top of the sequence is marked by several concurrent dinocyst, pollen and microfaunal last occurrences. These events are probably located within the high-gamma mudstone marking the sequence boundary. The Balder sequence has a thickness of 30-60 m in the Central Graben, Viking Graben and East Shetland Basin, where it is present in argillaceous facies (Fig. 12). The sequence thickens westwards on to the East Shetland Platform and into the Moray Firth Basin and Beryl Embayment, locally reaching > 270 m. The sequence onlaps older Paleocene sediments further west (Milton et al. 1990). A similar disposition of lithofacies can be mapped for both the Dornoch and Balder sequences, suggesting a continuation of the same tectono-sedimentary regime. Balder deposition also took place during a very short time period of c. 0.7 million years. The top Dornoch maximum condensation surface is marked by lignites in the Moray Firth Basin, suggesting that it was caused by eustatic sealevel rise associated with a second stage of the Ypresian transgression. Beauly shelf sandstones have a similar, but less widespread, distribution to that of the deeper Dornoch sandstones, pinching out eastwards into mudstones across the East Shetland Platform and Moray Firth Basin. Thick Odin sandstones occur within a narrow zone to the east of this pinchout line in the Beryl Embayment, where they are preserved as massive mounded bodies, and are believed to have originated by the same processes of slope failure and sliding that occurred during Dornoch times. In this area the sands are confined to a structural low between the shelf sand front and the Crawford Spur. Sediment flows may have been generated by failure of this sediment front, which would have rested on an unstable slope of soft muds and silts. The mounded, steep-sided morphology of the Odin sands may

result from dissection of sand-rich flows by later mud slides. Evidence of sediment disruption is seen on seismic data, which display abundant sedimentary faults that detach at the base of the Balder sequence. Odin sandstones also occur in the South Viking Graben close to the graben margin fault, and in the North Viking Graben in the Frigg field area. The widespread maximum condensation surface marking the top of the Balder sequence is believed to result from the final phase of the Ypresian marine transgression that drowned the shelf and hinterland and opened up the North Sea Basin to fully marine conditions.

Frigg Sequence The Frigg sequence (Mudge & Bujak 1994) is of early to mid Eocene (Ypresian to Lutetian) age. It includes the sand-rich lower Mousa Formation and the lower part of the Horda Formation, which contains the Frigg Sandstone, Skroo Sandstone and Tay Sandstone Members. Where the sequence is developed in mudstone facies, it is often possible to distinguish three subsequences separated by highgamma log peaks. In general there is an overall upward decrease in gamma ray response within the sequence. Frigg I, which consists of high-gamma mudstone, is often very condensed. Frigg II displays a lower gamma-ray response and there may be a marked step upwards into Frigg III, which has the lowest gamma-ray values. Where the sequence is thin, e.g. over the Forties-Montrose High, it is dominantly composed of high-gamma mudstone and the three subsequences cannot be distinguished. The presence of sandstones within the sequence in basinal areas interrupts but does not otherwise affect the bow-shaped gamma-sonic log patterns of the three subsequences. The areal distribution of Frigg lithofacies is very similar to that of the older Balder and Dornoch sequences, although deposition took place over a much longer time interval, and the volumes of sand emplaced were much less (Fig. 13). The sequence is generally 50-150 m thick on the East Shetland Platform, thickening to over 200 m in the Moray Firth Basin in Quadrant 15. In these areas the sequence contains lower Mousa shelf sandstones, which reach a maximum thickness of 180 m in the Beryl Embayment but are only 15-50 m thick in the Moray Firth Basin. They pinch out eastwards into Horda mudstone. The Frigg sequence thickens eastwards into the South Viking Graben where several depocentres are present in the hanging wall of the western boundary fault. These are sand prone in Quadrant 16, where Skroo sandstones can be mapped in a number of small fan deposits. In Quadrant 9, however, two depocentres are present containing over 400 m of mudstone. Further north,

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY

Fig. 13. Facies distribution Frigg stratigraphic sequence. Light stipple, sand; heavy stipple, sand thicks; grey shading, mudstone.

massive Frigg fan sandstones reach a thickness of 270 m in the Frigg field. The most northerly depocentre in this trend occurs on the western margin of the North Viking Graben in Quadrant 3, where up to 180 m of sandstones are present. The sequence is 50-100 m thick in the Central Graben, thinning to < 30 m over the Forties-Montrose High. Tay sandstones are present on the Forties Salt Platform and in the western Central Graben. Sand distribution is very variable because of the syndepositional effects of salt-induced highs (Armstrong et al. 1987). The Frigg sequence was marked by a long period (6.7 million years) of dominantly mud deposition. Coarse clastic input to the North Sea Basin reached its minimum for the Paleocene and Eocene, reflecting the high relative sea level at this time (Haq et al. 1987), allowing drowning of the western hinterland to take place. Sediment supply to the basin during a long period of relatively high sea level is likely to have been controlled by inter-

109

mittent tectonic uplift of the Scotland-Shetland landmass to the west. Basement influence over the siting of small basin-floor fans can be seen in the South Viking Graben. The persistence of thick sand deposition in the Frigg field area during the early Eocene and the presence of thick mud-prone depocentres in the Viking Graben are also indicative of local palaeotectonic control. The boundary between the Balder and Frigg sequences coincides with an abrupt change in the dinocyst and microfaunal assemblages reflecting a widepsread marine flooding event that correlated with the London Clay transgression in southern England (Powell et al. 1996). Typical Eocene dinocyst taxa were established as the basin deepened and southerly migration pathways opened, resulting in diverse late Ypresian assemblages including many taxa known outside the North Sea Basin. A minor late Ypresian regression is indicated by the dinocyst populations, with a succession of bioevents delimiting each of the Frigg subsequences. The most consistent Frigg bioevents are shown in Tables 1 & 2. The top of the Ypresian is marked by the top acme occurrence of E. ursulae, which is coincident with the top occurrence of the species in many wells. These two bioevents occur in the upper part of the Frigg (subsequence Frigg III). The lower part of the sequence contains abundant A. medusettiformis associated with various species of Dracodinium. The top acme occurrence of A. medusettiformis (bioevent Am) is located within the Frigg II subsequence. The microfaunal assemblages also changed abruptly at the base of the Frigg sequence, with diatom-dominated assemblages being replaced by foraminifera and locally common radiolaria. However, most of the microfaunal assemblage changes became diachronous as the basin deepened and a variety of coeval biotic niches were established. Thus, only a few microfaunal biomarkers can be used for regional North Sea correlation within the Frigg sequence (bioevents Sla, SI, Sn and Cs). Alba Sequence The Alba sequence (Mudge & Bujak 1994) is of mid-Eocene (Lutetian) age. It contains the middle part of the argillaceous Horda Formation and the sand-rich middle Mousa Formation. It also includes the Caran sandstone, which is an informally named unit within the Horda Formation in the Viking Graben and northern Central Graben (Knox & Holloway 1992). The sequence is dominated by argillaceous facies in the central North Sea, in the southern part of the Moray Firth Basin and along the Viking Graben. In these areas a threefold

110

D.C. MUDGE & J. P. BUJAK

division into Alba I, II and II subsequences may be distinguished. Alba mudstones have a lower gamma log signature than those of the underlying Frigg sequence; the mud-dominated facies shows very little variation in log character. Thick-bedded Caran sandstones with blocky log motifs and sharply defined tops and bases are present in parts of the South Viking Graben. On the East Shetland Platform, Middle Mousa sandstones are also thickly bedded and display similar log patterns. The two sandstone facies are difficult to separate along the platform margin in Quadrant 16. Dinocysts remained abundant and diverse during Alba deposition and each of the Alba subsequences is readily characterized by dinocyst biomarkers. These include bioevents Spa and Df (Tables 1 & 2). Systematophora placacantha is a characteristic species of both the Frigg and Alba sequences, displaying a succession of acme tops. The top occurrence of Diphyesficusoides (bioevent Df) is a consistent marker for the lower part of the Alba sequence. Microfaunas exhibited a variety of coeval populations indicative of local biotic niches

and provide few basinwide chronostratigraphic biomarkers (bioevents Cda and Sas). Compared to Frigg patterns of sedimentation, sand deposition was progressively more confined, being restricted to the East Shetland Platform and South Viking Graben (Fig. 14). Basement control of Eocene sedimentation reached a maximum during deposition of the Alba, with muddy depocentres containing 400-700 m of sediment present in the Witch Ground Graben and Viking Graben. The sequence thins on to the East Shetland Platform, Fladen Ground Spur and Utsira High, as well as southwards into the central North Sea. The Alba sequence reaches its maximum thickness in Quadrant 3 in the North Viking Graben, where a major muddy depocentre is developed in the hanging wall of the graben margin fault. In the South Viking Graben, smaller depocentres in a similar palaeotopographic setting are sand prone. These basinal Caran sandstones are thicker and more widespread than the older Skroo sandstones. They can be mapped southwards towards the Forties-Montrose High as a series of isolated sand-

Fig. 14. Facies distribution Alba and Grid stratigraphic sequences. Light stipple, sand; heavy stipple, sand thicks; grey shading, mudstone.

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY bodies. Middle Mousa shelf sandstones are present on the East Shetland Platform and in northern part of the Moray Firth Basin, where they have a thickness of 50-200 m, but the depocentre in the Witch Ground Graben contains 550 m of mudstone. The platform sandstones thin westwards as a result of erosion and shale out eastwards into Horda mudstones. The thick deposits of mud and sand in the Viking Graben are believed to have formed as a result of the same slope failure processes described for the older Eocene sequences. Middle Mousa shelf sands were moved downslope into the basin as part of a large submarine slide whose detachment was located to the west of the East Shetland Platform margin fault. These sands accumulated in bathymetric lows as basin-floor fans. Later mud flows cutting across sand-rich areas of seafloor may have isolated the sands, preserving them as discrete mounded bodies. The major sediment thick in Quadrant 3 may have resulted from a mud slide under similar circumstances of slope failure.

Grid Sequence The Grid sequence (Mudge & Bujak 1994) is of mid Eocene (Bartonian) to early Oligocene (Rupelian) age. In the Moray Firth Basin, Beryl Embayment and Viking Graben, the sequence is represented by arenaceous facies of the upper Mousa Formation. In the central North Sea and in the Norwegian sector, the sequence is argillaceous and contains the upper part of the Horda Formation. The informal Brodie sandstone of Knox & Holloway (1992) and the Grid Formation of Isaksen & Tonstad (1989) are included here within the upper Mousa Formation. In some wells the Grid sequence can be divided into two subsequences; these are often well defined in sandy sections, where thick-bedded sandstones are interrupted by a thin high-gamma mudstone. The upper Mousa sandstones are typically thickly bedded with upward-coarsening and -fining log cycles restricted to the Moray Firth Basin. The lower boundary of these sandstones displays a sharp contact with Alba mudstone; a high-gamma log peak is often present at this sequence boundary, and is well developed where both sequences are developed in argillaceous facies. The upper sequence boundary with Oligocene mudstones is also marked by a distinctive high-gamma log peak that is readily identifiable in most parts of the North Sea Basin. Dinocyst assemblages were diverse during Grid deposition and are very similar to those in the Hampshire Basin (Bujak et al. 1980). The most consistent biomarkers are the Dc, Hp and Ad bioevents (Tables 1 & 2). Both of the Grid subsequences can be delineated using these biomarkers,

111

but no microfaunal bioevents of proven chronostratigraphic significance have been established across the North Sea Basin. The top occurrence of A. diktyoplokus marks the top of the Eocene, and is present within the uppermost part of the Grid sequence in most wells. Occasionally it may be reworked into the overlying Oligocene section (wells 9/10b-l, 9/27-1 and 16/1-I). The Grid sequence had a duration of 7.7 million years, representing an episode of major sand input on to the North Sea Basin. The Grid sequence is thinner than the Alba, with a thickness not exceeding 360 m in the South Viking Graben and Beryl Embayment. However, the volume of coarse clastic influx was greater than that of the earlier Eocene sequences, rivalling that of the Paleocene Andrew-Heimdal sand system. The upper Mousa sands formed part of a single shelf sedimentary system, with basinal sands restricted to a small area of the North Viking Graben (Fig. 14). The sands form a broad envelope bounding the margin of the East Shetland Platform, limited by erosion to the west and by distal shale out to the east and south. The sandstones are generally 5 0 - 1 0 0 m thick, reaching more than 200 m in local depocentres in the hanging wall of the South Viking Graben boundary fault and in the Beryl Embayment, where the platform margin fault system still influenced sedimentation. They pinch out eastwards into mudstones along the margin of the Utsira High and southwards across the southern parts of the South Viking Graben and Moray Firth Basin. They are interpreted as the products of shallow marine deposition on a gently sloping shelf. The sequence thins eastwards on to the Utsira High and southwards into the Central Graben, where it is generally < 100 m thick and only 30 m thick over the Forties-Montrose High. It is represented by mudstone in these areas. The consistent position of the basal Grid sequence within the upper part of the E6b subzone and the persistence of a clastic starvation surface below the sands suggests geologically instantaneous initiation of sand deposition over the western part of the North Sea Basin. Shallow marine sands spread rapidly eastwards followed by aggradation and stacking of sands into extensive sheets. The length of the Grid depositional episode suggests that pulses of coarse clastic influx were followed by long periods of marine reworking on the shelf by storms and other processes. Sedimentation was terminated in early Oligocene times by the global Rupelian marine transgression (NP22 nannoplankton zone). This gave rise to a readily identifiable maximum condensation surface that caps the Grid sequence over the whole North Sea Basin. Where sands are present, the incoming of the transgression is marked by the sudden cut-off

112

D . C . MUDGE •

of sand, followed by a short, upward-fining mudstone section terminated by the high-gamma peak.

Application to exploration A high-resolution sequence stratigraphic framework based on the integration of electric log and biostratigraphic data has been erected for the Paleocene and Eocene succession of the North Sea Basin. This framework is based simply upon the recognition of log patterns; in particular, upon the identification of high-gamma peaks that can be correlated from well to well over the whole basin. The degree of confidence in using this scheme would however be low without the corresponding development of a detailed biostratigraphy that effectively constrains the stratigraphic position of the sequence boundaries. This precision is provided by a dinocyst zonation that has been integrated with bioevents based on other microfossil groups and correlated with onshore sections, permitting calibration with the standard nannoplankton and planktonic foraminiferal zones and hence the absolute timescale. The consistent relationship between sequence boundaries and bioevents indicates that these surfaces are effectively isochronous across the North Sea Basin. This time-stratigraphic scheme has immediate application in the exploration for Tertiary hydrocarbons, particularly for the more subtle plays lying close to or below seismic resolution. The remaining potential of individual plays can be assessed

J. P. BUJAK

by mapping reservoir and topseal packages at the sequence level on a regional scale, and subsequence packages at a more local sub-basin or sub-quadrant scale. The combination of reservoir and topseal maps with hydrocarbon charge distribution enables the construction of play fairway maps that highlight areas of low regional risk. This mapping is much more useful when underpinned by confident stratigraphy. Sequence mapping also provides a series of snapshots of the development of basin architecture through time, aiding the prediction of sand presence in areas of poor well data. At the block or prospect scale the stratigraphy can be tied to seismic data to locate areas with local stratigraphic potential. The availability of highresolution stratigraphy developed from a large basinwide well database enables confident recognition of the significant regional surfaces in individual wells and concentration on horizons with favourable play potential. The increasing use of 3D seismic data for exploration also requires precise well stratigraphy to gain the most information from this expensive technique. The authors wish to thank R. W. O'B. Knox and A. C. Morton for their constructive criticism and help in finalizing this manuscript. We are also indebted to the many oil companies that have provided well biostratigraphic and log data for this study, and to Amoco, BR Chevron, Conoco, Marathon, Phillips, Texaco, Total, Shell and Unocal for giving permission to publish data.

References ARMENTROUT, J. M., SHEPPARD, C. E., NAYLOR, P. H., MILES, A. W., DESMARAIS,R. E., MALECEK, S. J. & DESMARAIS,R. J. 1993. Log-motif analysis of Paleogene depositional systems tracts, Central and Northern North Sea; defined by sequence stratigraphic analysis. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 45-58. ARMSTRONG,L. A., TEN HAVE,A. & JOHNSON,H. D. 1987. The geology of the Gannet fields, central North Sea, UK sector. In: BROOKS,J. & GLENNIE,K. W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 533-548. BUJAK, J. P. & MUDGE, D. C. 1994. A high-resolution North Sea Eocene dinocyst zonation. Journal of the Geological Society, London, 151,449-462. , DOWNIE,C., EATON,G. L. & WmLIAMS,G. L. 1980. Dinoflagellate cysts and acritarchs from the Eocene of southern England. Special Papers in PalaeontoIogy, 24, 1-100. DEEGAN, C. E. & SCULL, B. J. 1977. A Standard Lithostratigraphic Nomenclature for the Central and Northern North Sea. Institute of Geological Sciences Report 77/25.

GALLOWAY,W. E. 1989. Genetic stratigraphic sequences

in basin analysis I: architecture and genesis of flooding-surface bounded depositional units. American Association of Petroleum Geologists Bulletin, 73, 125-142. --, GARBER, J. L., XIJIN LIU & SLOAN, B. J. 1993. Sequence stratigraphic and depositional framework of the Cenozoic fill, Central and Northern North Sea Basin. In: PARKER,J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 33-43. GRADSTEIN, F. M., KRISTIANSEN, I. L., LOEMO, L. & KAMINSKI,M. A. 1992. Cenozoic foraminiferal and dinoflagellate cyst biostratigraphy of the central North Sea. Micropaleontology, 38, 101-137. HARLAND, R., HINE, N. M. & WILKINSON, I. P. 1992. Paleogene biostratigraphic markers. In: KNOX, R. W. O'B. & HOLLOWAY, S. 1. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea-levels since the Triassic. Science, 235, 1156-1167.

INTEGRATED NORTH SEA PALEOGENE AND EOCENE STRATIGRAPHY HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg I borehole, central Jylland, Denmark. Danmarks Geologiske UndersCgelse, Serie A, 7, 1-69. ISAKSEN, D. & TONSTAD,K. 1989. A Revised Cretaceous and Tertiary Lithostratigraphic Nomenclature for the Norwegian North Sea. Norwegian Petroleum Directorate Bulletin, 5, 1-59. JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: Palaeogene stratigraphy and relative sea-level history of the outer Moray Firth, UK North Sea. Marine and Petroleum Geology, 11, 157-165. JOPPEN, M. & WHITE, R. S. 1990. The structure and subsidence of Rockall Trough from two-ship seismic experiments. Journal of Geophysical Research, 95, 19 821-19 837. KENNEDY, W. J. 1987. Late Cretaceous and Early Paleocene Chalk Group sedimentation in the Greater Ekofisk area, North Sea Central Graben. Bulletin des Centres de Recherches ExplorationProduction Elf-Aquitaine, 11, 91-126. KENNETT, J. P. & STOTT, L. D. 1991. Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229. KING, C. 1983. Cainozoic Micropalaeontological Biostratigraphy of the North Sea. Institute of Geological Sciences, Report 82/7, 1-40. 1989. Cenozoic of the North Sea. In: JENKINS,D. G. & MURRAY, J. W. (eds) Stratigraphical Atlas of Fossil Foraminifera 2rid Edition. Ellis Horwood Ltd, Chichester, 418-489. KNOX, R. W .O'B. & HOLLOWAY,S. 1992. Paleogene of the Central and Northern North Sea. In: KNox, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. MILTON, N. J., BERTRAM,G. Z. & VANN, I. R. 1990. Early Palaeogene tectonics and sedimentation in the Central North Sea. In: HARDMAN, R. E E & BROOKS, J. (eds) Tectonic Events Responsible for Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. -

-

113

MORTON,A. C., HALLSWORTH,C. R. & WILKINSON,G. C. 1993. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 73-84. MUDGE, D. C. & BUJAK, J. V. 1994. Eocene stratigraphy of the North Sea basin. Marine and Petroleum Geology, 11, 166-181. & COPESTAKE, P. 1992a. A revised Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. & -1992b. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. NEAL, J. E., STEIN, J. A. & GAMBER, J. H. 1994. Graphic correlation and sequence stratigraphy in the Palaeogene of NW Europe. Journal of Micropalaeontology, 13, 55-80. POWELL, A. J., BRINKHUIS, H. & BUJAK, J. P. 1996. Dinoflagellate cyst sequence biostratigraphy of the Thanetian in the type region. This volume. SCHRODER, Z. 1992. A palynological zonation for the Paleocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-126. STEWART,I. J. 1987. A revised stratigraphic interpretation of the Early Palaeogene of the central North Sea. In: BROOKS, J. & GLENNIE, K. W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 557-576. THOMAS, E. & SHACKLETON, N. J. 1996. PaleoceneEocene benthic foraminiferal extinction and stable isotope anomalies. This volume. VAN WAGONER, J. C., MITCHUM, R. M., CAMPION,K. M. & RAHMANIAN,V. D. 1990. Siliciclastic Sequence Stratigraphy in Well Logs, Cores and Outcrops: Concepts for High-Resolution Correlation of Time and Facies. American Association of Petroleum Geologists Methods in Exploration Series No. 7. WILLIAMS, G. L. & BUJAK, J. P. 1977. Cenozoic palynostratigraphy of offshore eastern Canada. In: ELSIK, W. E. (ed.) Contributions of Stratigraphic Palynology, Vol. 1, Cenozoic Palynology. American Association of Stratigraphic Palynologists, Contribution Series, 5A, 14-47. -

-

The occurrence of the dinoflagellate cyst Apectodinium (Costa & Downie 1976) Lentin & Williams 1977 in the Moray and Montrose Groups (Danian to Thanetian) of the UK central North Sea Extended Abstract J. E. T H O M A S

British Geological Survey, Keyworth, Nottingham NG12 5GG, UK

Apectodinium (Costa & Downie 1976) Lentin & Williams 1977, a genus of wetzeliellioid dinoflagellate cyst, has been widely reported from the Upper Paleocene (Thanetian) and lowermost Eocene of the UK North Sea and from onshore UK and continental Europe. Its distinctive appearance and common occurrence have led to the selection of members of the genus as index species in a number of dinoflagellate cyst zonation schemes covering the Upper Paleocene and Lower Eocene, e.g. by Costa & Downie (1976), Heilmann-Clausen (1985), Costa & Manum (1988), Powell (1988, 1992) and Schr6der (1992). Powell (1988) suggested that, when working with cuttings, the first downhole occurrence (FDO) of Apectodinium augustum (Harland 1979) Lentin & Williams 1981 can be used as a good approximation to the top of the Paleocene. The lowermost occurrence of the genus was reported by Costa & Downie (1976) as occurring at a level equivalent to the base of the late Paleocene calcareous nannoplankton biozone NP9 of Martini (1971) in NW Europe. This was referred to as the 'Base Apectodinium Datum' by Powell (1988). However, several authors have reported the occurrence of Apectodinium spp. at lower stratigraphic levels. Jan du Chine et al. (1975) recorded Apectodinium spp. (as Wetzeliella) in samples assigned to biozones NP5 to NP8 from Haute-Savoie, France. Powell et al. (1993) reported an early (pre-NP9) occurrence of the Base Apectodinium Datum in Thanetian strata at Herne Bay in Kent, within the Reculver Silts Member below the Reculver Tabular Band. These beds were assigned to biozone NP8 by Siesser et al. (1987). Brinkhuis et al. (1994) noted that the Base Apectodinium Datum in lower palaeolatitudes is generally calibrated against calcareous nannofossil zones NP6-NP8. In addition, Brinldauis et al. (1994) reported Apectodinium hyperacanthum (Cookson & Eisenack 1965) Lentin & Williams 1977 in strata as old as earliest Selandian (NP4) at E1 Kef in Tunisia. They

suggested therefore that chronostratigraphic correlation using the Base Apectodinium Datum is problematic. Additionally, they suggested that the taxon migrated northwards during the Late Paleocene, possibly indicating climatic warming, culminating in the worldwide acme of Apectodinium spp. near the Paleocene-Eocene boundary. Section studied

UK central North Sea Well 22/10a-4 (Fig. 1) has over 122 m of core covering an interval of 146 m representing the Sele, Lista and Maureen formations of Knox & Holloway (1992). However, the precise positions of the international Paleocene stage boundaries (including the base of the

m

Jan L e

Fig. 1. Location of UK central North Sea Well 22/10a-4 (after Knox & Holloway 1992).

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 115-120.

115

116

J.E. THOMAS

;>

"o

ga,

e,i

APECTODINIUM IN THE CENTRAL NORTH SEA

Selandian) are yet to be agreed and they are therefore not certain in well 22/10a-4. The well is located in a basinal setting (Fig. 1) and was recognized as providing a reference section for biostratigraphic, lithostratigraphic and sedimentological studies. The stratigraphy of the section is summarized in Fig. 2. Details of samples from the Lista and Maureen Formations are given in Table 1.

Palynology Apectodinium spp. Species of the genus Apectodinium are typically pentagonal in shape, covered in fine, nonparatabular processes and have a single-paraplate, quadra, intercalary archaeopyle. The genus comprises a plexus of forms with speciation chiefly based on the degree of development of the lateral, apical and antapical horns. In practice, many intermediate forms are found. Apectodinium species including A. augustum, A. homomorphum (Deflandre & Cookson 1955) Lentin & Williams 1977, A. hyperacanthum, A. paniculatum (Costa & Downie 1976) Lentin & Williams 1977, A. parvum (Alberti 1961) Lentin & Williams 1977 and A. quinquelatum (Williams & Downie 1966) Costa & Downie 1979 were found to be common to abundant throughout the Forties Sandstone Member of the Sele Formation and below that in unit S lb and the upper part of unit S I a to a depth of 2623.48 m. The percentage of the dinoflagellate cyst assemblage made up by Apecwdinium spp. is illustrated in Fig. 2. In addition to the expected occurrences of Apectodinium spp. in the Sele Formation, the genus was found to be present at 2658.35 m in the Lista Formation. The species present were A. augustum, A. homomorphum, A. paniculatum and A. parvum in a low-diversity dinoflagellate cyst assemblage. In addition, a large number of specimens transitional between A. augustum and A. paniculatum were encountered. Apectodinium spp. were also found to be present in the Maureen Formation, in basal unit M2 at 2701.68 m and in unit M1 at 2707.10 m.

Apectodinium augustum, A. homomorphum, A. parvum and A. quinquelatum plus some intermediate forms were present, again in lowdiversity assemblages. These results were considered sufficiently unusual to warrant resampling of the Maureen Formation core in much more detail. Apectodinium events were again encountered at similar levels, allowing for differences in sample quality and spacing. In the second sample set Apectodinium spp. were found in unit M2 at 2697.02 m and in unit M1 at 2706.40 m.

] 17

On the basis of current North Sea usage the

Apectodinium spp. events at 2706.40m and 2707.10 m are considered to be of latest Danian age.

Other dinoflagellate cysts Floods of dinoflagellate cysts belonging to the genera Areoligera and Glaphyrocysta, including Areoligera gippingensis Jolley 1992, were encountered at 2644.82, 2652.53 and 2655.48 m (Lista Formation). However, allocation of these strata to the A. gippingensis acme of Harland et al. (1992) cannot be made with confidence due to the generally poor productivity of the samples in question. Other palynological markers include the range base of Isabelidinium ? viborgense HeilmannClausen 1985 at 2701.22 m in unit M2. Since this species was found to be present in the uppermost sample of the Maureen Formation core but not in the lowermost sample of the Lista Formation core its range top is assumed to lie in the uncored interval. The range top of Alisocysta reticulata Damassa 1979 occurs at 2704.80 m in unit M1, although the range top of common or abundant A. reticulata lies at 2711.13 m in unit M1. The uppermost occurrence of this species was chosen as a marker for the top of the early Paleocene (Danian) by Harland et al. (1992). Additionally, HeilmannClausen (1994) stated that the lowermost occurrence of A. reticulata is in the upper Danian and that the uppermost occurrence of the species may prove useful for correlation with other regions although its precise position relative to North Sea zones is still uncertain. Another significant component of the dinoflagellate cyst assemblages in unit M1 is a distinctive gonyaulacacean species, Spiniferites sp. A, referred to as Spiniferites n. sp. by Stewart (1987) and Spiniferites cf. supparus by Powell (1992). Its range top at 2701.45 m lies just above the top of unit M1 in 22/10a-4. A number of authors, including Powell (1992) and Harland et al. (1992), have indicated that the uppermost occurrence of this species could be used to mark the top of the Danian in the North Sea Basin. In general, dinoflagellate cyst assemblages from the Maureen Formation of 22/10a-4 were dominated by Spiniferites spp., Cleistosphaeridium spp., Hystrichosphaeridium spp. and Areoligera spp. Also common was Palaeoperidinium pyrophorum (Ehrenberg 1838) Sarjeant 1967, its range top in Well 22/10a-4 occurring at 2625.00 m (unit Sla). Thalassiphora delicata Williams & Downie 1966 was present to a depth of 2703.13 m. Confirmation that the records of Apectodinium spp. in the Maureen Formation are true stratigraphic records, and not the result of misplaced core, is provided by the occurrence of Alisocysta

118

J . E . THOMAS

Table 1. Details of samples from the Lista and Maureen Formations of Well 22/10a-4 Core depth (ft)

8630.00 8636.00 8640.10 8645.00 8650.00 8655.00 8660.40 8665.00 8670.00 8675.00 8680.30 8685.00 8690.00 8695.75 8699.40 8704.00 8713.80 8723.40 8803.50 8805.00* 8807.00* 8809.00 8815.00" 8816.50 8819.75 8826.25* 8830.00 8837.00* 8839.00* 8840.00* 8840.75* 8841.50 8846.25* 8847.00 8849.50* 8851.75" 8855.00* 8857.00* 8859.00 8862.50* 8866.00 8869.00* 8871.00 8872.50* 8876.00 8879.00* 8881.50 8886.00 8888.00* 8891.00 8896.00

Log depth (ft)

Log depth (m)

Lithostrat unit

Dinocyst diversity (in No. taxa)

Apectodinium (%)

Marine (%)

8650.00 8656.00 8660.10 8665.00 8670.00 8675.00 8680.40 8685.00 8690.00 8695.00 8700.30 8705.00 8710.00 8715.75 8719.40 8724.00 8733.80 8743.40 8823.50 8825.00 8827.00 8829.00 8835.00 8836.50 8839.75 8846.25 8850.00 8857.00 8859.00 8860.00 8860.75 8861.50 8866.25 8867.00 8869.50 8871.75 8875.00 8877.00 8879.00 8882.50 8886.00 8889.00 8891.00 8892.50 8896.00 8899.00 8901.50 8906.00 8908.00 8911.00 8916.00

2637.19 2639.02 2640.27 2641.76 2643.29 2644.82 2646.46 2647.86 2649.39 2650.91 2652.53 2653.96 2655.48 2657.23 2658.35 2659.76 2662.74 2665.67 2690.09 2690.54 2691.15 2691.76 2693.59 2694.05 2695.04 2697.02 2698.17 2700.30 2700.91 2701.22 2701.45 2701.68 2703.13 2703.35 2704.12 2704.80 2705.79 2706.40 2707.01 2708.07 2709.15 2710.06 2710.67 2711.13 2712.20 2713.11 2713.87 2715.24 2715.85 2716.77 2718.29

Lista Lista Lista Lista Lista Lista Lista Lista Llsta Lista Lista Lista Lista Lista Lista Lista Lista Lista M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M2 M1 M1 M1 M1 M1 M1 M1 M1 MI MI M1 M1 M1 M1 M1 M1 MI M1 M1

24 11 2 Barren Barren 6 21 21 Barren Barren 6 19 8 Barren 9 19 9 2 Barren 30 37 29 34 29 1 10 Barren 37 37 24 38 21 34 25 34 4O 33 23 25 42 Barren 38 27 42 34 44 29 29 40 28 15

0 0 0

16.77 75.42

0 0 0

99.21 70.92 72.08

0 0 0

67.25 84.17 88.23

47.50 0 0 0

70.83 33.22 40.55 90.50

0 0 0 0 0 0 45.50

67.95 59.60 57.34 65.55 65.82

0 0 0 0 11.60 0 0 0 0 0 1.93 2.60 0

55.37 64.95 62.40 98.90 57.57 88.58 91.98 94.95 93.40 94.55 79.50 67.07 92.90

0 0 0 0 0 0 0 0 0 0

96.00 96.25 93.65 97.89 91.84 96.82 93.02 97.00 88.88 91.78

* Indicates resampling of the Maureen Formation core.

52.40

APECTODINIUM IN THE CENTRAL NORTH SEA

reticulata and Spiniferites sp. A in the same assemblage as Apectodinium augustum at 2706.40 rn. Additionally, Palaeoperidinium pyrophorum was found to be present in the same assemblages as A. augustum at 2701.68, 2706.40 and 2707.01 m. The absence of such diagnostic species in the other Apectodinium-bearing samples is ascribed to the low-diversity nature of the assemblages (see below). Additional factors The relative abundance of marine versus nonmarine palynomorphs is illustrated in Fig. 2. and detailed in Table 1. The diversity of the dinoflagellate cyst assemblage (in number of taxa) is also given in Table 1. Generally, the Lista and Maureen Formation Apectodinium spp. events correspond to increases in the relative abundance of terrestrially derived palynomorphs, mostly pollen, and a limited decline in the diversity of the dinoflagellate cyst assemblages accompanied by a reduction in numbers of skolochorate (spinebearing) cysts. Furthermore, the amounts of amorphous organic matter are greater in these samples. Taken together with the restriction of the Apectodinium events to more sandy lithologies, these factors could be interpreted as indicating that the events do not represent in situ assemblages, but basin-margin assemblages that were swept periodically into the basin centre. Indeed, the 'Hystrichosphaera Association' of Downie et al. (1971), consisting of skolochorate cysts, was taken to represent more open marine conditions whilst the 'Wetzeliella Association' was taken to represent an inner neritic environment. Brinkhuis et al. (1994) mentioned an influx of continentallyderived material accompanied by the dominance of marginal marine dinoflagellate cysts at the

119

Danian-Selandian Boundary. The occurrence of a similar association in Well 22/10a-4 points to a possible relationship between basinal occurrences of Apectodinium spp. and periods of lowered sea level in the North Sea Basin. Evidence that Apectodinium spp. occurred in marginal North Sea environments before the major flourishing of the genus in biozone NP9 is provided by the record of Apectodinium spp. in Powell et al. (1993) from the Reculver Silts (NP8) in Kent. Additional data from such marginal deposits is needed to fully establish the distribution and environmental preferences of the genus. Although Brinkhuis et al. (1994) mentioned only the possible influence of climatic warming on extending the geographical distribution of Apectodinium spp., it is evident that other factors, such as sea-level change and palaeogeographical configuration, may have played an equally important role. It is clear from this study that a pattern of facies control is superimposed on the temporal distribution of the genus and that caution should be exercised in using the Base Apectodinium Datum as a time-constant biostratigraphic marker, even within a single basin. For the same reasons, it is not yet possible to say whether the 'early' appearances of Apectodinium spp., now reported from the central North Sea, southern England, France and Tunisia, represent events with the potential for regional correlation or whether they are of local significance only. The author benefited greatly from discussions with R. Harland and R. W. O'B. Knox during the course of this work. The paper was substantially improved by the comments of the referees C. Heilmann-Clausen, D. W. Jolley and J. B. Riding. This paper is published with the approval of the Director, British Geological Survey (NERC).

References ALBERTI, G. 1961. Zur Kenntnis mesozoischer und altterti~irer Dinoflagellaten und Hystrichosphaerideen von Nord- und Mitteldeutschland sowie einigen anderen europaischen Gebieten. Palaeontographica A, 116, 1-58. BRINKHUIS,H., ROMEIN, A. J. T., SMIT, J. & ZACHARIASSE, J.-W. 1994. Danian-Selandian dinoflagellate cysts from lower latitudes with special reference to the E1 Kef section, NW Tunisia. GFF, 116, 46-48. COOKSON, I. C. & EISENACK, A. 1965. Microplankton from the Dartmoor Formation, SW Victoria. Proceedings of the Royal Society of Victoria, 79, 133-137. COSTA, L. I. & DOWNIE, C. 1976. The distribution of the dinoflagellate Wetzeliella in the Palaeogene of north-western Europe. Palaeontology, 19, 591-694. -& 1979. The Wetzeliellaceae; Palaeogene dinoflagellates. Proceedings of the Fourth Inter-

national Palynology Conference, Lucknow (19761977), 2, 34-46. & MANUM, S. B. 1988. The description of the interregional zonation of the Paleogene (D1-D15) and the Miocene (D16-D20). Geologisches Jahrbuch, A100, 321-330. DAMASSA, S. P. 1979. Danian dinoflagellates from the Franciscan Complex, northern California. Palynology, 3, 191-207. DEFLANDRE, G. & COOKSON, I. C. 1955. Fossil microplankton from Australian late Mesozoic and Tertiary sediments. Australian Journal of Marine and Freshwater Research, 6, 242-313. DOWNIE, C., HUSSAIN, M. A. & WILLIAMS, G. L. 1971. Dinoflagellate cyst and acritarch associations in the Paleogene of south-east England. Geoscience and Man, 3, 29-35. EHRENBERG, C. G. 1838. Ober das Massenverhfiltniss --

120

J.E. THOMAS

der jetzt lebenden Kiesel-Infusorien und fiber ein neues Infusorien-Conglomerat als Polirschiefer von Jastraba in Ungarn. Abhandlungen der Preussischen Akademie der Wissenschaften, 1836, 109-135. HARLAND, R. 1979. The Wetzeliella (Apectodinium) homomorphum plexus from the Palaeocene/earliest Eocene of northwest Europe. Proceedings of the Fourth International Palynology Conference, Lucknow (1976-1977), 2, 59-70. , HtNE, N. M. & WILKINSON, 1. P. 1992. Appendix. Paleogene biostratigraphic markers. In: KNOX, R. W. O'B. & HOLLOWAY, S. Paleogene of the Central and Northern North Sea. In: KNox, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham, A1-A5. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks Geologiske UndersOgelse, A7, 1~59. 1994. Review of Paleocene dinoflagellates from the North Sea region. GFF, 116, 51-53. JAN DU CHI~NE, R., GORIN, G. & VAN STUIJVENBERG,J. 1975. Etude g6ologique et stratigraphique (palynologie et nannoflore calcaire) des Gr~s des Voirons (Paleog~ne de Haute-Savoie, France). Geologie Alpine, 51, 51-78. JOLLEY, D. W. 1992. A new species of the genus Arealigera Lejeune Carpentier from the Late Palaeocene of the eastern British Isles. Tertiary Research, 14, 25-32. KNOX, R. W. O'B. & HOLLOWAY, S. 1992. Paleogene of the Central and Northern North Sea. In: KNox, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. LENTIN, J. K. & WILLIAMS, G. L. 1977. Fossil Dinoflagellates: Index to Genera and Species. 1977 Edition. Bedford Institute of Oceanography, Report Series, B1-R-77-8, 1-209. & 1981. Fossil Dinoflagellates: Index to

Genera and Species. 1981 Edition. Bedford Institute of Oceanography, Report Series, B1-R-81-12, 1-345. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: Proceedings of the H Planktonic Conference, Roma, 1970, Vol. 2, 739-785. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Palaeocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327-344. 1992. Dinoflagellate cysts of the Tertiary system. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. BMS Publications Series, Chapman & Hall, London, 155-251. --, BRINKHUIS,H. & BUJAK, J. P. 1993. Dinoflagellate cyst sequence biostratigraphy of the Thanetian in the type region. In: Correlation of the Early Paleogene in Northwest Europe, Programme and Abstracts. Geological Society, London. SARJEANT, W. A. S. 1967. The genus Palaeoperidinium Deflandre (Dinophyceae). Grana Palynologica, 7, 243-258. SCHRODER, T. 1992. A palynological zonation for the Palaeocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-126. SmSSER, W. G., WARD, D. J. & LORD, A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Journal of Micropalaeontology, 6, 85-102. STEWART,I. J. 1987. A revised stratigraphic interpretation of the Early Palaeogene of the central North Sea. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe, Volume 1. Graham & Trotman, London, 557-576. WILLIAMS, G. L. • DOWNIE, C. 1966. Wetzeliella from the London Clay. In: DAVEY, R. J., DOWNIE, C., SARJEANT, W. A. S. & WILLIAMS, G. Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bulletin of the British Museum of Natural History, Supplement 3,215-235. -

-

An integrated palynological-palynofacies approach to the zonation of the Paleogene in the Forties-Montrose Ridge area, central North Sea S U S A N E. W O O D & R I C H A R D V. T Y S O N

Newcastle Research Group, Fossil Fuels and Environmental Geochemistry (Postgraduate Institute), D r u m m o n d Building, University o f Newcastle, Newcastle upon Tyne NE1 7RU, UK Abstract: A palynological and palynofacies study of late Paleocene and early Eocene sediments

of the central North Sea has been undertaken using 467 core samples from 11 wells on the Forties-Montrose Ridge. Conventional dinoflagellate cyst (dinocyst) biostratigraphy gives poor precision within the narrow (c. 2.5 million years) time interval studied. Furthermore, dinocysts are often present in only very low numbers and are frequently poorly preserved. An alternative local zonation scheme is proposed on the basis of downwell variations in quantitative palynofacies and palynomorph occurrence data, and on angiosperm pollen ratios. Integration of these three sets of data indicates a potential sevenfold subdivision of the mid-Sele Formation ($2) to upper Lista Formation (L3) interval in this area, with up to five divisions within the Forties Sandstone Member. This composite zonation scheme appears to work reasonably well for the studied wells, and preliminary comparisons with independent biostratigraphic data from two of our wells indicate a partial correlation with the dinocyst zones.

The Paleocene Forties Sandstone Member (FSM) (sensu Knox & Holloway 1992, p. 53-55) is one of the major exploration targets in the North Sea, and has been one of the basin's most prolific oil reservoirs. As has been widely recognized for many years, this sandstone unit was deposited in a submarine fan setting, with the main reservoir units corresponding to lower to middle fan channel sands (e.g. Thomas et al. 1974; Carman & Young 1981; Kulpecz & Van Geuns 1990). Submarine fan sands represent particularly challenging reservoir targets because of the complexity of the vertical and lateral (radial and proximal-distal) facies changes that result from the interactions of palaeogeographic, tectonic, eustatic and sedimentological factors. Successful production strategies therefore require sophisticated depositional models, which in turn require detailed local correlation schemes utilizing as wide a range of stratigraphic tools as possible. The integration of seismic stratigraphy, biostratigraphy and geophysical log correlations has been especially valuable (e.g. Stewart 1987; Milton et al. 1990; Mudge & Copestake 1992). This integrated approach is now almost universal in North Sea Paleogene studies (e.g. Whyatt et al. 1992; Armentrout et al. 1993; D e n Hartog Jager et al. 1993; O'Connor & Walker 1993; Vining et al. 1993). In addition to the general difficulties associated with submarine fan sands, biostratigraphic corre-

lation in the FSM is complicated by a number of other factors. The dark-coloured shale facies (characteristic of the Sele Formation) was deposited under dysoxic-anoxic depositional conditions, resulting in a strongly facies-controlled assemblage of calcareous microfossils (nannoflora and foraminifera) that is both sparse and rather poorly preserved (Stewart 1987). Paleogene palynology also has its problems; the taxonomy of both the plankton (dinocysts) and sporomorphs (spores and pollen) of this interval are complex and incompletely described; most of the available zonation schemes produced are of a local nature, and the relative influence of palaeoecological and evolutionary factors is not yet entirely clear. Consequently, the petroleum industry makes pragmatic use of a whole range of micropalaeontological (M) and palynological (P) 'bioevents' to provide a biostratigraphic framework for the North Sea Paleogene (Stewart 1987, p. 562; Mudge & Copestake 1992, p. 55). A lithostratigraphicbiostratigraphic-seismic sequence correlation is given in Table 1. This framework features only two bioevents which relate specifically to the FSM, and these correspond only to its top and base (P3 and M5 of Mudge & Copestake 1992, p. 56, respectively); it cannot be used in intra-FSM correlations. The poor biostratigraphic subdivision of the FSM undoubtedly relates to the short time period

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Earl), Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 121-128.

121

122

S.E. WOOD • R. V. TYSON

during which this 200 m thick unit was deposited. The estimated duration of only c. 0.8-1.0 million years is too short for major evolutionary changes in either plankton or terrestrial floras, and therefore a small number of conventional biozones is inevitable. Other approaches, such as ecostratigraphy event stratigraphy, and palynofacies are therefore essential if greater resolution is to be achieved. Non-palaeontological approaches have also been applied. The FSM (S 1) can sometimes be locally subdivided into two units based on gammaray log signatures (Knox & Holloway 1992; O'Connor & Walker 1993): a lower and often more argillaceous unit (Sla) is separated from an upper unit (Slb) by a gamma-ray spike, especially in more distal sections. These gamma spikes relate to thin mudstones which are thought to reflect basinwide condensation events (Milton et al. 1990; Mudge & Copestake 1992, p. 68), and thus to be approximately isochronous, although the potential for local condensation due to sediment bypass cannot be ruled out. The commercial importance and value of Paleogene palynological zonation schemes mean that most of the intensive studies that have been undertaken within the petroleum industry are as yet unpublished. Some aspects of the palynological zonation schemes of Simon-Robertson, Mobil, BGS-Unocal and Esso have recently been very

briefly mentioned in papers in the Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference (Armentrout et al. 1993, p. 53; Morton et al. 1993, p. 75-76; Vining et al. 1993, p. 20). However, the only detailed publications available are those of Powell (1988) and Schrtider (1992). The dinocyst zonation scheme of Powell (1988) shows that the 'Forties Formation' corresponds to the Apectodinium augustum biozone. Three dinocyst biozones (Aau, Ahy and Ama) are shown for the 'Forties interval' in a table by Powell (1992, p. 15), but no details are given. The Area (Alisocysta margarita) biozone is usually placed in the top of the Lista Formation (L3), below the FSM (e.g. Knox & Holloway 1992, p. 30), so only the first two of Powelrs (1992) biozones appear to apply to the FSM. The most detailed published palynological scheme, that of Schr6der (1992), provides a fourfold palynological subdivision of the 'Forties Formation' (PT 19.1-PT 19.4, where PT stands for palynological time). However, the M5 bioevent of Mudge & Copestake (1992), which defines the base of the FSM, is correlated with the middle part of zone PT 19.2 (Schr6der 1992, p. 124), so only the upper two and a h a l l out of the four PT zones, relate to the FSM. Schr6der's subdivision of the FSM is based on one dinocyst-related boundary which defines the base of zone PT 19.2 (the base

Table 1. Lithostratigraphic, seismic and biostratigraphic subdivision of the Paleocene-Early Eocene (Thanetian-

Ypresian pars) of the central North Sea Basin Knox & Holloway 1992 Lithostratigraphy Sand Shale

Stewart 1987 Code

Sequence

B2 & B 1

9

Seismic System tracts

Mudge & Copestake 1992

SchriSder 1992

Bioevent

PT zone

Basin oxygen

21

Anoxic

20

Anoxic

19.4 19.2-

Anoxic?

M7 Balder

LST M6

Sele

$3 & $2

8

HST P3

Forties

Sele

S1

7

TST 4-- LST M5

Lista

L3 (pars)

6

HST

Oxic M4 & P2

Balm'l

Lista

L3 & L2

5/4

TST

19.1 15

Oxic

15

Oxic

13

Oxic

P1 Andrew

Lista

L1

3

LST M3

Maureen

M2 & M1

2

HST <---LST

The correlations between the various published schemes are sometimes approximate (especially for the Lista L3 & L2). The Mudge & Copestake (1992) bioevent scheme (based on Stewart 1987) refers to the top of events which mark the upper boundaries of the various intervals (in most cases locally correlated with thin high gamma-ray mudstone horizons). The correlation between bioevents and Schr6der's (1992) PT zones is as given by that author. LST, lowstand systems tract; TST, transgressive systems tract; HST, highstand systems tract; Balm'l, Balmoral Sandstone (Mey Sandstone Member, L2 & L3).

NORTH SEA PALEOGENE PALYNOFACIES of A. augustum and the lower limit of the common occurrence of other Apectodinium species), the stratigraphic range of the pollen Labrapollis labraferoides (which defines the base of PT 19.3), and on cyclic, apparently climate-related, changes in the abundance of certain other pollen species (see Schr6der 1992, p. 124). The dinocyst event may correlate with the Aau-Ahy dinocyst biozone boundary of Powell (1992). It is possible that this event may be present in four of our study wells, but the variable dinocyst recovery prohibited any definite conclusions. Although the use of relative abundance data for palynomorph taxa is increasingly common in commercial North Sea studies, little use has apparently been made of detailed quantitative palynofacies data. However, Boulter & Riddick (1986) have described some aspects of FSM palynofacies, and Enjolras et al. (1986, p. 172-174) have commented on some general palynofacies trends in the North Sea Paleogene. More recently, Schr6der (1992) has given descriptive notes on the palynofacies character of his PT zones, and Vining et al. (1993) have given a brief account of the overall palynofacies characteristics of the Lista and Forties Formations. As part of our own study, we have tried to independently assess the potential value of quantitative palynofacies data in intraFSM correlation studies.

Approach and methods A detailed palynology and palynofacies investigation of Paleocene-Eocene sediments was undertaken using cores from 11 wells from the FortiesMontrose Ridge area of the central North Sea (Fig. 1). The identity and exact location ofthe wells cannot be reported at this time, but they range from the Forties Field in the northwest to the vicinity of the Nelson Field in the southeast, and together provide an c. 70 km long proximal-distal (NW-SE) transect down the Forties-Montrose submarine fan system. The stratigraphic coverage of the cores varies from well to well, but in total ranges from the mid-Sele Formation ($2?), through the FSM, to the upper Lista (L3?). However, only two dinocyst biozones could be identified with confidence: the Deflandrea oebisfeldensis interval biozone (B2-$2) and the Apectodinium augustum acme biozone (S1), both sensu Powell (1988); dinocyst assemblages corresponding to the Ama zone were not positively identified in our well core-sections thought to contain the Lista Formation (L3). Basic lithological logs were made of the core sections, and the colour, modal grain size, fabric, and plant macrofossil content of each sample were recorded. Some 467 core samples, including 252 shale, 163 argillaceous sandstone/siltstone, and 46

123

Fig. 1. Location map for the Forties-Montrose Ridge area.

sandstone samples were analysed palynologically. The mean stratigraphic sample density within the argillaceous intervals was approximately one sample for every 3 m; 5 g of each sample were processed using standard palynological techniques and the residue sieved with a 10 gm nylon mesh before being strew-mounted in glycerin on glass slides. The overall (unoxidized) kerogen composition was determined using 500 counts per slide (relative numeric frequencies sensu Tyson 1993, 1995). Separate palynomorph counts (using unoxidized, 'panned' preparations) were then made using a target of 200 sporomorph counts per slide, plus whatever other palynomorphs were encountered during this procedure. The advice we received at the time indicated that oxidation treatments would distort the palynomorph assemblage, so panning was used instead of oxidation; however, we retrospectively consider this advice to have been overcautious, and that some oxidation and double-sieving (as now regularly used in the industry) would have produced better quality palynomorph data [e.g. less masking by amorphous organic matter (AOM)]. A tenfold classification scheme was used for the kerogen counts (with greatest emphasis on the phytoclast components), and an 18-fold classification was used for the palynomorph (sporomorph and microplankton) counts. The stratigraphic occurrence of sporomorph and microplankton taxa were also separately recorded. The quantitative data were recalculated as various percentages and ratios according to the

124

s.E. WOOD & R. V. TYSON

approach of Tyson (1995); the palynomorph data were also used in their raw form to indicate relative yields between samples. Our original intention to use the exotic Lycopodium-grain method to assess particle abundances in terms of numbers per gram was prevented by the temporary non-availability of calibrated Lycopodium tablets. The quantitative data were visually and statistically evaluated to identify stratigraphic and lateral variations, to identify relationships between the different parameters, and to evaluate the influence of both actual sample lithology, and the dominant lithology of the interval from which any sample was taken.

Lithological controls on palynomorph data Our data show that lithology exerts a strong control on the occurrence and relative abundance of palynomorphs. Whereas 85 and 95% of the shale samples yield sporomorphs and plankton, respectively, these figures fall to 67 and 35% in sandstone samples. The percentage of palynomorphs in the total kerogen assemblage is in the range of 3-33% for shale samples (mean 12%), 1-21% for argillaceous sandstone/siltstone samples (mean 4%) and 1-3% for the sandstone samples (mean 1%). The relative palynomorph yield of the shale samples shows major differences between the lithostratigraphic units: the kerogen assemblages contain 30% palynomorphs in the Sele ($2) shale samples, but only 6% in the FSM (S1) shales, and 2.5% in the Lista Formation (L3). Shale samples taken from shale-dominated intervals are generally more productive than those taken from sandstonedominated intervals, except for the Lista samples which generally yielded only poor amounts of residue (oxic highstand mudstone facies associated with low phytoclast input and little preservation of AOM of marine origin). The average recorded count of the total marine microplankton for all samples and all lithologies was only nine, relative to the 200 sporomorph count. Although the percentage of total marine microplankton in the palynomorph assemblage ranges up to 75% in isolated samples, in half of the productive samples it is < 7%, and in 80% of the samples it is < 25%. We found the abundance of dinocysts to be generally so low that it was not practical to compensate for this by increasing the size of the palynomorph count. Boulter & Riddick (1986, p. 876) have also observed that dinocysts are 'rarely abundant' and 'usually very badly preserved' in the FSM. This makes consistent use of dinocyst-based subdivisions rather difficult. There are several probable reasons for the low relative abundance of marine microplankton, but most relate to the basinal setting. The environment of deposition of the shaley hemipelagic units was

probably too deep to be associated with significant in situ dinocyst production (as dinocysts are mainly associated with shallow unstable shelf regimes, see reviews of Tyson 1993, 1995). The sand-rich intervals like the FSM are associated with extensive turbidite redeposition and strong terrestrial influence, and thus the palynomorphs are diluted by phytoclasts and the marine microplankton diluted by sporomorphs. In addition, in the Sele Formation ($2), the high relative abundance of AOM (mean 46% of kerogen assemblage v. 32% in the FSM) results in partial masking of palynomorphs, and the evident high relative imput of small angiosperm pollen strongly dilutes the few dinocysts that were redeposited. Some high dominance dinocyst 'flood events', most commonly of Apectodinium, were observed in specific samples, and here the percentage of dinocysts in the palynomorph assemblage was c. 25% or more. These events were found to be most frequent in the upper FSM (S 1) to lower mid-Sele ($2) interval, and may be associated with local condensation events related to the major change in depositional conditions at this time (the rising sea level), and the associated reduction in redeposition (less dilution by terrestrial components).

Individual zonation schemes We have used our quantitative palynofacies and palynomorph data to construct three independent types of zonation. The first is based upon the quantitative palynomorph data, both key ratios and raw count data (PNo zones); the second is based upon trends in the general composition angiosperm pollen assemblages (POL zones); the third is based upon quantitative kerogen data (PF zones). Examination of each type of data allows us to define three zones for the first two types of zonation and four for the last: PNo zones A-C, POL zones I-III and PF zones 1-4. The zonal boundaries were determined from visual examination of stratigraphic plots of each form of data, making appropriate allowances for localized lithological anomalies and the general 'noise' level in the data. The small number of zones differentiated reflects an attempt to distinguish only the largest, most consistently recognizable, and thus most significant, stratigraphic trends within the sequence.

PNo zones Zone A is characterized by the presence of deflandreoid dinocysts and a usual absence of Apectodinium spp, indicating the D. oebisfeldensis biozone of Powell (1988). A count of 200 sport-

NORTH SEA PALEOGENE PALYNOFACIES

morphs can usually be achieved, but this is strongly dominated by small angiosperm pollen; there is a lack of fungal spores and thick-walled or strongly-ornamented pteridophyte spores, and usually a low count of bisaccate pollen (1-5 per sample). Comparison with lithological logs suggests this zone correlates with the supraFSM mid-Sele Formation ($2). A transitional A-B zone of mixed characteristics was all that could be distinguished in some (especially distal) wells. Zone B is characterized by the presence of Apectodinium, sometimes in large numbers (socalled 'flood events'). Some deflandreoid dinocyst taxa are also present, but in generally lower numbers than in zone A; the A. augustum biozone of Powell (1988) is indicated. Variable counts of total palynomorphs are recorded, including relatively high counts of fungal spores and the presence of bisaccate pollen and thick-walled or strongly-ornamented pteridophyte spores. The freshwater Chlorophyte alga Pediastrum occurs sporadically, but is always rare. This zone appears to correlate with the upper part of the FSM. Zone C is characterized by either an absence of, or very low numbers of, deflandreoid and Apectodinium dinocysts. Insignificant counts of sporomorphs are also recorded. The zone may correspond to the lower part of the FSM (Sla) and the upper part of the Lista (L3).

POL zones Zone I is dominated by Inaperturopollenites, zone II is characterized by high numbers of triporate/ tricolpate pollen (e.g. Caryapollenites) and zone III shows a return to dominance by inaperturate pollen. Zone I appears to correspond to the midSele Formation ($2), zone II with the upper FSM and zone III with the lower FSM-upper Lista interval. Zone II may correspond to the PT 19.4 zone of SchrOder (1992), who also refers to abundant bisaccates, triporates and Caryapollenites, although these characteristics are said to be shared by zone PT 19.1. Greater use of specific pollen taxa would have been made during our study had convenient and up-to-date taxonomic sources been available, and had the validity of this approach been clearly demonstrated in the literature before the end of the study. Oxidation and additional sieving are advisable to optimize angiosperm pollen recovery, and thus a retrospective re-analysis of our slides is probably impracticable.

PF zones Zone 1 is characterized by an increase in the percentage of both palynomorphs (mostly angio-

125

sperm pollen) and AOM, with a sympathetic decrease in the percentage of phytoclasts. Prasinophyte phycomata are more common than in the lower zones, but never frequent; Schr0der (1992, p. 122) notes that they show an upward increase in the uppermost PT 19.4 zone (top FSM), and Knox & Holloway (1992) report an acme in large leiospheres (also prasinophyte algae) just above the FSM. This may reflect decreasing sediment accumulation rates (cf. Tyson 1995). Zone 1 probably coincides with the mid-Sele interval ($2). A transitional 1.5 zone of mixed character was all that could be distinguished in some wells. One well, with an atypically long-cored Sele sequence ($2-?$3), shows signs of renewed terrestrial progradation (increasing phytoclasts) in its upper part (a common feature in the later part of highstand systems tracts). This makes the subdivision of zone 1 a likely prospect. Zone 2 exhibits an upward decrease in the percentage of phytoclasts (by 60-90%), and a corresponding increase in the percentage of AOM, while the percentage of palynomorphs remains fairly constant. It appears to correspond to the upper FSM, especially where this becomes more shale rich. Schr6der (1992, p.l19) notes that the proportion of AOM (SOM) increases upwards from his zone PT 19.3 to PT 19.4. Zone 3 is characterized by a fairly stable kerogen composition, with high phytoclast and low palynomorph percentages. It appears to correspond to the lower FSM. Zone 4 is poorly represented in our data base, but the samples are generally characterized by very high percentages of phytoclasts (mean 87%) and very low percentages of AOM (mean 10%) and palynomorphs (mean 2.5%). The phytoclast fraction shows a relative increase in the proportion of both refractory opaque phytoclasts and resistant structured brown phytoclasts. The shales of zone 4 are distinct by virtue of their blue-green (rather than medium-dark grey) colour, and their low organic content. We believe zone 4 to be equivalent to the upper Lista Formation (L3).

Composite zonation Individually, the three- or fourfold subdivisions achieved with the three different types of zones are of limited application, as they provide only rather poor resolution. However, the three individual zonations can be integrated together to provide a more meaningful composite zonation that offers the prospect of significantly greater resolution. The distribution of the three individual types of zones was plotted v. depth for each well (e.g. Fig. 2), and the stratigraphic pattern of the zones examined for evidence of recurrent associations between

126

s . E . WOOD ~:; R. V. TYSON

n POL Zones O PF Zones [ z~ PNo Zones

E o. 100 o O O

_o 200

I.

II

cQ.

~

300

P O L = Pollen PF = Palynofacies PNo = P a l y n o m o r p h s

i0

i

2'.0

310

Stratigraphic order of zones Fig. 2. Example showing the distribution of the three types of individual zones versus depth in one of the 11 study wells. The PNo, POL and PF zones have all been placed in their stratigraphic order using a common relative numerical scale (1-3); the small lateral offsets between the different types of zones are for diagrammatic purposes only; in this well the first PNo zone is a transitional one (recorded as 1.5). Note how the different combination of the three types of zones defines four stratigraphic intervals.

D. oebisfeldensis

the zones. It was soon clear that there were seven recurrent associations, and that these always occurred in the same relative stratigraphic order (Fig. 3). These recurrent combinations are used to define seven composite zones (CZ1-CZ7). Although this procedure was initially done by eye, we have established that, once the allocation to the individual PNo, POL and PF zones has been achieved, the samples can be objectively assigned to the same composite zones using hierarchical cluster analysis of the individual zone assignments. This leads us to believe that the composite zones are a meaningful summary of our data. Furthermore, only one out of our 467 samples was misclassified (i.e. out of sequence) according to this procedure, and allowing for some reasonable extrapolations where information on only two out of the three types of zones were available for some samples, only 12 samples out of the total of 467 could not be classified. The composite zonation can be applied to nine out of 11 of our study wells, although no more than five CZ zones occur in any one well's cores; the two exceptions are both wells where our sample density was atypically low (38 samples from 500 feet of core). R e f e r e n c e to our lithological descriptions indicate that CZ1 corresponds to the supra-FSM black shales of the Sele Formation ($2?), and that

P~No

CZ1

SELE $2

IPOL

-PNo -[ Upper A. augustum

PFL CZ2

~

SELE $1 u. FORTIES

CZ3

PRo POL t

PF ] C Z 4

rPO,

~

El Lower A. augustum

A. margarita

L

L. FORTIES

cz~ , [PNo

PF ] C Z 7 ' L _ ~ - -

I i

LISTA L3

I Individual Zone Types:

PF = Palynofacies PNo = Palynomorphs POL = Pollen

PF PNo POL

1 A I

1.5 A/B

2 B

II

3 C III

4 C III

Fig. 3. Summary diagram showing how the different combinations of the individual PNo, POL and PF zones are used to define the composite zones CZ1-CZ7. The apparent correlation with the conventional dinocyst zones, based on two wells, is also indicated.

NORTH SEA PALEOGENEPALYNOFACIES CZ7 corresponds to the green claystones of the upper Lista Formation (L3); this leaves up to five zones in the FSM. However, a meaningful test of our approach requires that our composite zones are compared with other conventional subdivisions of the sequence. Unfortunately, we do not have access to the necessary independent information from enough of our study wells to perform a satisfactorily conclusive analysis at the present time. Until such a comparison can be undertaken we cannot be sure of the extent to which some of the composite zones may be influenced by lithofacies or palaeoenvironmental factors, nor can we carry out the necessary 'fine tuning' of the scheme. However, preliminary analysis of biostratigraphic summaries from two of our study wells (21/10-2 and 21/10-5), based on unpublished work by A. J. Powell and kindly provided by BP, confirm that CZ1 does correlate with the D. oebisfeldensis biozone (Sele $2), and also that zones CZ2-CZ5 correspond with the upper A. augustum biozone (upper FSM, PT 19.4-19.2), zone CZ6 corresponds to the lower A. augustum biozone (lower FSM, PT19.1; the Ahy biozone of Powell 1992), and that zone CZ7 mostly corresponds to the Alisocysta margarita (Ama) biozone (Lista L3, PT 15), as shown in Fig. 3. It should be noted that some consecutive pairs of composite zones (CZ3 and CZ4, also CZ6 and CZ7) differ only by virtue of their palynofacies characteristics (see Fig. 3); the distinction between these zones is probably more vulnerable to facies effects than are the other CZ zones. Zone CZ2 is also only distinct from CZ3 by virtue of the occurrence of a transitional PNo zone (A-B, rather than either A or B), which appears to be most commonly developed in distal wells; this zone may, therefore, also partially reflect facies. If these provisos are allowed for, the 'worst case scenario' still leaves two to three CZ zones for the FSM interval. Even if some of the CZ zones do not turn out to be of true chronostratigraphic value, they may still provide correlations which are of significance for facies models. Our palynofacies results, which will be reported in full elsewhere, appear to show a significant correlation between the individual and composite zonations and the four main phases of

127

development of the FSM fan system: namely, the pre-fan oxic, hemipelagic highstand mudstone phase (CZ7), the main sand-rich phase of fan development (lowstand systems tract, CZ6-CZ4), the progressive abandonment and 'shaling-out' of the fan (transgressive systems tract, CZ4-CZ2), and the post-fan anoxic, hemipelagic highstand mudstone phase (CZ1). These facies changes appear to be driven by regional or global relative sea-level changes and, on a local scale at least, can be reasonably expected to be broadly synchronous. However, in sedimentological systems as complex as submarine fans we must expect additional local complexities: channel 'switching' controlling the spatial and temporal distribution of fan lobes and areas of sediment bypass; within-lobe radial channel v. levee and interchannel facies transitions; proximal-distal variations in the scale, extent, and continuity of sand deposition; and potential stratigraphic omission and repetition due to slumping on the fan slope. Actual application of any scheme will therefore require an interdisciplinary approach and full integration with other independent sets of data. We consider that our integrated approach and the resulting composite zonation have the considerable advantage of making the maximum use of all the available types of palynological data. Our preliminary results lead us to believe that quantitative palynofacies investigations may have considerable potential in detailed correlation studies. The possibility of improved preparation and quantification methods, and thus better palynomorph data, mean there is considerable potential for further refinement of the proposed scheme. SEW wishes to thank NERC for receipt of a studentship, and the CASE partner, Enterprise Oil plc, for their additional assistance and support. Enterprise, BP and Shell are gratefully acknowleged for their permission to study and sample central North Sea Paleogene core material. Special thanks are due to BP for providing biostratigraphic summary figures from their two wells. Dr Jeff Goodall is thanked for valuable discussion throughout the study on which this paper is based. Barbara Brown (NRG, Newcastle) is thanked for preparation of the figures. The two anonymous reviewers are thanked for their constructive comments.

References ARMENTROUT,J. M., MALECEK,S. J., FEARN,L.

B. ET AL.

1993. Log-motif analysis of Paleogene depositional systems tracts, Central and Northern North Sea: defined by sequence stratigraphic analysis. In: PARKER,R. J. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 45-57. BOULTER, M. C. & RIDDICK, A. 1986. Classification

and analysis of palynodebris from the Palaeocene sediments of the Forties Field. Sedimentology, 33, 871-886. CARMAN, G. J. & YOUNG,R. 1981. Reservoir geology of the Forties Field. In.: ILLIN6, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of North-West Europe. Heyden & Son, London, 104-109.

128

s . E . WOOD 8r R. V. TYSON

DEN HARTOGJAGER, D., GILES, M. R. & GRIFFITHS,G. R. 1993. Evolution of Paleogene submarine fans of the North Sea in space and time. In: PARKER, R. J. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 59-71. ENJOLRAS, J. M., GOUADAIN, J., MUFTI, E. & PIZON, J. 1986. New turbiditic model for the Lower Tertiary sands in the South Viking Graben. In: SPENCER, A. M. ET AL. (eds) Habitat of Hydrocarbons on the Norwegian Continental Shelf. Graham & Trotman, London, 171-178. KNOX, R. W. O'B. & HOLLOWAY,S. 1992. 1. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the North Sea. British Geological Survey, Nottingham. KULPECZ, A. A. & VAN GEUNS, L. C. 1990. Geological modeling of a turbidite reservoir, Forties Field, North Sea. In: BARWIS,J. H., MCPHERSON, J. G. & STUDHCK, J. R. J. (eds) Sandstone Petroleum Reservoirs. Springer-Verlag, New York, 489-508. MILTON, N. J., BERTRAM,G. T. & VANN, I. R. 1990. Early Paleogene tectonics and sedimentation in the Central North Sea. In: HARDMAN, R. F. P. & BROOKS, J. (eds) Tectonic Events Responsible For Britain's Oil and Gas Reserves. Geological Society, London, Special Publication, 55, 339-351. MORTON, A. C., HALLSWORTH,C. R. & WILKINSON, G. C. 1993. Stratigraphic evolution of sand provenance during Paleocene deposition in the Northern North Sea area. In: PARKER, R. J. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 73-84. MUDGE, D. C. & COPESTAKE, P. 1992. Revised Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. O'CONNOR, S. J. & WALKER, D. 1993. Paleocene

reservoirs of the Everest trend. In: PARKER, R. J. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 145-160. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Palaeocene and earliest Eocene sediments from the Central North Sea. Review of Palaeobotany and Palynology, 56, 327-344. 1992. Making the most of microfossils. Geoscientist, 2 (1), 12-16. SCHRODER, T. 1992. A palynological zonation for the Paleocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-126. STEWART, W. 1987. A revised stratigraphic interpretation of the Early Palaeogene of the Central North Sea. In: BROOKS, J. & GLENNIE, K. W. (eds). Petroleum Geology of North-West Europe. Graham & Trotman, London, 1, 557-576. THOMAS, A. N., WALMSLEY, P. J. & JENKINS, D. A. L. 1974. Forties Field, North Sea. Bulletin of American Association of Petroleum Geologists. 58, 396-405. TYSON, R. V. 1993. Palynofacies analysis. In: JENKINS, D. G. (ed.) Applied Micropalaeontology. Kluwer, Dordrecht, 153-191. -1995. Sedimentary Organic Matter: Organic Facies and Palynofacies. Chapman & Hall, London. VINING, B. A., IOANNIDES,N. S. & PICKERING,K. T. 1993. Stratigraphic relationships of some Tertiary lowstand depositional systems in the Central North Sea. In: PARKER, R. J. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 17-29. WHYATT, M., BOWEN, J. M. & RHODES, D. N. 1992. Nelson - successful application of a development geo-seismic model in North Sea exploration. In: SPENCER, A. M. (ed.) Generation, Accumulation and Production of Europe's Hydrocarbons II. Special Publication of the European Association of Petroleum Geoscientists, 2, 3-21.

Chronostratigraphic framework for the Thanetian and lower Ypresian deposits of southern England J A S O N R. ALI 1 & DAVID W. J O L L E Y 2

1 Department of Oceanography, The University, Southampton S 0 9 5NH, UK (Present address: Earth Sciences, UNIMAS, 94300 Kota Samarhan, Sarawak, Malaysia) 2 Centre for Palynological Studies, The University, Sheffield S1 3JD, UK Abstract: Magnetostratigraphic and palynomorph data are used to assess the timing and nature

of the late Paleocene-earliest Eocene depositional sequences of SE England. Ormesby Clay Formation mudstones in east Norfolk are the oldest upper Paleocene deposits (C26r) of southern England. The base of the type Thanet Sand Formation (= start of the Thanetian Stage) is c. 0.65 million years younger (C26n). A subaerial erosion surface separates the Thanet-Ormesby package from the Lambeth Group sediments. The duration of the hiatus (> 0.5 million years) separating the two units can be calculated from the magnetostratigraphicdata; C25n is missing from the stratigraphic record across almost all of southern England. However, the Upnor Formation in central London contains a record of C25n. A hiatus of c. 0.4 million years separates the late Paleocene Lambeth Group from the early Eocene Harwich Formation. The Lambeth Group, Harwich Formation and lower London Clay Formation were all deposited during C24r. The start of Chron C24n.3n is positioned at the base of Division B of the London Clay Formation.

The majority of the internationally recognized Paleogene stages were originally defined in outcrops in the North Sea Basin. Each of the stages was related to a lithostratigraphic unit; however, the base of many of these units corresponds to an unconformity. Stage boundaries have been extended across the globe by relating the base of each of the stratotypes to the standard biostratigraphic zonation framework. In recent times, the increased precision required by stratigraphers has led to demands for boundary stratotypes that more easily facilitate global correlations. Thus it is likely that many, if not all, of the Paleogene stage boundaries will eventually be 'relocated" in sections outside the North Sea Basin. For example, the early Paleocene Danian Stage, originally defined by Desor (1847) in an outcrop at Stevns Klint, Denmark, is today located in an outcrop at E1 Kef, Tunisia (Jenkins & Luterbacher 1992). Minimizing the age discrepancy between the level originally used to define the start of a stage and the point chosen in an alternative section to mark the start of the newly defined standard stage will ensure that the stage remains a valuable chronostratigraphic concept. The late Paleocene Thanetian Stage (defined in Kent, England) and the early Eocene Ypresian Stage (defined in Belgium) bracket the PaleoceneEocene boundary (Hardenbol & Berggren 1979; Jenkins & Luterbacher 1992). Unfortunately, critical portions of the uppermost Thanetian and lowermost Ypresian stratotypes lack key calcareous

microlossils and nannofossils, and linking these intervals to the global marine record is problematic (see discussion in Ali et al. 1993). Nannofossil zone NP9 is poorly represented and NP10 is not known. Dinoflagellate cysts have been used to provide a second-order correlation with upper NP9 and NP 10 in the stage stratotype areas. Magnetostratigraphic studies potentially provide useful datums for linking the Paleocene-Eocene boundary to the North Sea Basin sequences, in particular the levels corresponding to the start and end of Chron C24r. Thus, if the boundary is eventually located in a section outside NW Europe, palaeomagnetic investigations of the Thanetian and Ypresian stratotypes will provide valuable reference markers. Magnetostratigraphic studies of the Thanetian and Ypresian stratotypes also provide a chronostratigraphic framework from which it is possible to assess the timing and duration of the upper Paleocene and lower Eocene depositional sequences of the southern North Sea Basin. One of the principal research programmes of the Southampton University Palaeomagnetic Group is to determine the synchroneity of late Paleocene and early Eocene depositional sequences from a number of passive margins around the globe. The object of this programme is to provide a rigorous test of the eustatic sea-level model for a 12-15 million years window of geological time. The present study marks the completion of the investigations for the upper Paleocene of the southern

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlationof the Early Paleogene in NorthwestEurope, Geological Society Special Publication No. 101, pp. 129-144.

129

130

J.R. ALl • D. W. JOLLEY

North Sea Basin. (Studies of the lower Eocene from the region have recently been presented by Ali et al. 1993.)

Depositional setting During the late Paleocene and early Eocene the North Sea extended across much of SE England and NW mainland Europe (see Ziegler 1982). As the sea level fluctuated, the shoreline periodically advanced and retreated across the region. The upper Paleocene and lowermost Eocene sediments preserved in southern England (Fig. 1) comprise outer neritic, inner neritic, marginal-marine and fluviatile sediments.

The upper Paleocene and lower Eocene of southern England Many of the lithostratigraphic units referred to in this paper were introduced last century (e.g. Prestwich 1850, 1852, 1854; Whitaker 1872). Cooper (1976), Curry et al. (1978) and King (1981) have revised and formalized some of these terms. A major review of the stratigraphic nomenclature of the upper Paleocene and lower Eocene of southern England (summarized in Fig. 2) has recently been carried out by Ellison et al. (1994). O r m e s b y Clay F o r m a t i o n

The oldest Tertiary sediments in southern England (Chron C26r) are assigned to the Ormesby Clay

Fig. 1. Outcrop and subcrop of the Paleogene deposits of SE England, northern France and Belgium, and the location of boreholes and outcrops referred to in this study.

THANETIAN/YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND

131

The formation is present in a triangular area between NE Kent, south Greater London and south Suffolk. The formation attains over 30 m thickness in north Kent, thinning to a few metres in south Suffolk, where it passes laterally into the Ormesby Clay Formation (middle and upper part). The Thanet Formation spans nannoplankton zones NP6-NP8 and magnetochrons C26n to C25r (Knox et al. 1994). Curry (1981) suggested that a significant part of the Thanet Formation was removed by erosion prior to deposition of the Woolwich & Reading Beds (Lambeth Group). L a m b e t h Group: U p n o r F o r m a t i o n

The Lambeth Group comprises three formations. The lowermost unit, the Upnor Formation, was previously referred to as the Woolwich Bottom Bed (Ellison et al. 1994). The Upnor Formation sediments were deposited in a shallow marine environment. The unit is typically 2-5 m thick unit and consists of fine- to medium-grained sands and silts with occasional pebbles and is rich in glauconite. It is present across the London Basin and the eastern part of the Hampshire Basin. Calcareous nannoplankton obtained from the Upnor Formation in both the Hampshire Basin (Siesser et al. 1988) and the London Basin (Ellison et al. 1996) indicate an NP9 age. Fig. 2. Summary of the upper Paleocene and lower Eocene stratigraphy of the southern England (based on Eilison et al. 1994 and Jolley 1996). Correlation with the Martini (1971) nannoplankton zonation-schemeis based on Aubry et al. (1986), Siesser et al. (1988), Knox et al. (1994) and Ellison et al. (1996).

Formation (Cox et al. 1985; Knox et al. 1990). The formation is restricted to eastern East Anglia where it is known only from boreholes. It consists of glauconitic mudstone. Sporadic, thin, ash beds occur in the lower part of the formation. The unit is over 25 m thick in east Norfolk, thinning to c. 10 m in Suffolk. A red-brown mudstone in the lower middle part of the formation in Norfolk constitutes a useful regional marker. It can be traced across Suffolk and southwards into Essex and Kent, where it is present as a sandstone just above the base of the Thanet Sand Formation (Knox et al. 1994). The upper part of the Ormesby Clay Formation is assigned to Chron C25r. Thanet S a n d F o r m a t i o n

The Thanet Sand Formation consists of bioturbated silts and fine sandstones which are rich in glauconite. The sediments were deposited in an inner-shelf to coastal setting (Ellison et al. 1994).

L a m b e t h Group: W o o l w i c h a n d Reading formations

The laterally equivalent Woolwich and Reading formations were deposited over a large part of southern England. The Woolwich Formation occurs in the eastern half of the London Basin and consists of grey and grey-brown interlaminated sands, silts and clays. The formation is rich in plant debris, and includes thin lignites. Shell beds, up to 10-30 cm, occur. The formation is typically 10-15 m thick and represents a variety of marginal marine, low to high energy environments, with occasional freshwater intercalations (Ellison et al. 1994). The Reading Formation (fluviatile facies) comprises red mottled clays with thin sandy units. In the western London Basin and the Hampshire Basin the formation is typically 25-40 m thick. The Woolwich and Reading Formations interdigitate in the London area. The Woolwich and Reading Formations are also represented in north France, where they are referred to as the Sparnacian. T h a m e s Group: H a r w i e h F o r m a t i o n

Ellison et al. (1994) include in the Harwich Formation all the sediments between the top of

132

J . R . ALI & D. W. JOLLEY

the Lambeth Group and the base of the London Clay Formation Walton Member. The dominant lithologies of the Harwich Formation are glauconitic sandy clays and glauconitic fine sands. Tephras, up to 4 cm thick, are common in the upper part of the formation in north Essex and Suffolk. Elsewhere, disseminated ash occurs sporadically throughout the formation. The Harwich Formation was deposited in a shallow-shelf environment. The formation is typically 15-20 m thick, although it attains over 40 m in eastern Norfolk. In parts of Greater London the Harwich Formation is absent. Correlation of the Harwich Formation tephras with those in Goban Spur DSDP sites provides a second order link to the early Eocene NP10 nannoplankton zone (Knox 1984). The unit is assigned to the uppermost part of the Apectodinium hyperacanthum dinoflagellate zone, which Powell (1992) considers to be of early Eocene age.

Thames Group: London Clay Formation (basal part) Ellison et al. (1994) positioned the base of the London Clay Formation at the base of the Walton Member. The transgression surface marking the base of this member has been identified over the whole of the London Basin and the eastern part of the Hampshire Basin (King 1981). The sediment consists of silty clays; no ash layers have been recorded in this unit. In the eastern London Basin the Walton Member is typically c. 15 m thick, thinning westwards and southwards to c. 5 m at Whitecliff Bay, Isle of Wight. The first appearance datum (FAD) of the dinoflagellate Wetzeliella astra is positioned at c. 2 m above the base of the Walton Member at the type locality (King 1981). The Walton Member is of early Eocene age (?NP10).

Palynomorph association sequences Interpretation of the palynological data used in correlation for this study followed the method outlined by Jolley (1992, 1996), involving the recognition of palynofloral associations. Associations, i.e. groups of assemblages with shared characteristics of composition, were defined by empirical inspection of the data, with the aid of cluster analysis and detrend correspondence analysis. This enabled identification of groups of successive assemblages which showed shared characteristics of composition. In some cases, lower density sampling resulted in assemblages from single samples being referred to as associations, where the preceding and succeeding assemblages showed a significant difference in composition. Associations were defined individually for each section, without reference to other

studied sections. Each association was subsequently assigned a name based on the specific epithets of two of its characteristic taxa. In correlating the individual sections, comparison of associations was applied only between nearest neighbours which provided the most complete record of the palynofloras. This method of comparison between adjacent sections was utilized to minimize the effects of any biofacies control on the palynofloral assemblages. Neighbouring sections contain palynofloras representative of similar biofacies, whereas those from a distant section may contain biofacies which would make reliable correlation difficult. From the comparisons of the associations it is possible to identify laterally continuous, stratigraphically comparable sequences of associations within the study area. These are termed 'association sequences', and are regarded as containing palynofloras deposited in sediments of equivalent age. Each association sequence identified here is numbered (oldest to youngest), the number being prefixed by the letter T (Thanetian) and Y (Ypresian). In all, nine T association sequences and seven subsequences were identified (Jolley 1992), together with nine Y association sequences and ten association subsequences (Jolley 1996).

Palaeomagnetic methods The samples used in this study were collected between 1987 and 1993. Most of the analysed specimens are either 8 or 14 cm 3 cubes, although the recent acquisition of a whole-core cryogenic magnetometer has allowed measurements to be carried out on core pieces of 150-300 cm 3. Samples were collected from each section with a stratigraphic spacing of 0.3-1.0 m. As most of the sections are from boreholes, the specimens could not be oriented with respect to north, and were 'way-up' oriented only. Approximately half of the specimens were measured using a Molspin spinner magnetometer. The alternating field (a.f.) method of demagnetization was applied to specimens at steps of 5 mT rising to peak values of 30-60 mT using a Molspin tumbling-specimen system. The remaining specimens were examined using a 2G Enterprises cryogenic magnetometer, which has an in-line threeaxes demagnetization system capable of generating fields of 60 roT. Both Zijderveld (1967) plots and equal-area stereographic plots were used to determine the stability of remanence of each specimen in order to define the characteristic direction. Examples of response to a.f. demagnetization are shown in Fig. 3. The data are categorized in the manner described by Ali et al. (1993). A stable end-point direction (SEPD) is defined when a high

13 3

THANETIAN/YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND

(a) BRADWELL: 217

j/jol~

0t 0

(b)

F

BD4

J l0

,

F

SAMPLENO Jo (mA/m) BD4 15.73

~

~

BRADWELL: 217

,

J

r

'

F

20 DcmagFicld(naT)

'

30

;

"

40

BD42

SAMPLENO Jo (naA/m) BD42 18.40

j/jol? ~ . . ~ . . . . X . _ . ~ _ X _ _ _ ~ X 0I 0 10 20 DernagField(mT) ,

,

~

J

,

X ~

30

X-----X J__~ 40

Fig. 3. Examples of responses of typical samples to alternating field demagnetization. (a) Stable end-point (SEP) behaviour; (b) directional trend but with no SEP. In each case the magnetic vector after each demagnetization step is plotted on a stereographic projection (declination is arbitrary because the specimens are from unoriented borehole cores). On the stereographic projections solid symbols represent positive inclinations (plotted in the lower hemisphere) and open symbols represent negative inclinations (plotted in the upper hemisphere). The values in the range 0-40 represent the applied field treatment (in mT). Also shown for each specimen is a plot of the normalized magnetic intensity (J/Jo) v. applied field (mT).

stability component of magnetization is isolated in a specimen. Specimens which do not achieve an SEPD, but for which a reliable polarity determination can be made based on the trajectory of the remanence vector towards a particular polarity state, are referred to as 'trending'. Specimens which exhibit wide fluctuations in both direction and/or intensity between steps are classified as 'erratic'. Such behaviour is most commonly observed in very weakly magnetized sediments where the magnetic signal may be of comparable magnitude to the noise level of the magnetometer. On the magnetostratigraphic log for each section SEPD, trending and erratic behaviour are indicated. As none of the borehole cores are oriented, the polarity assignment for each specimen is based on the dip of the characteristic remanence inclination angle (downward directed indicates normal polarity, whereas upward dipping indicates reverse polarity). Isothermal remanent magnetization (IRM)

analyses were performed on representative specimens in order to determine the principal remanence carriers in each of the boreholes sampled. A Molspin pulse magnetizer, with a peak direct field of 0.86 T was used to generate the IRM. The IRM was measured between each of the 14 progressive field increments using a Molspin spinner magnetometer. Two distinct types of behaviour were observed. In the first case (e.g. Fig. 4a & b) the magnetization saturates in an applied field of c. 0.3 T, suggesting that magnetite is present in the specimen. In the second case (e.g. Fig. 4c & d), the magnetization does not saturate at the peak field, indicating that the remanence is carried by iron oxide in a higher oxidation state. The 'IRM ratio', defined by Ali (1989) as the IRM at 0.3 T as a ratio of the IRM at the peak field, is a convenient method for quantifying the shape of the IRM curve. Magnetite-rich specimens have IRM ratios typically >0.9 whereas for hematite-bearing sediments the value is typically 0.6-0.9.

134

J.R. A L I & D . W. JOLLEY

4000

2000

0

,

0.0

0.2

i, 0.4

a.

b.

JU47

SZ32

(0.92)

(0.97)

i r i 0.6 0.8

0

0.0

i, 0.2

i i i, 0.4 0.6

C.

400

V 0

i 0.8

d.

(0.76)

V

I'

I'

I'

I'

I

0.0

0.2

0.4

0.6

0.8

APPLIED FIELD (T)

o

_ (0.70) .

l,

i,

i

0.0

0.2

0.4

, i 0.6

~ i 0.8

APPLIED FIELD (T)

Fig. 4. Examples of IRM acquisition curves for magnetite-bearing (a & b) and hematite-bearing (c & d) sediments. IRM is expressed in mA m2; the figure in parenthesis is the specimen's IRM ratio.

Palaeomagnetic results Ormesby, Norfolk

The stratigraphy of the Paleogene succession in the British Geological Survey (BGS) Ormesby borehole (TG 5148 1424) was reported on by Cox et al. (1985). In ascending stratigraphic order, the following units (names from Ellison et al. 1994 and Jolley 1995) were cored: Ormesby Clay Formation (27.45m), Hales Clay (Harwich Formation) (14.40m) and Wrabness Member (Harwich Formation) (27.80m). The palaeomagnetic interpretations published in Cox et al. (1985) were based on 94 8 cm 3 palaeomagnetic specimens collected from the Ormesby core. Following Cox et al. (1985), Townsend & Hailwood (1985) and Aubry et al. (1986), Knox et al. (1990) were able to provide a fuller interpretation of the Ormesby borehole magnetostratigraphic sequence. Their principal findings were: (1) the Ormesby Clay Formation spans Chrons C26r to C25r; (2) the unconformity separating the Ormesby Clay from the Hales Clay is marked by the absence of Chron C25n, indicating a stratigraphic break of > 0.5 million years; (3) the Hales Clay and Wrabness Member were deposited during the early part of Chron C24r. In this study, a further 19 specimens were collected from the Ormesby core in order to refine the magnetostratigraphy presented by Cox et al. (1985). Sixteen samples were collected from the Wrabness Member, one sample from the Hales Clay and two from the Ormesby Clay Formation. Twelve samples from the upper part of the Wrabness Member had initial natural remnant magnetization (NRM) intensities of < 0.6 m A m -1.

Samples from below c. 79 m below datum (m.b.d.) had values of 20-40 mA m -t. Ten specimens demagnetized to a stable end-point direction, with eight specimens exhibiting well-defined trend directions. The magnetostratigraphy of the Ormesby section is shown in Fig. 5 (modified from Cox et al. 1985, fig. 5). The revised polarity sequence includes a number of thin (c. 1 m) normal polarity intervals, based on one or two samples. These had previously been identified in a study by Johnston (1983), but were omitted from the Cox et al. (1985) synthesis. The normal polarity intervals identified in the Ormesby core have been relabelled OR-A to OR-H, in ascending stratigraphic order. However, while some of the thin normal polarity intervals reported by Johnston (1983) have been confirmed, several are somewhat thinner than previously thought. Johnston also identified a short reversed interval in the upper part of magnetozone OR-B (correlated with Chron C26n). This has now been confirmed by data from two samples. Hales, Norfolk

A multidisciplinary stratigraphic investigation of the BGS Hales borehole (TM 367 969) was carried out Knox et at. (1990). Here we present the details of the palaeomagnetic investigations. The Hales borehole recovered 45.77 m of Paleogene sediments. The following rock units (names from Ellison et al. 1994 and Jolley 1996) were present: Ormesby Clay Formation (25.59 m), Hales Clay (Harwich Formation) (15.62m), and Wrabness Member (Harwich Formation) (4.96 m). One hundred and eleven 8 cm 3 specimens were

135

THANETIAN/YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND FM. PAS POL. 70--

Y9

-

I

OR-H OR-G

Y8

co "-" 09 Y7c LM Y7b' Z

80--

133

,OR-F 9 OR-E

Y7a

90--

Y6b i Y6a

T

100 -

d

E

!y_~4

OR-D

O Y3c

G)

O

110--

._J _ I Y3b Y3a T9

~

,,= 120

OR-C

reverse polarity specimens within each formation record similar inclination values; (4) the Ormesby Clay Formation and the Hales Clay have quite different mean inclination angles (45.8 and 29.3 ~ respectively). The palaeogeographic maps of Smith et al. (1981) predict a southern UK late Paleocene palaeolatitude of 42 ~ N, which equates to a geomagnetic field inclination angle of 61 ~ followed by a steady northward drift to its present-day site (c. 51.5 ~ N). The mean inclination of the Hales Clay is approximately half that predicted by Smith et al. (1981). Assuming that the SEPD represent primary magnetizations (the Hales Clay and Ormesby Clay both include normal and reverse polarity specimens which suggests this is the case), and that geomagnetic secular variation has

T8

15 T7 i

f

OR-B >-

a-b T!

130 -

T3

T~I

UNIT PAS POL.

OR-A

I

-90

i

'

0

~

-r"

Y3a I

90

Fig. 5. Magnetostratigraphy of the upper Paleocene and lower Eocene at Ormesby, Norfolk. S, T and E denote stable end-point, trending or erratic directions, respectively. Data from Johnston (1983) are marked +. Palynomorph association sequences (PAS) from Jolley (1992, 1996). Normal polarity, black; reverse polarity, white. Normal polarity intervals are coded OR-AOR-H (ascending).

,.-.., vE 7I-O. UJ 0

HC1

Tx

OC4

T9 ~"

O

HL-E

HL-D

T7e

<~O TTd

o I i10~10

collected from the core, with an average spacing of 0.4 m. About one third of the specimens yielded stable end-point directions; almost all of the remaining specimens had polarity determinations based on the trend of their demagnetization vector. The magnetostratigraphy of the Hales borehole is summarized in Fig. 6. Table 1 shows the averaged inclinations for stable end-point specimens in the Ormesby Clay Formation and the Hales Clay. From Fig. 6 and Table 1, we draw attention to the following: (1) specimens with SEPD are fairly evenly spaced through the succession; (2) specimens with 'trending' directions generally have inclinations similar to those with stable directions, suggesting that the information recovered from them provides reliable polarity data; (3) normal and

HL-F

Y3b

I

Inclination

Y4~

IIO~1 "1- - - I ~ m Y3c

T1

140-

Y5 Y4b

TTb

HL-C

a-b I

T4

HL-B HL-A

-90 Inclination 0 90 Fig. 6. Magnetostratigraphy of the upper Paleocene and lower Eocene at Hales, Norfolk. S denotes a stable end-point direction, + indicates either trending or erratic directions. Palynomorph association sequences (PAS) from Jolley 0992, 1996). Normal polarity, black; reverse polarity, white; indeterminate polarity, hatched. Normal polarity intervals are coded HL-A-HL-E

136

J.R. ALI • D. W. JOLLEY Table 1. All stable end-point data from the Ormesby Clay Formation and Hales Clay (Harwich Formation)from the Hales borehole

Formation Ormesby Clay Hales Clay

Normal 46.7 (_+ 10.0)~ 33.5 (__3.1)~

been fully averaged, then possible explanations for the anomalously shallow values associated with the Hales Clay include: (1) northward offset of the geomagnetic dipole (Hailwood 1977) during the deposition of this unit; (2) shallowing of the primary remanence during deposition; and (3) preferential compaction causing flattening of the remanence of the Hales Clay. The third explanation is considered to be the most likely explanation for the anomalous inclinations (the possibility that the site underwent c. 15 ~ southward drift within the late Paleocene is not considered), although no information is known about the physical properties of the Hales Borehole material. The normal polarity intervals recognized in the Ormesby Clay Formation and Harwich Formation Hales Clay in the Hales borehole have been labelled HL-A to H L - E The normal polarity intervals have varying stratigraphic thicknesses, varying from several tens of centimetres (defined by one or two specimens) to several metres (based on tens of specimens). The lithostratigraphic succession in the Hales borehole is similar to that at Ormesby. On this basis, magnetozone HL-C is correlated with Chron C26n (see discussion in Knox et al. 1990). Deposition of the Ormesby Clay Formation spanned geomagnetic Chrons C26r to C25r. The unconformity which separates the Thanet SandOrmesby Clay Formation from younger sequences (see Cox et al. 1985; Aubry et al. 1986), is located between the Ormesby Clay and the Harwich Formation Hales Clay. The Hales Clay was deposited during Chron C24r. As is the case with the Ormesby borehole, the Hales core contains a number of thin (c. 1 m) of normal polarity intervals within reverse polarity magnetozones.

Reverse

Combined

-45.3 (_+ 10.3)~ -28.4 (+ 19.3)~

45.8 (_+ 10.3)~ 29.3 (+ 8.7)~

Clay (Harwich Formation) (3.75 m), and Wrabness Member (Harwich Formation) (5.94 m). Fortyseven 8 cm 3 specimens were collected from the succession with an average spacing of 0.5 m. The Ormesby Clay, Hales Clay and Wrabness Member had NRM intensities typically between 5 and 30 mA m -1 (Fig. 7). The Woolwich and Reading formations had notably lower intensities, with values typically 0.2-2.0 mA m -1. Thirty-seven specimens produced SEPD nine produced trending directions and one produced an erratic demagnetization. Two normal polarity magnetozones have been identified in an otherwise reversely magnetized sequence (Fig. 7). The lower one,

Halesworth, Suffolk

A borehole recently drilled by the BGS at Halesworth (TM 4178 7627) has provided a further section through the early Paleogene succession of East Anglia. The following lithostratigraphic units were encountered (A. N. Morigi, pers. comm.): Ormesby Clay Formation (6.16 m), interdigitated Woolwich and Reading Formations (7.34 m), Hales

Fig. 7. Magnetostratigraphy of the upper Paleocene and lower Eocene at Halesworth, Suffolk. S denotes a SEPD; + indicates either trending or erratic directions. Normal polarity, black; reverse polarity, white; indeterminate polarity, hatched. Also shown is downhole plot of NRM intensity.

THANETIAN/YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND HW-A, is positioned at the base of the Ormesby Clay Formation (36.16-35.50 m.b.d.). The second, unnamed, normal polarity magnetozone spans the Wrabness Member (13.81-18.24 m.b.d.). Magnetozone HW-A is considered to represent the latter part of Chron C26n as it is associated with a distinctive red mudstone, OC2, of the Ormesby Clay Formation (see Knox et al. 1990; Jolley 1992). The normal polarity magnetization identified in the lower part of the Wrabness Member is unexpected; litho- and biostratigraphic information suggests an earliest Eocene age. It is unlikely that this magnetozone represents either Chrons C25n (the magnetozone is too young, see Ellison et al. 1996) or C24n.3n (the magnetozone is too old, see Ali et al. 1993). Possible explanations are: (1) the magnetization results from a short-period geomagnetic field inversion or (2) the Wrabness Member carries a weathering induced chemical remanence. The former is considered unlikely as similar stratigraphic levels elsewhere do not record such an event. The latter explanation is more likely, although changes in magnetic properties

137

(see NRM intensity plot in Fig. 7), which are often diagnostic of such behaviour, were not observed. Sizewell, S u f f o l k

The BGS Sizewell core C3 (TM 475 620) recovered the following lithostratigraphic units: Ormesby Clay Formation (10.15m), Upnor Formation (3.0 m), Woolwich Formation ( 10.2 m), Orwell Member (Harwich Formation) (3.5 m) (Jolley 1996), and Wrabness Member (Harwich Formation) (8.9 m). Sixty-two 14 cm 3 palaeomagnetic specimens were collected from the core. Initial NRM intensities were typically 530 m A m -j except in the interval of 76.461.8 m.b.d. (from the upper part of the Thanet Sand Formation to the middle part of the Woolwich Formation), where values were typically < 1.0 m A m -1 (Fig. 8). Forty-four specimens produced SEPD, 13 produced trending directions and 5 produced erratic trajectories. The succession is of dominantly reverse polarity (Fig. 8). In the Thanet Sand Formation a normal

Fig. 8. Magnetostratigraphy of the upper Paleocene and lower Eocene at Sizewell, Suffolk. Orwell Member (of the Harwich Formation) from Jolley (1996). Orwell Member, PAS Y3b-Y5; Wrabness Member, PAS 6b-Y8. On the inclination plot, S denotes a stable end-point direction; + indicates either trending or erratic directions. Normal polarity, black; reverse polarity, white. A mixed interval of positive and shallow inclinations is shown hatched. Also shown are downhole plots of NRM intensity and IRM ratio.

138

J.R. ALl • D. W. JOLLEY

polarity magnetozone, SZ-A, is defined by six specimens spanning 82.07-80.61 m.b.d. This normal polarity magnetization is associated with the Ormesby Clay red mudstone. As is the case with the Hales, Halesworth and Bradwell cores the normal polarity remanence identified in the red mudstone is correlated with Chron C26n. Specimens from 75.03-72.32 m.b.d, in the upper part of the Thanet Sand Formation show either indeterminate or normal polarity magnetizations. Normal polarity magnetization is not predicted for this interval, which equates with the lower to middle part of NP8. Elsewhere (e.g. Herne Bay, Aubry et al. 1986; Hales, Knox et al. 1990) a record of Chron C25r has been identified at this level. As noted above, magnetic intensities in the upper part of the Thanet Sand Formation are somewhat lower than in the main part of the formation. Isothermal remnant magnetization (IRM) analysis was carried out on eight representative Thanet Sand Formation specimens (e.g. Fig. 4). Upsection, an abrupt change from low to high coercivity carriers occurs at 76.4 m.b.d. (Fig. 8). We suggest that the zone of mixed indeterminate and normal polarity specimens in the upper part of the Thanet Sand Formation is not a primary remanence. We postulate that the sediments in this interval were deposited during Chron C25r, but due to later weathering and alteration their remanence was acquired during Chron C25n. Such an explanation seems plausible because prior to deposition of the Lambeth Group the Thanet Sand Formation sequences were subaerially exposed (Curry 1981). Correlation of the reversely magnetized sediments in the Woolwich and Harwich Formations with Chron C24r is proposed.

81.38 m.b.d.) carry a normal polarity magnetization. In borehole 217, a normal polarity magnetozone was identified in the lower part of the Thanet Sand Formation, between 82.22 and 74.74 m.b.d. Specimens from 73.51 and 73.24mbd exhibit shallow inclinations (< 4~ Reversely magnetized specimens were identified from 72.28 m.b.d, to the top of the section. Data from boreholes 209 and 217 have been used to construct a composite magnetostratigraphic section (Fig. 9) The normal polarity magnetozone in the lower part of the Thanet Sand Formation in the two holes record portions of the same normal polarity magnetochron, referred to as BD-A. The top of magnetozone BD-A is positively identified in

Bradwell, Essex

The stratigraphy of the Thanetian succession at Bradwell was reported on by Knox et al. (1994). The Thanet Sand Formation was shown to contain nannoplankton zones NP6-8. A normal polarity magnetozone, c. 9 m thick, in the basal part of the sequence, was correlated with Chron C26n. In this paper we present the details of the palaeomagnetic investigations. The magnetostratigraphy of the lower part of the Thanet Sand Formation at Bradwell was determined from two cores (boreholes 209: TM 0103 0889 and 217: TM 0177 0916). Twenty hand-sized specimens were sampled from the two cores. The specimens had NRM intensities ranging between 8 and 8 5 m A m -l. Nine specimens produced SEPD. Demagnetization of the remaining 11 specimens produced trending directions. In borehole 209, all seven specimens from the lowest part of the Thanet Sand Formation (84.51-

Fig. 9. Composite magnetostratigraphyof the lower part of the Thanet Sand Formation in boreholes 209 and 217, Bradwell, Essex. No palynomorph data available. The Chalk/Thanet Sand Formation contact is used as datum. S and T denote stable end-point and trending directions, respectively.Normal polarity, black; reverse polarity, white; indeterminate polarity, hatched.

THANETIAN/YPRESIAN

MAGNETOSTRATIGRAPHY,

FM. POL.

hole 217, where it is arbitrarily positioned at 73.38 m.b.d., 8.84 m above the Thanet Sand Formation/Chalk unconformity. Knox et al. (1994) correlated BD-A with the upper part of Chron C26n, as the sediments at these levels contain Zone NP6 nannoplankton.

24 1--

g

As part of the civil engineering works related to London Underground's extension of the Jubilee Line, and the Channel Tunnel Rail Link, a number of boreholes were drilled through the lower Paleogene of central and south London. Material from four boreholes through the Lambeth Group and basal London Clay Formation have been the subject of a lithostratigraphic, palaeomagnetic and biostratigraphic investigation (Ellison et al. 1996), summarized in Fig. 10. Their principal findings were: (1) identification of the NP9 nannoplankton zone in the basal Upnor Formation; (2) identification of Chron C25n in the lower part of the Upnor Formation (the first time this magnetochron has been identified in SE England); and (3) the Woolwich and Reading Formations and the Harwich Formation were deposited during Chron C24r.

,,E

16

c-

o 12

o0. >, o t-~

/!

~ 8 ID :t:

4

<, --J

r

0

'1'1'111 -90

0

Inclination

Townsend & Hailwood (1985) presented palaeomagnetic data from the lower Paleogene succession at Alum Bay and Whitecliff Bay. A reverse polarity magnetization spanning the Reading, Harwich and lower London Clay Formations was identified in both outcrops. A normal polarity magnetization was identified in Division B of the London Clay Formation. Aubry et al. (1986) correlated the reverse polarity interval with Chron C24r and the normal polarity interval with Chron C24n.3n.

Varengeville, north France Palaeomagnetic data from the lower Eocene part of the section at Varengeville, north France, were presented by Ali et al. (1993). Ali (1989) presented data from Sparnacian sediments (= Woolwich Formation facies). A reverse polarity magnetization was identified in this formation; correlation with Chron C24r was proposed.

Belgium Palaeomagnetic data from the lower Eocene of Belgium were presented by All et al. (1993). Data from the basal Ieper Clay Formation are unreliable.

I I

f

20

Central London

Whitecliff Bay, Isle o f Wight

139

SE ENGLAND

90

0.2 0.4 0.6 0.8 1.0

IRM ratio

Fig. 10. Magnetostratigraphy of the upper Paleocene and lower Eocene in central and south London. + denotes a stable end-ponit direction, filled circles indicate either trending or erratic directions. Normal polarity, black; reverse polarity, white; indeterminate polarity, hatched. Also shown is a downhole plot of IRM ratio.

The normal polarity magnetozone in the Wardrecques Member type section is correlated with Chron C24n.3n.

Summary of the magnetochron correlations The magnetostratigraphy of the upper Paleocene and lowermost Eocene of southern England, northern France and Belgium is summarized in Fig. 11. Data from nine areas are presented, some representing composite sections (e.g. Essex), others individual outcrops or boreholes. The geomagnetic polarity timescale of Cande & Kent (1995) has been used to age-calibrate the magnetochron boundaries. Martini's (1971) nannofossil scheme has been positioned against this scale using the fossil zone-magnetochron correlations of Berggren et al. (1985, 1995).

140

J.R. ALI&

D. W. JOLLEY

~C 3NBOO3 A7HV"3 1

3N3OO:::17Vcl 31V7

ILl ~

0

rY w

~..

'E.~ ,..-L

IZ LU

0

LL! CD

~

:E

~

0

~Om~ >~>,,,~

-~E o o.,

I

~ > ~ Zw~OO~z

I--

~ ~ . ~ . ~. ~

m~ i-

~

l EISISS:IC)OVN I

+

~.~ ~ - ~

§

~vW

viva R78VI7BNNO

C]:~'IdI~IVS

ION

~zeo~>oo~ ~o~

~

~• .J 0

u~EO

us

~

0

"~

~

o~o~o

u9EO

EL I:L Z

G. Z

Z

13_ Z

Z

I

Z

~,~

""~ 0 ~

~

M

THANETIAN]YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND Chron C26r

The oldest upper Paleocene sediments in eastern England occur at subcrop in the Ormesby and Hales boreholes, Norfolk. There the lower part of the Ormesby Clay is correlated with Chron C26r. The base of the Ormesby Clay Formation at both Ormesby and Hales yields broadly similar ages; palynomoph association sequences T1 and T2, respectively (Jolley 1992 and unpublisfied data). The difference in thickness of C26r (cf. Figs 5 & 6) sediments indicates relative condensation at Hales with respect to the same interval at Ormesby. Assuming a fixed accumulation rate during deposition of the Ormesby Clay at Hales, we postulate that the base of the Ormesby Clay in Norfolk may be c. 0.65 million years older than the oldest part of the Thanet Sand Formation (deposited during Chron C26n). Chron C26n

In the Ormesby and Hales boreholes the start of Chron C26n is within palynomorpb association sequence T6 (Jolley 1992). In all sections between Halesworth (Suffolk) and north Kent, the basal Tertiary sediments carry a normal polarity magnetization. The majority of the sections that have been studied for palynomorphs indicate that the base of the Thanet Sand Formation (and the Ormesby Clay in Sizewell) is at a higher level within T6 (Jolley 1992). Nannoplankton studies of the Thanet Sand Formation at Bradwell (Knox et aL 1994) suggest an NP6 age for these levels, which is similar to that reported from the type section in Kent (Aubry et al. 1986; Siesser et al. 1988). The normal polarity magnetozone identified at the base of the Thanet Sand and in the lower half of the Ormesby Clay formation is correlated with Chron C26n. Chron C25r

The start of Chron C25r is recorded at a number of sections through the Ormesby Clay and Thanet Sand Formation. In the Thanet Sand Formation at Bradwell the reversal is positioned at a level which corresponds to the upper part of nannofossil zone NP6 (Knox et al. 1994). Recent palynological studies (Jolley 1992) indicate a mid T7 (T7c) age for the start of Chron C25r in all of the sections between Norfolk and Kent. The top of the Thanet Sand and Ormesby Clay formations are thought to have been eroded prior to deposition of younger sediments. In the Thanet Sand and Ormesby Clay Formation sections north of London the youngest preserved sediments are of T9 age. In London and

141

north Kent the top of the Thanet Sand Formation is assigned to association sequence T8, indicating a slightly deeper level of erosion. Chron C25n

Chron C25n is associated with the upper part of nannofossil zone NP8 and the lower part of NP9 (Berggren et al. 1985). The studies by Hamilton & Hojjatzadeh (1982), Aubry et al. (1986) and Siesser et al. (1988) indicate that the record of nannofossil zones NP8 and NP9 in southern England is rather poor. Aubry et al. (1986) deduced that Chron C25n was not preserved in southern England; the lack of such a record resulted from either non-deposition or erosion of material deposited during that interval. The nonsequence/erosional event was located at the unconformity separating the Thanet Sand Formation and the Upnor Formation. Biostratigraphic and palaeomagnetic studies of the upper Paleocene at Ormesby and Norfolk revealed a similar gap in the sedimentary record (Cox et al. 1985; Knox et al. 1990). (Although the paper of Cox et al. 1985 pre-dated that of Aubry et al. 1986, their interpretation relied heavily on the conclusions of the latter work.) Knox et al. (1990) recognized that the Hales Clay was somewhat younger than the Lambeth Group, implying that the Chron C25n unconformity separating the Ormesby Clay Formation and the overlying sediments (Harwich Formation) is greater in Norfolk than in the London Basin. Recent studies by Ellison et al. (1996) report a record of Chron C25n in a borehole section from central London; normal polarity sediments positively assigned to the lower part of nannoplankton zone NP9 were recovered. There the Upnor Formation comprises two distinct sediment packages separated by a pebble bed. The lowermost unit, which records the Chron C25n magnetization, extends across only a few square kilometres. It forms an isolated remnant preserved beneath sediments deposited during the main Upnor Formation transgression. In this paper we have reported an interval in the uppermost 4 m of the Thanet Sand Formation in the Sizewell borehole where samples carry either positive or shallow characteristic remanence inclinations. The 'anomalous' interval is considered to have resulted from the late Paleocene (C25n) weathering of the subaerially exposed Thanet Sand Formation prior to deposition of the Lambeth Group. However, although numerous cores have sampled this same stratigraphic interval, the C25n remagnetization event has not been seen outside of the Sizewell area.

142

J.R. ALl • D. W. JOLLEY

Chron C24r

The main body of the Upnor Formation is reversely magnetized. The formation includes the oldest Chron C24r sediments. The Woolwich, Reading and Harwich Formations, and lower London Clay Formation, were all deposited during the 2.56 million years of Chron C24r. Based on Berggren et al. (1985, 1995), these deposits straddle the Paleocene-Eocene boundary. Chron C24n.3n

The start of Chron C24n.3n coincides with the base of the London Clay Formation Division B at Alum Bay, Whitecliff Bay (Aubry et al. 1986) and Sheppey, Kent (Ali et al. 1993). At these sections the normal polarity magnetozone correlated with Chron C24n.3n terminates within Division B. At Varengeville, north France, a reverse polarity magnetization is associated with the highly glauconitic clay marking the base of Division B. However, Ali et al. (1993) considered this magnetization to be much delayed postdepositional remanence acquired during a later reverse polarity interval. In Belgium, a record of the middle and latter part of Chron C24n.3n is preserved in the type section of the Wardrecques Member (Ali et al. 1993). This normal polarity interval is from a level equivalent to Division B of the London Clay Formation (King 1990). It is not possible to locate the start of Chron C24n.3n in Belgium as intervals corresponding to the Division A-B junction (the level at which the start of Chron C24n.3n is positioned in southern England) are not exposed.

Chronostratigraphy of depositional units Integration of the palaeomagnetic and palynomorph data provides a chronostratigraphic framework for age calibration of the upper Paleocene and lowermost Eocene depositional units in southeast England (Fig. 12). The Ormesby Clay Formation was deposited as a series of south- and eastward onlapping claystone units which pass laterally into the Thanet Sand Formation, the basal Thanet Sands at Pegwell Bay being 0.65 million years younger than the basal units of the Ormesby Clay Formation in Norfolk. Within this interval two separate series of onlapping units are identified, one within the basal part of Chron C26n and the second initiated in the lowermost Chron C25r (Jolley 1992; Knox et al. 1994). The accumulation of sediment was terminated by a fall in relative sea level at c. 56.6 Ma, giving rise to a prominent sequence boundary, with lowstand deposits of equivalent age to the subsequent hiatus occurring in the North Sea

Fig. 12. Chronostratigraphy of the upper Paleocene and lowermost Eocene deposits of southern England.

Basin (Sele Formation unit Sla). A significant rise in relative sea level at c. 56.2 Ma initiated lower Upnor Formation glauconitic sand deposition (Chron C25n) across southeast England, of which only a single erosional remnant has been identified (i.e. central London). The hiatus between the termination of Ormesby Clay Formation sedimentation and this later transgression is thought to exceed 0.3 million years, a period marked by subaeriel weathering of the Thanet Sand Formation and erosion of its upper units. Deposition of the remainder of the Upnor Formation (Chron C24r) and the overlying Woolwich and Reading Formations was as a series of barrier-bar and floodplain units, at a time of increasing relative sea level. Within this barrier-bar depositional system, two significant flooding surfaces are evident, although the younger is of apparently lesser intensity. Termination of the final Lambeth Group flooding phase immediately prior to 55 Ma, by a significant fall in relative sea level, shifted sedimentation into the North Sea Basin, with no equivalents of the Sele Formation units S2a and S2b occuring in southeast England or East Anglia (55.0-54.5 Ma) (Jolley 1995). The uppermost occurrence of the

THANETIAN/YPRESIAN MAGNETOSTRATIGRAPHY, SE ENGLAND dinoflagellate cyst Apectodinium augustum around the Sele Formation unit Slb/S2a boundary has been used by Powell (1988, 1992) and Knox & Holloway (1992) to approximate to the P a l e o c e n e - E o c e n e boundary in the North Sea Basin. This datum is comparable to the upper limit of the Lambeth Group in southeast England. Sedimentation resumed in southeast England and East Anglia, at c. 54.5 Ma in north Norfolk, with basal sands of the Hales Clay, this basal sand facies onlapping onto the Lambeth Group, reaching Essex by 54.4 Ma. In all, seven units (parasequences) are determined in the Harwich Formation, occurring in two onlapping series, separated by a hiatus of c. 0.2 million years in the west London Basin and the Hampshire Basin. The highest of these parasequences provides evidence of late Harwich Formation prograding sedimentation prior to a major flooding surface at the base of the overlying Walton Member (London Clay Formation) which appears to have been deposited during considerably higher relative sea levels.

Conclusions The upper Paleocene deposits of SE England magnetobiostratigraphic important constraints on following.

and lowermost Eocene are now placed within a framework that provides our understanding of the

143

(1) How the original Thanetian and Ypresian stage stratotypes relate to the global marine record. (2) The nature and timing of the Thanetian and Ypresian depositional sequences of the southern North Sea Basin. During this interval, deposition of the shallow marine and continental sediments along the southwest margin of the North Sea was punctuated by two hiatuses, each > 0.4 million years, at c. 57.6 and 55.0 Ma. The high-resolution age calibration of the N W European Thanetian and Ypresian depositional sequences is a valuable contribution to our ongoing research aimed at providing a rigorous examination of the Exxon eustatic sea-level concept. Ultimately, the dataset we are constructing from a number of passive margins around the globe will be used to examine the nature and timing (particularly the synchroneity) of the model's third- and fourthorder cycles over a 12-15 million year window of the late Paleocene to early-mid Eocene. NERC is acknowledged for financial support to JRA during his PhD studies. We are grateful to various colleagues for helpful and stimulating discussions during the development of the research, in particular Ernie Hailwood, Robert Knox, Chris King, Richard Ellison, Nick Johnston, Tony Morigi and Norman Hamilton. Kate Davies helped with the drafting of figures. Robert Knox and Chris King provided constructive reviews of the manuscript. The British Geological Survey are thanked for providing access to the cores used in this study. This paper is a contribution to IGCP Project 308.

References ALI, J. R. 1989. Magnetostratigraphy of Early Palaeogene Sediments from N.W. Europe. PhD Thesis, University of Southampton. , HAILWOOD,E. A. & KING, C. 1996. The 'Oldhaven Magnetozone' in East Anglia: a revised interpretation. This volume. - - - , KING, C. & HAILWOOD, E. A. 1993. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD,E. A & KIDD, R. B (eds) High Resolution Stratigraphy Geological Society, London, Special Publication, 70, 99-125. AUBRY, M.-E, HAILWOOD,E. A. & TOWNSEND,H. A. 1986. Magnetic and calcareous-nannofossil stratigraphy of the lower Palaeogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735. BERGGREN, W. A., KENT, D. V & FLYNN, J. J. 1985. Palaeogene geochronology and chronostratigraphy. In: SNELLING, N. J. (ed.) Geochronology of the Geological Record. Memoir of the Geological Society London, 10, 141-95. , , SWISHER, C. C. III & AUBRY, M.-P. 1995. A revised Cenozoic geochronology' and chronostratigraphy. In: BERGGREN, W. A., KENT, D. V., AUI3RY, M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic

Correlation: Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Volume, 54, Tulsa. CANDE, S. & KENT, D. V. 1995. Revised calibration of the geomagnetic time scale for the Late Cretaceous and Tertiary. Journal of Geophysical Research, 100, 6093-6095. COOPER, J. 1976. British Tertiary Stratigraphical and Rock Terms. Tertiary Research Special Paper, 1. Cox, F. C., HAILWOOD,E. A., HARLAND,R., HUGHES, M. J., JOHNSTON,N. & KNOX, R. W. O'B. 1985. Palaeocene sedimentation and stratigraphy in Norfolk, England. Newsletters on Stratigraphy, 14, 169-185. CURRY, D. 1981. Thanetian. In: POMEROL, C. (ed.) Stratotypes of Paleogene Stages. M6moire Hors S6rie du Bulletin d'Information des G6ologues du Bassin de Paris, 2, 255-265. - - - , ADAMS, C. G., BOULTER, M. C., DILLEY, E C., EAMES,E E, FUNNELL,B. M. & WELLS,M. K. 1978. A Correlation of Tertiary Rocks in the British Isles. Geological Society, London, Special Publication, 12. DESOR, E. 1847. Sur le terrain Danien, nouvel 6tage de la craie. Bulletin de la Socigt~ G~ologique de la France, 2, 179-182. ELLISON,R. A., ALl, J. R., HrNE, N. M. & JOLLEY,D. W.

144

J . R . ALl 8z D. W. JOLLEY

1996. Recognition of Chron C25n in the upper Paleocene Upnor Formation of the London Basin, UK. This volume. , JOLLEY,D. W., KING, C. & KNOX, R. W. O'B. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. HAILWOOD, E. A. 1977. Configuration of the geomagnetic field in early Tertiary times. Journal of the Geological Society, London, 133, 23-36. HAMILTON, G. B. & HOJJATZADEH, M. 1982. Cenozoic calcareous nannofossils - a reconnaissance. In: LORD, A. R. (ed.) A Stratigraphical Index of Calcareous Nannofossils. Ellis Horwood, Chichester, 136-167. HARDENBOL, J. & BERGGREN, W. A. 1979. A new Paleogene numerical time-scale. American

Association of Petroleum Geologists Studies in Geology, 6, 213-34. JENKINS, D. G. & LUTERBACHER, H. 1992. Paleogene stages and their boundaries (introductory remarks).

Neues Jahrbuch fiir Paliiontologie, Abhandlungen, 186,

1-5.

JOHNSTON, N. 1983. Magnetostratigraphic Stud)' of Paleogene Sediments from SE England. MSc Thesis, University of Southampton. JOLLEY, D. W. 1992. Palynofloral association sequence stratigraphy of the Palaeocene Thanet beds and equivalent sediments in eastern England. Review of Palaeobotany and Palynology, 74, 207-237. 1996. The earliest Eocene sediments of eastern England: an ultra high resolution palynological correlation. This volume. KING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6. - 1990. Eocene stratigraphy of the Knokke borehole (Belgium). Toelichtingen, Verhandelingen Geologische en Mijnkaarten van Belgie, 29, 67-102. Knox, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141,993-999. - - & HOLLOWAY,S. 1992. 1. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. - - , HJNE, N. M. &ALI, J. R. 1994. New information on the age and sequence stratigraphy of the type Thanetian of Southeast England. Newsletters on Stratigraphy, 30, 45-60. - - , MORIGI, A. N., ALl, J. R., HAILWOOD, E. A. &

HALLAM,J. R. 1990. Early Palaeogene stratigraphy of a cored borehole at Hales, Norfolk. Proceedings of the Geologists' Association, 101, 15-151. MARTINI, E. 1971. Standard Tertiary and Quaternary calacareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the II Planktonic Conference, Roma 1970. Edizioni Tecnoscienza, Rome, 739-785. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for the latest Palaeocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327-342. 1992. Dinoflagellate cysts of the Tertiary system. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellete Cysts. Chapman & Hall, London, 155-252. PRESTWICH,J. 1850. On the structure of the strata between the the London Clay and the Chalk in the London and Hampshire Tertiary Systems. Part I. Quarterly Journal of the Geological Society of London, 6, 252-281. - 1852. On the structure of the strata between the the London Clay and the Chalk in the London and Hampshire Tertiary Systems. Part III. The Thanet Sands. Quarterly Journal of the Geological Society of London, 8, 235-264. - 1854. On the structure of the strata between the the London Clay and the Chalk in the London and Hampshire Tertiary Systems. Part II. The Woolwich and Reading Series. Quarterly Journal of the Geological Society of London, 10, 75-170. SIESSER, W. G., WARD, D. J. ~r LORD, A. R. 1988. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Journal of Micropalaeontology, 6, 85-102. SMITH, A. G., BRIDEN, J. C. 8z HURLEY, A. G. 1981. Phanerozoic Palaeocontinental World Maps. Cambridge Earth Science Series. TOWNSEND, H. A. & HAILWOOD, E. A. 1985. Magnetostratigraphic correlation of Palaeogene sediments in the Hampshire and London Basins, southern UK. Journal of the Geological Society, London, 142, 1-27.

WHITAKER, W. 1872. The Geology of the London Basin: Part 1. Memoir of the Geological Survey of the United Kingdom. ZIEGLER, R A. 1982. Geological Atlas of Western and Central Europe. Shell International Petroleum, Amsterdam. Z1JDERVELD, J. D. A. 1967. AC demagnetization of rocks: analysis of results. In: COLLINSON, O. W., CREER, K. M. & RUNCORN, S. K. (eds) Methods in Palaeomagnetism, Elsevier, New York, 254-286.

Upper Paleocene-Lower Eocene dinoflagellate cyst sequence biostratigraphy of southeast England A. J. POWELL t, H. BRINKHUIS 2 & J. R B U J A K 3

1Millennia Ltd, Unit 3, Weyside Park, Newman Lane, Alton, Hampshire GU34 2P J, UK 2Laboratory of Palaeobotany and Palynology, University of Utrecht, Heidelberglaan 2, 3584 CS Utrecht, The Netherlands 3The Lexis Group, Albion House, 9 Albion Avenue, Blackpool, Lancashire FY3 8NA, UK Abstract: Detailed study of the aquatic palynomorph assemblages, particularly dinoflagellate cysts, from the type-Thanetian and related sections in southeast England has enabled a detailed biostratigraphic and sequence biostratigraphic analysis to be carried out. The base of the Thanetian Stage at Pegwell Bay lies above the base of the Alisocysta margarita (Area) biozone; the overlying type-Thanetian is assigned to the Apectodinium hyperacanthum (Ahy) and Apectodinium augustum (Aau) biozones. The base of the Ypresian succession is drawn at the base of the Harwich Formation at Lower Upnor corresponding to the base of the Glaphyrocysta ordinata (Gor) biozone. The Wetzeliellaastra (Was) biozone is characteristic of the lower London Clay Formation at Herne Bay. Five Thanetian and three Ypresian sequences are identified through consideration of dinoflagellate cyst palaeoecology and sedimentological evidence. Deepening and shallowing trends enable transgressive and highstand systems tracts to be identified. The maximum flooding surfaces are identified and correlated into the North Sea Basin (Central Graben) as a series of eight primary condensed sections, two of which (characterized by the Areoligera and Apectodinium acmes) are of a second-order scale (the other six are third-order). Lowstand system tracts are represented onshore by a series of seven unconformities (type 1 sequence boundaries) with the amount of missing time below biozonal resolution. A single type 2 sequence boundary is also evident in the Ypresian succession at Wrabness. The unconformities may be correlated into more distal locations in the North Sea Basin where lowstand sandstone deposition is characteristic.

One of the objectives of the International Commission on Stratigraphy (of the International Union of Geological Sciences) is the development of a global standard stratigraphic scale. As part of this exercise, the remit of the International Subcommission on Paleogene Stratigraphy (ISPS) is to review and, if necessary, revise the definition of the ages/stages of the Paleocene Epoch/Series (see Schmitz 1994). The ISPS has decided that the Paleocene Series be divisible into three stages, namely Danian, Selandian and Thanetian (see Jenkins & Luterbacher 1992). The Paleocene Working Group of the ISPS has set itself the objective of assessing the relative merits of various global events which might be used to define the base of the Selandian and Thanetian stages, and to recommend the Global Stratotype Section and Point (GSSP) for each stage. In view of the fact that one of the requirements for a GSSP should be as complete a sedimentary succession as possible, it is important that the type-Thanetian should be reappraised from a sequence stratigraphic perspective. After all, from a classical perspective, the base of the Thanetian Stage can be no older than the

base of the type-Thanetian (i.e. the contact of the Thanet Sand Formation with the Chalk Group at Pegwell Bay, Kent, UK). It is primarily in this context that the present study has been undertaken.

Aims of the study The principal purpose of this study is to provide a detailed interpretation of the sequence stratigraphy of Upper Paleocene and Lower Eocene sediments exposed in southeast England on the basis of the aquatic palynological (primarily dinoflagellate cyst) stratigraphic record. The scale of inquiry is such that systems tracts may be deduced from the biostratigraphic data (see Powell 1992a), thus necessitating a closely-spaced suite of samples. By so doing, it is possible to test third-order cyclicity represented in the rock record. By considering the details of a well-exposed set of strata in a relatively proximal setting, it is possible to deduce the stratigraphic relations with those located more distally within the North Sea Basin. A number of predictions may be made concerning the proximal expression of primary condensed sections (Powell

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Earl)' Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 145-183.

145

146

A.J. POWELL

1992a) expressed more fully in distal locations in the basin. In addition, unconformities apparent proximally (onshore) may be traced towards their correlative conformities in distal localities offshore as secondary condensed sections (Powell 1992a) and basin-floor fans.

Scope of the study The sections examined are exposed in Kent and Essex (Fig. 1). The stratigraphic position of the samples is shown in the stratigraphic summary logs (Figs 2-5). During April 1991, two of us (AJP and HB, with the guidance of Mr David J. Ward) collected 28 samples from Pegwell Bay (Fig. 2); two samples were obtained from the Upper Chalk, 18 from the Thanet Sand Formation (Base-Bed, Stourmouth Clays and Pegwell Marls) at the cliffend section, Pegwell Bay, and eight further samples from the adjacent hoverport car-park section (Pegwell Marls and Reculver Silts). At Herne Bay (Fig. 3), the eastern cliff section (four samples from the Reculver Silts), the foreshore section (12 samples from the Reculver Silts, through the Woolwich Bottom Bed of the Upnor Formation to

ETAL.

the Oldhaven Member of the Harwich Formation) were sampled (AJP, HB and DJW). A further four samples were also obtained from the western cliff section from the Oldhaven Member into the London Clay Formation. At the Lower Upnor sandpit (Fig. 4) the Reculver Silts of the Thanet Sand Formation (two samples), the Lambeth Group (the Upnor and Woolwich Formations - 12 sampies) and the Oldhaven Member of the Harwich Formation (five samples) were sampled (AJP and HB). At the cliff section of Wrabness, Essex (Fig. 5), 15 samples from the Wrabness Member of the Harwich Formation were obtained (AJP and HB). The average gap between samples were: at the cliff-end section (Pegwell Bay) 45 cm, at the hoverport car-park (Pegwell Bay) 43 cm, at the eastern cliff section (Herne Bay) 105 cm, at the foreshore section (Herne Bay) 135 cm, at the western cliff section (Herne Bay) 136 cm, at the sandpit (Lower Upnor) 127 cm, and at the cliff section (Wrabness) 102cm. The key references were: for Pegwell Bay Ward (1977), for Herne Bay Ward (1978), for Lower Upnor Kennedy & Sellwood (1970), and for Wrabness King (1981).

~_~---- ; ~ ~

~"

so .,-,,

_=,~

gambridg~ 52"

Wr~ LONDON

Harwich

.~ ---~~_r--...

9

52'

Herne Bay

) eOover 51:

51" Newhaven V'-

,. i

I

i

Fig. 1. Location map of the Pegwell Bay, Herne Bay, Lower Upnor and Wrabness sections.

PALEOCENE-EOCENE DINOFLAGELLATESEQUENCE BIOSTRATIGRAPHY The Laboratory of Palaeobotany and Palynology, University of Utrecht, undertook all the preparations using standard techniques. Heavy liquid (ZnC12) separation of the material was applied and residues were sieved using a 10 gm precision mesh sieve. After mixing to obtain homogeneity, four

Fig. 2. Stratigraphic summary log for Pegwell Bay.

147

slides were prepared using glycerine jelly as the mounting medium. The slides were studied both qualitatively and quantitatively. The quantitative analysis comprised ,a count of 200 aquatic palynomorphs where possible. For this purpose, six broad palynomorph

148

A . J . POWELL ET AL.

Fig. 3. Stratigraphic summary log for Herne Bay.

PALEOCENE--EOCENE DINOFLAGELLATESEQUENCE BIOSTRATIGRAPHY

149

Fig. 4. Stratigraphic summary log for Lower Upnor. categories were recognized: (1) dinoflagellate cysts (dinocysts); (2) miscellaneous algae (e.g. Pediastrum spp., Tasmanites spp.); (3) acanthomorph acritarchs; (4) foraminiferal test linings; (5) Leiosphaeridia spp.; (6) Paralecaniella spp. The remaining material was scanned qualitatively for additional dinocyst species. The results are displayed in Figs 6-13 and Tables 1-4. The dinoflagellate cyst taxonomy corresponds to that cited in Lentin & Williams (1993). All material

is filed in the collection of the Laboratory of Palaeobotany and Palynology, University of Utrecht, The Netherlands.

Background Lithostratigraphy The Kentish sections at Pegwell Bay and Herne Bay are internationally important because they

150

A . J . POWELL E T A L .

Fig. 5. Stratigraphicsummarylog for Wrabness.

comprise the type section of the Thanetian Stage, the youngest stage of the Paleocene Epoch (Curry 1981). By definition, the Thanetian Stage can be no older than the age at its lower contact with the Coniacian-Santonian Chalk. The section at Lower Upnor provides a lateral comparison of the Upnor Formation of the Lambeth Group, and the Oldhaven Member of the Harwich Formation (JoUey 1996) exposed at Herne Bay. The Wrabness section comprises the type section of the Wrabness Member of the Harwich Formation (Jolley 1996). It is beyond the scope of the present study to consider in detail the lithostratigraphy of the exposed sections; a detailed review of the Pegwell Bay and Herne Bay sections is given by Siesser et al. (1987). A recent revision of the lithostratigraphic nomenclature for the Upper Paleocene and

Lower Eocene of southeast England has been presented by Ellison et al. (1994) and this scheme has been adopted in the present study. M a g n e t o strati g raphy

Aubry et al. (1986) assigned the Pegwell Bay section to Chron 26n, and that at Herne Bay to Chron 25r. The Lambeth Group (Upnor and Woolwich Formations) and the Thames Group (Harwich and London Clay Formations) belong to Chron 24r. Aubry et al. (1986) and Knox (1990) points out that Chron 25n is absent and that an unconformity must therefore lie between the Thanet Sand Formation and the Lambeth Group. More recently, however, Ali (1994) and Ellison et al. (1994) have reported Chron 25n from the

151

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

Table 1. Aquatic palynomorph distribution (counts) at Pegwell Bay SamplesPB

03 04 0 5 0 6 0 7 0 8 0 9

10 11 12 13

14

15

16 17 18 19 20 21 22 23 24 25 26 27 28

Yaxa

A. senonensiss.l. D. denticulata

200

26

Foram test linings

4

P. magnificum C. medcalfii O. centrocarpum Leiosphaeridia spp. undiff. S. pseudofurcatus S. ramosus S. sepmtus

1

H. tubiferum A. alcicornu Cordosphaeridium spp. undiff. C. fibrospinosum A. margarita Alisocysta sp. 2 t C. inodes C. speciosum A. gippingensis Fibrocysta spp. undiff.

* 189 45 35 143 185 134 * 66 47

1

14 11 62 4

3

14 25

6

2

10

12

5

1 2 * *

1 5

* * 3 28 29

3 5

40

1 9

2

*

5

1 1

4

9 1

2

*

9

1

*

9 9

Pterospermella spp. undiff. T. delicata H. membraniphorum D. oebisfeldensis S. ancyrea G. divaricata A. ramulifera Tasmanites spp. undiff. P. crenulatum Apteodinium spp. undiff.

13 2 6 9 21 15

*

3

* * * 10 10 9 1

* 1 * 1 *

22 47 65

1 2

Acanthomorph acritarchs C. gracile Spinidinium? spp. undiff. Cerodinium spp. undiff. Hystrichokolpoma spp. undiff. L paradoxum Cribroperidinium spp. undiff. C. speciosum glabrum Gerdiocysta? sp. indet. G. intricata s.l. Subtilisphaera ? spp. undiff. t Source: Heilmann-Clausen 1985. * present outside of count.

2 13 14

2 26

1 14 * 2 10 10 7 3 2 5 8 7 4 10 5 2 1 3 5 5 3 5 25 56 44 48 49 21 32 3 9 10 10 8 3 1

2 1 *

6

*11

2 1 1

10 7

1

3

* 5

2

4

5 6

1 3 9 * 1 1 * * 19 11 6 10 9 11 11 17 20 60 23 42

*

4 2

1

*

1

1

*

1

3

1 1 7

* 3 3

5 *

9 6

12 7 7 24 10 10 15 11 5 12 37 10 11 40 27 5 1 1

19 38

3 *

1 20

lmpletosphaeridium spp. undiff. S. bentorii Rhombodinium sp. Lejeunecysta spp. undiff. Phelodinium ? sp. P. lidiae L. communis R. borussica Glaphyrocysta spp. undiff. M. pseudorecurvatum

2

*

2 4 1 11

*

1

1

2

1 21 22 23 24 25 26 27 28 1 2 2 8 6 12 8 2

1

21 28 2 3 2

3

3

1

1

1 4

7

5 2 2 1 I1 9

3 5

3

1

2 10 31 2

152

A.J.

POWELL ETAL.

%

%

\

\ A

A

i'

\ ! ~

I~ ~ ~ I~

~ l ~ l I ~ 9.., ~-

,....,

O i I

I

I

O

I r

,ta r

L~

.,,.~

~5 ,d

.a

PALEOCENE--EOCENE

DINOFLAGELLATE

SEQUENCE

153

BIOSTRATIGRAPHY

%

\

%

ZN

\

%

~

,4,,

%

\ %

===~====

======

s

fa0 o

.f..j.

~~i ~ ~

"

I

I

I

I

I

o .=,

o ,.Q

2

154

A.J.

P O W E L L ET AL.

%

%

\

\

\ \ !

/"

/~J

i

i

\

:! I J J

o

~

i

"~

~Z

I

I

I

I

I

F

O

.u b-, O

.a

PALEOCENE-EOCENE

DINOFLAGELLATE

SEQUENCE

BIOSTRATIGRAPHY

155

\ ~% .%

/\ % %

\

F~

\

/F, %

\ 'a

o o

E

o'~

T

T

]

.,..., "d

O

O

~,..., r

.a

156

A . J . POWELL E T A L .

\

%

\ %

\

\

.a

~, -~

"ua.I

:

NOIs

HDIM~[VHINO~VDINO:IHDIMqOOM

~ONdFI o

=:

g

.~!

o .,..~

g U

I

I

,,

{_ L

_

.,...

C2 ,m,M

PALEOCENE-EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

%

157

\ A

%

\

%

\ J

\

% o

"us=I

NOI,LVI/TSO,..q HDIMqOOM

H:~IM~IVI-I

NOI,LVIAI~O.,-I ~ I O N d f ' l

o

o

E

0

, i

I

"

1

I t.,,t

9- '

I o . ....~

t,,..

o

~

t~~

~l

_L~_L

l

1

I

A. J. P O W E L L ET AL.

158

%

.\ \

%

\

\\ \ \ u~d X'r D N C K I N O q

NOI.I.VIAt~IOd HDI/VIHVI--I

~: ,...1

I

+

, -,...

PALEOCENE-EOCENEDINOFLAGELLATESEQUENCEBIOSTRATIGRAPHY

159

\ A

% ua

tj

\ \

t~ ~3

\ v--1

% .%

\ \

I

A - T T l I ~-v-~

'a

~u~alXV'IDNCKINO3

NOIZVI,~OJ

HDIM~IVH 2

! \..

e~o o 9 E

l

1

~

[..,

I

t

I

J

160

A . J . POWELL ET AL.

Table 2. Aquatic palynomorph distribution (counts) at Herne Bay SamplesHB Taxa A. gippingensis A. robustum Alisocysta sp. 2t Gerdiocysta ? sp. indet. M. pseudorecurvatum Thalassiphora spp. undiff. S. pseudofurcatus T. pelagica P. crenulatum S. septatus

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20

7 * 1 * 4 * * 1 * * 3 * 8 1 24 13

2

1 19 1 2 *

3

2

1 1

1 :#

*

3

1

6 4 21

1 2

3

*

1 1

5 11

3

9

6 18 13

23 5 6 1 * * 1 2 * * * * * * 1

Foram test linings O. centrocarpum Fibrocysta spp. undiff. G. divaricata s.1. G. ordinata Acanthomorph acritarchs C. gracile C. inodes Spinidinium? spp. undiff. C. speciosum

13 76 40 1 30 28 9 1 5 11 21 31 8 15 * 11 12 10 3 14 17 21 2 9 18 15 19 11 10 9 * 1 * * * 26 4 13 71 19 24 34 46 13 1 8 14 29 17 2 1 * * * 9 2 1 2 2 6 10 13 4 7 5 34 * 5 1 * 2 5 3 2 * 4 2 2 * 2 7 4 6 1 2 16 13 16 1 6 10 7 13 12 7 2 6 2 7 5 10 4 2 3 5 5 10 17 17 11 8 14 19 3 7 7 12 15 *

4 1 32 1 9

A. cornufruticosum S. ramosus A. ramulifera H. tubiferum A. alcicornu

1 58 31 43 23 25 47 42 31 61 56 54 27 39 39 42 5 13 19 13 16 12 8 5 3 6 11 20 18 29 10 6 * 1 1 3 2 5 1 4 1 * 3 1 1 3 6 3 1 1 2 1 1 7

2 12 1 37 21 10 1 11 3

*

9

9 39 15 23

*

9

*

1

6

1 1 4

A. senonensis s.l. G. pastielsii Operculodinium spp. undiff. C. ? minimum Polysphaeridium spp. undiff. R. borussica Hystrichokolpoma spp. undiff. L. communis M. fimbriatum C. fibrospinosum

*

7

1 6

*

*

3

5 iI

1

5

*

*

1

1

1

1 1

4

*

* 4

*

7

5

8

4

*

2

2 1 3 1 * 1 1

1 5 1 8 2

* present outside of count.

l o w e r m o s t U p n o r F o r m a t i o n ( L a m b e t h Group). Knox et al. (1994) presented a reappraisal of the m a g n e t o s t r a t i g r a p h y o f the t y p e - T h a n e t i a n succession.

Biostratigraphy

Apart from dinoflagellate cysts, the most intensively studied microfossils from the type-Thanetian have been the calcareous nannofossils and a full review is given b y Siesser et al. (1987). These authors have also carried out the most detailed analysis o f the stratigraphic distribution o f these microfossils from P e g w e l l and H e r n e Bays. Their findings suggest that the l o w e r m o s t 4.2 m of the

Thanet Sand Formation is assignable to biozones N P 6 - 7 , and the overlying part o f the formation belongs to biozone NP8. Knox (1990) believed an N P 7 biozonal allocation to be likely for the l o w e r Thanet Sand Formation. H o w e v e r , more recently, K n o x et al. (1994) assigned the l o w e r m o s t Thanet Sand Formation (Stourmouth Clays) to b i o z o n e N P 6 and the R e c u l v e r Silts to b i o z o n e N P 8 (the P e g w e l l Marls being unassignable). K n o x (1990) suggested that the U p n o r Formation o f the L a m b e t h G r o u p m a y be assigned to b i o z o n e N P 9 (with an u n c o n f o r m i t y at its base). A c c o r d i n g to K n o x (1990), the ' O l d h a v e n B e d s ' o f the H a r w i c h Formation, and the overlying L o n d o n Clay Formation, belong to b i o z o n e N P 1 0 (with an u n c o n f o r m i t y at the b a s e o f the ' O l d h a v e n B e d s ' ) .

161

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY Table 2. Continued SamplesHB

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20

Taxa

A. homomorphum C. depressum P. lidiae Apteodinium spp. undiff. G. intricata Pterospermella spp. undiff. Leiosphaeridia spp. undiff. H. campanula Tectatodinium? sp. indet. C. medcalfii S. ancyrea L. cf. hyalina 14. membraniphorum Subtilisphaera? spp. undiff. D. simplex P. magnificum C. speciosum glabrum Homotryblium spp. undiff. Qyclopsiella spp. undiff. D. denticulata

2 1

9 1

5 *

*

1 4

*

1

6

2 2 2

19 5 2 2 7 3 2 5 1 1

2

1 2

7 15 20 15 7 2 1 10 2 1

7

8 18 4

9

9

1 14

1

4 16 36 28 * 2 13 *

3

1

6 20 1

1 *

44

3

3

2

6

*

*

3

1

3

*

2

2

*

1

1 1

4

6

4

8

2

1

3

3

3

4

1

P. golzowense L. hyalina Tasmanites spp. undiff. Phelodinium ? sp. indet. D. oebisfeldensis A. summissum T. delicata A. cf. quinquelatum A. cf. parvum A. parvum

8

3

1

1

4

1 1 7

1

2

3

1

1 46 *

1

1 * 2 * 1 1 1 4 1 5

1

4

2 1

2 9 1 1

1

1

24

1

4 8 7 42 12

Apectodinium spp. undiff. A. quinquelatum W. astra t Source: Heilmann-Clausen 1985. * present outside of count.

Previous dinoflagellate cyst studies A number of studies have been carried out on the dinoflagellate cyst assemblages from the typeThanetian and these have been reviewed by Powell (1992b). These previous studies had biostratigraphic or palaeoenvironmental objectives and were conducted on a relatively coarse scale. Thus, although broad correlations are possible throughout northwest Europe (and second-order palaeoe n v i r o n m e n t a l trends are apparent), detailed sequence biostratigraphic analysis has hitherto not been possible. To summarize, it is well k n o w n that Areoligera spp. and Alisocysta margarita are characteristic of the Thanet Sand Formation (Husain 1967; Downie

et aL 1971; Allen 1982; Jolley 1992). From a biostratigraphic perspective, both Costa & Downie (1976) and Costa et al. (1978) assigned the Thanet Sand F o r m a t i o n to the Deflandrea speciosa biozone and the Lambeth Group to the Wetzeliella (Apectodinium) hyperacantha biozone. Costa et al. (1978) allocated the L o n d o n Clay Formation of Kent to the Wetzeliella (W.) astra and W. (W.) meckelfeldensis biozones. Costa et al. (1978) were unable to allocate the 'Oldhaven Beds' of the Harwich Formation to any biozone due to the absence o f key species. K n o x et al. (1981) suggested that the 'Oldhaven Beds' lie within the A. hypercanthum biozone. H e i l m a n n - C l a u s e n (1985) was able to correlate the Thanet Sand Formation at Pegwell Bay (his biozone 4, and the

162

A.J. POWELL ET AL.

Table 3. Aquatic palynomorph distribution (counts) at Lower Upnor Samples LU

01 02 03 04 05 06 07 08 09

10

11 12 13 14

I5

16 17 18 19

Taxa

O. centrocarpum Leiosphaeridia spp. undiff. Acanthomorph acritarchs Pediastrum spp. undiff. H. abbreviatum A. hyperacanthum H. tenuispinosum A. quinquelatum A. parvum Apectodinium spp. undiff. A. homomorphum A. cornufruticosum S. ramosus L. communis M. fimbriatum Deflandreoids undiff. Batiacasphaera spp. undiff. G. divaricata Polysphaeridium spp. undiff. L. machaerophorum A. senonensis s.L G. ordinata G. intricata M. pseudorecurvatum P. indentata A. robustum Adnatosphaeridium spp. undiff. Cordosphaeridium spp. undiff. D. colligerum H. tubiferum C. inodes F. ferox H. plectilum Fibrocysta spp. undiff. Cribroperidinium spp. undiff. Apteodinium spp. undiff. S. membranaceus G. pastielsii T. delicata Pterospermella spp. undiff.

14 14 9 44 50 14 29 6 ! 11 7 1 1 8 41 50 7 33 15 5 53 40 18 12 9 3 8 46 11 4 46 7 15 1 5 1 7 * * 1 * 9 2 5 12 9 9 12 17 1 3 7 57 30 2 4 3 14 65 86 5 8 86 11 17 5 2 3 27 20 5 9 1 * 1 1 3 * 2 3 7 13 11 5 * 3 42 15 22 3 4 2 5 2 2

1 1 1

1 *

*

* *

5 4

2 1 48

1

3 9 1

3 1

1

7

1 11 14 6 4

3 13 1 3 3

22 17 11

1 :#

12 7

2 12 l 7 3

*

* present outside of count.

Ama biozone o f Powell 1992b) with the Holmehus Formation of the Viborg 1 borehole in Jutland, Denmark. Similarly, he correlated the formation at Herne Bay (his biozone 5, and the Ahy biozone of Powell 1992b) to the 'Grey Clay' lying between the Holmehus and Olst Formations in the same borehole, despite the apparent absence of Apectodinium spp. Powell (1992b) assigned the Thanet Sand Formation at Pegwell Bay to the Alisocysta margarita (Ama) biozone, and at Herne Bay to

the Apectodinium hypercanthum (Ahy) biozone. He placed the Lambeth Group in the Apectodinium augustum (Aau) biozone, and the 'Oldhaven Beds' of the Harwich Formation in the Glaphyrocysta ordinata (Got) biozone. The lower London Clay Formation was assigned to the Wetzeliella astra (Was) and W. mecketfeldensis (Wme) biozones. The only previous palynological studies of the Harwich Formation at Wrabness have been those

PALEOCENE-EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY of Jolley & Spinner (1989, 1991). They assigned the Harwich Formation to the Apectodinium hypercanthum biozone (sensu Costa & Downie 1976 and Costa et al. 1978) and restricted the succession above the Harwich Stone Band to the Deflandrea oebisfeldensis acme biozone.

Dinoflagellate cyst biostratigraphy of the studied sections Although the lowermost samples at Pegwell Bay and Lower Upnor are virtually barren of palynomorphs, most other samples contain reasonably well-preserved aquatic palynological associations suitable for quantitative analysis. The results show that dinoflagellate cysts dominate the associations in most cases. Other categories of aquatic palynomorphs are only common to abundant in particular intervals of the Lower Upnor section and in the upper part of the Wrabness section. Samples from the Oldhaven Member at Herne Bay contain virtually only Leiosphaeridia spp. and Pterospermella spp. In the dinoflagellate cyst assemblages usually only a few taxa or 'complexes' of morphologically similar taxa are quantitatively important. In most samples, Spiniferites ramosus, Achomosphaera ramulifera, Areoligera senonensis group,

Glaphyrocysta intricata, G. divaricata, Cerodinium speciosum or C. medcalfii are common to abundant, and range throughout the investigated interval. In addition, Apectodinium spp. (mainly A. homomorphum) and Deflandrea oebisfeldensis are common in particular intervals. Cordosphaeridium spp., usually represented by C. fibrospinosum, C. gracile or C. inodes, may be common in some samples, as are representatives of Operculodinium centrocarpum, Homotryblium spp., Impletosphaeridium spp. and Polysphaeridium spp. The biozonation scheme applied in the present study is that of Powell (1992b). This scheme is defined by the range bases of particular taxa; it has the additional advantage in that range tops may be extrapolated into the North Sea Basin. The major latest Paleocene and earliest Eocene palynological (dinoflagellate cysts) bioevents occurring in the North Sea Basin are shown in Fig. 14, while a comparison of published latest Paleocene and earliest Eocene North Sea palynological biozonations is presented in Fig. 15. The acronyms FO and LO stand for first occurrence and last occurrence respectively. The relative abundance categories used in the present study are: rare (< 6%), common (6-20%), abundant (20.5-60%), superabundant (> 60%) and present (outside of count).

163

Pegwell Bay dinoflagellate cyst biostratigraphy Barren interzone Age: Indeterminate. Samples: PB1 and PB2. Lithostratigraphy: Chalk Group. Comments: The two samples from the Chalk proved to be barren of palynomorphs and may not be assigned to any biozonei' The Chalk Group is Coniacian to Santonian in age (Siesser et al. 1987). Alisocysta margarita (Ama) interval biozone

Age: Thanetian (pars), late Paleocene (pars). Samples: PB3-PB28. Lithostratigraphy: Thanet Sand Formation (BaseBed, Stourmouth Clays, Pegwell Marls and Reculver Silts, pars). Comments: The succession in the Base-Bed of the Thanet Sand Formation contains neither the index taxon of the Ama biozone (Deflandrea denticulata) nor the nominate taxon (Alisocysta margarita). However, the superabundant occurrence and total dominance (100%) of members of the Areoligera senonensis group in sample PB4 is strong evidence for allocation of the Base-Bed to the Ama biozone. Furthermore, indicators of older biozones, notably Palaeoperidinium pyrophorum, are absent. The sample at PB6 from the base of the Stourmouth Clays contains a highly impoverished assemblage which includes representatives of the Areoligera senonensis group. Other samples from the lower Stourmouth Clays (PB7-PB10) are also impoverished in terms of both specimen and species numbers. However, the occurrence of Deflandrea denticulata at PB7 confirms allocation to the Area biozone. Phelodinium magnificum,

Cerodinium medcalfii, Spiniferites pseudofurcatus and S. septatus all have their FO at PB7. Alisocysta margarita, the nominate taxon for the Area biozone, has its FO at P B l l (upper Stourmouth Clays) together with those of Cordosphaeridium fibrospinosum, Alisocysta sp. 2 of Heilmann-Clausen (1985), Achomosphaera alcicornu, the Cerodinium speciosum and the Hystrichosphaeridium tubiferum groups. The assemblage at PB 12 is impoverished, but contains a reasonable recovery of both Areoligera senonensis and Cerodinium medcalfii. Moving into the Pegwell Marls (sample PB13), a number of FOs are apparent, including those of

Diphyes colligerum, Hystrichostrogylon coninckii (only present in this sample), H. membraniphorum, Thalassiphora delicata, Glaphyrocysta divaricata

164

A.J. POWELL ETAL.

Table 4. Aquatic palynomorph distribution (counts) at Wrabness SamplesWB

01

02

03

04

05

1 2 2

*

06

07

08

*

2 1 2 2 2 5

3 3

13 4

1 8 1 6 2 2 13 2 7 3

5 3 1 8 2 1 11 4 11 1

1 2 4 5 43 18 8 7 7

* 2 8 4 46 18 12 3 2

1 4 3

09

10

1 1

1 1 1 1 1

9

4

11

12

13

14

15

4 4 5

5 5 4 1

2 4 4

7 15 10 1

Taxa

G. exuberans O. complex P. subtile L. disjuncture D. cladiodes s. Morg. E bipolaris C. multispinosum S. cf. chlamydophora C. giuseppei S. septatus

* 1 2 3 * 1 8 1 2 5

* 2

2 4

4

3 3

2 3

5 2

11 2 14 5

5 5 5 8 2 * 11 3 12 4

Hystrichokolpoma sp. A. C. gracile L. wetzelii O. centrocarpum A. alcicornu A. multispinosum A. ramulifera S. cf. membranaceus S. monilis Senegalinium? sp. A.

2 6 * 18 1 1 8 4 9 4

T. cf. pelagica T. delicata G. ordinata A. senonensis s.l. S. ramosus H. tubiferum G. divaricata D. oebisfeldensis Acritarch sp. A Foram test linings

1 * 9 3 38 23 2 15 4 5

1 14 6 37 22 7 9 3 4

Michrystridium spp. undiff. Acritarch sp. B. Leiosphaeridia spp. undiff. P. indentata

13 2 4 5

3 3 3 6

12 2

6

2 2 3 5

8 4 4 11 1

6 3 4 12 1

11 4 10 3

9 4 12 5

2 1 11 4 39 16 5 4 7 2

1 13 3 46 17 9 4 3 2

1 8 2 48 16 4 2 6 4

8 5

4 2

3

4

6 2 3 2

and Deflandrea oebisfeldensis. The Areoligera senonensis group is superabundant at PB15. Phthanoperidinium crenulatum has its FO at PB 18. A generally similar assemblage is present at PB19 and it is at this sample position that the top of an acme (> 20%) of Areoligera may be placed (equivalent to bioevent P1 of Mudge & Copestake 1992; biochronoevent BC-7 of Armentrout et al. 1993; and bioevent Aga of M u d g e & Bujak 1996). The assemblage at PB20 is highly impoverished but includes Areoligera senonensis. A large number of taxa have their FOs at PB21 (uppermost Pegwell Marls), the most significant being Cerodinium depressum, Lejeunecysta communis, Rottnestia borussica, Palaeocystodinium lidiae, Melitasphaeridium p s e u d o recurvatum and the Glaphyrocysta pastielsii group. The assemblage at PB22 (lowermost Reculver

* 3 4

4 5 2 7 2 * 9

4 3 3 5

4 4 1 3

49 18 4 5 18

2 5 5 5

1

1

2 1

4 4

3 2 3 3

4 1

Silts) is i m p o v e r i s h e d in terms o f s p e c i m e n numbers, but nevertheless is generally similar in species composition to that at PB21. The assemblage at PB23 contains the FO of Cordosphaeridium gracile, while that at PB24 contains the LO of Cerodinium medcalfii and the FO of Spiniferites cornutus. A generally similar assemblage is present in PB25 but with the additional FOs of I m p a g i d i n i u m p a r a d o x u m , Gerdiocysta cassiculus and Cerodinium speciosum glabrum. The LO of Alisocysta margarita lies at PB26 (equivalent to bioevent P2 of M u d g e & Copestake 1992; biochronoevent BC-8 of Armentrout et al. 1993; and bioevent A l m of M u d g e & Bujak 1996), while an impoverished assemblage is present at PB27 broadly similar to that at PB26 in terms of species composition. The richer assemblage

165

PALEOCENE-EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY Table 4. Continued SamplesWB

01

02

03

1

1

04

05

06

* 2

1 4

2 3

2

* 1 2

3 4 3

07

08

09

10

11

12

13

14

15

5

2

4

3

7

7

14

1

2

Waxa

A. biformoides E cf. vectense D. phophoritica C. pannuceum M. pseudorecurvatum T. pellitum C. depressum D. colligerum P crenulatum R. borussica

2 4 2 2 9 2

D. pastielsii G. pastielsii H. membraniphorum Pteroapermella spp. undiff. O. ? severinii S. cornutus I. californiense S. pseudofurcatus A. hyperacanthum A. homomorphum

2 * 3 2

1 1

1

3 1 9 2

5 2

2

2

2 3

6 2

4

4 1 1

1 1

Lejeunecysta spp. undiff. S. densispinatum A. parvum P. golzowense P.. magnificum P. inversibuccinum P. minusculum 11. rigaudiae C. speciosum C. inodes

*

6 2 2 1

* 1

1

1 2

1

7 * * 2 *

*

2 3 1 9 9 3 1

Pediastrum spp. undiff. Cymatiosphaera spp. undiff. Tasmanites spp. undiff.

* 1 * 1 3 3

1 2

1 * 2 1 1

2 1

2

*

4

5

3

* present outside of count.

at

PB28

is

characterized

by

abundant

Glaphyrocysta spp.

Herne

Bay dinoflagellate

cyst

biostratigraphy

A p e c t o d i n i u m h y p e r a c a n t h u m (Ahy) interval

biozone Age: Thanetian (pars), late Paleocene (pars). Samples: HB 1-HB 14. Lithostratigraphy: Thanet Sand F o r m a t i o n (Reculver Silts, pars); Lambeth Group, U p n o r Formation (Woolwich Bottom Bed, pars). Comments: The FO of Apectodinium homomorphum at HB 1 is good evidence for allocation to the Ahy biozone (calibrated close to bioevent M5

of M u d g e & Copestake 1992; biochronoevent BC-9 of Armentrout et al. 1993; and bioevent IA of M u d g e & Bujak 1996); Apectodinium cornufruticosum also has its FO at HB1, as do Adnatosphaeridium robustum and Muratodinium fimbriatum. Other FOs of note include Palaeocystodinium lidiae at HB2, P. golzowense at HB9 and Lejeunecysta hyalina at HB10, all within the Reculver Silts of the Thanet Sand Formation. There are, in addition, a number of LOs within the succession. At HB10, the LOs of Gerdiocysta

cassiculus, Melitasphaeridium pseudorecurvatum, Adnatosphaeridium robustum, Alisocysta sp. 2 of Heilmann-Clausen (1985) and Areoligera gippingensis (equivalent to bioevent P2 of Mudge & C o p e s t a k e 1992; B i o c h r o n o e v e n t BC-8 of Armentrout et al. 1993; and bioevent Ag o f Mudge & Bujak (1996) are all present, while Spiniferites

A.J. POWELL ET AL.

166 CHRONO-,BIOZONES STRAT. (Powell '92)

w

~

9 Was

z w

Gor

,5 ~

MAJOR PALYNOLOGICAL (DINOFLAGELLATE CYST) BIOEVENTS ~

'9'

...i

9

uJ z W

<_

Aau Z

~ Ahy ttl Z < "1-

w

n

o~ -~

Ama Ppy

~'~ 'r r ,, t.~ r ,, r E-" t-! t_J

t.J ~

Wetzeliella astra Deflandrea oebisfeldensis/Glaphyrocysta ordinata (acme) Cerodinium wardenense (acme) Pterospermella/Leiosphaendia (acme) Glaphyrocysta ordinata (acme) Deflandrea oebisfeldensis (acme) Cerodinium dartmoofium/Cerodinium speciosum glabrum Apectodinium augustum/Apectodinium Complex (persistent) Apectodinium Complex (acme) (P3) Apectodinium augustum s. Wills & Peattie (1990) Muratodinium fimbriatum Apectodinium Complex (acme), A. augustum s. Schroder (1992) Apectodinium parvum Adnatosphaefidiumrobustum Apectodinium Complex (persistent), L. machaerophorum Areoligera gippingensis (P2) Rottnestia borussica (re-appearance) Alisocysta margarita (P2) Phthanoperidinium crenulatum Areoligera senonensis/Areoligera gippingensis (acme) (P1) Areoligera senonensis/Areoligera gippingensis (acme) Deflandrea denticulata, Areoligera gippingensis Palaeoperidinium pyrophorum (persistent)

Biochrono- Bioevents ~vents(Armen- (Mudgeand trout el a1.'93) Bujak,in prep)

t-~ j--~

i--~

-(E)r7 t..x t.x ~ ~ ~ -.-)

~ ~ t_~ r'7 - @ - - - @ - "

Fig. 14. Major latest Paleocene and earliest Eocene palynological (dinoflagellate cysts) bioevents occurring in the North Sea Basin.

pseudofurcatus has its LO at HB 11 at the top of the Thanet Sand Formation. Within the Upnor Formation of the Lambeth Group (Woolwich Bottom Bed), the LO of Phthanoperidinium crenulatum is at HB13, while

Achomosphaera alcicornu, Spiniferites septatus, Glaphyrocysta pastielsii, Areoligera senonensis, Palaeocystodinium lidiae and Lejeunecysta hyalina all have their LOs at HB14.

Apectodinium augustum (Aau) interval

biozone Age: Thanetian (pars), late Paleocene (pars). Sample: HB 15. Lithostratigraphy: Lambeth Group; Upnor Formation (Woolwich Bottom Bed, pars). Comments: Although the nominate taxon, Apectodinium augustum, is absent in sample HB 15 (it is believed to be restricted to more offshore parts of the basin), the Apectodinium complex accounts for 34.5% of the assemblage. This constitutes the FO of the Apectodinium acme and is taken as evidence (see Heilmann-Clausen 1985) for the assignment to the Aau biozone. The assemblage at HB15 also contains the FO (and LO) of Apectodinium parvum. This taxon is accompanied by other Apectodinium species including A. summissum (FO and LO) and

(abundant) A. homomorphum. A number of species have their LOs at HB15. These include Rottnestia

borussica, Lejeunecysta communis, Muratodinium fimbriatum and Thalassiphora delicata. Barren interzone Age: Indeterminate. Samples: HB 16 and HB 17. Lithostratigraphy: Thames

Group; Harwich Formation (Oldhaven Member). Comments: These two samples from the Oldhaven Member proved to be barren of palynomorphs and assignment to any biozone is not possible. Glaphyrocysta ordinata (Gor) chronozone

Age: Ypresian (pars), early Eocene (pars). Samples: HB18 and HB19. Lithostratigraphy: Thames Group; London Clay Formation, Basement Bed and Walton Member (pars). Comments: The sample (HB18) at the base of the London Clay Formation contains neither the index taxon (Phelodinium magnificum) nor the nominate taxon (Glaphyrocysta ordinata) of the Got biozone. However, because the assemblages at HB18 and (especially) HB19 are characterized by the presence of Pterospermella spp. (common at HB19) the base of the Gor chronozone (i.e. the

167

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

~

~!

..

::}

~un~eucpao

,

wnqlu~o~ad/q cun!u!poloadv

9

~ ~ wn!leqdauolo/O

:O~--

--'~i

c~

eso!oads eaapueliaQ

+

I'"

X

•

o_-.

aw0v

wnqzu~o~ad/fq tun!upozaadv

s!suaplajs!qao ea~puelJaO s!suauap~A4 s!sdo!le~aO

,,,,.,., . T OO"-"

w

wnlsn~n~ tun!u!poloadv

T

e~p~6i~w

s oo~

awov

tunlsn~ne wn!u!poloadv

s!suaPle/s!qeo ea~pu~lte(7

,

e~!~e6aetu ~s/oos!lV

(5

k

O ca

c-

cO

~-4

~!a~fS~etu e~s/oosflv

o

'--a . ,,,,,.,

( ~ co =,...

o ,-g >,..

s

c}..

C3

;E C)

s!suauapa~ .-----_'s

9

r~

I ~

~1

.

o

~ ~ ~ ~,~ tunlsn~n~ "V

Z

~s/oos!lV o m

.c_ ~"

m

,=r.

I

CO I--

O3 F-

I

0 OJ

t--

r~,r-

_

~.

(.9

z • . NVIS3Ua~k ~

=INgOO=I t:i=IMO7

< 04 t---

./Z-

--_/

r

== c,)

tO 11"-

h-. O_

< NVIIBNVHI

laS

E

168

A . J . POWELL E T A L .

Leiosphaeridia biofacies of Bujak & Mudge 1994) probably lies at HB18. Glaphyrocysta ordinata does, however, occur at HB 19. Wetzeliella astra (Was) interval biozone

Age: Ypresian (pars), early Eocene (pars). Sample: HB20. Lithostratigraphy: Thames Group; London Clay Formation, Walton Member (pars). Comments: The occurrence of Wetzeliella astra in the stratigraphically highest sample examined at Herne Bay (HB20) indicates assignment to the Was biozone. Also characteristic of the assemblage is the common occurrence of Deflandrea oebisfeldensis together with abundant Cerodinium speciosum, common Apectodinium species, notably A. homomorphum and A. quinquelatum, and Pterospermella spp.

LU10 in an assemblage dominated by superabundant Apectodinium spp., notably A. homomorphum. This is taken as evidence for allocation to the Aau biozone despite the absence of the nominate taxon (see Heilmann-Clausen 1985). A generally similar assemblage is present at L U l l although Muratodinium fimbriatum and Lejeunecysta communis are additionally present. A. parvum also occurs at both LU12 and LUI3, although Apectodinium spp. as a whole are no longer dominant. The FO of Lingulodinium machaerophorum lies at LU10. The sample at LU14 is impoverished, but Apectodinium spp. are again very well represented (particularly A. homomorphum) at LU15, which probably marks the LO of the Apectodinium acme (equivalent to bioevent P3 of Mudge & Copestake 1992; biochronoevent BC- 10 of Armentrout et al. 1993; and bioevent Aa? of Mudge & Bujak 1996). Adnatosphaeridium robustum occurs at LU15 as well as Diphyes

colligerum.

Lower Upnor dinoflagellate cyst biostratigraphy Barren interzone Age: Indeterminate. Samples: LU1-LU8. Lithostratigraphy: Thanet

Sand Formation (Reculver Silts, pars); Lambeth Group, Upnor Formation (Woolwich Bottom Bed) and Woolwich Formation (Woolwich Beds, pars). Comments: The eight samples obtained from the base of the Lower Upnor section proved to be barren of microplankton and may not therefore be assigned to any biozone. Samples from the Upnor Formation at Herne Bay are productive and are assigned to the Ahy Biozone.

Impoverished interzone Age: Indeterminate. Sample: LU9. Lithostratigraphy: Lambeth Group; Woolwich Formation (Woolwich Beds, pars). Comments: The sample at LU9 is impoverished both in terms of specimen and species numbers. Pediastrum spp., Leiosphaeridia spp. and undifferentiated acanthomorph acritarchs are present. Apectodinium augustum (Aau) interval

biozone Age: Thanetian (pars), late Paleocene (pars). Samples: LU 10-LU 15. Lithostratigraphy: Lambeth Group; Woolwich Formation (Woolwich Beds, pars).

Comments: Apectodinium parvum is present at

Glaphyrocysta ordinata (Gor) chronozone

Age: Ypresian (pars), early Eocene (pars). Samples: LU 16-LU 18. Lithostratigraphy: Thames Group; Harwich Formation (pars). Comments: Although

the

marker

species

(Phelodinium magnificum) of the Gor biozone is absent, Leiosphaeridia spp. are particularly dominant in the two samples (LU16 and LU17) at the base of the Harwich Formation (biochronoevent BC-11 of Armentrout et al. 1993; Leiosphaeridia biofacies of Bujak & Mudge 1994). Allocation to the Gor chronozone is therefore appropriate. Also characteristic of the assemblages is superabundant Paralecaniella indentata; Apectodinium homomorphum is present at LU16. Glaphyrocysta ordinata occurs commonly at LU18 in an assemblage with Leiosphaeridia spp. and Pterospermella spp. present, but not common. Also occurring at LU18 is Apectodinium homomorphum and common Hystrichosphaeridium

tubiferum. Impoverished interzone Age: Indeterminate. Sample: LU19. Lithostratigraphy: Thames

Group; Harwich Formation (Oldhaven Member). Comments: The highest sample at Lower Upnor (LU19) contains a highly impoverished assemblage which precludes assignment of it to any biozonal unit; Leiosphaeridia spp. and Pterospermella spp. are present.

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

Wrabness dinoflagellate cyst biostratigraphy Glaphyrocysta ordinata ( Gor) interval

biozone Age: Ypresian (pars), early Eocene (pars). Samples: WB 1-WB 10. Lithostratigraphy: Thames Group; Harwich Formation (Wrabness Member, pars). Comments: The assemblage in sample WB1 at the base of the succession examined (50 cm above the Harwich Stone Band) contains common Deflandrea oebi~feldensis and Glaphyrocysta ordinata (thus indicating assignment of the D. oebisfeldensis acme of Knox & Harland 1979). Also common in WB1 is Hystrichosphaeridium

tubiferum. Other events of note within the Got biozone include: the FOs of Fibrocysta bipolaris and Lentinia? wetzelii at WB1; the FOs of Achilleodinium biformoides and Deflandrea phosphoritica at WB2 (together with Diphyes colligerum, Phthanoperidinium crenulatum and Melitasphaeridium pseudorecurvatum); and the FO of Palaeotetradinium minusculum at WB6. Phelodinium magnificum, the index taxon for the Gor biozone has its LO at WB7. Apectodinium

homomorphum, Deflandrea oebisfeldensis, Glaphyrocysta ordinata and Hystrichosphaeridium tubiferum have their LOs at WB 10. The Leiosphaeridia biofacies of Bujak & Mudge (1994) was not encountered in our sample set from Wrabness. However, the data of Jolley & Spinner (1989, fig. 3) indicates that Deflandrea oebisfeldensis occurs at levels < 1% below the Harwich Stone Band, which suggests that the succession below the Stone Band could be assigned to the Leiosphaeridia biofacies, albeit on negative evidence.

Impoverished interzone Age: Indeterminate. Samples: WB 11-WB 15. Lithostratigraphy: Thames

Group; Harwich Formation, Wrabness Member (pars). Comments: The samples between WB 11 and WB 15 contain impoverished microplankton assemblages including low levels of Leiosphaeridia spp., acanthomorph acritarchs, Pediastrum spp., Pterospermella spp., Tasmanites spp. and

Paralecaniella indentata.

Sequence biostratigraphy The biostratigraphic analysis of the sections has demonstrated that a refined breakdown is possible

169

through the utilization of the dinoflagellate cyst record. Previous studies of these sections have been correlatable on a broad scale throughout northwest Europe (see, for example, Costa & Downie 1976; Costa et al. 1978; Heilmann-Clausen 1985) and the North Sea Basin (see, for example, Costa & Manure 1988; Powell 1988, 1992b). However, the present study makes it possible for finer-scale correlations to be made with the North Sea Basin where similarly detailed (and largely unpublished) analyses have been carried out since the early 1980s (e.g. Stewart 1987; Powell 1988; Wills & Peattie 1990; Mudge & Bujak 1996). Correlations into northwest Europe, however, remain broad because detailed biostratigraphic data either require reappraisal (having being published prior to 1980) or are unavailable through lack of study. A number of hiatuses are apparent in onshore southeast England, as might be expected in a more proximal setting. Where a hiatus is contained within a particular biozone, sedimentological and lithological characteristics have been taken into consideration. The hiatuses are manifest as a series of eight transgressive-on-highstand systems tract unconformities (type 1 sequence boundaries) marking the bases of five Thanetian sequences and two Ypresian sequences. A third Ypresian sequence has its base marked by a shelf margin systems tract (type 2 sequence boundary).

North Sea sequences Correlation between the onshore (proximal) sequences of the present study and their offshore (distal) equivalents is dependent upon the recognition of linking bioevents. Recent developments in North Sea lower Paleogene biostratigraphy (e.g. Mudge & Copestake 1992; Mudge & Bujak 1996) suggest that particular bioevents (often acme occurrences) are associated with high-gamma shales. The nature of high-gamma shales is not always apparent; some represent condensed sections associated with downlap surfaces. As noted by Powell (1992a), the amount of time represented within a particular condensed section will increase progressively in an offshore direction. As a result, condensed sections cannot be regarded to represent true time lines. Condensed sections are not only characteristic of maximum flooding surfaces (primary condensed section of Powell 1992a), but also develop at the base of lowstand and shelf margin wedges (secondary condensed section of Powell 1992a). Care must therefore be taken when interpreting the correlative significance of high-gamma shales. Nevertheless, many sequences developed for the lower Paleogene of the North Sea (e.g. Jones & Milton 1994; Mudge & Bujak 1996) are based upon the recognition and

170

A.J. POWELL ET AL.

correlation of condensed sections. Because a number of bioevents recognized in our study are also present in the North Sea (Fig. 14), the dinocyst record provides an excellent means of relating the proximal and distal expressions of sequences.

Dinoflagellate cyst palaeoecology and relative sea-level fluctuations Studies of Recent dinoflagellate cyst distribution patterns indicate that water-mass composition (roughly comparable to distance from shore), seasurface temperature, salinity and productivity are the principal controlling parameters. An actualistic approach would thus suggest that a similar situation existed in the past. On this basis, it has been possible to interpret the late Paleocene to early Eocene relative sea level and climatic history of southeast England. Lithological, sedimentological and biostratigraphic data indicate that the sections investigated represent marginal marine environments. Such considerations are supported by the absence of oceanic dinoflagellate cysts like Impagidinium and Nematosphaeropsis spp. One would expect that relative sea-level fluctuations in such a marginal marine setting would be reflected in varying abundances of taxa that typically occur in restricted marine or inner neritic water masses as opposed to those that characteristically occur in outer-neritic waters. Moreover, it might be anticipated that only the transgressive and highstand phases would be present at the margins, while the lowstand deposits are only well developed in the basin, i.e. the North Sea. In a manner similar to that applied by Brinkhuis (1994), the successive shifts in distribution of the principal dinoflagellate cyst groups may be used to reconstruct the late Paleocene to early Eocene relative sea-level history of southeast England. The distribution of co-occurring aquatic palynomorphs may be used to further support interpretations. As a result of the scale of the present investigation, only second- and third-order sequences, sensu Haq et al. (1987), may be identified. Because the third-order signal is superimposed upon the second-order, high energy conditions become progressively more open marine through each second-order cycle. The recognition of higher order sequences would require even more closely-spaced samples. The palaeoecological affinity of the dinoflagellate cyst groups may be established on the basis of published Recent distribution patterns available for those components of the assemblages that have modern counterparts (or that are morphologically comparable to extant taxa). The palaeoecology of extinct taxa may be interpreted from

various, less reliable, sources, including information about other fossil organisms. Changes in average water temperature may be traced by comparing the relative contribution of typical high, middle and lower latitude dinoflagellate cysts to successive assemblages. However, the generally marginal marine nature of the investigated deposits obscures temperature-related signals significantly, since the onshore-offshore and facies-related trends are dominant. Moreover, most species occurring in the sections are cosmopolitan in nature. The principal dinoflagellate cyst complexes (Evitt 1985) and the significance of the other aquatic palynomorphs are summarized below. Deflandrea Complex of Pp-cysts. The motile dinoflagellate stages that produced cysts assignable to Cerodinium may well represent heterotrophic peridinioids. High frequencies of heterotrophic peridinioids are characteristic of areas with high primary production related to increased nutrient availability in upwelling areas and river mouths (see Powell et al. 1992). Raised quantities of Cerodinium spp. may therefore, like Deflandrea spp. (Brinkhuis et al. 1992), be tentatively linked to such depositional settings; their motile stages may have grazed on diatoms and other phytoplankton (including other dinoflagellates) as well as on bacterially decayed material. Lejeunecysta, Palaeocystodinium and Phelodinium are included in this complex, as well as Cerodinium and Deflandrea. The factors controlling the distribution of Deflandrea oebisfeldensis were probably similar to those acting upon Cerodinium spp. Apectodinium Complex of Pq-cysts. Apectodinium spp. may be compared with other peridinioids, such as Cerodinium spp., and may have a similar palaeoecology. When present, Apectodinium spp. often dominate assemblages completely. This phenomenon suggests that the motile stages producing Apectodinium spp. were highly tolerant towards varying water-mass conditions and could be successful where others failed, e.g. in highly restricted marine settings. Furthermore, there is evidence to suggest (Jan du Chine et al. 1975; Heilmann-Clausen 1985) that Apectodinium spp. appeared earlier in the Tethyan Realm than in the North Sea Basin, or were more consistently present in the older Paleocene (Selandian) at lower latitudes (Brinkhuis et al. 1994). During the latest early Paleocene, they appear to have migrated into higher latitudes (they would have migrated into higher latitudes initially in a pulsating manner associated with relative sea-level highstands and relatively high temperatures), including the North Sea Basin (Thomas 1993), and reached high abundances as far north as the Barents Sea by

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

the latest Paleocene (Brinkhuis, pers. obs.). Information from Thomas (1993) indicates that 'middle' Paleocene Apectodinium occurrences are typically present at or near condensed intervals immediately above sediments representative of short periods of sand-prone, high terrestrial flux. During the late Paleocene they became increasingly common, culminating in worldwide, nearly monotypic, assemblages at the Paleocene-Eocene boundary (Powell 1988). The present study shows that Apectodinium spp. gradually disappeared through the course of the early Eocene, at least in the basin margins, to be replaced by the Wetzeliella lineage. Areoligera Complex of Gv-cysts. This complex, which includes Gerdiocysta as well as Areoligera, is thought to represent marginal marine, innerneritic settings, often in association with relatively coarse-grained deposits (hydrodynamically high energy), indicating a tolerance for such settings (Downie et al. 1971; Brinkhuis 1994). However, one species, Areoligera gippingensis, is indicative of offshore conditions (Heilmann-Clausen 1994). Representatives of the Glaphyrocysta group, most of which are included within the Areoligera Complex, may represent open marine, inner-neritic water masses (Downie et al. 1971; May 1980; Brinkhuis & Zachariasse 1988; Brinkhuis et al. 1992; Brinkhuis 1994). Homotryblium Complex of Gq-cysts. Representatives of the Homotryblium Complex, include Polysphaeridium as well as Homotryblium itself, are regarded to represent restricted marine to inner neritic water masses (Bradford & Wall 1984; Brinkhuis 1994). Spiniferites Complex of Gs-cysts. This complex, which includes Achomosphaera, Hystrichostrogylon, Rottnestia and Spiniferites, is regarded to represent open marine, neritic water masses (Wall et al. 1977; Head & Wrenn 1992; Brinkhuis 1994).

171

Miscellaneous algae. Although published records of Tasmanites spp., Pterospermella spp. and Cyclopsiella spp. show them to have cosmopolitan distributions extending from freshwater to oceanic settings, high relative abundances almost exclusively occur in quite nearshore, inner-neritic to brackish settings (Chateauneuf 1980). Pediastrum spp. have a definite fresh- to brackishwater origin.

Acanthomorph acritarchs.

Published records of these taxa show them to have cosmopolitan distributions from freshwater to oceanic. Highest relative abundances almost exclusively occur in relatively nearshore, inner-neritic to brackish settings.

Foraminiferal test linings. Remains of the inner walls of foraminifera have been suggested to be derived from benthic foraminifera in relatively inshore habitats at the change-over from normal to restricted marine settings (G. J. Van de Zwaan, pers. comm.). They are also known to occur in areas of upwelling (Powell et al. 1992). Leiosphaeridia spp. This category, often referred to as leiospheres, probably represents a heterogeneous group of freshwater to marine organisms. However, they often dominate in sediments independently characterized as brackish or freshwater environments. They are often accompanied by Pterospermella spp. which may support attribution to restricted marine to freshwater environments. Bujak & Mudge (1994) referred to this assemblage as the Leiosphaeridia biofacies of late Paleocene to early Eocene age in the North Sea Basin. Paralecaniella spp. Although Paralecaniella spp. have been reported from Upper Cretaceous and Tertiary successions from around the world, their palaeoenvironmental significance is poorly understood. Elsik (1977) considered them to represent either schizosporous algae or acritarchs, and mentioned that they were most abundant in marginal marine successions, indicating a tolerance for inner-neritic, probably brackish, conditions.

Cordosphaeridium Complex of Gi-cysts. This complex, only represented by Cordosphaeridium in the present study, is regarded to represent open marine, neritic water masses (Downie et al. 1971; May 1980; Brinldauis 1994).

Dinoflagellate cyst sequence biostratigraphy of the studied sections

Operculodinium Complex of Gn-cysts. The Operculodinium Complex, which includes Dapsilidinium, Diphyes and Lingulodinium and Operculodinium is thought to represent restricted to open marine, neritic water masses (Wall et al. 1977).

By using the above listed palaeoecological characteristics of the principal dinoflagellate cyst groups and the other accompanying aquatic palynomorphs, five Thanetian and three Ypresian sequences have been identified in the studied sections. Figure 16 shows a summarized sequence stratigraphy of studied sections, while the predicted sequence

172

A . J . POWELL E T A L .

-"~ ~STRAT" Was

CHRONO- BIOZONES LITHOSTRAT. (Powell'92) (Ellison et .

SEOOE"OESl SEO~E"C~SiSEOOE"OES I SEOOE.OESlSEouE.oEs Pegwell Bay HerneBay LowerUpnor Wrabness (Hardenbol'94

Yp-2

Gor

Th-6

Aau Th-5

Ahy

Th-4

Th-3

Ama

Th-2

Th-1

Fig. 16. Summarized sequence stratigraphy of studied sections.

PALEOCENE--EOCENEDINOFLAGELLATESEQUENCEBIOSTRATIGRAPHY

173

FortiesField SEQUENCESSEQUENCESLITHOSTRAT. Field SEQUENCES Predictedstratigraphicalextent Forties Biozonati on Units Stewart(1987)! Jones& CentralGraben ThisStudy of condensedsections ] W&P W&P(1990) W&P(1990) M&B(thisvol) !Milton (1994) I (K&H 1992) 3rdOrder

~

Ypr-3

a2

0

d~

9 (Balder)

~

ShelfMarginWe

T50

~ - -

~m B1

3rd Order PrimaryCD

Ypr-2

CromartYMemberSandsto~ 3rd Order

$3 ~~... r

M

Ypr-1

8 T45 (Dornoch)

S2b

CromartySandstone , ~ M e ~

Deflandrea 2ndOrderPrimaryt~ dartmooria CondensedSection

y

cO

. ~

S2a

.

Tht-5 F~

~ 3rd Order PrimaryCD

E

L N J

3rd Order PrimaryCD

Sa e S*O j

M

-.~ S l b

7 (Forties)

Forties:a~

Forties y

0 LL

H

Tht-4

Tht-3

E

~

F

Sla

T40

J:::

, "6 E

~~.~

E

3rd Order

Tht-2 UpperBalmoral

,

~_

6

(Lista III)

E

D y ~

Tht-1

2ndOrderPrimary Condensed Sectii LowerBalmoral ~ Sandstone

g) :~.(b

~P.pyrophorum Fig. 17. Summarizedsequence stratigraphyof North Sea Central Graben.

t--

._o L3 0 It

5 (Lista II) T30 AndrewTuff 3 (Lista I)

co

",

1_2 L1

174

A . J . POWELL E T A L .

stratigraphy of North Sea Central Graben is shown in Fig. 17.

Thanetian Sequence 1 (Tht-1) Samples: PB3-PB21 (Pegwell Bay). Biozonation: Alisocvsta margarita (Area) biozone (pars).

Magnetostratigraphy: Chron 26n to 25r (pars) in Knox et al. (1994). Lithostratigraphy: Thanet Sand Formation (BaseBed, Stourmouth Clays, Pegwell Marls).

Basal unconformity: The Base-Bed of the Thanet Sand Formation lies unconformably upon the Chalk Group with the Campanian-Maastrichtian, the whole of the Lower Paleocene (biozones Cco, Tin, Xlu, Scr and Cst of Powell 1992b) and much of the Upper Paleocene (biozones Sde, Csp and Ppy of Powell 1992b) successions missing. Recognized systems tracts: TST, PB3-PB 15; HST, PB 16-PB21. Comments: Areoligera spp. exhibit a strong dominance in the assemblages between PB3 and PB15 (superabundant), but show a steady decline above PB16-PB 19 (abundant). This succession is therefore interpreted as a restricted high-energy marginal marine setting typical of a transgressive regime. Samples PB14 and PB15 are by far the richest samples of the succession in terms of concentration and are thought to lie close to the most condensed interval or maximum flooding surface (m.f.s.). Conspicuously, Areoligera gippingensis has its peak occurrence in these samples; this species may represent a morphotype of Areoligera adapted to more offshore conditions (HeilmannClausen 1994). The decline in the relative percentage of Areoligera spp. above sample PB15 is accompanied by an increase in the recovery of the Spiniferites Complex as well as the introduction of Cordosphaeridium spp. In addition, relative numbers of the Glaphyrocysta group start to increase in this interval. This succession is interpreted to represent the introduction of hydrodynamically less energetic, neritic, open marine conditions in response to a relative sea-level highstand. A possible underlying second-order trend of relative sea level rise through subsidence may have caused the significant sediment accumulation of the main body of the Pegwell Marls during this HST phase. Equivalence: The lowstand systems tract (LST) of sequence Tht-1 is absent. However, in the central North Sea this part of the succession may correspond to the Lower Balmoral Sandstone within the Lista Formation (L2) (on the basis that it lies above the LO of Palaeoperidiniurn pyrophorum according to Knox & Holloway 1992).

The m.f.s, of the Tht-I sequence is represented by a second-order primary condensed section (Fig. 17) at the 5-6 sequence boundary of Stewart (1987), the T30-T40 sequence boundary of Jones & Milton (1994) and the Lista II-Lista III sequence boundary of Mudge & Bujak (1996). The Tht-1 sequence corresponds to the Th-I sequence of Hardenbol (1994) and also equates with sequence BT1 of Knox et al. (1994).

Thanetian Sequence 2 (Tht-2) Samples: PB22-PB28 (Pegwell Bay). Biozonation: Alisocysta margarita (Area) biozone (pars).

Magnetostratigraphy: Chron 25r (pars) in Knox et al. (1994). Lithostratigraphy: Thanet Sand Formation (Reculver Silts, pars).

Basal unconformity: The Reculver Silts lies unconformably upon the Pegwell Marls with part of the succession attributable to the Area biozone missing. Recognized systems tracts: TST, PB22-PB24; HST, PB25-PB28. Comments: The Spiniferites Complex, being the most 'offshore' dinoflagellate cyst group present in the assemblages, exhibits a strong dominance in the assemblages in samples PB22-PB24, but declines steadily above. This succession is interpreted to represent open marine, neritic conditions marking the onset and development of a transgressive phase. The '6 inch' shell bed (unit 15) at the base of the Reculver Silts is thought to mark the transgressive surface (TS) overlying the highstand deposits of the underlying Tht-1 sequence. Samples PB25 and PB26 are the richest samples in this sequence (in terms of concentration) and are thought to lie close to the most condensed interval or m.f.s. The discontinuous nodule layer (unit 17) between these two samples is considered to represent this horizon. The decline in relative percentages of the Spiniferites Complex above PB25 is accompanied by a marked relative increase in the Glaphyrocysta group, reflecting increased influence of innerneritic water masses. This part of the sequence is therefore interpreted to represent the introduction of gradually more restricted, inner neritic conditions in response to a subsequent relative sea-level highstand (HST). The LO of Alisocysta margarita (PB26) seems to be associated with the relative sea level fall of the Tht-2 sequence and may be used to correlate the m.f.s, into other sections. Equivalence: The missing succession at the base of the Tht-2 sequence represents the LST. This probably corresponds to the Upper Balmoral

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY Sandstone within the Lista Formation (L3) in the central North Sea (on the basis that it lies below the LO of Alisocysta margarita according to Knox & Holloway 1992). The m.f.s, of the Tht-2 sequence is represented by a third-order primary condensed section (Fig. 17) within sequence 6 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Lista III sequence of Mudge & Bujak (1996). The Tht-2 sequence corresponds to the Th-2 sequence of Hardenbol (1994) and also equates with sequence BT2 of Knox et al. (1994).

Thanetian Sequence 3 (Tht-3) Samples: HB 1-HB 11 (Herne Bay). Biozonation: Apectodinium hyperacanthum (Ahy) biozone (pars).

Magnetostratigraphy: Chron 25r (pars) in Knox et al. (1994). Lithostratigraphy: Thanet Sand Formation (Reculver Silts, pars).

Basal unconformity: Because there is no overlap between the sections at Pegwell and Herne Bays, the presence or absence of an unconformity between samples PB28 and HB 1 cannot be proven. The argument for the presence of a sequence boundary between the sections is based upon the composition of the dinoflagellate cyst assemblages (discussed below). Recognized systems tracts: TST, HB I-HB4; HST, HB5-HB 11. Comments: The whole aspect of the assemblages from the lower part of the Herne Bay section (notably the presence of Apectodinium spp.) are markedly different to those from the upper part of the section at Pegwell Bay and supports the view that a discontinuity is present between the sections. The Spiniferites Complex exhibits a strong dominance in the assemblages in samples HB 1-HB2 and HB3, but has lowered relative percentages in overlying samples, until sample HB 11. The lower part of the Herne Bay succession is therefore interpreted to represent open marine conditions marking the onset and development of a transgressive phase (and an underlying TS is therefore predicted). Sample HB4, taken at the 'Reculver Tabular Band' (unit B), conspicuously displays a marked increase in the Glaphyrocysta group, reflecting increased influence of inner-neritic water masses. The Reculver Tabular Band is therefore thought to represent the condensed interval or m.f.s. The succession overlying the Reculver Tabular Band is therefore interpreted to represent the introduction of more restricted inner-neritic conditions in response to a subsequent relative sea-level HST. This culminates higher up in the renewed introduction of increased relative percentages of the even more restricted Areoligera spp. (found in

175

sediments with high levels of continentally-derived plant material). The highest sample taken from the Reculver Silts, sample H B l l , unexpectedly contains high relative numbers of the offshore Spiniferites Complex. It is possible that these may have been introduced from the overlying Beltinge Fish Bed (unit J) of the Upnor Formation through burrowing. An alternative explanation may be sampling error as the samples were collected from the covered succession in the intertidal zone of the foreshore. Equivalence: It must be considered likely that no lowstand phase of the Tht-3 sequence was developed at the basin margins; in the central North Sea the equivalent lowstand deposits probably lie within the (lower) Forties Sandstone Member within the Sele Formation (S 1a) (on the basis that they lie above the LO of Alisocysta margarita according to Knox & Holloway 1992). The m.f.s, of the Tht-3 sequence is represented by a third-order primary condensed section (Fig. 17) within sequence 7 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Forties sequence of Mudge & Bujak (1996). The Tht-3 sequence corresponds to the Th-3 sequence of Hardenbol (1994) and possibly equates to sequence BT3 of Knox et al. (1994).

Thanetian Sequence 4 (Tht-4) Samples: HB12-HB15 (Herne Bay); LU3-LU9 (Lower Upnor).

Biozonation: Apectodinium hyperacanthum (Ahy) biozone (pars).

Magnetostratigraphy: Chron 24r (pars), inferred from Ali (1994).

Lithostratigraphy:

Lambeth Group; Upnor Formation (Woolwich Bottom Bed). Basal unconformi~': The Upnor Formation lies unconformably upon the Reculver Silts at Herne Bay, and upon the Thanet Sand Formation at Lower Upnor. The unconformity is expressed by the Beltinge Fish Bed (unit J) at Herne Bay and the 'basal conglomerate' (unit 2) at Lower Upnor. Recognized system tracts: TST, HB12-HB14, LU2-LU5; HST, HB 15, LU6-LU9. Comments: The Spiniferites Complex exhibits a strong dominance in the assemblages in samples HB12-HB14 while relative numbers of the Glaphyrocysta group decline. As in older sequences, this unit is interpreted to represent open marine, neritic conditions marking the onset and development of a transgressive phase; the base of the Beltinge Fish Bed (unit J) is seen to represent the TS of the Tht-4 sequence. Since the samples overlying sample HB 12 from the Upnor Formation (up to HB15) display no marked compositional changes, the entire interval is interpreted to represent the TST of the sequence.

176

A. J. P O W E L L E T A L .

Samples LU1-LU8 from the lower part of the Lower Upnor section are barren of palynomorphs. However, the sedimentological development of this part of the succession (Kennedy & Sellwood 1970) may be interpreted to represent the Tht-4 Sequence. The 'basal conglomerate' (unit 2) at the base of the Upnor Formation at this site may be regarded as being equivalent to the Beltinge Fish Bed (unit J) at Herne Bay, its base indicating the TS. The overlying Upnor Formation (Woolwich Bottom Bed: unit 3 - grey and yellow, sometimes glauconitic sands) may therefore represent the TST of the Tht-4 sequence. The overlying Woolwich Formation, characterized as a 'massive sandstone' (unit 4) succeeded by unit 5 ('sandy ironstone'), 'cross-bedded white sands and Ophiomorphaburrows' (unit 6) and the overlying 'iron-cemented sandstone' (unit 7) may be tentatively regarded to represent the highstand phase of the sequence. In contrast to the underlying samples, LU9 does contain palynomorphs; a few specimens of Operculodinium centrocarpum form the dinoflagellate cyst population, although they may have been washed in from more offshore settings. Sample HB15 from Herne Bay is tentatively assigned to the Tht-4 sequence (HST) on the basis of its dominance by Apectodinium spp., including

A. parvum. Equivalence: The succession attributable to the lowstand phase of the Tht-4 sequence is absent but is probably equivalent to the (middle) Forties Sandstone Member within the Sele Formation (Sla) in the central North Sea (on the basis that it lies above the LO of Alisocysta margarita according to Knox & Holloway 1992). The m.f.s, of the Tht-4 sequence is represented by a third-order primary condensed section (Fig. 17) within sequence 7 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Forties sequence of Mudge & Bujak (1996). The Tht-4 sequence corresponds to the Th-4 sequence of Hardenbol (1994).

Comments: Apectodinium spp. exhibit a strong dominance in the assemblages in samples LU10LU12 and are accompanied by a few other taxa. This nearly monotypic assemblage indicates a tolerance for high energy, shallow marine conditions, indicative of the TST. This succession is followed by an assemblage where the more offshore, open marine Spiniferites Complex is abundant in association with the Operculodinium Complex (in sample LU12). Above LU12, Polysphaeridium spp. (Homotryblium Complex) become abundant indicating the increasing influence of restricted, probably hypersaline, waters and shallowing. This succession is interpreted to reflect relative sea-level HST conditions following a m.f.s, around sample LU12. The return of abundant Apectodinium spp. in LU15 may reflect the renewed introduction of restricted, high energy, shallow marine waters at the final phase of the HST. Equivalence: The lowstand phase of the Tht-5 sequence is missing onshore. It is probably represented in the central North Sea by the (upper) Forties Sandstone Member within the Sele Formation (Slb) (on the basis that it lies below the top of the acme occurrence of Apectodinium spp. according to Knox & Holloway 1992). The m.f.s, of the Tht-5 sequence is represented by a second-order primary condensed section (Fig. 17) at the sequence 7-8 boundary of Stewart (1987), T40-T45 sequence boundary of Jones & Milton (1994) and Forties-Dornoch sequence boundary of Mudge & Bujak (1994). The Tht-5 sequence corresponds to the Th-5 sequence of Hardenbol (1994). Ypresian sequence 1 (Ypr-1) Samples: LU16-LU18 (Lower Upnor). Biozonation: Glaphyrocysta ordinata (Gor) biozone.

Magnetostratigraphy: Chron 24r (pars), inferred from Ali (1994).

Thanetian Sequence 5 (Tht-5) Samples: LU10-LU15 (Lower Upnor). Biozonation: Apectodinium augustum (Aau) biozone.

Magnetostratigraphy: Chron 24r (pars), inferred from Ali (1994).

Lithostratigraphy: Lambeth Group; Woolwich Formation (Woolwich Beds, pars). Basal unconformity: The unconformity at the base of the Tht-5 sequence is indicated by the abrupt incoming of shell beds at the base of the unit 10 ('Woolwich Shell Beds') at Lower Upnor. Recognized systems tracts: TST, LU10-LU11; HST, LU12-LU15.

Lithostratigraphy: Thames

Group,

Harwich

Formation, pars.

Basal unconformity: This unconformity at the base of the Harwich Formation is marked by a distinct shell bed with flint pebbles and an erosional surface at its base (unit 16), exposed near the top of the Lower Upnor Section. The base of this bed is taken to represent a TS. Recognized systems tracts: TST, LU16; HST, LU17-LU18. Comments: In terms of dinoflagellate cysts, sample LU16 contains a mixture of taxa, which typically tolerate restricted marine conditions, including members of the Apectodinium Complex., Polysphaeridium spp. (Homotryblium Complex)

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

and the Operculodinium Complex. In addition, the sample is characterized by abundant Paralecaniella spp. and Leiosphaeridia spp. which also point towards strong restricted marine influences. Unit 16 from which sample LU16 was taken is thus thought to reflect the early transgressive phase of the Ypr-1 sequence (TST). Sample LU17, taken from the overlying unit, is characterized by assemblages where the more offshore, open marine Spiniferites Complex is abundant and is in association with the Operculodinium Complex. This is taken to indicate the return of open marine, neritic waters (HST) with a m.f.s, between sample LU16 and LU17. However, the restricted marine influence persists into the HST, as indicated by the still relatively high numbers of Paralecaniella spp. A subsequent trend towards more restricted marine conditions may be inferred from the composition of sample LU18 (increasing representation of the Glaphyrocysta group and the Homotryblium Complex). Equivalence: The lowstand phase of the Ypr-1 sequence is absent and is possibly equivalent to the (lower) Cromarty Sandstone Member within the Sele Formation (S2b) (on the basis that it lies above the acme occurrence of Apectodinium spp. according to Knox & Holloway 1992). The m.f.s, of the Ypr-1 sequence is represented by a third-order primary condensed section (Fig. 17) within sequence 8 of Stewart (1987), sequence T45 of Jones & Milton (1994) and the Dornoch sequence of Mudge & Bujak (1996). The Ypr-1 sequence corresponds to the Th-6 sequence of Hardenbol (1994), although he suggests that it has a Thanetian rather than Ypresian age.

Ypresian sequence 2 (Ypr-2) Samples: HB16 and HB17 (Herne Bay); LU19 (Lower Upnor); WB 1-WB 10 (Wrabness).

Biozonation:

Glaphyrocysta

ordinata

(HB16 and HB17) are barren of palynological assemblages. Sample LU19, from unit 19 (Oldhaven Member of the Harwich Formation) is impoverished in terms of its palynomorph content. This may reflect the return of extremely marginal marine, high energy conditions, leading to wellsorted and oxidized deposits. At Wrabness, the lowest sample was taken from just above the Harwich Stone Band and contains an assemblage with abundant Spiniferites Complex, Glaphyrocysta group with Apectodinium Complex, and Deflandrea oebisfeldensis. The composition of the overlying assemblages remain more or less the same until sample WB10 (see below). The sequence between WB 1 and WB 10 is interpreted to represent the HST of the Ypr-2 sequence. The Harwich Stone Band (stratigraphically below sample WB1) may well represent the horizon of maximum condensation and m.f.s, of this sequence. Jolley & Spinner (1991) interpreted this succession as a transgressive-regressive cycle. Equivalence: The Oldhaven Member comprises proximal deposits of the lowstand phase (incised valley fill of Hardenbol 1994). Although parts of the lowstand of the Ypr-2 sequence are preserved, the early parts of the lowstand succession (attributable to the Got biozone) are probably missing and may correspond to the (upper) Cromarty Sandstone Member within the Sele Formation (S2b) in the central North Sea (on the basis that it lies above the acme occurrence ofApectodinium spp. according to Knox & Holloway 1992). The m.f.s, of the Ypr-2 sequence is represented by a third-order primary condensed section (Fig. 17) at the 8-9 sequence boundary of Stewart (1987), T45-T50 sequence boundary of Jones & Milton (1994) and the Dornoch-Balder sequence boundary of Mudge & Bujak (1994). The Ypr-2 sequence corresponds to the Yp-1 sequence of Hardenbol (1994).

(Gor)

biozone (pars).

Magnetostratigraphy: Chron 24r (pars), inferred from Ali (1994).

Lithostratigraphy:

177

Thames Group; Harwich Formation (Oldhaven Member, pars; Wrabness Member, pars). Basal unconformity: The unconformity at the base of the Ypr-2 sequence is marked at Herne Bay by the 'basal pebble bed' (unit L) at the base of the Oldhaven Member of the Harwich Formation. At Lower Upnor, it is represented by the base of bed 19. The contact is not exposed at Wrabness. Recognized systems tracts: LST, HB16 and HB17 (Herne Bay), LU19 (Lower Upnor); HST, WB1WB 10 (Wrabness). Comments: The samples from the Harwich Formation (Oldhaven Member) at Herne Bay

Ypresian Sequence 3 (Ypr-3) Samples: HB 18-HB20 (Herne Bay); WB 11-WB 15 (Wrabness).

Biozonation:

Glaphyrocysta ordinata (Gor) biozone (pars), Wetzeliella astra (Was) biozone (pars). Magnetostratigraphy: Chron 24r (pars), inferred from Ali (1994). Lithostratigraphy: Thames Group; Harwich Formation (Wrabness Member, pars); London Clay Formation (Walton Member, pars). Basal contact: The base of the Ypr-3 sequence is not expressed at Wrabness by a major lithological change. It is, however, clearly marked by a major change in the palynological composition of the assemblages between samples WB10 and WBll

178

A . J . POWELL E T AL.

(see below) and a type 2 sequence boundary is envisaged. Recognized systems tracts: SMST, WBI 1-WB 15 (Wrabness); TST, HB18-HB 19 (Herne Bay); HST, HB20 (Herne Bay). Comments: The samples between WB 11 and WB 15 contain impoverished palynological assemblages (terrestrial, sporomorph palynomorphs dominate the assemblages); aquatic palynomorphs, other than dinoflagellate cysts, are present, indicating restricted marine, possibly brackish, conditions. Jolley & Spinner (1989) reported near monotypic assemblages of Deflandrea oebisfeldensis from some samples in this interval which were not recorded in the present study. In view of the absence of a major lithological change at the base of this unit, it is thought to represent a shelf margin wedge (with a type 2 sequence boundary at its base) almost continuous with the underlying HST of the Ypr-2 sequence (although there is a marked change in the palynological assemblages). Jolley & Spinner (1991) envisaged a transgressive unit at the top of the Harwich Member which is not apparent from our data. The overlying Walton Member in Essex (not sampled by us) is regarded to be the subsequent TST of the Ypr-3 sequence (acknowledged as a transgressive unit by Jolley & Spinner 1991). The lowermost sample from the London Clay Formation at Herne Bay (HB18) is also impoverished but does contain Leiosphaeridia spp. and Pterospermella spp. This assemblage is regarded to represent the early transgressive deposits containing restricted marine material, possibly redeposited from the older lowstand phase of the Ypr-2 sequence (the London Clay Basement Bed being the transgressive surface). The overlying sample at HB19 contains abundant specimens of the Spiniferites Complex indicating the return of more open marine, neritic conditions. Samples HB 18 and HB 19 are thought to be part of the TST of the Ypr-3 sequence. The Wetzeliella astra (Was) biozone was observed at sample HB20 but was not encountered in the Wrabness samples; Jolley & Spinner (1989), however, reported the base of the Was biozone coincident with the base of the Walton Member of the London Clay Formation (exposed in cliffs at Felixstowe), which is tentatively interpreted to be the TST of the Ypr-3 sequence. Sample HB20 contains an assemblage typical of shallow marine, inner-neritic conditions (abundant Apectodinium spp. with Deflandrea oebi~feldensis and Cerodinium spp.) which is taken to be indicative of the HST part of the sequence (a m.f.s, therefore lies between sample HB19 and HB20). Equivalence: The m.f.s, of the Ypr-3 sequence is represented by a third-order primary condensed

section (Fig. 17) within sequence 9 of Stewart (1987), sequence T50 of Jones & Milton (1994) and the Balder Sequence of Mudge & Bujak (1994). The Ypr-3 sequence corresponds to the Yp-2 Sequence of Hardenbol (1994).

Discussion The results of the present study have implications on a global, regional and local scale. In terms of its merits as a global stratotype section and point, the type-Thanetian is clearly easily accessible. However, since it is located in a relatively proximal setting, the lowstand phases of the sequences comprising the type-Thanetian are represented by unconformities. The base of the type-Thanetian lies within the Alisocysta margarita (Area) biozone of Powell (1992b); the base of the Ama biozone lies at the base of the lowstand phase of the Tht-1 sequence in the North Sea Basin (the base of the L2 unit of the Lista Formation at the LO of Palaeoperidinium pyrophorum). One option, therefore, is for the global stratotype point for the base of the Thanetian Stage to be placed at the base of the Ama biozone in a suitable location in the North Sea region (e.g. the Viborg 1 borehole, Jutland, Denmark, which contains the reference section for the Ama biozone), if not at Pegwell Bay. From a regional perspective, the present study reveals that the unconformities identified onshore may be extrapolated into more distal locations in the North Sea Basin. Here they are manifest as phases of lowstand submarine fan deposition, or as lateral conformities (secondary condensed sections). Published sequence stratigraphic schemes for the North Sea Basin do not provide sufficient resolution to allow the third-order sequence to be identified. This is due to the fact that the unconformities are below traditional biozonal resolution, and that the sequences are, generally, defined by so-called high gamma shales (mostly second-order primary condensed sections). It is predictable that the number of sequences in the North Sea Basin is somewhat higher than is currently acknowledged in the published literature (i.e. there are predictably five Thanetian sequences). It is only on the reservoir scale that sufficient resolution will be possible to identify these sequences, primarily through the integrated use of sophisticated biostratigraphic techniques. The present study also has implications for local stratigraphy. First, there is no overlap between the Pegwell Bay and Herne Bay sections judging by the recovery of Apectodinium specimens below the Reculver Tabular Band (Bed B at Herne Bay). Second, a sequence (Tht-5) is recognized at Lower Upnor with its base at the base of the Woolwich Shell Beds, i.e. within the Woolwich Formation.

PALEOCENE-EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

Third, the oldest Ypresian sequence (Ypr-1) lies below the Oldhaven Member at Lower Upnor with its base coincident with the base of the Harwich Formation; this sequence had previously been considered to be Thanetian in age (sequence Th-6 of Hardenbol 1994). Fourth, the succession at Wrabness has no direct equivalent at Herne Bay; the Oldhaven Member being considered to be the lowstand systems tract of the Ypr-2 sequence exposed at Wrabness. Fifth, the Wetzeliella astra (Was) Biozone was identified at Herne Bay, but not Wrabness.

Conclusions

Thanetian Sequence 1 (Tht-1) The unconformity at the base of the Tht- 1 sequence lies between samples PB2 and PB3 (Pegwell Bay) at the base of the Area biozone (at the contact between the Chalk and the Base-Bed of the Thanet Sand Formation). The type-Thanetian is therefore no older than the Ama biozone. The lowstand phase of the Tht-1 sequence is not represented at outcrop. It is probably equivalent to the Lower Balmoral Sandstone within the Lista Formation (L2) in the central North Sea. The m.f.s, of the Tht-1 sequence lies between samples PB 15 (LO of superabundant Areoligera senonensis) and PB16 (Pegwell Bay) within the Ama biozone (intra-Pegwell Marls, Thanet Sand Formation). In the central North Sea the m.f.s, is represented by a second-order primary condensed section at the 5-6 sequence boundary of Stewart (1987), the T30-T40 sequence boundary of Jones & Milton (1994) and the Lista II/Lista III sequence boundary of Mudge & Bujak (1996). The Tht-1 sequence corresponds to the Th-1 sequence of Hardenbol (1994), although he suggests that it ranges down into the Ppy biozone which cannot be substantiated. The Tht-1 sequence also equates with sequence BT1 of Knox et al. (1994).

Thanetian Sequence 2 (Tht-2) The unconformity at the base of the Tht-2 sequence lies between samples PB21 and PB22 (Pegwell Bay) within the Ama biozone (at the contact of the Pegwell Marls with the Reculver Silts, Thanet Sand Formation). The lowstand phase of the Tht-2 sequence is not represented at outcrop. It is probably equivalent to the Upper Balmoral Sandstone within the Lista Formation (L3) in the central North Sea. The m.f.s, of the Tht-2 sequence lies between samples PB26 (LO of Alisocysta margarita) and PB27 (Pegwell Bay) within the Area biozone

179

(intra-Reculver Silts, Thanet Sand Formation). In the central North Sea, the m.f.s, is represented by a third-order primary condensed section within sequence 6 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Lista III sequence of Mudge & Bujak (1996). The Tht-2 sequence corresponds to the Th-2 sequence of Hardenbol (1994) and also equates with sequence BT2 of Knox et al. (1994).

Thanetian Sequence 3 (Tht-3) The unconformity at the base of the Tht-3 sequence cannot be proven. However, our palaeoecological interpretation suggests that samples PB28 (Pegwell Bay) and HB1 (Herne Bay) belong to different sequences, with a possible unconformity between the Area and Ahy biozones (intra-Reculver Silts, Thanet Sand Formation). The lowstand phase of the Tht-3 sequence is not represented at outcrop. It is probably equivalent to the (lower) Forties Sandstone Member (Forties Field unit E and pre-Apectodinium subzone of Wills & Peattie 1990) within the Lista Formation (L3) in the central North Sea. The m.f.s, of the Tht-3 sequence lies between samples HB4 and HB5 (Reculver Tabular Band, Herne Bay) above the FO of Apectodinium homomorphum within the Ahy biozone (intra-Reculver Silts, Thanet Sand Formation). In the central North Sea, the m.f.s, is represented by a third-order primary condensed section within sequence 7 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Forties sequence of Mudge & Bujak (1996). The Tht-3 sequence corresponds to the Th-3 sequence of Hardenbol (1994) and possibly equates to sequence BT3 of Knox et al. (1994).

Thanetian Sequence 4 (Tht-4) The unconformity at the base of the Tht-4 sequence lies between samples HB 11 and HB 12 (Herne Bay), and between samples LU2 and LU3 (Lower Upnor) within the Ahy biozone (at the contact of the Thanet Sand Formation with the Upnor Formation). The lowstand phase of the Tht-4 sequence is not represented at outcrop. It is probably equivalent to the (middle) Forties Sandstone Member (Forties Field unit F and Apectodinium homomorphum subzone of Wills & Peattie 1990) within the Sele Formation (Sla) in the central North Sea. The m.f.s, of the Tht-4 sequence is not well expressed at outcrop since it is only questionably present at Herne Bay (between samples HB 14 and HB15) and lies in a barren interzone at Lower Upnor (at the base of a unit with Ophiomorpha

180

A . J . POWELL ET AL.

burrows). In the central North Sea, the m.f.s, is represented by a third-order primary condensed section (the Apectodinium parvum acme of Wills & Peattie 1990) within sequence 7 of Stewart (1987), sequence T40 of Jones & Milton (1994) and the Forties sequence of Mudge & Bujak (1996). The Tht-4 sequence corresponds to the Th-4 sequence of Hardenbol (1994).

Thanetian Sequence 5 (Tht-5) The unconformity at the base of the Tht-5 sequence lies between samples LU9 and LU10 (Lower Upnor) at the base of the Aau biozone (at the base of the Woolwich Shell Beds within the Woolwich Formation). The lowstand phase of the Tht-5 sequence is not represented at outcrop. It is probably equivalent to the (upper) Forties Sandstone Member (Forties Field unit J and Apectodinium augustum subzone of Wills & Peattie 1990) within the Sele Formation (Slb) in the central North Sea. The m.f.s, of the Tht-5 sequence lies between samples L U l l and LU12 (below the LO of the Apectodinium acme at sample LU15) within the Aau biozone (at the top of the Woolwich Shell Beds within the Woolwich Formation). In the central North Sea, the m.f.s, is represented by a secondorder primary condensed section at the 7-8 sequence boundary of Stewart (1987), T40-T45 sequence boundary of Jones & Milton (1994) and the Forties-Dornoch sequence boundary of Mudge & Bujak (1994). The Tht-5 sequence corresponds to the Th-5 sequence of Hardenbol (1994).

The Ypr-1 sequence corresponds to the Th-6 sequence of Hardenbol (1994), although he suggests that it has a Thanetian rather than Ypresian age.

Ypresian Sequence 2 (Ypr-2) The unconformity at the base of the Ypr-2 sequence lies between samples HB 15 and HB16 (Herne Bay) and between samples LU18 and LU19 (Lower Upnor) at the base of a barren interzone (at the base of the Oldhaven Member, intra-Harwich Formation). The lowstand phase of the Ypr-2 sequence is represented at outcrop by the Oldhaven Member of the Harwich Formation. These beds are probably equivalent to the (upper) Cromarty Sandstone Member within the Sele Formation (S2b) in the central North Sea (Ceratiopsis wardenensis biozone of Wills & Peattie 1990; subzone E l a of Mudge & Bujak 1994). The m.f.s, of the Ypr-2 sequence lies below sample WB1 (Wrabness) at the Harwich Stone Band, within the Harwich Formation. In the central North Sea, the m.f.s, is represented by a third-order primary condensed section at the 8-9 sequence boundary of Stewart (1987), T45-T50 sequence boundary of Jones & Milton (1994) and the Dornoch-Balder sequence boundary of Mudge & Bujak (1994). The Ypr-2 sequence corresponds to the Yp-1 sequence of Hardenbol (1994), although he places the base of the Eocene succession at the base of the Oldhaven Member.

Ypresian Sequence 1 (Ypr-1)

Ypresian Sequence 3 (Ypr-3)

The unconformity at the base of the Ypr-1 sequence lies between samples LU15 and LU16 (Lower Upnor) at the base of the Gor biozone (at the base of the Harwich Formation). The lowstand phase of the Ypr- 1 sequence is not represented at outcrop. It is probably equivalent to the (lower) Cromarty Sandstone Member within the Sele Formation (S2b) in the central North Sea (Ceratiopsis wardenensis biozone of Wills & Peattie 1990; E l a Subzone of Mudge & Bujak 1994). The m.f.s, of the Ypr-1 sequence lies between samples LU 16 (LO of the Leiosphaeridia biofacies) and LU17 (Lower Upnor) within the Gor biozone (intra-Harwich Formation). In the central North Sea, the m.f.s, is represented by a third-order primary condensed section within sequence 8 of Stewart (1987), sequence T45 of Jones & Milton (1994) and the Dornoch sequence of Mudge & Bujak (1994).

The base of the Ypr-3 sequence lies between samples WB10 and W B l l (Wrabness) at the base of an impoverished interzone (intra-Wrabness Member of the Harwich Formation). A type-2 sequence boundary is envisaged at the base of the Ypr-3 sequence (a lowstand phase is therefore not present). The shelf margin wedge phase is probably represented by the Balder Formation (B 1) in the central North Sea. The m.f.s, of the Ypr-3 sequence is tentatively placed between samples HB19 and HB20 (Herne Bay) at the base of the Was biozone (within the Walton Member of the London Clay Formation). In the central North Sea, the m.f.s, is represented by a third-order primary condensed section within sequence 9 of Stewart (1987); sequence T50 of Jones & Milton (1994) and the Balder sequence of Mudge & Bujak (1994). The Ypr-3 sequence corresponds to the Yp-2 sequence of Hardenbol (1994).

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY

Recommendations

181

from onshore southeast England and northwest Europe, as well as in the North Sea, at reservoir scale. On this basis, there should be greater confidence in relating events onshore (proximal setting) to those in the North Sea Basin (distal setting). The dinoflagellate cyst stratigraphic record has the refinement to allow such undertakings only through the application of sophisticated techniques.

The present study has revealed that it is possible to use the aquatic palynomorph (particularly dinoflagellate cyst) stratigraphic record to recognize unconformities and m.f.s, in proximal settings. The merits of the type-Thanetian as a candidate for the Thanetian global stratotype section and point may therefore be more carefully considered. If it is decided that the base of the Thanetian Stage should be placed at the base of a sequence (i.e. the base of a lowstand systems tract), then a section should be proposed for consideration in which the base of the Alisocysta margarita (Area) biozone may be identified. From a practical stand point, in order to test the validity of the sequence configurations, it will be necessary to re-examine a number of key sections

Stimulating discussions with J. R. Ali, J. Hardenbol, C. Heilmann-Clausen, R. W. O'B. Knox, D. C. Mudge, H. Visscher and D. J. Ward and are gratefully acknowledged. HB acknowledges support from The Netherlands Foundation for Geosciences (GOA) and financial aid from The Netherlands Organization for the Advancement of Scientific Research (NWO), the LPP Foundation, and The Netherlands Research School of Sedimentary Geology (NSG). This is NSG publication No. 950804.

References ALl, J. R. 1994. Magnetostratigraphy of the upper Paleocene and lowermost Eocene of SE England. Geologiska F6reningens F6rhandlingar, 116, 41-42. ALLEN, L. O. 1982. Palynology of the Palaeocene and Early Eocene of the London Basin. PhD Thesis, University of London. ARMENTROUT,J. i . , MALECEK,S. J., FEARN,L. B. ET AL. 1993. Log-motif of Paleogene depositional systems tracts, Central and Northern North Sea: defined by sequence stratigraphic analysis. In: PARKER, J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 45-57. AUBRY, M.-P., HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous-nannofossil stratigraphy of the lower Palaeogene formations of the Hampshire and London basins. Journal of the Geological Society, London, 143, 729-735. BRADFORD, M. R. & WALL, D. 1984. The distribution of Recent organic-walled dinoflagellate cysts in the Persian Gulf, Gulf of Oman and Northwestern Arabian Sea. Palaeontographica Abt. B, 192, 16-84. BRINKHUIS, H. 1994. Late Eocene to Early Oligocene dinoflagellate cysts from the Priabonian type-area (Northeast Italy): biostratigraphy and paleoenvironmental interpretation. Palaeogeography, Palaeoclimatology, Palaeoecology, 107, 121-163. & ZACHARIASSE,J.-W. 1988. Dinoflagellate cysts, sea level changes and planktonic foraminifers across the K/T boundary at E1 Haria, Northwest Tunisia. Marine Micropaleontology, 13, 153-191. , POWELL, A.J. & ZEVENBOOM, D. 1992. Highresolution dinoflagellate cyst stratigraphy of the Oligocene/Miocene transition interval in NorthWest and central Italy. In: HEAD, M. J. ~: WRENN, J. H. (eds) Neogene and Quaternary Dinoflagellate Cysts and Acritarchs. American Association of Stratigraphic Palynologists Foundation, Dallas, 219-258.

-

-

ROMEIN, A. J. T., SMIT, J. & ZACARIASSE, J.-W. 1994. Danian-Selandian dinoflagellate cysts from lower latitudes with special reference to the E1 Kef section, NW Tunisia. Geologiska FOreningens FOrhandlingar, 116, 46-48. BUJAK, J. P. • MUDGE, D. 1994. A high-resolution North Sea Eocene dinocyst zonation. Journal of the Geological Society, London, 151, 449-462. CHATEAUNEUF, J.-J. 1980. Palynostratigraphie et Paldoclimatologie de l'Eoc~ne Supdrieur et de l'OligocOne du Bassin de Paris. Bureau de Recherches G6ologiques et Minibres, M6moire, 116. COSTA, L. I. & DOWNIE, C. 1976. The distribution of the dinoflagellate Wetzeliella in the Paleogene of northwestern Europe. Palaeontology, 19, 591-614. & MANUM, S. B. 1988. The description of the interregional zonation of the Paleogene (D1-D15) and the Miocene (D16-D20). Geologisches Jahrbuch, Reihe A, 100, 321-330. - - - , DENISON, C. N. D. & DOWNIE, C. 1978. The Paleocene/Eocene boundary in the Anglo-Paris Basin. Journal of the Geological Society, London, 135, 261-264. CURRY, D. 1981. Thanetian. Bulletin d'Information Gdologique du Bassin de Paris, M~moire, 2, 255-266. DOWNIE, C., HUSSAIN, M. A. & WILLIAMS,G. L. 1971. Dinoflagellate cyst and acritarch associations in the Paleogene of southeast England. Geoscience and Man, 3, 29-35. ELLISON, R. A., KNOX, R. W. O'B., JOLLEY, D. W. & KING, C. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata of the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. ELSIK, W. C. 1977. Paralecaniella indentata (Deft. & Cooks. 1955) Cookson & Eisenack 1970 and allied dinocysts. Palynology, 1, 95-102. EVlTT, W. R. 1985. Sporopollenin Dinoflagellate Cysts their Morphology and Interpretation. American --,

-

-

A . J . POWELL ET AL.

182

Association of Stratigraphic Palynologists Foundation. GRADSTE1N, E M., KR1STIANSEN, I. L., Loemo, L. & KAMINSK1, M. A. 1992. Cenozoic foraminiferal and dinoflagellate cyst biostratigraphy of the central North Sea. Micropaleontology, 38, 101-137. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. HARDENBOL, J. 1994. Sequence stratigraphic calibration of Paleocene and Lower Eocene continental margin deposits in NW Europe and the US Gulf Coast with the oceanic chronostratigraphic record. Geologiska F6reningens FOrhandlingar, 116, 49-51. HEAD, M. J. & WRENN, J. H. 1992. Neogene and

Quaternary Dinoflagellate Cysts and Acritarchs. American Association of Stratigraphic Palynologists Foundation. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks

Geologiske UndersCgelse, Series A, 7. 1994. Review of Paleocene dinoflagellates from the North Sea region. Geologiska FOreningens FOrhandlingar, 116, 51-53. HERITIER, E E., LOSSEL, E & WATHNE, E. 1979. Frigg Field - Large submarine-fan trap in Lower Eocene rocks of North Sea Viking Graben. American Association of Petroleum Geologists Bulletin, 63, 1999-2020. , & -1981. The Frigg Gas Field. ln: ILLING, L. V. & HOBSON, G. D. (eds) Petroleum Geology

-

-

of the Continental Shelf of North-West Europe. Heyden & Son, London, 380-391. HUSAIN, M. A. 1967. Dinoflagellates and Acritarchsfrom

the Eocene and Paleocene of Southeast England. PhD Thesis, University of Sheffield. IOAKIM, C. 1978. Etude Comparative des Dinoflagell6s du

Tertiaire lnf6rieur de la Mer du Labrador et de la Mer du Nord. Universit6 Pierre et Marie Curie (Paris VI). JAN DU CHI~NE, R., GORIN, G. & STUIJVENBERG,J. VAN 1975. Etude g6ologique et stratigraphique (palynologie et nannoflore calcaire) des GrOs des Voirons (Pal6ogene de Haute-Savoie, France). Gdologie Alpine, 51, 51-78. JENKINS, D. G. & LUTERBACHER, H. 1992. Paleogene stages and their boundaries. Neues Jahrbuch fiir Geologie und Paliiontologie, Abhandlungen, 186, 1-5. JOLLEY, D. W. 1992. Palynofloral association sequence stratigraphy of the Palaeocene Thanet Beds and equivalent sediments in eastern England. Review of Palaeobotany and Palynology, 74, 207-237. 1996. The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation. This volume. & SPINNER, E. G. 1989. Some dinoflagellate cysts from the London Clay (Paleocene-Eocene) near Ipswich, Suffolk, England. Review of Palaeobotany and Palynology, 60, 361-373. & -1991. Spore-pollen associations from the lower London Clay (Eocene), East Anglia, England. Tertiary Research, 13, 11-25. -

-

JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: Palaeogene stratigraphy and relative sea-level history of the Outer Moray Firth, UK North Sea. Marine and Petroleum Geology, 11, 157-165. KENNEDY, W. J. & SELLWOOD,B. W. 1970. Ophiomorpha nodosa Lundgren, a marine indicator from the Sparnacian of south-east England. Proceedings of the Geologists' Association, 81, 99-110. KING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6, 1-158. KNox, R. W. O'B. 1990. Thanetian and early Ypresian chronostratigraphy in south-east England. Tertiary Research, 11, 57-64. & Harland, R. 1979. Stratigraphical relationships of the early Palaeogene ash-series of NW Europe. Journal of the Geological Society, London, 136, 463-470. & HOLLOWAY,S. 1992. 1. Paleogene of the Central and Northern North Sea. In: KNOX, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea, British Geological Survey, Nottingham. , HINE, N. M. & ALl, J. R. 1994. New information on the age and sequence stratigraphy of the type Thanetian of southeast England. Newsletters on Stratigraphy, 30, 45-60. , MORTON, A. C. & HARLAND, R. 1981. Stratigraphical relationships of Palaeocene Sands in the UK Sector of the Central North Sea. In: ILHNG, L. V. & HOBSON, G. D. (eds) Petroleum -

-

Geology of the Continental Shelf of North-West Europe. Heyden & Son, London, 267-281. LENTIN, J. K. & WILLIAMS,G. L. 1993. Fossil dinoflagellates: index to genera and species 1993 edition. American Association of Stratigraphic Palynologists Contribution Series, 28, 856 pp. MAY, E E. 1980. Dinoflagellate cysts of the Gymnodiniaceae, Peridiniaceae, and Gonyaulacaceae from the Upper Cretaceous Monmouth Group, Atlantic Highlands, New Jersey. Paalaeontographica, Abt.B., 172, 10-116, pl. 1-23. MUDGE, D. C. & BUJAK,J. E 1994. Eocene stratigraphy of the North Sea basin. Marine and Petroleum Geology, 11, 166- 181. & -1996. An integrated stratigraphy for the Paleocene and Eocene of the North Sea. This volume. & COPESTAKE, E 1992. Revised Lower Palaeogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Palaeocene and earliest Eocene sediments from the central North Sea.

-

-

-

-

Review of Palaeobotany and Palynology,

-

-

-

-

--,

5 6 ,

327-344. 1992a. Making the most of microfossils. Geoscientist, 2(1), 12-16. 1992b. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. British Micropalaeontological Society Publication Series, Chapman & Hall, 155-251. LEWIS, J. & DODGE, J. D. 1992. The palynological

PALEOCENE--EOCENE DINOFLAGELLATE SEQUENCE BIOSTRATIGRAPHY expressions of post-Palaeogene upwelling: a review.

In: SUMMERHAYES,C. P., PRELL, W. L. & EMEIS, K.-C. (eds) Upwelling Systems: Evolution Since the Early Miocene. Geological Society, London, Special Publication, 64, 215- 226. SCHMITZ,B. 1994. The Paleocene Epoch - stratigraphy, global change and events. Geologiska FOreningens FOrhandlingar, 116, 39-41. SCHRODER, T. 1992. A palynological zonation for the Palaeocene of the North Sea. Journal of Micropalaeontology, 11, 113-126. SIESSER, W. G., WARD, D. J. • LORD, A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Micropalaeontology, 6, 85-102. STEWART,I. J. 1987. A revised stratigraphic interpretation of the Early Palaeogene of the central North Sea. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology' of North West Europe, Graham & Trotman, London, 557-576. THOMAS, J. E. 1993. The occurrence of the dinoflagellate

183

cyst Apectodinium in a cored section through the Montrose and basal Sele groups (Paleocene) of the Central North Sea. In: Correlation of the Early

Paleogene in Northwest Europe, 1-2 December 1991, (Programme and Abstracts). Geological Society, London. WALL, D., LOHMANN, G. E & SMITH, W. K. 1977. The environmental and climatic distribution of diDoflagellate cysts in Modern marine sediments from regions in the North and South Atlantic Oceans and adjacent seas. Marine Micropaleontology, 2, 121-200. WARD, D. J. 1977. The Thanet Beds exposure at Pegwell Bay, Kent. Tertiary Research, 1, 69-76. 1978. The Lower London Tertiary (Palaeocene) Succession of Herne Bay, Kent. Institution of Geological Sciences Report, 78110. WILLS,J. M. & PEATTIE,D. K. 1990. The Forties Field and the evolution of a reservoir management strategy.

In: North Sea Oil and Gas Reservoirs, Volume 2. Graham & Trotman, London, 1-23.

Recognition of Chron C25n in the upper Paleocene Upnor Formation of the London Basin, UK R I C H A R D A. E L L I S O N 1, J A S O N R. ALI 2, N I C O L E T T E M. H I N E 3 & DAVID W. J O L L E Y 3

1 British Geological Survey, Kingsley Dunham Centre, Keyworth, Nottingham NG12 5GG, UK 2 Department of Oceanography, University of Southampton, Southampton S 0 9 5NH, UK 3 Industrial Palynology Unit, Sheffield University, Mappin Street, Sheffield S1 3JD, UK Abstract: Lithostratigraphic studies of four borehole cores drilled through the late Paleocene Lambeth Group (Upnor, Woolwich and Reading Formations) and basal London Clay Formation of central London have been supplemented with palaeomagnetic, calcareous nannoplankton and palynological data. The Woolwich and Reading Formations and the lower London Clay Formation are reversely magnetized and were deposited during the early part of Chron C24r. The first record of both NP9 and Chron C25n, hitherto missing from the Paleogene record in southern England, has been identified in the Upnor Formation (formerly the Woolwich Bottom Bed). It provides a key reference marker for linking events associated with the Paleocene-Eocene boundary (positioned within Chron C24r) to the type area of the internationally recognized Thanetian and Ypresian Stages.

Work carried out to establish the chronostratigraphy of the Paleogene in southern England (Fig. 1) has been concerned principally with the late Paleocene Thanet Sand Formation and the early Eocene London Clay Formation (e.g. Aubry et al. 1986; Ali et al. 1993; Knox et al. 1994). The Thanet Sand Formation lies within calcareous nannoplankton zones NP6-NP8 and magnetochrons C26r to C25r (magnetochron nomenclature from Cande & Kent 1992). The base of Division B of the London Clay Formation (sensu King 1981) coincides with the start of Chron C24n.3n. The intervening Lambeth Group (Upnor, Woolwich and Reading Formations), Harwich Formation (Ellison et al. 1994) and lower part of the London Clay Formation are allocated to Chron C24r. Aubry et al. (1986) indicated that an hiatus in excess of 0.6 million years, marked by the absence of Chron C25n, separates the Thanet Sand Formation from the Lambeth Group. In recent years the drilling of fully cored boreholes (Fig. 2) through the Lambeth Group in central London has provided a unique opportunity for multidisciplinary study. One borehole in particular, Jubilee 404T (BGS Registration number TQ 37 NW 2118; National Grid Reference TQ 3363 7960), has become a key reference section for the Lambeth Group in central London (Ellison et al. 1994). The main reasons for its importance are firstly the wide variety of Lambeth Group facies, including marine sands, brackish lagoonal muds and pedogenic horizons, and secondly the presence

e'CO

LONDON ,DIV.Bt c,Au III111I[IIlII111 N IEl[III/I[IIlII/I[[l]IIIIIIlI11111

HARWICH

I,.,..

~o

55

56-o

0_

57--

icq

o. NP7

ORMESBY CLAY

58--

THANET SAND/

NP6

m

59-Fig. 1. Summary of the upper Paleocene stratigraphy of the London Basin based on Cox et al. (1985), Aubry et al. (1986), Ellison et al. (1994) and Knox et al. (1994); timescale following Cande & Kent (1992, 1995).

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 185-193.

185

186

R . A . ELLISON ET AL.

N THE CITY

f

eA4A

BEXLEY

LEWISHAM 9A6 BROMLEY 0 9

5km I

Fig. 2.

Location maps showing the distribution of Paleocene deposits in southeast England and boreholes referred to in the text.

of a calcareous nannofossil fauna in the lowest part of the Upnor Formation.

Stratigraphy Figure 3 shows the general late Paleogene stratigraphy in four central London boreholes from which palaeomagnetic and palaeontological data have been obtained. The Upnor Formation, from which most of the data are derived, occurs throughout the London Basin, overlying the Thanet Sand Formation in the east and overstepping on to Chalk in the west (see, for example, Curry 1992). The overlying Woolwich and Reading Formations interfinger in central London, the Reading Formation being split into two leaves by the Woolwich Formation which thins westwards (see Ellison et al. 1994 for a review). In general, the Upnor Formation consists of 2-4 m of medium-grained, glauconitic, quartzose sands with a variable proportion of well-rounded flint pebbles. In Borehole 404T (Fig. 4) the Upnor Formation, 6.04 m thick, is intensely bioturbated and with a variable glauconite content, estimated as up to 20% of the sediment. The grains are subrounded, amorphous, dark green and medium to coarse sand grade and give rise to shades of green coloured sediment. Bioturbation has resulted in mottling which, between 40.5 and 38.4 m, consists of relatively pale khaki sand in burrow infills surrounded by a highly glauconitic, dark green, structureless matrix. The top 0.8 m of the formation consists of sand and scattered glauconite grains in a mottled green and dark red-brown clay matrix. The clay is thought to be derived by eluviation

during contemporary pedogenesis of the overlying Reading Formation (Buurman 1980). Irregularshaped calcium carbonate concretions up to c. 0.1 m in diameter in the uppermost 2 m of the formation are thought also to be pedogenic in origin. A flint-pebble bed up to 2 m thick occurs in southeast London at the top of the Upnor Formation but is not usually recovered in cores. In Borehole 404T it is represented by c. 0.1 m of pebbly clay with disarticulated oyster shells. Significant features of this pebble bed are the presence of deeply-weathered flint pebbles, many of them red-stained and shattered, and clay clasts < 10 mm in diameter presumed to be derived from the Reading Formation to the west. They are interpreted as evidence for emergence and a hiatus in deposition between the Upnor Formation and succeeding Reading Formation.

Calcareous nannoplankton biostratigraphy The only significant biostratigraphical information hitherto obtained from the Lambeth Group is the identification of calcareous nannoplankton referred to NP9 in a section at Clarendon Hill in the Hampshire Basin (Siesser et al. 1987), whereas the well-exposed sections on the Isle of Wight and the coast of north Kent have proved barren. Ten samples from the Upnor Formation in Borehole 404T yielded calcareous nannofossils (Fig. 4). Assemblages range from low to moderate diversity (6-30 taxa per sample) and are generally of moderate preservation. A number of them contain rare specimens of Discoaster multiradiatus

CHRON C25N IN THE LONDON BASIN

187

Fig. 3. Correlation of boreholes in central London; filled circles identify palaeomagnetic sample positions.

Bramlette & Reidel (1954), although the assemblages are dominated by species of Toweius [Toweius tovae Perch-Nielsen (1971), Toweius eminens (Bramlette & Sullivan) Perch-Nielsen (1971), Toweius pertusus (Sullivan) Romien (1979)]. Other common species which occur include Discoaster lenticularis Bramlette & Sullivan (1961), Discoaster delicatus Bramlette & Sullivan (1961), Fasciculithus tympaniformis Hay & Mohler (1967), Placozygus sigmoides (Bramlette & Sullivan) Romien (1979), Prinsius bisulcus (Stradner) Hay & Mohler (1967), Fasciculithus cf involutus Bramlett & Sullivan (1961), Heliolithus kleinpellii Sullivan (1964), Chiasmolithus bidens (Bramlette & Sullivan) Hay & Mohler (1967), Coccolithus pelagicus (Wallich) Schiller (1930) and Zygodiscus plectopons Bramlette & Sullivan (1961). Rare Helicosphaera

and other small noelaerhabdaceae species are also recorded. D. multiradiatus, recovered from samples between 39.72 and 41.12 m, has been recorded by Perch-Nielsen (1985) as ranging from the base of NP9 (Discoaster multiradiatus zone) through to zone NP10 (Tribrachiatus contortus zone) or N P l l (Discoaster binodosus zone). According to PerchNielsen, T. tovae is restricted to zone NP9, although as Siesser et al. (1987) recorded this species from the Thanet Sand Formation, which is now attributed to nannoplankton zones NP6-NP8, its full range must be regarded as NP6-NP9. The cooccurrence of D. multiradiatus and T. tovae in Borehole 404T core indicates an NP9 age assignment. The Discoaster multiradiatus zone (NP9) of Martini (1971) is defined as the interval between

188

R . A . ELLISON E T AL.

the first occurrence (FO) of D. multiradiatus to the FO of Tribrachiatus bramlettei. The absence in the samples of species of Tribrachiatus and Discoaster diastypus or any other NP10 indicators, further supports the assignment of the lower part of the Upnor Formation to nannoplankton zone NP9. In the Okada & Bukry (1980) zonation scheme (prefixed CP) the time interval equivalent to NP9 is divided into two subzones (CP8a and CP8b), the boundary between them being based on the FO of either Campylosphaera eodela or Rhomboaster species. Samples from Borehole 404T contain neither of these species, suggesting that they belong to the Chiasmolithus bidens subzone (CP8a), which is equivalent to the lower part of NP9. This assumption is supported by the persistent occurrence of P sigmoides, the last occurrence of which was used in the mid- to low-latitude zonation scheme of Varol (1989) to define the top of his zone NTpl7 which is also equivalent to lower NP9. In conclusion, the calcareous nannofossil assemblages of the Upnor Formation in Borehole 404T are assigned to the standard nannofossil zone NF'9 (CP8a) on the basis of the co-occurrence of

Formation in Borehole 404T (Fig. 3) are comparable with the records of abundant specimens of Apectodinium species recorded in zone 6 of Powell (1988) and the Apectodinium augustum zone of Powell (1992) from the lower Sele Formation and the Forties Sandstone Member of the Central Graben, North Sea Basin (sensu Mudge & Copestake 1992). This suggests that the sporadic occurrences of the genus recorded in the Upnor Formation may be equivalent to zone 5 of Powell (1988) or the Apectodinium hyperacanthum zone of Powell (1992), which is regarded as containing the oldest records of Apectodinium spp. in the North Sea Basin. This zonal attribution places the Upnor Formation of Borehole 404T in the latest part of the late Paleocene (Thanetian), the Apectodinium hyperacanthum zone of Powell (1992) being correlated with the lower part of standard nannofossil zone NP9. Thus, the dinoflagellates are ascribed to the Apectodinium hyperacanthum zone of Powell (1992), which is correlated with nannoplankton zone NP9.

D. multiradiatus, D. mohleri, T. tovae, Discoaster delicatus and H. kleinpellii, and by the absence of species of Tribrachiatus and Discoaster diastypus.

Environment of deposition of the Upnor Formation

Palynology The palynoflora of the Upnor Formation in Borehole 404T is characterized by rich and diverse assemblages from 38.25 to 41.13m and sparse palynomorphs from 35.15 to 38.25 m (Fig. 4). The lower interval is dominated by the dinoflagellate cyst taxa Areoligera cf. coronata (Wetzel) LejeuneCarpentier 1938, Paralecaniella indentata (Deflandre & Cookson) Cookson & Eisenack emend. Elsik 1977 and Spiniferites ramosus subsp. ramosus (Ehrenberg) Loeblich & Loeblich 1966, and the gymnospermous pollen grains Inaperturopollenites hiatus (Potoni6) Pflug & Thomson 1953 and Pityosporites spp. The age of this palynoflora is best described in terms of the dinoflagellate cystbased zonations of Powell (1988, 1992), rather than pollen- and spore-based zonation schemes which are too broad to assist in interpretation of the stratigraphy (see Jolley & Spinner 1991). The presence of Apectodinium homomorphum (Deflandre & Cookson) Lentin & Williams emend. Harland 1979 and A. quinquelatum (Williams & Downie) Costa & Downie 1979 between 39.86 and 40.22 m indicate that the beds lie in the Apectodinium hyperacanthum zone of Costa & Downie (1976), suggesting a late Paleocene to early Eocene age. Abundant occurrences of Apectodinium species in the overlying shelly clays of the Woolwich

The lower part of the formation in Borehole 404T contains an abundant and diverse microplankton flora typical of a relatively normal marine environment. High frequencies of Areoligera spp. were interpreted by Islam (1983) as being typical of turbid water, shallow marine environments. The uppermost part of the formation in the borehole contains a sparse microplankton flora with occurrences of Lingulodinium machaerophorum, Leiosphaeridia spp., Paralecaniella indentata, Botryococcus braunii and Apectodinium spp., assemblages suggestive of deposition in a marine environment with high stress, such as lower than normal salinity. The pollen and spore assemblages are impoverished and dominated by Pityosporites spp., a small grain that may have been concentrated by winnowing in a relatively high energy, nearshore environment. The pollen and spore assemblages of the Upnor Formation as a whole contain taxa from Montane, Arborescent Peat Swamp, Mixed Mesophytic Forest and Nyssa-Taxodium swamp parent plant communities (see Jolley & Spinner 1991). This combination suggests that the hinterland during deposition of the Upnor Formation contained extensive areas of wet lowland.

Magnetostratigraphy Sampling and analysis for the present palaeomagnetic investigations were principally from

CHRON C25N IN THE LONDON BASIN

189

Fig. 4. Graphical section of the Upnor Formation in Jubilee 404T Borehole, showing location of samples and interpretation of magnetostratigraphic and biostratigraphic data.

Borehole 404T, supplemented by data from three other cored boreholes, namely A6 (BGS Registration No. TQ47SW/lI7; National Grid Reference 4260 7207), A4A (BGS Registration No. TQ37NE/789; National Grid Reference 3569 7588) and Jubilee 410T (BGS Registration No. TQ37NW/2124; National Grid Reference 3444 7941). Forty 8 cm 3 cubes and five 150 cm 3 half-round specimens were collected from the four borehole cores (sampling was restricted to consolidated

intervals from which specimens could be cut). All the specimens were analysed using a 2G Enterprises cryogenic magnetometer, with an in-line three-axis demagnetization system capable of generating fields of up to 60 mT; the palaeomagnetic properties are listed in Table 1. The specimens had initial natural remanent magnetization (NRM) intensities typically between 0.5 and 2.0 m A m -1. Vector end-point (Zijderveld 1967) and equal area stereographic plots were used to assess the stability of remanence of each specimen.

190

R. A. ELLISON ET AL.

Table 1. Summary of the palaeomagnetic data Hole

Spec.

Depth (m)

Rel. ht (m)

NRM (mA m -1)

Max (mT)

MDF (mT)

RI

Pol

DEC

INC

IRM ratio

Peak IRM

404T

JU49 JU48 JU47 JU46 JU45 JU44 JU43 JU42 JU41 JU40 JU39 JU38 JU37 JU36 JU77 JU76 JU20 JU20 JU35 JU20 JU34 JU20 JU20 JU33

22.56 23.02 23.53 25.97 26.18 31.56 32.05 32.45 32.75 33.13 35.28 35.77 36.61 37.36 39.00 39.52 39.59 39.67 39.72 39.86 40.26 40.33 40.40 40.70

18.57 18.11 17.60 15.16 14.95 9.57 9.08 8.68 8.38 8.00 5.85 5.36 4.52 3.77 2.13 1.61 1.54 1.46 1.41 1.27 0.87 0.80 0.73 0.43

3.540 1.580 1.670 0.396 0.439 0.305 4.020 0.341 2.224 0.407 3.050 0.310 0.360 4.690 1.777 0.736 2.102 0.853 0.977 2.109 0.197 1.110 0.658 0.173

50 50 25 55 50 55 55 55 30 55 55 30 55 25 30 30 60 60 53 60 53 60 60 55

45 22 20 NR NR 45 NR 32 NR NR 20 NR 15 13 9 14 17 7 40 11 NR 9 60 30

S S T S S S S S S T S T E S T T S T S S T S S T

R R R R R R R R '~ '~ R R R R N N N N N N 9 N N N

204.1 18.8 101.8 257.4 171.1 158.6 294.7 89.8 88.7 264.5 1.2 12.3 177.5 205.6 244.0 255.0 183.3 201.1 0.2 258.9 102.8 241.0 252.5 351.6

-19.8 -19.9 -29.9 -45.9 -45.1 -47.1 -15.1 -21.8 10.2 -6.2 -64.8 -47.8 -54.5 -75.2 67.3 45.6 76.1 35.9 58.7 64.1 -3.2 44.4 58.7 72.7

0.93

4040

0.92 0.70

3401 354

0.72 0.72 0.63 0.34

284 460 424 1647

0.96 0.95

3294 1291

0.89

214

0.91

283

0.85 0.86

243 194

JU93 JU92 JU91 JU78 JU79 JU80 JU81 JU82 JU83

11.54 13.14 14.83 17.38 18.02 25.96 26.40 27.16 31.50

23.67 22.07 20.38 17.83 17.t9 9.25 8.81 8.05 3.71

1.873 2.122 2.290 1.104 1.213 1.559 0.396 0.391 0.348

55 55 55 55 55 55 55 45 55

20 8 8 5 17 17 NR 16 9

S T T T T T S T T

9 R R R R ? R R R

197.8 64.7 154.0 335.6 142.1 195.1 6.9 218.3 251.9

-1.2 -22.6 -28.3 -35.1 -14.7 7.9 -32.7 -29.8 -32.1

0.83

197l

0.70

1062

0.76

230

0.87

184

JU53 JU52 JU51 JU54 JU50

7.58 8.24 15.38 15.79 16.89

16.92 16.26 9.12 8.71 7.61

1.185 0.797 0.418 0.444 0.450

25 55 55 55 55

20 37 NR NR NR

T S S S S

R ? R R R

15.2 119.3 248.7 281.2 46.1

-15.0 --6.6 -28.8 -32.5 -58.3

0.82

301

0.73

362

JU69 JU70 JU71 JU72 JU75 JU73 JU74

19.83 20.57 20.73 21.83 23.38 24.23 24.52

18.60 17.86 17.56 16.56 15.05 14.20 t3.9I

2.11 5.03 3.61 4.430 3.600 1.553 0.598

55 55 55 25 55 55 55

20 5 12 15 17 8 12

T T S T S S S

R R R N R R N

92.0 192.1 184.3 299.8 249.7 46.3 53.1

-72.7 -33.2 -42.2 25.2 -33.4 -44.4 70.6

0.90

2976

0.94

1433

0.93

1152

A6

410T

A4

Sample depths for each borehole are relative to the ground surface. In Borehole 404T the top of the Thanet Sand is at 41.13 m, the top of the Upnor Formation at 35.09 m and the base of the London Clay at 23.64 m. In Borehole A6 the base of the London Clay is at 18.08 m; in Borehole 410T the top of the Thanet Sand is at 24.50 m and in Borehole A4A the base of the London Clay is at 21.30 m. Relative heights are defined using the top of the Thanet Sand in Borehole 404T as the main datum and the base of the London Clay as a secondary datum. The relative height of each sample has been used to construct the composite magnetostratigraphy in Fig. 5. NRM, intial NRM intensity; Max, maximum applied alternating demagnetization field; MDF, median destructive field (NR indicates MDF not reached); RI, demagnetization reliabilty index (S, stable end point; T, trend; E, erratic); Pol, magnetic polarity (N, normal; R, reverse; ?, indeterminate); IRM ratio, IRM at 0.3 T as a ratio of the IRM at 0.86 T. The peak IRM is measured at 0.86 T.

CHRON C25N IN THE LONDON BASIN As none of the borehole cores was oriented, the polarity assignment of each specimen is based on the sign of its inclination angle (downward dipping indicates normal polarity; upward dipping indicates reverse polarity). Twenty-five of the specimens demagnetized to a stable end-point (SEP) direction (see Table 1). Twenty-two of these were assigned a polarity; the remainder had characteristic inclination angles of between _+ 10~ and were classified as indeterminate polarity. Eighteen specimens had demagnetistion trajectories which did not produce an SEP, but a reliable polarity determination has been made for 15 of them based on the 'trend' of the magnetic vector towards a particular polarity. One specimen (JU37) showed large fluctuations in direction and intensity during demagnetization and the polarity determination (reverse) is tentative. Isothermal remanent magnetization (IRM) analyses was performed on 20 representative specimens in order to determine the principal remanence carriers at different levels within each of the cores. A Molspin pulse magnetizer was used to generate the IRM in each specimen, and was measured between each of the 14 progressive field increments using a Molspin spinner magnetometer. The IRM ratio, defined by Ali (1989) as the IRM at 0.3 T as ratio of the peak field IRM (at 0.86 T), is used to provide a quantitative estimate of each specimen's IRM acquisition behaviour. Generally magnetite-rich specimens have IRM ratios typically > 0.9 whereas for hematite-bearing samples the value is typically 0.6-0.9 (see Fig. 5). IRM ratios and peak IRM values are listed in Table 1. IRM ratio data are also shown in Fig. 6. A composite magnetostratigraphic section (Fig. 5) has been constructed from polarity data from the four cores (Table 1), using the top of the Thanet Sand Formation as datum. It shows that the bulk of the Lambeth Group is of reverse polarity, but a normal polarity interval, coded CL-A, is identified in the lower part of the Upnor Formation in Jubilee borehole 404T. Two normal polarity levels identified in the Woolwich Formation and upper leaf of the Reading Formation in Borehole A4A are based on single specimens only, and the stratigraphic significance of these data is not yet fully understood. In order to assess the reliability of the magnetostratigraphic data, in particular the normal polarity magnetozone CL-A, the magnetic properties of each formation are considered. The five specimens from CL-A that were included in the IRM analyses have ratios between 0.85 and 0.95 (Fig. 6a & b). This suggests that a low-coercivity mineral, probably magnetite, is the dominant remanence carrier at these levels. The specimen (JU34) immediately above the top of CL-A has an IRM ratio of 0.96,

191

CENTRAL

24

20

LONDON

--

I

I

-

I

I

m

FM. POL.

t,.,I--

s

r I

-

<,

-90

90 Inclination

,,,,,,h

..A O

0.2 0.4 0.6 0.8 1.0 IRM ratio

Fig. 5. Composite magnetostratigraphy of the upper Paleocene sequence in central London. The top of the Thanet Sand Formation and base of the London Clay Formation in Jubilee 404T Borehole are used as datum levels in order to position samples collected from boreholes A4A and A6. On the inclination plot filled circles denote stable end pont directions, + indicates either trending or erratic directions. In the polarity column, normal polarity is shown black, reverse polarity is shown white, and indeterminate polarity is shown dotted.

indicating a similar mineralogy. The majority of specimens from the lower part of the Upnor Formation have median destructive fields (MDF) of < 20 mT (Table 1), confirming the interpretation of the IRM data that the remanence of this interval is carried by a low-coercivity mineral. In contrast, specimens from the top part of the Upnor Formation (which is affected by pedogenic processes operating during deposition of the overlying Reading Formation) (Fig. 6c), the Woolwich Formation (Fig. 6d) and the Reading Formation (Fig. 6e) have IRM ratios between 0.7 and 0.85, and MDF in excess of 30mT (Table 1). The remanence at these levels is probably being carried by hematite. Samples from the basal part of the London Clay Formation have IRM ratios of c. 0.91,

192

R.A. ELLISON ET AL.

Conclusions looo

a.

200

b. JU35

o

I''''1''''1 0.0

~

1.0

0.0

0.5

C.

1000

o

05

0.0

1.0

200

I"''1''''1 0.0

05

JU46 __

1

(01701

0.5

(0.92)

r,

I''''l''''l 0.0

v 1.0

1.0

f.

2000

J /

1.0

d.

~ I''''1''''1

o

='''1

o

0.5

I''''l''''] 0.0

0.5

1.0

APPLIED FIELD (T) APPLIED FIELD (T) Fig. 6. Examples of IRM acquisition for specimens from the central London boreholes. (a) (JU77), (b) (JU35) and (c) (JU39) are for specimens from the Upnor Formation; (d) (JU81) and (e) (JU46) are from the Woolwich and Reading formations, respectively; (f) (JU47) is from the basal London Clay Formation. IRM is expressed in m Am-2, the number in parenthesis is the specimen's IRM ratio. Magnetite-bearing specimens (a, b and f) saturate in relatively low fields and typically have IRM ratios > 0.9. Other samples, from the top part of the Upnor Formation and the Woolwich and Reading formations do not saturate and have IRM values which suggest that hematite is their principal remanence carrier.

suggesting that the remanence at these levels is dominated by magnetite. The magnetite-dominated remanence of magnetozone CL-A is likely to be primary rather than secondary (oxidation) magnetization, a conclusion strengthened by the presence of a wellpreserved nannofossil assemblage which indicates that oxidation has not occurred.

The Upnor Formation (formerly the Woolwich Bottom Bed) of central L o n d o n has yielded calcareous nannofossils indicative of the lower part of zone NP9 and a normal polarity interval, CL-A. Nannofossil zone NP9 is generally correlated with Chron C25n (see Berggren et al. 1985) and thus a first-order correlation can be made between the normal polarity interval CL-A and Chron C25n. The recognition of Chron 25n sediments in London reveals a more complete upper Paleocene succession there than elsewhere in southeast England, despite a search in the classic Paleogene outcrops at Alum, Whitecliff and Herne Bays (Aubry et al. 1986). The Upnor Formation is interpreted as highstand deposits in which flooding events are marked by pulses of glauconite deposition, winnowing and low sediment input. These flooding events may be eustatic or a response to minor local tectonic activity. As the southern margin of the London basin lies above major, roughly east-west trending, structures (Lee et al. 1990) the possibility of Paleogene synsedimentary tectonic activity, although not documented, seems likely. Of the four flooding events recognized, shown by increased glauconite abundance in Fig. 4, the third one is probably the most important having apparently been preceded by a period of erosion that removed the Chron C25n sediments from most of southern England. In addition, palynological data indicate that this erosion surface marks a change from a shallow marine, normal salinity environment to a lower salinity stressed environment. The enhanced chronological framework for the lower Paleogene of London presented here permits more precise timing of Paleocene sea-level fluctuations across SE England, and can be used to assist correlation of onshore sequences with the more complete North Sea succession. We wish to thank Ernie Hailwood, Chris King and Robert Knox for their advice and comments in discusions, and Jerry Hooker for helpful comments in reviewing the paper. NMH wishes to thank Conoco (UK) Ltd for their support during this work. The paper is published with the permission of the Director, British Geological Survey. This paper is a contribution to IGCP 308.

References ALI, J. R. 1989. Magnetostratigraphy of Early Palaeogene Sediments from N.W. Europe. PhD Thesis, University of Southampton. , KING, C. & HAILWOOD, E. A. 1993. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD,E. A & Kn3D, R. B. (eds) High

Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. AUBRY, M.-E, HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous-nannofossil stratigraphy of the lower Palaeogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-35.

CHRON C25N IN THE LONDON BASIN BERGGREN, W. A., KENT, D. V & FLYNN, J. J. 1985. Palaeogene geochronology and chronostratigraphy. In: SNELLING, N.J. (ed.) Geochronology of the Geological Record. Memoir of the Geological Society of London, 10, 141-95. BUURMAN, P. 1980. Palaeosols in the Reading Beds (Paleocene) of Alum Bay, Isle of White, U.K. Sedimentology, 27, 593-606. CANDE, S. 8L KENT,D. V. 1992. A new geomagnetic time scale for the Late Cretaceous and Tertiary. Journal of Geophysical Research, 97, 13917-13951. & 1995. Revised calibration of the geo-magnetic time scale for the late Cretaceous and Tertiary. Journal of Geophysical Research, 100, 6093-6095. COSTA, L & DOWNIE, C. 1976. The distribution of the dinoflagellate Wetzeliella in the Palaeogene of North-Western Europe. Palaeontology, 19, 591-614. COX, F. C., HAILWOOD, E. A., HARLAND, R., HUGHES, M. J., JOHNSTON, N. & KNOX, R. W. O'B. 1985. Palaeocene sedimentation and stratigraphy in Norfolk, England. Newsletters in Stratigraphy, 14, 169-185. CURRY, D. 1992. Tertiary. In: DUFF, P. McL.D. & SMITH, A. J. (eds) Geology of England and Wales. Geological Society, London, 389-408. ELLISON, R. A. KNox, R. W. O'B, JOLLEY, D. W. & KING, C. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists'Association, 105, 187-197. ISLAM, M. A. 1983. Dinoflagellate cysts from the Eocene cliff sections of the Isle of Sheppey, southeast England. Revue de Micropal~ontologie, 25, 231250. JOLLEY, D.W. & SPINNER, E. 1991. Spore-pollen associations from the lower London Clay (Eocene), East Anglia, England. Tertiary Research, 13, 11-25. KING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6. KNOX, R. W. O'B, HINE, N. M. & ALI, J. R. 1994. New information on the age and sequence stratigraphy of the type Thanetian of Southeast England. Newsletters in Stratigraphy, 30, 45-60.

193

LEE, M. K., PHARAOH, T. C. & SOPER, N. J. 1990. Structural trends in central Britain from images of gravity and aeromagnetic fields. Journal of the Geological Society, London, 147, 241-258. MARTINI, E. 1971. Standard Tertiary and Quaternary calacareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the II Planktonic Conference, Roma 1970. Edizioni Tecnoscienza, Rome, 739-785. MUDGE, D, C. 8~; COPESTAKE,P. 1992. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. OKADA, H. • BUKRY, D. 1980. Supplementary modification and Introduction of code numbers to the low-latitude coccolith biostratigraphic zonation. Marine Micropalaeontology, 5, 321-325. PERCH-NIELSEN, K. 1985. Cenozoic calcareous nannofossils. In: BOLLI, H. M., SAUNDERS,J. B. & PERCHNIELSEN, K. (eds) Plankton Stratigraphy. Cambridge University Press, Cambridge, 427-554. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for the latest Palaeocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327342. 1992. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. Chapman & Hall, London, 155-252. SXESSER W. G., WARD, D. J. & LORD. A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Journal of Micropalaeontology, 6, 85-102. TOWNSEND, H. A. & HAILWOOD, E. A. 1985. Magnetostratigraphic correlation of Palaeogene sediments in the Hampshire and London Basins, southern UK. Journal of the Geological Society, London, 142, 1-27. VAROL, O. 1989. Palaeocene calcareous nannofossil biostratigraphy. In: CRUX, J. A. & VAN HECIr S. E. (eds) Nannofossils and their Applications. Ellis Horwood, 267-310. ZIJDERVELD,J. D. A. 1967. AC demagnetization of rocks: analysis of results. In: COLLINSON, D. W., CREER, K. M. & RUNCORN, S. K. (eds) Methods" in Palaeomagnetism, Elsevier, New York, 254-286. -

-

The 'Oldhaven magnetozone' in East Anglia: a revised interpretation J. R. A L I 1, E. A. H A I L W O O D 1'2 & C. K I N G 3

1 Department of Oceanography, The University, Southampton S 0 9 5NH, UK (Present address: Earth Sciences, UNIMAS, 94300 Kota Samarhan, Sarawak, Malaysia) 2 Core Magnetics Ltd, The Green, Frostrow Lane, Sedbergh, Cumbria LAIO 5JS, UK 3 41 Montem Road, New Malden, Surrey KT3 3QU, UK

Abstract: A palaeomagnetic study has been carried out on the type Harwich Formation and basal London Clay Formation in the Ispwich area of East Anglia. Both units were deposited during the middle part of geomagnetic Chron C24r. The normal polarity magnetization previously identified in the upper part of the Harwich Formation at Wrabness (Townsend & Hailwood 1985, Journal of the Geological Society, London, 142, 1-27), and called the 'Oldhaven magnetozone' by these authors, is now regarded as a recent overprint. The new information is critical in linking the Paleocene-Eocene boundary to NW Europe where the internationally recognized early Eocene Ypresian stage stratotype is defined. The Paleocene-Eocene boundary is positioned within geomagnetic Chron C24r (e.g. Berggren et al. 1995). The first magnetostratigraphic studies of the late Paleocene-early Eocene formations of southern England were reported by Townsend & Hailwood (1985). Using calcareous nannofossils, Aubry et al. (1986) correlated the magnetostratigraphic record obtained by Townsend & Hailwood (1985) with the geomagnetic polarity timescale. Normal polarity Chron C25n was not identified in the succession. Aubry et al. (1986) suggested that this magnetochron coincided with the unconformity below the base of the Woolwich Bottom Bed (now named the Upnor Formation: Ellison et al. 1994). Aubry et al. (1986) located the start of Chron C24n.3n (previously called C24BN) at the base of 'Division B' (King 1981) of the London Clay Formation in exposures on the Isle of Wight, a finding recently confirmed by Ali et al. (1993) working on borehole material from north Kent.

The 'Oldhaven magnetozone' Townsend & Hailwood's (1985) palaeomagnetic studies of the lower Eocene in East Anglia identified a normal polarity magnetozone in the upper part of the Harwich Formation at Wrabness. They correlated this with a normal polarity magnetozone they identified both in the Oldhaven Beds (now included in the Harwich Formation) at Herne Bay, Kent, and a broadly coeval Harwich Formation section at Harefield, Middlesex. In their regional synthesis they termed this normal polarity interval the 'Oldhaven magnetozone'. Aubry et al. (1986) argued that the 'Oldhaven magnetozone' must be a short period event within Chron C24r; the suggestion that it represented Chron C25n was

disregarded since it was considered too young (?late NP9 age). However, others have informally suggested that the 'Oldhaven magnetozone' might represent Chron C25n. Today, a decade after the pioneering work of Townsend & Hailwood (1986), a re-evaluation of the reliability and significance of the 'Oldhaven magnetozone' is called for in the light of more recent biostratigraphic and palaeomagnetic data. Palynology data (Jolley 1995), indicate problems correlating the 'Oldhaven magnetozone' between Kent and East Anglia (Fig. 1). At Herne Bay, Kent, the normal polarity magnetization corresponds to palynomorph association sequences (PAS) Y4b toY5, whilst at Wrabness the supposedly correlative normal polarity magnetozone corresponds with PAS Y7b to Y9. New magnetostratiographic information relevant to the 'Oldhaven magnetozone' problem have been obtained from palaeomagnetic studies of a number of British Geological Survey borehole cores from East Anglia (summarized by Ali & Jolley 1996). The new work indicates that the Harwich Formation carries a reverse polarity magnetization throughout, which is correlated with Chron C24r. In an attempt to clarify the status of the 'Oldhaven magnetozone', at least in the typeHarwich Formation area, we have examined the magnetic polarity stratigraphy of two key outcrops in East Anglia.

Stratigraphic summary The outcrop and subcrop of Paleogene sediments in southern England is shown in Fig. 2. The lithostratigraphic nomenclature of the late Paleocene and early Eocene of SE England has recently been

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 195-203.

195

196

J.R. ALl ET AL.

Fig. 1. Summary of the stratigaphy of the upper Paleocene and lower Eocene formations of southern England (based on Ellison et al. 1994 and Jolley 1995). Geomagnetic polarity timescale from Cande & Kent (1995). PaleoceneEocene boundary from Berggren et al. (1995). Correlation with the Martini (1971) nannoplankton zonation scheme (zones NP6-NPll) is based on Aubry et al. (1986), Siesser et al. (1988), Knox et al. (1994) and Ellison et al. (1996). Note the problem of correlating the 'Oldhaven magnetozone' between north Kent and north Essex [based on the palynomorph association sequences (PAS) of Jolley 1996].

revised by Ellison et al. (1994) (Fig. 1). These authors defined a new formation, the Harwich formation, which includes all of the sediments between the top of the Woolwich-Reading Formations and the base of the London Clay Formation Walton Member. In East Anglia, the Wrabness Member of the Harwich Formation reaches a total thickness of c. 16 m. It consists of bioturbated silty clays and sandy, clayey silts with subordinate glauconitic, sandy silts and silty sands. Volcanic ash beds (up to 4 cm) are common in the upper part but occur only sporadically in the lower part. In the type area, at c. 6 m above the base of the Wrabness Member, is the distinctive Harwich Stone Band, a 15 cm calcite-cemented ash layer. Jolley (1996) places the Harwich Stone Band within PAS Y6b, and the upper part of the Wrabness Member within PAS Y9. A number of thinner concretionary bands occur within the member, but these are less persistent and cannot be used as reliable stratigraphic markers

between exposures. Correlation of the Wrabness Member tephra layers with those of DSDP sites in the NE Atlantic provides a second-order link to the NP10 nannoplankton zone, indicating an early Eocene age (Knox 1984). The member is assigned to the uppermost part of the A p e c t o d i n i u m h y p e r a c a n t h u m dinoflagellate zone which, according to Powell (1992), is of early Eocene age. Ellison et al. (1994) repositioned the base of the London Clay Formation, placing it at the base of the Walton Member. The transgression surface marking the base of this member has been identified over the whole of the London Basin and the eastern part of the Hampshire Basin (King 1981). The sediment consists of silty clays; no ash layers have been recorded in this unit. In the eastern part of the London Basin the Walton Member is typically c. 15 m thick, thinning westwards and southwards to c. 5 m at Whitecliff Bay (Isle of Wight). Jolley (1996) places the basal sediments of the Walton Member in PAS Y10. The first

'OLDHAVEN MAGNETOZONE' IN EAST ANGLIA

197

The response of each specimen to a.f. demagnetization was analysed from the changes in intensity and direction of remanence, using a combination of stereographic and orthogonal vector plots (Zijderveld 1967). On this basis the behaviour of the specimens was divided into three classes. Class 1: SEP behaviour. This refers to specimens which show progressive changes in remanence direction during the early stages of AF treatment (as one or more low-stability components of remanence were removed) but which reach a directional stable end-point (SEP) at higher demagnetization levels (when only a single characteristic component remains, e.g. Fig. 3a).

Fig. 2. Outcrop and subcrop of Paleogene deposits in southern England.

appearance of the dinoflagellate Wetzeliella astra (? = NP10) occurs at c. 2 m above the base of the Walton Member at the type locality (King 1981).

Palaeomagnetic methods Palaeomagnetic samples were obtained using the 'copper-tube' method employed by Townsend & Hailwood (1985). Two separately oriented specimens (25 mm in diameter and c. 100mm in length) were taken at each stratigraphic level. Each section was usually sampled at 1 m intervals, reduced to 0.3 m intervals across important lithostratigraphic boundaries. Natural remanent magnetization

The natural remanent magnetization (NRM) of each specimen was measured using a Molspin spinner magnetometer. Specimens had intensities typically between 0.5 and 3 0 r o A m -1. The alternating field (a.f.) method of demagnetization was applied to specimens at steps of 5 mT rising to peak values of between 30 and 50 mT. The more weakly magnetic specimens, with intensities < 0.3 mA m -z, were analysed on a CCL cryogenic magnetometer, which has an in-line three-axis stationary demagnetization system (Riddy & Hailwood 1989). Specimens with intensities > 0.3 mA m -1 were measured using the Molspin magnetometer and demagnetized using a Molspin AF demagnetizer with a two-axis tumbler.

Class 2: Trend behaviour. This refers to specimens which show a progressive directional change (i.e. a directional trend) towards a SEP, as a low stability magnetic overprint is progressively removed, but which do not reach this end-point before the magnetization becomes too weak to measure and/or erratic behaviour sets in (Fig. 3b). Class 3: Erratic behaviour. This refers to specimens which show erratic changes in direction and intensity of magnetization during demagnetization, without a clear directional trend or SEP being definable. For this sampling region a normal magnetic polarity is defined by a characteristic remanence direction with a northerly declination and positive (downward dipping) inclination (e.g. Fig. 3a). A reverse polarity is defined by a southerly declination and negative inclination (e.g. Fig. 3b). The polarity of the characteristic magnetization of each specimen was assigned on the basis of the direction of the SEP for Class 1 specimens or the direction of convergence of great circle trends for Class 2 specimens. No polarity assignment could be made for Class 3 specimens. About 40% of the specimens exhibited a SEP, and c. 55% showed clear directional trends. IRM analysis

Isothermal remanent magnetization (IRM) analysis was carried out on representative specimens in order to determine the principal magnetic minerals (normally either hematite or magnetite) contributing to the specimen's remanence. Specimens were incrementally magnetized using a Newport electromagnet (in peak fields of up to 0.86 T) and measured between steps using a Molspin spinner magnetometer. Two parameters were used to help identify the dominant magnetic mineral. These were the peak IRM value and the IRM ratio. The peak IRM is the magnetic moment generated in a

198

J.R. ALl E T A L .

(a) LEVINGTON

(b) LV 1. 1

WALTON-ON-THE-NAZE

SAMPLE NO Jo (rnA/rn) LV1, 1 136.6

1 ~ , ~ J/Jo

WN3. 1

SAMPLE NO Jo (mA/m) WN3. 1 10.945

OI J/Jo

01 0

t

t ~ 10 20 Demag Field (roT)

30

Xl

0

10

20 ' 310 ' Demag Field (mT)

40

'

50

Fig. 3. Examples of responses of typical samples to a.f. demagnetization. In each case the magnetic vector after each demagnetization step is plotted on a stereographic projection on which solid symbols represent positive inclinations (plotted in the lower hemisphere), and open symbols represent negative inclinations (plotted in the upper hemisphere). Numbers in the range 0-50 represent the a.f. treatment (in mT). Also shown for each specimen is a plot of the normalized magnetic intensity (J/Jo) v. a.f. (mT). (a) Illustrates SEP behaviour; (b) illustrates a directional trend, but with no SEE

standard-sized specimen at the maximum applied field (typically c. 0.1-1.0 A m 2 for hematite, and 5-20 A m 2 for magnetite). The IRM ratio is defined as the ratio of the IRM at 0.3 T to the peak IRM value. Magnetite-rich specimens have IRM ratios > 0.9 (e.g. Fig. 4a & b), whereas hematiterich specimens have values typically between 0.7 and 0.9 (e.g. Fig. 4c & d).

Palaeomagnetic results The Harwich Formation and lower London Clay Formation are exposed at a number of localities including river cliffs, coastal cliffs and beaches in north Essex and south Suffolk. Most of these sections typically expose < 10 m or so of section and all are partly affected by weathering. Townsend & Hailwood (1985) sampled the Wrabness and Walton Members at Wrabness, and the Harwich Stone Band at Harwich (Fig. 5). A normal polarity zone was identified in the upper part of the Wrabness Member at Wrabness. New results from nearby sections at Levington and Walton-on-the-

Naze are presented below and compared with the existing data from Wrabness and Harwich. Results of a magnetomineralogical investigation based on an IRM study of material from three of the sections are also given.

Levington A river cliff section on the east bank of the River Orwell at Levington (TM 225 387) near Ipswich was sampled (Fig. 5). This poorly documented exposure of the Wrabness Member was described briefly by George & Vincent (1976). The 10 m section begins 1.2 m below the Harwich Stone Band and extends almost to the top of the Wrabness Member. A number of degraded ash beds occur in the section. The section was sampled at 11 stratigraphic levels (palaeomagnetic sampling 'sites') including one in the Harwich Stone Band. NRM intensities vary considerably throughout the section (1-250 mA m-l), probably as a result of sampling a mixture of ash-bearing and ash-free levels. H o w e v e r there is an overall upwards

199

'OLDHAVEN MAGNETOZONE' IN EAST ANGLIA

5000-r~

J '

0.0 20000

f

a.

WN10.1 (0.77)

VVN4.1 (0.94)

I

'

0.2

I

'

0.4

I

'

0.6

I

I 0.8

0.0

I

'

0.2

I

'

0.4

I

'

0.6

I 0.8

l J

--

S

rr

~

i

1000--

o

, 0.0

i

t

0.2

1000

J

--

C,

WR1.8 (0.71)

WR1.23 (0.95)

I 0.4

'

I

'

0.6

,

I 0.8

APPLIED FIELD (T)

0.0

I'

,

0.2

i 0.4

i "'1 0.6

,

i 0.8

APPLIED FIELD (T)

Fig. 4. Examples of IRM acquisition curves for magnetite-bearing (a & b) and hematite-bearing (c & d) sediments. IRM is expressed in mA m2; the figure in parenthesis is the specimen's IRM ratio.

decrease in magnetic intensity in the section. SEP directions were defined for eight of the 21 specimens analysed. The remaining specimens exhibit well-defined demagnetization trends. The charac-

t

teristic magnetization has a normal polarity for all the sites sampled except for the Harwich Stone Band site which has a reverse polarity (Fig. 6). Four specimens from the section were subject to IRM analysis (see Table 1). The lower part of the section shows quite different IRM properties from those of the upper part, with much greater IRM ratios and peak IRM values. This behaviour is probably related to a change of the magnetic mineralogy in the section, the lower part being characterized by magnetite (or maghemite)-rich sediments and the upper part by hematite-rich sediments.

Walton-on-the-Naze

Wr:i: ion Fig. 5. Harwich Formation and London Clay Formation outcrops sampled in the Ipswich area of East Anglia.

The beach and cliff section at Walton-on-the-Naze (TM 267 238) exposes the upper part of the Wrabness Member and the lower part of the typeWalton Member of the London Clay Formation. The top 4 m of the Wrabness Member was sampled at six levels and the lower 8.3 m of the Walton Member at 11 levels. Sediments from the ash-bearing Wrabness Member had NRM intensities of c. 10 mA m -1, whereas the ash-free Walton Member had typical values of c. 1 mA m -~. The magnetostratigraphic results are presented in Fig. 6. Both members are of reverse polarity.

200

J.R.

ALl E T A L .

KEY .

(RECENT OVERPRINT)

1=

REVERSE POLARITY

10

H HARWICH STONE BAND .

.

.

WALTON-ON-THE-NAZE t w l

l/

.T

I"

'-

WRAaNESS"~ ' :

.

ii ~

'

tY

CO

1 ~

\

/

ti

--

'T

]

............................................ LEVINGTON POE~--~t-,....... 4)~\,~.ik~f....f_"-~---'--,_ _! -~-- ~- --~-~, i -8

-

:

,

6 -

',.

4-

~

-

20

i T

o -

.

2

HARWICH

H~

T-FT-f ~ ~ 0 180 -90 0 90 1

DEC

!:

INC

10 100

~ 0,8

1.0

H

E:::

[~H

1 10 100 0.8 1.0 INT (mA/m)IRM RATIO

INT (mMm) IRM RATIO

Fig. 6. Magnetostratigraphy of the sections at Levington and Walton-on-the-Naze (this study) and Wrabness and Harwich (Townsend & Hailwood 1985). Values in the range 0-12 are heights (in m) above a local datum. On the declination and inclination plots, S denotes a SEP direction and T trending behaviour. In the latter case, the declination and inclination values plotted are those corresponding to the highest demagnetization step. Palynomorph association sequence datums (based on Jolley 1996): Levington: base section = Y5, top of section = YTc; Wrabness: base of section = Y6b, base of Y7b at c. 2.4 m, Y9-Y10 boundary at Harwich Formation-London Clay junction; Walton-on-the-Naze: Y9-Y 10 boundary at Harwich Formation-London Clay junction.

Three specimens from the Wrabness Member and three from the Walton Member were subjected to IRM analysis (Table 1). The specimens from the Wrabness Member generally have greater IRM ratios and slightly higher peak IRM values (cf. Fig. 4a & c). The two specimens adjacent to the Wrabness-Walton Member junction exhibit intermediate IRM behaviour. NRM intensities decrease upsection in a similar fashion to those at Levington.

Wrabness Townsend & Hailwood (1985) examined the section at Wrabness (TM 172 323). They reported a normal polarity zone in the upper part of the Wrabness Member, commencing 2.6 m above the Harwich Stone Band and terminating at the base of the Walton Member (Fig. 6). Six of the specimens sampled by Townsend & Hailwood (1985) from the Harwich Formation and two of their specimens from the Walton Member were subjected to IRM

analysis as part of the present study (Table 1; Figs 4b & d). The Harwich Stone Band (middle) is at 0.41 m in the section and the base of the Walton Member is at 10.2 m. An abrupt change in the magnetic properties of the sediments occurs at 2.3 m (Fig. 6). Below this level the IRM ratio and peak IRM values are typically c. 0.95 and > 10 A m 2, respectively. Above this level the values are much lower (0.75 and < 5 A m 2, respectively). The NRM intensity data suggest that the transition takes place at c. 2.3 m above the base of the section (Fig. 6). This behaviour is again probably related to changes in the magnetic mineralogy, with the magnetic properties of the lower part of the section being dominated by magnetite and those of the upper parts by hematite.

Harwich The Harwich Stone Band is exposed at low tides at Dovercote near Harwich. Townsend & Hailwood (1985) drilled seven specimens from two sites

'OLDHAVEN MAGNETOZONE' IN EAST ANGLIA

201

Table 1. NRM intensity and IRM acquisition data for the Harwich Formation Wrabness Member and the London Clay Formation Walton Member in East Anglia Locality Levington

Walton-on-the-Naze

Wrabness

Unit

Spec.

Pol

Ht (m)

NRM (mA m-1)

IRM ratio

Peak IRM

H H H H W W W W H H W W H H H H H H

LV11.2 LV7.1 LV5.2 LV2.2 WN16.2 WN 10.1 WN7.1 WN6.1 WN4.1 WNI. 1 WR2.4 WR2.2 WR1.2 WR1.3 WR1.8 WRl.14 WR1.19 WR1.23

N N N N R R R R R R R R N N N N R R

8.15 4.65 2.85 0.40 11.20 5.40 4.05 3.90 2.75 0.00 11.68 10.63 9.18 8.33 5.99 3.46 1.29 0.00

2.11 3.09 12.39 30.16 0.31 0.38 2.14 5.25 7.92 15.13 1.14 1.45 1.82 8.57 1.10 2.41 26.43 30.96

0.74 0.79 0.87 0.97 0.73 0.77 0.90 0.85 0.94 0.96 0.83 0.74 0.70 0.81 0.71 0.76 0.95 0.95

1427 1866 3512 12502 1311 1132 4658 3437 8446 1320 682 1201 1167 264 1214 1951 14749 20920

H, Wrabness Member; W, Walton Member; N, normal polarity; R, reverse polarity. in the stone band. NRM intensities averaged 300 mA m -I, and all specimens exhibited a reverse polarity characteristic magnetization (Fig. 6). Summary

Magnetostratigraphic data have now been obtained from the Wrabness Member at Levington, Waltonon-the-Naze, Wrabness and Harwich (Fig. 6). The two most distant localities are only 25 km apart (Levington and Walton-on-the-Naze) and there are only minor variations in the lithology between the sections. The Harwich Stone Band in the Wrabness Member and the glauconitic base of the overlying Walton Member are the most important lithostratigraphic markers. The core-drilled specimens taken from the Harwich Stone Band at Levington, Wrabness and Harwich all exhibit a reverse polarity characteristic magnetization. However, at Levington, all sediments sampled above and below the Harwich Stone Band display normal polarity. At Wrabness, a normal polarity zone commences at c. 3.0 m above the Harwich Stone Band and continues to the top of the Wrabness Member. Two reverse polarity sites were identified in the lower part of the Walton Member. At Walton-on-the-Naze, the Wrabness Member displays reverse polarity, as did the Walton Member. At Harwich, only the reversely magnetized Harwich Stone Band was sampled. At first sight, the normal polarity zones at Levington and Wrabness appear to be at least

partially correlative. However, for reasons discussed below, we believe that they probably owe their normal polarity to recent weathering, which oxidized original magnetite in these sediments to hematite and resulted in a magnetic overprint in the recent normal polarity geomagnetic field. Firstly, the few metres of sediments above and below the distinctive and readily correlatable Harwich Stone Band are of normal polarity at Levington, but reverse polarity at Wrabness. Both of these sections are in degraded river cliffs, where the rates of erosion are likely to be much lower than in the sea cliff at Walton-on-the-Naze (where the cliffs are receding at c. 1 m per year). Consequently, there is a higher probability of recent weathering processes affecting the magnetism of the Levington and Wrabness sections than that of the Walton-on-theNaze section. However, it is necessary to explain why the reverse polarity Walton M e m b e r at Wrabness appears to have escaped weathering re-magnetization processes, even though it is located near the top of the section, where such processes might be expected to be more intense than at depth. Results from the less weathered section at Walton-on-the-Naze provide a possible explanation. The Harwich Formation and Walton Member in this section are of reverse polarity. However, the NRM intensity and IRM data indicate that the magnetic properties of the two members are quite different; the Wrabness Member is dominated by magnetite and the Walton Member by hematite. Assuming a hematite-rich original composition for

202

J.R. ALI ET AL.

the Walton Member at Wrabness, the magnetic minerals of this unit would be more resistant to further oxidizing processes and more able to withstand the normal polarity overprint which appears to have affected the upper part of the Wrabness Member which is believed to have been originally magnetitie-rich. The argument is strengthened further by the fact that in the section at Wrabness the level of the downward change from normal to reverse polarity (c. 3.0 m above the top of the Harwich Stone Band) occurs very close to that of the downward transition from hematite-rich to magnetite-rich sediments (c. 2.3 m above the top of the Harwich Stone Band) based on the IRM analyses. The normal polarity magnetization observed at Levington is believed to be the result of weathering also. The top of the Levington section is more altered (the magnetic mineralogy being dominated by hematite), with the base being only partially oxidized to maghemite. The lithified Harwich Stone Band appears to have resisted weathering and retained its original reverse polarity magnetization.

Borehole data from the Harwich Formation, East Anglia Since the pioneering studies of Townsend & Hailwood (1985) and Aubry et al. (1986) on outcrop sections, a large volume of lower Tertiary magnetostratigraphic information has become available from the East Anglia region through study of cores obtained by the British Geological Survey (Cox et al. 1985; Knox et al. 1990; Ali & Jolley, 1996). A major advantage of studying drill cores is that the recovered sediments are usually in pristine condition, having been protected from surficial weathering. All published data (summarized in Ali & Jolley 1996) indicate that across southern England the Harwich Formation Wrabness Member carries a dominantly reverse polarity magnetization (acquired during Chron C24r).

Conclusions It is concluded that both the Harwich Formation Wrabness Member and the lower part of the Walton Member of the London Clay Formation in the Ipswich area of East Anglia are characterized by reverse magnetic polarity. Recent studies of the borehole sections through the Harwich Formation of East Anglia support this observation. The preliminary result of Townsend & Hailwood (1985), indicating a normal polarity for the upper part of the Wrabness Member at Wrabness, is now believed to reflect recent weathering processes on this part of the section. Since the Wrabness Member and Walton Member together span the Apectodinium hyperacanthum to Wetzeliella astra dinoflagellate zones they can be correlated with the middle part of Chron C24r. The results presented in this paper refer to the Harwich Formation of the East Anglia region. There remains uncertainty over the reliability of the 'Oldhaven magnetozone' in Herne Bay, Kent (and Harefield, Middlesex). The magnetozone at Herne Bay was defined in lithologies which are not ideal for palaeomagnetic investigations (cross-bedded sands) and only a.f. demagnetization was used on the samples from there. It would be beneficial for the section to be re-studied using thermal demagnetization and detailed magneto-mineralogical anlyses to explore the validity of the 'Oldhaven' normal polarity magnetozone in this region.

The Natural Environment Research Council is gratefully acknowledged for financial support to JRA during his PhD studies. We would like to thank Nick Johnston and Kevin Padley for their invaluable help in the laboratory and field, and Kate Davis for drafting the figures. We are grateful to various colleagues for helpful and stimulating discussions during the development of the research, particularly Robert Knox and Norman Hamilton. Robert Knox, Niels Abrahamsen and Paul Montgomery are thanked for their constructive reviews of the manuscript.

References ALI, J. R. & JOLLEY, D. W. 1996. Chronostratigraphic framework for the Thanetian and lower Ypresian deposits of Southern England. This volume. , K1NG, C. & HAILWOOD, E. A. 1993. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD,E. A & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. AUBRY, M.-E, HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous-nannofossil stratigraphy of the lower Palaeogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735.

BERGGREN, W. A., KENT, D. V., SWISHER, C. C. HI & AUBRY, M . - R 1995. A revised Cenozoic geochronology and chronostratigraphy. In:

BERGGREN, W. A., KENT, D. V., AUBRY,M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Volume, 54, Tulsa. CANDE, S. C. t~ KENT, D. V. 1995. Revised calibration of the geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100, 6093-6095. Cox, F. C., HAILWOOD,E. A., HARLAND,R., HUGHES,M.

'OLDHAVEN MAGNETOZONE' IN EAST ANGLIA J., JOHNSTON, N. & KNOX, R. W. O'B. 1985. Palaeocene sedimentation and stratigraphy in Norfolk, England. Newsletters on Stratigraphy, 14, 169-185. ELLISON, R. A., ALl, J. R., nINE, N. M. & JOLLEY,D. W. 1996. Recognition of Chron 25n in the upper Paleocene Upnor Formation of the London Basin, UK. This volume. , JOLLEY, D. W., KING, C. & KNOX, R. W. O'B. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. GEORGE,W. & VINCENT,S. 1976. Some river exposures of the London Clay in Suffolk and Essex. Tertiary Research, 1, 25-28. JOLLEY, D. W. 1996. The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation. This volume. KING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper 6. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. - - , HINE, N. M. & Ali, J. R. 1994. New information on the age and sequence stratigraphy of the type Thanetian of Southeast England. Newsletters on Stratigraphy, 30, 45-60.

203

MORIGI, A. N., ALI, J. R., HAILWOOD, E. A., & HALLAM, J. R. 1990. Early Palaeogene stratigraphy of a cored borehole at Hales, Norfolk. Proceedings of the Geologists' Association, 101, 145-151. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the lI Planktonic Conference, Roma 1970. Edizioni Tecnoscienza, Rome, 739-785. POWELL, A. J. 1992. Dinoflagellate cysts of the Tertiary system. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellete Cysts. Chapman & Hall, London, 155-252. RIDDY, P..J. & HAILWOOD,E. A. 1989. Stationary-sample A.E demagnetisation using a coupled cryogenic magnetometer/demagnetiser system. (Abstract). Geophysical Journal Royal Astronomical Society, 88, 597. SIESSER,W. G., WARD, D. J. & LORD, A. R. 1988. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Journal of Micropalaeontology, 6, 85-102. TOWNSEND, n. A. & HAILWOOD, E. A. 1985. Magnetostratigraphic correlation of Palaeogene sediments in the Hampshire and London Basins, southern UK. Journal of the Geological Society, London, 142, --,

1-27.

ZIJDERVELD,J. D. A. 1967. AC demagnetization of rocks: analysis of results. In: COLLINSON,D. W., CREER, K. M. & RUNCORN, S. K. (eds) Methods in Palaeomagnetism. Elsevier, New York, 254-286.

Mammalian biostratigraphy across the Paleocene-Eocene boundary in the Paris, London and Belgian basins J. J. H O O K E R

Department of Palaeontology, Natural History Museum, Cromwell Road, London SW7 5BD, UK

Abstract: Problems of resolution and poor superpositional evidence in mammalian biostratigraphy through Paleocene-Eocene boundary strata in NW Europe are solved by applying parsimony analysis to taxa shared between localities. On this basis, five biozones are established in the area for the interval formerly delineated by mammalian biostratigraphers as MP7-MP9. Integration with other biostratigraphies (dinocyst, calcareous nannoplankton, charophytes) aids correlation between the London, Belgian and Paris Basins, and supports the earlier idea of diachronism of the 'argile ~t lignites' facies. The advent of 'Sparnacian' mammal faunas in Europe may coincide with a carbon isotope excursion recently recognized in the Paris Basin. This would support recent views on essential synchronism of the beginnings of both the North American Wasatchian and European 'Sparnacian' land mammal ages.

One of the most important events in mammalian history during the Cenozoic, and certainly the most important within the Northern Hemisphere Paleogene, was that which took place at or around the Paleocene-Eocene boundary. This event was a rapid faunal turnover with large numbers of extinctions in mammal groups that had been dominant in the Paleocene, accompanied by origins at ordinal and family level. The event is best represented and documented in western North America, where long continental sequences contain an essentially continuous record of mammalian fossils (Gingerich 1989; Gingerich et al. 1980; Rose 1981; Schankler 1980). In Asia, the event is best documented in Mongolia, where continental sequences have a more sporadic mammalian record (Dashzeveg 1988). Europe has the most disjointed mammalian record (Russell 1975; Russell et al. 1982a,b; Hooker 1991), but the event is striking and the area is classic for containing all the stratotypes of the globally recognized Paleocene and Eocene stages (Pomerol 1981). In Europe (as in North America), the main Paleocene groups to suffer decimation were the order Multituberculata, the Plesiadapiformes (primate relatives) and the archaic ungulates (paraphyletic order 'Condylarthra'). The incoming groups in both continents were the orders Perissodactyla, Artiodactyla, Primates (s.s.), probably Chiroptera (although not recorded in the very earliest post-event faunas) and the family Hyaenodontidae (order Creodonta). The suddenness and morphological distinctness of the appearances imply dispersal from elsewhere, but the source has not been identified, although 'the

south' is usually invoked, e.g. Africa, Central America (Gingerich 1976) or India (Krause & Maas 1990). Other incoming European groups, the orders Rodentia and Apatotheria, the marsupial family Didelphidae (Paleocene records no longer upheld: Gheerbrant 1991) and the pantodont genus Coryphodon, are thought to have their origin in North America, because of distinctly earlier appearances there (Gingerich 1989; Rose 1981). Interchange was probably via land bridges connecting Greenland to each continent (McKenna 1983). The new fauna in Europe is often termed the Hyracotherium-Coryphodon fauna, after the dominant elements in old collections, and is taken to characterize the 'Sparnacian Stage', but there are problems with this definition (see below).

Biostratigraphic problems In 1987, at the International Symposium on Mammalian Biostratigraphy and Palaeoecology of the European Paleogene, in Mainz, a mammalian biochronology was established for the Paleogene of Europe (Brunet et al. 1987). It consists of numbered units with the prefix MP. Workers are unanimous that MP6 is Paleocene and MP10 is Eocene on the criteria of any of the main organisms used to define the Paleocene-Eocene boundary (i.e. planktonic or benthic forams, calcareous nannoplankton, dinocysts, mammals). MP7-9 lie in a transition zone, with the major mammalian faunal turnover between MP6 and MP7. There is currently poor biostratigraphic resolution within the important MP7-MP9 interval mainly for two reasons. Firstly, although workers have normally

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 205-218.

205

206

J.J. HOOKER

accepted that the fauna from the Paris Basin locality of Mutigny is older than that of nearby Avenay, the consensus opinion in Mainz was that the differences were minor and that the combined faunas should be designated MP8+MP9 (Godinot in Brunet et al. 1987). Secondly, the Paris Basin localities of Pourcy and Meudon were placed in MPT, but the former was already known to have yielded several MP8+MP9-defining taxa and the latter was later shown to yield such taxa too (Russell et al. 1988) (see Hooker 1991). Russell et al. (1982b, fig. 2) show the location of all the major mammal localities of late Paleocene and early Eocene age in the London, Belgian and Paris Basins. The Paris Basin has the largest cluster, but despite this the superpositional evidence for faunal succession here is extremely difficult to find. This is because of a combination of rapid lateral facies change and poor exposure. Nevertheless, certain superpositional evidence is well established. C o r y p h o d o n , whose earliest occurrence is in MP7, was recorded in shelly lignitic sands and clays stratigraphically well above the Conglom6rat de Cernay, yielding the main MP6 fauna, in the small outlier of Mont de Berru (Dep6ret 1906). The Sables ~t Unios et T6r6dines, which at several sites in the vicinity of Epernay yield MP 10 faunas, consistently overlie an 'argile ~t lignites' facies, which in its upper part yields the MP8+MP9 fauna of Mutigny (Feugueur 1963; Riveline 1984). A complication is that the MP8+MP9 Avenay fauna occurs in sands of Sables Unios et T6r6dines facies, immediately overlying the argile ~t lignites. Michaux (1964), however, considered that the lithofacies at Avenay was subtly

different from that of typical Sables ~t Unios et T6r6dines.

Biostratigraphic solutions Methods

To avoid circular reasoning in considering the age relationships of these MP7-MP9 faunas, I have subjected them to parsimony analysis, and subsequently examined the available evidence for superposition to assess its support or otherwise for the analysis. Alroy (1992) has introduced the use of parsimony into taxonomic distributional studies. His statistical method involves the distinction between overlapping (conjunct) ranges and nonoverlapping (disjunct) ranges, together with the creation of hypothetical distributional spaces to overcome the inaccuracies caused by absent records ('apparent disjunctions') due to taphonomic or collecting biases. I have instead used a program called Phylogenetic Analysis Using Parsimony (PAUP 3.0: Swofford 1990). This program is much employed in phylogenetic analysis, but has been adopted for ecological analysis too (Lambshead & Patterson 1986). It avoids the need for hypothetical distributional spaces by simply expressing 'apparent disjunctions' as homoplasies. In the data matrix (Table 1), in contrast to a phylogenetic analysis, the locality names take the place of taxa, and taxa (numbered) take the place of characters, as in an ecological analysis. Only taxa that occur in more than one and fewer than all the localities have been used, as it is the principle of shared taxa that is being applied in order to relate the localities. A

Table 1. Data matrix of taxa and localities 00000000011111111112222222222333333333344444444 12345678901234567890123456789012345678901234567 Sezanne-B. Conde-en-Brie Avenay Mutigny Pourcy Abbey Wood Soissons Meudon Dormaal Suffolk P.B. Erquelinnes Try Berru

00000000000000000000001011000011000001001111111 00000000000000000000011111000101001101111111111 00000000000000000000111111000101111111011111110 00010000000000101101110111001011110111111100000 00100000111011100001110110111111111111000000000 00000000110000100000010100111111000000000000000 000000010110000100?0000101000000000000000000000 00010001111000010011111110000000000000000000000 000001110011110Ill00000000000000000000000000000 00001111101111100000000000000000000000000000000 00001101011000000000000000000000000000000000000 11101111110000000000000000000000000000000000000 11110000000000000000000000000000000000000000000

Localities span MP6-MP9. Taxa are restricted to those which occur in more than one and fewer than all localities within the MP7-MP9 group. Numbers attached to taxa relate to those listed in Fig. 3.

MAMMALIAN BIOSTRATIGRAPHY AND PALEOCENE--EOCENE BOUNDARY

taxon occurring at only one locality would simulate an autapomorphy in phylogenetic analysis and thus would not aid the analysis, but misleadingly increase the consistency index. Thus, the relationships between localities are established on the basis of taxa shared amongst them, minimizing the number of 'apparent disjunctions' that need to be invoked (i.e. it identifies the most parsimonious pathway linking localities). The localities are grouped into a tree, which is subsequently rooted by selection of one or more localities known to be stratigraphically the oldest (i.e. by outgroup). In this analysis, the site of Berm is used as the outgroup. This is justified because Berru together with Cernay are MP6 sites within the Sables de Rilly of the Mont de Berru outlier, which have been demonstrated to be stratigraphically below an MP7-MP9 fauna (Dep6ret 1906). Use of the Dollo-up character type in PAUP 3.0 (Swofford 1990, pp. 9-12) is essential since it ensures that all homoplasy takes the form of reversals, preventing a taxon from originating more than once in parallel. Thus, a synapomorphy simulates an origination and a reversal simulates an extinction. More than one reversal of the same taxon on different branches indicates either a local extinction or a collection failure due to taphonomic or methodological bias, within the total range of that taxon (i.e. = 'apparent disjunction'). Choice of taxa or taxonomic rank depended largely on whether there had been a recent revision and to an extent on reliability of occurrence. For instance, tillodonts were recently revised by Baudry (1992), but the occurrence of each species is so sporadic that they have been lumped here as Esthonychidae. Carnivores have been omitted. MP7-MP9 multituberculates are only partially described and have thus not been included in the analysis. Lophiodon is dealt with at genus level, at which it is readily recognizable, but its species require extensive revision (Marandat 1987). The new Meudon fauna is undescribed and I here rely on the published list (Russell et al. 1988). Clearly, much taxonomic work remains to be done and future additions to faunal lists will improve resolution.

Results Analysis, by means of a branch-and-bound search, of 13 localities and 47 taxa results in three maximum parsimony trees, each with 125 steps. The consistency index excluding uninformative taxa is 0.371. The successive nesting of the crown localities Stzanne-Broyes, Condt-en-Brie, Avenay and Mutigny, respectively, and the pairing of Pourcy and Abbey Wood at the next lower node are constant in all. Dormaal branches off at a node

207

above the Suffolk Pebble Beds in one tree, but the two form sister localities in the other two. Soissons is the most unstable, being relatively poorly represented faunally. It is the sister locality to Meudon in two trees, but sister locality to Erquelinnes in the third. An Adams consensus of the three trees shows Soissons and Meudon on the one hand and Dormaal and the Suffolk Pebble Beds on the other as forming trichotomies with the respective crown groups (Fig. l a). Analysis of the same taxa, but omitting the Soissons locality, results in four maximum parsimony trees each with 116 steps. The consistency index excluding uninformative taxa is 0.400. The only differences between them are that the relationship of Dormaal and the Suffolk Pebble Beds varies as in the original analysis and that in two trees Pourcy branches off at a node higher than Abbey Wood, the two forming sister localities in the other two. The Adams consensus is shown in Fig. lb. Removal of Soissons has slightly destabilized the relationship between Abbey Wood and Pourcy. An Adams consensus of 16 trees including all those of 116, 117 and 118 steps is shown in Fig. lc. It shows the three following locality pairs, Pourcy and Abbey Wood, Dormaal and the Suffolk Pebble Beds, and Erquelinnes and Try, as forming trichotomies with the respective crown groups, indicating the relative weakness of the evidence for the hierarchy of these localities in the maximum parsimony trees. Table 2 is a chart of the mammal occurrences used in the analysis plus some MP10 ones linking St Agnan with the Sables ~t Units et Ttrtdines localities in the vicinity of Epernay. It can be used as a range chart provided that it is recognized that the order of the Suffolk Pebble Beds and Dormaal on the one hand and of Meudon and Soissons on the other is interchangeable. The chart shows that Coryphodon does not typify the entire MP7-MP9 span, but becomes extinct part way through the sequence. The genus Hyracotherium has in the past been used in a grade sense for almost any primitive horse-like perissodactyl. These are here segregated amongst the genera Cymbalophus, Pliolophus, Propachynolophus, 'Propachynolophus" and Hyracotherium s.s. (see Hooker 1994b). The implied time order of species of the plesiadapid Platychoerops coincides with the order of nodes in an independent cladistic character analysis recently conducted on this genus (Hooker 1994a), suggesting that it is as important biostratigraphically as its Paleocene precursor, Plesiadapis (Gingerich 1976).

Biozonation To attempt some objectivity in constructing a biozonation from these data, I have summed the last

208

J.J. HOOKER

a

SEZANNE-BROYES CONDE-EN-BRIE AVENAY MUTIGNY POURCY ABBEY WOOD SOISSONS MEUDON DORMAAL S U F F O L K P.B. ERQUELINNES TRY BERRU

I

m

b

C

I

I

SEZANNE-BROYES CONDE-EN-BRIE AVENAY MUTIGNY POURCY ABBEY WOOD MEUDON DORMAAL SUFFOLK P.B. ERQUELINNES TRY BERRU SEZANNE-BROYES CONDE-EN-BRIE AVENAY MUTIGNY POURCY ABBEY WOOD MEUDON DORMAAL S U F F O L K P.B. ERQUELINNES TRY BERRU

Fig. 1. Adams consensus trees derived from analysis of the data matrix in Table 1. (a) From three maximum parsimony trees obtained from the full data matrix; (b) from four maximum parsimony trees excluding the Soissons locality; (c) from 16 trees of 116, 117 and 118 steps excluding Soissons.

occurrences of one locality and first occurrences of the next to obtain a turnover figure between each successive pair (Fig. 2). There would be little change if the order of the Suffolk Pebble Beds and Dormaal were reversed or if Soissons were deleted. These peaks also coincide with the more robust nodes of the cladograms. I have chosen the peaks to determine the zone boundaries. The zones are concurrent range zones, named below and numbered for ease of reference as PE (for Paleocene-Eocene) I-V. They differ from the MP system in that they are biozones (sensu Hedberg 1976), not reference levels (sensu Thaler 1966), and they apply to N W Europe (onshore North Sea Basin) only, not the whole of Europe. Overlapping first appearance datum (FAD) and last

appearance datum (LAD) are indicated in the definitions.

P E I - Platychoerops georgei-Cymbalophus cuniculus C o n c u r r e n t R a n g e Z o n e

Definition: Total range of Platychoerops georgei, coincident with that of one or more of the following: Cymbalophus cuniculus, Atria cf. junnei and Microparamys nanus. The zone can also be recognized by concurrence of Teilhardina belgica, Cantius eppsi and Coryphodon (FAD) with Pleuraspidotherium aumonieri and Orthaspidotherium edwardsi (LAD), provided that the last two taxa are truly contemporaneous (see

Table 2. Occurrence chart for main localities ranging from MP6 to MP IO in the Paris, London and Belgian Basins

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.

Pleuraspidotherium aumonieri Orthaspidotherium edwardsi Plesiadapis remensis Berruvius lasseroni et cf. Cymbalophus cuniculus Platychoerops georgei Arfia cf. junnei Teilhardina belgica Cantius eppsi Coryphodon Paschatherium dolloi Microparamys nanus Microhyus musculus Landenodon woutersi Hyopsodus wardi Palaeonictis gigantea Peratherium constans Amphiperatherium brabantense Platychoerops russelli Hyracotherium aft. leporinum Apatemys sigogneaui et cf. Paramys ageiensis et cf. Peratherium matronense Neomatronella Amphiperatherium maximum Peradectes louisi Palaeonictis cf. occidentaIis Pliolophus vulpiceps Arcius fuscus Microparamys russelli s.s. Phenacodus lemoinei Esthonychidae

Bunophorus cappettai Placentidens lotus Microparamys chandoni Platychoerops daubrei Apatemys mutiniacus Diacodexis varleti et cf. Lophiaspis maurettei Peradectes mutigniensis 'Propachynolophus' maldani et aft. Cantius savagei Propachynolophus levei Arcius lapparenti Donrussellia gallica Amphiperatherium bourdellense Lophiodon Nannopithex zuccolae Ailuravus michauxi Buxolestes Propachynolophus gaudryi Cuisitherium lydekkeri Protodichobune oweni Platychoerops richardsonii

B

T

E

S

D

M

S

A

P

M

A

C

S

S

E

R

R

U

0

E

0

B

0

U

V

0

E

T

R

R

Y

Q

F

R

U

I

B

U

T

E

N

Z

A

A

X

X

X

X

X

X

X

X

X X

X

X

X

X

X

X

X

X

X

X

X X

X

X X

G

X

X X

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X X X X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

X

X

X X X

X X

X

X

X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

X

X X

X

X X

X X

X

X

X

X

X

X

X

X

X

X

X X

X

X

x

X

X

X

X

X

X X

x

X

X

X

x

X

X

x

X

X

x

x

X

X

x

x

X

X

x

X

x

x x x x H

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

Numbered taxa are those used in the cladistic analysis (Table 1). Those unnumbered are additional MPI 0 taxa. H means unrecorded but occurs in higher strata. Palaeonictis gigantea and Coryphodon are added to the Soissons list as they are recorded from the Argile ~t Lignites du Soissonnais in the Soissons area, although not specifically from Soissons. Relevant taxonomic works are listed in Hooker (1991, p.77). More recent additions are: Baudry (1992), Gunnell & Gingerich (1991) and Hooker (1994a, b). Abbreviations: BER, Berru; ERQ, Erquelinnes; SUE Suffolk Pebble Beds localities (Kyson and Ferry Cliff); DOR, Dormaal; MEU, Meudon; SOI, Soissons; ABB, Abbey Wood; POU, Pourcy; MUT, Mutigny AVE, Avenay; CON, Condd-en-Brie; SEZ, Sdzanne-Broyes; STA, St Agnan; GRA, Grauves and other MPI 0 localities in the Epemay area in the Sables ~ Unios et Tdrddines.

210

J.J. HOOKER FA

GRAUVES etc

o

ST AG NAN

73 o3

SEZANNE-B.

PE III - Platychoerops daubrei-Cantius eppsi Concurrent Range Zone

LA

Definition: Coincident ranges of Palaeonictis cf. occidentalis and Pliolophus vulpiceps. Concurrence of Arcius fuscus, Phenacodus lemoinei, Microparamys russelli (s.s.), M. chancloni, Esthonychidae, Bunophorus cappettai, Diacodexis varleti, Placentidens lotus, Platychoerops daubrei and Apatemys mutiniacus (FAD) with Plesiadapis remensis, Cantius eppsi, Coryphodon, Paschatherium clolloi, Microhyus musculus and Landenodon woutersi (LAD). Reference localities: Pourcy (Marne), France

CONDE~n-BRIE AVENAY

4

3

MUTIGNY

4

6

POURCY

6

8

ABBEY WOOD

6

o

SOISSONS

1

3

MEUDON

~

o

DORMAAL

3

3

SUFFOLK P.B.

4

1

ERQU ELII~.'NES

1

0

TRY

6

2

BERRU

~

(

~

33

PTj MP6

40

_~0

'~0

fo

I.A+FA

(Falun de Pourcy, within the Argile h Lignites d'Epernay); Abbey Wood, England (Blackheath Beds); Harwich, England (Harwich Stone Band in Harwich Member, London Clay Formation).

Fig. 2. Graph of taxonomic turnover between sites as listed in Table 2. Turnover figures are obtained by summing last appearances (LA) at one site with first appearances (FA) at the next. That between Berru and Try includes the entire fauna of Berru; that between the other localities is derived from Table 2. seudoextinctions are not accounted for as it is the morphological change that is here considered important for biostratigraphic purposes.

below). FAD and acme of Paschatherium dolloi occurs in this zone. Reference localities: Try (Marne), France ('Conglom6rat h Coryphodon', at the base of the Marnes Blanches de Dormans); Erquelinnes, Belgium (Erquelinnes Sand Member, Landen Formation); Kyson/Ferry Cliff, England (Suffolk Pebble Beds); Dormaal, Belgium (Dormaal Sand Member, Landen Formation). (Note: only Platychoerops georgei occurs at all the localities.)

PE II -Platychoerops russeUi-Teilhardina

belgica Concurrent Range Zone Definition: Total range of Platychoerops russelli. Concurrence of Hyracotherium aft. leporinum, Apatemys sigogneaui, Paramys ageiensis, Peratherium matronense, Amphiperatherium maximum and Neomatronella (FAD) with Teilhardina belgica and Palaeonictis gigantea (LAD). Reference localities: Meudon (Paris), France (Conglom6rat de Meudon at the base of the Argile Plastique Bariol6e); Soissons (Aisne), France (Argile ~ Lignites du Soissonnais). Sinceny (Aisne) (Sables de Sinceny) may also belong to this zone although it contains none of the zonal indicators (see below).

PE IV - Cantius savagei-Arcius fuscus Concurrent Range Zone

Definition: Concurrence of 'Propachynolophus' aft. maldani, Lophiaspis maurettei, Peradectes mutigniensis and Cantius savagei (FAD) with Hyopsodus wardi, Peratherium constans, Amphiperatherium brabantense, Arcius fuscus and Hyracotherium aft. leporinum (LAD). Reference locality: Mutigny (Marne), France (near the top of the Argile h Lignites d'Epernay).

PE V -Donrussellia gallica-Apatemys sigogneaui Concurrent Range Zone

Definition: Total range of Propachynolophus level Concurrence of Arcius lapparenti, Donrussellia gallica and Amphiperatherium bourdellense (FAD) with Bunophorus cappettai, Apatemys sigogneaui, Placentidens lotus, *Peradectes mutigniensis, *Phenacodus lemoinei, *Lophiaspis maurettei, *Platychoerops daubrei, *Microparamys chandoni, *M. russelli (s.s.) and *Paramys ageiensis (LAD); and of Lophiodon (FAD) with the asterisked taxa only.

Reference localities: Avenay (Marne), France (Sables ~ Unios et T6r6dines); Cond6-en-Brie (Aisne), France (Sables de Cuise); S6zanne-Broyes (Marne) (Sables h Unios et T6r6dines).

Minor faunas Harwich.

The Harwich Stone Band within the Harwich Member of the London Clay (= Wrabness Member, Harwich Formation of Jolley 1996) has yielded Pliolophus vulpiceps (the holotype) which is restricted to PE III. Between Harwich and St Osyth, the holotype jaw of Coryphodon eocaenus

MAMMALIAN BIOSTRATIGRAPHY AND PALEOCENE--EOCENE BOUNDARY

was found by offshore dredging. Although no horizon is recorded for this specimen, dredging was in the last century used to mine the Harwich Stone Band for cement manufacture (Whitaker 1885, p. 17), so it is likely that the Coryphodon came from a nearby horizon within the Harwich Member.

Sinceny. Amongst the small fauna from the Sables de Sinceny, only Pliolophus aft. vulpiceps (see Hooker 1994b) and Dipsalidictis cf. transiens (see Gunnell & Gingerich 1991, p. 177) give a clue to the age. P aft. vulpiceps otherwise occurs only at Soissons, whilst D. cf. transiens otherwise occurs in Europe only at Meudon, both PE II (personal observation).

Evidence for superposition Paris Basin The Avenay (PE V) fauna occurs in Sables h Unios et T6rddines facies immediately overlying Argile Lignites facies (Guernet 1981); the Mutigny (PE IV) fauna occurs 3 m below the top of the Argile Lignites (Riveline 1984, p. 145-146) at a lateral distance of 1.5 km from the Avenay quarry. The Pourcy (PE III) fauna occurs in a sandy coquina (falun) within Argile ~ Lignites facies (Laurain & Barta 1985, p. 42-43) 13 km from the Mutigny locality and so is impossible to stratify with respect to the latter. The best way of demonstrating mammal succession seems to be by means of associated dinocyst and charophyte zonal taxa, whose succession is documented. Thus, Pourcy (PE III) has yielded two species of dinocyst (D. E. Russell, pers. comm.) that, according to Powell (1992), occur no earlier than zone D7B (= W5), the highest dinocyst zone recorded from the Argile a Lignites facies of the Montagne de Reims area (i.e. at Verzenay and Mailly; Gruas-Cavagnetto et al. 1980). On this basis, it is likely to be close in level to Mutigny. The suggested partial reworking of the Pourcy fauna, thus giving it an overall older aspect (Cavelier 1987, p. 263-264) is a possibility, but the rolled appearance of the isolated teeth is insufficient evidence on its own and should in any case affect only certain faunal elements. Mutigny (PE IV) has yielded the charophyte Peckichara piveteaui (see Riveline 1984). This has also been recorded in the upper part of the Argile Lignites of the Fosse-Parisis quarry at Mt Bernon, Epernay (Grambast 1977), where, in a borehole, a different species, P disermas, occurs throughout the underlying Marnes Blanches du Mt Bernon (Riveline 1984). P. disermas also occurs in the Cendrier and Argile Plastique Bariolde at Passy in

211

the Paris area (Riveline 1984) just above the horizon with the Meudon mammal fauna (PE II). At Soissons, the Argile ~ Lignites with the PE II mammal fauna near the top (apparently in the Sables h Paludines at the Grande Sdminaire pit; de Lapparent 1939) are overlain by the Sables de Sinceny (Faluns Sableux) with a diversity of dinocysts of the genus Apectodinium. In the overlying Falun Supdrieur (Argiles ~t Cyrbnes et Huitres), Apectodinium dominates the dinocyst assemblage (Bignot et al. 1981). The same stratigraphic distribution of dinocysts is present at Sinceny (Gruas-Cavagnetto 1968, p. 21-22), where two mammals uniquely shared with the PE II Soissons locality occur in the Sables de Sinceny. At neither locality is there any sign of Wetzeliella, which occurs at Mt Bernon near the base of the Argile ~ Lignites (namely, W. meckelfeldensis, a D6B indicator; Laurain et al. 1983); so these strata at Soissons and Sinceny must pre-date D6. The dinocysts therefore demonstrate that PE II is below PE III. Try (PE I) is the most difficult site to relate stratigraphically. Its fauna is presumed to have come from the vertebrate-rich 'conglomdrat ~t Coryphodon', although, apart from this taxon, the remaining elements were found in quarry spoil (Louis et al. 1983). Suggestions by these authors of mixed ages for the fauna were based on the association of three Cernaysian and two Sparnacian taxa. One of the former, Plesiadapis tricuspidens, has now been reidentified as Platychoerops georgei (Hooker 1994a) and typifies PE I. Bearing in mind the rarity of mammaliferous horizons and the consistency of preservation type in the assemblage, it seems equally likely that the fauna is homogeneous and that the two remaining MP6 representatives (Pleuraspidotherium aumonieri and Orthaspidotherium edwardsi), as well as a champsosaur (D. E. Russell, pers. comm.), are survivors from an earlier time. A final resolution to the problem can only come from recollecting in situ. Whichever the outcome, a PE I locality with MP6 survivors, or closely superposed MP6 and PE I faunas at the same site, Try has great biostratigraphic potential. The 'conglom6rat b, Coryphodon' is sandwiched between 15 m of Marnes Blanches de Dormans above and the marnes calcaires Paludina aspersa (a probable equivalent of the Calcaire de Rilly) below (Hdbert 1853; Feugueur 1963, p. 334). The Marnes Blanches are capped by Argile h Lignites with brackish molluscs (Hdbert 1853). The succession of thick white marls followed by lignitic shelly clays is similar to that documented at Mt Bernon (Laurain et al. 1983). The 'conglom6rat ~t Coryphodon' has been equated with the Conglom6rat de Meudon (Feugueur 1963) as it occurs at the base of a 'Sparnacian' succession.

<++

+

~.

+

++

++ .+-+: ~ +~]+ +

+-~j ~"

~

,,+, <~

<,<

~

~

z

+++_~ ~,.,,,,

,",

II1

m

o9

w

m

m

5 Z


Z

CO

+,,,~ II ~,~0++++,<+

co

I

z

Z z

0

<

m

I'%

r

,1V70 ~3d31

~t

,<

I~+,> S_++:+~) 3+<) )~ zp u~"

o z


o<

~m

0

I~I-.I

o,.

AV-IO

NOaNOI

Lu

ci.

z v

i 0~I
l:1

~ ~Jt,~

m

/

~lJ~

c~:, ~ u l ~

T

T

,o~

1 /

O'J

O_~u-

~;

MAMMALIAN BIOSTRATIGRAPHY AND PALEOCENE--EOCENE BOUNDARY However, the m a m m a l faunas are distinctly different (as noted above). An alternative interpretation has been offered by Laurain & Meyer (1986, p. 108-110). These authors equated the 'conglom6rat ~t Coryphodon' with the Marnes ~t Rognons, which at various sites overlie Calcaire de Rilly/Marnes de Chenay with an MP6 mammal fauna and Sphaerochara edda zone charophytes. The Marnes ~ Rognons themselves can frequently be shown to belong to the base of the Peckichara disermas charophyte zone by the occurrence of P. microcarpa, which is restricted to this horizon (sites of Banthelu and Guitrancourt: Riveline 1984). P microcarpa also occurs in the apparently lateral equivalent, the Calcaire de Mortemer. Both t h e - M a r n e s h Rognons and the Calcaire de Mortemer underlie (depending on the site) either Argile Plastique Bariol6e or Argile ~t Lignites du Soissonnais, both of which contain PE II mammals. It is probable therefore (although unproven) that the Try (PE I) mammals come from the Marries Rognons Calcaire de Mortemer level. Discovery of charophytes at Try could provide the necessary tests.

L o n d o n Basin At Ferry Cliff the Harwich Member with the Harwich Stone Band (containing the PE III indicator Pliolophus vulpiceps at Harwich) directly overlies the Suffolk Pebble Beds with a PE I fauna. However, I have already argued (Hooker 1991) that the latter fauna is reworked on the basis of an associated assemblage of charophytes and bithyniid

213

opercula which otherwise co-occur in a clay lens within a cross-bedded sand unit (the 'brown sand' of Hester 1965, fig. 3) overlying the Bottom Bed at Harefield, Berkshire. The reworking hypothesis is supported by the presence of large clay clasts in the Suffolk Pebble Beds (Hooker 1991) and an abundance of Woolwich Beds pollen (Jolley & Spinner 1991). Another PE III locality is Abbey Wood. The mammals are in sandy shelly Blackheath Beds that fill channels, or at least valleys, eroded in the underlying Woolwich Beds. Similar but better exposed Blackheath Beds channel fills at Swanscombe, Kent, have a basal pebble bed that is confluent with that at the base of the Oldhaven Member that overlies them. The Blackheath Beds can thus be shown to be younger than the Woolwich Beds, which contain a disermas zone charophyte (Stephanochara curryi), and slightly older than at least one of the pre-Wetzeliella London Clay members (namely the Oldhaven Member) (Costa & Downie 1976; Costa et al. 1978, Knox et al. 1983). Thus, both the Harwich and Abbey Wood PE HI faunas are relatively close in age but apparently much older than Pourcy. The Woolwich Shell Beds also contain the mammal Coryphodon, so indicate no greater accuracy than PE I-III.

Belgian Basin The only accurately datable Belgian Basin mammal faunas are from Dormaal and Erquelinnes (PE I), so no superposition is demonstrable here.

Fig. 3. Correlation scheme for lithostratigraphic units in the London, Belgian and Paris Basins, calibrated to the geomagnetic polarity timescale of Cande & Kent (1992). Magnetostratigraphy has so far only been documented in the London Basin and in the leper Clay of the Belgian Basin (Ali et al. 1992; Ellison et al. 1996) and correlation elsewhere is by means of biostratigraphy alone. Time calibration of London Basin strata within Chron 24r is based largely on Berggren & Aubry (1996). The correlation framework relies heavily on the dinocyst zonation (Costa & Manum 1988; Chateauneuf & Gruas-Cavagnetto 1978; Gruas-Cavagnetto 1968, 1974, 1976; Knox et al. 1983; Laurain et al. 1983; Powell 1992) backed up by calcareous nannoplankton (NP) (Aubry 1983; De Coninck et al. 1981; Siesser et al. 1987) and charophytes (Riveline 1984). Mammal zones (PE) and relevant reference levels (MP) are added to this framework, which they support and in the lower part of the sequence help to refine. In particular they support the idea of diachronism of the Argile ~tLignites facies (Cavelier 1987). Lateral equivalence of Argile Lignites du Soissonnais with Argile Plastique Bariolre and Marnes Blanches du Mont Bernon in the Pads Basin is supported by charophytes and mammals and parallels that of the Woolwich Shell Beds and Reading Mottled Clay in the London Basin. However, whereas the tectonically controlled hiatus between the Woolwich Shell Beds and the London Clay in the London Basin is well established, the equivalent hiatus shown in the Paris Basin may alternatively have been partly or completely filled by the Falun Suprrieur and/or the Sables d'Auteuil. Although the complete thickness of the London Clay is shown, erosion has reduced its thickness progressively northeastwards until only the lower units are present in the Harwich area. Major unconformities are indicated by a wavy line. Abbreviations: A, Argiles; BB, Blackheath Beds; C, calcaire; HSB, Harwich Stone Band; L, lower; M, Marries; Mbr, Member; S, Sables; SL, striped loams; SPB, Suffolk Pebble Beds; U, upper; WSB, Woolwich Shell Beds. Charophyte zones: EDDA, Sphaerochara edda; DIS, Peckichara disermas; (M), range of P. microcarpa within DIS; PIV, P. piveteaui. Mammal localities: (Av), Avenay; (AW), Abbey Wood; (CB), Cond6-en-Brie; (Me), Meudon (Conglom6rat de Meudon); (Mu), Mutigny; (Py), Pourcy; (SB), Srzanne-Broyes; (SH), Studd Hill; (Shp), Sheppey; (So), Soissons.

214

J.J.

HOOKER

Integrated zonation Thus, the order of mammal zones indicated by the PAUP analysis fits with the known sequence of dinocyst and charophyte zones. Only two slight anomalies exist. Firstly, PE III appears very long, spanning five dinocyst zones (D5B to D7B). One possible explanation that would be worth investigating is as follows. The absence of a record of zone D7A between D6B and D7B in either the Fausses Glaises or Argile ~ Lignites d'Epernay (Gruas-Cavagnetto 1976; Gruas-Cavagnetto et al. 1980; Laurain et al. 1983) might mean a major hiatus in this sequence, which could have allowed reworking of much of the Pourcy fauna from D6B into D7B sediments, thus potentially shortening the timespan of PE III. Secondly, of two mammal species known from London Clay B at Studd Hill, Herne Bay, one Platychoerops richardsonii occurs for the first time in the Paris Basin in MP10, following its possible ancestor, P. daubrei in PE III-V. According to the dinocyst zones, this should equate with about the middle of the Argile h Lignites d'Epernay below the level of Pourcy (PE III). There seems a much stronger case for the integrity of the dinocyst zones than for one hypothetical mammalian ancestor-descendant relationship; and it is likely that P. richardonii originated earlier in southern England and dispersed to France in MP10 times, where it first occurs at Saint-Agnan (Louis et al. 1983). In fact, a degree of endemism in southern England is suggested by the presence of Hyracotherium leporinum s.s. at both Studd Hill and higher parts of the London Clay at Sheppey, but not in the Paris Basin (Hooker 1994b). Figure 3 shows correlation of the major lithostratigraphic units of the London, Belgian and Paris Basins in the late Paleocene and early Eocene, based on integration of the dinocyst, charophyte, calcareous nannoplankton and mammal zonations. The dinocysts form the major framework, but the mammals have the potential to improve resolution in the region of zones PE I-II if their stratigraphical relationships can be better established. This could be important as this is a time interval when dinocysts seem less reliable, because of relatively long ranges, and are lacking from the more alkaline sediments where mammals are often best represented. Figure 4 shows calibration of the mammal zones with the other zonations based on this study. The proposed correlation of PE I faunas across the three basins as discussed above (see Fig. 3) leads to several observations. In the London and Belgian Basins they are associated with crossbedded fluvial sands (Dormaal Sand and Erquelinnes Sand Members of the Landen Formation and the 'brown sand' of Hester 1965). It

NP

DINOCYSTS

MAMMALS

CHAROPHYTES

MP 10 D8 PE V PE I f f D7B

-PIVETEAUI _

D7A 11 D6B

PE III

~A

PE II

D5

DISERMAS

PE I

2MI_.CROCARP._A-

MP 6

EDDA

9 D4 8

Fig. 4. Calibration of mammal zones with those of dinocysts, calcareous nannoplankton (NP) and charophytes against the timescale as in Fig. 3.

implies that there is a large hiatus in the Dormaal and Erquelinnes areas of Belgium between the Landen and leper Clay Formations. This gap is probably filled, in part at least, further north in Belgium by the Knokke Member, which is of Woolwich Shell Beds facies (Laga & Vandenberghe 1990).

Wider implications Correlation has long been made between the early Wasatchian of western North America and the 'Sparnacian' (as defined by mammal faunas here distributed between zones PE I-V) of Europe on the basis of many shared mammal genera (e.g. Savage 1971). Very few species, however, have been recognized to be in common [until recently only Pachyaena gigantea (Cope); see Boule 1903], which is strange in view of the degree of intercontinental interchange invoked for this period of time. One reason for this could be a slow diachronous dispersal event, allowing time for speciation. Diachronism has often been invoked for the dispersal event leading to the advent of the earliest Wasatchian or 'Sparnacian' faunas. It has often been claimed that base 'Sparnacian' is significantly earlier than base Wasatchian, overlapping most of the late Paleocene North American Clarkforkian age (e.g. Godinot 1982; Hooker 1980; Rose 1980).

MAMMALIAN BIOSTRATIGRAPHYAND PALEOCENE--EOCENEBOUNDARY Table 3. Closely related species in Wasatchian 0 (WaO in the Clarks Fork Basin, Wyoming, North America, and in zone PE I in NW Europe (see Gingerich, 1989, 1993; Hooke 1994a). North America Cantius torresi Teilhardina brandti Arfia junnei Diacodexis ilicis 'Hyracotherium' sandrae

Europe Cantius eppsi Teilhardina belgica Arfia cf. junnei Diacodexis sp. (Dormaal) Cymbalophus cuniculus

The discovery of an earlier fauna of Wasatchian type (Wa0) with some species either very similar or identical to ones in Europe reinstated the idea of synchrony of Wasatchian-base 'Sparnacian' (Gingerich 1986, 1989, 1993). In Europe, close relatives of some of these restricted Wa0 species occur in zone PE I (Table 3). However, the extrapolation of the NP9-NP10 boundary from the North American Atlantic coast to the Bighorn Basin, Wyoming, via the first occurrence of Platycarya pollen (Wing et aL 1991) seemed to favour once again the diachronism idea. This is because this datum in the Bighorn Basin lies just below the base of Wa0, whereas the NP9-NP10 boundary of the North Sea Basin, as extrapolated from the Atlantic via the -17 ash, lies at the base of the London Clay Formation (Knox 1990) approximately coincident with the PE II-III boundary. A hopefully independent source of correlation comes from carbon isotope stratigraphy. The brief strong negative excursion documented in the deep sea by Kennett & Stott (1991) has been recognized from tooth enamel and soil carbonate at the Clarkforkian-Wasatchian boundary in the Bighorn Basin (Koch et al. 1992). A similar excursion has been recognized in a sandy calcrete (Marne Rognons) at the base of the main Argile Plastique Bariol6e sequence at Limay in the Paris Basin (Sinha & Stott 1994; Stott et al. 1996; Thiry 1981, pp. 19-21, pers. comm.). Three kilometres away

215

at Guitrancourt, the Marne ?~ Rognons contains Peckichara microcarpa (Thiry 1981, p. 23; Riveline 1984, pp. 101-102), thus providing a tentative link between the carbon isotope excursion and mammal zone PE I and supporting essential synchroneity of the mammal dispersal event in North America and Europe. The short-lived nature of the warming event deduced from the excursion could have allowed the rapid northward expansion of mammalian ranges, normally invoked to allow the intercontinental exchange via Greenland. It could equally rapidly have caused their contraction when temperatures cooled again a few thousand years later, resulting in isolation and rapid speciation in the respective continents. The MP6 mammal faunas occur in the Marnes de Chenay at Montchenot (Louis in Laurain & Meyer 1986, p. 117) and in the underlying Sables de Rilly (including the Conglom6rat de Cernay) at Cernay and Berru (Russell 1964). The Sables de Rilly at Montchenot are recorded as containing A p e c t o d i n i u m h o m o m o r p h u m and A. p a r v u m (Gruas-Cavagnetto 1974, p. 5). The first occurrence ofA. parvum is associated with D5A (Aau) (Powell 1992, pp. 169, 177) and in the London Basin (Powell et al. 1996) the Bottom Bed, thus within the lower part of NP9. As the bases of both NP9 (Aubry 1985) and the Clarkforkian (Butler et al. 1987) occur within Chron 25N, MP6 must be coeval with a part of the Clarkforkian.

I would especially like to thank Dr D. E. Russell for much helpful advice and discussion on Paris Basin mammal faunas, and for providing access to collections in the Laboratoire de Pal6ontologie, Mus6um National d'Histoire Naturelle, Paris. I am grateful to him and to Dr A. B. Smith for critically reading the manuscript and for discussion of methodologies. Drs P. D. Gingerich and D. E. Russell provided important casts. Drs M.-P. Aubry, W. A. Berggren, J. Hardenbol and M. Thiry created stimulating field discussion in the Paris Basin. Ms B. West carefully drafted Fig. 3. This paper forms part of ongoing research within IGCP Project 308 and is a contribution from the NHM/UCL-BkB project on Global Change and the Biosphere.

References

ALl, J. R., KING, C. & HAILWOOD, E. A. 1992. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD,E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. ALROV, J. 1992. Conjunction among taxonomic distributions and the Miocene mammalian biochronology of the Great Plains. Paleobiology, 18, 326-343. AUBRY, M.-E 1983. Biostratigraphie du Pal6og~ne 6picontinental de l'Europe du nord-ouest. Etude

fond6e sur les nannofossiles calcaires. Documents du Laboratoire de Gdologie de la Facultd des Sciences de Lyon, 89, 1-317. - 1985. Northwestern European Paleogene magnetostratigraphy, biostratigraphy, and paleogeography: calcareous nannofossil evidence. Geology, 13, 198-202. BAUDRY, M. 1992. Les tillodontes (Mammalia) de l'Eocbne inf6rieur de France. Bulletin du Musdum National d'Histoire Naturelle, Paris, (C4)14, 205-243.

216

J.J. HOOKER

BERGGREN, W. A . & AUBRY, M.-P. 1996. A late Paleocene--early Eocene NW European and North Sea magnetobiochronological correlation network: a sequence stratigraphic approach. This volume. BIGNOT, G., GRUAS-CAVAGNETTO, C., GUERNET, C., PERREAU, M., RENARD, M., RIVELINE, J. t~ TOURENQ, J. 1981. Le Sparnacien de Soissons (Aisne, France): 6tude srdimentologique et palrontologique. Bulletin d'Information des Gdologues du Bassin de Paris, 18(4), 3-19. BOULE, M. 1903. Le Pachyaena de Vaugirard. Mdmoires de la Soci~t~ G~ologique de France, 10(4), 1-16. BRUNET, M., FRANZEN, J. L., GODINOT, M., HOOKER, J. J., LEGENDRE, S., SCHMIDT-KITTLER, N. t~ VIANEY-LIAUD, M. (coordinators) 1987. European reference levels and correlation tables. Miinchner Geowissenschaftlicher Abhandlungen, (A)10, 13-31. BUTLER, R. E, KRAUSE, D. W. & GINGERICH,P. D. 1987. Magnetic polarity stratigraphy and biostratigraphy of Middle-Late Paleocene continental deposits of south-central Montana. Journal of Geology, 95, 647-657. CANDE, S. C. & KENT, D. V. 1992. A new geomagnetic polarity timescale for the late Cretaceous and Cenozoic. Journal of Geophysical Research, 97(B10), 13 917-13 953. CAVELIER, C. 1987. Diachronisme des faci6s et discontinuitrs dans le Palrogbne. Exemple du Thanrtien et de l'Ypr6sien en lie-de-France. Essai de grnrralisations et corrrlations dans les bassins anglais, beige et parisien. Bulletin d'Information des G~ologues du Bassin de Paris, M~moire Hors S~rie, 6, 243-270. CHATEAUNEUF,J.-J. & GRUAS-CAVAGNETTO,C. 1978. Les zones de Wetzeliellaceae (Dinophyceae) du bassin de Paris. Bulletin du Bureau de Recherches G~ologiques et Minikres, (2)4(2), 59-93. COSTA, L. I. & DOWNIE, C. 1976. The distribution of the dinoflagellate Wetzeliella in the Palaeogene of north-western Europe. Palaeontolog); 19, 59-64. & MANUM, S.B. 1988. The description of the interregional zonation of the Paleogene (D1-D15) and the Miocene (D16-D20). In: V1NKEN,R., VON DANIELS, C. H., GRAMANN, E, KOTHE, A., KNOX, R. W. O'B., KOCKEL, F., MEYER, K.-J. & WEISS, W. (eds) The Northwest European Tertiary Basin. Results of the IGCP Project No 124. Geologisches Jahrbuch, (A)100, 321-330. --, DENISON, C. & DOWNIE, C. 1978. The Palaeocene/Eocene boundary in the Anglo-Paris Basin. Journal of the Geological Society, London, 135, 261-264. DASHZEVEG, D. 1988. Holarctic correlation of nonmarine Palaeocene-Eocene boundary strata using mammals. Journal of the Geological Society, London, 145, 473-478. DE CONINCK, J., DE DECKER, M., DE HEINZELIN, J. WmLEMS, W. 1981. Uage des faunes d'Erquelinnes. Bulletin de la Soci~t~ Beige de G~ologie, 90, 121-154. DE LAPPARENT,A. E 1939. Une coupe du Sparnacien

Soissons (Aisne). Annales de la Socidtd gdologique du Nord, 64, 7-18. DEPERET, C. 1906. Relations stratigraphiques des faunes de Cernay et de Meudon au mont de Berru. Bulletin de la Soci~t~ Ggologique de France, (4)6, 442-443. ELLISON, R. A., ALI, J. R., HINE, N. M. & JOLLEY, D. W. 1996. Recognition of Chron 25n in the upper Paleocene Upnor Formation of the London Basin, UK. This volume. FEUGUEUR,L. 1963. L'Yprgsien du Bassin de Paris. Essai de monographie stratigraphique. M6moire du Service de la Carte Grologique de France, Paris, 1-568. GHEERBRANT, E. 1991. Bustylus (Eutheria, Adapisoriculidae) and the absence of ascertained marsupials in the Palaeocene of Europe. Terra Nova, 3, 586-592. GINGERICH, P .D. 1976. Cranial anatomy and evolution of early Tertiary Plesiadapidae (Mammalia, Primates). Papers on Paleontology, Museum of Paleontology, University of Michigan, 15, 1-145. -1986. Early Eocene Cantius torresi - oldest primate of modern aspect from North America. Nature, 320, 319-321. - 1989. New earliest Wasatchian mammalian fauna from the Eocene of northwestern Wyoming: composition and diversity in a rarely sampled highfloodplain assemblage. Papers on Paleontology, Museum of Paleontology, University of Michigan, 28, 1-97. 1993. Early Eocene Teilhardina brandti: oldest omomyid primate from North America. Contributions from the Museum of Paleontology, University of Michigan, 28, 321-326. --, ROSE, K. D. & KRAUSE, D. W. 1980. Early Cenozoic mammalian faunas of the Clark's Fork Basin-Polecat Bench area, northwestern Wyoming. Papers on Paleontology, Museum of Paleontology, University of Michigan, 24, 51-68. GODINOT, M. 1982. Aspects nouveaux des 6changes entre les faunes mammaliennes d'Europe et d'Amrrique du Nord h la base de l'Eocbne. Gdobios, Mdmoire Spdcial, 6, 403-412. GRAMBAST, L. 1977. Etude sur les charophytes tertiaires d'Europe occidentale. II Esp~ces nouvelles de l'Eoc~ne infdrieur. Pal~obiologie Continentale, 12(2), 1-35. GRUAS-CAVAGNETTO, C. 1968. Etude palynologique des divers gisements du Sparnacien du Bassin de Paris. M(moires de la Socidtd Ggologique de France, 110, 1-114. -1974. Associations sporopolliniques et microplanctoniques de l'Eoc~ne et de l'Oligocrne infrrieur du Bassin de Paris. Pal~obiologie Continentale, 5(2), 1-20.

1976. Etude palynologique du sondage de la Drfense (Paris). Revue de Micropaliontologie, 19, 27-46. , LAURAIN, M. & MEYER, R. 1980. Paysage vrgdtal et position stratigraphique du sommet des Lignites du Soissonnais dans la Montagne de Reims (Yprrsien, Bassin de Paris). G~obios, 13, 947-952. GUERNET, C. 1981. Ostracodes sparnaciens du Bassin de

MAMMALIAN BIOSTRATIGRAPHY AND PALEOCENE-EOCENE BOUNDARY Pads (France). Revue de Micropaldontologie, 24, 51-56. GUNNELL,G. E & GINGERICH,P. D. 1991. Systematics and evolution of late Paleocene and early Eocene Oxyaenidae (Mammalia, Creodonta) in the Clarks Fork Basin, Wyoming. Contributions from the Museum of Paleontology, University of Michigan, 28, 141-180. HEBERT, E. 1853. Note sur l'age des sables blancs et des marnes h Physa gigantea de Rilly, en rrponse ~ la communication faite par M. Prestwich, dans la srance du 21 frvrier 1853. Bulletin de la Socigtg Gdologique de France, (2)10, 436-454. HEDBERG, H. D. 1976. International Stratigraphic Guide. John Wiley & Sons, New York. HESTER, S. W. 1965. Stratigraphy and palaeogeography of the Woolwich and Reading Beds. Bulletin of the Geological Survey of Great Britain, 23, 117-137. HOOKER, J. J. 1980.The succession of Hyracotherium (Perissodactyla, Mammalia) in the English early Eocene. Bulletin of the British Museum (Natural History) (Geology), 33, 101-114. 1991. The sequence of mammals in the Thanetian and Ypresian of the London and Belgian Basins. Location of the Palaeocene-Eocene boundary. Newsletters on Stratigraphy, 25, 75-90. 1994a. A new species of Platychoerops (Plesiadapiformes, Mammalia), from the latest Palaeocene of the Paris, London and Belgian Basins. Ggobios, 27, 343-352. 1994b. The beginning of the equoid radiation. Zoological Journal of the Linnean Society, 112, 29-63. JOLLEY, D. W. 1996. The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation. This volume. & SPINNER, E. 1991. Spore-pollen associations from the lower London Clay (Eocene), East Anglia, England. Tertiary Research, 13, 11-25. KENNETT, J. P. & STOTT, L. D. 1991. Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229. KNOX, R. W. O'B. 1990. Thanetian and early Ypresian chronostratigraphy in south-east England. Tertiary Research, 11, 57-64. --, HARLAND,R. & KING, C. 1983. Dinoflagellate cyst analysis of the basal London Clay of southern England. Newsletters on Stratigraphy, 12, 71-74. KocH, P. L., ZACHOS, J. C. & GINGERICH, P. D. 1992. Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary. Nature, 358, 319-322. KRAUSE, D. W. & MAAS, M. C. 1990. The biogeographic origins of late Paleocene-early Eocene mammalian immigrants to the Western Interior of North America. Geological Society of America Special Paper, 243, 71-105. LAGA, P. & VANDENBERGHE,N. (eds) 1990. The Knokke well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M~moires pour servir gl l'Explication des Cartes Ggologiques et Minikres de la Belgique, 29, 1-118. -

-

217

LAMBSHEAD, E J. D. & PATERSON, G. L. J. 1986. Ecological cladistics - an investigation of numerical cladistics as a method for analysing ecological data. Journal of Natural History, 20, 895-909. LAURAIN, M. & BARTA,L. 1985. Journres d'rtude 13-14 Octobre 1984, Palrog~ne de la Montagne de Reims et de la Montagne d'Avize. Bulletin d'Information des Ggologues du Bassin de Paris, 22(1), 17-45. & MEYER,R. 1986. Stratigraphie et palrogrographie de la base du Palrogbne champenois. Ggologie de la France, 1986, 103-123. --, BARTA,L., BOLIN, C., ETAL. 1983. Le sondage et la coupe du Mont Bernon ~ Epernay (Marne). Etude srdimentologique et palrontologique du stratotype du Sparnacien et de la srrie 6oc~ne. G~ologie de la France, 3, 235-254. LouIs, P., LAURAIN, M., BOLIN, C. & BARTA, L. 1983. Nouveau gisement de vertrbrrs dans le Cuisien sup6rieur de Saint-Agnan (Aisne) ses relations stratigraphiques avec les autres gisements yprrsiens du Bassin parisien. Bulletin d'Information des Gdologues du Bassin de Paris, 20(4), 3-20. MCKENNA, M. C. 1983. Cenozoic paleogeography of north Atlantic land bridges. In: BOTT, M. H. P., SAXOV, S., TALWANI,M. & THIEDE,J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Press, New York, 351-399. MARANDAT,B. 1987. La biostratigraphie mammalienne de l'Eoc~ne continental du versant sud de la Montagne Noire. Reconsidrration d'apr~s les nouvelles donnres. Miinchner Geowissenschaftlicher Abhandlungen, (A)10, 149-158. MICHAUX, J. 1964. Age des Sables ~ Unios et Trrrdines (Eocene infrrieur) d'Avenay (Marne) et leurs relations avec les sables du meme nom des gisements classiques d'Epernay. Comptes Rendus Sommaires des S~ances de la Socigtg Ggologique de France, 1964, 103. POMEROL, C. (ed.) 1981. Stratotypes of Paleogene stages. Bulletin d'Information des Gdologues du Bassin de Paris, Mdmoire Hors Sgrie, 2, 3-301. POWELL, A. J. 1992. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. Chapman & Hall, London, 155-229. , BRINKHUIS, H. & BUJAK, J. P. 1996. Upper Paleocene - Lower Eocene dinoflagellate cyst sequence biostratigraphy southeast England. This volume. RJVELINE, J. 1984. Les gisements ~t charophytes du Crnozoique (Danien h Burdigalien) d'Europe occidentale: lithostratigraphie, biostratigraphie, chronostratigraphie. Bulletin d'lnformation des Ggologues du Bassin de Paris, Memoire Hors S~rie, 4, 1-523. ROSE, K. D. 1980. Clarkforkian Land-Mammal Age: revised definition, zonation, and tentative intercontinental correlations. Science, 208, 744-746. 1981. The Clarkforkian land-mammal age and mammalian faunal composition across the Paleocene-Eocene boundary. Papers on Paleontology, Museum of Paleontology, University of Michigan, 26, 1-197. RUSSELL, D. E. 1964. Les mammif~res palroc~nes -

-

218

J.J. HOOKER

d'Europe. Mdmoires du Musdum National d'Histoire Naturelle, Paris, (C)13, 1-321. 1975. Paleoecology of the Paleocene-Eocene transition in Europe. Contributions to Primatology, 5, 28-61. --, GALOYER,A., LouIs, P. & GINGERICH,P. D. 1988. Nouveaux vertrbrrs sparnaciens du Conglomrrat de Meudon h Meudon, France. Comptes Rendus Hdbdomadaires des S~ances de l'Acaddmie des Sciences, Paris, (I1)307, 429-433. & 32 others. 1982a. Tetrapods of the Northwest European Tertiary Basin. Geologisches Jahrbuch, (A)60, 5-74. HARTENBERGER, J.-L., POMEROL, C., SEN, S., SCHMIDT-KITTLER, N. & VIANEY-LIAUD,M. 1982b. Mammals and stratigraphy. Palaeovertebrata, M(moire Extraordinaire, 1-77. SAVAGE, D. E. 1971. The Sparnacian-Wasatchian mammalian fauna, Early Eocene, of Europe and North America. Abhandlungen des Hessischen Landesamtes fiir Bodenforschung, 60, 154-158. SCHANKLER, D. M. 1980. Faunal zonation of the Willwood Formation in the central Bighorn Basin, Wyoming. Papers on Paleontolog); Museum of Paleontology, University of Michigan, 24, 99-114. SIESSER, W. G., WARD, D. J. & LORD, A. R. 1987. Calcareous nannoplankton biozonation of the -

-

,

Thanetian Stage (Palaeocene) in the type area. Journal of Micropalaeontology, a, 85-102. SINHA, A. & STOTT,L. D. 1994. New atmospheric pCO 2 estimates from paleosols during the late Paleocene/early Eocene global warming interval. Global and Planetary Change, 9, 297-307. STOTT, L. D., SINHA, A., THrRu M., AUBRY, M.-P. & BERGGREN, W.A. 1996. Global 813C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine correlations. This volume. SWOFFORD, D. L. 1990. PAUP, Phylogenetic Analysis Using Parsimony, Version 3.0. User's manual, Illinois Natural History Survey, Champaign. THALER, L. 1966. Les rongeurs du Bas-Languedoc dans leurs rapports avec l'histoire du Tertiaire d'Europe. Mdmoires du Musdum National d'Histoire Naturelle, Paris, (C)17, 1-296. THIRY, M. 1981. Srdimentation continentale et altrrations associres: calcitisations, ferruginisations et silicifications. Les Argiles Plastiques du Sparnacien du Bassin de Paris. Sciences Gdologiques, 64, l-173. WHITAKER,W. 1885. The Geology of the Country around Ipswich, Hadleigh, and Felixstow. Memoir of the Geological Survey of England and Wales, 1-156. WING, S. L., BOWN, T .M. & OBRADOVICH,J. D. 1991. Early Eocene biotic and climatic change in interior western North America. Geology, 19, 1189-1192.

The earliest Eocene sediments of eastern England: an ultra-high resolution palynological correlation DAVID W. J O L L E Y

Centre for Palynological Studies, University of Sheffield, Mappin Street, Sheffield S1 3JD, UK Abstract: The acritarchs, dinoflagellate cysts, algae, pollen and spores from the Harwich Formation sediments in 28 sections from southeastern England have been quantitatively studied to assist in the stratigraphic correlation of sedimentary units. Application of published zonations based on dinoflagellate cyst zonations does not enable a subdivision of the Harwich Formation to be proposed; to address this problem, quantitative palynological data has been studied to define associations which have been grouped into association sequences. Thirteen subdivisions of the Harwich Formation palynofloras have been achieved, enabling an accurate correlation of units with an c. 38 500 year duration across southeastern England. This correlation has identified two major depositional sequences, deposited during separate periods of rising relative sea level during the 54.5-54.0 Ma period. Application of the association sequence technique to three equivalent sections from the North Sea Basin and one from the eastern margin of the WyvilleThompson Ridge, has allowed a comparison of the southeastern England onlap model with depositional sequence models proposed for the outer Moray Firth by other authors. Correct identification of the relationship of the deposits in southeastern England to the Dornoch, Balder and Sele Formations of the North Sea, and identification of the stratigraphic level of tephra layers -17 and +19, enables comments to be made about the position of the Paleocene-Eocene boundary in the North Sea.

Sandy and silty sediments which occur between the base of the London Clay (Prestwich 1846) and the top of the Lambeth Group (Ellison et al. 1994) in the London and Hampshire Basins of southeastern England were first identified as a sedimentary unit by Prestwich (1850), who referred them to his London Clay basement-bed (Fig. 1). Subsequently, Whitaker (1866) separated the laminanted, crossbedded and pebbly sands of north Kent, which he named the Oldhaven Beds, from the glauconitic sandy silts that composed the majority of the basement-bed sensu Prestwich (1850). This terminology received widespread use (e.g. Stamp 1921; Curry 1965) until a review of the lithologies and the lithostratigraphic terminology used to describe them was undertaken by King (1981). He defined the London Clay Formation to include the dominantly argillaceous sediments which lie unconformably on the Oldhaven Beds, or older Tertiary sediments. In an expansion of the concept of Whitaker (1866), the Oldhaven Beds of Kent were included with the sandy basement-bed of the western London and Hampshire Basins (divided into the Tilehurst and Twyford Members), in the newly defined Oldhaven Formation. King followed the work of Curry (1965) in regarding the pebbly sands occurring in the east London area (the Blackheath Beds) as stratigraphically older than the basement-bed and deposited as part of a separate, and earlier, post-Lambeth Group transgression.

However, unlike Curry, who regarded the Oldhaven Beds as equivalent to the Blackheath Beds on faunal grounds, King (1981) suggested that the latter were deposited during a post-Lambeth Group-pre-London Clay transgression. This suggestion was subsequently questioned by Ellison (1983) who interpreted the Blackheath Beds as representing the deposits of fluvial channels within the Lambeth Group. In the northern part of the London Basin, in the area of the Stour and Orwell estuaries of Suffolk, a unit of ash-bearing silty claystones and sandy siltstones had been recorded as the 'ash series' by Knox & Ellison (1979) and compared with the Danish Fur and Olst Formations (BOggild 1918; Andersen 1937). King placed the Suffolk ash series, which he named the Harwich Member, within his Division A1 of the London Clay Formation, along with the Swanscombe Member, a glauconitic silty sand occurring in the south and central London Basin. However, it is apparent from the description given by King in 1981, that the lower part of the ash series was not examined due to lack of exposure at Wrabness. All of these basement-bed sediments are disconformably overlain by the blue-grey claystones with white siltfilled burrows referred to the Walton Member, London Clay Formation (Division A2) by King (1981), In the far west of the Hampshire Basin these blocky claystones are not recorded; here King

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 219-254.

219

220

D.W. JOLLEY

Fig. 1. The historical development of lithostratigraphic terminology for the Haxwich Formation and component beds.

described claystones of Division A3 resting directly on the glauconitic silty sands of Division A1 (the basement-bed). Recently, a series of workers have made modifications to the lithostratigraphic nomenclature of King (1981). The Tilehurst Member (King 1981) was transferred to the London Clay Formation by King & Curry (1992), along with the Herne Bay Member (renamed the Oldhaven Member); resulting in the abandonment of King's (1981) Oldhaven Formation, which previously contained them. However, this reattribution did not clarify the status of the Twyford Member, which was apparently unattributed or abandoned. This review was followed by a comprehensive modification of the hierarchical lithostratigraphic terminology of the early Eocene sediments by Ellison et al. (1994). Their proposal erected a Harwich Formation which incorporated sediments previously included in the Harwich, Swanscombe and Tilehurst Members of King (1981), together with the Oldhaven Beds and Blackheath Beds. This new formation roughly

equates to the basement-bed of Prestwich (1850) and, as yet, has not been subdivided into members; in particular, the claystones present in north Norfolk and recorded by Knox et al. (1990) as the Hales Clay remain in informal nomenclature. In this study, two new members of the Harwich Formation are defined, the Wrabness and Orwell Members (see Appendix A). Overlying the Harwich Formation, the London Clay Formation has remained in the definition of King (1981), with the oldest unit, the Walton Member, resting disconformably on Harwich Formation sediments. The base of the Harwich Formation (referred to as Division A1) in the Hampshire and London Basins was regarded as being diachronous by King (1981), a suggestion previously made by both Stamp (1921) and Wrigley (1940) who had argued for a steady transgression to the west. However, Curry (1965) pointed out that evidence for this was lacking. Later, using the occurrence of an ostracod, Cytheridia unispinae, in the Swanscombe Member, together with the geo-

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND graphical distribution of the sediments of the Oldhaven Formation, King suggested that two transgressions were responsible for the deposition of the Oldhaven Formation and later London Clay Formation. In addition, he compared the London Clay sea transgression to a dated Quaternary example, suggesting that the base of each transgressive cycle was virtually isochronous. This double transgression model was not supported by dinoflagellate cyst analyses conducted by Knox et al. (1983) on samples taken from five sections of the Harwich Formation (Division A1), Walton Member and Division A3 of the London Clay Formation across the London and Hampshire Basins (Fig. 1). These authors reported evidence for significant diachronism within the members of King (1981). The dinoflagellate cyst assemblage of the Swanscombe Member (Harwich Formation) ranged in attribution from the Apectodinium hyperacanthum zone of Costa & Downie (1976) in the east to the Wetzeliella astra zone of Costa et al. (1978) in the west London Basin. Similarly, the dinoflagellate cysts present in the Tilehurst Member were attributed to the A. h y p e r a c a n t h u m zone in the east and to the Wetzeliella m e c k e l f e l d e n s i s zone of Costa & Downie (1976) in the south of the Hampshire Basin. The authors remained in favour of a twophase transgression across the London and Hampshire Basins from the east; significant lithological diachronism was invoked to account for the differing ages of the sediments. Since the study of Knox et al. (1983), research into the stratigraphy of the Harwich and London Clay Formations has not clarified the relative age of the lithological units. Earliest Eocene sediments in the North Sea Basin have been correlated with those of southeastern England on the basis of biostratigraphy and lithostratigraphy (e.g. Knox et al. 1981) and by tephrostratigraphy (Knox & Morton 1983, 1988), but the detail of this relationship remains unclear. It is the aim of this study to present a detailed, palynology-based correlation of sediments in southeastern England and a comparison with selected North Sea Basin sections to assist in the understanding of depositional sequences in the earliest Eocene.

Lithologies, southeastern England At outcrop in south Suffolk and north Essex (Fig. 2), and subcropping in Norfolk, are sandy silts of the Harwich Formation, with numerous tephra layers. The formation stratotype at Wrabness (TM 1724 3260, Fig. 3) in north Essex is the most extensive exposure of this unit in eastern England. Here, 32 ash beds have been recorded (author) in the silty claystones. The Harwich Stone Band, a

221

tabular concretionary horizon, outcrops at the base of the exposure and the upper limit of the formation is marked by sands which are disconformably overlain by claystones of the Walton Member. Originally, the definition of the Harwich Member given by King incorporated all the sediments between the base of the Suffolk Pebble Bed (Boswell 1916, 1927; Figs 4-6), which rests on the Reading Formation, and the base of the Walton Member. This included silty sands and sandy siltstones cored in the British Geological Survey (BGS) Shotley Gate Borehole (TM 2437 3460; Fig. 4) below the Harwich Stone Band. King has recently identified (pers. comm. in Jolley & Spinner 1991) a disconformity just below the Harwich Stone Band, which can be recognized at the type section (Fig. 3). Below this disconformity the sediments are dominantly sandy silts with only disrupted tephra layers, and were referred to the Swanscombe Member of King (1981) by Jolley & Spinner (1991). However, the non-glauconitic nature of these sediments in comparison with the glauconitic Swanscombe Member of Kent and the west London Basin, mitigates against the use of this name for the deposits in East Anglia, as does the apparent lack of lithological differentiation of the Tilehurst and Swanscombe Members. Although the proposed lithostratigraphic nomenclature proposed by Ellison et al. (1994) contributes to the resolution of the previously confused terminology, the absence of a formal name for the distinctly tuffaceous sediments of East Anglia and the silty sand with disrupted tephra layers underlying them, is a handicap in descriptive lithostratigraphy. It is addressed here with the proposal of the Wrabness and Orwell Members, formal lithostratigraphic subdivisions of the Harwich Formation of Ellison et al. (1994). Observations on the lithologies exposed, cored and angered at localities in Norfolk, Suffolk and Essex (Figs 3-8) has enabled the recognition of two sedimentary units in the Wrabness Member, and of three units within the Orwell Member. The relationship of these to other units is illustrated later in the text, and described with the Wrabness and Orwell Members in Appendix A. In the Norfolk area (Fig. 2), the Harwich Formation is recognized in the BGS Hales Borehole (TM 3671 9687) and Ormesby A Borehole (TG 5148 1424). Here (Figs 9 & 10), sandy silts contain up to 87 tephra layers in the most complete section recorded (Ormesby A) and rest disconformably on silty claystones which are not recognized further to the south. These sediments rest unconformably on the late Paleocene Ormesby Clay Formation (Ellison et al. 1994) and were named the Hales Clay by Knox et al. (1990). Volcanoclastic layers were recorded by Cox et al.

222

Fig. 2. Location map showing the studied sections.

D.W. JOLLEY

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

223

Fig. 4. Summary log, BGS Shotley Gate Borehole, Suffolk. See Fig. 3 for key.

(1985) at the base of the Hales Clay in the Ormesby Borehole and subsequently four bentonized ash layers were identified in the Hales Borehole (Knox et al. 1990). The Hales Clay was also recorded in the offshore-Norfolk BGS borehole 79/07a ( 5 8 ~ 1~ resting disconfomably on the Ormesby Clay Formation (see Jolley 1992 and Fig. 11). Examination of the gamma-ray log for the Hales Borehole shows three sedimentary sequences capped by claystones with a high gamma-ray log response, these are partly equivalent to units A - C of the Orwell Member in the Ipswich area, their correlation is discussed below. Sediments equivalent to unit A of the Wrabness Member are clearly recognized in the Ormesby Borehole, where they are 24 m thick and

Fig. 3. Summary log, Wrabness section, Essex. The lithostratigraphic units are shown on the left, a graphic lithology in the centre with the sample positions indicated by ticks, and the association sequences on the right. The far right column indicates the number of associations present in the section.

Fig. 5. Summary log, Nacton section and auger Hole, Suffolk. See Fig. 3 for key.

224

D.W. JOLLEY

Fig. 8. Summary log, Walton Naze section, Essex. See Fig. 3 for key.

Fig. 6, Summary log, Ferry Cliff section, Woodbridge, Suffolk. See Fig. 3 for key.

Fig. 7. Summary log, Sizewell C3 Borehole, Suffolk. See Fig. 3 for key.

are overlain by a 4 m thick, tephra-layer free, sandy siltstone assigned to unit B of the Wrabness Member. In south Essex the base of the Harwich Formation is not currently exposed, and records therefore rely on borehole data, some of which were available for this study. In southeast Essex a series of boreholes drilled at Bradwell (TM 600208) encountered silty sands of the Orwell Member with units A--C recognizable (Fig. 12), resting disconformably on Lambeth Group sediments. Here, a more arenaceous representation of unit C lacks the tephra layers seen in the Ipswich area, and is overlain by the tuffaceous siltstones of Wrabness Member unit A (7 m in thickness). This has ten tephra layers capped by ash-free siltstones and an unstructured sand attributed to unit B (0.30 m in thickness). Disconformably overlying this are blue-grey claystones with white silt-filled burrows assigned to the Walton Member. In west Essex the base of the Harwich Formation is composed of sandy silts which have been referred to the Swanscombe Member (King 1981). These rest unconformably on the Lambeth Group and are overlain by claystones of the Walton Member and the London Clay Formation. Borehole coverage in west Essex is sporadic but the Stanford Rivers Borehole (TQ 5399) penetrated this section (Fig. 13); a similar lithological succession was also recorded in the BGS Dowsett's Farm Borehole, Collier's End (TL 38062079). This borehole penetrated 3.75 m of silty sands resting unconformably on the Reading Formation and overlain by the Walton Member (Fig. 14). In the Greater London area the basal beds of the Harwich Formation are attenuated or missing. In the Channel Tunnel Rail Link Borehole A4A (TQ 53561758; Fig. 15) and in the Jubilee Line Extension Borehole 810 (TQ 37NE 1593), a thin

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND

225

Fig. 9. Summary log, BGS Hales Borehole, Norfolk; after Knox et aI. (1990), details of the ash distribution within the Harwich Formation were not given by these authors. See Fig. 3 for key.

representation of the Harwich Formation can be seen resting disconformably on the Lambeth Group (Fig. 16). Within A4A a single volcanoclastic layer is recorded in the siltstones, which are overlain disconformably by the Walton Member. In northeast London, the Chingford South Borehole (TQ 380920) recorded c. 1.5 m of sandy silts (Fig. 17) which can be attributed to the Harwich Formation. A similar sequence was cored in the Sheppey 1 Borehole (TR 015665) on the Isle of Sheppey (Figs 2 & 18), referred to as the Harty Borehole by Ali et aL (1993). Here, a thin representation of the Oldhaven Beds is overlain by sandy silts of the undifferentiated Harwich Formation. However, further to the east at Herne Bay (TR 206687) and Shelford Pit (TR 160600), and to the west at Upnor Pit (TQ759711), the Harwich Formation section is composed solely of the cross-laminated sands and mud lenses of the Oldhaven Beds (Figs 19-21). North of London the arenaceous units of the Harwich Formation are thicker. At South Mimms (TL 227 007; Fig. 22), a section exposed during the cutting of the M25 motorway showed silty claystones of the Walton Member overlying 1.90 m of silty sandy claystones with scattered

pebbles at the base attributed to the Swanscombe Member by King (1981). In turn, these overlie glauconitic sandy, clayey siltstones with a distinctive shelly fauna and sporadic well-rounded flint pebbles at the base. These sediments where referred to the Tilehurst Member by King (1981) and rest unconformably on the Lambeth Group. Neither the Swanscombe or Tilehust Members of King (1981) are utilized in this study, due to difficulties in identification. The descriptions of King (1981) demonstrate little difference between the two units, other than those likely to be affected by decalcification and weathering. As no significant lithological difference exists between the two units at South Mimms, both units are attributed to the undifferentiated Harwich Formation. Further to the west, in Buckinghamshire a similar section was exposed at Fulmer (TQ 020840) in cuttings for the M40 motorway (Fig. 23). Here, the Walton Member rests with a pebbly disconformity on a problematic unit referred by King (1981) to the Walton-Swanscombe Member which also shows a basal accumulation of well-rounded black flint pebbles. A sharp junction separates this unit from the underlying sandy siltstones attributed to the Swanscombe Member by King (1981), here some

226

D.W. JOLLEY

1.40 m in thickness. A further sharp lithological boundary separates this unit from the underlying sandy siltstones attributed to the Tilehurst Member by King, which is recorded as being 1.I0 m in thickness, resting disconformably on the Lambeth Group. In the western part of the London Basin the section at Tilehurst Quarry (SU 683 734) in Berkshire

Fig. 11. Summary log, BGS Borehole 79/07A. See Fig. 3 for key.

Fig. 10. Summary log, BGS Ormesby A Borehole, Norfolk; after Cox et aL (1985). See Fig. 3 for key.

Fig. 12. Summary log, Bradwell 205 Borehole, Essex. See Fig. 3 for key.

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

227

Fig. 16. Summary log, 810 Borehole, London, See Fig. 3 for key.

Fig. 13. Summary log, Stanford Rivers Borehole, Essex. See Fig. 3 for key.

Fig. 14. Summary log, BGS Dowsett's Farm Borehole, Colliers End, Hertfordshire. See Fig. 3 for key.

Fig. 15. Summary log, A4A Borehole, London. See Fig. 3 for key. Fig. 17. Summary log, Chingford South Borehole, London. See Fig, 3 for key.

228

D.W. 3OLLEY

Fig. 18. Summary log, Sheppey 1 Borehole, Kent. See Fig. 3 for key.

provided the stratotype section of the Tilehurst Member of King (1981), which is some 3.82 m in thickness (Fig. 24). Here, the Harwich Formation rests disconformably on the Lambeth Group and contains sandy and silty lenses with a shelly fauna, becoming increasingly sandy in the upper part. Overlying the Harwich Formation is the Walton Member which has a disconformable contact with scattered pebbles. The nearby section at Knowl Hill, Berkshire (SU 819798), contains a similar succession. Here, 4.8 m of silty sands are overlain by < 1 m of siltstones, referred to as the Swanscombe Member by King (1981), which have a glauconite-rich bed associated with scattered pebbles above a disconformity (Fig. 25). This unit is in turn overlain by claystones of the Walton Member. In the Hampshire Basin the silty sands of the

Fig. 19. Summary log, Herne Bay section, Kent. See Fig. 3 for key.

Fig. 20. Summary log, Shelford Pit section, Kent. See Fig. 3 for key.

Harwich Formation unconformably overlie the Reading Formation in several localities. The Harwich Formation in the BGS Shamblehurst Borehole (SU 493146) and the sections at Chichester Harbour (SU 829081) and Whitecliff Bay (SZ 640860) (Figs 26-28) are silty sandy clays with shelly lenses, often decalcified. In the sections at Chichester Harbour and Whitecliff Bay these

Fig. 21. Summary log, Upnor Pit section, Kent. See Fig. 3 for key.

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

229

Hg. 22. Summary log, South Mimms section, London. See Fig. 3 for key.

Fig. 25. Summary log, Knowl Hill section, Twyford, Berkshire. See Fig. 3 for key.

Fig. 23. Summary log, Fulmer section, Buckinghamshire. See Fig. 3 for key.

Fig. 26. Summary log, BGS Shamblehurst Borehole, Hampshire. See Fig. 3 for key.

Fig, 24. Summary log, Tilehurst section, Berkshire. See Fig. 3 for key.

Fig. 27. Summary log, Chichester Harbour section, Sussex. See Fig. 3 for key.

230

D.W. JOLLEY

Fig. 28. Summary log, Whitecliff Bay section, Isle of Wight. See Fig. 3 for key.

sediments are overlain by the Walton Member (Division A2), but at Shamblehurst, they are overlain by Division A3 claystones. Many of the records of this interval given by King (1981) and Knox et al. (1983) show Division A3 resting directly on the Harwich Formation (e.g. Alum Bay and Bunker's Hill Borehole) at the western margin of the Hampshire Basin.

Biostratigraphic record The application of biostratigraphy to the lower London Clay Formation has proved in the past to be problematical; the macrofossil studies of Wetherall (1836) and Wrigley (1940) subdivided the London Clay into three zones, which cannot be extended from the central London area, even as far as the Isle of Sheppey (Curry 1965). The first micropalaeontological studies of the lower units of the London Clay Formation include that of Venebles (1962) on the London Clay of Bognor Regis, but the identification of the 'planktonic datum' above the base of Division B by Wright (1972) was the first stratigraphic application. Unfortunately, as demonstrated by King (1981), planktonic and benthonic foraminifera are usually absent from the Harwich Formation, although ostracodes and diatoms are present. The erection of a zonation by King (1983) for the Tertiary of the North Sea Basin enabled Cox et al. (1985) to provide diatom evidence for the age of the Harwich Formation of the Ormesby Borehole, with planktonic foraminifera Zone NSP4 and benthonic zone NSB2 being identified. This zonation has unfortunately not been utilized by workers on North Sea Basin sections (e.g. Mudge & Copestake 1992a, b), who have used zonations derived from petroleum exploration company schemes. Such schemes may provide a framework for the subdivision of the Balder and Sele Formations, but lack of evidence in southeastern England prevents comparison.

Previous palynological studies of the Harwich Formation have been more successful due to the favourable preservation and abundance of this fossil group. Initial studies by Davey et al. (1966) resolved primary problems with the taxonomic nomenclature that were expanded on by Costa & Downie (1976) and Costa et al. (1978), who produced a zonation based on the 'Wetzeliellaceae'. Unfortunately, the Harwich Formation yielded assemblages assignable only to the Apectodinium hyperacanthum zone, which was also recorded in the underlying Lambeth Group sediments. Knox & Harland (1979) were the first to report on the dinoflagellate cysts of the Harwich Formation (above the Harwich Stone Band), identifying an acme of Deflandrea oebisfeldensis which was utilized to create an 'informal subzone' of the A. hyperacanthum zone of Costa & Downie (1976). However, it was the zonation of Costa & Downie (1976) that was used by Knox et al. (1983) in their study of the basement-bed. These authors reported that the dinoflagellate cysts assemblages of the Harwich Formation at Tilehurst were attributable to the Wetzeliella astra zone, while those in the western Hampshire Basin were attributable to the Wetzeliella meckelfeldensis zone, indicating considerable diachronism. In the study of the dinoflagellate cyst assemblages of the Ormesby A Borehole by Cox et al. (1985), it proved possible to identify only the Apectodinium hyperacanthum zone within the Harwich Formation. The Deftandrea oebisfeldensis subzone was not recognized, due to low frequencies of the taxon in the samples studied, regarded by the authors as possible evidence for biofacies control. More recently, Jolley & Spinner (1989) confirmed the extent of the A. hyperacanthum zone, the D. oebisfeldensis subzone, and the W. astra and W. meckelfeldensis zones in the Harwich Formation and Walton Member of the Ipswich area. They demonstrated that the D. oebisfeldensis acme was restricted to the sediments above the Harwich Stone Band and did not occur in the underlying more arenaceous units. A subsequent pollen and spore study (Jolley & Spinner 1991) recognized five associations within the Harwich Formation and Walton Member in the Ipswich area. This was primarily an environmental study, although some stratigraphic correlations were considered. Dinoflagellate cyst zonations similar to that of Costa & Downie (1976) have been used in the North Sea Basin (e.g. Knox et al. 1981), but they have been superseded by schemes based on both first downhole occurrences (FDO) and abundances of dinoflagellate cysts. The most recent of these are the zonations presented in Stewart (1987) and a similar scheme was proposed by Powell (1988).

231

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND These schemes utilize the (FDO) occurrence of Cerodinium wardenense, Cerodinium dartmoorium and Apectodinium augustum in the Balder and Sele Formation interval. Unfortunately, the comparison of zonations provided by Powell (1988) indicates that this scheme provides insufficient resolution within the southeastern England palynofloras for a correlation more detailed than that given by Costa et al. (1978). Lack of detail in the biostratigraphic database has been the primary cause of uncertainty over correlation of sections exposing the Harwich Formation (Fig. 29). Consequently, attempts at comparison have been based largely on lithologies.

Palynology methods Samples prepared for this study were subjected to treatment with hydrofluoric acid to remove silicates

Dinoflagellate Cyst Zonations

and a dilute nitric acid wash to remove pyrite where present. No alkali was used during processing, but in some rare cases a 2% solution of sodium hyperchlorite in water was used. The resultant residues, sieved at 10 pm were strew mounted in petropoxy or glycerine jelly and studied using transmitted light. Analysis of the palynological strew mounts involved recording occurrences of all microplankton, pollen and spores in optimum counts of 250 microplankton plus associated numbers of pollen and spores. If the terrestrial component was dominant, then 300 pollen and spores were counted, together with associated microplankton. Taxa not recorded during these counts, but present in a subsequent scan of other material, were also noted as records. From this a database was prepared using the Tilia data-plotting programme; data was produced showing the quantitative variations in taxa for all sections

Localities and Lithological Units >, C~

Q


Costa & Downie 1976 Costa et al 1978

Powell 1992 o

W. meckelfeldensis

Wme

W. astra

Was

_J

ow

~-m

F--~

O~

I LU Z UJ 0 O W

~

E

E

(b t~

e~

Gor i

s

e i

i

E

I i

8

~,,

:~

=1

6

Aau

Fig. 29. Chart showing the application of the published dinoflagellate cyst based zonations of Costa & Downie (1976), Costa et al. (1978) and Powell (1992) to the localities studied on the left of the diagram.

232

D.W. JOLLEY

analysed in raw numbers and normalized formats. This data is presented here in selective normalized frequency plot charts (Appendix B), except where the total number of specimens counted fell below 100 per sample in impoverished material, where raw data plots are substituted. Interpretation of the data followed the method outlined by Jolley (1992), involving the recognition of palynofloral associations in each of the studied sections. Associations were defined by empirical inspection of the data, with the aid of cluster analysis and detrended correspondence analysis, enabling the identification of groups of successive assemblages which showed shared characteristics of composition. In some cases, lower density sampling resulted in assemblages from single samples being referred to as associations, where the preceding and succeeding assemblages showed a significant difference in composition. As in the previous study (Jolley 1992), associations were defined individually for each section without reference to other sections. Each association was given a name using the specific epithets of two of its dominant, or characteristic, taxa. Where a characteristic could not be identified to specific level, e.g. Leiosphaeridia spp., then the generic name was substituted.

Application of established biozonations The correlation of Thanetian sediments suggested by Jolley (1992) utilized published dinoflagellate cyst-based biozonations as a method of calibrating the study to previously published work, and to provide an age framework for the association-based stratigraphy. This methodology is followed here, where the zonations of Costa & Downie (1976), emend. Costa et al. (1978) and Powell (1992), are used. The stratigraphic extent of these zones in the palynofloras recovered during this study is controlled by the occurrence of Apectodinium spp., and by the acme occurrences of Deflandrea oebisfeldensis recorded in the Harwich Formation sections (Appendix B). Application of these zonations does little to assist in our understanding of the detailed depositional geometries of the Harwich Formation, because of their broad resolution (Fig. 29). Application of the association sequence technique has been adopted to give additional information on these depositional relationships.

sections have been compared with those in neighbouring sections and the resulting correlation used to establish a series of association sequences. For example, the associations defined in the Wrabness Member at Wrabness were compared with those in Bradwell 205 (the nearest complete section to the south), to Sizewell C3 and Ormesby A (to the north). From these comparisons, laterally continuous, stratigraphically predictive sequences of associations were identified. These were termed association sequences (Jolley 1992), and are regarded as containing palynofloras deposited in sediments of equivalent age and as such should closely follow depositional units. In this study, nine association sequences are recognized. Some of these showed clear subdivisions on empirical and multivariate analysis; these lower level subdivisions are termed 'association subsequences'. The opportunity has been taken to use association subsequences to describe correlative associations which show a degree of similarity with their neighbours. These have been used to subdivide association sequences where appropriate. In all, nine association sequences have been identified incorporating ten association subsequences. Neighbouring sections generally contain palynofloras deposited in similar environments, whose composition may change in a separate part of the basin. This nearest-neighbour correlation minimizes biofacies effects on the data, but despite this checks must be provided against which the stratigraphic data can be compared. In an earlier study (Jolley 1992), magnetostratigraphy provided a valuable comparitor, but no significant magnetostratigraphic events are present in the study section. However, this association sequence stratigraphy can be compared to tephrostratigraphy and to the third-order depositional units. These depositional units are presented on the correlation diagrams referred to below, seven parasequences (Van Waggoner et al. 1988) being identified within the Harwich Formation on the basis of lithological and electrical log evidence. The correlation of the association sequences and comparison with other data are discussed in order below, (oldest to youngest) for the area south of the mid North Sea High. The precise definition of the association sequences and subsequences (Tables 1-4) is presented in the frequency plot diagrams of Appendix B, and the correlation discussed below is shown in Fig. 30a & b.

Association sequences

Association sequences Y1 and Y2

A detailed stratigraphy for the Harwich Formation sediments is proposed here based on palynoforal associations. Associations defined in individual

These association sequences are encountered in sections north of the Mid North Sea High and are considered later.

i~ ~ ~

~j

q~

d~ Z

~li ~

~J

% ~J

J

234

D . W . JOLLEY

eZ

e'., q..,

~._= Z

%

e4

e;

235

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

I

~5 I

e;

~4

r

e; <~

236

D.W. JOLLEY

Association sequence Y3 This is composed of three association subsequences.

Association

subsequence Y3a. Comparisons between the lmpletosphaeridium ligospinosum dominated associations of 81/46a and Ormesby A and the Fromea fragilis-Pterospermella spp. association at the base of the Hales Clay in the short core of 79/07a is not direct, but the increased frequencies of F.fragilis seen in Y3a of Ormesby A strongly support this correlation. The position of 79/07a at the southern end of the Sole Pit structure probably accounts for the increased dominance of environmentally stressed microplankton taxa by the effects of turbid shallow water. Further south in the Hales Clay of Hales, palynofloras of Y3a also demonstrate a greater degree of proximality in comparison to those seen in Ormesby A. This represents the southernmost record of association subsequence Y3a in the study area, this change in association composition reflecting the position of this site at the edge of the depositional margin. Association subsequence Y3b. In both 81/46a and Ormesby A, Cometodinium comatum, Spiniferites ramosus subsp, ramosus, S. multibrevis and Impletosphaeridium ligospinosum are significant conaponents of the associations. This dominance of Spiniferites is less clearly demonstrated in 79/07a, where again a shift in composition to an environmentally stressed Subtilisphaera sp.A dominated association is recorded, a further reflection of the 'shoal' effect of the Sole Pit structure. The corresponding association in both Hales and that in the Orwell Member unit A at Sizewell are similar to those in Ormesby A, with Impletosphaeridium ligospinosum and Spiniferites ramosus subsp, rarnosus being dominant. In the Ipswich area, Paralecaniella indentata, Subtilisphaera sp.A and S. ramosus subsp, ramosus dominate associations in Orwell Member unit A at Ferry Cliff, Nacton and Shotley Gate, the last of these containing increased frequencies of Plicapollis pseudoexcelsus. This change in association composition is related to increasing

proximity to the palaeoshoreline. Further to the south, in Bradwell Borehole 205, the southernmost record of this association subsequence shows further evidence for increasing proximality, with Leiosphaeridia spp. replacing P. indentata as the dominant taxon.

Association subsequence Y3c. In 81/46a, Ormesby A, Hales and 79/07a, Apectodinium homomorphum, A. parvum, Spiniferites multibrevis and lmpletosphaeridium ligospinosum are the dominant taxa seen in associations referred to Y3c. These associations contain the greatest frequency and diversity of Apectodinium species within the Sele Formation unit $3. In both 81/46a and Hales this association subsequence is coincident with a high gamma-ray log reading, marking an interval of condensed sedimentation at the top of parasequence 1. Further to the south, in Sizewell C3, the frequency of Apectodinium is lower, and Cometodinium comatum together with Spiniferites ramosus subsp, ramosus dominate the association referred to Y3c. The case is clearer in the Ipswich area, where Apectodinium homomorphum and A. quinquelatum occur in significant frequencies with lmpletosphaeridium ligospinosum, at Ferry Cliff, Nacton and Shotley Gate. An association of similar composition is also recorded at Bradwell 205, which currently represents the most southerly and westerly record of this association subsequence.

Association sequence Y4 This is composed of two association subsequences.

Association subsequence Y4a. An Impletosp.2 sphaeridium ligospinosum-Alisocysta Heilmann-Clausen (1985) association in 81/46a is referred to this association subsequence, although the equivalent association in Ormesby A is missing due to sample spacing and core loss. The palynofloras of 79/07a show significant frequencies of both Subtffisphaera sp.A and Fromea fragilis, reflecting the environmental stress exerted by the Sole Pit structure, and they are referred to Y4a by reason of the position directly above the distinctive

Fig. 30. (a) Diagram showing the correlation of sections from 81/46a in the north to Whitecliff Bay in the south. Parasequences 1-3 are separated from parasequences 4-7 because of the evident disconformity present in Ormesby A and 81/46a at the base of parasequence 4. Heavy stipple indicates sands and sandy siltstones, medium stipple indicates siltstones and claystones. Parasequence boundaries are marked by solid thick lines, association sequence boundaries by solid thin lines. The lithological and stratigraphic information contained in the sections is as in Figs 3-28, spacing between sections is proportional to distance. (b) Correlation diagram showing the relationships of the sections in London and on the north Kent coast. The ornament is as for Fig. 30a, the vertical separation between parasequence 3 and parasequence 6 is as in Fig. 30a to allow comparison.

o

i~iiiiiiiii~i~i~i~ili~i~iii~iiiiiiiii~i~i~!~iiiiiili!i!ii~iiiiiiiiiiii!ii!i ........................................................ :.:.::,:.::.:.:. :+:<:.+:.:.:+:.:+:+:<<+::+:,:.:.:.: ................... ii~ililililililili!i:iliE~i~i~iliEiiiililili!i~ilili E!ilililili!ili~

...................................................... :+:,:.::+::

I

................................................. ....,.............. ......................................................... :+:+:+::.:~

i~i~i~i~i~iiiWi~iii~iii~iiiiiiiiiiiiiiiiiii!ili~i~iiiiiiili~iiiii ~iiiiiiiiiii~Oiiiiliiiiiii!i!i~i!ili~i!i~iii!iii!ili!iliiiiiii~ii ii!ii!iliWiiiiiiiiiiiiiiiiiiiiiiiiiil]i!i!iiiiii!!!!i! i iiiiii ~iiiiiiiiiiiiiiiiii!iiiiiii ~iiiiiiii(Jiiiiiiiiiiii[iiiii!i)iiiiiiiiii iiiiii U.I iiiiiiiiiiiiiili iiiiiiiir.~iiiiiiiiiiiiiii!iiiiili!iiiiiiii iiiiiiiiiiii,~ iiiiiiiiiiiili!iiiii!iiiii!ili!i i!iiiiiiiiiin"}iii[iiiiiiiiiiiiiiiiiiiiiiiiiii

;ii!i!iiiiiii~i!i!i!i!iii!i!!iii!!ii!!iiiiiiiii!i!iiiii!iii!iiiii

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

!iiiiiiiiiiiii!ii:iiiiiiiii!iiii!iiiiiii;iiii~iii~i~i~iiiiiiiiiiii?iiiiiiiii

C5

;~i~iiil}i!ii!i!;iii!i:ii~i!i!i:i~ii!ii!i~i!~ililiiii!ili!ilili !ii!i;!!ii!!ii!:i ~i?!!i!!!ii !i :!:i:i!!~i!i !i!:i!i!i!!!l)i !;i !;!i i i!il ................................................... ::::::.:::

I

r-

!iiiiii:iiiiiiiiiliilii!!:ii!;i!!;i!ii!!ii!!ii!!ii!!i~i!!!ii~?ii:~i~i?i!i~i!:ii!~i~!i~i:i!i!iii~ iiii!ili!iii!iiiii!iiiiiiii!iiiiiiiiiiii;i}ii!iiiiiii!i?}

!izi!ili:ilili!i~i~ili!~ilili!ili!i!ilililil

0

"o rn

.>

OC: 7O

o

0

0

4.-

..c

E E c-

q~ eli "1-

E I1.

c~ c0~

cl

.c_ c-

i _

E "o 0 I11

J~ 0

o c

o rn

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND Y3c palynofloras. This association also compares well with the Spiniferites ramosus subsp, ramosusF. fragilis association recorded in Hales. Further to the south, in Sizewell C3, Y4a shows an increased influence of proximality on the associations, with significant occurrences of Paralecaniella indentata and Impletosphaeridium ligospinosum. A similar palynoflora is recorded at Ferry Cliff, where Spiniferites multibrevis and I. ligospinosum dominate an association which contains increased frequencies of F. fragilis. At Nacton and in spot field samples from Levington, significant frequencies of Polysphaeridium subtile occur with dominant Paralecaniella indentata and Impletosphaeridium ligospinosum, a similar association also being recorded in the Shotley Gate Borehole. At the southern- and westernmost limits of Y4a, both P. indentata and Leiosphaeridia spp. dominate the association in Bradwell 205, reflecting the increase in proximal influence at the depositional margin.

Association subsequence Y4b. Species of Apectodinum and Spiniferites dominate the associations referred to Y4b in both 81/46a and Ormesby A. This is poorly defined in 79/07a, which shows a high dominance of Inaperturopollenites hiatus with records of Apectodinium homomorphum, but the case is clearer in Hales, which has an association rich in both Fromeafragilis and A. homomorphum. Further south in Sizewell C3, an association comparable to those in Norfolk, dominated by Spiniferites and Apectodinium is referred to Y4b. At Ferry Cliff, these taxa are supplemented by increased frequencies of Paralecaniella indentata, this being closely similar to the Y4b association at Nacton which is also rich in Impletosphaeridium ligospinosum and Glaphyrocysta ordinata. A similar P. indentata-G, ordinata dominated association is recorded at Shotley Gate, although the influx of Apectodinium spp. is suppressed and restricted to the lower part of the interval. Frequencies of Apectodinium spp. are higher in Bradwell 205, here the association also shows significant frequency increases in Spiniferites species and P indentata. This is the oldest association subsequence recorded further to the west and south than north Essex. It is discussed below in two parts; localities on the north Kent coast and in the London area, and those in Essex, north of London and the western London and Hampshire Basins. At Stanford Rivers in Essex a Paralecaniella indentata and Deflandra oebisfeldensis characterized association containing Apectodinium homomorphum is referred to Y4b, but further to the south in the Greater London area, Paralecaniella indentata dominates all the associations in the

237

Oldhaven Member of the Chingford South, Jubilee 810 and A4A Boreholes, making differentiation between Y4b and Y5 difficult. At Upnor, a clear Y4b is recognized by the presence of two associations, but in Sheppey Borehole 1 a monospecific P. indentata association is recorded at this level. At Herne Bay, the majority of the Oldhaven Member section yields monospecific P. indentata associations, but at the base of the section the more argillaceous sediments yield a diverse palynoflora containing significant frequencies of Areoligera cf. senonensis and Deflandrea oebisfeldensis, referable to Y4b. A similar case is seen at Shelford Pit where the basal part of the section yields a Leiosphaeridia spp.-Eocladopyxis peniculata association. In the more silty sands, variously assigned to the Swanscombe and Tilehurst Members, north and west of London, Y4b is recognized in the Dowsett's Farm Borehole as an association containing significant frequencies of both Impletosphaeridium ligospinosum and P. indentata together with Fromea fragilis and Apectodinium homomorphum. This composition is closely similar to those recorded at South Mimms and Fulmer in Buckinghamshire and at Knowl Hill in Berkshire, which contain significant frequencies of Apectodinium spp., although the increased frequencies of Leiosphaerida spp. in the latter appears to show increased proximal influence. A more diverse association recorded at the base of the Harwich Formation at Tilehurst contains significant frequencies of Intratriporopollenites microreticulatus, a composition also seen in the Harwich Formation of the Shamblehurst Borehole. The Fromea fragilis-Lentinia? wetzelii association recovered at the base of the Harwich Formation in Chichester Harbour is similar to other associations referred to Y4b at Fulmer, Knowl Hill and Tilehurst in the frequency of F. fragilis and in consistently low frequencies of Apectodinium spp. At Whitecliff Bay a Paralecaniella indentata-'Platycarya' platycaryoides association is recorded and referred to Y4b, by virtue of the significant frequencies of Lingulodinium machaerophorum in the microplankton.

Association sequence Y5 This association sequence shows a distribution pattern similar to that of subsequence Y4b. In 81/46a, Ormesby A and Hales, Glaphyrocysta ordinata, Spiniferites multibrevis and Subtilisphaera sp.A occur commonly in the associations assigned to Y5. Sizewell C3 shows a shift in association composition to the south, with an influx of Paralecaniella indentata reflecting increasing proximality. At Ferry Cliff, Shotley

238

D.W. JOLLEY

Gate Borehole and Wrabness, Glaphyrocysta ordinata, Cometodinium comatum, Deflandrea oebisfeldensis, Impletosphaeridium ligospinosum, Caryapollenites circulus and Intratriporopollenites microreticulatus are recorded in significant frequencies in Y5. Similar dominance of /. ligospinosum and P. indentata is also recorded further south at Bradwell 205, while in the Stanford Rivers Borehole a single sample contains a P. indentata-l, ligospinosum association similar to that of Bradwell 205, the greater dominance of P. indentata reflecting increasing proximity. In the Greater London area, the Chingford South, Jubilee 810 and A4A sections all contain P. indentata dominated undifferentiated Y4bY5 associations, but at Upnor, a P. indentataGlaphyrocysta ordinata association and an overlying monospecific P. indentata association are both thought to be referable to Y5. This is also the case at Sheppey 1, where the greater diversity of the P. indentata-S, ramosus subsp, ramosus association suggests an affinity with Y5. A similar diverse assemblage can be found in a silt lens in the middle of the Oldhaven Beds at Herne Bay. The remainder of the section at this locality is composed of a monospecific P. indentata association, a similar pattern of occurrences also being seen at Shelford Pit. In the Dowsett's Farm Borehole, an impoverished P. indentata dominated association contains significant frequencies of Intratriporopollenites microreticulatus and Caryapollenites circulus. However, further west at South Mimms and Tilehurst, numbers of Apectodinium spp. rise in conjunction with significant frequencies of C. circulus and L microreticulatus. The section at Knowl Hill is impoverished at this level, but Apectodinium spp. are again recorded in significant frequencies at Chichester Harbour. Less clearly attributable to Y5 is the Leiosphaeridia spp. and Spiniferites ramosus subsp, ramosus dominated association seen in Shamblehurst, but its position above Y4b makes it appear a likely correlative event. This association in Shamblehurst has high frequencies of Inaperturopollenites hiatus, a feature seen in the P. indentata-l, hiatus association present in the Harwich Formation at Whitecliff Bay. This is apparently a local feature related to increasing proximality in sections in the southwest of the Hampshire Basin.

Association sequence Y6 This comprises two association subsequences.

Association subsequence Y6a. A Cerodinium wardenense-lnaperturopollenites hiatus association is recorded at the base of the Balder

Formation in 81/46a and is comparable with a Cometodinium comatum-l, hiatus association at the base of Wrabness Member unit A of the Harwich Formation in Ormesby A. The distribution of C. wardenense appears to reflect biofacies variation, the high frequencies seen in 81/46a apparently not extending south of the mid North Sea High. The association attributed to Y6a at the base of Wrabness Member unit A in Hales is similar in character to that seen in Ormesby A, but the abundance of P. indentata implies increasing proximality to the depositional limit. Indeed, this occurrence in the Hales Borehole is the southernmost record of the association subsequence.

Association subsequence Y6b. Within the associations referred to Y6b from the southern North Sea and Norfolk, Glaphyrocysta ordinata, C. wardenense, Deflandrea oebisfeldensis and Spiniferites ramosus subsp, ramosus are recorded in significant frequencies. Associations of similar composition are also recorded in the Shotley Gate Borehole and at Wrabness, where the common occurrence of P. indentata is indicative of increasing proximity in this the southernmost occurrence of this association subsequence. Association sequence Y7 This is composed of three association subsequences.

Association subsequence Y7a. Cometodinium comatum, Hystrichosphaeridium tubiferum and Caryapollenites circulus are characteristic of this association subsequence in 81/46a and Ormesby A. A similar association with significant frequencies of Paralecaniella indentata and Glaphyrocysta ordinata is seen in the uppermost part of the Hales core, and may be attributable to this association subsequence (the termination of the core leaves this as questionable). However, a similar P. indentata-G, ordinata association is recorded in the Wrabness section and is here assigned to Y7a, although further north at Sizewell the association attributed to Y7a is more closely similar to those seen in north Norfolk. An association similar to this is also seen in Bradwell 205, which represents the southernmost limit of Y7a. Association subsequence Y7b. Cometodinium comatum, Hystrichosphaeridium tubiferum and Impletosphaeridium ligospinosum are the dominant taxa in this association sequence in 81/46a and Ormesby A; but a peak of Membranosphaera maastrichtia is recorded, which is also seen in the Y7b association of Sizewell C3. The most complex record of associations which can be referred to Y7b

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

is in the relatively expanded section at Wrabness, here associations dominated by Spiniferites ramosus subsp, ramosus, S. multibrevis, I.

ligospinosum, M. maastrichtia, Paralecaniella indentata and Fromea fragilis are recorded. The S. multibrevis-M, maastrichtia association recorded here seems to be most similar to that in Ormesby and Sizewell further to the north, while further to the south Y7b is represented by a P. indentata-C, comatum association at Bradwel1205. Here, an increase in the frequency of S. multibrevis at the upper limit of the association may indicate a comparison with the S. multibrevis-M. maastrichtia association recorded in the middle of Y7b at Wrabness. No records of associations referable to Y7b are made south or west of Bradwell, which may account for the increased frequencies of P. indentata in this more proximallyinfluenced section.

Association subsequence Y7c. In 81/46a this is represented by a Glaphyrocysm ordinataMicrodinium cf. ornatum association, while in Ormesby A it is represented as a Spiniferites multibrevis-Apectodinium homomorphum association. Sizewell C3 shows a less distinctive palynoflora with a Deflandrea oebisfeldensis-Spiniferites ramosus subsp, ramosus association referable to Y7c. The section at Wrabness clearly demonstrates two associations which are attributable to Y7c, a Glaphyrocysta divaricata-Tectatodinium sp. WB association and a G. divaricata-Deflandrea oebisfeldensis association. The former of these two associations is closely similar to an association of the same name in another Bradwell borehole, Bradwell 209. Nearby in Bradwell 205, sampling or local variation means that the Tectatodinium sp WB influx is not recorded, instead a P. indentataC. comatum association is present, representing the southernmost record of Y7c in the study area.

Association sequence Y8 This association sequence is not further subdivided, the palynofloras attributed to Y8 being recognized widely across the basin. In 81/46a and the Norfolk area, Deflandrea oebisfeldensis, Spiniferites ramosus subsp, ramosus and Intratriporopollenites microreticulatus form significant parts of the associations attributable to Y8. Similar associations are recovered at Sizewell C3, Walton Naze and Wrabness, rich in Cometodinium comatum,

Deflandrea oebisfeldensis, Glaphyrocysta divaricata and Pterospermella spp. However, the section at Walton Naze contains lower frequencies of D. oebisfeldensis than other nearby sections, having palynofloras of a distal aspect. Further to

239

the south at Bradwell 205, a Paralecaniella indentata-D, oebisfeldensis association is recorded and can be traced into a P. indentata dominated association (L microreticulatus-D, oebisfeldensis) at Stanford Rivers. A similar P. indentata dominated association is seen at Chingford South (Fromeafragilis-I. microreticulatus), but at Jubilee 810, A4A and Upnor no representation of any association sequence younger than Y5 is seen. However, at Sheppey 1, a D. oebisfeldensis-Hystrichospaeridium tubiferum association similar to other associations referred to Y8 from Suffolk and Essex is recorded, resting directly on palynofloras referred to Y5. Sections at Herne Bay and Shelford Pit show a similar succession to those in Greater London, lacking any palynofloras younger than Y5. North of London, the Dowsett's Farm Borehole shows impoverished microplankton assemblages in a Pityosporites spp.-P, indentata association which contains significant frequencies of /.

microreticulatus. A P. indentata-C,

comatum

association attributable to Y8 at South Mimms suggests that numbers of D. oebisfeldensis are suppressed to the west of London. A Lingulodinium machaerophorum-foraminifera test linings association at Fulmer suggests that Y8 becomes increasingly proximal to the west, a trend continued in the P. indentata-Leiosphaeridia spp. association at Knowl Hill and the Leiosphaeridia spp.-Pterospermella spp. association at the top of the Harwich Formation at Tilehurst. In the Hampshire Basin, the Shamblehurst Borehole has a P. indentata-Achomosphaera alcicornu association at the top of the Harwich Member. Overlying siltstones referred to Division Alb, i.e. Wrabness Member equivalent, by King (pets. comm.) contain a palynoflora typical of the lower part of the Walton Member (Division A2). The more easterly location of the Chichester Harbour section results in a higher microplankton diversity, giving a C. comatum-H. tubiferum association in a palynoflora dominated by P. indentata. No representation of Y8 is recorded at Whitecliff Bay.

Association sequence Y9 This is the youngest association sequence included in the study and has a distribution restricted to the east of the area. It has only a thin representation in 81/46a due to local erosion at the top of the Balder Formation, but in Ormesby A the sandy siltstones of unit B of the Wrabness Member and at Wrabness in the fine sands of Wrabness Member unit B, rich associations dominated by Pityosporites spp. occur. At Walton Naze, the greater marine influence gives

a Pityosporites spp.-Deflandrea oebisfeldensis

240

D.W. JOLLEY

association with a higher diversity. The southernmost record of this association sequence is at Bradwell, where the fine sands at the top of the Wrabness Member also contain a P i t y o s p o r i t e s spp. dominated association.

Stratigraphic correlation The proposal of this correlation (Fig. 30a & b) raises several points of interest with regard to the relationships of lithostratigraphic units. It is clear from the correlation that only the Sele Formation unit $3 (Knox & Holloway 1992) is represented in 81/46a, a lithological unit which is directly correlative with the Hales Clay (unit HC2) of the Harwich Formation (Knox et al. 1990; Ellison et al. 1994). Originally proposed by Cox et al. (1985) as a correlative of the Lambeth Group (Ellison et al. 1994), Knox et al. (1990) subsequently proposed that the Hales Clay was an early deposit of the transgression of the London Clay Sea. As supporting evidence they cited the presence of this unit in a confidential borehole 35 km south, which has subsequently been made available for study (Sizewell C3). Lithological comparison of this unit in the Sizewell cores with the Orwell Member exposed and augered in Suffolk and Essex indicates little difference in appearance, accordingly these lower Harwich Formation beds in Sizewell are here referred to the Orwell Member units A--C. In addition, data from electrical logs, particularly the neutron log, and lithological data from the Hales core provide evidence for the upward extension to Hales Clay unit HC 1 from the original definition of Knox et al. (1990). This is substantiated by palynological evidence (see Appendix B), the whole of HC1 containing shallow marine palynofloras comparable to those of the Upnor Formation. The palynological evidence for the correlation of the Hales Clay unit HC2 with the Orwell Member is strong, and is supported by a lateral continuity of these lithological units. The fine sands at the base of Hales Clay unit HC2 contain palynofloras attributed to Y3a and Y3b, while the fine sands of Orwell Member unit A in Sizewell C3 contain palynofloras attributed to Y3b and basal Y3c; a feature also seen in Ferry Cliff, Nacton and Shotley Gate. Further south at Bradwell 205, these sands extend into upper Y3c indicating a slight diachronism from Norfolk to Essex which is supported by the evidence for onlap at this level (Fig. 31). The occurrence of tephra layers in Orwell Member unit C has previously been recorded by Knox & Ellison (1979) in the Shotley Gate Borehole, such layers also being visible in the upper part of the Orwell Member at Nacton, Levington, Ferry Cliff and Sizewell C3. However,

the degree of preservation of the tephras is variable, most are represented by discontinuous layers. In south Suffolk these tephra layers are mostly recorded in sediments of Y4b-Y5 age, a feature which is thought to relate to the change from sandy siltstones to clayey siltstones evident across the area. However, the Hales Clay has been reported to contain only a few discontinuous tephra layers in the Ormesby A Borehole (R. W. O'B. Knox, pers comm.), although this record is incomplete due to significant core loss at the top of the unit. However, the Hales Clay at Hales also contains only a few argillized tephra layers, which, in view of the argillaceous lithology and the strength of the palynological correlation, makes the absence of well- defined tephras at this level an enigma. Correlation of the Balder Formation with the Wrabness Member commented on by Knox & Harland (1979) is supported by the palynological correlation presented here (Fig. 30a & b). Comparison of tephra layer distribution within the sections studied is now possible on a refined scale using the association sequence stratigraphy as a framework. Major concentrations of tephra layers within the Balder Formation of 81/46a and the Wrabness Member of Ormesby A occur in the interval between the top of Y6a to below Y8. The most persistent laterally traceable group of tephra layers occurs in Y7c in Bradwell 205 and Wrabness, but is less prominent in the slightly condensed Sizewell C3 section. A lower Y7a to basal Y7b group of ashes is also traceable from Bradwell 205 through Wrabness, but are again lacking in Sizewell C3. Sediments potentially containing tephra layers -17 and + 19 are present as far south as Essex, although the lithologies containing Y3c and Y6b may preclude their preservation. In the more arenaceous lithofacies to the south and west of East Anglia discrete tephra layers are not preserved, but Knox (1983) has reported the presence of disseminated volcanic ash in the Harwich Formation at Aveley (TQ 576802) and Hadleigh (TQ 8002 8654) and in the Oldhaven Beds at Aveley and Herne Bay. In the light of the palynological evidence it can now be demonstrated that the disseminated ashes of the Oldhaven Beds are correlative with the tephra layers in Harwich Formation sediments of Y4b-Y5 age in East Anglia (Figs 30a & b & 31).

Evidence for relative sea-level change Analysis of the palynological data by simple statistical means has been used before to assist in the interpretation of palaeoenvironments (e.g. Goodman 1979). Here, the diversity of the pollen, spore and microplankton floras has been used to identify increasing proximality as an aid to the

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND

241

Fig. 31. Diagrammatic representation of the onlap model for the Harwich Formation incorporating evidence for relative sea-level trends from the statistical data.

recognition of relative sea level changes within association sequence intervals. It has previously been demonstrated that both the diversity and numbers of pollen and spores increase with proximity to a land mass (Muller 1959). Examination of diversity plots should enable the identification of periods of relative sea-level rise when terrestrial influence would be declining. The same principle can be used with regard to microplankton and has been used before with particular emphasis on dinoflagellate cysts (Goodman 1979). The work of Harland (1983) has established the outline of dinoflagellate cyst distribution patterns in recent marine sediments. Dinoflagellate diversity appears to reach a maximum at around the outer-neritic-shelf-slope break area, declining both inshore and offshore of this location. In particular, low diversity assemblages may be found in the inner-neritic zone. Using this information, it is possible to relate changes in microplankton diversity in the fossil record as evidence for relative sea-level change. In addition to this diversity data, plots of the variance of the microplankton taxa have been undertaken. Instead of using the number of taxa present in each sample, i.e. the diversity, the variance of each assemblage takes into account the number of taxa present and the frequency in which they occur, giving a representative picture of the

dominance of the sample by any particular species. This approach is adopted in preference to Goodman's (1979) dominance index which uses the sum of the two most abundant species divided by the diversity of the sample, an approach not suited to this study where many assemblages contain more than two taxa of equal dominance. The selected results of the statistical tests are shown in Figs 32 & 33, and the conclusions summarized in the figure captions. Only the more argillaceous sections have been treated in this manner; they are the most complete and have minimal potential control of lithofacies on palynomorph assemblages. Of particular interest in this data is the lack of direct correspondence between diversity and the presence of flooding surfaces which is apparent in Fig. 32. Although such intervals of maximum relative sea level in Y3c and Y4a do show high diversities they are not exceptional, the variance most clearly demonstrating the relationships between dominance within microplankton populations and flooding surfaces in the studied sections. Depositional history In constructing a depositional history for the Harwich Formation and associated sediments, information was drawn from several sources. Primarily, the association sequence stratigraphic

242

D.W. JOLLEY

framework enabled an onlap curve to be established, based on the current geographic extent of the rock units constrained by the model. Refinements to this basic approach were added to establish relative sea level trends within each depositional unit and to identify intervals of contemporaneous erosion or condensed sedimentation. This was achieved by incorporating the microplankton and pollen and spore statistics together with basic observations on the microplankton biofacies composition, such as those of Islam (1984). Deposition of sediments containing association sequences Y1 and Y2 was apparently restricted to areas in the centre of the basin, presumably deposited during a lowstand period of lower relative sea levels. Using the chronology proposed by Beggren (1995), from evidence in Goban Spur Hole 550, sedimentation on the shelf in Norfolk was initiated c. 54.5 Ma by a rise in sea level which

caused the deposition of Y3a-Y3b age shelf sheet sands of the basal Harwich Formation (Figs 30a & b & 31). The lower boundary of Y3a is regarded as the transgressive surface at the base of a transgressive sequence, this sequence occupies the time interval from 54.5 to 54 Ma, giving an estimated association sequence/subsequence duration of 38 500 years. This conforms with a third order sequence duration of c. 50 000 years given by Van Waggoner et al. (1988), although the time period represented by the studied interval and the number of parasequences present in the section suggests a duration of c. 40 000- 71 000 years for each third order sequence. Relative sea level continued to rise during deposition of parasequence 1 (Y3b-Y3c times), shifting the deposition of inner neritic sands further south to Suffolk (Orwell Member unit A, the Suffolk Pebble Bed) and eventually into eastern Essex. Sand deposition gave way to sandy silts as

Fig. 32. Graph showing diversity and variance trends in the Hales Borehole; note the overall decrease in variance upsection indicating rising relative sea level in the Y3-Y5 interval. In addition, it is important to note that the maximum diversity peaks do not correspond to the high gamma-ray record claystones on the electric log which instead correspond closely to the intervals of highest variance.

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND

243

Fig. 33. Graph showing diversity and variance trends in the Sizewell C3 Borehole; note the overall decreasing variance in Y3-Y5, and the slight decrease in variance from Y7 to Y8.

relative sea levels rose throughout Y3c times. During the Y3b-Y3c period, mudstone deposition was initiated in Norfolk (Hales Clay), with a distinctive claystone at the top of the lowermost unit, which has a high gamma-ray log response (see Hales section, Fig. 9). This fining-upward sequence is represented in the Ipswich area by a decrease in the amount of sand followed by a laminated claystone/siltstone bed (see Nacton section, Fig. 5) at the base of Orwell Member unit B. In other localities this condensed section appears to be missing through contemporaneous erosion or non-deposition (see abrupt upper boundary of Orwell Member unit A on Sizewell C3 log, Fig. 7). Tephra layer -17, an important correlation datum between Hole 550 and 81/46a (see Figs 34 & 35; Knox 1984, 1985) occurs in the upper part of this unit, suggesting that this horizon is comparable with immediately above the base of nannofossil Zone NP10 which is used to approximate to the Paleocene-Eocene boundary.

Higher relative sea levels over the study area during deposition of sediments with Y4a palynofloras ensured continuance of claystone sedimentation in Norfolk and sandy silts in southern East Anglia. A second high gamma-ray response claystone bed in Hales corresponds to association sequence Y4b which occurs at the top of Orwell Member unit B and in the base of the finer-grained tuffaceous claystones of Orwell Member unit C in Suffolk and Essex. This is interpreted here as a second fining upwards sequence deposited in response to rising relative sea levels. This transgressive phase reached its maximum in latest Y4b, times initiating arenaceous sedimentation in the western part of the London and Hampshire Basins which had not experienced deposition since cessation of Reading Formation sedimentation - a hiatus from 55.2 to 54.5 Ma. In the Ipswich area this maximum flooding phase is represented by the shift to finer grained sediments at the base of unit C.

244

D . w . JOLLEY

Deposition of glauconitic siltstones and glauconitic shelly sands initiated during the Y4b transgression across the enclosed basin continued into Y5 times as the sea levels rose to their maximum, sedimentation probably being terminated by the late Y5 sea-level fall. In north Kent and the Greater London area deposition of inner-neritic cross-bedded sands and pebble beds (Oldhaven and Blackheath Beds), initiated by the Y4b transgressive maximum, occurred against the flank of the Weald-Artois axis and on to the London Platform. Sediments containing association sequences Y3a-Y5 are attributable to three retrograding stacked parasequences (Van Waggoner et al. 1988). They all represent periods of relative rising sea level terminated by condensed sedimentation at a transgressive maximum and are seen in the studied sections as south (and westward?) onlapping, fining-upward sequences from shelf sheet sands to silty sands, to sandy silts and claystones. The three parasequences identified do not correspond directly to units A-C of the Orwell Member identified in Suffolk. These lithological units have boundaries defined by changes upsection to finer sedimentation. From the palynological correlation it appears that the transgressive maximum at the end

o

of the each parasequence caused a shift to finergrained sedimentation which is used to define the base of the succeeding lithological unit (Fig. 31). A widespread disconformity is apparent between the top of the sedimentary sequence that contains Y5 assemblages, and the base of the overlying Wrabness Member, containing Y6a assemblages. It is apparent in the cores of 81/46a, Ormesby A and Hales, but it is not suggested that at any time these areas were subaerally exposed. This may have been the case in north Kent, the London Platform, the west London and Hampshire Basins, but the lack of evidence for terrestrial weathering of the Y4b-Y5 age units in these areas suggests continuous submergence during this period. Accordingly, it is possible that the whole of the western London-Hampshire Basin area remained a shallow inner-neritic embayment at times of low sea level. Using the duration of the association sequence/subsequence units it is possible to estimate the Balder/Sele and Wrabness/Hales (only in Norfolk) boundaries as being at 54.27 Ma, an age similar to the estimation of the P5-P6 planktonic foraminifera zone boundary (Fig. 34) given by Beggren & Aubry (1995). Deposition resumed in the north Norfolk area with the onset of tuffaceous siltstones of Wrabness

~ o

_

,-

~

~.~. -~g"~ EasternEngland Onlap , ~

~g

Y10

54 0 -

/l/l~', Y9

f

(5

.

7

f

Y8

6

__ b

Y7

9

T50

8

T45

5

a b -+19

~

,d.

Y6

iii o

o,,

a

~

f

Y5 Y4

c

z

(5

3

a b

Y3

b

- -17

a

i4.5 -

Y2

Fig. 34. Table showing the comparison of magnetostratigraphy, foraminiferal, nannofossil and dinoflagellate zones with the palynofloral association sequences and the southeastern England onlap model. The columns on the right of the diagram show the proposed comparison to the depositional sequence units of Stewart (1987) and Jones & Milton (1994) for the Outer Moray Firth.

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND

Fig. 35. Summary log, BGS Borehole 81/46a, Mid North Sea High; after Lott et al. (1983).

245

246

D.W. JOLLEY

Member unit A containing Y6a palynofloras. Sediments laid down at the depositional margin during this period have not been encountered, but unless removed by erosion should occur between Hales and Sizewell, where Y6a is absent. A rise in relative sea level initiated tuffaceous siltstone deposition at Sizewell with Y6b palynofloras, indicating coastal onlap of retrograding depositional units of parasequence 4, at least as far south as Wrabness. It is this tuffaceous siltstone unit that contains the second stratigraphically important tephra bed, +19. Subsequently, relative sea levels rose significantly, initiating Wrabness Member unit A sedimentation at Bradwell where sandy silts contain a Y7a palynoflora. During deposition of Y7b-Y7c palynofloras, tuffaceous siltstones continued to be laid down in the East Anglia area. Palynological statistical evidence (see above) suggests that relative sea levels fell to the end of Y7a times with the establishment of prograding sedimentation (Figs 32 & 33). However, this appears to have been short lived, as the palynofloras of Y7b suggest resumption of the retrograding, onlapping sedimentation regime

Fig. 36. Summary log, Shell/Esso Well 14/25-1.

throughout the remainder of parasequence 5. The record at Wrabness identifies three separate transgressive phases during deposition of Y7b, some of which are not represented in other sections. This can be attributed to the presence of additional accommodation space on the fault-controlled Ipswich-Felixstowe feature (Jolley 1992), while other areas may have undergone sediment starvation. This phase of retrograding sedimentation continued into Y7c; the marginal sediments of this Y7 interval must occur abutted against the apparent break between the extension of the southern North Sea Basin and the eastern margin of the enclosed west London-Hampshire Basin area. Maximum relative sea levels appear to have been reached during deposition of parasequence 6 (Y8 times) with major onlap and resumption of sedimentation in the area west of Stanford Rivers (excluding the London Platform). In this area, glauconitic siltstones and sandy siltstones were deposited as far west as Knowl Hill, where they pass laterally into shelly sands. Sedimentation also resumed in the north Kent coast area, with sandy

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND siltstone deposition at Sheppey, although it appears probable that much of this sediment was removed by later erosion. In the Norfolk area, parasequence 6 shows a marked progradational profile in contrast to the more southerly sections. As this is at variance with the other data, it is suggested that a local increase in accommodation space or sediment source may have been responsible for this feature. A relative fall in sea level during deposition of parasequence 7 in the region caused the cessation of sedimentation over a wide area south and west of Bradwell. Progradational shelf sand deposits are recorded containing Y9 palynofloras at Bradwell, Walton Naze and Wrabness, while in Norfolk the upper part of the Wrabness Member becomes increasingly sandy. Immediately following this period of prograding sedimentation a major hiatus was entered, until deposition of the Walton Member was initiated, possibly as a highstand deposit.

During this break it is again unknown whether the western London-Hampshire Basin was subarealy exposed, or whether a shallow coastal platform existed whose deposits (if any) where subsequently removed by the following transgression.

Comparison with the North Sea Basin Data from Shell/Esso 14/25-1 (58~ 00~ ' W) and BNOC 15/28a-3 (58~ 00~ in the Outer Moray Firth (Figs 36 & 37) supplement the evidence derived from BGS Borehole 81/46a (54~ 00032.275 ' E) which cored the Tertiary sediments on top of the mid North Sea High (Lott et al. 1983; Fig. 35). The lithostratigraphic nomenclature for the Tertiary sediments of the North Sea Basin proposed by Knox & Holloway (1992) is used here.

,~Lithostratigraphy

247

Gamma

_- ~c

-

O_

O O ~ --g t~

5~ =

Sonic

O r-

a

15O

C

>, O

O"

.J

140

40


zz

Y9

i-

Y8

Horda Formation -

3451 Py 6100-

v v

B2

Py v

t.flO

v

6200 - Py

iio "13 m

Y7b-c

Py v

B1

6300 - v Py v v

Y6a-b

Py -6400-

v

~" .,=

~-

o

o

u_

Y5-4

S3

PYPY i Y3a-c 6500 -

S2b

m

py <=='-

.3

(D

(/3

S2a

4)

-

Py

Y1-2

J~._ -

Fig. 37. Summary log, BNOC 15/28a-3.

Tx

-

248

D.W. JOLLEY

A direct application of the association sequence framework established in the eastern England area was possible with the palynofloras recorded from the ditch cuttings and limited sidewall core samples available from the two commercial wells. Application of the scheme is primarily restricted by sample quality and cavings which have the effect of blurring the boundaries between associations. Together with the wide sample spacing (30 ft cuttings), this leads to a lower resolution than is possible in cored sections. However, the comparability of the floras from the dominantly terrestrial sediments of 14/25-1, where the subdivision depends primarily on pollen associations to those floras recovered from the marine sections of 15/28a-3 and 81/46a, allows identification of the association sequence framework in a range of facies. The results of the application of the association sequence framework to the North Sea Basin sections is illustrated in the individual well summary logs (Figs 35-37) and comparison

between the lithostratigraphic units of eastern England and the North Sea is presented in Fig. 38. Comparison of the association sequence subdivision established for eastern England with the North Sea sequences shows that the oldest association sequence occurring in Norfolk (Y3a) occurs in upper units S2b and $3 of the Sele Formation. Below Y3 palynofloras, referred here to association sequences Y1 and Y2, overlie other associations similar to those characteristic of the Lambeth Group in eastern England (here labelled Tx). The paucity of good quality samples available for this study from the Sele Formation units 2a and 2b interval restricts further description of the palynofloras included in them. However, they appear to represent the palynofloras of lowstand sediments deposited away from the margins of the basin. Comparison of the onlap model proposed for the 54.5-54.0 Ma period here (Fig. 31) with the stratigraphic studies of Stewart (1987) and Jones

Fig. 38. Comparison of lithological units in the Outer Moray Firth to those in southeastern England based on the palynologial results of this study. Note that the boundary of the Harwich Formation extends upsection to become partly equivalent to the Walton Member; UGS is used to indicate the unnamed glauconitic sands and silty sands attributed to the Tilehurst and Swanscombe Members by King (1981). The subdivisions of the Balder Formation used by Malm et al. (1984) are shown on the right of the diagram; LTZ, lower tuff zone; MTZ, main tuff zone; UTZ, upper tuff zone. The other subdivisions used are those of Knox & Holloway (1992).

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND & Milton (1994) are possible with the aid of the association sequences determined in 81/46a, 14/25-1 and 15/28a-3. The initial phase of Harwich Formation onlap (Y3b-Y5) is apparently represented in unit 8 of Stewart (1987) and package T45 of Jones & Milton (1994). As these Outer Moray Firth units include all of the Sele Formation units $2a-$3 it is only the upper part of the section that is comparable. Jones & Milton (1994) suggest that this is a period of gently rising relative sea level in the closing phase of lowstand deposition, followed by a significant rise in relative sea level (500 m), causing the Balder Formation and Beauly Member to onlap westward. While this sequence 9 or T50 phase can be related to the Y6-Y9 major onlap in southeastern England (Figs 30a & b & 31), the absence of an earlier rise in relative sea level equivalent to the parasequence 1-3 interval (Y3-Y5) in the Outer Moray Firth implies significant differences in tectonic history. It is however of greater significance to note the palynofloral evidence for rising relative sea levels in the Y6-Y9 interval in southeastern England, the Outer Moray Firth and west of the Shetland Islands (see below). This record is too widespread to be attributed to local tectonic factors, and is rather a reflection of the rifting of the North Atlantic and the peak of phase 2 volcanism (Knox & Morton 1988). Recognition of association sequences in the three North Sea Basin sections can also assist in our understanding of the relationship of these deposits to the Paleocene-Eocene boundary. Palynofloras from Sele Formation units S2a and S2b are older than those occurring at the position of tephra layer -17 in 81/46a at the base of the Sele Formation. Unit 3. As sediments in the London and Hampshire Basins, equivalent on palynological grounds to Sele unit 1, have been shown to contain calcareous nannofossils attributable to zone NP9 of Martini (1971) (Siesser et al. 1987; Ellison et al. 1996), it is probable that the base of nannofossil zone NP10, which is used to approximate to the Paleocene-Eocene boundary, occurs between the upper limit of the Sele Formation unit S 1 and the upper limit of the Sele Formation unit S2b. Whether its position is coincident with the major sequence boundary at the base of the Sele Formation unit 2a and the upper limit of the dinoflagellate cyst Apectodinium augustum, as suggested by Powell (1988) and Knox & Holloway (1992), remains unproven.

Borehole 90/3 (59~ ' N 5~ ' W), drilled on the eastern end of the Wyville-Thompson Ridge. This cored borehole penetrated 91.5 m of early Eocene deltaic sediments overlain by Miocene marine strata. No formal lithostratigraphic division exists for this area; accordingly the informal subdivision used in Hitchen et al. (1995) is used (Fig. 39). The application of the association sequence method to this section was dependent on the rich pollen floras in a section devoid of in situ microplankton. The high quality of the samples has allowed a full subdivision from Y6b to Y9 to be identified in a sequence closely similar, both lithologically and florally, to the Beauly Formation of the Moray Firth area (Figs 36, 38 & 39). Together with the three examples from the North Sea Basin, the short section of deltaic sediments in 90/3 demonstrates the potential for a regional application of association sequences Y3-Y9. The possibility of identification of association sequences outside the area of their original definition (i.e. eastern England) itself points to evidence for sedimentary sequence, sea level and climatic controls over their extent and distribution, giving them an inherent reliability for stratigraphic application.

West of Shetland The stratigraphy of the deposits recorded and analysed in southeastern England are addition compared to results derived from BGS cored

249

Fig. 39. Summary log, BGS Borehole 90/3.

250

D.W. JOLLEY

The collection of sample material for this study would have been impossible without the help of R. W. O'B. Knox, C. King and J. R. Ali. The author would also like to thank R. A. Ellison and P. Hopson for assisting in the acquisition of samples. London Underground Ltd, British Rail, Shell/Esso and BP are all thanked for their permission to sample borehole material. Extensive discussions with R. W. O'B. Knox, C. King and J. R. Ali have been

formative in the conclusions of this study. E. G. Spinner is thanked for his constructive comments during the course of this work, and J. P. G. Fenton and R. W. O'B. Knox are thanked for their constructive referees' comments. This study would not have been possible without access to BGS material and is published with the approval of The Director, British Geological Survey (NERC).

References ANDERSEN, S. A. 1937. De vulkanske Askelag i Verjgennemskaeringen ved Olst og deres Udbredelse i Danmark. Danmarks Geologiske UndersCgelse, 59, 1-50. ALI, J. R. & JOLLEY, D. W. 1996. Chronostratigraphic framework for the Thanetian and lower Ypresian deposits of southern England. This volume. KING, C. & HAILWOOD, E. A. 1993. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD, E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy, Geological Society, London, Special Publication, 70, 99-125. BERGGREN, W. A. & AUBRY, M. P. 1996. A late Paleocene--early Eocene NW European and North Sea magnetobiochronological correlation network.

This volume. , KENT, D. V., SWISHER, C. C. III & AUBRY, M.-P. 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A., KENT, D. V., AUBRY, M.-E & HARDENBOL, J. (eds)

Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Volume, 54, Tulsa. BOGGLED, O. B. 1918. Den vulkanske Aske i Moleret samt en Oversigt over Danmarks ~eldre Tertiaerbjaergarter. Danmarks Geologiske UndersCgelse, 33, 1-159. BOSWELL, P. G. H. 1916. The stratigraphy and petrology of the Lower Eocene deposits of the north-eastern part of the London Basin. Journal of the Geological Society, London, 71,536-591. 1927. The Geology of the Country around Ipswich.Memoirs of the Geological Survey England & Wales. COSTA, L. I. & DOWNIE, C. 1976. The distribution of the dinoflagellate Wetzeliella in the Palaeogene of north-western Europe. Palaeontology, 19(4), 591-614. , DEN1SON, C & DOWNIE, C. 1978. The Paleocene/Eocene boundary in the Anglo-Paris Basin. Journal of the Geologial Society, London, 135, 261-264. COX, F. C., HAILWOOD,E. A., HARLAND,R., HUGHES, M. J., JOHNSTON, N. & KNOX, R. W. O'B. 1985. Palaeocene sedimentation and stratigraphy in Norfolk, England. Newsletters on Stratigraphy, 14(3), 169-185. CURRY, D. 1965. The Palaeogene Beds of South-East England. Proceedings of the Geologists' Association, 76, 151-174. DAVEY, R., DOWNIE, C., SARJEANT,W. A. S. & WILLIAMS,

G. L. 1966. Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bulletin of the British Museum (Natural History) Geology Supplement, 3, 1-248. ELLISON, R. A. 1983. Facies distribution in the Woolwich & Reading Beds of the London Basin, England. Proceedings of the Geologists' Association, 94, 311-319. ~, KNOX, R. W. O'B, JOLLEY, D. W. & KING, C. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata of the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. --, ALI, J., HINE, N. & JOLLEY, D. W. 1996. Recognition of Chron C25n in the upper Paleocene Upnor Formation of the London Basin, UK. This

volume. GOODMAN, D. K. 1979. Dinoflagellate 'communities' from the Lower Eocene Nanjemoy Formation of Maryland, U.S.A. Palynology, 3, 169-190. HARLAND, R. 1983. Distribution maps of Recent dinoflagellate cysts in bottom sediments from the North Atlantic Ocean and adjacent seas. Palaeontology, 26, 321-387. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks Geologiske Underscgelse, Series A, 7, 1-39. HITCHEN, K., JOLLEY, D. W. & HARLAND, R. 1995. Micropalaeontological and palaeo-environmental investigation of the lowermost interval in BGS borehole 90/3, offshore north-west Scotland: a case history with regional implications. Scottish Journal of Geology, in press. ISLAM, M. A. 1984. A study of early Eocene palaeoenvironments in the Isle of Sheppey as determined from microplankton assemblage composition. Tertiary Research, 6(1), 11-21. JOLLEu D. W. 1992. Palynofloral association sequence stratigraphy of the Palaeocene Thanet Beds and equivalent sediments in eastern England. Review of Palaeobotany and Palynology, 74, 207-237. -& SPINNER,E. 1989. Some dinoflagellate cysts from the London Clay (Palaeocene-Eocene) near Ipswich, Suffolk, England. Review of Palaeobotany and Palynology, 60, 361-373. & -1991. Spore-pollen associations from the lower London Clay (Eocene), East Anglia, England. Tertiary Research, 13, 11-25. JONES, R. W. & MILTON, N. J. 1994. Sequence development during uplift: Palaeogene stratigraphy and relative sea-level history of the Outer Moray Firth, UK North Sea. Marine & Petroleum Geology, 11, 157-165. -

-

EARLIEST EOCENE PALYNOSTRATIGRAPHY, E ENGLAND KING, C. 1981. The stratigraphy of the London Clay and associated deposits. Tertiary Research Special Paper, 6, 1-158. 1983. Cainozoic micropalaeontological biostratigraphy of the North Sea. Institue of Geological Sciences Report, 82/7, 1-40. -& CURRY, D. 1992. Molluscs from the Tilehurst Member (London Clay Formation, Early Eocene) at Crondall and Up Nately (Hampshire). Tertiary Research, 13, 141-146. KNOX, R. W. O'B. 1983. Volcanic ash in the Oldhaven Beds of southeast England, and its stratigraphical significance. Proceedings of the Geologists' Association, 94, 245-251. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe; an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. 1985. Stratigraphic significance of volcanic ash in Palaeocene and Eocene sediments at Sites 549 and 550. In: GRAC1ANSKY,P .C., POAG, C. W. DE et al. Initial Reports of the Deep Sea Drilling Project, 80, 845-849. -& ELLISON, R. A. 1979. A lower Eocene ash sequence in SE England. Journal of the Geological Society, London, 136, 251-253. -& HARLAND,R. 1979. Stratigraphical relationships of the early Palaeogene ash-series of NW Europe. Journal of the Geological Society, London, 136, 463-470. & HOLLOWAY,S. 1992. 1. Paleogene of the Central and Northern North Sea. In: KNox, R. W. O'B. & CORDEY, W. G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey. & MORTON, A. C. 1983. Stratigraphical distribution of early Palaeogene pyroclastic deposition in the North Sea Basin. Proceedings of the Yorkshire Geological Society, 44(3), 355-363. -& 1988. The record of early Tertiary N Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. , HARLAND,R. & KING, C. 1983. Dinoflagellate cyst analysis of the basal London Clay of southern England. Newsletters on Stratigraphy, 12, 71-74. , MORIGI, A. N., ALl, J. R., HAILWOOD, E. A & HALLAM, J. R. 1990. Early Paleogene stratigraphy of a cored borehole at Hales, Norfolk. Proceedings of the Geologists'Association, 101(2), 145-151. , MORTON, A. C. & HARLAND, R. 1981. Stratigraphical relationships of Paleocene sands in the UK sector of the Central North Sea. In: ILLING,L. V. & HOBSON, G. D. (eds) Petroleum Geology of North-west Europe. Heyden & Son, London, 267-281. LENTIN, J. K. & WILLIAMS,G. L. 1993. Fossil dinoflagellates: index to genera and species, 1993 edition. AASP Foundation, Contribution Series, 28, 1-855. LOTT, G. K., KNOX, R. W. O'B., HARLAND,R. & HUGHES, M. J. 1983. The Stratigraphy of Palaeogene -

-

-

-

-

-

251

Sediments in a Cored Borehole off North-east Yorkshire. Institute of Geological Sciences Report, 83/9,

1-9.

MALM, O. A., CRISTENSEN,O. B., FURNES,H., LOVLIE, R., RUSELATTEN, H. & OSTBY, K. L. 1984. The Lower Tertiary Balder Formation: An organogenic and tuffaceous deposit in the North Sea region. In: SPENCER, A. M. ET AL. (eds) Petroleum Geology of the North European Margin. Norwegian Petroleum Society, Graham & Trotman, 149-170. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings II Planktonic Conference Rome 1970, Edizioni Technoscienza, Rome, 739-783. MUDGE, D. C. & COPESTAKE,P. 1992a. A revised lower Paleogene lithostratigraphy for the Outer Moray Firth, North Sea. Marine & Petroleum Geology, 9, 53-69. & 1992b. Lower Paleogene stratigraphy of the northern North Sea. Marine & Petroleum Geology, 9, 287-301. MULLER, J. 1959. Palynology of Recent Orinoco delta and shelf sediments: Reports of the Orinoco Shelf Expedition; Volume 5. Micropaleontology, 5, 1-32. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for the Latest Paleocene and Earliest Eocene sediments from the Central North Sea. Review of Palaeobotany and Palynolgy, 56, 327-342. 1992. Dinoflagellate cysts of the Tertiary System. In: POWELL, A. J. (ed.) A Stratigraphic Index of Dinoflagellate Cysts. Chapman & Hall, London, 155-252. PRESTWICH, J. 1846. On the Supracretaceous Formations of the Isle of Wight as exhibited in the section at Alum Bay and Whitecliff Bay. Quarterly Journal of the Geological Society, London, 2, 223-259. 1850. On the structure of the Strata between the London Clay and the Chalk in the London and Hampshire Teriary systems. Part 1. Quarterly Journal of the Geological Society, London, 6, 252-258. SIESSER, W., WARD, D. J. & LORD A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian stage (Palaeocene) in the type area. Journal of Micropalaeontology, 6(1), 85-102. STAMP,L. D. 1921. On the beds at the base of the Ypresian (London Clay) in the Anglo-Franco-Belgian Basin. Proceedings of the Geologists' Association, 32, 57-108. STEWART,I. J. 1987. A revised stratigraphic interpretation of the Early Paleogene of the central North Sea. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 557-576. VAN WAGONER,J. C., POSAMENTIER,H. W., MITCHUM,R. M. JR, VAIL, P. R., SARG, J. F., LOUTIT, T. S. & HARDENBOL, J. 1988. An overview of the fundamentals of sequence stratigraphy and key definitions. In: WmGUS, C. K., HASTINGS,B. S., KENDALL, C. G. STC., POSAMENTIER,H. W., ROSS, C. A. & VAN WAGONER, J. C. (eds) Sea Level Changes:An Integrated Approach. Society of Economic -

-

-

-

252

D.W. JOLLEY

Paleontologists and Mineralogists, Special Publication, 42, 39-45. VENEBLES, E. M. 1962. The London Clay of Bognor Regis. Proceedings of the Geologists' Association, 73, 245-271. WETHERELL, N. T. 1836. Observations on some of the Fossils of the London Clay, and in particular those organic remains which have recently been discovered in the tunnel from the London and Birmingham railroad. Philosophical Magazine, (3)9, 462-469.

WHITTAKER,W. 1866. On the Lower London Tertiaries of Kent. Quarterly Journal of the Geological Society, London, 22, 404-435. WRIGHT, C.A. 1972. The recognition of the planktonic formainiferid datum in the London Clay of the Hampshire Basin. Proceedings of the Geologists' Association, 83, 413-420. WRIGLEY,A. 1940. The faunal succession in the London Clay, illustrated by some new exposures near London. Proceedings of the Geologists' Association, 51, 230--245.

Appendix A Harwich Formation (Ellison et al. 1994) Two new members of the Harwich Formation of the East Anglia area are proposed here.

Orweli M e m b e r The Orwell Member is described here as a new lithological member of the Harwich Formation as defined by Ellison et al. (1994). Name: From the River Orwell, which flows through south Suffolk from Ipswich to the sea at Felixstowe. Type section: Low cliffs on the north bank of the River Orwell at Bridge Wood (TM 182 408 to 189 401) expose the middle and upper part of the member; the lower part and contact with the underlying Reading Formation can be augered at the same site and is present at the reference sections. Reference sections: BGS Shotley Gate Borehole (Knox & Ellison 1979) and the low cliff and foreshore exposures at Ferry Cliff near Woodbridge, Suffolk. Lithology: The member is composed of three lithological units which are described below, oldest to youngest. Unit A. Fine buff sands with common wellrounded black flint pebbles and faunal debris at the base, which rest with evident disconformity on the Lambeth Group. These were described by Boswell (1916) as the Suffolk Pebble Bed, reaching up to 1 ft (0.33 m) in thickness. From personal observation, the pebbles are concentrated at the base of the unit which may reach up to 95 cm in thickness. The section at Bramford Brick Pit and the section at Nacton are regarded as having typical representations of this unit. Unit B. Grey-brown micromicaceous, intensely bioturbated silty sands and sandy silts with shell fragments and discontinuous laminations of dark grey-black claystone, especially at the base.

Tuffaceous material is microscopically visible in this unit which reaches c. 1.75 m in maximum thickness at Nacton. Unit B rests with apparent conformity on unit A. The section at Nacton has a typical exposure of this unit. Unit C. A unit of tuffaceous sandy silts c. 2.50 m in thickness, with fine sand laminations and burrow fills. The tephra layers are largely discontinuous, nine being recorded by Knox & Ellison (1979) in the Shotley Gate Borehole; disseminated ash is common microscopically. Sections at Levington and Nacton also expose this unit with numerous discontinuous and some continuous tephra layers, these sections are regarded as being typical of this unit. Unit C rests with apparent conformity on unit B. Boundaries: The Orwell Member marks the base of the Harwich Formation in Suffolk and north Essex; it rests disconformably on the uppermost sands of the Reading Formation. Exposures of the Suffolk Pebble Bed are now few, but show an undulating erosion surface with abundant reworking of Reading Formation sands into the basal Orwell Member, a factor referred to by Boswell (1916, 1927) (who gives probably the best account of this, then extensively exposed boundary). The upper boundary of the member is coincident with the lower boundary of the Wrabness Member, which is marked by a sharp disconformity surface with a glauconite concentration first identified by King (pers. comm. in Jolley & Spinner 1991). Regional correlation: To the north, palynological evidence suggests that the Orwell Member passes laterally into the Hales Clay, although the lowermost units of the latter are missing in Suffolk (see Fig. 30) demonstrating onlap. The Orwell Member onlaps to the southwest before passing into the glauconitic silty sands referred to the Swanscombe and Tilehurst Members north and west of London, respectively. To the south, unit C of the Orwell Member passes laterally into the Oldhaven Beds of the north Kent coast and London area.

EARLIEST EOCENE PALYNOSTRATIGRAPHY,E ENGLAND

Wrabness Member This new member is erected to describe the tuffaceous siltstones found in East Anglia that lie between the Orwell Member and the base of the London Clay Formation. Name: From the best exposed section of these sediments on the River Stout at Wrabness, Essex. Type section: The cliff section at Wrabness, Essex (TM 172 323), which exposes the complete member from the contact with the underlying Orwell Member to the base of the overlying Walton Member. Reference section: The upper part of the member is well exposed in the cliffs and on the foreshore at Walton on the Naze, Essex (TM 267 244). Lithology: This member is composed of two distinct units which are described below in stratigraphic order. Unit A. Unit A is a tuffaceous siltstone c. 10 m in thickness with 32 complete tephra layers recorded in the section at Wrabness, disseminated ash also being recognizable microscopically. Concentrations of tephra layers are observed at the outcrops on the Stour and Orwell Estuaries, in particular a group of 12 closely-spaced layers near the top of the unit is highly distinctive. Tephra layers become less prominent towards the upper

253

limit of the unit which is taken at the base of the fine sand beds at the top of the section. Unit B. Unit B rests with apparent disconformity on unit A and reaches 1.60 m in the Ipswich area. It is composed of a bed, or beds of fine, intensely bioturbated, sand showing no apparent structure. The upper limit of this unit is marked by the unconformity between the Harwich Formation and the overlying Walton Member of the London Clay Formation. Boundaries: The lower boundary of the Wrabness Member is marked by the concentration of glauconite above the basal disconformity with the Orwell Member. The upper limit is marked by the disconformably overlying Walton Member claystones, which contain sporadic small, well-rounded black flint pebbles. These claystones rest on the upper fine sandy beds of Wrabness Member unit B. Regional correlation: To the north and east the Wrabness Member passes laterally into the Balder Formation (both units B1 and part of B2; see Fig. 38). Further south, basal onlap is recorded (Fig. 30a & b), only the uppermost beds of unit A passing laterally into the glauconitic siltstones seen above the Oldhaven Member at Sheppey (Fig. 30b). To the west the siltstones of the member pass into the glauconitic silty sands of the undifferentiated Harwich Formation north and west of London (Swanscombe and Tilehurst Members).

Appendix B Selective rnicroplankton and pollen/spore frequency diagrams for the key sections in the study, describing the association sequence correlation, are available as Supplementary Publication No. SUP18104 (36pp) from the Society Library or the British Library Document Supply Centre, Boston Spa, Wetherby, W. Yorks LS23 7BQ, UK. The systematic palynology of the dinoflagellate cysts is that of Lentin & Williams (1993); that of the other algae, acritarch, pollen and spore taxa named is given below.

Pollen and Spores Alnipollenites verus (Potoni6) Potoni6 1931 Baculatisporites primarius (Wolff) Pflug & Thomson 1953

Caryapollenites circulus (Pflug) Krutzsch 1961 C. triangulus (Pflug) Krutzsch 1961 C. veripites (Wilson & Webster) Nichols & Ott 1978

Compositoipollenites rhizophorus subsp, rhizophorus (Potoni6) Krutzsch & Vanhoorne 1977

Cicatricosisporites dorogensis Potoni6 & Gelletich 1933

Inaperturopollenites

distichiforme (Simpson) Jolley & Morton 1992 I. dubius (Potoni6 & Venitz) Thomson & Pflug 1953 L hiatus (Potoni6) Pflug & Thomson 1953 Intratriporopollneites microreticulatus Mai 1961 Interpollis supplingensis (Pflug) Krutzsch 1961 Laevigatosporites haardti (Potoni6 & Venitz) Thomson & Pflug 1953

Leiotriletes adriennis Leiotriletes wolffi Krutzsch 1962 Liquidambarpollenites stigmosus (Potoni6) Raatz 1937

Milfordia

hungarica (Kedves) Krutzsch & Vanhoorne 1977 M. incerta (Thomson & Pflug) Krutzsch 1961 Momipites anellus Nichols & Ott 1978 M. tenuipolus Anderson 1960 M. triradiatus Nichols 1973 Monocolpopollenites tranquilis (Potoni6) Thomson & Pflug 1953

254

Nyssapollenites

D . W . JOLLEY

kruschi

subsp,

analepticus

(Potoni6) Thomson & Pflug 1953 Pistilipollenites mcgregorii Rouse 1962 'Platycarya' platycaryoides (Roche) Frederiksen & Christopher 1978 Platycaryapollenites swastacoidus (Elsik) Frederiksen & Christopher 1978 Plicapollis pseudoexcelsus (Krutzsch) Krutzsch 1961 Sparganiaceaepollenites cuvillieri (GruasCavagnetto) Roche 1968 S. magnoides Krutzsch 1970

Subtriporopollenites anulatus subsp, anulatus (Pflug & Thomson) Krutzsch 1961

Thomsonipollis magnificoides Krutzsch 1960 Triatriopollenites roboratus Pflug 1953 T. subtriangulus (Stanley) Fredericksen 1979 Tricolpites hians Stanley 1965 Tricolpopollenites retiformis Thomson & Pflug 1953

Triporopollenites coryloides Pflug 1953 T. robustus Pflug 1953 Acritarchs and other algae Micrystridium bigotii Deflandre 1937 Micrhystridium fragile Deflandre 1947 Pediastrum bifidites Wilson & Hoffmeister 1953

Palaeoenvironments in the North Sea Basin around the Paleocene-Eocene boundary: evidence from diatoms and other siliceous microfossils A L E X A N D E R G. M I T L E H N E R

Micropalaeontology Unit, Research School of Geological & Geophysical Sciences, University College London & Birkbeck College, Gower Street, London WCIE 6BT, UK (Present address: Department of Botany, The Natural History Museum, Cromwell Road, London SW7 5BD, UK) Abstract: Assemblages of siliceous microfossils (diatoms, silicoflagellates, radiolaria, and ebridians) have been examined from calcareous concretions (cementstones)in the Fur Formation, an ash-bearing diatomite which crops out around the Limfjorden region of Jutland, northern Denmark. Although microfossil assemblages (particularly diatoms) change character through the deposit, only one major turnover occurs that is reflected by all of the microfossil groups examined. This biotic event indicates a marked increase in marine influence, and probably climatic warming. It appears to mark the beginning of the early Eocene transgressive phase that affected the whole of the North Sea Basin and eventually re-established this isolated basin's links with the North Atlantic and Tethys. Other diatom assemblages examined from offshore samples towards the centre of the basin show strong similarities with those from the Fur Formation and confirm that the lower part of this deposit correlates to the Sele Formation, whilst the upper part correlates to the Balder Formation. The lowest occurrence of the widely occurring diatom Fenestrella antiqua is suggested as a practical marker for the base of the Eocene in both offshore and onshore sediments in the North Sea Basin. Restricted circulation processes involving seasonal, monsoon-driven upwelling leading to sporadic basin eutrophication and water-column stratification, are considered to account for the widespread proliferation of siliceous microfossils in the North Sea Basin during the late Paleocene to early Eocene. The precise biostratigraphic determination of the Paleocene-Eocene boundary in NW Europe has long proved problematic, as the sediments preserved there are devoid of the calcareous nannofossils normally utilized for age determination (Knox 1984; Gallagher 1990). Although some progress has been achieved via the use of organicwalled microfossils, chiefly dinoflagellates (e.g. Mudge & Bujak 1994), correlation across the basin has been hampered by the restricted marine nature of the North Sea Basin during the period in question, with the time-transgressive nature of species occurrences being increasingly recognized (Schr/Sder 1992). This paper seeks to examine the evidence provided by siliceous microfossils (diatoms, radiolaria, silicoflagellates and ebridians) which occur in abundance across the PaleoceneEocene boundary around the North Sea Basin, in terms of biostratigraphic potential and palaeoenvironmental reconstruction.

Previous studies: siliceous microfossils in the Paleocene-Eocene of the North Sea Basin

Diatoms The widespread occurrence of diatoms in upper Paleocene and lower Eocene sediments from the

North Sea Basin area has long been recognized (Heiberg 1863; Grunow 1866; Kitton 1871; Shrubsole & Kitton 1881; Schulz 1927; Staesche & Hiltermann 1940; Bettenstaedt et al. 1962; Benda 1965, 1972; Fenner 1988; Homann 1991). In addition to the well-documented onshore diatomites which crop out in the Limfjorden region of Jutland, northern Denmark (designated the Fur Formation by Pedersen & Suflyk 1983), hydrocarbon exploration led to the recognition that fossil diatoms (mainly replaced or infilled by pyrite) are also widespread in offshore sequences within the North Sea (Jacqu6 & Thouvenin 1975; Hughes 1981; King 1983; Maim et al. 1984). However, little detailed taxonomic, biostratigraphic or palaeoenvironmental work has been published on the Tertiary diatom assemblages of the central North Sea. This information is held within the confines of the oil companies and their ancillary service companies, and the diatoms are poorly preserved, replaced and/or infilled by pyrite. These two factors have hindered the study of these diatom floras, which have nevertheless been found to be of great use in identifying certain stratigraphic intervals (Knox & Holloway 1992; Mudge & Copestake 1992a, b; Mitlehner 1994), in particular within the Balder and Sele formations, at around the Paleocene-Eocene boundary. These

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 255-273.

255

256

A.G. MITLEHNER

assemblages and their onshore equivalents are the subject of this study. Previous studies on the Fur Formation diatomite have been primarily lithostratigraphic (BCggild 1918), sedimentological (Pedersen 1981; Pedersen & Surlyk 1983), or taxonomic (Heiberg 1863; Grunow 1866; Kitton 1871; Homann 1991); limited biostratigraphic studies involving diatoms were conducted by Benda (1972) and Fenner (1985, 1994). However, to date there has been no integrated study which involves all the siliceous microfossil groups occurring in the Fur Formation and their correlation with offshore sections in the North Sea. Silicoflagellates a n d ebridians

Silicoflagellates have a more sporadic occurrence than diatoms in the North Sea area, as their fragile skeletons are prone to dissolution. They are therefore absent from most of the clastic post-Danian sequence in the North Sea Basin, except for the Fur Formation diatomite where they occur in abundance (Loeblich et al. 1968; Martini 1974; PerchNielsen 1976). Previous studies have shown that silicoflagellate assemblages within this deposit show definite changes in assemblage composition which have enabled its division into biozones. The first silicoflagellate-based division of the Fur Formation was attempted by Martini (1974), who referred the deposit to the Dictyocha deflandrei and D. naviculoidea zones. A more refined scheme was later devised by Perch-Nielsen (1976), who placed the Fur Formation within the Naviculopsis constricta zone, and further subdivided it into five subzones. Sediments deposited in the central parts of the North Sea were strongly affected by diagenesis, and the only published occurrence of silicoflagellates in this region is from calcareous concretions within the Balder Formation. These concretions, recovered from a core in the Norwegian sector of the Viking Graben, contained diagnostic silicoflagellates which enabled Maim et al. (1984) to place the lower part of the Balder Formation within the N. aspera subzone of Perch-Nielsen (1976). Ebridians have also been described from the Fur Formation (Loeblich et al. 1968), but their occurrence patterns through this deposit have not been analysed. Radiolaria

Radiolaria occur at certain intervals in the offshore Paleogene sequence of the North Sea, often in high abundance, but are commonly poorly preserved. Spumellarian forms, generally referred to as Cenosphaera spp. (see King 1983, pl. 1) are often

abundant, identifying important marker horizons, which may indicate a deepening of the basin and a more oceanic connection. Other morphologies occur sporadically through the sequence, but poor preservation often precludes accurate identification. Onshore, Cenosphaera sp. has been recorded in the 'Unter-Eoz~in' of northern Germany (Bettenstaedt et al. 1962). The presence of rare but well-preserved radiolaria in the Fur Formation was noted by a number of workers (Perch-Nielsen 1976; Homann 1991, p.12), but these have never been documented in any detail.

Sampling procedure Siliceous microfossils (diatoms, silicoflagellates, ebridians, radiolaria) were examined from six calcareous concretions (cementstones) collected by Homann (1991) from within the Fur Formation, from the Limfjorden region of northern Denmark (Fig. 1). The position of these samples within the sequence of 179 volcanic ash layers (Bcggild 1918) is given in Fig. 2. Material used in this study took the form of cleaned diatomaceous residue, previously prepared by M. Homann and sent in phials to The Natural History Museum, London. The extreme purity of many of the layers of diatomite from which these samples came meant that minimal processing was required (Homann 1991, p. 13). Therefore, it was only necessary to take a pipetted sample from each phial, and dry it on a mica strip at normal room temperature, before picking and/or coating the sample for scanning electron microscope (SEM) examination. Strewn slides from each of M. Homann's six samples were mounted and are now housed in the diatom collections at The Natural History Museum, London. Diatom identifications were carried out with reference to Van Heurck (1896), Sims (1989, 1990) and Homann (1991). SEM study of some taxa also occurring in pyritized form in offshore sediments led to some taxonomic revision; these are the subject of a separate publication (Mitlehner 1995). For the purposes of this paper, these species have been left in open nomenclature. A minimum of 400 diatom specimens were counted from each of these slides; silicoflagellates, ebridians and radiolaria were also recorded. Silicoflagellates and ebridians were identified with reference to Loeblich et al. (1968) and PerchNielsen (1976); radiolaria were identified with reference to Bj~3rklund (1976). Statistical analysis was not attempted on the assemblages, due to the effects of preservational bias and relatively small sample size (Andrews 1972; Battarbee 1986). The opal-A skeletons of siliceous microfossils are prone to the effects of dissolution, both within

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN

257

Fig. 1. Localities mentioned in text. (a) North Sea Basin. NB: localities 1, 2 and 5 are the subject of this study; other localities from Mitlehner (1994, 1995). (b) Limfjorden area, northern Denmark. Underlined place names indicate sample localities. Harle: samples from here analysed by Fenner (1988, 1994).

the water column (where 80-90% of skeletons may be dissolved before settling, Calvert 1974) and subsequently during burial and diagenesis (Berner 1968; Blome 1984; Kakuwa 1984). The former may be offset by the incorporation of diatom

frustules into the faecal pellets of copepods and other zooplankton (Schrader 1971); recent studies of sediment traps on the deep ocean floor and beneath coastal upwelling zones have shown that there is a correspondence between fluxes of diatom

258

A.G. MITLEHNER

Fig. 2. Position of siliceous microfossil samples within the sequence of ash layers (numbers assigned by B0ggild 1918) preserved in the Fur Formation. After Homann (1991), with lithostratigraphy after Pedersen & Surlyk (1983).

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN and radiolarian skeletons and their abundances in the overlying water column (e.g. Schrader & Sorknes 1991; Thunell et al. 1994). Concretions, normally of calcite or phosphate, may arrest the effects of burial and diagenesis by the early precipitation of carbonate minerals, which protect the siliceous skeletons from the effects of compaction and destructive pore fluids percolating through the host sediment (Kakuwa 1984). Microfossils contained within these concretions are much better preserved than those in the surrounding rock, and are often remarkable for the preservation of delicate structures not normally preserved in diatoms and radiolaria which are older than late Neogene in age (Blome 1984).

Results and interpretation Results from microfossil counts of the strewn slides are given in Table 1. One noticeable assemblage change is evident in all of the siliceous fossil groups, which can be used to identify several assemblages. These changes correlate well with the boundaries of the silicoflagellate zones of PerchNielsen (1976). The chief characteristics of each assemblage are given below, and represented graphically in Fig. 3a & b. Author citations for all of the species encountered are given in Table 2.

259

elegans. Pennate diatoms occur in lower numbers than the underlying assemblage, and there is an absence of the neritic species Actinoptychus senarius. Two radiolarian taxa, Ceratocyrtis hystricosus and Peridium longispina are noted for the first time from the Fur Formation. The silicoflagellate assemblage is dominated by Corbisema hastata globulata. Ebridians are absent from all of these samples. The diatom assemblage changes composition in this interval, as fluctuations in dominant taxa are probably due to localized blooms, caused by seasonal variations in nutrient levels (J. Fenner, pets. comm.). Therefore, this assemblage was not subdivided into diatom subzones. The abundance of diatom resting spores (Round et al. 1990, p. 52) indicates environmentally stressed conditions related to seasonal nutrient depletion and/or other stress; some workers (e.g. Kitchell et al. 1986; Harwood 1988) have suggested that this survival mechanism may explain why diatom taxa did not suffer the same extinction rates of other marine planktonic organisms across the K-T boundary. It is possible that the proliferation of diatoms across the Paleocene-Eocene boundary in the North Sea Basin may be similarly explained by this adaptation. Naviculopsis aspera assemblage

Naviculopsis danica assemblage

[ash layer +291

[ash layer-20]

The uppermost sample analysed in this study shows broad similarities with the preceding assemblage, but is distinguished by the predominance of the diatom Stephanopyxis turris, a taxon which may be a resting spore genus (Round et al. 1990, p. 158). Other resting spore taxa show a marked increase, including Goniothecium odontella, Pterotheca aculeifera and Pyxilla carinifera. The silicoflagellates are dominated by Corbisema glezerae, a characteristic marker for the N. aspera subzone of Perch-Nielsen (1976).

The dominant diatom taxa in sample -20 are benthic forms such as the pennate species Rhaphoneis lancettula and Rutilaria sp., and the neritic genus Actinoptychus senarius. There is an absence of the larger planktonic centric and chainforming species found in the succeeding samples, with the notable exception of the centric taxon Craspedodiscus moelleri, which does not occur in any of the other samples studied. This sample is also characterized by large numbers of silicoflagellates (mainly Naviculopsis danica), and fairly abundant ebridians; radiolaria are absent. This assemblage probably indicates a shallow, neritic environment with depths of < 50 m (R. Ross, pers. comm.). Corbisema naviculoidea assemblage

[ash layers -11, -10, - 8 a n d - 6 ] The four samples in this assemblage are characterized by a dominance of planktonic centric diatom taxa ( Coscinodiscus morsianus, Fenestrella antiqua, and Stellarima microtrias) and the chain-forming diatom species Trinacria regina,

T. excavata, Solium exsculptum and Hemiaulus

Comparison with previous work correlation

North Sea Basin and adjacent areas Fur Formation. In a thorough taxonomic survey of diatoms from the Fur Formation (with transmitted light micrographs only), Homann (1991) considered that it was not possible to erect diatom zones for the deposit as various taxa appear, vanish and reappear through the samples (possibly due to localized blooms and sediment reworking). Pedersen (1981) studied the sedimentology of the Fur Formation, and found that laminae within

260

A.G. MITLEHNER

, Sll!lqns/sFtaol, snas!polos s m v m q a f snlno!maH snlnlomna snlnm, ruaH HallaOUt gnas!popadsv.O ,moadsng *nas!poot?In g gnuvuas gmlo&dou!lo V vumyaom ~!auo~ldmRt oguolqo s!xs YnlOal!d m.aovu!.zs oo!aaq?, O!lmVd sng~o "J~ xa snas!pou!aso D s!l!ls~ snlnvlUZaH mag!aVla v!ualosoz!q~t ??8aaq!aq ~tlaluopo ~ l s . ~ a s?doatotuopo

<

OlOtdlgUO0

s!sdolno!aVN

saaods ~u!lsaa

0Z 8I

I I ~ ~

~, 6 g 17 9 I I ~ g

E OI 6

8

L

8

17I

6

L II

g 9

~[ t,g

91 II

[ Jr 17 [ E[ L

L EI ~ 17 P [E 9E

8 ~ 9~ OOE 6 fg

6 ! g 17E # EI 9[ E~ g9 Eli OOE 08[ 9I ~ 17 6E OI ~I

s~. loaa!m ouz.tZ•llal S

saaods ~UllSaa (I661 uu~tuoH) aaqtunu oldme S

~2

86E g f f P6E Off f6g g6g

mvlnuaaa s!uuau! outas!qao o vlnqff "G "jje "ds mtao~OajG mvgu~ vqo~ vap!~ ~ s!aaluvaa~td mlaO&a!G ma!.tlsuoa s!,dolno!aVN ~omm ,Vu~aau! "D s!tuaau! s!utaau! "D a~aazalg'D mmsvq mm~m/ "D mvlnqol~ vl~l~vq outa~!q.toD '

L 9

~ OI E

I O[

gI L

E I gI /7

# fI E E1

(uet.p!aq~t) ong!qtuv t?t..tqaslVd ~!a~l~ "l~ (ueI-relo!pe~I) snsoo!algdq s!ladooloaaD i (ue!ael~ umu,ds!gUOl um!pHad

1

9

f L 8 Og

8 fI

8

~

8

g

fI 9 17

6

IE

LE

8

s so!aad S Of 8~ 9f 9~ 6E 9E palunoaaaqmnu l~lO,L IfP f81~ glTg Olg Llg EEP soaods BuBsoa "lopuI fE "lopu! so!aods lo snuoO L !!~aaq!aq mm.lv~aotas oum.ll!~t f s!sdoa!so!ssolvqs "ds m aVl!ln~l f t~Inlla~176 s!aU~ g6 (I66 [ uumuoH) Joqmnuoldmes

86E1 g f f P6E Off f6EIE6E

(uo!~'oJ uapJofjtuiq) x~!Ieaoq

(816I Pl!~Ofl) sJas

qsv

(f86I ~[filJnS ~ uoszopod) Aq&zag!lealsoql!q

~ "g=

i~

6E+ 9lt/fllAI dnalslt.S

~ "g=

~

=

8- O f

~

~

onbyu~ ~llaatsauad !!&aq!azl oHavu!~s snumsaom snas!pou?oso o !uollp[ snlnv!maH soft zan!Ivaao!.zs snuv!s.touz gn3s!polo!lg s!aaz?1 "A vaktaq!s O!lOaVd v!qnp o!ualosoz!zt~t valjm snlno!maH oa!uop m.uo~oumld~t S v.tafzuH~a nl!!Xgd vtvu.~va sldoato~uopo anb!~uV vllallsaua d snlolnguv snos!pOlOllsopnas d ~a!puvlmf ~aatltOlmldmo snoluvp snlnv!uzaH mvavoxav!aovu!as vudaa v!aavu!a~ Vllatuopo mn!aazpo!uoD suvgala snlno!maH smaadsns snas!poavln V !aallaou~ sna~!pou!a,VOD S!l!qva!zu m!soztoo~s suagund snlnm, uzaH vaafzalnav ~aatpoaakI smm.pva snas!pou!aso o smt&~176 snlnv!waH sv?~lo~a!tu mu!.tVllalg umldln.3*xa ran!loS v~outuzag slauoaldaa S 17 'ds s,xgdouwldals s.u,tn~ "S ja s~.xs

(I66I uuetuoH) aaqtunu aldme S

E

~ ~ I "6I gg g Of gE

aoqtuolAI lu[plopun)i

f L~ E

[ E E 6 ~E gg LI Ig L~

If I

E I I

6Z 6 17 0g

8I OL EI 6I EI

17 1717

gf~ ~6~ 0 f f f6~ ~6~

i

~

II-- OE-

9 17 ~ I [ I[ 8[ 9g 17E I~ IE

I

(816I PI.tffSr

saogeI qsv

(f86I ~Is ~ uasaopod) ,~qdea~gealsoqlIq

9~1~t~ dn.~lsI!S

8- 0 I - I I - OgaaqtuolAI lu!plopun)t

Table 2. Author citations for species of siliceous microfossils encountered during this study Diatoms Actinoptychus senarius (Ehrenberg) Ehrenberg, emend. Andrews & Abbott Aulacodiscus hirtus Barker & Meakin Aulacodiscus suspectus A. Schmidt Coscinodiscus ex gr. argus Ehrenberg sensu Homann Coscinodiscus morsianus (Grunow) Sims Coscinodiscus moelleri A. Schmidt Craspedodiscus moelleri A. Schmidt, emend. Homann Fenestrella antiqua (Grunow) Swatman Goniothecium odontella Ehrenberg oar. danica Grunow Hemiaulus cf. H. curvatulus Strel'nikova Hemiaulus danicus Grunow Hemiaulus elegans Heiberg Hemiaulus februatus Heiberg Hemiaulus hostilis Heiberg Hemiaulus kittonii Grunow Hemiaulus mitra Grunow, emend. Homann Hemiaulus muticus Strel'nikova Hemiaulus polymorphus Grunow Hemiaulus polymorphus Grunow oar. morsiana Grunow Hemiaulus proteus Heiberg Hemiaulus pungens Grunow, emend. Homann Hyalodiseus ex gr. laevis/subtilis sensu Grunow Odontella heibergii Grunow Odontotropis carinata Grunow Odontotropis cristata Grunow Odontotropis hyalina Witt, emend. Homann Omphalotheca jutlandica Grunow, emend. Homann Paralia siberica (A. Schmidt) Crawford & Sims Paralia siberica oar. laevis Crawford Pseudostictodiscus angulatus (Grunow) Grunow Pterotheca aculeifera (Grunow) Grunow, emend. Homann Rhaphoneis lancettula Grunow var. juetlandica Grunow, emend. Homann Rhizosolenia dubia (Grunow) Homann Rutilaria sp. sensu Homann Sceptroneis gemmata (Grunow) Van Heurck Solium exsculptum Heiberg Stellarima microtrias (Ehrenberg) Hasle & Sims Stephanogonia danica Grunow Stephanopyxis cf S. turris (Greville & Arnott) Ralfs Stephanopyxis sp. 4 sensu Homann Thalassiosiropsis wittiana (Pantocsek) Hasle Triceratium flos Ehrenberg Trinacria excavata Heiberg Trinacria heibergii Kitton, emend. Homann Trinacria pileolus (Ehrenberg) Grunow vat. jutlandica Grunow Trinacria regina (Heiberg) Homann Trinacria regina Heiberg var. tetragona Grunow Trochosira mirabilis Kitton Xanthiopyxis oblonga Ehrenberg Radiolaria Ceratocyrtis hystricosus (Jcrgensen) Peridium longispinum Jcrgensen Ebridian Falsebria ambigua Deflandre Silieoflagellates Corbisema naviculoidea (Frenguelli) Perch-Nielsen Corbisema hastata globulata Bukry Corbisema hastata hastata (Lemmermann) Bukry Corbisema glezerae Bukry Corbisema inermis crenulata Bukry Corbisema inermis inermis (Lemmermann) Bukry Corbisema inermis minor (Glezer) Bukry Dictyocha elongata Glezer Dic~ocha sp. aft. D. fibula Ehrenberg Dictyocha precanteris Bukry Naviculopsis constricta (Schulz) Frenguelli Naviculopsis danica Perch-Nielsen

262

A.G. MITLEHNER

Fig. 3. Graphic representation of selected diatom and radiolarian taxa in Table 1. See Fig. 2 for position of ash layers (x-axis). (a) Indicators of increasing open marine conditions. Note: Peridium longispinum is a radiolarian; other named species are diatoms. (b) Benthic diatom taxa, indicating shallow conditions for the lowermost sample studied.

calcareous concretions were of four types, each being dominated by specific diatom taxa. Fenner (1988) conducted a preliminary quantitative study on diatoms from the Harre Borehole (20 km south of the Limfjorden localities; Fig. 1) and found fluctuations in taxonomic abundances up through the section which are in broad agreement with those in the present study. However, silicoflagellates, ebridians and radiolaria were not included in her counts.

London Basin; northern Germany; The Netherlands; Belgium; Paris Basin. The diatom assemblages from the Fur Formation have similarities with pyritized assemblages recovered from onshore localities in NW Europe. Although poor preservation generally precludes accurate taxonomic identification, enough morphological features are

preserved to permit a confident correlation of these assemblages. The most diverse and best preserved of these assemblages is from the lower Eocene of northern Germany (Benda 1965), which has a very similar floristic content to the Fur Formation. A remarkable feature of the German assemblage is that, although it is almost entirely pyritized or phosphatized, the diatoms have not suffered secondary pyrite overgrowth (in contrast to offshore diatom material, see below), and are translucent to a degree which allows accurate identification. Pyritized diatoms found in a restricted interval at the base of the London Clay (now Harwich Formation) were the first to be identified as such (Shrubsole & Kitton 1881), and again were sufficiently well preserved for taxonomic identification in a number of cases. A century later, King (1981) confirmed the widespread occurrence of pyritized diatoms, mainly Fenestrella antiqua,

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN Coscinodiscus morsianus and Trinacria regina (= Coscinodiscus spp. 1 & 2 and Triceratium sp. l of Bettenstaedt et al. 1962), in a number of sections around the London Basin, suggesting a close correlation with the Corbisema naviculoidea assemblage in the Fur Formation. Other pyritized assemblages, again bearing similarities to the Fur Formation, have been recorded from the Netherlands (Ten Dam 1944), Belgium (King 1990) and the Paris Basin (Bignot 1983).

North Sea offshore diatom occurrences around the Paleocene-Eocene boundary. Diatoms occur sporadically throughout Paleogene sediments in the North Sea, often in consistently large numbers to form important biostratigraphic markers. They are the dominant microfossil group encountered in the volcaniclastic Sele and Balder Formations (King 1983; Mudge & Copestake 1992a, b; Knox & Holloway 1992). Despite poor preservation compared to their onshore equivalents, diatom assemblages recovered from exploration wells in the central North Sea (Fig. 4) have general similarities with those from the basin margins, in particular the predominance of Fenestrella antiqua (= Coscinodiscus sp. 1 of Bettenstaedt et al. 1962) and Coscinodiscus morsianus (= Coscinodiscus sp. 2 of Bettenstaedt et al. 1962). These and other Paleogene diatoms occurring in and around the North Sea Basin are the subject of a study by the present author (Mitlehner 1994), in which the taxonomic identity of a number of diatom morphotypes was established, cited previously in open nomenclature (King 1983). The identity of some of the more important marker species is given in Table 3. The formal description of these morphotypes is to be published separately (Mitlehner 1995). Figure 4 gives the distribution of prominent marker diatom taxa in two offshore wells studied by the present author; this information, together with that from Fur and other localities around the North Sea Basin, forms the basis for a siliceous microfossil correlation of upper Paleocene and lower Eocene diatomaceous intervals in and around the North Sea Basin (Table 4). Previous workers (Mudge & Copestake 1992a; Mudge & Bujak 1994) have documented a widespread diatom marker horizon (dominated by morphotypes referred to Coscinodiscus spp. 2, 4 & 7) for the top of the Sele Formation; of these, Coscinodiscus sp. 7 has been identified (via SEM study) as a resting spore of the species Fenestrella antiqua (Mitlehner 1994, 1995), and therefore correlates with the Corbisema naviculoidea assemblage in the Fur Formation. It is herein proposed that the first acme appearance of this species be regarded as a practical basinwide marker for the base of the Eocene (Table 4).

263

Diatomaceous sequences in other areas Localities in the former Soviet Union. Diatom assemblages from the Volga Basin (Glezer et al. 1974) show affinities with the assemblages from north Denmark (Fur) and the Norwegian Sea. This floral similarity suggests at least some connection across the Russian Platform during this period, probably in the form of a shallow, epicontinental sea (Barron 1987). Benda (1972) proposed a 'eurasiatic diatom-province' for the Paleogene of this region. Unfortunately, much of the Tertiary sequence is missing in the intervening area and in Fennoscandia, due to erosion during subsequent uplift and continental glaciation during the Pleistocene (Miller 1979). In addition, many of the diatom specimens from Russian localities housed in The Natural History Museum were collected in the last century when detailed information on locality and stratigraphic position was not collected. These deposits are consequently of negligible value for precise correlation (Ross & Sims 1985, pp. 279-282). Arctic Ocean. Biogenic silica accumulation occurred during at least three main phases during the Late Cretaceous and Paleogene in the Arctic Ocean (KitcheU & Clark 1982; Barron 1985). Recent data (Dell'Agnese & Clark 1994) suggest the close correlation of a rich siliceous interval (containing diatoms, silicoflagellates, chrysophycean cysts and ebridians) from the lower Eocene in the central Arctic Ocean, with that from the Fur Formation. This indicates at least an intermittent connection of the restricted North Sea Basin-Barents Sea with the Arctic Ocean during the early Eocene. Deep Sea Drilling Project~Ocean Drilling Program sites. Few continuous diatomaceous sections were recovered from oceanic Paleogene sediments (Fenner 1985), in marked contrast with the relatively complete Neogene record (Barron 1985). This is a reflection partly of age, consequent likelihood of dissolution and/or replacement, and also of the fact that there appear to have been fewer areas of upwelling (i.e. sites of diatom production, Barron 1993) during the Paleogene. These factors are partially offset by the robustness of Paleogene diatoms, so that where sections with preserved floras do exist they are a good representation of the original assemblage. Diatom species used in the biostratigraphic zonation schemes of the Deep Sea Drilling Project (DSDP) and its successor the Ocean Drilling Program (ODP) were mainly recovered from sequences in the open oceans, from the low and southern high latitudes, e.g. the Indian Ocean zonation of Fourtanier (1991) and the

MICROPALAEONTOLOGY

MICROFOSSIL MARKER SPECIES

Cenosphaera sp.

MAIN DIATOM EVENTS

- 6260

acme

6280 - 6300

-

Subbotina patagonica Fenestreila antiqua Stellarima microtrias

- 6340 -

Bettenstaedt et al. 1962

Fenestrella antiqua

Abundant

Craspedodlscus modleri 4a

i

L

=Coscinodiscussp.

4,

T h o m a s & G r a d s t e m 1981

6380

- 6400 -

t

6320

- 6360

Fenestrella antiqua =Coscinodiscus sp. 1,

4b

-

6420

- 6440 - 6450

Craspedodiscus moelleri Coscinodiscus morsianus - 6460 Stellarima microtrias Ttinacria regina i

a) 15/28a-3

MICROPALAEON'IDLOGY

@

~

r~

MARKERSPECIES

MAIN'DIATOM EVENTS .,

- 2050 <_

Subbotina patagonica

5

Stellarima microtrias

-2073

- 2104 Fenestrella antiqua ~Coscinodiscus sp.

4b "---I" - - -

-

-

-

w

- 2110 -

2116

-

2122

-

2128

Bettenstaedt et al. 1962

~

m .--.------.*t..-_.-

1,

Coscinodiscus morsianus Fenestrella antiqua Stellarima microtrias

m

m

m

9

,~

Craspedodiscus moelleri 4a

.---T------

~-Coscinodiscussp.

4,

T h o m a s & Gradstein 1981

Abundant

Craspedodiscus moelleri Coscinodiscus morsianus Stellarima microtrias Trinacria regina

b) 21/9-1

oy,o,i,hologics: Olivotoolivereycays,ono S,l o,ays,o o V ,e a,edsub0 ,lemods,ooe with sporadic pyrite

with volcanic glass shards

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN

265

Table 3. Taxonomic identity of selected diatom morphotypes from the Paleogene of the North Sea Basin Taxonomic status (this work)

Published reference (pyritized occurrences)

British Petroleum

Aulacodiscus suspectus

' Coscinodiscus

A. Schmidt Ref: Homann 1991, p. 36, pl. 5

sp. 9'

Coscinodiscus morsianus (Grunow)

Sims Ref: Sims 1989, p. 354, figs 8-15 Craspedodiscus moelleri (Schmidt)

Homann Ref: Homann 1991, p. 47, pl. 17, 1-5

SPT (formerly Robertson)

Stratigraphic range (North Sea and onshore)

Lista; Balder; Sele; Horda; Fur; Ieper

Coscinodiscus

Balder; Sele; Horda;

sp. 2. Bettenstaedt et al. 1962, pl. 52

Harwich; Fur; Olst; leper; U. Eozan 1

Coscinodiscus

' Coscinodiscus

Sele; Balder;

sp. 2. Thomas & Gradstein 1981, pl. 3.1, figs. 5, 6

sp. 77'

Harwich; leper ('Unit X'); Fur; 01st

Fenestrella antiqua

Coscinodiscus

Balder; Sele;

(Grunow) Swatman Ref: Sims 1990, p. 279, pls. 1 & 2

sp. 1. Bettenstaedt et al. 1962, pl. 52

Harwich; Fur; Olst; Ieper ('Unit X'); U, Eozan 1

F. antiqua (Gr.) Swat.

'Coscinodiscus barreliformis '

Balder; Sele

' Coscinodiscus

' Coscinodiscus

Balder; Sele

sp. 6; N6'

sp. 7'

Fur

'Coscinodiscus barreliformis '

auxospores Ref: Mitlehner 1994, p. 101, pl. 4, figs 1 & 2 E antiqua (Gr.) Swat. resting spores Ref: Mitlehner 1994, p. 103, pl. 4, figs 3-5 Stellarima microtrias

1. Coscinodiscus

' Coscinodiscus

"Coscinodiscus

Maureen; Lista;

Hasle & Sims Ref: Hasle & Sims 1986, p. 111

sp. B. Hughes 1981, pl. 3;

cf sp. 1'

cf sp.l'

Balder; Sele; Horda; Lark Fur; Olst; Harwich; Ieper ('Unit X)

2. Coscinodiscus

sp. 3. Thomas & Gradstein 1981, pl. 3.1, figs 7-12 Trinacria regina

(Heiberg) Homann Ref: Homann 1991, p.p. 124-5, pl. 50

Triceratium sp. 1 Ref: Bettenstaedt et al. 1962, pl. 52

Southern Ocean schemes of Mukhina (1976), Gombos (1977, 1984) and Fenner (1991). Sections from the northern high latitudes are discontinuous, restricted to certain parts of the Paleogene sequence, as in the Norwegian Sea (Dzinoridze et al. 1978). This makes inter-regional correlation difficult, and is a reflection of the tectonicallyrelated isolation of northern high-latitude basins

'Triceratium

Lista (top);

sp. 30'

Balder; Sele; Horda; Fur; Olst; Harwich

from the low-latitude Tethyan seaway (Haq 1981; Kitchell & Clark 1982; Rea et al. 1984). Other pyritized diatom assemblages: Eastern C a n a d i a n shelf; B e a u f o r t - M a c k e n z i e Basin.

Pyritized diatoms have also been recovered from hydrocarbon exploration wells offshore of Eastern Canada (Thomas & Gradstein 1981) and in the

Fig. 4. Micropalaeontology of the upper Paleocene to lower Eocene section in BP Wells (a) 15/28a-3 and (b) 21/9-1, central North Sea.

266

A.G. MITLEHNER

Table 4. Lower Paleogene stratigraphy of the central North Sea with equivalent units around the basin margins, including taxonomic status of main diatom marker species Age

Biozonation Silicoflagellates

Early Eocene

Lithostratigraphy

Diatoms and Foraminifera

Central North Sea

Southern England

Globigerina linaperta

Horda Formation

London Clay Formation

Naviculopsis aspera

ka

Corbisema naviculoidea

~a

Late Paleocene ;a

Naviculopsis danica

Denmark

Ieper Formation

Rosnaes Clay Formation

(9 Balder Formation

Fenestrella antiqua "~

Belgium

Craspedodiscus moeUeri Coscinodiscus morsianus Trinacria regina

Impoverished agglutinated foraminifera assemblage

Sele and Dornoch Formations

Harwich Formation

F

Maureen Formation

s t o r m a

'F o r

m

t

a t

o n

o n

i Landen Formation

Lista Formation Thanet (includingMey Formation SandstoneMember)

Bolivinopsis spectabilis

1

'Unit X'

Wooiwich and Reading Formations Upnor Formation

]o

u

i

Holmehus Formation

(Absent)

Kerteminde Formation

'Montian'

Danske-Kalke Formation

(Absent) Early Paleocene

Globigerina simplicissima

Ekofisk Formation

G. trivialis

Diatomaceous units shown stippled. Compiled from: Heilmann-Clausen(1985); Mudge & Copestake (1992a); Knox & Holloway ( 1992); Von Salis (1993); Ellison et al. (1994); Mitlehner (1994, 1995). (9, Knudshoved Member; (~), Silstrup Member; @, Knudeklint Member; @, Diverse diatom assemblage.

Beaufort-Mackenzie Basin, Arctic Canada (McNeil 1990). The former is closest to the North Sea Basin assemblage in terms of age and composition, whilst the latter is described as midEocene in age. Both these and unpublished reports of relatively rich pyritized diatom assemblages from the upper Paleocene-lower Eocene in the Barents Sea (Nagy et al. 1995) suggest that similar conditions prevailed in partially or fully restricted basinal settings in high northern latitudes during the Early Paleogene.

Discussion: North Sea Basin palaeoenvironments during the late Paleocene and early Eocene Palaeogeography

The main abundance peak of diatoms and silicoflagellates in the North Sea Basin (including the

Fur Formation) coincides with a particularly active episode in the opening of the North Atlantic (Roberts et al. 1984; Ziegler 1988), with widespread vulcanicity in northwest Britain, the Faeroe Islands and eastern Greenland (the Brito-Arctic Igneous Province of Boulter & Kvacek 1989); and the associated rapid uplift of the Scottish Highlands and the Shetland Platform. This led to the widespread deposition of volcanic ash and tuff in and around the North Sea, forming a prominent marker interval. This volcanic activity was comparatively explosive, as attested by the high proportion of volcanic glass present in the Balder Formation, a paradoxical feature as the ashes are largely basaltic in nature. Knox & Morton (1988) have suggested that this situation may have arisen from the reaction of the magma with seawater. The restricted nature of the North Sea Basin at this time (as evidenced by the complete absence of calcareous plankton and the widespread occurrence of pyrite and finely-

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN

laminated muds in the Balder Formation (Maim et al. 1984) lends support to this hypothesis. Palaeocirculation

Bonde (1979) suggested a circulation model for the North Sea Basin during the latest Paleocene, involving a prolonged period of upwelling along the coast of Norway, driven by prevailing northwesterly winds. This was held to account for both the increased siliceous productivity in the North Sea Basin, as well as for the accumulation of volcanic ash and tuff, present in the Fur and Olst Formations of Denmark, hundreds of kilometres from the nearest volcanic source (HeilmannClausen 1985). The abundance of silicoflagellates in the Fur Formation would seem to support the upwelling hypothesis, as intervals of very strong upwelling in modern systems are often initially accompanied by a dominance of silicoflagellates (Loeblich et al. 1968; Von Salis 1993), and quantitative analysis of samples for the present study certainly accords with this pattern (Table 1). The increasing predominance of both chain-forming diatoms (such as Trinacria, Hemiaulus and Solium) and diatom resting spore taxa (Goniothecium, Pterotheca and Stephanopyxis) through the same sample sequence (Fig. 3) suggests a coastal neritic upwelling environment (Beers et al. 1971). The absence of ebridians in samples above -20 may indicate increased warming (Loeblich et al. 1968).

Problems o f interpretation: upwelling in a 'greenhouse w o r M '

The Paleocene-Eocene boundary is known to have been a period of increased climatic warming, with enhanced humidity in high latitudes (Robert & Chamley 1991). Therefore, possible mechanisms for strong wind stress and resultant upwelling need to be considered for a warm, relatively isothermal setting dominated by low temperature gradients and an equable climate (with high levels of CO 2 leading to greenhouse warming, Kitchell & Clark 1982; Roberts et al. 1984; Rea et al. 1990). Miller et al. (1987) gave evidence for decreased wind strengths near the Paleocene-Eocene boundary, which would appear to discount the possibility of wind-driven upwelling. However, their data were from the open ocean, essentially isolated from the North Sea area during this period (Mudge & Copestake 1992a, b). Kitchell & Clark (1982) examined palaeogeographical and palaeoclimatic factors promoting polar upwelling during the Late Cretaceous and early Paleogene, and considered that a circulation pattern dominated by cyclonic conditions was

267

responsible for prolonged upwelling, with a semipermanent atmospheric low centred over the Alpha Ridge. Tectonic events were thought to have played a key role in the timing of transfers of global silica by the opening and closing of exchange routes for silica-rich deep waters. Similar, relatively stable, though more markedly, seasonal atmospheric conditions (Robert & Chamley 1991) occurred over the North Sea Basin during the early Paleogene. Currents necessary for upwelling were derived from seasonal monsoonal-type pressure systems, which became more intensified with the early Eocene transgression (Jolley & Spinner 1991) and global climatic warming (Robert & Chamley 1991). Bonde (1979) suggested that the pronounced laminations present within the Fur Formation reflected seasonal variations in sediment input, but subsequent detailed studies (Pedersen 1981; Pedersen & Surlyk 1983) showed that these laminations are not consistent through the deposit, and therefore favoured more sporadic and localized periods of upwelling. Recent work by Thunell et al. (1994) from the Gulf of California suggests a close relationship between the seasonal pattern of biogenic silica flux and phytoplankton biomass levels, and is controlled by a monsoon-like climate system which leads to marked seasonal changes in weather and hydrographic conditions. The gulf is a major silica sink, playing an important role in the global silica cycle; the Fur diatomite may have played a similar role in latest Paleocene to earliest Eocene times. Palynological investigations from the uppermost Paleocene of Mull, Scotland (Boulter & Kvacek 1989), and the lower Eocene basal London Clay (now Harwich Formation, see Table 3) by Jolley & Spinner (1991) point to the widespread occurrence of paratropical forests during this period with increased waterborne (rather than windborne) transport from a rapidly-eroding hinterland. This would account for the large amount of terrestriallyderived organic matter in sediments from the basin centre (Sele and Balder formations) during latest Paleocene times (Schr6der 1992), as well as at its eastern margin, in the Fur Formation (Larsson 1975). The presence of monsoonal conditions would further explain this situation.

Diatom ecology

Ethmodiscus: a possible modern analogue. Recent work by Villareal (1993) on the abundance of the giant diatom Ethmodiscus in the modern southwest Atlantic and central Pacific indicates that an abundance of large, solitary diatoms does not in itself indicate strong upwelling. This contradicts the earlier suggestion of Gardner & Burckle (1975)

268

A.G. MITLEHNER

that Ethmodiscus ooze was accumulated during intense upwelling associated with glacial periods; in which case the apparent dominance of the similarly-sized Fenestrella antiqua in North Sea sediments of late Paleocene and early Eocene age would not necessarily support the upwelling hypothesis. However, it must be stressed that this dominance is greatly distorted by the effects of differential preservation, as the pyritized assemblages common in the deeper parts of the basin are only a relict assemblage. The more delicate diatom species, which often possess spines, are not preserved here, although their original presence, initially in at least the centre of the North Sea Basin, is borne out by their presence in calcareous concretions in some areas (Malm et al. 1984).

Abundance peak of Fenestrella antiqua: a possible explanation. The great abundance of Fenestrella antiqua (= Coscinodiscus sp. 1 of Bettenstaedt et al. 1962) in the Sele and Balder Formations may in part be explained by its ability to form resting spores (= Coscinodiscus sp. 7 of BP), which enabled its survival during periods of nutrient depletion. Its large, convex, heavily silicified frustule may also have aided its preservation, providing a receptacle for pyrite infilling during the early stages of diagenesis. In the Fur Formation, F. antiqua is not a major assemblage constituent, although it forms blooms in some horizons, with both vegetative cells and resting spores cooccurring. It is probable that these blooms were often basinwide in scale, as a number of abundance peaks are known to occur which can be correlated closely in some areas (M. A. Charnock, pers. comm.). These blooms signify periods of high productivity leading to basin eutrophication, brought about by rapid influxes of both terrestrially-derived and upwelled nutrients such as iron, nitrates and phosphorus (Codispoti 1989). These nutrients are now known to be of greater importance in diatom productivity than the availability of silica, which was formerly held to account for the association of diatomites with volcanic deposits. It is now thought that this association is due to the generation of high pore-water silica concentrations by volcanic material, which favour diatom preservation (Fenner 1991). The abundance of fish and fish debris in coeval strata around the North Sea Basin (Bonde 1979) is a further indication of eutrophic conditions at this time. Basin stratification Several workers have commented on the water column stratification which the North Sea area underwent during the late Paleocene to early

Eocene, with an upper layer of less saline water (resulting from increased run-off from the rapidly uplifting Scottish land mass to the west, Mudge & Copestake 1992a, b) being separated from deeper, more saline and stagnant water by a pronounced halocline. Such an environment would favour plankton which were able to sink to deeper parts of the water column during pronounced periods of freshwater influx (such as monsoonal episodes), which would affect the upper water layers in the basin. Evidence from dinoflagellates suggests that upwelling did not occur throughout the North Sea Basin (J. Bujak, pers. comm.). This appears to conflict with the abundance of chain-forming diatoms throughout the North Sea Basin, in particular Trinacria regina (= Triceratium sp. 1 of Bettenstaedt et al. 1962), widely held to be indicative of upwelling conditions. This situation may be explained by water circulation patterns. In an already oxygen-depleted waterbody, the advected organic-rich anoxic water (including robust chain-forming diatoms) would extend farther from the area of intensified upwelling (i.e. northern Denmark) than would be the case in today's colder, relatively well-oxygenated marine waters, before it lost its identity due to mixing (Parrish & Curtis 1982). This would be partially offset by dissolution through the water column beneath the relatively low salinity upper water layer, so that only the most robust diatom species would be preserved and consequently pyritized. These conditions, in conjunction with differing environments of diagenesis and subsequent burial, may account for the lower diversity of diatoms preserved in the sediments of the North Sea proper from this period, in comparison with the rich, wellpreserved floras found in the Fur Formation.

Summary: the North Sea Basin during the latest Paleocene A model for circulation in the North Sea Basin during the latest Paleocene is suggested, modified from an earlier scenario proposed by Bonde (1979). A representative cross-section through the North Sea Basin, from SE England to Denmark, is presented in Fig. 5. To the southwest, increased freshwater run-off may have contributed to lowered salinity in the upper part of the water column, whilst the Coriolis effect helped to 'bank-up' water at this end of the basin. Downwelling of denser, more saline water in this part of the basin created further accommodation space for the less saline water near the surface. Upwelling of cool, nutrientrich water in the area of present-day Denmark

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN

269

Fig. 5. Cross-section through the North Sea Basin, latest Paleocene, showing tentative circulation regime (modified after the model of Bonde 1979). Note that the vertical scale is schematic: water depths in the Central Graben were probably of the order of 300 m, while the abundance of neritic benthic diatoms in the Knudeklint Member of the Fur Formation suggests a water depth not exceeding 50 m in the area of northern Denmark.

contributed to enhanced productivity in the surface waters which, together with terrestrially-derived nutrients (nitrogen, phosphorus, silica) promoted the blooming of diatoms. Nutrient-poor water sank beneath this organic-rich surface layer, and a lack of circulation led to anoxic bottom conditions (Pedersen 1981). Few benthic organisms could survive, apart from a depauperate assemblage of low pH-tolerant agglutinating foraminifera (zone NSB 2 of King 1983). Sedimentation was slow, predominantly clastic in nature, and characterized by laminated muds and pyrite (the latter replaced and infilled diatom frustules soon after they were deposited, and again during diagenesis). Both nonmarine and submarine fans were deposited in some areas (e.g. the Beauly, Dornoch, Forties, Forth and Gryphon Sands), which later became the sites of hydrocarbon reservoirs (Glennie 1990; Bain 1993). Moving across the basin, the RingkCbing-Fyn High formed a sill, partially isolating northern Denmark from the rest of the North Sea Basin. This further restricted circulation which, coupled with more nutrient-rich surface waters in the area of northern Denmark, led to the proliferation of diatoms, silicoflagellates and ebridians. Highly stagnant bottom conditions, with little or no benthos to bioturbate and mix the sediments, contributed to the accumulation and preservation of laminated diatomite. This situation changed during the earliest Eocene, as marine waters began to influence the basin at the beginning of a major transgressive phase with eustatic sea-level rise.

Conclusions In summary, this study has shown that siliceous microfossils document the unusual palaeoenvironmental conditions which existed during the upper Paleocene to lower Eocene interval in the North Sea Basin. The Fur Formation of Denmark can be regarded as a microcosm of the North Sea proper during this period, having been deposited in a highly restricted sub-basin within the essentially restricted North Sea. The circulation changes outlined provide insight into water chemistry and circulation processes during the warm Early Cenozoic, before Antarctic ice build-up heralded the 'ice-house' conditions (Baldauf & Barron 1990; Dell'Agnese & Clark 1994) which dominate water circulation patterns in today's oceans. Despite wide variations in preservation across the North Sea Basin, limited correlation via the use of siliceous microfossils is possible. A problem besetting the accuracy of biostratigraphic determination of the Paleocene-Eocene boundary in North Sea offshore sections, has been the high incidence of microfossil cavings; further detailed studies, on cored sections across the PaleoceneEocene boundary from a number of sites in and around the North Sea, would greatly help to refine the siliceous microfossil stratigraphy of this otherwise poorly-defined interval. This work was carried out as part of a PhD research project in the Micropalaeontology Unit of the Research School of Geological and Geophysical Sciences,

270

A.G. MITLEHNER

University College London, during the tenure of a Natural Environment Research Council CASE award with British Petroleum plc, both of whom are gratefully acknowledged for financial assistance. M. Homann (Bundesanstalt ftir Geowissenschaften und Rohstoffe, Hannover) gave permission to use her samples for further analysis. Invaluable guidance throughout this work was provided by A. R. Lord (UCL), P. A. Sims (The Natural History Museum, London) and R. W. Battarbee (Environmental Change Research Centre, UCL). Helpful comments were received from J. Fenner (Bundesanstalt fiir Geowissenschaften und Rohstoffe, Hannover), C. Heilmann-Clausen

(Aarhus Universitet) and G. Stenstrop (Fur Museum, Denmark). I also wish to acknowledge the following people for many discussions and comments: friends and colleagues in the Department of Geological Sciences, UCL, especially C. Kavouras, L. McCarthy and P. Chambers; J. Hinchey (Oxford University), K. Kennington (University of East Anglia). Technical assistance was provided by J. Davy and T. Stiles (Micropalaeontology Unit, UCL), and K. Childs (The Natural History Museum, London). I would also like to thank my family and friends for their support and encouragement throughout this work.

References ANDREWS, G. W. 1972. Some fallacies of quantitative diatom paleontology. Nova Hedwigia, Beiheft, 39, 2285-2295. BAIN, J. S. 1993. Historical overview of exploration of Tertiary plays in the UK North Sea. In: PARKER, J. R.(ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 5-13. BALDAUF, J. G. & BARRON, J. A. 1990. Evolution of biosiliceous sedimentation patterns - Eocene through Quaternary: paleoceanographic response to polar cooling. In: BLEIL, U. & Theide, J. (eds) Geological History of the Polar Oceans: Arctic Versus Antarctic. Kluwer, Amsterdam, 575-607. BARRON, J. A. 1985. Miocene to Holocene planktic diatoms. In: BOLLI, H. M., SAUNDERS, J. B. & PERCH-NIELSEN, K. (eds) Plankton Stratigraphy. Cambridge University Press, 763-810. 1987. Diatomite: Environmental and geologic factors affecting its distribution. In: HEIN, J. R. (ed.) Siliceous Sedimentary Rock-hosted Ores and Petroleum. Van Nostrand Reinhold, New York, 164-178. 1993. Diatoms. In: LIPPS, J. H. (ed.) Fossil Prokaryotes and Protists. 155-167, Blackwell, Boston, 155-167. BATTARBEE,R. W. 1986. Diatom analysis. In: BERGLUND, B. E. (ed.) Handbook of Holocene Paleoecology and Paleohydrology. Wiley & Sons, Chichester, 527-570. BENDA, L. 1965. Diatomeen aus dem Eoz~in Norddeutschlands. Palgiontologisches Zeitschrift, 39 (3/4), 165-187. 1972. The Diatoms of the Moler Formation of Denmark (Lower Eocene). A preliminary report. Nova Hedwigia, Beiheft, 39, 251-266. BEERS, J. R., STEVENSON, M. R., EPPLEY, R. W. & BROOKS, E. R. 1971. Plankton populations and upwelling off the coast of Peru, June 1969. Fishery Bulletin, 69, 859-876. BERNER, R. A. 1968. Calcium carbonate concretions formed by decomposition of organic matter. Science, 159, 195-197. BETTENSTAEDT, F., FAHRION, H., HILTERMANN, H., t~z WICK, W. 1962. Tertiar Norddeutschlands. In: SIMON, W. & BARTENSTEIN, H. (eds) Arbeitskreis deutscher Mikropali~'ontologen: Leitfossilien der Mikropalaontologie. Gebruder Borntraeger, Berlin, 2 vols.

BIGNOT, G. 1983. Les Lagunes Sparnaciennes: Une 6tape dans la conqu&e des eaux douces par les diatom~es. Revue de Micropalgontologie, 26 (1), 15-21. BJORKLUND, K. R. 1976. Radiolaria from the Norwegian Sea, Leg 38 of the Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 38, 1101-1168. BLOME, C. D. 1984. Middle Jurassic (Callovian) radiolarians from carbonate concretions, Alaska and Oregon. Micropaleontology, 30 (4), 343-389. BOGCLrD, O. B. 1918. Den vulkanske Aske i Moleret samt en Oversigt over Danmarks ~eldre Terti~erbjaergarter. Danmarks Geologiske UndersCgelse, Rcekke 2, 33. BONDE, N. 1979. Palaeoenvironment in the "North Sea" as indicated by the fish-bearing Mo-Clay deposit (Palaeocene/Eocene), Denmark. Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie, 16 (1), 3-16. BOULTER,M. C. & KVACEK,Z. 1989. The Palaeocene flora of the Isle of Mull (incorporating unpublished observations by A. C. Seward & W. N. Edwards). Special Papers in Palaeontology, 42, 1-149. CALVERT, S. E. 1974. Deposition and Diagenesis of Silica in Marine Sediments. Special Publications of the International Association of Sedimentology, 1, 273-299. CODISPOT1,L. A. 1989. Phosphorus vs. nitrogen limitation of new and export production. In: BERGER, W. H., SMETACEK, V. S. & WEFER (eds) Productivity of the Ocean: Present and Past. Wiley & Sons, Chichester, 377-394. DELL'AGNESE, D. J. & CLARK, D. L. 1994. Siliceous microfossils from the warm late Cretaceous and early Cenozoic Arctic Ocean. Journal of Paleontology, 68 (1), 31-47. DZINORIDZE, R. N., JOUSE, A. P., KOROLEVA-GOLIKOVA, G. S., KOSLOVA, G. E., NAGAEVA, G. S., PETRUSHEVSKAYa, M. G. & STREL'NIKOVA,N. I. 1978. Diatom and radiolarian Cenozoic stratigraphy, Norwegian Basin: DSDP Leg 38. Initial Reports of the Deep Sea Drilling Project, Supplement, 38-41, 289-427. ELLISON, R. A., KNOX, R. W. O'B., JOLLEY, D. W. & KING, C. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists'Association, 105, 187-197. FENNER, J. M. 1985. Late Cretaceous to Oligocene planktic diatoms. In: BOLLI, H. M., SAUNDERS,J. B.

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN & PERCH-NIELSEN, K. (eds) Plankton Stratigraphy, Volume 2. Cambridge University Press, 713-762. 1988. Occurrences of pre-Quaternary diatoms in Scandinavia reconsidered. Meyniana, 40, 133-141. 1991. Taxonomy, stratigraphy and paleoceanographic implications of Paleocene diatoms. In: CIESIELSKI, P. F., KRISTOFFERSEN, Y. E T AL. Proceedings of the Ocean Drilling Program, Scientific Results, 114, 123-154. 1994. Diatoms of the Fur Formation from Harre drillhole, and their interpretation (abstract). In: PEDERSEN, S. (ed.) Geology of the Mo-Clay. Abstracts and excursion guide for the Mo-clay symposium on Mors, June 13-15 1994. Danmarks Geologiske UndersCgelse Service Report, 34. FOURTANIER,E. 1991. Paleocene and Eocene diatom biostratigraphy and taxonomy of eastern Indian Ocean site 752. In: WE~SSEL, J., PIERCE, J., TAYLOR, E., ALT, J. eT AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 121, 171-187. GALLAGIaER, L. T. 1990. Calcareous nannofossil biozonation of the Tertiary of the North Sea Basin. Newsletters on Stratigraphy, 22, 21-41. GARDNER,J.V. & BURCKLE,L. H. 1975. Upper Pleistocene Ethmodiscus rex oozes from the eastern equatorial Atlantic. Micropaleontology, 21, 236-242. GLENNIE, K. W. 1990. Introduction to the Petroleum Geology of the North Sea 3rd Edition. Blackwell, Oxford. GLEZER, Z. I., JOUSE, A. P., MAKAROVA,I. P., PROSHKINALAVRENKO,A. I. & SHESHUKOVA-PORETSKAYA,g. S. 1974. The Diatoms of the USSR Fossil and Recent. Vol.1. Akademii Nauk SSSR, Botanical Institute, Leningrad (in Russian). GOMBOS, A. M. Jr 1977. Palaeogene and Neogene diatoms from the Falkland Plateau and Malvinas Outer Basin: Leg 36, Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 36, 575-687. 1984. Late Paleocene Diatoms in the Cape Basin. Initial Reports of the Deep Sea Drilling Project, 73, 495-511. GRUNOW, A. 1866. Moler aus Jtitland, eingesandt von Th. Jensen, analysirt von A. Grunow. Hedwigia, 5 (10), 145-146. HASLE, G. R. & SIMS, P. A. 1986. The diatom genera Stellarima and Symbolophora with comments on the genus Actinoptychus. British Phycological Journal, 21, 97-114. HAQ, B. U. 1981. Palaeogene paleoceanography - Early Cenozoic oceans revisited. Oceanologica Acta, 4, 71-82. HARWOOD, D. M. 1988. Upper Cretaceous and lower Paleocene diatom and silicoflagellate biostratigraphy of Seymour Island, eastern Antarctic Peninsula. In: FELDMAN,R. & WOODBURNE,D. (eds) Geology and Paleontology of Seymour Island. Geological Society of America Memoir, 169, 55-129. HEmERG, P. A. C. 1863. Conspectus Criticus Diatomacearum Danicarum: Kritisk Oversigt over de Danske Diatomeer. William Priors Forlag, Copenhagen. HEILMANN-CLAUSEN, C. 1985. Lithostratigraphy and -

-

-

-

-

-

271

depositional environments in the Upper Paleocene and Eocene of Denmark. Bulletin of the Geological Society of Denmark, 33, 287-323. 1994. Depositional history of the PalaeoceneEocene transition in Denmark (abstract). In: PEDERSEN, S. (ed.) Geology of the Mo-Clay. Abstracts and excursion guide for the Mo-clay symposium on Mors, June 13-15 1994. Danmarks Geologiske Unders~gelse Service Report, 34. HOMANN, M. 1991. Die Diatomeen der Fur-Formation (Alt-tertiiir) aus dem Limfjord-Gebeit, Nordjutland/ D~inemark. Geologisches Jahrbuch, Reihe A, 123. Hu~nEs, M. J. 1981. Contribution on Oligocene and Eocene microfossils from the southern North Sea. In: NEALE, J. W. & BRASmR, M. D. (eds) Microfossils from Fossil and Recent Shelf Seas. Ellis Horwood, Chichester, 186-294. JACQUI~, M. & THOUVENIN,J. 1975. Lower Tertiary tufts and volcanic activity in the North Sea. In: WOODLAND, A. W. (ed.) Petroleum and the Continental Shelf of Northwest Europe, Vol. 1: Geology. Applied Science Publishers, Barking, 455--465. JOLLEY, D. W. & SP1NNER, E. 1991. Spore-pollen associations from the lower London Clay (Eocene), East Anglia, England. Tertiary Research, 13 (1), 1 1 - 2 5 .

KAKUWA, Y. 1984. Preservation of siliceous skeletons in siliceous rocks. Scientific Papers of the College of Arts and Sciences, University of Tokyo, 34, 43-61. KING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6, Rotterdam. 1983. Cainozoic Micropalaeontological Biostratigraphy of the North Sea. Institute of Geological Sciences Report 82/7, HMSO, London. 1989. Cenozoic of the North Sea. In: JENKINS, D. G. & MURRAY, J. W. (eds) A Stratigraphical Atlas of Fossil Foraminifera 2nd Edition. Ellis Horwood, Chichester, 418-489. 1990. Eocene stratigraphy of the Knokke borehole (Belgium). In: LAGA, P. & VANDENBERGHE,N. (eds) The Knokke Well (11E/138) with a Description of the Den Haan (22W/276) and Oostduinkerke (35E/142) Wells. Toelichten, Verhandelingen Geologische en Mijnkaarten van Belgie, 29, 67-102. KITCHELL, J. A., & CLARK, D. L. 1982. Late CretaceousPaleogene paleogeography and paleocirculation: evidence of North Polar upwelling. Palaeogeography, Palaeoclimatology, Palaeoecology, 40, 135-165. & GOMBOS, A. M. Jr 1986. Biological selectivity of extinction: a link between background and mass extinction. Palaios, 1, 504-511. KITTON, E 1871. Diatomaceous deposits from Jutland. Journal of the Quekett Microscopical Club, 2, 99-102, 168-171. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of northwest Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society of London, 141, 993-999. & HOLLOWAY,S. 1992. 1. Paleogene of the Central -

-

-

-

-

-

272

A.G. MITLEHNER and Northern North Sea. In: KNOX, R. W. O'B. & Cordey, W.G. (eds) Lithostratigraphic Nomenclature of the UK North Sea. British Geological Survey, Nottingham. & MORTON, A. C. 1988. The record of early Tertiary volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early

--

Tertiary Volcanism and the Opening of the North Atlantic. Geological Society, London, Special Publication, 39, 407-419. LARSSON, S. G. 1975. Palaeobiology and mode of burial of the insects of the lower Eocene Mo-clay of Denmark, Bulletin of the Geological Society of Denmark. 26, 175-184. LOEBLICH, A. R. III, LOEBLICH, L. A., TAPPAN, H. & LOEBLICH, A. R. Jr 1968. Annotated Index of Fossil

and Recent Silicoflagellates and Ebridians with Descriptions and Illustrations of Validly Proposed Taxa. Geological Society of America Memoir, 106. MALM, O. A., CHRISTENSEN,O. B., FURNES, H., LOVLIE, R., RUSELATI'EN, H. 8,: ~STBY, K. L. 1984. The Lower Tertiary Balder Formation: an organogenic and tuffaceous deposit in the North Sea region. In: SPENCER, A. M. ET"AL. (eds) Petroleum Geology of the North European Margin. Graham & Trotman, London, 149-170. MARTINI, E. 1974. Silicoflagellate zones in the Eocene and early Oligocene. Senckenbergiana Lethaia, 54, 527-532. MCNEIL, D.H. 1990. Stratigraphy and paleoecology of the Eocene Stellarima assemblage zone (pyrite diatom steinkerns) in the Beaufort-Mackenzie Basin, Arctic Canada. Bulletin of Canadian Petroleum Geology, 38, 17-27. MILLER, K. G., JANACEK,T. R., KATZ, M. E. & KEIL, D. K. 1987. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary. Paleoceanography, 2, 741-761. MILLER, U. 1979. Reworked lower Tertiary/upper Cretaceous diatoms in the Pleistocene deposits of Scania, southern Sweden, and their stratigraphic application. Nova Hedwigia, Beiheft, 64, 473-488. MrrLEHNER, A. G. 1994. The occurrence and preservation

of diatoms in the Palaeogene of the North Sea Basin. PhD Thesis, University of London. 1995. The taxonomy and biostratigraphy of pyritised diatoms in the Palaeogene of the North Sea basin. Special Papers in Palaeontology, in press. MUDGE, D. C. & BUJAK, J. P. 1994. Eocene stratigraphy of the North Sea basin. Marine and Petroleum Geology, 11 (2), 166-189. & COPESTAg.E,E 1992a. Revised lower Palaeogene lithostratigraphy for the outer Moray Firth, North Sea. Marine and Petroleum Geology, 9, 53-69. & -1992b. Lower Palaeogene stratigraphy of the northern North Sea. Marine and Petroleum Geology, 9, 287-301. MUKHINA, g. 1976. Species composition of the late Paleocene diatoms and silicoflagellates in the Indian Ocean. Micropaleontology, 22 (2), 151-158. NAGY, J., KAMINSKI,M. A., JOHNSEN, K. & MITLEHNER, A. G. 1995. Foraminiferal, palynomorph and diatom stratigraphy and paleoenvironments of the

-

-

-

-

-

-

Paleogene Torsk Formation: A reference section for the western Barents Sea. Palaeogeography, Palaeoclimatology, Palaeoecology, in press. PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene benthic foraminiferal isotopes and assemblages: implications for deepwater circulation. Paleoceanography, 7 (4), 405-422. PARRISH, J. T. & CURTIS, R. L. 1982. Atmospheric circulation, upwelling, and organic-rich rocks in the Mesozoic and Cenozoic eras. Palaeogeography, Palaeoclimatology, Palaeoecology, 40, 31-66. PEDERSEN, G. K. 1981. Anoxic events during sedimentation of a Paleogene diatomite in Denmark. Sedimentology, 28,487-504. & SURLYK, E 1983. The Fur Formation, a late Paleocene ash-bearing diatomite from northern Denmark. Bulletin of the Geological Society of Denmark, 32, 43-65. PERCH-NIELSEN, K. 1976. New silicoflagellates and a silicoflagellate zonation in north European Paleocene and Eocene diatomites. Bulletin of the Geological Society of Denmark, 25 (1/2), 27-40. 1985. Silicoflagellates. In: BOLLI, H. M., SAUNDERS, J. B. & PERCH-NIELSEN, K. (eds) Plankton Stratigraphy, Volume 2. Cambridge University Press, Cambridge, 811-846. REA, D. K., ZACHOS, J. C., OWEN, R. M. & GINGERLICH, E D. 1990. Global change at the Paleocene-Eocene boundary: climatic and evolutionary consequences of tectonic events. Palaeogeography, Palaeoclimatology, Palaeoecology, 79, 117-128. ROBERT, C. & CHAMLEY, n. 1991. Development of early Eocene warm climates, as inferred from clay mineral variations in oceanic sediments. -

-

Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change), 89, 315-331. ROBERTS, D. G., MORTON, A. C. & BACKMAN, J. 1984. Late Paleocene-Eocene volcanic events in the northern North Atlantic Ocean. Initial Reports of the Deep Sea Drilling Project, 81, 913-923. Ross, R. & SIMS, P. A. 1985. Some genera of the Biddulphiaceae (diatoms) with interlocking linking spines. Bulletin of the British Museum, Natural History (Botany), 13 (3), HMSO London, 277-381. ROUND, E E., CRAWFORD,R. M., & MANN, D. G. 1990.

The Diatoms. Biology and Morphology of the Genera. Cambridge University Press, Cambridge. SCHRADER,H. 1971. Faecal pellets: Role in sedimentation of pelagic diatoms. Science, 174, 55-57. -• SORKNES,R. 1991. Spatial and temporal variation of Peruvian coastal upwelling during the latest Quaternary. In: SUESS, E., VON HEUNE, R. ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 112, 391-406. SCHRODER, Z. 1992. A palynological zonation for the Palaeocene of the North Sea Basin. Journal of Micropalaeontology, 11 (2), 113-126. SHRUBSOLE,W. H. & KITTON, E 1881. The diatoms of the London Clay. With a list of species, and remarks. Journal of the Royal Microscopical Society, 11, 381-387. SCHULZ, E 1927. Diatomeen BUS norddeutschen

PALEOCENE--EOCENE SILICEOUS MICROFOSSILS, NORTH SEA BASIN Basalttuffen und Tuffgescheiben. Zeitung Gescheibeforschung, 3, 66-78. SIMS, P. A. 1989. Some Cretaceous and Palaeocene species of Coscinodiscus: a micromorphological and systematic Study. Diatom Research, 4 (2), 351-371. 1990. The fossil diatom genus FenestrelIa, its morphology, systematics and palaeogeography. Nova Hedwigia, Beiheft, 100, 277-288. STAESCHE, K. & HILTERMANN,H. 1940. Mikrofaunen aus dem Tertitir Nordwest-deutschlands. Abhandlungen Reichsanstalt Bodenforschung, Berlin, 201. TEN DAM, A. 1944. Die stratigraphische Gliederung des niederlandischen Palaeozans und Eozans nach Foraminiferen. Mededelingen van de Geologische Stichting, C, 3. THOMAS, E C. & GRADSTEIN,E M. 1981. Tertiary subsurface correlations using pyritised diatoms, off-

273

shore eastern Canada. In: Current Research, Part B, Geological Survey of Canada, 81-1 B, 17-23. THUNELL, R. C., PRIDE, C. J., TAPPa, E. & MULLERKARGER, E E. 1994. Biogenic silica fluxes and accumulation rates in the Gulf of California. Geology, 22, 303-306. VAN HEURCK, H. 1896. A Treatise on the Diatomaceae. London. VoN SALIS, K. 1993. First Oligocene silicoflagellates from N. Europe (Silstrup, Denmark). Zitteliana, 20, 79-86. VILLA~AL, T. A. 1993. Abundance of the giant diatom Ethmodiscus in the southwest Atlantic Ocean and central Pacific gyre. Diatom Research, 8 (1), 171-177. ZIEGLER, P. A. 1988. Evolution of the Arctic - North Atlantic and the western Tethys. American Association of Petroleum Geologists Memoir, 43.

Stable isotope and biotic evolution in the North Sea during the early Eocene: the Alb~ek Hoved section, Denmark B. SCHMITZ 1, C. H E I L M A N N - C L A U S E N 2, C. KING 3, E. STEURBAUT 4, F. P. A N D R E A S S O N l, R. M. CORFIELD 5 & J. E. CARTLIDGE 5 1 Department of Marine Geology, Earth Sciences Centre, University of G6teborg, S-41381 Giiteborg, Sweden 2 Department of Geology, Aarhus University, C.E Mr All~ Bygn. 120, DK 8000 Aarhus C, Denmark 3 41 Montem Road, New Malden, Surrey KT3 3QU, UK 4 Royal Belgian Institute of Natural Sciences, rue Vautier 29, B-1040 Bruxelles, Belgium 5 Department of Earth Sciences, University of Oxford, Parks Road, Oxford OX1 3PR, UK Abstract: Stable isotope (813C, 8180) and biostratigraphic data are presented for a 20 m thick section spanning the R0sn~es Clay Formation at Alba~k Hoved in Denmark. This early Eocene formation is the first calcareous deposit in the North Sea after a several million year period of non-calcareous sedimentation. Nannofossil and dinoflagellate data indicate that the section is unusually complete, spanning lower zone NPll to lower NP13. Throughout the section, Subbotina spp. dominate the planktonic foraminiferal assemblages. Benthonic foraminiferal assemblages indicate middle bathyal water depths (600-1000 m). Water exchange between the semi-enclosed North Sea and the open ocean has been estimated by comparing the North Sea isotopic records with coeval records for DSDP Hole 550 in the northeastern Atlantic Ocean. Anomalously low 8180 values (-4 to -5%& for bulk samples and planktonic deep-dwelling Subbotina from the R0sn~es Clay indicate a significant freshwater component in the North Sea. Average salinities in the euphotic zone ranged between 26 and 30 ppt throughout the early Eocene. The benthonic foraminiferal 8180 values indicate generally somewhat more saline and stable conditions in the water mass near the seafloor. During the early Eocene, three principal conditions altemated in the North Sea, depending on regional sea level, position of critical sills and the extent of water exchange with the open ocean. (1) At times of strongly restricted water exchange, calcite dissolution was complete. Grey clays formed and sediment oxygen content was low. Non-calcareous agglutinated foraminifera dominated. This condition prevailed in the earliest Eocene (NP10), during deposition of the lowermost R#sn~es Clay Formation, and at the end of the early Eocene. (2) At times of moderately restricted water exchange, calcite dissolution was important. Calcite content and planktonic/benthonic foraminiferal ratios in the sediment were low. Different grey or reddish brown clays formed. Oxygen content at the seafloor was low to intermediate. Subbotina 8180 values (-2 to -4%0) were generally a few per mil lower than in the coeval open ocean and fluctuated dramatically, due to freshwater admixture. Subbotina-benthonic A813C gradients were high, because of low biological productivity and slow renewal of bottom water in connection with temporarily strongly density-stratified water masses. This condition prevailed during the later half of Biochron NP11, and possibly in latest NP12. (3) At times of more open water exchange, calcite-rich sediments dominated. Deep-dwelling planktonic foraminifera invaded the North Sea and planktonic/benthonic foraminifera ratios were high. Surface-thriving morozovellids, however, were absent, probably because of reduced surface salinities. Subbotina 8180 values (-2 to -3%0) indicate that a freshwater component was present at mid-depth, but salinities were higher and more stable than during more restricted water exchange. Subbotina-benthonic A~13C gradients were low, reflecting higher productivity and invigoration of bottom-water circulation. Water mass density stratification was less profound. Reddish brown marls dominated, and oxygen content in the sediment was moderately high. This condition prevailed during most of NP 12 and probably during early NP 11. High-resolution isotopic profiles over a 5 m thick interval in the upper NP12 part of the R0sn~es Clay Formation reveal that three distinct lithological 'event beds' are associated with profound short-term negative shifts (1-2%0 in 813C and 8180. The isotopic shifts are of the same magnitude in bottom waters as at mid-depth, implying that they reflect rapid changes in the chemistry of the entire water mass of the North Sea. The events may reflect short-term sea-level falls and/or rapid water mass exchange with other semi-enclosed basins to the north.

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 275-306.

275

276

B. SCHMITZ E T AL.

The evolution of the North Sea Basin from its formation in the Permian to the present has been determined primarily by vertical crustal movements related to the development of the North AtlanticArctic rift system (Ziegler 1988; King 1989). A crucial period in its history was the late Paleocene and early Eocene, when rifting of the northeastern Atlantic commenced. This was a violent event with possibly major global environmental consequences (Roberts et al. 1984; White 1988; Eldholm 1990; Rea et al. 1990; Eldholm & Thomas 1993). During an estimated 2-3 million years period in connection with this rifting, the so-called Thulean volcanism generated a total volume of extrusive basalts of 1 - 2 x 106km 3 (White 1988), which is about the same volume of basalts as produced by the Deccan Traps volcanism at the Cretaceous-Tertiary transition. Palaeoreconstructions suggest that at this

time the North Sea was a semi-enclosed sea with restricted contact with the Atlantic (Fig. 1; Ziegler 1988). Water exchange with the Atlantic may have occurred primarily across the Shetland-Faeroe Channel, but also across the English Channel (Berggren & Schnitker 1983; King 1989, 1993). The North Sea stood in open contact with the Norwegian-Greenland Sea to the north and, during some periods, there may also have been an indirect connection with the Tethys across north Germany and Poland (King 1993). The extent of water exchange with the open ocean must have varied considerably with time, depending on sea level and the development of regional geography. During periods of regional low sea level and elevated topography, as during the latest Paleocene, the North Sea was almost completely isolated (Knox et al. 1981), and inhabited by an impoverished

Fig. 1. The northeastern Atlantic in the late Paleocene-earliest Eocene. After Ziegler (1988).

277

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

fauna and flora. In the early Eocene, following the opening of the northeastern Atlantic, sea level rose, leading to increased water exchange with the world ocean and an influx of a more diverse marine fauna and flora. A remarkable feature of upper Paleocene-lower Eocene sediments in the North Sea and adjacent areas is the widespread ash layers that resulted from unique explosive basaltic volcanism, probably at the onset of separation of Greenland from Europe (Knox & Morton 1988; Morton & Knox 1990). More than 200, predominantly tholeiitic ash layers, with provenance in the Faeroe-Greenland area, were spread over the North Sea, Denmark, England, offshore mid-Norway and as far south as the Bay of Biscay, covering a total area of c. 6 x 106 km 2 (Knox & Morton 1988). It has been suggested that the unusually warm climate in the early Eocene was triggered by greenhouse gases that derived from the Thulean and other coeval volcanism (Owen & Rea 1985; Rea e t al. 1990; Eldholm & Thomas 1993). The sediments that formed in the North Sea during the latest Paleocene and earliest Eocene mostly lack biogenic calcite, making global correlations through biostratigraphy or stable isotopes difficult. However, in the early Eocene (early Biochron NP11), after several million years of non-

Ma

Age Middle Eocene

Danish dinoflagellate zones

NP zones 14

D.

North Sea

pachydermum

Late P a l e o c e n e to early E o c e n e s e d i m e n t s in D e n m a r k The explosive episode of Thulean volcanism has been subdivided into four phases, 2a-2d, of which 2a and 2b represent the major phases of very frequent ash ejection (Knox & Morton 1988). The Resnms Clay Formation formed during phases 2c and 2d, when volcanic activity had decreased substantially (Fig. 2). In total, about 20 ash layers are found within the ROsnms Clay Formation. This formation overlies the extremely ash-rich sediments of the major volcanic phases, represented by

A.

diktyoplokus

numbers

C.

coleothrypta

Horda land Group

12"

Lithostratiflraphy Denmark NNW (Fur)

Ash layers

of B~ggild

13"

Early Eocene

Volcanic phase

calcareous sedimentation, the calcareous RCsnms Clay Formation began to form. Recently, a c. 20 m thick, unusually complete, section spanning this formation has been described from Albmk Hoved in Denmark (Heilmann-Clausen 1990). We present stable-isotope data, together with biostratigraphic information, from this section, and discuss the early Eocene evolution of the North Sea shortly after the opening of the northeastern Atlantic. We also compare the North Sea isotopic data with coeval data from the North Atlantic DSDP Hole 550, in order to determine the extent of water exchange between the two regions.

(1918)

=

~

itlebaelt

~ <

IIIIIV

V

(part) D.

SSE (Rojle)

Clay Fm. (part)

Resna~s Clay Fm.

varielongitudum

11" 19. solidum ......

5 4 - ~

10

W. astra -E.

7 6

9 Late Paleocene

8

7

5 4

ursulae

2c 2b 2a

Balder Form. Sele Form. Lista Form. (part)

~ Knudshoved Mb. ~ 1 1 1 I1111111111111111111111111111111111111111111

+140-I d Silstrup Mb. Werum Mb. '+series' +1-Fur Fro. ~)lst F o r m . - KnudeklintMb. > u~,..~.~ LAk x l 1 ~ '-series'-33 -39~ Grey Clay

fJll llrl

i

Holmehus Form. , .

Fig. 2. Stratigraphy and age of the Danish ash-bearing deposits and their North Sea equivalents (based on information from Knox 1984; Heilmann-Clausen 1985, 1988; Heilmann-Clausen et al. 1985; Knox & Morton 1988). Only NP Zones (Martini 1971) marked with an asterisk are identified as biozones in Denmark. Other NP Zones are suggested from dinoflagellate correlation between Denmark and other West European areas and on an ash layer correlation (see main text). Hatched area indicates the interval in which the Paleocene-Eocene boundary occurs. Chronology after Aubry et al. (1988) and Swisher & Knox (1993).

278

B. SCHMITZ ET AL.

the Fur Formation in northwestern Denmark and the Olst Formation in the southeast (HeilmannClausen et al. 1985; Nielsen & Heilmann-Clausen 1988). The former is c. 60 m thick and contains about 170 ash layers of predominantly basaltic composition (Pedersen et al. 1975; Pedersen & Surlyk 1983). The Olst Formation contains the same ash layers with a further six at its base. This formation varies in thickness, usually from 9-29 m (Heilmann-Clausen et al. 1985). In Denmark the ash layers of volcanic phases 2a and 2b have been numbered consecutively from -39 to +140 (BCggild 1918; Gry 1941). Biostratigraphy for the non-calcareous sediments underlying the RCsna~s Clay Formation relies on dinoflagellate and ash-layer correlations (Fig. 2). Relying on an ash layer correlation, the boundary between nannoplankton zones NP9 and NP10 is probably close to the base of the Fur Formation (Nielsen & Heilmann-Clausen 1988). An ash layer identical to the Danish l a y e r - 1 7 is identified in the lowermost part of Zone NP10 (c. 8 m above the base of the zone) in DSDP Hole 550 (Goban Spur, SW of Ireland) (Knox 1984; Aubry 1995; Berggren & Aubry 1996). The Fur Formation was probably deposited at a higher sedimentation rate, and the base of the formation is located 13 or 14 m below the ash layer -17. By means of 4~ dating an age of c. 54.5 Ma has been obtained on single feldspar crystals from layer -17 (Swisher & Knox 1993). A nannoplankton (NP) zonation for the noncalcareous Danish deposits is suggested by means of dinoflagellate correlation between Denmark and other West European areas (Heilmann-Clausen 1985) (Fig. 2). For example, the Danish dinoflagellate Zone 6 (the acme of Apectodinium spp.) is correlated with the Woolwich Beds in southern England. At one locality the basal Upnor Formation (formerly the Woolwich Bottom Bed) yields calcareous nannoplankton indicating Zone NP9 (Siesser et al. 1987). Dinoflagellate Zone 6 also correlates with the lower part of the Alpine 'Association ~t W. homomorpha' (Jan du Chine et aL 1975), of which the base is in the lower part of Zone NP9. The middle part of the 'zone W. hyperacantha' defined in the Pyrenees also correlates with dinoflagellate Zone 6 and is suggested to be restricted to Zone NP9 (Caro 1973; Caro et al. 1975). In conclusion, there is clear evidence that the base of Danish dinoflagellate Zone 6 occurs in Zone NP9 and is possibly confined to it. The RCsn~es Clay Formation consists mainly of very fine-grained, reddish or yellowish brown, calcareous clay (Heilmann-Clausen et al. 1985). It has previously been assigned to the early Eocene nannoplankton zones NPI1 and NP12 (Thiede

et al. 1980; Heilmann-Clausen et al. 1985). The

formation occurs throughout the distribution area of the Danish Eocene. A comparable fine-grained red-brown clay unit also occurs over extensive areas of the North Sea and in northern Germany. In northern Germany, south of a line connecting Bremen and Hamburg, the red-brown clay is laterally replaced by a more sandy, glauconitic facies, reflecting a position closer to land. The London Clay Formation in England and the leper Formation in Belgium represent more landproximal sediments that formed in the North Sea Basin at the same time as the RCsna~s Clay. These formations are more coarse-grained and about an order of magnitude thicker than the Rcsn~es Clay Formation. The thickness of the Rcsna~s Clay Formation in Denmark varies between 3.25 and 20 m. The formation has previously been divided into seven distinct beds (Heilmann-Clausen et al. 1985). The lowermost bed constitutes the Knudshoved Member, which is restricted to the western Limfjord area. The remaining six beds have been named consecutively, from the base upwards, R1 to R6. Some of the beds are quite thin but, despite this, show only minor lateral variation over wide areas. A detailed description of the subdivision of the RCsna~s Clay Formation is given in Heilmann-Clausen et al. (1985). The Rcsn~es Clay Formation grades upwards into the mainly grey-green, non-calcareous and extremely finegrained Lilleb~elt Clay Formation (Fig. 2).

The Paleocene-Eocene boundary in Denmark The placement of the Paleocene-Eocene (P-E) boundary in the sedimentary record of Denmark must await the completion of the ongoing international revision of the definition of this boundary. Since the introduction of the term Paleocene in the nineteenth century, the position of the P-E boundary has been controversial. Various stratigraphic levels have been suggested in the classical Paleogene sections of northern France, Belgium and southern England. These levels include the base of the Sparnacian in northern France, the base of the Ieper Clay in Belgium and the base of the London Clay/Oldhaven Beds in southern England. A biostratigraphic correlation of these boundary levels to deposits in Denmark was made by Heilmann-Clausen (1985) on the basis of dinoflagellates. Thus, the base of the Sparnacian coincides with the base of the Danish dinoflagellate zone 6, which is at the base of the Olst Formation (Fig. 2). The base of the Ieper Clay coincides with the base of the Wetzeliella astra zone (De Coninck

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

1990), which in Denmark occurs at the base of the Knudshoved Member (Fig. 2). The base of the Oldhaven Beds is diachronous according to Knox et al. (1983), and seems to span strata ageequivalent to the Danish dinoflagellate zone 7. Hence the various suggestions for the P-E boundary in NW Europe may be traced to Denmark where they correlate with either the base or the top, or some intermediate level, of the r Formation. On the other hand, if the P-E boundary is placed at the level of the global benthonic foraminiferal extinction event and associated stable isotopic shifts (see Kennett & Stott 1991; Berggren & Aubry 1996), this would mean that the boundary in Denmark lies somewhere in the lower part of the Olst Formation or in the upper Grey Clay (Fig. 2). Here we will consider the P-E boundary as corresponding to the NP9-NPI0 boundary (see, for example, Berggren et al. 1985; Hardenbol 1994), which in Denmark lies close to the base of the Fur Formation.

The late Paleocene--early Eocene section at Alb~ek Hoved RCsnces Clay and Olst Formations For this study we have collected samples from the RCsna~s Clay Formation in ice-rafted clay mounds at Alba~k Hoved (55042 ' N, 9058' E) on eastern central Jutland. Here the RCsna~s Clay Formation is particularly expanded and well exposed (HeilmannClausen 1990). In the basal part of the mounds there are a few metres of the non-calcareous Holmehus Formation (Fig. 3). A hiatus, probably representing a substantial amount of time, separates the Holmehus Formation from the overlying ashrich Olst Formation. The Olst Formation is only 2 m thick at Alba~k Hoved and contains about 40 of the ash layers (+79 to + 118) that were deposited during volcanic subphase 2b (Fig. 2). The RCsna~s Clay Formation overlies the ~lst Formation disconformably at this site and attains a thickness of c. 20 m. The lowermost member of the RCsna~s Clay Formation, the Knudshoved Member (Fig. 2), is missing, indicating the presence of a hiatus. In NW Denmark the non-calcareous Knudshoved Member is only about 4 m thick (HeilmannClausen et al. 1985). The member is predominantly made up of silty clay, implying that it was deposited relatively rapidly. The biostratigraphic succession across the RCsna~s Clay Formation at Alba~k Hoved has recently been studied in detail (Heilmann-Clausen, pers. comm.). The most important results as regards nannoplankton zonation, foraminiferal distribution, and local Danish dinoflagellate zonation are summarized in later sections and in Fig. 3.

279

Both the detailed nannoplankton and the dinoflagellate (Fig. 3) stratigraphies indicate that the RCsn~es Clay Formation at Alba~k Hoved is relatively complete over the interval from lower NP11 to the base of NP13. This stratigraphic interval corresponds to a time period of somewhat more than 2 million years, using the timescale of Aubry et al. (1988). The 20 m thick section thus appears very condensed, having formed at sedimentation rates as low as c. 1 cm per thousand years.

The +11 to +16 m interval o f the RCsnces Clay Formation The interval from 11-16 m above the base of the RCsna~s Clay Formation was studied for highresolution isotopic distribution because some prominent lithological 'event-beds' occur in this interval (Fig. 4). The major part of the interval contains abundant, well-preserved foraminifera. In the lower 3 m of the interval there is highly calcareous, light reddish brown clay. Above this follows 2 m of calcareous, light greenish grey clay (Heilmann-Clausen etal. 1985). Heilmann-Clausen (1990) identified three distinct lithological beds in the interval, numbered in ascending order from 6 to 8. Layer 6 is a 10 cm thick ash layer (no. V16 in Fig. 3) below a white clay layer, c. 15 cm thick. The white clay occurs c. 10-20 cm above the ash and may reflect a separate event. We therefore identify the ash as no. 6a and the white layer as no. 6b. Layer 7 is a prominent 20 cm thick white clay layer. Layer 8b is a black clay layer, extremely rich in dinoflagellates and other organic materials. The layer probably formed in connection with a plankton bloom. It is an important marker bed, that occurs also at other sites in Denmark where the RCsna~s Clay has been found. The layer is similar in appearance to the Cretaceous-Tertiary boundary clay (the Fish Clay) at Stevns Klint (Schmitz 1988). In the interval just below layer 8b there are abundant pyrite spherules similar to those that occur in the Fish Clay (Schmitz et al. 1988). We analysed samples from layer 8b with the Iridium Coincidence Spectrometer at the Lawrence Berkeley Laboratory (see Schmitz et al. 1991) and found 0.4 ppb Ir. This is a prominent enrichment compared to average shale (< 0.05 ppb Ir), but orders of magnitude lower than the Ir content of the Fish Clay at Stevns Klint (c. l l 0 p p b Ir; Schmitz 1988). The small Ir enrichment in layer 8b probably represents volcanic Ir that has been remobilized from surrounding sediments. The adjacent sediments are also generally high in Ir (0.05-0.2 ppb), because they contain reworked volcanic Ir-rich ash from the earlier explosive basaltic volcanism (see Schmitz 1994). Layers 8a and 8c are less distinct

280

B. SCHMITZ ETAL. AG[ FORMATIONI NP I DINO

AS~ES I zo.E ZONE

~T

.

.

.

~13 C

.

I

.

.

.

~18 O

.

.

INP13!

+18 ,

+16

c~o~eo c.

__I

NP12

D

NP11

=

.

1,.-.

.

.

.

.

.

.

.

.

.

/J'"

12~ .- . . . .

.

.

.

--,.-o-

o~

.

L e n t i c u l i n a spp. :- O r i d o r s a l i s u m b o n a t u s = S u b b o t i n a spp.

NP8

.._i

.._.1 O 'l-

Fig. 3. Stratigraphy and foraminiferal stable-isotopic results for clay section at Albmk Hoved. Dinoflagellate zones refer to local Danish zonation (Heilmann-Clausen 1985, 1988); NP zonation refers to Martini (1971). Units A to G refer to planktonic/benthonic ratio variations described in Fig. 17. In the 2 m-thick Olst Formation 44 ash layers have been found, whereas the overlying RCsnaes Clay Formation contains about 20 ashes.

beds, and are characterized by streaks of black organic-rich clay. The sediments surrounding layers 8a-8c are light grey to whitish grey. The white colour is a weathering phenomenon, and occurs only at the surface of the outcrop. At a few centimetres depth the sediment is instead darker grey, indicating high organic matter content. The same applies for the white layers 6b and 7.

Isotopic studies: materials and methods Samples for stable isotopic studies were collected at 0.1-0.5 m intervals through the major part of the RCsnms Clay Formation at Albmk Hove& Many of the samples, however, could not be used because of insufficient calcite content. In addition, a highresolution isotopic profile was established for the interval from 11-16 m above the base of the formation. The major part of this interval was sampled at 5 cm resolution, and some parts at 1 0 c m resolution. The R0sn~es Clay at Albmk Hoved appears generally in a very fresh, unweathered state at some centimetres depth below the outcrop surface.

Erosion by sea waves constantly exposes new clay. Tree and plant roots, a feature characteristic of many weathered land sections, are generally absent. The isotopic analyses of the low-resolution profile spanning the major part of the R0snms Clay Formation were performed on picked foraminiferal samples and on bulk-rock samples. The former included monospecific assemblages of the benthonic foraminifera Oridorsalis umbonatus and of planktonic Subbotina spp., respectively (all or most of the Subbotina spp. belong to Subbotina ex gr. linaperta, i.e. Subbotina patagonica of Gradstein et al. 1992). In addition, we picked monogeneric samples of the benthonic foraminifera Lenticulina spp. We chose O. umbonatus because it has a very long range in the RCsnms Clay Formation. Lenticulina spp. also occurs throughout most of the section, and has the additional advantage of being thick-shelled, which facilitates examination for possible diagenetic alteration and infillings. For isotopic analysis, 0 . 3 - 0 . 4 m g foraminiferal calcite was used. We used about 30 individuals of O. umbonatus in the size range 125250 ~tm and 60-80 individuals of Subbotina spp. in

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

281

powdered rock were separated for isotopic analyses. For the high-resolution profile in the +11 to +16m interval we analysed assemblages of the benthonic foraminifera Cibicidoides ungerianus and planktonic Subbotina spp. For each analysis, we used three to seven individuals of C. ungerianus, which corresponds to 0.05-0.1 mg calcite, and ten to 15 individuals (0.04-0.09 mg calcite) of Subbotina spp. (185-250 ~tm size). Isotopic analyses were also performed on assemblages of O. umbonatus (125-250 ktm) and Subbotina spp. (185-250mm) from the early Eocene part of DSDP Hole 550. For these analyses small samples (0.03-0.08 mg) were used. Foraminifera were examined in detail with the light microscope for calcite infillings and overgrowths. Only clean, well-preserved individuals were used for establishing the isotopic records presented here (see next section). The isotopic analyses of foraminifera from the profile spanning the major part of the ROsna~s Clay Formation were performed according to standard procedures using a VG Micromass 903E mass spectrometer at the Department of Marine Geology in Grteborg (see Schmitz et al. 1992). (From a few levels, smaller foraminiferal samples, 0.030.05 mg calcite, were analysed with a Finnigan Mat 251 at the Institute of Energy Technology, Kjeller, Norway; Table 1.) The foraminiferal samples of the high-resolution profile were analysed with a VG Prism Series II mass spectrometer installed at the Department of Marine Geology in Grteborg in November 1993. Bulk-rock samples from the Rcsmes Clay Formation and DSDP 550 foraminiferal samples were analysed with a VG Prism mass spectrometer at the Department of Earth Sciences in Oxford. Intercalibration between the laboratories was attained by analyses of NBS standards 19 and 20. All isotopic data are given as 6 per mil values normalized v. the PDB standard.

Preservation of foraminifera Fig. 4. Lithology of the interval from 11-16 m above the base of the ROsn~esClay Formation. For highresolution stable isotopic profiles, see Figs 9-12. The numbering of the event beds refer to that introduced by Heilmann-Clausen (1990).

the size range 185-250 ktm. For the analyses of Lenticulina spp., six to eight individuals (250500 ktm large) were used. For the bulk-rock isotopic analyses 3-4 g of dried sediment was ground in an agate mortar. After calcite determination by EDTA titration, aliquots of 0.3-0.5 mg of

The effects of diagenetic processes may obscure original isotopic signals in fossil foraminiferal tests. The most common problems are: (1) replacement of original test calcite; (2) isotopic exchange and re-equilibration between test and pore water; (3) secondary infillings and/or encrustations attached to original tests (Barrera et al. 1987; Barrera & Huber 1990). Prior to analysis, foraminifera from all sampled levels were examined under the binocular microscope for possible diagenetic alteration. In addition, tests from some levels were studied by scanning electron microscope (SEM) (Fig. 5). We examined the outer surface and, after breaking the tests, also

282

B.

SCHMITZ ET AL.

Table 1. Isotopic results (%0 v. PDB) for foraminiferal assemblages from ROsnces Clay Fm Metres above base

O. umbonatus ~513C

1.0 1.5 2.1 2.5 3.4 3.9 4.7 4.7 4.8 4.9 5.2 5.5 5.7 5.9 6.1 6.4 6.6 6.7 6.9 7.2 7.4 8.5 9.0 9.2 9.2 9.7 9.7 10.2 10.7 11.2

~18O

-2.70 -0.81

-2.03 -2.12

-0.73 -1.92 -1,8" -1.17

-1.78 -2.03 -3.1" -1.64

-1.13 -1.15 -1.6" -1.13 -1.22 -1.46 -1.24 -0.97 -0.65 -0.98 -0.72

-1.74 -1.50 -1.8" -1.61 -1.49 -1.74 -1.61 -1.34 -1.08 -1.37 -1.42

-0.84

-1.34

-0.60 -0.89

-1.29 -1.43

11.7

-0.7*

-1.7*

12.2 12.7 13.2 13.4 13.5 13.5 13.7 14.2 14.7 14.8 14.9 15.2 15.7 15.7 16.1 17.6

-0.67 -0.91 ~).56

-1.49 -1.57 -1.26

-1.2" -0.59 -0.9* 0.0" -0.9*

-1.34 -1.9" -1.7" -2.4*

-0.54

-1.44

Lenticulina spp. ~13C

~18O

-1.60 -1.50 -1.89 -0.63 -2.45 -2.91 -1.09 -0.67

-1.70 -1.65 -1.68 -1.57 -2.15 -2.55 -2.08 -1.94

-1.75 -1.8"

-2.21 -2.0*

-1.84

-2.11

-1.62

-1.83

-0.63 -1.61 -1.20 -0.79 -1.28 -1.23 -1.27 -1.72 -2.02 - 1.73 -1.17 -1.40 -0.74 -2.65 -2.65 -3.05 -0.91

-1.46 -1.95 -1.81 -1.84 -1.88 -1.98 -1.93 -2.25 -2.39 -2.26 -2.03 -2.20 -1.80 -2.73 -2.76 -2.83 -1.85

-0.60 -1.04 -1.87 -1.1" -1.22 -0.97 -2.30

-1.89 -2.19 -2.66 -1.8" -1.81 -1.87 -2.72

Subbotina spp. ~13C

~180

-0.64 +0.47

-4.12 -2.46

+0.82 +0.90

-2.17 -2.32

+1.19 +0.79 +1.24 +0.36 +0.29 +0.59 +0.48 +0.52 +0.76 +1.80 +0.41 +0.59 +0.85

-3.96 -3.63 -3.25 9-2.01 -3.64 -2.03 -2.13 -1.96 -2.65 -3.95 -1.93 -2.53 -2.92

+0.61

-2.07

+0.62 -0.35 +0.04 +0.10 +0.49 -0.04 +0.58 -0.33 -0.25

-1.78 -2.45 -2.41 -2.32 -2.25 -2.67 -2.12 -2.62 -2.87

+0.70 +1.12 +1.15 +0.46 +0.20 +1.11 +0.91

-2.57 -3.04 -2.40 -2.62 -2.81 -3.22 -3.21

+0.10 -0.19

-3.11 -3.08

* = Sample analysed at Institute of Energy Technology, Kjeller, Norway

their interiors. Throughout the major part o f the R0sn~es Clay Formation, from 4 m a b o v e its base to its top, the tests appear excellently preserved, and infillings or encrustations are absent (Fig. 5b, c, e & f). The benthonic foraminifera typically s h o w an enamel-shining surface. Pores and other minute

c h a m b e r surface features appear as new. Original growth-related textural features are present, such as the prismatic structure of Lenticulina walls (Fig. 5c), and growth-lamellae parallel to test surface in O. umbonatus (Fig. 5a & b). In fact, the foraminifera o f the R0sna~s Clay Formation in

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

283

E

Q ~

;,,..= ~ +

"~+~ ~..=

9~

ej

~eq ID

~

e~

'-"

r.~ , - ~

--t

284

B. SCHMITZ ET AL.

general represent some of the best preserved early Paleogene specimens that can be found. Within the basal 4 m of the formation the tests also appear well-preserved, but in some specimens there are secondary, euhedral calcite crystals or encrustations on the interior chamber walls (Fig. 5d). In some cases thin layers of compressed coccoliths occur on the inner walls (Fig. 5a). The interior walls of tests devoid of overgrowths differ from those with secondary calcite, by being covered by a thin brown organic coating. This coating is easily detached from the wall, and most likely represents the remains of the soft tissue of the foraminifera All picked Lenticulina tests were individually broken so that their interior chamber walls were exposed. The test fragments were treated with ultrasound for 10-20 s, and thereafter examined in detail under the binocular microscope. Only fragments that were definitely free from secondary calcite were used for isotopic analyses. Many specimens of O. umbonatus, C. ungerianus and Subbotina from different levels were broken and examined for occurrence of secondary calcite. It was soon realized that secondary calcite is only present at levels where the Lenticulina specimens have infillings. Secondary calcite infillings were only observed at some levels in the lower 4 m, although we extensively searched for them throughout the entire formation. Isotopic results for O. umbonatus or Subbotina from levels where calcite infillings of Lenticulina tests were observed, have been disregarded.

Stable isotope records ~13C variations through the RCsnces Clay Formation

Both the planktonic and the benthonic foraminiferal 513C records through the RCsn~es Clay Formation show substantial, rapid and seemingly random fluctuations. The strongest variations are observed for Subbotina and Lenticulina, whereas Oridorsalis shows a more stable trend (Fig. 3; Table 1). The benthonics Lenticulina and Oridorsalis give relatively similar values throughout most of the section, but at some levels such as between +10 and +12 m relative to the base of the RCsna~s Clay Formation the two records diverge. At the maximum divergence in this interval, Lenticulina shows 1.3%o lower ~513C values than Oridorsalis. The latter shows a stable trend over the interval, whereas Lenticulina displays a gradual, negative 813C excursion. At 13.4m, where the distinct greyish-white layer 7 occurs (Fig. 4), there is also a pronounced divergence. Here Lenticulina shows

a dramatic negative ~13C excursion, reaching values as low as -3%o, whereas Oridorsalis only shows a small negative shift from values of c. -0.6 to -1.3%o. Most benthonic foraminifera do not secrete calcium carbonate in isotopic equilibrium with ambient pore or bottom waters (Woodruff et al. 1980; Graham et al. 1981; Vincent et al. 1981; Grossman 1987). So called 'vital effects' lead to isotopic fractionations, mostly resulting in depletions in the heavy isotopes. Incorporation of metabolic carbon-oxygen compounds in the calcium carbonate is the most likely explanation to these 'vital effects' (Grossman 1987). Different species show more or less strong fractionation. For carbon isotopes it has been shown that both Oridorsalis and Lenticulina strongly discriminate 13C relative to 12C when precipitating their calcite. Both typically show 13C depletions of 0.5-1.5%~ compared to ambient water (Grossman 1987). In general, the extent of isotopic fractionation in a benthonic foraminiferal species is constant, leading to parallelism in multiple monospecific isotopic curves through time. Microhabitat environmental differences, however, may distort this parallelism (Grossman 1987). For example, below the sediment surface, bacterial oxidation of 13C-depleted organic matter may induce lower pore-water 513C values than at the seafloor. Infaunal species may therefore register more negative 513C values than epifaunal species (McCorkle et al. 1990). Recent findings show that O. umbonatus is a transitional infaunal taxon living in the sediment from 0 to c. 4 cm below the surface (Rathburn & Corliss 1994). Lenticulina show similar depth habitats, being either epifaunal or shallow infaunal, thriving at 1-2 cm depth in the sediment (Corliss & Chen 1988; Corliss 1991; Corliss & Fois 1991). Thomas & Shackleton (1996) found that Oridorsalis generally shows lower 513C values than typically epifaunal species, supporting a habitat at some depth. The great similarity in 813C values for Lenticulina and Oridorsalis at most levels in the ROsna~s Clay Formation, indicate that they fractionate carbon isotopes in a similar manner. It also indicates generally the same depth habitat in the sediment, or possibly different depth habitats, but absence of a vertical ~13C gradient in the uppermost sediment. The divergences between the two records at some levels reflect changes in depth habitat or 513C gradient. These changes may be primarily related to variations in the flux rate of organic matter to the seafloor. Subbotinid foraminifera lived in the deep part of the planktonic realm according to isotope depthranking approaches (Boersma & Premoli Silva 1983; Shackleton et al. 1985; Corfield & Cartlidge 1991). Subbotina ~513C values fluctuate consider-

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK ably between extreme values o f - 1 and 1.8%o (Fig. 3). The foraminifera specimens giving extreme values are excellently preserved, supporting that short-term isotopic variations occurred, possibly related to an unstable upper water mass. The subbotinids in the RCsn~es Clay Formation give about one to two per mil more positive 813C values than Oridorsalis, reflecting a well-developed, vertical negative 813C gradient between intermediate depths and the seafloor. Benthonic foraminifera data in this study indicate water depths of c. 600-1000 m during deposition of the RCsna~s Clay Formation. The generally great difference in 813C between Subbotina and Oridorsalis is consistent with such water depths. In shallow water (< 200 m) environments, the deep-dwelling planktonic foraminifera typically register similar 813C values as the bottom-dwelling foraminifera (Schmitz et al. 1992). At many places in the present-day ocean there is a vertical negative 813C gradient over the upper 500-1000 m in the water column (e.g. Kroopnick 1980, 1985). The vertical 813C profile in the North Sea during the early Eocene may have been similar to, for example, the present-day type carbon-isotopic profile at GEOSECS stations 212 and 213 at mid-

latitudes in the North Pacific (Kroopnick 1985; McNichol & Druffel 1992). At these stations, there is a gradual one to two %o negative 813C shift over the interval from where the subbotinids dwelled down to 600-1000 m depth. Decreasing 813C values with depth are related to the sinking and decomposition of ]3C-depleted organic matter (813C c. -25%o). During the decomposition of the organic material oxygen is utilized, thus the 813C depth profile in general parallels the oxygen curve (Kroopnick 1974, 1980, 1985). The water mass pH is inversely related to the amount of CO 2 and carbonic acid, which primarily derive from the decomposed organic detritus. Therefore, the 813C depth profile also gives an indication of how corrosive bottom waters were. In the North Sea in the early Eocene oxygen content and pH apparently decreased with depth, whereas CO 2 and carbonic acid increased. The oxygen minimum level probably lay just above the seafloor. This indicates that replenishment of deep water was generally slow, or that deep waters entering the North Sea were 'aged', with already low oxygen content and being corrosive. The Subbotina-Oridorsalis 813C difference (A~13C) fluctuates throughout the section (Fig. 6).

A 8~3C

v~ iii z

:~

w

z

I

NP12

I

A 8~80

~

S

y14--

Q r"r v.

NP11

-2

" ~

03 I

O uJ "J ._i

--~ NP81 -r" LLI _ _ ~ ~; O :l-

0

1

2

285

3

0

-1

Fig. 6. Subbotina-Oridorsalis A8]3C and A8180 variations through the RCsn~esClay Formation.

286

13. SCHMITZ ET AL.

The Ag13C values are generally higher in the interval from +4 m up to c. +9 m, than in the interval from +9 to +15.8 m. In the lower interval most of the A~13C values lie in the range 1.6-2.4%o, compared with 0.8-1.4%o in the upper interval. The higher A813C values in the lower interval reflect higher 813C values in Subbotina and lower in Oridorsalis compared to the upper interval. The higher ~513C values for Subbotina probably reflect that addition of decomposing organic material sinking from the surface was slower than when the upper interval formed. Thus, in the earlier interval, biological productivity in the upper water mass was low compared with later interval. During the earlier interval bottom waters were also lower in ~513C because of strong density stratification of the water mass. The renewal of bottom waters was slow and 13C-depleted carbon diffusing out from the sediment and from sinking plankton accumulated in the bottom water. The presence of corrosive, low-pH bottom waters in the North Sea during this period is also supported by the low planktonic/benthonic foraminifera ratios and low CaCO 3 content observed in the lower interval (see later sections). The smaller A~513C gradients in the upper interval can be explained by increasing

~3 (%)

addition of decomposing organic matter at middepth related to higher surface water productivity. Bottom water (Oridorsalis) ~13C values increase because of invigorated bottom water circulation. The reduction in Subbotina 513C values and the inferred increase in surface water productivity in the upper interval coincides with the major divergence between the Lenticulina and Oridorsalis 513C records, which may reflect a change in response to increased influx of organic matter to the seafloor. Possibly Lenticulina moved deeper into the sediment, whereas Oridorsalis stayed at or near the sediment-water interface. Bulk sample ~513C and CaCOa results are presented in Fig. 7. With respect to 8r3C values, the RCsn~es Clay Formation can be divided in three pans. In the lower 7 m of the formation 813C values generally lie in the interval -1.0 to 0%0. From c. +7 to +10 m, 513C values are higher, in the range +0.4 to +2. In the upper third of the section values are again lower, fluctuating in the range -0.2 to +0.6. Analyses of different aliquots of ground bulk samples from levels with low calcite content give sometimes widely scattered results, indicating inhomogeneties. In particular, in the samples with low calcite content, the 813C signal of

I

813C (bulk)

$18 0 (bulk)

~ 0...4

.._t~

I

~,.eo

g tl,---,O

H M

"El e

e.e J.e

I

",'r--"

o.~-. N

_C Z-"

le

I

4i

I

25

50

75

-1

0

+T~

-2

j

-1

-2

-3

-4

-5

Fig. 7. CaCO3 content and stable isotopic composition of bulk samples through the RCsn~esClay Formation. (For additional information, see also Figs 3 and 17.)

287

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

heterogeneously distributed, diagenetic and authigenic calcite may sometimes overprint the nannofossil-calcite isotopic signal. In general, bulk-sample calcite is primarily made up of coccoliths. Coccolithophores grow and calcify primarily in the upper 100-200 m of the oceans. In the present-day Pacific, around the equator and north of 30~ ~ N, they are most abundant within the upper 50-100 m of the water column (Okada & Honjo 1973). In the midlatitudes the standing crop is more or less evenly distributed over the upper 200 m. The bulk-sample values most likely give a rough estimate of the average 813C composition of the water mass over the entire depth habitat of the coccolithophores, and therefore gives an idea of the isotopic composition closer to the surface than where the deep-dwelling Subbotina lived. This is supported by the oxygen isotopic results, discussed in the next section. The bulk-sample 813C values, however, are relatively similar to the Subbotina values. Usually, in the water column, there is a strong vertical 8~3C gradient between the surface and the base of the euphotic zone (Kroopnick 1974, 1980, 1985). Biological productivity in surface waters leads to transport of light 12C with decaying organic tissue from the surface to the deep. Reduced salinities or any other property of surface waters in the North Sea, however, may have suppressed productivity near the surface, leading to a reduction of the 813C gradient within the uppermost water mass. Wellpreserved nannofossil calcite usually registers the higher 813C values in the upper water mass (Margolis et al. 1975), and there is no strong reason to believe that influence of 'vital effects' has severely distorted the record. Considering the spread in bulk sample 813C values at some levels it

seems likely, however, that diagenesis has obscured original features of the record to some extent. The generally low 813C values in the lower 6 m, for example, may reflect diagenesis. Over the lower 4 m we found diagenetic 12C-enriched calcite infillings in foraminifera tests. The CaCO 3 content in the RCsn~es Clay Formation correlates with planktonic/bentonic foraminifera ratios. Based on planktonic/benthonic ratios, the RCsn~es Clay Formation has been divided into seven units, A-G (Fig. 7 and later section). Units with low planktonic/benthonic foraminifera ratios, such as unit D, also show low calcite content. This indicates that both parameters are related to calcite dissolution rates in the water column. The reason for the large 813C fluctuations in the foraminiferal assemblages in the RCsn~es Clay Formation is not clear. Present-day northwestern European freshwater is enriched in light carbon, usually showing 813C values between-9 and-13%o (Mook 1968). In The Netherlands, at the southern shore of the present North Sea, estuarine waters with salinities of c. 31 ppt, show 813C values two to three per rail lower than fully marine waters off the coast. One possibility is that the fluctuations in 813C in the R~sn~es Clay Formation are partly related to variations in organic carbon influx from rivers and estuaries. There is, however, no clear positive correlation between 813C and 8180 in support of this view (Fig. 8).

~ 8 0 variations through the RCsnces Clay Formation The foraminiferal oxygen isotopic records also show considerable, rapid fluctuations in a

O. umbonatus

Subbotina spp. []

%.

B 9 [] mm nnn 9

[] []

[] []

-,z

[]

[]

[]

mm

9

[]

[]

[]

in[]

oo -3

[]

m []

-2

[]

[]

[]

I

-3 Fig. 8. 813C versus

I

-2 5180

I

I|

813C

I

"I

0

'I

-1

I

I

0 813C

correlation plots for O. umbonatus and Subbotina spp., respectively.

288

B. SCHMITZ ET AL.

seemingly random fashion (Fig. 3; Table 1). The isotopic results most likely reflect rapidly changing conditions in a marine-marginal, semi-enclosed basin. It is possible that minor salinity changes, and not temperature changes, are the prime cause of the isotopic variations. Throughout the section there appears to be a relatively constant isotopic offset between Lenticulina and Oridorsalis. Lenticulina generally gives c. 0.3-0.8%~ more negative values than Oridorsalis, although in a few cases the values coincide. The pattern is consistent with the data about foraminiferal 'isotopic behaviour' compiled by Grossman (1987), suggesting that Lenticulina may record lower 8180 values than Oridorsalis due to 'vital effects'. Detailed studies by Vincent et al. (1981) and Graham et al. (1981), show that Oridorsalis precipitates calcite close to oxygen isotopic equilibrium with ambient water. Belanger et al. (1981), however, present results indicating disequilibrium precipitation. Based on extensive studies of fossil Oridorsalis, Shackleton et al. (1984) suggest that this genus is close to equilibrium for 8180, but erratic for 813C. Both the Oridorsalis and the Lenticulina records are relatively stable over the interval from +8 to +13 m, but below and above this the values fluctuate a great deal. In the lower interval Oridorsalis values fluctuate between -1.8 and -3.1%o. The oxygen isotopic record for Subbotina spp. shows even greater fluctuations than the benthonic records. The Subbotina values range between -1.8 and -4. l%o, with the most negative values and the most dramatic fluctuations registered between +4 and +8 m. The larger isotopic variations for Subbotina most likely reflect more unstable conditions at mid-depth than at the seafloor. The fluctuations may, however, also to some extent reflect foraminiferal blooms at different seasons or variations in the depth of test precipitation (Bouvier-Soumagnac & Duplessy 1985; Corfield & Cartlidge 1991). Throughout most of the section there is a strong Subbotina-Oridorsalis 8180 difference (A8180; Fig. 6). The A8180 gradient fluctuates considerably without apparent systematic trend, and varies between -0.3 and -2.6%o, with most values in the interval -0.6 to -1.5%o. This mid-depth to bottom 8180 gradient indicates 2-6~ lower temperatures or 1-2 ppt more saline waters at the seafloor than at mid-depth. Alternatively, a mixed effect of decreasing temperature and increasing salinity with depth may account for the 8180 gradient. The oxygen isotope values in the RCsn~es Clay Formation are lower than most published 8180 values for contemporaneous fully marine environments. Oridorsalis 8180 values lie in the range

-1.1 to -3.1%~, whereas Subbotina varies between -1.8 and -4.1%~. If these data are interpreted in terms of palaeotemperatures, and assuming that fully marine conditions prevailed and that Eocene sea water had a 8180 composition of-1.28%o (icefree world; Shackleton & Kennett 1975), this would imply that bottom and subsurface waters had temperatures in the range 16-25~ and 19-30~ respectively. These calculated high temperatures may, however, be artefacts due to lower salinities. In the present North Atlantic a 1%~ reduction in salinity corresponds to a 0.6%~ decrease in 8180 (Craig & Gordon 1965), which is equivalent to a 2-3~ increase in temperature. The lowest Oridorsalis 8180 values, in the range -2 to -3.1, most likely reflect temporary salinity reductions rather than unusually warm water masses. Subbotina temperatures of 30~ appear unrealistically high, indicating that very negative 8180 values reflect a freshwater admixture. The majority of the bulk samples give 8180 values in the range -3 to -5%o, which is very low. In open marine environments, such as at DSDP Leg 74 sites in the southern Atlantic, early Paleogene bulk samples give 8180 values typically in the range -0.8 to 0.6%0 (Shackleton & Hall 1984). Bulk-sample 8180 values at Alb~ek Hoved are generally one to two per mil lower than those measured in Subbotina. Most likely the coccoliths, which dominate the bulk-rock calcite, preferentially formed somewhat higher up in the water column, thus registering lower salinities than the Subbotina tests. Coccolithophores have a tolerance for somewhat reduced salinities (Okada & Honjo 1973). Alternatively, the low values could be related to disequilibrium isotopic fractionation during coccolith formation. Dudley et al. (1986) have grown different species of coccolithophores in the laboratory and show that there are profound species effects in the uptake of oxygen isotopes. Some species precipitate calcite enriched 1%,~ in 180, whereas others are depleted as much as 2.5%o relative equilibrium. It is not likely, however, that the coccolithophore flora was dominated by 18tdiscriminating species to the extent that it can explain the strong 180 depletions throughout the R0sna~s Clay Formation. Low salinity in the upper 200 m of the water mass is also consistent with other findings in this study, such as Subbotina low 8180 values and the absence of surface-dwelling Morozovella foraminifera (see later section). Meteoric diagenesis may also lead to low 8180 values. Although diagenesis may explain some features of the bulk-sample 8180 record, it is not likely that diagenesis accounts for a several per mil negative shift in 8t80 for all the calcite throughout the section. Only in sections that have been strongly affected by diagenesis do such major

289

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK shifts occur. In sections with very negative diagenetic 8180 values it is impossible, or very difficult, to find the kind of well-preserved foraminifera that characterize the Rcsn~es Clay Formation. The fact that Buchardt (1978) measured 8180 values as low as -4.7%0 in well preserved mollusc shells formed at 30-150 m depth in the Eocene North Sea, gives further support for reduced salinities.

High-resolution 613C and 6180 profiles in +11 to +16 m interval In the +11 to +16 m interval both the oxygen and carbon isotopes show prominent negative excursions at the three major lithological event levels (Layers 6b, 7 and 8b; Figs 4, 9 and 10; Table 2). The 813C negative shifts are on the order of 1.5 to 2%0 for both Subbotina and C. ungerianus, which implies that the entire water column experienced

m

8180 ---..-a.--

.

~.

.

813C Subbotina

16-

C. ungerianus . .

i.

o

8c

8b 15

14[

~ 8a

13] ..........................................

iiiiiiiiiiiiiii

m

spp.

Subbotina

16

......

7

"

6u 6a

12

spp. I

11 15-

!

-1

-3

-4

Fig. 10. High-resolution 8180 profiles for Subbotina spp. and C. ungerianus through interval from 11-16 m above the base of the R0snees Clay Formation. For lithology and event-bed denotation, see Fig. 4.

14-

13-

12-

> 11-2

-2

-I

0

I

Fig. 9. High-resolution 813C profiles for Subbotina spp. and C. ungerianus through interval from 11-16 m above the base of the R0sn~es Clay Formation. For lithology and event-bed denotation, see Fig. 4.

dramatic changes in its carbon isotopic composition. The negative ~13C shifts are almost as large as the famous negative carbon isotopic (2 to 2.6%o) excursion at the P-E boundary (Kennet & Stott 1991). The Subbotina-C. ungerianus A813C gradients show no clear relation to the lithological event levels (Fig. 11). Instead, there is a gradual change, with A513C gradients generally increasing upward in the section. This change is also observed in the Subbotina-Oridorsalis A813C curve (see Fig. 6). As will be discussed later, the gradual increase in A813C is probably related to increasingly restricted water exchange between the North Sea and the open ocean, which is also reflected in the lithological change over the interval, from highly calcareous, reddish brown to calcareous, greenish light-grey clay (Fig. 4). The oxygen isotopic excursions at the lithological event beds are not as large as the carbon

290

B. SCHMITZ ET AL.

A~I3C

8c 8b 8a

7

6b 6a

r

0

'T

I"

!

1

2

Fig. 11. Subbotina-C. ungerianus high-resolution ASI3C profile through interval from 11-16 m above the base of the R0snres Clay Formation. For lithology and event-bed denotation, see Fig. 4.

this is not seen. Nor is there any clear positive correlation between 813C and 6180 (Fig. 13). The lithological event beds and associated isotopic shifts appear to reflect short-term (103-105 year) events when conditions in the North Sea were quite unusual. The chemistry of the entire water mass changed, which could only have happened over a time period longer than the residence time of the water mass in the North Sea Basin. If unrestricted water exchange existed with the Norwegian-Greenland Sea to the north (and perhaps with the Arctic Ocean) a much larger water mass was involved. Thus the isotopic shifts may reflect large-scale regional palaeoceanographic events. In the newly rifted area between Greenland and Norway, there may have been small semienclosed basins with unusual water-mass chemistry. The isotopic shifts in the upper RCsnres Clay Formation may reflect that connections evolved between any of these basins and the North Sea, leading to mixing of two different water masses. Alternatively, the excursions may reflect intermittent water exchange with a semi-isolated Arctic Ocean, with unusual water-mass chemistry. The three lithological event beds are all characterized by a dark grey or black colour in unweathered condition. We think that the dark colour reflects high organic matter content (and/or occurrence of monosulfides, related to reduced oxygen concentrations). In Layer 8b the high organic-matter content is most likely related to a distinct, prominent plankton bloom, whereas, for layers 6b and 7 increasing surface-water biological productivity in general can account for the reducing conditions in the sediments.

Comparisons with DSDP Hole 550 isotopic shifts. There seems to be short lags in the 8180 shifts with respect to the 813C shifts (Figs 9, 10), and the peak negative ~5180 values occur typically 5-10 cm above the lithological event beds The largest ~5180 excursion occurs at the black Layer 8b, where Subbotina 8180 values change from -2.5 to -4%o. (We have no data for benthonic foraminifera at this level, because of insufficient recovery.) At Layers 6b and 7 the shifts are smaller, in the range 0.5 to 1%,~.The 8s80 shifts are of the same size in the benthonics and the planktonics, showing that deep and bottom waters were influenced to the same extent. This is seen also in the A8180 values that do not increase in connection with the isotopic excursions (Fig. 12). A short-term episode of freshwater influx, from for example, river run-off, would have resulted in a larger 8180 change for Subbotina than for the benthOnic foraminifera, leading to an increase in A8180, but

The stable isotopic records for the early Eocene part of DSDP Hole 550 in the North Atlantic are presented in Fig. 14, and comparisons with the North Sea are presented in Figs 15-16. Hole 550 is located in the Goban Spur area southwest of Ireland (Fig.l). The present-day water depth is 4400 m. Palaeodepth backtracking indicates a depth of 4000-4300 m in the early Eocene (see Miller et al. 1985). Nannofossil data indicate a relatively continuous record in the lower Eocene at Hole 550 (Aubry 1995). For stratigraphic correlations we have used the NP Zone boundaries (this study, and Aubry 1995) as reference levels, and depth-age interpolations in the intervals between. The isotopic records for benthonic foraminifera at the two sites (Fig. 15), show that bottom-water chemistry was more stable at the abyssal DSDP Site 550, than in the semi-enclosed North Sea. This is natural considering the much larger volume and longer mixing time of the North Atlantic deep-

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

29 1

Table 2. Isotopic results (%~ v. PDB) for high-resolution profiles 10.90-15.85 m above the base of RCsnces Clay Fm Metres above base

10.90 11.00 11.10 11.20 11.30 11.40 11.50 11.60 11.70 11.80 11.90 12.00 12.05 12.10 12.15 12.20 12.25 12.30 12.35 12.40 12.45 12.50 12.55 12.65 12.75 12.85 12.95 13.05 13.15 13.25 13.35

Subbotina spp.

C. ungerianus

~13C

~180

~13C

-0.285 -0.363 -0.093 +0.123 +0.348 +0.497 +0.266 -0.149 +0.123 +0.131 +0.581 +0.098 +0.160

-2.590 -2.612 -2.618 -2.534 -2.416 -2.370 -1.987 -2.278 -2.329 -2.434 -2.019 -2.376 -2.657

+0.621 +0.623 +0.726 +0.715 +0.462 +0.504 -0.118 -0.922 -0.359 -0.048 +0.136 +0.388 +1.019 +0.846 +0.541 +0.718 -0.274

-2.100 -2.202 -2.073 -2.233 -2.363 -2.417 -2.622 -2.825 -3.183 -2.912 -2.618 -2.192 -1.872 -2.224 -2.527 -2.425 -2.974

-0.135 --0.171 -0.492 -0.373 -0.091 -0.048 +0.231 -0.227 -0.448 -0.108 +0.129 -0.341 -0.357 +0.226 +0.048 +0.268 +0.103 +0.107 +0.051 +0.200 -0.148 -1.483 -0.943 -0.873 -0.449 -0.285 +0.293 +0.401 -0.048 +0.018 -1.286

81SO -2.127 -2.186 -2.042 -1.417 -2.261 -2.230 -2.160 -2.425 -2.522 -2.558 -2.192 -2.200 -2.541 -2.102 -2.242 -2.194 -2.216 -2.319 -2.128 -2.157 -2.398 -1.937 -2.864 -2.744 -2.465 -2.395 -1.955 -1.877 -2.473 -2.178 -2.708

water reservoir compared with North Sea bottom waters. During Biochron NP12 Oridorsalis show similar 813C values in the North Sea and in the northern Atlantic, however, in the N P I 1 interval the North Sea values are generally lower and more variable. Because of a core loss at DSDP Hole 550 comparisons are difficult across the upper NP12 interval. The benthonic 8180 values in the North Sea are throughout 0.6 to 2.5%0 lighter than in the northern Atlantic, with the greatest 6180 difference in the middle and upper N P l l interval, and a convergence near the NP 11-NP 12 boundary. The generally more negative 6180 values of bottom waters in the North Sea may reflect a combined salinity/temperature effect. The Subbotina isotopic records also show m u c h more stable trends at D S D P Hole 550 compared with the North Sea (Fig. 16). Unfortunately, Subbotina specimens are poorly preserved in the uppermost NP12 part (above the core loss) at DSDP Hole 550, which excludes detailed com-

Subbotina spp. Metres above base 13.45 13.55 13.65 13.75 13.85 13.95 14.05 14.15 14.25 14.35 14.45 14.55 14.65 14.70 14.75 14.85 14.90 15.00 15.05 15.10 15.15 15.20 15.25 15.30 15.35 15.40 15.45 15.55 15.65 15.85

813C

C. ungerianus

8180

~13C

~180

+0.065 +1.110 + 1.047 +1.361 +1.172 +0.590 +0.991 +1.441 +0.941 +1.329 +1.213 +1.204 +0.477 +1.078

-2.806 -2.473 -2.766 -2.636 -2.735 -2.045 -3.085 -3.441 -2.429 -3.046 -2.963 -2.495 -2.617 -2.939

-1.272 -0.275 -0.163 +0.610 +0.690 +0.760 +0.268

-2.817 -2.470 -2.069 -1.820 -1.960 -2.114 -2.225

+0.116 +0.214 +0.572 +1.157 +0.774 +0.291 -0.013 -0.168 -0.049 +0.337 +0.165 +0.305 +0.668 +1.133 +0.656

-2.981 -3.099 -2.832 -3.039 -2.361 -2.468 -2.820 -3.577 -4.100 -3.665 -3.505 -3.608 -3.198 -3.042 -3.195

+0.236 +0.403 +0.390 +0.083 -0.380 +0.621 -0.511 -0.616 -0.322 +0.104 +0.205 -0.706 -1.063

-2.271 -2.160 -2.170 -1.714 -1.311 -1.922 -2.441 -2.578 -2.734 -2.458 -2.485 -1.699 -1.891

parisons in this interval. In the lower NP12 part Subbotina 8180 values appear to be generally 1 to 2%0 lower in the North Sea than at DSDP Hole 550, whereas in the upper N P l l part the North Sea values are often as m u c h as 3 to 3.5 lower. The Subbotina records are of particular interest, because they probably reflect the isotopic composition of the water mass at similar depths. A s s u m i n g that water mass temperatures at Subbotina depths were the same at the two sites, the difference in 8180 between the two sites can be attributed to salinity differences alone. Palaeosalinity estimates using this approach indicate generally 2 to 3 ppt, and intermittently 5 to 6 ppt lower salinity at Subbotina depths in the North Sea than in the North Atlantic. The bulk sample 8180 values are 1 to 2 ppt lower than Subbotina values. Thus North Sea upper water mass (i.e. coccolithophore-depths) m a y have had salinities approximately ranging between 26 and 30 ppt. Latitudinal temperature gradients were extremely

292

B. SCHMITZ ETAL.

m

A~I80

16-

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

8c 8b

15, ................

-

- .................

a 8a

low in the early Eocene, thus there is no reason to believe that the Subbotina ~180 difference between the two sites is related to temperature differences, although Alb~ek Hoved lies 600 km further to the north than DSDP Hole 550. A possibly complicating circumstance in the salinity reasoning, however, may be that Subbotina had adjusted to different depths in the North Sea and the North Atlantic (see Corfield & Cartlidge 1991), but there is no way to test whether this was the case.

Biostratigraphic results Foraminifera distribution

14-

13-" .......

....

6b 6a 12-

i

i -I

0

i

Fig. 12. Subbotina-C. ungerianus high-resolution ~]80 profile through interval from 11-16 m above the base of the R0sn~es Clay Formation. For lithology and eventbed denotation, see Fig. 4.

The foraminifera in the Rc~sna~s Clay Formation can be assigned to four broad groups: noncalcareous agglutinating, calcareous agglutinating, calcareous benthonic and planktonic foraminifera. There are marked vertical changes in the relative proportions of these groups, which can be used to subdivide the R0sn~es Clay Formation at Alb~ek Hoved into seven successive biostratigraphic units, here lettered A to G in ascending order (Fig. 17). These units can be identified in all sections of the ROsn+es Clay Formation throughout Denmark, and in subsurface sections in similar facies in the southern North Sea. The vertical assemblage changes are identified by plotting two percentages for each sample analysed: (1) percentage of planktonic foraminifera in the total foraminifera assemblage (P% of King 1989); and (2) percentage of benthonic foraminifera which are non-calcareous agglutinants (NCA% of King 1989). Non-calcareous foraminifera form 100% of the assemblage in the lowest fossiliferous interval

C. ungerianus

Subbotina spp. []

-2 [] []

~m

[]

9

I

In

m.~ []

[] ~ . r % , r 9

[]

9 oo

[]

-2

-3

&

[]

n

9 []B

D me I

[]

9 []

[]

m

m

[]

m

[] B

I

[]

-4

[]

[]

i|

I

-2

I

0

~513C

I

I

-i

!

[]

I

o

I

!

~13C

Fig. 13. 6J3C versus 5]80 correlation plots for C. ungerianus and Subbotina spp., respectively.

293

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

I DEPTH mbsf i

NP ZONE

320-

NP 13

8 180

8 130

I

2

330 NP 12 340 -

350 1

NP 11

360 -2

-1

0

+1

+1 =

I

I

'

0

-1

-2

Subbotina spp. umbonatus

= O.

Fig. 14. Stable isotopic results for DSDP Hole 550.

NP Zone

(~ 1 8 0

8 13C /"

k\

l

12

q,,;P

Q:

\

I

I

!

I

-3

-2

-1

0

-

+1

,I

--9

b

I

I

0

-1

'

I

I

t

I

-2

-3

-4

-5

= DSDP 550 =NorthSea

Fig. 15. Comparison of benthonic foraminiferal isotopic results at DSDP Hole 550 and Alb~ek Hoved. Basal five data points for the North Sea are measured on Lenticulina spp. All other analyses on O. umbonatus.

294

B. S C H M I T Z E T A L .

8 130

8180

,/ k

/ 4 //

-3

-2

-1

0

+1 - ~ - =

+2

i +1

0

-1

-2

-3

.4

-5

DSDP 550 = North Sea

Fig. 16. Comparison of Subbotina isotopic results at DSDP Hole 550 and Albaek Hoved. (unit A). The low-diversity assemblage is characteristic of the 'Rhabdammina-biofacies" (King 1989), including Ammodiscus cretaceus, Glomospira charoides and Glomospirella sp. The 'Rhabdammina-biofacies' is characteristic of restricted (usually poorly oxygenated) carbonatepoor bathyal and abyssal environments in extratropical areas (see Charnock & Jones 1990). The assemblages in unit A resemble Late Cretaceous (Maastrichtian) assemblages (Kuhnt et al. 1989) interpreted as 'middle slope' (middle bathyal, 500-1500 m water depth). Their occurrence in brown clay with low organic carbon content at Alb~ek Hoved indicates adaption to an oxygenated environment. The absence of calcareous foraminifera in this unit is therefore probably due to deposition below the CCD. This unit can probably be correlated with Division A of the London Clay Formation in southern England (King 1981 ). In units B to G, NCA% is very low (generally < 1%), and assemblages are composed predominantly of calcareous benthonic and planktonic foraminifera. Foraminiferal abundance is moderately high (typically > 10 specimens/g of sediment), and the benthonic associations are of moderately high diversity (20-40 taxa/sample). The benthonic foraminiferal associations are relatively similar in composition throughout

units B to G. They are characterized by the consistent occurrence of Gaudryina hiltermanni, Anomalinoides aft. capitatus, Angulogerina abbreviata, Cibicidoides eocaenus, C. aft. ungerianus and Oridorsalis umbonatus. In addition to these, there are some common taxa that are relatively short-ranging and can be used in biostratigraphic correlation (see Fig. 17). Based on the data summarized in Morkhoven et al. (1986, fig. 6), the occurrence of Anomalinoides (aft.) capitatus, Gaudryina hiltermanni, Turrilina brevispira, Vaginulinopsis decorata and Valvulina haeringensis, which have their upper depth limit near the neritic/bathyal boundary (c. 200 m), indicates bathyal depths for the entire interval. Nuttalides truempyi (common in parts of units D and E) and Hanzawaia ammophila (present in unit G) have their upper depth limit within the upper bathyal zone (c. 500 m), while Aragonia aragonensis, which is frequent in a number of samples from unit B to unit G, has an upper depth limit in the lower bathyal zone (c. 1000m) according to the same source. Assessment of the overall assemblages seems to favour a middle bathyal environment (600-1000 m depth) for most of the interval. More precise data on absolute and relative depths of deposition are under study. The changes in the relative abundance of plank-

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

295

Fig. 17. Distribution of some key foraminifera across the RCsn~es Clay Formation. Percentages of planktonic (P) and non-calcareous agglutinated (NCA) foraminifera. Palaeounits A to G represent a subdivision of the R0sn~es Clay Formation, based on variations in the relative proportions of these groups.

tonic foraminifera, used to define units C to G, are believed to reflect changes in water-depth and circulation patterns in the North Sea Basin, as well as extent of water exchange with the world ocean. The latter may be related both to sea level variations in the northeastern Atlantic region (King

1989), as well as to the tectonic history of gateways to the open ocean. The semi-enclosed geography of the basin in the early Cenozoic (Ziegler 1988, 1990; Fig. 1), led to restricted circulation in deep water environments with only intermittent exchange of water between the Basin and the North

296

B. SCHMITZ ET AL.

Atlantic Ocean. In late Paleocene to earliest Eocene times, during deposition of the Sele Formation, low sea levels, perhaps associated with tectonic uplift, led to the almost complete isolation of the North Sea Basin (Knox et al. 1981). The resulting anoxic seafloor environments led to the almost complete extinction of marine benthic faunas in all areas below wave-base. Sea levels began to rise in the earliest Eocene, with each significant rise in sealevel leading to inflow of oceanic waters and an associated influx of benthonic and planktonic foraminifera. Thus foraminifera evolving elsewhere were able to enter the North Sea Basin intermittently. Most biostratigraphic events (first occurrences) identified within the basin reflect episodes of sea level rise, or regional tectonic events leading to connections with the world ocean. Therefore many of the events may be significantly later than apparently correlative events in the open oceanic environments. Unit B is characterized by an influx of calcareous benthonic foraminifera, with very low proportions (< 10%) of planktonics. Non-calcareous agglutinated foraminifera continue from unit A, but comprise only c. 10% of the assemblage. The total faunal turnover at the base of this unit reflects a lowering of the CCD. It is believed on the basis of regional evidence to represent a sea level rise, and can be correlated with the base of Division B in the London Clay Formation. It marks the first major early Eocene influx of planktonic and benthonic taxa from open oceanic environments into the North Sea Basin. Unit B is very thin, indicating either a high degree of condensation, or the presence of a stratigraphic break between this unit and unit C. The base of unit C is marked by a major influx of planktonics, which comprise over 80% of the foraminiferal assemblage. The planktonics are dominated by Subbotina ex gr. linaperta and allied taxa. This influx can be identified at the base of Division C in England, and probably represents the highest sea-levels attained during the early Eocene. Subbotina-dominated planktonic assemblages of this age occur throughout the North Sea Basin (King 1989). The base of unit D is marked by an abrupt decrease in the proportion of planktonics, which comprise less than 10% of the assemblage through most of the unit. A correlative event can be identified throughout the North Sea Basin, reflecting possibly a major fall in sea level or a rise in CCD. Nuttalides truempyi appears abundantly within the upper part of this unit, reflecting a water depth in the southern North Sea greater than 500 m. There is a brief increase in NCA% at the top of unit D, which may prove to be of regional significance. The base of unit E is defined at an increase in P

to between 10-50%. The base of unit F is defined at a further rise in P to between 50-80%. At or just below this event, the first occurrence of the planktonic foraminifera Pseudohastigerina wilcoxensis and Planorotalites spp. are recorded. This event marks a major rise in sea level, which is well-documented from contemporaneous neritic environments, and correlates with the marine flooding events at the base of Division E in southern England (King 1981). The first occurrences of the benthic taxa Bulimina sp. A (King 1989) and Cancris sp. A (King 1989), within unit F, are coincident biostratigraphic events which can be recognized throughout the North Sea Basin (King 1989). The base of unit G is defined at a decrease in P to less than 50%. In the lowermost part of the succeeding Lilleb~elt Clay Formation, P falls to less than 10%, reflecting a fall in sea level marked at the basin margin by the abrupt replacement of open marine by marginal-marine sediments at this time (King 1981). This sea-level fall terminated deposition of the London Clay Formation in southern England.

Calcareous nannoplankton distribution The calcareous nannoflora of the RCsna~s Clay Formation at Alb~ek Hoved is highly diversified. Eighty-five species have been recognized, of which approximately 25% are useful for the biostratigraphic subdivision of the formation and for correlation within the Ypresian of the North Sea Basin (Fig. 18). The underlying Olst Formation and the overlying Lilleba~lt Clay Formation are decalcified, except for the lowermost 1 m of the latter, which presents poorly preserved, low to moderately diversified assemblages assignable to Zone NP13. In the present paper King subdivides the RCsn~es Clay Formation into seven biostratigraphic units, labeled A to G in ascending order, on the basis of vertical changes in the relative proportions of the four main groups of foraminifera (the noncalcareous agglutinating, the calcareous agglutinating, and the calcareous benthonic and planktonic forms). These units, except G, can also be differentiated through their calcareous nannoflora. The boundaries between units A to F correspond to substantial vertical changes in the nannofossil abundance or to special events in the vertical distribution of biostratigraphically important taxa (first or last occurrences, recurrences, base or top of peak abundances) (Fig. 18). They coincide with the boundaries between Steurbaut's nannoplanktonunits I to VI, defined in the proximal shallow-water facies of the southern North Sea Basin (Steurbaut

297

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK

'

~

! J

" ~ ' numbe~ ot specime= ",,..,~,~ - . . . . . . . , ~ ! ~. per mm "~ .t~L,r~u L,I~I~I~U[ m

E b

I [

2'5

'

~

7~s l& UNITS

'

UNITS ]

=

]

CALCAREOUS NANNOFOSSILS

I (1) Steurbaut, 1988 & 1991

,.a~

--1

I~u ,.,1

I i

--20

IX

~.

~ --18

.......(.IL _

VIIlb

[~

=

~

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

i .-~ [ l .

60

sa

--is

Z0

~t

]

~

~'i ~

~ ~=

12

44

-lo!

F

'VI

E

V

36 ~

4 D

~

C

~\ \ \ \ ~ ' ~

~

I -Ill ,

;;;iiiii "

~..1~1

1 IIIa2

,-,-., ~o

'

IVmb

--e

= -

.~

,

~

*

_It

I-

--14

--

~

VIIIa

Illal ~

[~

| "

9 ~ !.,~,",~ 9

--

0

1

~

,

...~2 ;.'.'." . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 'J.L'.'~

l

i

Fig. 18. Qualitative and quantitative distribution of calcareous nannofossils across the Rosn~es Clay and basal Lilleb~elt Clay Formations at Alb~ekHoved (sample numberings after Heilmann-Clausen, pers. comm.; position in metres above the base of the Rcsmes Clay Formation).

1988, 1991; Steurbaut & King 1994). This indicates that these changes in nannoflora have a basin-wide biostratigraphic significance and reflect variations in water exchange with the world ocean. The lowermost calcite-containing sample studied

(A1-2; standard numbering according to HeilmannClausen, pers. comm.) occurs 0.33 m above the base of the RCsnaes Clay Formation. This sample, which probably marks the base of unit A, is characterized by a rich nannoflora, with rather

298

B. SCHMITZ ET AL.

common Discoaster diastypus, D. binodosus, D. kuepperi and Tribrachiatus orthostylus. The presence of these species, together with the absence of Tribrachiatus nunni, T. contortus and Rhabdosphaera sola, indicate the lower part of NP Zone 11 (nanno-unit I of Steurbaut 1991). How far down in N P l l is difficult to say, as in the underlying sediments in the North Sea Basin there is a rather large interval (about 1 to 2 million years) without any calcareous nannofossils (NP10 has never been recorded). With reference to the Goban Spur region (off Ireland) and Spain, however, it appears that the lowermost R~snaes Clay Formation does not correlate with the basal part of NP11, but formed somewhat later (D. kuepperi does not occur in the basal part of NP 11). Unit A has a maximum thickness of 20 cm at Alba~k Hoved. The composition of its nannofossil assemblage (only sample A1-2 was recovered) suggests correlation with the upper part of nannounit I, which, in Belgium, is over 15 m thick. The base of unit A is believed to correspond to the lower boundary of the third order depositional sequence Y-C of Steurbaut, which has been identified in the upper part of the Orchies Clay Member in Belgium and in the upper part of Division A in England (Steurbaut, pers. comm.). Unit B lies approximately between 0.5 m and 0.8 m above the base of the R~sna~s Clay Formation. Only one sample has been recovered (A1-3). Its nannoflora is marked by the incoming of rare Rhabdosphaera sola, which was used to define the base of Steurbaut's nanno-unit II. The nannofossil assemblage of unit B is less diversified and less rich than that of the underlying unit A, although there are many similarities. This suggests that only minor fluctuations must have occurred in the chemistry and temperature of the North Sea surface waters at the start of the formation of unit B. The base of unit C (AI-4) is marked by a sharp rise in nannofossil abundance and by the incoming of the various species, among which the marker Chiphragmalithus calathus, which have not been recorded from the underlying strata. This major influx of nannofossils is known to occur at the base of nanno-unit IIIal. It has been identified within the Roubaix Clay and Mons-en-P6v~le Sand Members in Belgium, just above a major omission surface, representing the main flooding surface of the fourth order depositional sequence (sub-sequence) Y-D1 (Steurbaut & King 1994). Throughout unit C the nannofossil associations are moderately well-preserved, quantitatively rich (over 2000 specimens/mm2 glass-slide; see Steurbaut & King 1994, for details on the counting techniques) and of high diversity (35 to 40 taxa/ sample).

The nannofossil associations in unit D present a rather high, but variable degree of decalcification. The base of unit D is characterized by an abrupt decrease in nannofossil abundance (over 90%), by the end of the acme of Rhabdosphaera sola and by the disappearance of Nannifula dupuisii. These events, which can be identified throughout the North Sea Basin, mark the boundary between nanno-units IIIal and IIIa2 (see Fig. 18). The reason for this drastic fall in nannofossil abundance (?productivity) is not clear, although there is strong evidence for calcite dissolution. In Belgium this event occurs at a medium-grained glauconite horizon, which separates underlying calcareous, nummulitic silty fine sands from overlying less fossiliferous silty clays. Unit D can be subdivided in four sub-units on the basis of qualitative and quantitative differences in the nannoplankton associations. These of unit D1 are poorly preserved and of low diversity. They belong to nanno-unit IIIa2 (absence of common R. sola; presence of Neochiastozygus rosenkrantzii). Sub-unit D2 is marked by a 100% calcite dissolution. Sub-unit D3 contains partially dissolved, low-diversity assemblages (20 taxa/ sample). The first occurrence of Micrantholithus mirabilis lies within this interval. In the less decalcified assemblages from Belgium this event, which occurs at the base of the third order depositional sequence Y-E (Steurbaut, pers. comm.), is slightly posterior to the first consistent occurrence of Discoaster lodoensis, and, thus, falls within Zone NP12. Assemblages from D4 are less dissolved, and characterized by frequent M. mirabilis, Imperiaster obscurus and Pontosphaera sp. The first D. lodoensis are recorded at the base of this interval at Alb~ek Hoved. The overall aspect of the nanno-assemblages of units D3 and D4 refers to periods with more restricted conditions and lower salinities, occurring during periods of relatively low sea-level. The base of unit E is marked by an abrupt increase in nannofossil abundance, by the first occurrence (also acme) of Chiphragmalithus barbatus and by a first marked rise in abundance of D. lodoensis. In Belgium, these events are known to occur at the base of the fourth order depositional sequence Y-E2. The nannofossil associations of unit E show considerable calcite dissolution, except for the base and the top. The base of unit F is characterized by a major influx of nannofossils and by the last occurrence of Toweius pertusus, defining the base of nanno-unit VI. This coincides with the start of the peak abundance of D. lodoensis and the entry of very rare, atypical Rhabdosphaera crebra. These events, which mark a major rise in sea level, have been recorded at the base of the Aalbeke Clay Member

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK in Belgium and at the base of Division E in southern England. The quantitatively and qualitatively richest associations are recorded at 13.75 m above the base of the RCsn~es Clay Formation (Sample A1-49: c. 9000 specimens per m m 2 glass slide; 41 taxa). This level is marked by the start of the peak abundance of R. crebra, which defines the base of Steurbaut's nanno-unit VII, by the common occurrence of Helicosphaera seminulum, and by the incoming of the genus Scyphosphaera. The first occurrence of the benthonic foraminifera Bulimina sp. A (sensu King 1989) and Cancris sp. A (sensu King 1989) seem to be coincident with these events, which all can be recognized throughout the North Sea Basin. Higher up in the R~sn~es Clay Formation, between 14 m and 18 m above the base, the assemblages are moderately diversified (20 to 30 taxa/ sample), although partially dissolved at some levels. The main events within this interval are the first consistent occurrence of Discoaster cruciformis at 15.55 m and the first occurrence of Nannoturba robusta at 17.60 m, which respectively define the base of nanno-units Villa and VIIIb. In Belgium, nanno-units VII and VIII correlate with the shallow marine Egem Sand Member (see Steurbaut & Nolf 1986; and Steurbaut 1995), which has been formed during a sea-level lowstand. Just above volcanic ash-bed V18, at c. 2 m below the top of the RCsna~s Clay Formation, nannofossil abundance and diversity increase again (> 35 taxa).

299

This might reflect a next sea-level rise, after a substantial period of low sea levels. This rise is slightly posterior to the first occurrence of N. robusta and slightly prior to the last occurrence of Tribrachiatus orthostylus, and consequently falls within the extreme top of NP12. In Belgium it corresponds to the transgressive event at the base of the Hyon Sand Formation (Steurbaut & King 1994). Within this interval with progressively increasing nannofossil abundance and diversity, the highest values are recorded at the top of the ROsna~s Clay Formation (A1-67). In the lower part of the succeeding Lilleb~elt Clay Formation the nannofossil abundance falls to less than 10%, reflecting the general fall in sea level.

A semi-enclosed North Sea in the early Eocene The results in this study indicate that in the early Eocene North Sea the major characteristics of water mass chemistry, sedimentation and biology, and changes in these conditions, were determined by the extent of water exchange with the open ocean. Water exchange was controlled by regional sea level and tectonic changes affecting the width and depth of important gateways. We discern principally three different conditions (strongly, moderately and somewhat restricted water exchange) that existed alternatingly in the North Sea (Table 3). These conditions can be related to the division of

Table 3. The North Sea in the early Eocene - three principal conditions

Condition I: Strongly restricted water exchange with open ocean (very low regional sea level) * Strong calcite dissolution; calcite is absent or very rare in sediment. * Non-calcareous agglutinated foraminifera dominate. * Grey or greenish grey clays dominate; dark brown clays occur; low oxygen content in sediment.

Condition II: Moderately restricted water exchange with open ocean (low regional sea level) * Calcite dissolution; low to intermediate calcite content in sediment. * Impoverished planktonic foraminiferal assemblages; low planktonic/benthonic foraminifera ratios. * Surface salinities 4-8 ppt lower than in the open ocean; fluctuating and repeatedly very negative Subbotina 5180 values (-2 to -4%0, related to reduced salinities also at mid-depths. * High Subbotina-benthonic A513C gradients; low surface-water biological productivity and reduced supply of organic 12C at mid-depths; temporarily strong density stratification of water mass, bottom waters corrosive because of slow renewal rates. *Variable lithology; different varieties of reddish brown clays and greenish or whitish grey clays; generally low to intermediate oxygen content in sediment. Condition III: Somewhat restricted water exchange with open ocean (high regional sea level) * Calcite-rich sediments dominate. * Rich planktonic foraminiferal assemblages; high planktonic/benthonic foraminifera ratios. * Surface salinities a few per rail lower than in the open ocean; Subbotina 8180 values (c. -2 to -3%0 indicate generally higher salinities and more stable conditions at mid-depths compared with condition II. * Low Subbotina-benthonic AS13C gradients, reflecting high surface-water productivity as well as invigorated bottomwater circulation; less strong water mass density stratification; less corrosive bottom-water. * Homogeneous reddish brown clay or marl; relatively high oxygen content in bottom-water.

300

B. SCHMITZ ET AL.

the RCsn~es Clay Formation in the units A to G, which mainly reflects the variations in planktonic/ benthonic foraminifera ratios (Fig. 17). The same variations are observed at many sites in the southern North Sea (King 1989). According to King (1989) the variations in planktonic/ benthonic ratios can be explained by migration of planktonics via the Faeroe Trough or the English Channel into the North Sea in connection with high sea levels. The results in this study, however, show that instead the principal governing factor may have been variation in calcite dissolution rates. This is supported by the strong positive correlation between bulk-sample CaCO 3 content and planktonic/benthonic ratios in the Alb~ek Hoved section (Figs 7, 17). In the intervals with low planktonic/benthonic ratios nannofossils also show stronger influence of dissolution. Thus the planktonic/benthonic ratios may mainly reflect the position of the lysocline, which in the North Sea was related to the rate of water exchange with the open ocean. With slow rates of water mass replenishment, there was a CO 2 build-up in the North Sea, inducing low pH and leading to calcite dissolution and reduced planktonic/benthonic ratios. If increasing calcite dissolution is related to a reduction in water exchange with the open ocean this would be reflected in lower salinities and more negative 8180 values. Indeed, the calcite-poor unit D corresponds to the interval with fluctuating and unusually negative foraminiferal 5180 values (Figs 3 and 6). The isotopic comparisons with DSDP Hole 550 also indicate that unit D formed while the North Sea was more isolated (Figs 15 and 16). In the major part of unit F, where planktonic/ benthonic ratios and calcite content generally are high, 5180 values are more positive and the isotopic pattern is more stable than in unit D. However, in the upper part of unit F calcite concentrations become lower (Fig. 7). (This decrease coincides with a lithological change from reddish brown to whitish grey clays.) In this interval 5180 values are again more negative and show fluctuations similar to those in unit D. Most likely, the upper part of unit F represents the beginning of a return to conditions similar to those in unit D. Eventually this gradual change led to the deposition of the non-calcareous Lillebzelt Clay, reflecting a return to a strongly isolated North Sea. Unit E is intermediate and transitional between units D and F in 8180 behaviour, calcite content as well as planktonic/benthonic ratios. In the dissolution unit D the S u b b o t i n a Oridorsalis A813C values are generally about a factor 1.5 to 3 higher than in the more calcite-rich parts of unit F (Fig. 6). In the upper, less calcareous part of unit F, the A813C gradients gradually

decrease towards values similar to those in unit D. Unit E shows an intermediate range of A813C values. The higher vertical 813C gradients in unit D and upper unit F are related both to higher Subbotina 813C values and lower Oridorsalis 813C values. An increase in 813C for surface-dwelling foraminifera is generally interpreted in terms of an increase in productivity. With increasing productivity the light carbon isotope, 12C, is preferentially removed from the water mass into organic tissue of plankton. With increasing productivity in the euphotic zone, the downward flux of decomposing 12C-rich organic matter increased at mid-depth where Subbotina lived. Thus we interpret the higher Subbotina 813C values as reflecting reduced surface-water productivity. The overall sedimentological and chemical data indicate that high Subbotina 13C values coincide with increased isolation of the North Sea and possibly low sea levels. Basin compartmentalization and freshwater admixture induced a more rigid density-stratification. This inhibited surfacewater productivity because of a decrease in upwelling of nutrients. Slow replenishment of the bottom water led to accumulation of CO 2 and carbonic acid from organic matter, and intensification of the oxygen minimum zone near the sea floor. This also explains why benthonic 813C values are low when Subbotina 813C is high. The lowermost part of the RCsn~es Clay Formation in Denmark is non-calcareous. This applies for the Knudshoved Member (missing at Alba~k Hoved) and the 0.3 m interval below unit A at Alb~ek Hoved. In the dark brown clay of unit A at Alba~k Hoved only non-calcareous agglutinated foraminifera are present. Gradstein & Berggren (1981) show that water depth is not the principal factor determining the occurrence of agglutinated foraminifera. More crucial is instead the deposition of 'fine-grained organic-rich carbonate-poor clastics under somewhat restricted bottom water circulation in compartmental basins'. The agglutinated foraminifera of unit A may be relics from the latest Paleocene and earliest Eocene (NP10), when the North Sea was almost completely isolated. At this time organic-rich, noncalcareous sediments with agglutinated foraminifera were ubiquitous in the North Sea. Bottom water 'aging' and CO 2 build-up was even more pronounced than during deposition of calcitepoor unit D in the early Eocene. In Table 3 the three different principal conditions of the North Sea in the early Eocene are summarized. Condition I reflects strongly restricted water exchange with the open ocean. It is represented by the sediment of the lowermost RCsna~s Clay Formation and the Lilleb~elt Clay Formation that began to form at the end of the early Eocene

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK (Fig. 2). Condition II, with moderately restricted water exchange, typically prevailed in connection with deposition of unit D. In the upper part of unit F there is a gradual return to these conditions (Figs 6 and 11). Condition III, with more open water exchange, correlates with the major part of unit E Unit E represents a transitional state between conditions II and III. Unit C, below the low-calcareous unit D, is rich in calcite and shows high planktonic/benthonic foraminifera ratios. Possibly this unit reflects a period of more open water exchange with the world ocean before a return to more isolated conditions. Unfortunately, diagenetic calcite inflllings in the foraminifera prevented detailed reconstruction of water mass conditions at this time. Analyses of well preserved foraminifera from unit C may add important information on the relation between isolation of the North Sea and water-mass chemistry.

Palaeogeography of the North Sea and adjacent areas in the early Eocene The palaeogeographical reconstructions indicate that in the late Paleocene-early Eocene the North Sea and its northward extension, the embryonic Norwegian-Greenland Sea, were delimited from the North Atlantic to the west by the British Isles and plateau basalts between Scotland and Greenland (Fig. 1; Ziegler 1988; King 1989). Based on ridge subsidence calculations it has been argued that water exchange between the North Sea and the Atlantic across the Greenland-Scotland volcanic ridge did not begin until the middle Eocene (Thiede & Eldholm 1983). This is consistent with land mammal distributions (McKenna 1983) indicating that a land bridge existed between Scotland and Greenland until the middle Eocene. On the other hand, Berggren & Schnitker (1983) found that early Eocene plankton faunas and floras in the Norwegian-Greenland Sea and the North Sea were similar to those at the Rockall Plateau in the northeastern Atlantic. They suggest that already at the beginning of the Eocene there was substantial water exchange through the Faeroe Channel, which overlies continental crust and may be a graben structure. Hulsbos et al. (1989) noted that early Eocene calcareous plankton in the Norwegian Sea, the North Sea and the North Atlantic belong to the same biogeographical province. Because of poverty of fauna and flora in the Norwegian Sea, however, they argued that migration did not occur across the Greenland-Scotland Ridge, but through the epicontinental seas of NW Europe. Major water exchange at this time may have occurred along the line of the present English Channel (King 1989,

301

1993). Via the Norwegian-Greenland Sea to the north there may also have been a shallow water connection with the Arctic Ocean (Marincovich et al. 1990). This is supported by the similarity of post-Danian to early Eocene faunas of molluscs and ostracods in the Arctic Ocean with coeval North Sea Basin faunas. Thick sediment accumulations in the Lofoten Basin, west of northernmost Norway, indicate sediment contribution from the Barents Shelf, and an open connection also with the Barents Sea (Thiede et al. 1986). There may also have been a connection with the Tethys across North Germany and Poland into southern Russia (King 1989). The North Sea Basin and its northern extensions were surrounded by vast continental drainage areas (Fig. 1). Rivers from Greenland, Fennoscandia, Britain and central Europe all transported fresh water into this relatively small semi-enclosed sea. Despite this, the fossil microfauna and -flora throughout the R0sna~s Clay Formation reflects more or less typical marine conditions. The isotopic data also indicate almost fully marine conditions. The average salinity in the euphotic zone of the southern North Sea may have been around 2630 ppt through most of the early Eocene. These salinities are still high compared to recent seas like the Baltic Sea, Hudson Bay or Black Sea, that are similarly surrounded by vast land areas. Accordingly, a substantial ocean-water influx must have balanced the riverine input in the North Sea in the early Eocene. Probably saline waters entered continuously across epicontinental shallow water (< 200 m) connections, such as the vast Barents Shelf and the English Channel. However, deepwater connections, where the rich cosmopolitan, bathyal, benthonic foraminiferal fauna could enter, must have existed, at least intermittently. Oxygen isotopic data for the North Sea in the late Maastrichtian and early Danian (Schmitz et al. 1992), the early Selandian (Mattsson & Schmitz 1994) and the late Eocene/early Oligocene (Burman & Schmitz, unpublished data) indicate fully marine conditions (salinity c. 34 ppt) at these times. From the mollusc isotopic data of Buchardt (1978) it appears that, at least in the southern North Sea, reduced surface salinities existed during most of the Eocene. Most likely, the North Sea had unusually low salinities in the latest Paleocene, but the absence of biogenic calcite in sediments from this time makes this difficult to test. The reduced salinities measured at Alba~k Hoved may be representative of the entire North Sea, and possibly also the Norwegian-Greenland Sea (and perhaps the Arctic Ocean). Early Eocene microplankton assemblages at Alb~ek Hoved are similar to faunas and floras of other areas of the North Sea (Gradstein et al. 1992), indicating representative conditions at Alba~k Hoved. Based on studies of agglutinated

302

B. SCHMITZ E T AL.

foraminifera in ODP Hole 643 Kaminski et al. (1990) suggest a few per mil reduced salinity for surface waters in the Norwegian-Greenland Sea in the early Eocene. The RCsn~es Clay Formation at Alb~ek Hoved formed at considerable distance from any shore, which explains why it is very fine grained, resembling pelagic red clays. The closest shore was probably situated 300-400 km to the south in northern Germany. The benthonic foraminifera fauna as well as the stable isotope depth gradients indicate substantial water depths, possibly corresponding to a middle bathyal environment, which means depths in the range 600-1000m. The typically reddish brown colour of the sediment and neglible amounts of organic carbon (HeilmannClausen et al. 1985), indicate mostly oxygenated bottom conditions, which is consistent with voluminous inflow of ocean water. The decrease in benthonic 813C values in connection with episodes of calcite dissolution are indicative of higher CO 2 pressure and lower oxygen content in the bottom water at these times. Low oxygen content would have promoted formation of anoxic laminated clays, rather than red clays. The formation of red clays under such conditions, however, could have taken place if sedimentation rates were low. Such a general situation is indicated both by grain-size and the completeness of the biostratigraphic record. Moreover, although red-brown clays dominate, there is a large variation in the lithology across the RCsn~es Clay Formation. Some intervals indeed consist of greyish or laminated sediments, indicative of suboxic conditions. Pyrite occurs in many samples throughout the section. Probably the seafloor redox conditions balanced on the boundary between suboxic and oxic. Even a minor change in sedimentation rates or organic matter influx would have shifted sedimentation conditions in a new direction. The water flow across different sea ways connecting the North Sea with the open ocean may have varied with time, depending on palaeogeographical configurations and regional isostasy. In particular the isotopic profiles for Oridorsalis, but also for Subbotina, show a trend towards convergence with North Atlantic data in latest Biochron NP11 and early NP12. This indicates that the Faeroe and the English Channel connections may have been deepened or widened at this time, which may be related to subsidence of the British Isles and the Greenland-Scotland Ridge, or to a general sea-level rise in the northeastern Atlantic. In the middle and late Biochron NPll, during formation of unit D, water exchange with the North Atlantic may have been more restricted and other water passages may have been of more importance. The stable isotope composition at mid-depth as

well as in deep waters in the North Sea fluctuated significantly and depletions occurred in the heavy isotopes (13C and 180). The instability of the isotopic trends may reflect greater local or regional influence on the water-mass chemistry. Only little is known about the northward and the eastward connections of the North Sea. These connections may have been of greater relative importance for North Sea water chemistry at times when water exchange with the North Atlantic resumed. The sometimes very low 813C values compared with the North Atlantic, could reflect influx of CO 2- and nutrient-rich waters into the North Sea. Waters intermittently entering the North Sea from north via the Norwegian-Greenland Sea could also have been low in 813C, because of admixed magmatic CO 2 (813C c. -5%0) or organic carbon (813C c. -25%0) from river influx. There is no correlation between occurrences of volcanic ash in the RCsna~s Clay Formation and negative 813C, arguing against significant admixture of magmatic CO 2 with the water. Gas emissions, however, may have occurred in connection with episodes of flood basalt formation, which may not have given any clear imprints in the sedimentary record. The more negative 813C and 8180 values during middle and late Biochron N P l l compared with later could reflect a higher influence of water deriving from the Arctic Ocean or any newly evolved semienclosed basin to the north. Such basins and the semi-enclosed Arctic Ocean probably had water mass chemistries quite different from the world ocean (see Marincovich et al. 1990). The prominent negative isotopic excursions in late NP12 may be related to water exchange with such a semienclosed basins. A substantial influx of water from the eastern Tethys does not appear likely considering, for example, that keeled planktonic foraminifera, such as morozovellids, are absent from the RCsmes Clay Formation. These foraminifera occur at quite high latitudes on the eastern North Atlantic margin during the Eocene (King 1993). If voluminous eastern Tethys water masses entered the North Sea across Poland and Germany, one would expect to find in the North Sea abundant morozovellids, a characteristic element of Tethys waters. However, one may wonder why the morozovellids did not invade the North Sea from the North Atlantic across the English Channel. One possibility is that reduced surface salinities prevented the surfacedwelling morozovellids from invading.

Summary During the early Eocene, the North Sea was semienclosed with restricted contact with the open ocean. The benthonic foraminiferal fauna indicate

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK water depths of 6 0 0 - 1 0 0 0 m where present D e n m a r k lies. Throughout the early Eocene, conditions in the southern North Sea were essentially marine, but salinities were somewhat reduced. Average salinities in the upper 100 m of the water column may have fluctuated between 26 and 30 ppt. Water exchange with the open ocean determined the water mass properties in the North Sea, inducing principally three different types of sedimentation: non-calcareous, low-calcareous and calcareous. During periods of restricted water exchange, density stratification of low- and highsalinity waters was more rigid, inhibiting water circulation and effective replenishment of bottom waters. Low sea levels and associated basin compartmentalization added to this effect. Slow bottom-water renewal led to CO 2 build-up, more strongly developed vertical ~13C gradients between mid-depth and the seafloor and a rise in the lysocline. Biological productivity decreased due to reduced nutrient upwelling. Non-calcareous sediments occur in the lowermost part of the RCsna~s Clay Formation, reflecting restricted water exchange with the world ocean. This was followed in early Biochron NP11 by more open water exchange. Calcite-rich sediments with marine nannoplankton and Subbotina-dominated planktonic foraminiferal faunas began to form. In late N P l l , water exchange was somewhat restricted and calcite dissolution increased. For this interval, Subbotina 5180 trends are very unstable, with repeated positive excursions because of fresh water influence. Calcite-rich sedimentation began

303

again at the N P l l - N P 1 2 boundary and prevailed until late NP12. At this time Subbotina 5180 values are more positive with a more stable trend, indicating less freshwater admixture at mid-depth. At the end of the early Eocene there was a gradual return towards a more isolated North Sea. Three prominent lithological event beds occur in the upper NP12 part of the Rcsna~s Clay Formation. Associated with these beds are major negative shifts (1 to 2%,~) in carbon and oxygen isotopes. The chemistry of the entire water column changed, implying that important regional palaeoceanographic events are registered. It is possible that water masses entering the North Sea from the north (embryonic NorwegianGreenland Sea or Arctic Ocean) explain some of the differences in water mass chemistry between the North Sea and the Atlantic Ocean. The shortterm isotopic events in late N P I 2 may be related to rapid water exchange between the North Sea and other semi-enclosed basins with unusual water chemistry to the north. Financial support for this study was obtained from the Swedish Natural Science Research Council, The Bank of Sweden Tercentenary Foundation, and the Erna and Victor Hasselblad Foundation. Samples from DSDP Hole 550 were provided by the Deep Sea Drilling Project. We thank O. Gustafsson for isotopic analyses and E Asaro for iridium analyses. Laboratory assistance was provided by T. Alavi, T. Andinsson and J. Burman, and M. Eliasson did the artwork. Reviews by M.-P. Aubry, W. B. Berggren and E. Thomas greatly improved this paper. The paper is a contribution to IGCP Project 308.

References AUBRY, M.-E 1995. From chronology to stratigraphy: assessing the continuity of the stratigraphic record. In: BERGGREN,W. A., KENT, D. V., AUBRY,M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlations: Framework for an Historical Geology. Society for Economic Paleontologists and Mineralogists, Special Publication, 54. ~, BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD, K. D., OBRADOVICH,J. D. & PROTHERO, D. R. 1988. Paleogene geochronology: An integrated approach. Paleoceanography, 3, 707742. BARRERA,E. & HUBER,B. T. 1990. Evolution of Antarctic waters during the Maestrichtian: Foraminifer oxygen and carbon isotope ratios, Leg. 113. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 813-827. - - , - - , SAVIN,S. M. & WEBB, P.-N. 1987. Antarctic marine temperatures: Late Campanian through early Paleocene. Paleoceanography, 2, 21-47. BELANGER,P. E., CURRY,W. B. & MATTHEWS,R. K. 1981. Core-top evaluation of benthic foraminiferal isotopic ratios for paleoceanographic interpretations,

Palaeogeography, Palaeoclimatology, Palaeoecology, 33, 205-220. BERCGREN, W. A., & AUBRY, M.-P. 1996. A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network. This volume. -& SCHNITKER,D. 1983. Cenozoic marine environments in the North Atlantic and NorwegianGreenland Sea. In: BOTT, M. H., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland RidgeNew Methods and Concepts. Plenum, New York, 495-548. - - - , KENT, D. V. & P-LYNN, J. J. 1985. Jurassic to Paleogene: Part 2. Paleogene geochronology and chronostratigraphy. In: SNELL~G, N. J. (ed.) The Chronology of the Geological Record. Geological Society, London, Memoir, 10, 141-195. BOERSMA, A. & PREMOLI SILVA, I. 1983. Paleocene planktonic foraminiferal biogeography and the paleoceanography of the Atlantic Ocean. Micropaleontology, 29, 355-381. BOGGILD,O. B. 1918. Den vulkanske Aske i Moleret samt en Oversigt over Danmarks ~eldreTerti~erbj~ergarter.

304

B. SCHMITZ ET AL.

Danmarks Geologiske UndersCgelse, Raekke 2, 33. BOUVIER-SOUMAGNAC, Y. & DUPLESSY, J.-C. 1985. Carbon and oxygen isotopic composition of planktonic foraminifera from laboratory culture, plankton tows and recent sediment: Implications for the reconstruction of paleoclimatic conditions and of the global carbon cycle. Journal of Foraminiferal Research, 15, 302-320. BUCHARDT, B. 1978. Oxygen isotope palaeotemperatures from the Tertiary period in the North Sea area. Nature, 275, 121-123. CARO, Y. 1973. Contribution h la connaisance des dinoflagellrs du Palrocene-Eocene des Pyrenfes espagnoles. Revista Espanola de Micropaleontologia, 5, 329-372. --, LUTERBACHER, H., PERCH-NIELSEN, K., PREMOLISILVA, I., RIEDEL, W. R. & SANFILIPPO, A., 1975. Zonations h l'aide de microfossiles prlagiques du Palrocbne suprrieur et de l'l~ocbne infrrieur. Bulletin de la Socidt~ G~ologique de France, S~rie 7, 17, 125-147. CHARNOCK, M. A. & JONES, R. W. 1990. Agglutinated foraminifera from the Palaeogene of the North Sea. In: HEMLEBEN, C., KAMINSKI, M. A., KUHNt, W. & ScoTT, D. B. (eds) Paleoecology, Biostratigraphy, Paleoceanography and Taxonomy of Agglutinated Foraminifera. NATO Advanced Study Institutes, Series C, 327, 139-244. CORFIELD, R. M. & CARTLtDGE, J. E. 1991. Isotopic evidence for the depth stratification of fossil and recent Globigerinina: A review. Historical Biology, 5, 37-63. CORLISS, B. H. 1991. Morphology and microhabitat preferences of benthic foraminifera from the northwest Atlantic Ocean. Marine Micropaleontology, 17, 195-236. & CHEN, C. 1988. Morphotype patterns of Norwegian Sea deep-sea benthic foraminifera and ecological implications. Geology, 16, 71 6-719. & FoIs, E. 1991. Morphotype analysis of deep-sea benthic foraminifera from the northwest Gulf of Mexico. Palaios, 5, 589-605. CRAIG, H. & GORDON, L. I. 1965. Deuterium and oxygen 18 variations in the ocean and the marine atmosphere. In: TONGIORGI,E. (ed.) Stable Isotopes in Oceanographic Studies and Pateotemperatures. Consiglio Nazionale delle Ricerche Laboratorio di Geologia Nucleare, Pisa, Italy, 1-122. DE CONrNCK, J. 1990. Ypresian organic-walled phytoplankton in the Belgian Basin and adjacent areas. Bulletin de la Soci~t~ Belge de G~ologie, 97, 287-319. DUDLEY, W. C., B1ACKWELDER, P., BRAND, L. & DUPLESSY, J.-C. 1986. Stable isotopic composition of coccoliths. Marine Micropaleontology, 10, 1-8. ELDHOLM, O. 1990. Paleogene North Atlantic magmatictectonic events: Environmental implications. Memorie della Societ~ Geologica Italiana, 44, 13-28. & THOMAS, E. 1993. Environmental impact of volcanic margin formation. Earth and Planetary Science Letters, 117, 319-329. GRADSTEIN,E M. & BERGGREN,W. A., 1981. Flysch-type agglutinated foraminifera and the Maestrichtian to -

-

Paleogene history of the Labrador and North Seas. Marine Micropaleontology, 6, 211-268. ~ . , KRISTIANSEN,I. L., LOEMO, L. & KAMINSKI,M. A. 1992. Cenozoic foraminiferal and dinoflagellate cyst biostratigraphy of the central North Sea. Micropaleontology, 38, 101-137. GRAHAM, D. W., CORLISS, B. C., BENDER, M. L. & KEIGWlN, L. D. JR. 1981. Carbon and oxygen isotopic disequilibria of recent deep-sea benthic foraminifera. Marine Micropaleontology, 6, 483-497. GROSSMAN, E. L. 1987. Stable isotopes in modern benthic foraminifera: A study of vital effect. Journal of Foraminiferal Research, 17, 48-61. GRY, H. 1941. De istektoniske Forhold i Moleromraadet. Meddelelser fra Dansk Geologisk Forening, 9, 586-627. HARDENBOL, J. 1994. Sequence stratigraphic calibration of Paleocene and Lower Eocene continental margin deposits in NW Europe and the US Gulf Coast with the oceanic chronostratigraphic record. GFF, 116, 49-51. HEILMANN-CLAUSEN,C. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central Jylland, Denmark. Danmarks Geologiske UndersCgelse, Series A7. 1988. The Danish Subbasin: Paleogene dinoflagellates. Geologisches Jahrbuch Reihe A, 100, 339-343. -1990. Tertia~ret ved Alb~ek Hoved: Et fedt profil. Varv, 1990 (4), 99-118. --, NIELSEN, O. B. & GERSNER, E 1985. Lithostratigraphy and depositional environments in the Upper Paleocene and Eocene of Denmark. Bulletin of the Geological Society of Denmark, 33, 287-323. HULSBOS, R. E., KROON, D., JANSEN, H. S. M. & vAN HINTE, J.E. 1989. Lower Eocene benthic foraminifera and paleoenvironment of the outer V0ring Plateau, Norwegian Sea (DSDP Site 338). Micropaleontology, 35, 256-273. JAN DU CHI~NE, R., GORIN, G. & VAN STUIJVENBERG,J. 1975. Etude g6ologique et stratigraphique (palynologie et nannoflore calcaire) des Gr~s des Voirons (Pal6og6ne de Haut-Savoie, France). G~ologie Alpine, 51, 51-78. KAMINSKI, M. A., GRADSTEIN, E, GOLL, R. M. & GREIG, D. 1990. Biostratigraphy and paleoecology of deep-water agglutinated foraminifera at ODP Site 643, Norwegian-Greenland Sea. In: HEMLEBEN,C., KAMINSKI, M. A., KUHNT,W. • SCOTT, D. B. (eds) Paleoecology, Biostratigraphy, Paleoceanography and Taxonomy of Agglutinated Foraminifera. NATO Advanced Study Institutes Series C, 327, 345-386. KENNETT, J. P. & STOry, L. D. 1991. Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229. KING, C. 1981. The stratigraphy of the London Clay and associated deposits. Tertiary Research, Special Paper, 6, Backhuys, Rotterdam. 1989. Cenozoic of the North Sea. In: JENKINS,D. G. & MURRAY, J. W. (eds) Stratigraphical Atlas of Fossil Foraminifera. Ellis Horwood Limited, Chichester, 418-489. -

-

STABLE ISOTOPE AND BIOTIC EVOLUTION IN NORTH SEA, DENMARK 1993. The channel connection: Tropical foraminiferids in the early Eocene of the southern North Sea Basin. In: IGCP 308 Meeting: Correlation of the Early Paleogene in Northwest Europe, London, 1-2 December 1993, Programme and Abstracts. KNox, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141,993-999. -& MORTON,A. C. 1988. The record of early Tertiary N Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. , HARLAND,R. & ICdNG,C. 1983. Dinoflagellate cyst analysis of the basal London Clay of southern England. Newsletters on Stratigraphy, 12, 71-74. , MORTON, A. C. & HARLAYD, R. 1981. Stratigraphical relationships of Palaeocene sands in the UK Sector of the Central North Sea. ln: ILLING, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of North-West Europe, Heyden & Sons, London, 267-281. KROOPNICK, E 1974. The dissolved O2-CO2-13C system in the eastern equatorial Pacific. Deep-Sea Research, 21, 211-227. - 1980. The distribution of 13C in the Atlantic Ocean. Earth and Planetary Science Letters, 49, 469-484. - 1985. The distribution of 13C of ~CO 2 in the world oceans. Deep-Sea Research, 32, 57-84. KUHNT, W., MORLOTTI, E., WINKLER, W. & KAMINSKI, M. A. 1989. Alpine late Mesozoic and early Cenozoic deep-water sequences and their agglutinated foraminifera. Excursion Guidebook, Ttibingen, Germany. MCCORKLE, D. C., KEIGWIN, L. D., CORLISS, B. H. & EMERSON, S. R. 1990. The influence of microhabitats on the carbon isotopic composition of deep sea benthic foraminifera. Paleoceanography, 5, 161-185 MCKENNA, M. C. 1983. Cenozoic paleogeography of North Atlantic land bridges. In: BOTT, M. H., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge - New Methods and Concepts. Plenum, New York, 351-399. McNICHOL, A. E & DRUFFEL, E. R. M. 1992. Variability of the 813C of dissolved inorganic carbon at a site in the north Pacific Ocean. Geochimica et Cosmochimica Acta, 56, 3589-3592. MARGOLIS, S. V., KROOPNICK, P. M., GOODNEY, D. E. DUDLEY, W. C. & MAHONEY, M. E. 1975. Oxygen and carbon isotopes from calcareous nannofossils as paleoceanographic indicators. Science, 189, 555-557. MARINCOVICH,L. JR., BROUWERS,E. M., HOPKINS, D. M., & MCKENNA, M. C. 1990. Late Mesozoic and Cenozoic paleogeographic and paleoclimatic history of the Arctic Ocean Basin, based on shallow-water marine faunas and terrestrial vertebrates. In: GRANTZ, A., JOHNSON, L. & SWEENEY,J. E (eds) The Geology of North America,

305

Vol. L. The Arctic Ocean region. Geological Society of America, Boulder, Colorado, 403-426. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the 2nd Planktonic Conference, Rome 1970. Editizione Technoscienza, Rome, 739-785. MATTSSON, E. I. & SCHMITZ, B. 1994. Stable isotopic study of the basal type Selandian (Viborg borehole 5): preliminary results. GFF, 116, 58. MILLER, K. G., CURRY,W. B. & OSTERMANN,D. R. 1985. Late Paleogene (Eocene to Oligocene) benthic foraminiferal oceanography of the Goban Spur region, Deep Sea Drilling Project Leg 80. Initial Reports of the Deep Sea Drilling Project, 80, 505-538. MOOK, W. G. 1968. Geochemistry of the stable carbon and oxygen isotopes of natural waters in the Netherlands. PhD Thesis, Rijksuniversiteit te Groningen. MORKHOVEN, E P. C. M. VAN, BERGGREN, W. A. & EDWARDS, A. S. 1986. Cenozoic Cosmopolitan Deep-Water Benthic Foraminifera, Bulletin des Centres de Recherches Exploration-Production Elf Aquitaine, Memoir, 11, Pau, France. MORTON, A. C. & KNOX, R. W. O'B. 1990. Geochemistry of late Palaeocene and early Eocene tephras from the North Sea Basin. Journal of the Geological Society, London, 147, 425-437. NIELSEN, O . B . & HE1LMANN-CLAUSEN, C. 1988. Palaeogene volcanism: The sedimentary record in Denmark. In: MORTON,A. C. & PARSON,L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 395-405. OKADA, n. & HONJO, S. 1973. The distribution of oceanic coccolithophorids in the Pacific. Deep-Sea Research, 20, 355-374. OWEN, R. M. & PEA, D. K. 1985. Sea-floor hydrothermal activity links climate to tectonics: The Eocene carbon dioxide greenhouse. Science, 227, 166169. PEDERSEN, A. K., ENGELL,J. & RONSBO,J. G. 1975. Early Tertiary volcanism in the Skagerrak: New chemical evidence from ash layers in the mo-clay of northern Denmark. Lithos, 8, 255-268. - & SURLYK, E 1983. The Fur Formation, a late Paleocene ash-bearing diatomite from northern Denmark. Bulletin of the Geological Society of Denmark, 32, 43--65. RATHBURN,A. E. & CORLISS, B. H. 1994. The ecology of living (stained) deep-sea benthic foraminifera from the Sulu Sea. Paleoceanography, 9, 87-150. REA, D. K., ZACHOS, J. C., OWEN, R. M. & GINGERICH, E D. 1990. Global change at the Paleocene-Eocene boundary: Climatic and evolutionary consequences of tectonic events. Palaeogeography, PaiReDclimatology, Palaeoecology, 79, 117-128. ROBERTS, D. G., MORTON, A. C. & BACKMAN, J. 1984. Late Paleocene-Eocene volcanic events in the northern North Atlantic Ocean. Initial Reports of the Deep Sea Drilling Project, 81, 913-923. SCHMrrZ, B. 1988. Origin of microlayering in worldwide distributed It-rich marine Cretaceous-Tertiary boundary clays. Geology, 18, 87-94.

306

B. SCHMITZ ETAL.

1994. Iridium enrichments in volcanic ash layers from the r Formation (Lower Eocene) in Denmark. GFF, 116, 62. --, ANDERSSON, P. & DAHL, J. 1988. Iridium, sulfur isotopes and rare earth elements in the CretaceousTertiary boundary clay at Stevns Klint, Denmark. Geochimica et Cosmochimica Acta, 52, 229-236. --, ASARO, E, MICHEL, H. V., TH1ERSTEIN, H. R. & HUBER, B. T. 1991. Element stratigraphy across the Cretaceous/Tertiary boundary in Hole 738C. Proceedings of the Ocean Drilling Program, Scientific Results, 119, 719-730. --, KELLER, G. & STENVALL,O. 1992. Stable isotope and foraminiferal changes across the CretaceousTertiary boundary at Stevns Klint, Denmark: Arguments for long-term oceanic instability before and after bolide-impact event. Palaeogeography, Palaeoclimatology, Palaeoecology, 9 8 , 233-260. SHACKLETON,N. J. & HALL, M. A. 1984. Carbon isotope data from Leg 74 sediments. Initial Reports of the Deep Sea Drilling Project, 74, 613-619. -& KENNETT, J. P. 1975. Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: Oxygen and carbon isotope analyses in DSDP sites 277, 279, and 281. Initial Reports of the Deep Sea Drilling Project, 29, 743-755. , CORFIELD, R. M. & HALL, M. A. 1985. Stable isotope data and the ontogeny of Paleocene planktonic foraminifera. Journal of Foraminiferal Research, 15, 321-336. --, HALL, M. A. & BOERSMA, A. 1984. Oxygen and carbon isotope data from Leg 74 foraminifers. Initial Reports of the Deep Sea Drilling Project, 74, 599-612. SIESSER, W. G., WARD, D. J. & LORD, A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage (Paleocene) in the type area. Journal of Micropaleontology, 6, 85-102. STEURBAUT, E. 1988. New Early and Middle Eocene calcareous nannoplankton events and correlations in middle to high latitudes of the northern hemisphere. Newsletters on Stratigraphy, 18, 99-115. 1991. Ypresian calcareous nannoplankton biostratigraphy and palaeogeography of the Belgian Basin. In: DuPuIS, C., DE CONINCK, J. & STEURSAUT, E. (eds) The Ypresian stratotype. Bulletin de la Socidt~ Beige de Gdologie, 97 (1988), 251-285. -1995. Calcareous nannoplankton, a major key for developing high-resolution biochronologies and for unravelling Earth's Tertiary geological history. Nouvelles de la Science et des Technologies, 13 (2-3), 4p. Brussels.

& KING, C. 1994. Integrated stratigraphy of the Mont-Panisel borehole section (151E340), Ypresian (Early Eocene) of the Mons Basin, SW Belgium. Bulletin de la Socidtg Belge de Gdologie, 102 (1-2) (1993), 175-202. - & NOLF, D. 1986. Revision of Ypresian stratigraphy of Belgium and Northern France. Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie, 23, 115-172. SWISHER,C. C. & KNOX, R. W. O'B. 1993. Single-crystal laser-fusion Ar4~ 39 dating of Early Eocene tephra layers from the North Sea Basin: Calibration points for the Paleogene time-scale. Correlation of the Early Paleogene in Northwest Europe, Programme and Abstracts. Geological Society, London, 1-2 December 1993. THIEDE, J. & ELDHOLM, O. 1983. Speculations about the paleodepth of the Greenland-Scotland Ridge during late Mesozoic and Cenozoic times. In: BOTT, M. H., SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge - New Methods and Concepts. Plenum, New York, 445-456. - - . , DIESEN, G. W., KNUDSEN,B.-E. & SNARE, T. 1986. Patterns of Cenozoic sedimentation in the Norwegian-Greenland Sea. Marine Geology, 69, 323-352. , N I E L S E N , O . B . & PERCH-NIELSEN, K. 1980. Lithofacies, mineralogy and biostratigraphy of Eocene sediments in northern Denmark (Deep Test Viborg 1). Neues Jahrbuch fiir Geologie und Paliiontologie Abhandlungen, 160, 149-172. THOMAS, E. & SHACKLETON,N, J. 1996. The PaleoceneEocene benthic foraminiferal extinction and stable isotope anomalies. This volume. VINCENT, E., KILLINGLEy,J. S. & BERGER, W. H. 1981. Stable isotope composition of benthic foraminifera from the equatorial Pacific. Nature, 289, 639-643. WHITE, R. S. 1988. A hot-spot model for early Tertiary volcanism in the N Atlantic. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 395-405. WOODRUFF, E, SAVIN, S. M. & DOUGLAS, R. G. 1980. Biological fractionation of oxygen and carbon isotopes by recent benthic foraminifera. Marine Micropaleontology, 5, 3-11. ZIEGLER, P. A. 1988. Evolution of the Arctic-North Atlantic and the Western Tethys. American Association of Petroleum Geologists, Memoir, 43. -1990. Geological Atlas of Western and Central Europe (second edition). Shell Internationale Maatschappij B.V., Den Haag, Netherlands. - -

A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network W. A. B E R G G R E N

1 & M.-E

AUBRY 2

1 Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA 2 Institut des Sciences de l'Evolution, Universit~ Montpellier II, 34095 Montpellier Cedex 5, France Abstract: Published and unpublished data on bio-, chemo- and magnetostratigraphic events spanning the late Paleocene-early Eocene are reviewed and calibrated to a revised Geomagnetic Polarity Time Scale (GPTS), itself already in need of revision. This timescale serves as a template for placing the upper Paleocene-lower Eocene stratigraphic succession of NW Europe (Anglo-Paris-Belgian Basins) in a sequence stratigraphic framework. It is clear that an approach that integrates sequence- and magnetobioisotope stratigraphy provides a unifying correlation framework within which to delineate the disjunct depositional histories of marginal and deep marine basinal stratigraphies and allows the stratigrapher to move from rock to time. Significant conclusions of this integration include the following: 1. The major climatic warming, weakening of atmospheric circulation and faunal extinction events, which are seen in the deep sea stratigraphic record, are seen to be closely associated with the mid-part of the (redefined) planktonic foraminiferal Zone P5 and calcareous nannoplankton Zone NP9. 2. A relatively rapid (c. 1000 years) decrease of c. 3 to 4 per mil in the ~13C of marine carbonates has now been recognized in high southern and northern latitude sites and at several scattered, intermediate locations. A similar excursion has been observed in mammalian teeth and soil carbonates near the base of the Wasatchian North American Land Mammal Age in the Big Horn Basin a few metres above the Argile plastique bariolre ('type' Sparnacian) in the Pads Basin. This 813C excursion is seen to be a truly global event which occurs in both marine (c. midZone NP9) and terrestrial stratigraphies at a level estimated at c. 55.5 Ma. 3. An evaluation of the calcareous plankton biostratigraphy and stable isotope records at several deep sea sites suggests the presence of multiple unconformities in the Paleocene/Eocene boundary interval. There would appear to be no unequivocally demonstrated continuous stratigraphic section through the c. 2.55 million year interval of Chron C24r in the deep sea record or from outcrops on land, and a composite record is required to construct the sequence of events that occurred during this interval. 4. The Paleocene/Eocene boundary (which awaits the determination of a Global Stratigraphic Section and Point [GSSP]) is bracketed by the base of the leper Clay Formation (= Ypresian Stage), estimated here at > 54.6 Ma, and the P5/P6a zonal boundary (LAD Morozovella velascoensis) at c. 54.7 Ma (above), and by the ~13C spike (and associated events) in midZone NP9 at c. 55.5 Ma (below). The 'boundary interval' encompasses the NP9/10 zonal boundary at 55 Ma and the base of the Harwich Formation (base of the London Clay Formation Oldhaven Beds of previous authors) at c. 54.8 Ma. 5. Calcareous nannoplankton studies of recently recollected Thanetian localities in the eastern part of the Paris Basin have revealed that the Thanetian Sables de Ch~lons-sur-Vesles are older (Zones NP6 and NP7) than the Sables de Bracheux s.l. (= Sables du Tillet; Zone NP8). This leads to a revised framework of correlation between the upper Paleocene formations of northwest Europe. 6. The base of the Sparnacian base, taken as the Paleocene/Eocene boundary by most vertebrate palaeontologists, is correlative with a level within Zone NP9, c. 0.5 million year older than the NP9/10 zonal boundary (used by some marine micropalaeontologists to delineate the Paleocene/Eocene boundary) and > one million years older than the base of the Ieper Clay (base Ypresian) in Belgium.

The marine stratigraphic succession d e v e l o p e d in the sedimentary basins on the passive margins o f the Baltic Plain surrounding the North Sea (AngloBelgian-Paris-Danish-North G e r m a n Basins) has

c o m e to s e r v e as the standard for m o n d i a l P a l e o g e n e cbronostratigraphy. I G C P Project 308 ( P a l e o c e n e / E o c e n e B o u n d a r y Events in Space and Time) is currently examining representative marine

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 309-352.

309

310

W.A. BERGGREN & M. E AUBRY

and terrestrial stratigraphic sequences spanning this boundary in an attempt to provide criteria suitable for locating and positioning a global boundary stratotype section and point (GSSP). Within the framework of this project we present a comparative anatomy and correlation of the stratigraphic sequences in this region spanning an approximately six million year interval from early Thanetian to early Ypresian. In the first part of this paper we review the calcareous plankton and benthic foraminiferal biostratigraphy, as well as stable isotope and magnetostratigraphic data in deep sections spanning the Paleocene-Eocene boundary interval in order to erect a global magnetobiostratigraphic correlation network within which to place the standard NW European stratigraphic succession. We have examined samples from some deep sea drilling holes (in particular 549, 550 and 690) and have evaluated the published calcareous microfossil stratigraphies for others. The main lithostratigraphic units are then placed in a geochronological framework based on a newly constructed Geomagnetic Polarity Time Scale (GPTS) (Cande & Kent 1992, 1995) which is derived from an analysis of (primarily) South Atlantic seafloor anomaly profiles in which anomaly spacings are constrained by finite rotation poles and averages of stacked profiles. The revised magnetochronology was generated by using a spline function to fit a set of nine calibration points (including anchoring calibrations of 55.0 Ma on the supposed NP9/NP10 zonal boundary at/near the Paleocene/Eocene boundary and 66.0 Ma (Cande & Kent 1992; 65 Ma in Cande & Kent 1995) at the K/P boundary plus the zero age ridge axis to the composite polarity sequence. This magnetochronology has served as the template for a revised magnetobiochronology of the standard calcareous biostratigraphy of the Paleocene/Eocene boundary interval, based primarily on data from DSDP Hole 550 in the NE Atlantic, and for the Paleocene based primarily on a magnetobiostratigraphic record recently obtained at DSDP Hole 384 (NWAtlantic). Cross-correlation of the recently 4~ dated -17 and +19 ashes (of Denmark and North Sea area) to Chron C24r in Hole 550 and to the base of the Wrabness Member (ex Harwich Member) and Hales Clay of the Harwich Formation (ex London Clay division A1 - see Ellison et al. 1994; Jolley 1996), respectively, of the London-Hampshire Basin provides a direct tie between the deep sea and marginal marine record at the Paleocene/Eocene boundary. We then place the stratigraphic succession of NW Europe and the North Sea in a revised sequence stratigraphic framework and correlate this record to the newly revised geochronology. This

makes it possible to estimate the duration of the hiatuses associated with the major/minor unconformities separating the various lithic units, to situate the associated terrestrial mammalian levels within a chronostratigraphic framework and to make reasonable estimates of the timing/duration of major prochoresis/faunal turnover events.

Paleocene/Eocene b o u n d a r y sections In this section we present a review of data on bio-, magneto- and isotope stratigraphy and climatic events spanning the Paleocene/Eocene boundary observed in a number of deep sea drilling sites. Chronology is that of Berggren et al. (1985) with exceptions noted below. In the second part of the paper dealing with geochronology of the Paleocene/Eocene boundary and with NW European chronostratigraphy we switch to a revised chronology based on the newly developed Geomagnetic Polarity Time Scale (GPTS) by Cande & Kent (1992, 1995). With regard to the discussion of planktonic foraminiferal biostratigraphy below it is important to point out that a basic modification to planktonic foraminiferal Zones P5 and P6 (spanning the Paleocene/Eocene boundary) has been made in the revised Paleogene integrated magnetobiochronology of Berggren et al. (1995). This has been necessitated by the recognition that Zones P5 and P6a as defined in Berggren & Miller (1988) are essentially equivalent owing to the concurrent range of Morozovella subbotinae and M. velascoensis following the LAD of Globanomalina pseudomenardii, nominate taxon of the underlying Zone P4. The newly defined/emended zonation is presented here so that the reader may make the transition from the zonal definition of Berggren & Miller (1988) and that currently in use in the review presented below (Fig. 1). The magnetochronology used in the section on zonal definitions (below) is that of Berggren et aI. (1995) and it is taken directly from that paper. Zonal definitions from Berggren et al. (1995) Zone P5. Morozovella velascoensis. Interval Zone (Bolli 1957; P5 and P6a of Berggren & Miller 1988) Definition: biostratigraphic interval between the LAD of Globanomalina pseudomenardii and the LAD of Morozovella velasacoensis. Magnetostratigraphic calibration: Chron C25n(y)Chron C24r (midpart). Estimated age: 55.9-54.7 Ma. Remarks: Zone P5 with a different denotation

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE

Berggren & Miller (1988)

311

Berggren et al. (1995)

_.

Zones

Criteria

Z one

P7

P7

i.!

_2_

P6c

6b

8 P6b

P6a -v- -

T :,~

P6a r~

P5

P5

r~ ,.M

D

P4

!

P4

Fig. 1. Comparison of biostratigraphic criteria used in defining planktonic foraminiferalZones P5-P7 in Berggren & Miller (1988) and revised zonation used in this paper and Berggren et al. (1995).

(partial range of the nominate taxon between the LAD of Gl. pseudomenardii and the FAD of Morozovella subbotinae) was defined (Berggren & Miller 1988) before definitive evidence of the juxtaposition/overlap in range of Gl. pseudomenardii and M. subbotinae, nominate forms for Zone 1'4 (top) and P6 (base), respectively, became available (cf. Blow 1979, p. 265-267, although we would disagree with Blow on the upper limit of pseudomenardii in his Zone P7 = P6b of Berggren & Miller 1988). However, in some instances, the FAD of M. subbotinae has been observed to be delayed because of the widespread occurrence of (a) distinct dissolution event(s) that span(s) the lower part of magnetozone 24.3r and Zone P6a. Thus the chronology of a zone based on the FAD of M. subbotinae is obviously very approximate. For this reason we have decided to revert to the previous, relatively unequivocal usage of Bolli (1957) in which two sequential LADs are used to define a distinct interval which would appear to span the Paleocene/Eocene boundary as currently recognized by at least some (bio)stratigraphers. As a result of this modification to Zone P5, Zone P6 of Bergren & Miller (1988) is also modified. Zone P5 (as revised) essentially contains the concurrent range of M. subbotinae (FAD) and M. velascoensis

(LAD), but it is defined as an interval zone because the definition of the top of P4 is the LAD of Gl. pseudomenardii. Characteristic features of this subzone include the relatively closely spaced appearances of Morozovella marginodentata, M. formosa gracilis, Igorina broedermanni, Acarinina wilcoxensis, Turborotalia pseudoimitata, and the relatively common occurrence of strongly muricate 'large acarininids' (soldadoensis, coalingensis-triplex group). The Paleocene/Eocene boundary is usually correlated with the P5/P6 (= P6a/P6b of Berggren & Miller 1988) boundary by planktonic foraminiferal specialists and is estimated here at 54.8 Ma (cf. discussion in Berggren et al. 1995, on radioisotopic calibrations for the Paleogene chronology adopted here). Calcareous nannoplankton specialists usually consider the NP9/10 zonal boundary as definitive for this boundary. The base of the Harwich Formation (Oldhaven Beds = Hales Clay; = base Eocene = Thanetian/Ypresian boundary in some usage) has been cross correlated to the -17 ash in DSDP Hole 550 and dated in NW Europe at 54.5 Ma (see discussion in Berggren et al. 1995). However, we choose to use an age of 54.8 Ma for the base of the Harwich Formation

312

w.A. BERGGREN • M. P. AUBRY

based on sedimentation rates in DSDP Hole 550, rather than on the -17 ash date. This is discussed in greater detail below. The base of the leper Clay (= base Ypresian s.s.) in Belgium is located one fourth order cycle higher in the stratigraphic record with an estimated age of c. 54.6 Ma here (see discussion below). The problems associated with the identification and delineation of events suitable for the determination of an appropriate Paleocene/ Eocene boundary GSSP are being currently examined by IGCP 308 (Paleocene/Eocene Boundary Events in Time and Space) and are discussed in greater detail below and elswhere (Aubry et al. 1996).

manner that Zone P6 (as redefined in Berggren

et al. 1995) corresponds to Subzones P6b and P6c of Berggren & Miller (1988, p. 371).

Subzone P6a. Morozovella velascoensis-Morozovella formosa formosa and/or M. lensiformis. Interval Subzone (P6b of Berggren & Miller 1988; emended in Berggren et aL 1995) Definition: biostratigraphic interval between the LAD of Morozovella velascoensis and the FAD and/or of Morozovella formosa formosa M. lensifonnis Magnetostratigraphic calibration: mid to late part of Chron C24r.

Age estimate: 54.7-54 Ma; earliest Eocene (earliest Zone P6. Morozovella subbotinae. Partial Range Zone (redefined in Berggren et al. 1995) Definition: biostratigraphic interval characterized by the partial range of the nominate taxon between LAD of Morozovella vetascoensis and FAD of

Morozovella aragonensis Magnetostratigraphic calibration: Chron C24r (midpart)-Subchron C23n.2r (earliest part).

Estimated age: 54.7-52.3 Ma. Remarks: Berggren & Miller (1988, p. 370) defined the Morozovella subbotinae Partial Range Zone (P6) as the partial range of the nominate taxon between the FAD of the nominate taxon and that of M. aragonensis. Investigations on a number of deep sea sites and outcrop sections have now shown that the FAD of M. subbotinae essentially coincides with the LAD of Gl. pseudomenardii and that the supposed stratigraphic gap between the LAD of Gl. pseudomemardii and the FAD of M. subbotinae is illusory, although delayed entry of the latter taxon is often caused by strong dissolution in sections within the upper part of Zone NP9. Thus P5 and P6a, as defined by Berggren & Miller (1988, p. 370) are essentially equivalent. As emended here, Zone P6 coincides essentially with the M. subbotinae Zone of Premoli Silva & Bolli (1973) and Luterbacher & Premoli Silva in Caro et al. (1975) except for the (apparently brief) temporal interval between the LAD of M. velascoensis and the LAD of the small, enigmatic taxon M. edgari (Primoli Silva & Bolli 1973) (? = M. finchi Blow 1979) which was shown by Toumarkine & Luterbacher (1985, fig. 5, p. 100) to occur only slighly below the simultaneous FADs of M. formosa formosa and M. lensiformis. (cf. Blow 1979, figs 48 and 50 in which the LAD of M. finchi is shown to occur at essentially the same level). The temporal span between the LAD of M. velascoensis and the FAD of M. formosa formosa is estimated here at c. 0.8 million years In order to maintain numerical and biostratigraphic continuity with the zonation of Berggren & Miller (1988), we redefine Zone P6 in such a

Ypresian).

Remarks: In open ocean stratigraphic successions the sequence of FAD of M. subbotinae, LAD of M. velascoensis acuta, FAD of M. formosa formosa/M, lensiformis and FAD of M. aragonensis serve as a means of providing a discrete biostratigraphic subdivision of Zone P6 (Berggren & Miller 1988) and that subdivision is followed here with minor modification. This subzone has been redefined as an interval subzone to avoid conceptual confusion/overlap with the use of M. subbotinae as nominate form of both Zone P6 and Subzone P6a in its original definition (see Berggren & Miller 1988, p. 370, 371) and for Zones P6 and Subzone P6a (emended in Berggren et al. 1995). At the same time we remove Pseudohastigerina wilcoxensis as one of the nominate forms of Subzone P6a (P6b in Berggren & Miller 1988); this form appears under ideal conditions at the P5/6 zonal boundary in fully tropical assemblages but has a demonstrably delayed entry in mid-high latitude regions within the P6b-7 (or correlative) biostratigraphic interval. The LAD of Subbotina velascoensis occurs within this subzone. Additional comments on this subzone (as P6b) are to be found in Berggren & Miller (1988, p. 371).

Subzone P6b. Morozovella formosa formosa/ M. lensiformis-Morozovella aragonensis. Interval Subzone (P6b, emended in Berggren et al. 1995 = P6c of Berggren & Miller 1988, p. 371; P8a of Blow 1979). Definition: biostratigraphic interval between the essentially simultaneous FADs of M. formosa and/ or M. lensiformis and the FAD of Morozovella aragonensis. Magnetostratigraphic calibration: Chron C24r (late) to Chron C23r.

Estimated age: 54-52.1 Ma; early Eocene (early Ypresian).

Remarks: This is a biostratigraphically distinct interval (see also Primoli Silva & Bolli 1973; Blow

LATE PALEOCENE-EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE 1979) characterized by the essentially simultaneous FADs of the n o m i n a t e taxa and the L A D s o f

M. subbotinae, M. marginodentata and M. aequa.

Review of data from DSDP-ODP sites In the discussion below, we utilize the revised chron nomenclature of Cande & Kent (1992, 1995) as well as that of Berggren et al. (1985) where appropriate. Correlation of nomenclature and estimated ages of the Chron C24n to C25n are shown below: Berggren et al. 1985 Nomenclature

Estimated age (Ma)

Subchron C24 (younger) (--- Subchron C24A) Subchron C24 (older) (= Subchron C24B) Chron C25n

55.14-55.37 55.66-56.14 58.64-59.24

Cande & Kent (1995) Nomenclature

Estimated age (Ma)

Subchron C24n. In Subchron C24n.2n Subchron C24n.3n Chron C25n

52.364--52.663 52.757-52.801 52.903-53.347 55.904-56.391

Site 5 4 9 ( G o b a n Spur, Irish continental margin, N E Atlantic) (Fig. 2) Biostratigraphic data (Snyder & Waters 1985 and our own re-examination of material from this site) indicate that Zones P5-P7 (or their correlative biostratigraphies) are represented at this site. Globanomalina peudomenardii is present in the stratigraphic interval encompassed by 549/20/1-5 (c. 370376 m), absent in 549/19/2 (c. 361+ m) as well as in the overlying stratigraphic interval between cores 19/1 and 16/6 (c. 360-340 m) which is characterized by strong dissolution (Snyder & Waters 1985, 448, 449, fig. 6); a lone occurrence is noted at 16/5:57-60 cm (c. 338 m) located in the middle part of a reversed interval (= Chron C24r), in the upper part of Zone NP9 (Mtiller 1985; Aubry et al. 1996) and c. 3 m below the FAD of Morozovella subbotinae, M. marginodentata and Acarinina wilcoxensis at c. 335 m (associated with an unconformity; see below). Chron C25n is situated between c. 349.5-354 m (within the dissolution interval and is associated with the NPS/NP9 boundary). The LAD of pseudomenardii may lie within the dissolution interval and near the Chron C25n interval; we have not found Gl. pseudomenardii despite a diligent search of samples above the dissolution interval in cores 16 and 17. Thus the P4/5 zonal boundary cannot be determined precisely at this site.

313

The FAD of Morozovella lensiformis denotes the Subzone P6a/P6b boundary (Berggren et al. 1995). In Hole 549 the first occurrence of this taxon (Snyder & Waters 1985) is within a c. 8 m interval between samples in core 15R/1 and 15R/6 (c. 332-340 m). There is an unconformity at c. 335 m separating upper Zone NP9 from upper Zone NP10 and it is most likely that M. lens(formis (and Subone P6b) extends down to the level of the unconformity. This is consistent with the LAD of M. aequa between core 15R/6 AND 14R/6 (c. 331 m: Snyder & Waters 1985) which occurs in nearby Hole 550 in Subzone P6b). It is also consistent with correlations of the 13C record in Holes 549 and 550 (Aubry et al. 1996; Stott et al. 1996) which make sense over the stratigraphic interval of c. 335-332 m in Hole 549 and c. 380-370 m in Hole 550, within upper Zone NP10. The FAD of Morozovella aragonensis occurs in Core 14R between c. 313 and 320 m and delineates Zone P6/P7 boundary. At sites 549 and 550, the P6a/b zonal boundary occurs in the upper part of Chron C24r and within Zone NPI 1. The FAD of Pseudohastigerina wilcoxensis occurs in Core 14R/1 (c. 313 m) in the mid-part of Subzone P6b, c. 20-21 m above that of M. subbotinae and Ac. wilcoxensis in Core 16. Elsewhere the FAD of Ps. wilcoxensis has been shown to occur within the uppermost part of Zone NP9 (e.g. Site 605), or NP10 (e.g. Sites 401, 690). It is difficult to determine whether this difference is due to diachrony or taxonomic discrimination of the planispiral pseudohastigerinid morphology. The equivocal nature of the palaeomagnetic records at these sites adds to the problem as well. The calcareous nannofossil succession in the NP9NP10 zonal interval is discussed in detail in Aubry et al. (1996). These authors show the presence of two successive unconformities at and close to the NP9/NP10 zonal boundary as indicated by the close stratigraphic succession of the LAD of Fasciculithus tympaniformis (at 335.50 m), the FAD of Tribrachiatus bramlettei (at 335.30 m) and that of T. contortus (Morphotype B, see Aubry et al. 1996) (at 335.16 m). The presence of unconformities near the Paleocene/Eocene boundary in Hole 549 is supported by the absence in Hole 549 of the series of more than 55 bentonitic ashes (distal equivalents of North Sea-Danish ash series) which occur in the lower to mid-part of Zone NP10 in Hole 550 (from c. 386-403 m) and by studies on the benthic foraminiferal fauna of Hole 549 by Reynolds (1992, MSc Thesis, Univerisity of Maine, Orono) which suggest that c. 12.6 m of sediments equivalent to the lower part of Zone NP10 are missing at c. 335 m. Principal component analysis of the benthic foraminiferal fauna (Reynolds 1992, MSc Thesis, University of Maine, Orono) reveals a definite Paleocene fauna (PC2), a definite Eocene fauna (PC1), and two short-lived transitional faunas. Allowing for a hiatus of c. 0.8 million years at c. 335.40 m (near the NP9/10 boundary), the rate of faunal turnover was shown to be essentially synchronous in the high latitude North Atlantic and Antarctic (Thomas 1990). We would interpret the data of Reynolds to indicate that the LAD of the Stensoina beccariformis fauna occurs between 336 m and 332 m and is associated with the hiatuses at c. 335.40 m and 335.22 m; sporadic specimens of St. beccariformis above this level are interpreted as having been reworked.

314

w . A . BERGGREN & M.P. AUBRY

Fig. 2. Stratigraphic framework of events spanning the Paleocene/Eocene boundary: DSDP Hole 549.

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE The 813C spike has been identified at 338 m in Hole 549 (Sinha & Stott 1993a, b; Stott etal. 1996), only 2.6 m below the unconformity (c. 335.40 m) which separates (upper) Zone NP9 from Zone NP10 (see Aubry et al. 1996).

Site 5 5 0 ( P o r c u p i n e A b y s s a l Plain, S W o f s e a w a r d edge o f G o b a n Spur, N E Atlantic) (Fig. 3) Subchrons C24n.ln, C24n.3n and ?Chron C25n (partim) bracket a relatively long (62.85m) Chron C24r (Townsend 1985) making this a supposedly ideal site for detailed, integrated magnetobiostratigraphy (Table 1). Calcareous nannofossil (MUller 1985) and planktonic foraminiferal (Snyder & Waters 1985) data indicate the presence of Zones NP9-NP12 and P4 (equivalent), P5, P6a, P6b and P7, respectively (Tables 3-5). We have examined material from Hole 550 kindly provided by S. Snyder as well as material studied for stable isotopes by A. Sinha and L. Stott and our results are incorporated here in the text as well as in tables below. The FAD of D. lodoensis (346.73m) ( = N I l / 1 2 = CP9/10 zonal boundary) occurs in Subchron C24n.l-2r, consistent with earlier records. The stratigraphic range of Tribrachiatus contortus (which defines Subzone CP9a of Okada & Bukry (1980) in the upper part of Zone NP10 of Martini (1971) was shown by Mailer (1985) to span a 40.62 m-thick interval of Chron C24r, between 406.35 m and 365.72 m. Reexamination of the NP9-NP10 zonal interval (Aubry et al. 1996) has shown that the range of T. contortus is restricted to the upper part of Zone NP10 between 381.9m and 378.4m (Morphotype A) and between 372.65 m and 375.72 m (Morphotype B).

315

The NP9/NP10 zonal boundary occurs between 4 0 8 . 0 2 m and 407.75 m based on the FAD of Tribrachiatus bramlettei at this latter level (Aubry et al. 1996). The LAD of Fasciculithus tympaniformis at 408.02 m immediately precedes the FAD of T. bramlettei. Integration of calcareous nannofossil and carbon isotope stratigraphies (Aubry et aL 1996) shows that the NP9/NP10 zonal contact is unconformable, and that the unconformity corresponds to the lithological contact at 408 m. The sequential FADs of M. subbotinae (407 m in Snyder & Waters 1985; 409 m, this paper; but see below), Ac. wilcoxensis (404.5 m), M. formosa gracilis, Igorina broedermanni (395 m) and LAD ofM. acuta (396 m), and FAD of M. lensiformis (377 m) and M. aragonensis (341 m) support recognition of Zones P5, P6a, P6b and P7 respectively. The P5/P6a and P6a/b zonal boundaries lie within the mid- and upper part of chronozone C24r, respectively. However the FAD of M. subbotinae is certainly delayed owing to the presence of an intense dissolution zone which spans the interval of cores 36/2 to 34/4 (early Citron C24r interval). The P6a/b boundary (using a midpoint of 377 m for the FAD ofM. lensiformis) lies in the upper part of chronozone C24r. The LAD of the Stensioina beccariiformis benthic foraminiferal fauna occurs between samples 35/1: 62-65 cm and 34/4: 62-6 cm (413.62-408.65 m; personal observation, WAB) within a dissolution facies (characterized by abundant radiolarians) which spans the interval from Core 36/2 to Core 34/5 (c. 425-410 m), in lower Chron C24r, within Zone NP9 and (apparently) mid-Zone P5. Snyder & Waters (1985, fig. 6) place the FAD of M. subbotinae in sample 34/2:62-65 cm (405.65 m), c. 3 m above the top of the dissolution facies. Our study (WAB) of samples from cores 33-35 indicates that specimens transitional between M. aequa and

Table 1, Estimated (minimum) thickness of Chron C24r in some representative deep sea sections Hole

Depth (m)

Thickness (m)

577

79.95-85.45

5.50

690B

185.42-135 (uncons

50.42

549

317.5-350

37.5

550

359.65-422.50

62.85

605

557-598-7

>41.3

Remarks

Stratigraphic section possibly missing at/between core breaks of core 9 and 10 (= 82.8 m) and perhaps in lower part of core 9 (absence/nominal representation of Zone CP9a (c. 81 m). Probably minimum estimate. 'Normal events' between 154-144 m interpreted as C24R based on biostratigraphy. Unconformity between NP11/12 at about 135 m in reversed interval interpreted as C23R/C24R boundary. Estimate is considered a minimum. >56 bentonitic ashes in NP10 at Site 550 not found in NP10 at 549. NP10 unusually short (c. 4 m) compared to NP11 above (c. 30 m) suggesting lower NP10 missing at Site 549. Estimate considered a minimum. Unconformity at 408 m between NP9 and NP10. Strong dissolution facies from 422-410 m (_=_Zone NP9). Estimate considered a minimum. No data above c. 557 m within C24R, NP10 and CP9a. Estimate is minimum value.

Data source: DSDP/ODP Initial/Science Reports

316

W.A. BERGGREN & M . E AUBRY

Fig. 3. Stratigraphic relationship of events spanning the Paleocene/Eocene boundary: DSDP Hole 550.

M. subbotinae and referable to M. subbotinae occur in the lowest sample examined (34/4:62-65 cm = 408.65), c. 1.5 m above the dissolution facies. This sample belongs to Zone P5 and the LAD of the St. beccariiformis fauna would be consistent with (most) occurrences elswhere, including several outcrop sections discussed below. The benthic foraminiferal event in Hole 550 consists in the replacement of a relatively small, but typical, Paleocene

fauna characterized by i.al., Stensioina beccariiformis, Cibicidoides velascoensis, Nuttallinella florealis (up to sample 3 5 / 1 : 6 2 - 6 5 c m = 4 1 3 . 6 1 m), by an influx of minute abyssaminids, pleurostomellids, pulleniids and Cibicidoides eocaenus. (sample 34/4:62-65 m = 408.65 m). Nuttallides truempyi is a characteristic component of both the late Paleocene and early Eocene faunas. Dissolution is so intense in the intervening 5 m

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE that all calcareous benthic and planktonic foraminifera have been dissolved. Thus if our interpretation is correct the FAD of M. subbotinae is a minimum estimate and the LAD of the St. beccariiformis fauna is a maximum estimate. The true FAD of M. subbotinae may lie lower within the dissolution facies interval; indeed it probably does inasmuch as its FAD has been recorded in Chron C25n in Hole 577A by Corfield 1987 and by Liu & Olsson (pers. comm. 1992); see also Miller et al. 1987; p. 744, fig. 2, who record its FAD in mid-Chron C24r (based on lower sampling density and/or perhaps slightly different taxonomic concept). The true LAD of the St. beccariiformis fauna may lie within the interval of intense dissolution represented by samples 34/6: 62-65cm and 3 4 / 5 : 6 2 - 6 5 cm (411.62-410.12m) in which no benthic fauna was observed. A series of c. 40-50 volcanic ash beds span the interval of Cores 27-32, but are primarily developed in Cores 32-33 (c. 384.85-404.6 m; Fig. 3). These ashes are distal representatives of phase 2a and 2b ashes known from the North Sea and NW Europe, particularly Denmark. The distinctive -17 and +19 ash beds occur within the range of Tribrachiatus bramlettei, bracket the P5/P6a zonal boundary and lie approximately one-third up within Chron C24r. The -17 ash occurs in Zone P5, at the same level (400 m) as the FAD of M. marginodentata, 6 m above the FAD of M. subbotinae 18.1 m below the FAD of Morphotype A of Tribrachiatus contortus, 27.35 m below that of Morphotype B of this species, and c. 22.5 m above the Chron C24r/?C25n boundary. The +19 ash occurs at c. 393 m in Zone P6a. It is located c. 2 m above the FADs of M. formosa gracilis and Ac. broedermanni, 4 m above the LAD of M. acuta (397 m), 11.1 m below the FAD of Morphotype A of T. contortus, 20.35 m below that of Morphotype B of this species, and at c. 29.5 m above the Chron C24r/?C25n boundary. Recent 4~ dates of 54.5 and 54.0 Ma have been obtained on the -17 and +19 ash Beds, respectively (see discussion in Berggren et al. 1995; and below in this paper). Owing to the intense dissolution in the lower part of Chron C24r in Hole 550 it has proven impossible to obtain a complete ~13C record; however a portion of the major excursion has been recognized at c. 409 m (Sinha & Stott 1993a, b; Stott et al. 1996), c. 1 m below the unconformity delineated in Aubry et al. (1996) in what we would interpret as the mid-part of the truncated Zone NP9. Finally, the inability to derive a precise chronology in Hole 550 is due to the fact that the normal event identified at c. 425-435 m may not be Chron C25n. At Hole 549 a rhyolitic ash layer occurs at 352.66 m in the upper part of Chron C25n and basal Zone NP9 (Knox 1985). This ash does not occur at the corresponding level in Hole 550 at the base of NP9. It is uncertain whether the magnetic reversal between 419.51-425.5 m (Snyder & Waters 1985; Townsend 1985) corresponds to the Chron C25n/C24r boundary as indicated by Snyder & Waters 1985). The lowest occurrence of D. multiradiatus, normally associated with a level within Chron C25n, is at 424.85 m (Mtiller 1985) in an (unsampled) interval of unknown polarity. In addition the sharp lithological boundary at 426.5 m between Units 2b and 3a (Graciansky et al. 1985) occurs within Zone NP5, not at

317

the NP5/NP9 zonal boundary. Thus the upper 50 cm of the normal polarity interval which extends from 425.5434.04 m may not represent Chron C25n. We suspect that the upper surface of the unconformity between Zones NP5 and NP9 in Hole 550 is younger than the Chron C25n/C24r reversal and that Chron C25n is not represented in Hole 550. If this is true, we have lost an important calibration point in estimating the chronological position of the -17 and +19 ashes and the 813C spike in this hole (see further discussion below and in Aubry et al. 1996).

Site 401 ( M e r i a d z e k Terrace, B a y o f Biscay, N E A t l a n t i c ) (Fig. 4) Relatively incomplete recovery, sparse sampling and lack of stratigraphic range data (Krasheninnikov 1979) preclude accurate delineation of biostratigraphic datum events and zonal boundaries. Neither Tribrachiatus bramlettei nor T. contortus are reported from Hole 401. Zone NP10 in this hole is delineated by Mtiller (1979, table 7) based on the LAD of Fasciculithus tympaniformis (between 204.90 m and 204.15 m) and the FAD of T. orthostylus (between 193.75 m and 193.08 m). The association of Morozovella velascoensis, M. subbotinae and Ps. wilcoxensis in Core 13 CC (196 m) is used to differentiate Zones P6a and P6b which are equivalent to P5 and P6a here, and occur within Zone NP10. The FAD of M. formosa (core 13,191 m) delineates the P6a/P6b zonal boundary within Zone N P l l (cf. Aubry et al. 1988; Berggren & Miller 1988). The FAD of M. aragonensis in the lower part of core 12 delineates the Zone P6b/P7 boundary, approximately equivalent to the NPll/NP12 zonal boundary (cf. Aubry et al. 1988; Berggren & Miller 1988). The LAD of the St. beccariiformis fauna occurs between 202 m and 202.60 m (Schnitker 1979; Pak & Miller 1992) and is thus located in the lower part of Zone NP10 as delineated at its base by the LAD of E tympaniformis. (We recognize, however, that this constitutes only an approximation of this base which would perhaps more appropriately be taken at the FAD of D. diastypus at c. 198.5 m.) If the calcareous nannofossil biozonal interpretation is correct, we infer the presence of an unconformity in core 14, section 3, between 202.60 m and 202.15 m from the close association of the benthic foraminiferal extinction and the NP9/NP10 boundary. Sporadic bentonitic ashes (equivalent to those in Hole 550) occur at c. 195-196 m within Zone NP10 near the level of the LAD of M. velascoensis and FAD of M. subbotinae and above the LAD of the St. beccariiformis benthic foraminiferal association, generally consistent with observations elsewhere.

Site 6 0 5 ( N e w J e r s e y C o n t i n e n t a l Rise, NW Atlantic) Rather sparse sampling (Saint-Marc 1987) precludes precise determination of planktonic foraminiferal datum

318

W.A. BERGGREN & M . E AUBRY

Fig. 4. Stratigraphic relationship of events spanning the Paleocene/Eocene boundary: DSDP Hole 401.

levels/zonal boundaries, whereas relatively close sampling allows more accurate delineation of the calcareous nannofossil zonal boundaries (Applegate & Wise 1987; Lang & Wise 1987). The NP9/NP10 zonal boundary, as recognized by the FAD of T. bramlettei, lies between 558 and 559.5 m (Applegate & Wise 1987, table 1). Zone NP9 is c. 43 mthick whereas Zone NP10 is less than 9 m-thick. The FAD of Discoaster diastypus is immediately above the boundary.

The LAD of GL pseudomenardii is located within core 46 in a reversed interval identified as Chron C24r, c. 28 m above a normal event identified as Chron C25n, and within the middle of Zone CP8 (= Zone NP9). The virtually simultaneous LAD of M. velascoensis and FAD of M. subbotinae (c. 577 m) and subsequent (c. 562 m) FADs of Ps. wilcoxensis and M. formosa gracilis, the LAD of M. acuta in the upper part of Zone CP8 (= Zone NP9) and the FADs of M. marginodentata and Ac. wilcoxensis (577 m) support distinction of Zones

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE P5 and P6a at c. 577 m within a reversed polarity interval assigned to Chron C24r. The extreme thinness of Zone P5 suggests an unconformity at c. 577 m. There are no planktonic foraminiferal data above core 43 CC (c. 557 m) owing to poor preservation. The LAD of the St. beccariiformis fauna (core 44, c. 565 m) ocurs within the upper part of Zone P6a and lies within the upper part of Zone NP9 in Chron C24r, consistent with observations elsewhere.

Hole 690B (SW Flank of Mause Rise, S o u t h A t l a n t i c ) (Fig. 5) The absence of tropical (low latitude) markers precludes direct correlation with standard planktonic foraminiferal biostratigraphic zonal scheme(s); instead a high latitude zonation was developed by Stott & Kennett (1990). In contrast, the standard calcareous nannofossil zonation of Martini (1971) is applicable to this site/region (see Aubry et al. 1996). The FAD of Globanomalina australiformis (in core 19H, at 170.64 m in Stott & Kennett 1990, p. 556 but at 171.38 m, in Stott & Kennett 1990, p. 555) is closely associated with the LAD of St. beccariiformis fauna at 170 m (Thomas 1990) and two short normal polarity 'events' within the mid-part of Chron C24r. By correlation with the sequence of events at other sites this level should be stratigraphically equivalent to a level within Zone P5 (see discussion of Hole 550 above). The FAD of Ac. wilcoxensis berggreni (core 18H, c. 159 m) and the sporadic occurrence of Ps. wilcoxensis in core 16 supports assignment of the 'normal-reversed' polarity sequence of cores 16H and 17H to Chron C24r (and not to Subchrons C24An and C24Bn (Berggren et al. 1985; = Subchrons C24n.1n to 3n; Cande & Kent 1992). The interpretation of the magnetic polarity stratigraphy for the upper Paleocene-lower middle Eocene section in Hole 690B is discussed in Aubry et al. (1996). Because the magnetic polarity stratigraphy in Hole 690B is so important to regional and global correlations across the Paleocene/Eocene boundary, we comment further on the integration of magneto- and biostratigraphy in this interval. A detailed magnetic polarity stratigraphy was developed in the lower Paleogene by Spiess (1990). A series of short normal events was recorded in this expanded stratigraphic section over the interval of cores 16H-17H (top) and correlated with Subchrons C24An and C24Bn of Berggren et al. (1995) (Spiess 1990, fig. 9, pp. 288, 289 and fig. 10, p. 292; using the new terminology of Cande & Kent (1992), this would correspond to Subchron C24n.ln and Subchron C24n.3n, respectively). A hiatus (of c. 1.5 million years duration) was shown to be present within core 15H eliminating the younger part of Chron C23n to the uppermost part of Chron C22r (Spiess 1990, p. 290) based on magnetobiostratigraphic age-depth relationships and sedimentation rates. The Chron C23r/C24n boundary was shown to lie at the normal event in the basal part of core 15H. By contrast Stott & Kennett (1990) and Pospichal & Wise (1990, fig. 4, p. 632) identified the normal polarity events in mid-

319

core 15H as Chron C22n entirely (we would identify these two normal events as Chrons C22n and C23n, respectively, and place an unconformity between the two, see Aubry et al. 1996), and another unconformity at c. 137.8 m between Zones NP10 and NPll. Kennett & Stott (1991, p. 555) indicated that the magnetic polarity stratigraphy of Hole 690B between cores 17H and 25H matches clearly the standard magnetostratigraphic scale of BKF85. Reference to their fig. 2 (1990, p. 552), however, shows that Chron C24r is drawn from mid-17H to 21H (top), whereas Spiess (1990, fig. 9, p. 289 and appendix B, p. 312-314) shows that the long reversed interval of Chron C24r extends from mid-17H to base 21H (not to 23H (Stott & Kennett 1990, p. 555) or top 21H (Stott & Kennett 1990, fig. 2)). On the assumption that the base of the series of normal events at c. 154.2 m and the top of the normal event at 185.2 m are Subchron C24n.3n and Chron C25n, respectively, Stott & Kennett (1990:555) estimated the position of the Paleocene/ Eocene boundary at c. 173.2 m (Berggren et al. 1985 = 57.8 Ma) or at 165 m (Aubry et al. 1990 ; 57.0 Ma) using the correct depth for the Chron C24r/C25n of 185.48 m (see Stott etal. 1990, p. 851, table 1; Spiess 1990, p. 279, table 4). The former level is near the FAD of G1. australiformis (171.38 m) and the LAD of the St. beccariiformis benthic faunal assemblage at c. 170 m (Thomas 1990). The latter lies approximately half way between the FAD of Gl. australiformis (171.38 m) and the FAD of Ac. wilcoxensis berggreni at 159.45 m. In contrast, Pospichal & Wise (1990) drew the Paleocene/Eocene boundary at c. 150m (core 17H/2: 28-30 cm), at the level of the CP8/CP9 zonal boundary that they delineate based on the FAD of T. bramlettei, not on that of Discoaster diastypus (the marker for the zonal boundary) at 148.89 m. The 150 m level (= base NP10) lies within the series of short 'normal' events correlated by Spiess (1990, fig. 10, p. 292) with Subchrons C24An and C24Bn (Berggren et al. 1985). Pospichal & Wise (1990) report the LAD of T. contortus (= CP9a/b boundary = NP10/NPll boundary) at 137.4 m (but see Aubry et al. 1996), in the basal part of the reversed interval identified as Chron C23r by Spiess (1990). However, the association of Zone NP10 with normal polarity events identified as Chron C24n is completely anomalous inasmuch as Zones NP10 and CP9a are known to lie within Chron C24r and the NP10/NPll = CP9a/b boundary (= LAD of T. contortus) to occur a short distance below the Chron C24n/C24r boundary (see Berggren et al. 1985 and discussion above and below). The identity/reality of the series of normal 'events' in cores 16H and 17H remains to be resolved. Pending resolution of this discrepancy we consider the magnetic polarity signatures in this part of the stratigraphic record as questionable (possibly a result of normal overprinting) and not suitable for magnetobiostratigraphic correlation and/or calibration. A marked 813C excursion has been identified over an apparently brief (c. 0.02 million years interval) centred at 170 m (essentially contemporaneous with the FAD of Globanomalina australiformis and the LAD of the St. beccariformis benthic foraminiferal association: Thomas 1990; Thomas et al. 1990; Thomas & Shackleton 1996) with an age estimated of 57.3 Ma (Berggren et al. 1985; Stott et al. 1990; Kennett & Stott 1991). This level

320

w . A . BERGGREN (~ M. P. AUBRY

Fig. 5. Stratigraphic relationship of events spanning the Paleocene/Eocene boundary: ODP Hole 690B.

was considered just to predate the Paleocene/Eocene boundary in Hole 690B, based on a magnetobiostratigraphically interpolated age (from the Subchron C24n.3n/C24r boundary) of 57.0 Ma for the 166 m level (Kennett & Stott 1991). However it will be seen that the 166-171 m interval in Hole 690B (spanning the benthic foraminiferal extinction event, the carbon isotope

excursion and the Paleocene/Eocene boundary as estimated by Berggren et al. (1985) at 57.0 Ma) lies well down within Zone NP9, c. 20 m below the NP9/NP10 boundary. Elsewhere Zone NP9 lies within the mid-to lower part of chronozone C24r. The PaleoceneflEocene boundary (revision by Aubry et al. 1988: defined as the base of the London Clay Formation, not approximated by

LATE PALEOCENE-EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE the NP9/NP10 zonal boundary) with an estimated age of 57.0 Ma (Berggren et al. 1985) was suggested to lie within mid-upper Zone NP10, approximately correlative with the CP8/CP9 zonal boundary (= FAD of D. diastypus), which was believed earlier (Berggren et al. 1985) to correspond to a level in the mid to upper part of chronozone C24r. The reviews presented above (Holes 549, 550) are seen to compound the complexity of the issue because of the presence of an unconformity at both locations, rendering it difficult to determine the precise position of the ~13C excursion, the NP9/NP10 zonal boundary and the -17 ash within Chron C24r, and hence to establish a precise chronology for Chron C24n as well as for the Paleocene itself (see discussion below and in Aubry et al. 1996). The benthic foraminiferal extinction and carbon isotope events in Hole 690B are clearly older than 57.3 Ma in the magnetobiochronology/correlationof Aubry et al. (1988). The estimated age of the 170 m and 166 m levels in Hole 690B are c. 57.6 and c. 57.4 Ma, respectively, if one uses only the duration/length of Chron C25n (assuming that the entire normal interval between 185.48 m and 195.94 m represents Chron C25n) and extrapolates upward (using Berggren et al. 1985). The latter estimate is still well within the interval of Zone NP10 in the chronology of Aubry et al. (1988) indicating further problems with the calcareous nannoplankton biostratigraphy and/or the magnetostratigraphy of Hole 690B. The estimate of c. 57.4 Ma for the benthic foraminiferal extinction event is within previous estimates of the chronology of the lower part of Zone NP10, but this event has been shown to be linked with mid-Zone NP9 here in Hole 690B.

Site 577 (Shatsky Rise, NW Pacific) (Fig. 6 ) There are a relatively large number of datum events in both the planktonic foraminifera (Corfield 1987; Liu & Olsson, pers. comm. 1993) and calcareous nannoplankton (Monechi 1985) and a good magnetostratigraphy (Bleil 1985) spanning the interval of c. 75-85 m in Hole 577 and 70-80 m in Hole 577A, which allows the development of an integrated magnetobiostratigraphy (Monechi et aL 1985). However, it should be borne in mind that the c. 2.5 million year interval of Chron C24r is only c. 5 m thick at Site 577 precluding a clear separation of biostratigraphic events spanning the Paleocene/Eocene boundary. Unless otherwise specified the following data refer to Hole 577. Data on the relationship between the LAD of Globanomalina pseudomenardii and FAD of Morozovella subbotinae, nominate taxa for the Zone P4/5 and P5/6 boundaries, respectively, in the zonal scheme of Berggren & Miller (1988), are somewhat equivocal. Liu & Olsson (pers. comm. 1993 and verified by WAB by examination of their samples) indicate the stratigraphic overlap of these two taxa within Chron C25n, whereas Corfield (1987) indicated their overlap at the top of Chron C25n and basal Chron C24r. These observations have led to the revision by Berggren et al. (1995; see also above) to the Zone P4-6 part of the zonal scheme of Berggren & Miller (1988). Miller et al. (1987, p. 744, fig. 2) record the FAD

321

of M. subbotinae in mid-Chron C24r and the LAD of Gl. pseudomenardii in Chron C25n (the latter being consistent with both Corfield (1987) and Liu & Olsson (pers. comm. 1993)), indicating the presence of a 3--4 mthick Zone P5. However, sample spacing in these studies was relatively large; this was improved upon in the study by Pak & Miller (1992). The overlap of M. subbotinae and M. velascoensis (Zone P6a in Berggren & Miller 1988; Zone P5 of this work) over the interval of c. 86-82 m (upper core 10) extends to slightly above the midpoint of the reversed interval identified as Chron C24r, in close proximity to the LAD of Fasciculithus spp. (81 m), considered generally to occur at, or near, the top of Zone NP9 (= CP8). The LAD of M. velascoensis at c. 82-83 m is estimated to be at c. 57.8Ma (Berggren et al. 1985) by Miller et al. (1987, p. 745, table la) and 56.63 Ma by Corfield (1987, p. 100, table 6.2), apparently by interpolation between magnetostratigraphic datum levels (Bleil 1985). The benthic foraminiferal extinction occurs between 82 and 83 m. However, we believe that there is truncation at the level of the extinction inasmuch as a number of events occur near this level and there is a suggestion that Subzone CP9a is missing in this sequence (Monechi 1985; Monechi et al. 1985, fig. 2, who place an unconformity at the LAD of F. tympaniformis). However, Pak & Miller (1992) emphasize that closer sampling indicates that the LAD of M. velascoensis is within 577/10/1 and not between cores 9 and 10. The FAD of M. formosa (base of Zone P6b, this work) at c. 81 m (lower part of core 9) occurs just below a normal event identified as Subchron C24n.3n, consistent with earlier records (Berggren et al. 1985; Berggren & Miller 1988). The juxtaposition of the nominate criteria for the top of Zone P5 (LAD M. velascoensis: c. 82-83 m) and the base of Zone P6b (FAD M. formosa) in the upper half of Chron C24r reflects truncation by an unconformity. This is well supported by calcareous nannofossil stratigraphy. The NP10/NP11 zonal boundary is almost coincident with the Subchron C24n.3n/Chron C24r reversal (compare with Aubry et al. 1988). This, and the fact that the FAD of Discoaster kuepperi is at the same level as the FAD of Tribrachiatus orthostylus (see Monechi 1985, table 5), indicate that only the upper part of Zone N P l l is present. The mean age of the LAD of M. velascoensis has been estimated at 56.79 Ma (in the chronology of Berggren et al. 1985) based on data from Hole 577 and Site 527 (Corfield 1987), i.e. in the upper half of chronozone C24r, although this younger age estimate relative to Berggren et al. (1985) may reflect (in part) on the presence of unconformities at these sites. At Site 605 (see above) the LAD of M. velascoensis is well down within chronozone C24r, c. 23 m above the top of chronozone C25n. It may prove necessary to adjust the age of the chronozone P5/P6a boundary to younger within Chron C24r, but by what amount remains unclear. At Site 550 (above) we have used the LAD of M. acuta (just below the midpoint of chronozone C24r) as a proxy for the P5/P6 zonal boundary. The FAD of M. aragonensis (Zone P7) at c. 79 m (core 9) is associated with a normal event identified as Subchron C24n.3n, somewhat older than its suggested correlation with Subchron C24n.l-2n based on data from

322

W.A. BERGGREN & M . E AUBRY

rz Ig3

O

r 9

o

ID

g, ex0

~ ,....~

9O ,...i

z r .,,.~ ,.t=

,d

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE the Gubbio section(s) (Berggren et al. 1985) and basal Subchron C23n.2r at Sites 550 and 549 (see above). At Site 527 the FAD of this taxon is shown to correlate with Subchron C23n.2r (Corfield 1987), but identification of the magnetostratigraphy at that site is not unequivocal. The correlation of the FAD of M. aragonensis (Zone P6/P7) with Subchron C24n.3n, if correct, would suggest that Zone P6b is relatively short (c. 0.3 million years). The FAD of D. multiradiatus at c. 86 m is correlative with the upper part of chronozone C25n, consistent with some records elsewhere, although there are records of a correlation with the lower part of chronozone C25n (Site 605; see also Berggren et aL 1985). At Hole 690B this datum event is correlated to a level in uppermost chronozone C25n, virtually at the subchronozone C24n.3r/ chronozone C25n boundary. The lone, rare occurrence of T. contortus in sample 577/9/5:117 cm (c. 81 m, in the reversed section just below Subchron C24n.3n and only slightly above the FAD of M. formosa (= Zone P6b) precludes precise recognition of the T. contortus Subzone (CP9a) of the D. diastypus (CP9) Zone (Monechi 1985, p. 307; Monechi et al. 1985, p. 794). The latter authors point out, however, that the LAD of Fasciculithus spp. occurs slightly below (c. 82 m) the lone occurrence of T. contortus and within the upper part (but not top) of Zone CP8 (= NP9), consistent with records in the South Atlantic (Shackleton et al. 1984). The relatively consistent LAD of Fasciculithus spp. within the lower part of chronozone C24r suggests that Subzone CP9a spans the middle to upper part of chronozone C24r, inasmuch as the LAD of Tr. contortus (nominate form for the CP9a/b boundary) is consistently reported from the upper part of chronozone C24r (Berggren et al. 1985; Monechi et al. 1985; see also above). Indeed Monechi et al. (1985, p. 793; table 3) estimated the duration of the T. contortus (CP9a) Subzone at 0.3 million years (56.3-56.0 Ma) at Site 577, although it is difficult to understand how this estimate was made inasmuch as the taxon is recorded only from a single sample at Hole 577 and does not occur in Hole 577A (Monechi 1985). Also, these authors indicate their belief that Subzone CP9a is absent at Site 577 (Monechi et al. 1985, p.789, fig. 2). The FAD of D. lodoensis is located at a level (Subchron C23n.2r) just above a normal event in both Hole 577 (74.6 m) and Hole 577A (74.43 m) identified as Subchron C24n. 1-2n, slightly younger than its reported association with the top of subchronozone C24n. 1-2n in the Gubbio section(s) (Berggren et al. 1985) but consistent with other reports from subchronozone C24n.2r (Berggren et al. 1985) and Hole 550 (above). A major (c. 3) 6~3C shift (lowering) is seen to essentially span the Chron C25n to C24n interval with a midpoint between 83-81 m in the middle of Chron C24r (within Zones NP9 and P5 as currently recognized) at c. 57.8 million years (Shackleton et al. 1985; Miller et al. 1987; chronology of Berggren et al. 1985) and assuming that the sequence is essentially continuous in Hole 577 (see remarks above). A sequential series of benthic foraminiferal extinctions occur over the interval of 89-83 m, but the major St. beccariiformis faunal extinction event is located between 83 m and 82 m apparently, in close proximity to

323

the carbon isotope shift and the Zone P5/P6a boundary (Miller et al. 1987, p. 749, 750; fig. 3; Pak & Miller 1992).

Sites 698, 700 (NE Georgia Rise) and 702 (Islas Orcadas Rise) (Figs 7, 8) Incomplete recovery precludes precise delineation of biostratigraphic subdivisions at these sites in the Atlantic region of the Southern Ocean (Crux 1991; Nocchi et al. 1991; Katz & Miller 1991). The interval between 235.7 m and 198 m in Hole 702 is characterized by the occurrence of Tribrachiatus orthostylus without Discoaster lodoensis Crux 1991). As this latter species occurs at c. 191.7 m (at a level which yields D. lodoensis and D. sublodoensis (Cieselski, Kristoffersen et al. 1988), thus assignable to Subzone NP14a), its absence between 235.7 m and 198 m indicates that the interval belongs to Zones NP10 and NP11 and does not include Zone NP12. Furthermore, since T. contortus and T. bramlettei occur at other high latitude sites (e.g. ODP Site 690), their absence in this hole can be taken to indicate Zone N P t l . This interpretation is supported by the similar correlations between planktonic foraminifera and calcareous nannofossil zones seen in Holes 549 and 550. The situation is similar at Sites 698 and 700. The FAD of Gl. australiformis at c. 70.5 m in Hole 698B, 220 m in Hole 700B and 245 m in Hole 702B allows approximate correlation of these levels with the 170m level in Hole 690B which occurs near the St. beccariiformis benthic foraminiferal extinction event and the carbon isotope shift, and in mid-Chron C24r. In Holes 698A and 702B the FAD of Gl. australiformis (Nocchi et al. 1991, p. 247, fig. 9) coincides closely with the LAD of F. tympaniformis (used as proxy for a level equivalent to the Zones NP9/10 zonal boundary) and the FAD of Tr. orthostylus (Crux 199l:169) which supports our interpretation of an unconformity at this level. The St. beccariiformis faunal extinction event is approximately correlative with the FAD of Gl. australiformis, the LAD of Fasciculithus spp. and lies within a major carbon isotopic excursion (as seen elsewhere) which would be equivalent to a level within chronozone C24r (see remarks above). A significant minimum peak in the carbon isotope record at c. 195 m (core 21) in Hole 700B within Zone NP 11 (Zone NP10-NP12 undifferentiated in Crux 1991) may be correlative with one of the two peaks associated with Chron C24n in Hole 577 and the composite Bottacione-Contessa Highway sections (Corfield et al. 1991).

The benthic foraminiferal extinction event in some outcrop sections We have examined the sequence of biostratigraphic events which occur in the calcareous nannoplankton (MPA) and planktonic and benthic foraminifera (WAB) in four outcrop sections from Spain, Israel and the North Caucasus (Georgia) in order to assess the relationship of the extinction of the St. beccariiformis fauna to calcareous

613CpDB o I

•

1 I

BFE

I N truompyj

Fig. 7. Stratigraphic relationship of events spanning the Paleocene/Eocene boundary at ODP Hole 698A.

2 J

325

LATE PALEOCENE-EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

,

_

,.~ r

~"

I~ o o

1

,:::::,

2 ,

ti m = .~

3

6

i,

~13CpDB

} I 0! 1

t 2.

1 1

2 I

3 ,1 J

?

~205.8 ~ i ~

!

t t:: I

/

r

E

/ 2 2 , 1 8 - -

B

~ / j

r

i

/

r /

//

234.3 ~ C4"

1

,~= ~...,.v-.~ ~ ~ 9 ~,

2430,

BFE

J

~ ~

/

/ ! P6a "

253.3-

---j /

7

'

/

Cibicidoides

spp.

-~. ~

9

N_ truempy~

TT

,/ /

// //

/

/ )" / /

272.3

z

o

Fig. 8. Stratigraphic relationship of events spanning the Paleocene/Eocene boundary; ODP Hole 702B.

326

w . A . BERGGREN ~; M. P. AUBRY

planktonic biostratigraphy. Our results are reviewed below.

Zumaya section (NW Spain; samples kindly furnished by E. Molina and correspond to those used in the study by Canudo & Molina 1992). In this deepwater flysch facies the LAD of the St. beccariiformis fauna occurs (sample 18) within Zone NP9, essentially equivalent to the top of the concurrent range of M. subbotinae and M. veIascoensis-acuta (= Zone P5/6 boundary) and within a dissolution facies (see also Canudo & Molina 1992 where this event is said to occur in sample 18, c. 26 m above the base of the local base line of this section). Calcareous microfossils are generally strongly recrystallized and poorly preserved. Caravaca section (SE Spain; samples kindly provided by E. Molina; another set of samples has been donated by Shell International Oil Company). In this more carbonate rich section the LAD of the St. beccariiformis fauna apparently occurs near the base of a dissolution interval (within which planktonic foraminifera are dissolved and calcareous nannoplankton are too heavily recrystallized to allow definitive identification), within the concurrent range of M. subbotinae and M. velascoensis-acuta (= Zone P5) and within Zone NP9. Zin Valley section (Negev, Israel; samples kindly provided by C. Benjamini). In this section the LAD of the St. beccariiformis fauna occurs in a dissolution interval characterized by a flood of radiolarians (extensively developed in the Tethyan seaway and extending from the Mediterranean to Caucasus region), at the top of the concurrent range of M. subbotinae and M. velascoensisacuta (= Zones P5/6 boundary) and (apparently) upper Zone NP9. Khieu River section (NW Caucasus, Georgia; samples kindly provided by N. Muzylov). The LAD of the St. beccariiformis fauna occurs in sample 321, within Zone P5 (c. 10 m above the FAD of M. subbotinae), within a dissolution facies spanning most of Zone NP9 characterized by silicified radiolarians and a sporadic agglutinated benthic foraminiferal fauna, c. 20 m below the FAD of Ps. wilcoxensis (= Zone P6a) and c. 28 m below the FAD of M. lensiformis (= Zone P6b) in basal Zone NP11. The results from a study of these four widely separated sections spanning the upper Paleocene-lowermost Eocene are consistent and indicate a close association between the LAD of the St. beccariiformis fauna and a level within Zones P5 and NP9.

Discussion of DSDP/ODP and onshore data There is a generally consistent correlation of the St. beccariiformis benthic foraminiferal faunal association extinction event to a level within Zone P5 and with a level within Zone NP9, the mid-point of the carbon excursion event and a level within the lower third of chronozone C24r. In

Hole 549 the St. beccariiformis faunal association extinction essentially coincides with the FAD of M. subbotinae and is closely bracketed by the LADs of F. tympaniformis (below) and the FAD of T. bramlettei (above), at an unconformity between Zones NP9 and NP10 and P5 and P6. In Hole 550 this event occurs in a dissolution facies just below an unconformity separating Zone NP9 and NP10 and within Zone P5. In Hole 690B this event coincides with the 513C spike and the FAD of Gl. australiformis and occurs within mid-Zone NP9 and Chron C24r. In Hole 577 these events are separated by a few metres and a stratigraphic gap occurs at the NP9/10 zonal boundary. Correlation of these events to southern high latitudes is afforded by the FAD of Gl. australiformis which occurs in Zone NP9 in Hole 690B at the level of the St. beccariiformis assemblage extinction and the carbon isotope shift, whereas in Hole 702B it appears to coincide with the Zones N P 9 / 1 0 boundary (as delineated by the L A D of E tympaniformis) and the FAD of T. orthostylus (the juxtaposition of which suggests an unconformity). The North Atlantic ash series occurs within the lower half of Zone CP9a and NP10. It spans the upper Zone P5 to the lower half of Zone P6a in Hole 550. The - 1 7 and +19 ashes, recently dated at 54.5 and 54.0 million years, respectively, lie within the lower part of Zone NP10 and CP9a, a short distance above the L A D of F. tympaniformis and the FAD of T. bramlettei, and bracket the Zone P5/P6a zonal boundary a short distance above the (delayed) entry of M. subbotinae and the FAD of Ac. wilcoxensis. The two ashes also bracket the FADs of M. formosa gracilis and Ig. broedermanni (below) and the L A D o f M. acutavelascoensis (above). The position of these events within Chron C24r is the remaining fly in the ointment, their determination being hampered by the presence of unconformities at Sites 549 and 550 and the lack of a reliable magnetostratigraphy at Hole 690B. The magnetobiostratigraphic correlation of the NP9/NP10 zonal boundary remains problematic but crucial to the questions surrounding the placem e n t of the P a l e o c e n e / E o c e n e boundary. The problem is exacerbated by, and probably owing to factors that include: (a) d i s c o n f o r m i t i e s / paraconformities within chronozone C24r juxtaposing various biostratigraphic datums; (b) inconsistent boundary determination owing to difficulty of determining the L A D of F. tympaniformis, uncertainty of the relationship of the base of Zone NP9a (= FAD of Discoaster diastypus) and the base of Zone NP 10 ( - F A D of T. bramlettei). An interpretation of the composite records of the magnetobiostratigraphic correlations in Holes 549, 550 and 690B (Aubry et al. 1996) suggests that the

LATE PALEOCENE-EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

various events observed near the Paleocene/Eocene boundary (i.e. carbon isotope minimum spike, oxygen isotope warming spike, Stensioina beccariiformis benthic foraminiferal association extinction, reduction in grain-size of wind blown dust) are closely aligned and occur within Zone P5 and within mid Zones CP8 and NP9. Romein (1979, fig. 49) demonstrated an evolutionary sequence linking T. bramlettei-contortusorthostylus in the Nahal Avdat section of Israel where these forms replace each other over a 4 m interval (fig. 13, p. 31; fig. 14, p. 34, 35). A similar sequence was observed in DSDP Hole 550 in the upper part of Zone NP10 (see above). However, two morphotypes of T. contortus (or two different species of Tribrachiatus) with disjunct ranges occur in this 62 m-thick section. The evolutionary sequence concerns only Morphotype B. The short range of Morphotype A is entirely comprised within that of T. bramlenei and no evolutionary link between the two taxa is apparent. These and related taxonomic problems as well as magnetostratigraphy and magnetochronology of the Chron C25n-C24r interval are crucial in deriving an integrated magnetobiochronology for Biochrons NP9 and NP10. If one uses the FAD of D. multiradiatus (sample 102) and the FAD of D. lodoensis (sample 123) in the Nahal Avdat section, equates them with the mid-point of chronozone C25n and the chronozone C23r/C24n boundary (Berggren et al. 1985), respectively, assumes continuity of the section (Romein 1979, p. 31, 34, 35) and calculates age~lepth relationships for the datum levels concerned, one finds the following: (a) the FADs of T. bramlettei, T. contortus and T. orthostylus fall at 57.8, 57.2 (56.8 Ma in Berggren et al. 1985) and 57.0 Ma (56.3 in Berggren et al. 1985), respectively; (b) these values imply a duration for lower NP10 (prior to the entry of T contortus) of c. 0.6 million years and a duration of c. 0.2 million years for the T. contortus (CP9a) Subzone; (c) the LAD of most Fasciculithus species was shown to occur in sample 106, 3 m below the FAD of T. bramlettei (Romein 1990, figs 13, 14) (although rare specimens of F. tympaniformis were recorded throughout the T. contortus Subzone), consistent with data reported in Berggren et al. 1985 and here. In view of the records compiled here, it would appear that the use of the LAD of F. tympaniformis to approximate the NP9/NP10 zonal boundary is not satisfactory, and that subzone CP9a is indeed not the precise correlative of Zone NP10 at its lower limit as suggested by some workers (e.g. Okada & Bukry 1980; Monechi 1985). In a companion paper to this paper (Aubry et al. 1996) we have estimated the duration of Zone NP10 to be 1.45 million years, and that of Zone NP9 to be > 0.904 million years

327

(within the constraints of the CPTS of Cande & Kent 1995). In our earlier work Zone NP10 was estimated to have had a duration of c. 1.3 million years (Berggren et al. 1985: Aubry et al. 1988) based on the proxy use of the LAD of Fasciculithus spp. for the NP9/10 zonal boundary in the absence of a magnetostratigraphic calibration for the nominate boundary taxon, T. bramlettei. The data reviewed here (and above) suggest that there is a general succession of events from LAD of Fasciculithus spp., FAD of T. bramlettei, FAD of D. diastypus, FAD of T. contortus and FAD of T. orthostylus. This sequence is seen to occur in Holes 550 and 577. However, there is a stratigraphic gap at 408 m in Hole 550, which encompasses the upper part of Zone NP9 and the lowermost part of Zone NP10. Similarly, there is a gap at c. 82.5 m in Hole 577, which encompasses upper Zone NP9 and a large part of Zone NP10. It would appear that the FAD and LAD of M. subbotinae and G1. pseudomenardii, respectively, are juxtaposed in chronozone C25n(_, and Y) that the delayed entry o f M. subbotinae, caused by the widespread dissolution event in early Chron C24r is the reason for the erratic/anomalous records (i.e. stratigraphic overlap or separation) of the relationship between these two taxa. An assessment of available deep sea drilling sites which span the Paleocene/Eocene boundary stratigraphic interval suggests that there is no complete/ continuous (i.e. ideal) section present in which one can discern the complete succession of either bioor magnetostratigraphic events from Chrons C25n to C23n.

The Paleocene/Eocene boundary: identity and geochronology N W E u rope

The marine stratigraphic succession developed in the sedimentary basins on the passive margins of the Baltic Plain surrounding the North Sea (the Anglo-Belgian-Paris-Danish-North German) has come to serve as the standard for mondial Paleogene chronostratigraphy (Berggren 1971; Pomerol 1973; Curry 1965; Curry et al. 1978). Unfortunately exposures in this region are often poor and stratigraphic sequences are limited in both vertical and lateral extent and riddled with unconformities. Repeated transgressions and regressions have left a characteristic imprint of rapid lateral and vertical facies changes, and late Neogene glacial tectonism has further added a final coup de grace in some areas (e.g. Denmark) in the form of physical dislocation of important parts of the stratigraphic record. Nevertheless a network of biostratigraphic

328

W.A. BERGGREN & M.R AUBRY

and magnetostratigraphic studies (including Aubry 1983, 1985, 1986; Aubry et al. 1986; Townsend & Hailwood 1985; Steurbaut & Nolf 1986; Knox et al. 1990, 1994; Ali et al. 1992, 1996), coupled with petrological studies on volcanoclastic deposits intercalated in the uppermost Paleocene-lowermost Eocene sedimentary sequence of northern Europe and the North Sea Basin (Morton et al. 1983; Knox & Morton 1988; Morton & Knox 1990; Knox 1990) is now providing a framework for an integrated Paleogene chronostratigraphy and an estimate of the position and numerical chronology of the Paleocene/Eocene boundary relative to the geomagnetic polarity time scale (GPTS). In the remaining part of the paper the chronology of the late Paleocene-early Eocene is based on the revised magnetochronological scale developed by Cande & Kent (1992, 1995) and not on that of Berggren et al. 1985, unless otherwise indicated. In epicontinental northwest Europe, the Paleocene Series is represented by the Danian, Selandian and Thanetian Stages. The base of the Thanetian (Pegwell Marls and stratigraphic equivalents in a borehole in south Essex: Knox et aL 1994) is within the Zone NP6-NP7 interval and C26n, and the succeeding Thanet Sands are referable to Zone NP8 (see Aubry 1983, 1985, 1986). The overlying Woolwich-Reading Beds belong to Zone NP9 (Siesser 1987). The lowest calcareous fossiliferous horizons in marine sediments representative of the Ypresian Stage have been assigned to Zone N P l l in the London Clay Formation in England (Aubry 1983, 1985, 1986), the Formation de Varengeville at Varengeville, Normandy, France (see Aubry 1983, 1985, 1986) and Belgium (Steurbaut & Nolf 1986; Steurbaut 1988). Zone NP 10 has not been recognized in NW Europe owing to the absence of calcareous facies. However, dinoflagellate biostratigraphy has enabled a correlation between calcareous and noncalcareous facies, and this has been discussed in detail elsewhere (Berggren et aL 1985). Various lithostratigraphic schemes have been proposed to describe the Paleocene series in the North Sea. Five lithostratigraphic units (A1 to E3) were described by Knox et al. (1981) to further subdivide the Ekosfisk Formation, the Montrose Group and Moray Group (both groups being equivalent to the Rogaland Group), which Deegan & Scull (1977) had differentiated. These lie between the top of the Cretaceous chalk and Balder Formation. These, in turn, have been correlated with the shallow water deposits of England (e.g. Cox et al. 1985). The position and correlation of the North Atlantic volcanoclastic tuff series with regard to standard biostratigraphy and magnetostratigraphy is critical to an evaluation of the position and age

of the Paleocene/Eocene boundary. The record of at least two phases of early Paleogene explosive volcanism in the NE Atlantic is found in the form of volcanoclastic deposits intercalated in marginal marine sediments of NW Europe (Knox & Morton 1983, 1988). Distal, attenuated representatives of the younger (Phase 2) ash series have been recorded over 1000 km to the southeast in deep sea sediments of the Goban Spur (Hole 550) and Bay of Biscay (Hole 403) (Knox 1985) in the NE Atlantic. The source of the NW European and eastern Atlantic pyroclastic deposits is thought to have been the so-called Hebridean-Thulean volcanic province of Scotland, the Faeroes and East Greenland (Morton & Parsons 1988, table 7). The pyroclastic deposits of the first (older) Phase 1 were restricted to volcanic activity in the British and Faeroe-Greenland province (Knox & Morton 1988); those of the younger Phase II have been linked with the Faeroe-Greenland igneous province, more specifically the proto-GreenlandScotland Ridge (Morton & Knox 1990; Miller & Tucholke 1983; Berggren & Schnitker 1983). Phase I tuffs occur in stratigraphic units C 1 and C2 and the lower part of Unit C3 (Lista Clay Formation) in Norfolk (Cox et al. 1985), correlated in turn with the upper part of Chron C26n. Calcareous nannoplankton biostratigraphy (Siesser 1987) of outcrops in SE England enabled Knox & Morton (1988) to suggest extension of Unit C3 into Zone NP8, indicating that the Forties Sand (= Unit D) is no older than Zone NP8. Phase I ashes have also been recorded from the lower Thanet Beds at Pegwell Bay which also belong to Chron C26n (Townsend & Hailwood 1985; Aubry et al. 1986) which, in turn, appears to be correlative with the younger half of Chron 26N (Berggren et al. 1985; Berggren et al. 1985, 1995). These ashes have also been correlated with the upper part of the A l i s o c y s t a m a r g a r i t a Zone by Knox & Morton (1988) (see also Powell 1988). It would appear that the Phase I ashes (at least those recorded from the lowermost Thanet Sands at Pegwell Bay) are to be more appropriately correlated with the lower part of the A. m a r g a r i t a Zone if the correlation of that zone with Zone NP7 is correct (Powell 1988). Phase 2 pyroclastic deposits have been divided into 4 subphases (Knox & Morton 1988) and occur in the Sele Formation (stratigraphic unit E of Knox et al. 1981) and Balder Formation of the North Sea and at equivalent stratigraphic levels in onshore and offshore boreholes, as well as in NE Atlantic drillsites (e.g. Sites 403, 550). In Denmark, a remarkable series of almost 200 ashes corresponding to subphases 2a and 2b occur in a finegrained diatomaceous unit known as the Mo Clay or Fur Formation (BCggild 1918; Andersen 1937). These ash beds are divided into a lower (negatively

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE numbered series (-1 to -39) of mixed basalticrhyolitic composition (corresponding to phase 2a) and an upper (positively numbered) series (+1 to +140) of predominantly basaltic composition (corresponding to phase 2b). The negative series has been correlated with the following silicoflagellate zones (in ascending order): the Naviculopsis constricta Zone (ashes -39 to -35), the N. danica Zone (ashes -4 to -21A), the Dictyocha elongata Zone (ashes -19 to -17) and the Corbisema naviculoidea (partim) Zone (ashes -17 to 0), whereas the positive series has been correlated with the C. naviculoidea Zone (partim) (ashes +1 to +19) and the N. aspera Zone (ashes +20 to +130) (Perch-Nielsen 1976). The lower part of the negative ash series has been correlated with the upper part of the Apectodinium hyperacanthum (dinoflagellate) Zone. The upper part of the negative series, beginning with the -19 ash layer, and the positive ash series have been equated with the (?) local Acme Zone of Deflandrea oebisfeldensis in the upper part of the A. hyperacanthum Zone (Hansen 1979; Heilmann-Clausen 1982; Nielsen & Heilmann-Clausen 1988). The lowermost negative ashes have been (tenuously) correlated with the Sables d'Erquelinnes (Belgium) assigned to Zone NP9 (Heilmann-Clausen 1985). Subphase 2a ashes occur extensively in the North Sea Basin (Sele Formation and in onshore beds stratigraphically equivalent to the WoolwichReading Beds (Cox et al. 1985) as well as at deep sea drilling sites from Goban Spur (550) and the Bay of Biscay (403). In Hole 550 Knox (1984) has shown that a series of over 40 ashes (correlated with the Phase 2 ashes of the North Sea area) are restricted to Zone NP10. The -17 and +19 ashes are seen to lie within the lower third of Zone NP10 and the lower half of Chron C24r (see discussion above). Elsewhere the -17 ash has been shown to lie within the D. oebisfeldensis Acme Zone, equivalent with a level in the upper part of the Sele Formation, within the Hales Clay unit of the Harwich Formation and (probably) with a level close to (but above) the Woolwich-Reading/ Oldhaven Beds boundary (Knox 1990). There is probably a significant unconformity between the Woolwich/Reading Beds and the base of the Harwich Formation, however (Knox 1994). Recent magnetostratigraphic studies (All et al. 1992, 1996) have shown that the lower part of the Harwich Formation (Ellison et al. 1994, including the Oldhaven Beds and Division A1 of King 1981) and the lower part of the London Clay Formation (Divisions 2 and 3 (partim) of King 1981) can be assigned to Chron C24r. Until recently there was no documented record of Chron C25n in SE England (Townsend & Hailwood 1985; Aubry et al. 1986), its absence apparently owing to the uncon-

329

formity/hiatus between the Woolwich Shell Beds and the Oldhaven Beds (Harwich Formation). However, a recent study has revealed the presence of an areally restricted c. 2 m thick normal polarity interval assignable to Chron C25n at the base of the Upnor Formation (Bottom Bed) of the London Basin (Ellison et al. 1996; Ali & Jolley 1996). The presence at Sheppey of a (lower) normal polarity interval in Divisions B1-2 and an (upper) normal polarity event in Divisions C1-2 of the London Clay Formation (King 1981) assigned to Zones NPll and NPll/NP12, respectively, and in stratigraphically equivalent levels in Belgium, provides (familiar) constraints for determining the position of the Paleocene/Eocene boundary. The two normal events are assignable to Subchrons C24An and C24Bn (=Subchron C24n.ln (and probably C24n.2n) and Subchron C24n.3n, respectively; Ali et al. 1992). The lowest part of the Harwich sequence has reversed polarity (= Chron C24r; Ali et al. 1992). Knox (1990) has reported volcanic ash layers in sandy mudstones (Hales Clay of Knox et al. 1990) from Norfolk and Suffolk. A comparable mudstone with ash layers, including the -17 ash, was said to occur along the western margin of the southern North Sea. The similarity in stratigraphic position of the two mudstone sequences relative to the ash horizons led Knox (1990) to suggest correlation of the base of the 'London Clay' (now base Harwich Formation) to a level correlative with Zone NP10. Although the -17 ash has not been positively identified in outcrop sections of southern England, basaltic ashes equivalent to those of the North Sea and NE Atlantic (where they occur in Zone NP 10) have been identified in the Hales Clay (King's Division A1) of East Anglia. This latter is generally considered stratigraphically equivalent or only marginally younger than the Oldhaven Beds of the London-Hampshire Basins (King 1981; Knox 1990). Finally, Ali et al. (1992; fig. 19) correlated the base of the 'London Clay' (now Thames Group) (Division A1) transgression with the rapid rise in sea level in the middle part of cycle T.2.4 of Haq et al. (1987) and showed that it was correlative with the early part of Chron C24r (Ali et al. 1992; fig. 17). The position and age of the Paleocene/Eocene boundary in terms of the Geomagnetic Polarity Time Scale (GPTS) continues to elude stratigraphers. A significant component to the solution of this problem lies in the development of a correlation network which integrates data on the distinctive -17 (54.5 Ma) and +19 (54.0 Ma) ashes within the calcareous plankton biostratigraphic framework of DSDP Hole 550 (Fig. 9) and the combined ~13C record in Holes 549 and 550, the global deep sea stratigraphic record and placement

330

w . A . BERGGREN (~ M. P. AUBRY

~I

P7 340-.

o~

+ .-.....--...

.~

c"B'~Z.

"

!'~

-o

,~

o,,I

o

EL

lL

',r-

" I

.-...

-~-

350-~

EL

"5 t-. Q. t,_ 0

I:~ "~-

(b

0

t_ 0 -It-

9

,

"5 tv u)

370.~ r~

E 9

:

(b

: m"

380"

O~1

lL

13_

:

i Q}

::

~

9

~ -vvv.

i"

"

--

.,-vvvv

:

~

Lo 13..

. -

i

I

I

,-v-~ ,v.~ w~.,v,

I i

: :

L,vv.~ ,.,.,v~ ,.,.: .,..~-~

: :

o o o ~@o o ~ '~ o ,' o d ~ .o O'L - o '~ ~ S ~ " ^ o = o o .0 o o" 'oOoOoOo o OoO o oo%,^~ Oo( o PoO^o O o O otDOO'3:"~'C, O o O o ~ o" o o , o o o o~176 q|u"'o 6 o ~o~o 1 7 6 o o " oo ~, o'~----:,;~,~ OoO . ooo , o oo, o ooo~o ,.o o c ~o1 7 o~O r _ o o o ~o1 7 6o [ ~"-. ~ G ' ~ o ~o o o~176 o ~ 1 7 6 o o~176 o=r176176176176162176176176 ~oo~ 'o ~o ~)o?o ~ ~ 1 7 6 ~ . ~ o ~ ~o ~ 1 7 6 ~ . ~ . , ~ o ~ ~ ~ ~ ~~176 ~ ~oo ~ "~0~ o~ -

lL

".,.M

or

o~

! r i l l

-

420----

Positive ash

I~

NP5

series Negative ash series 0 r 1 6 2 +19 ash

o0o

-17 ash

i::::= ~

~

n

,-.

!

~-

:

LC

"'~(I)--

a:

L----

(P4)430"~.c2so

~

00

! ~ '1 !

44O

Z

Z

CP4

+

LAD 7 I n c l u d i n g ~" bracketing

} FAD . ) l i m i t s D

Zonal b o u n d a r y uncertainty

'~ Ckl

Benthic foraminifer extinction event

i (5

Fig. 9. Magnetobiostratigraphic correlations spanning the Paleocene/Eocene boundary in DSDP Hole 550.

LATE PALEOCENE-EARLYEOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE of the (standard) stratigraphic record in NW Europe in a sequence stratigraphic framework. Recent studies by J. R. Ali, E. A. Hailwood, J. Hardenbol, C. King, R. W. O'B. Knox, us and others have shown the following. 1. The base of the Harwich Formation (Oldhaven Beds = Hales Clay) is within early Biochron NP10 and probably within early to mid Chron C24r. 2. The North Sea Phase 2 ash series occurs primarily within Zone NP10, but may extend downwards into Zone NP9. The -17 ash is within the Deflandrea oebisfeldensis Acme Zone which is biostratigraphically equivalent to lower Zone NPI0. 3. In DSDP Hole 550 the -17 ash (54.5 Ma) lies at c. 400 m, about a third of the way up in Chron C24r; the +19 ash (54.0Ma) lies at c. 393 m slightly below the mid-point of Chron C24r (Fig. 8). 4. In DSDP Hole 550 the -17 and +19 ashes are bracketed by the (delayed) entry of Morozovella subbotinae and the FAD of Acarinina wilcoxensis (below) and the FAD of M. lensiformis (above). The FADs of Morozovella formosa gracilis and Igorina broedermanni and the LAD of Morozovella acuta (above) occur between the two ashes, i.e. they lie within (revised) basal Zone P6a and the P5/P6 zonal boundary, respectively. The -17 ash lies 8 m above the FAD of T. bramlettei, c. 27.8 m below its LAD and c. 27.35 m below the FAD of T. contortus Morphotype B. The +19 ash lies 15 m above the FAD of T. bramlettei, c. 20.8 m below its LAD, and c. 20.35 m below the FAD of T. contortus Morphotype B (Fig. 9). 5. The composite 8~3C records of Holes 549, 550 and 690B suggest that the prominent spike associated with the major turnover in deep water benthic foraminifera at other deep sea sites occurs in the early part of Chron C24r, within Zone NP9 and P5. There are problems with this straightforward interpretation, however, as underlined below and delineated in Fig. 10 (see also Aubry et al. 1996 for a more thorough discussion of this problem). These problems are associated with dissolution and an unconformity/hiatus at c. 408 m in Hole 550. We discuss the problem of dissolution first. Dissolution is moderate (upper half of chronozone C24r) to strong (lower half of chronozone C24r) in Hole 550. By carefully examining the c. 63 m-thick chronozone C24r stratigraphic interval in the course of a stable isotope study, Ashish Sinha (personal communication 1993) has estimated that c. 4.7 m (12.5%) of the stratigraphic interval between the -17 ash and the base of Subchron C24n.3n is a dissolution facies. Corresponding estimates for the interval from

331

the -17 ash to the top of C25n are 12-14m (60%). Sediment acumulation rates reflect this phenomenon: above the +19 ash rates are estimated at c. 5.1 cm/103 years, whereas below this level (in the interval of strong dissolution) they are c. 14-16 cm/103 years. However, these rates are quite unrealistic and alternative estimates are discussed in greater detail below. In an attempt to restore the c. 63 m stratigraphic interval of Chron C24r to its 'true' thickness, so as to determine the position of the -17 and +19 ash beds, that of the relevant biostratigraphic datum levels, and thus establish the appropriate ages for these datum events and, eventually, the age of the Paleocene/ Eocene boundary, we have added c. 5 m to the upper part of the Chron C24r interval and 13.5 m to its lower part. This results in a total thickness of c. 81 m for Chron C24r and a net change of c. 18.5 m (Fig.10). If we now assign the -17 and +19 ashes to their 'correct' (i.e. restored) level, we see that they lie at 398 m and 401 m, respectively (c. 5 m lower than their former (unreconstructed) level. Within the context of the revised thickness of Chron C24r, these now lie at c. 44% and 53%, respectively, of the way up in chronozone C24r, i.e. slightly higher than in the unrestored section (see Fig. 10). The net change in the position of the +19 ash is seen to be insignificant, whereas that of the -17 ash is relatively more significant (a 10% change upwards in chronozone C24r owing to the fact that the lower part of chronozone C24r has been expanded by c. 60% (Fig. 10). The second part of the conundrum is the fact that Chron C24r is not completely represented at Hole 550. We have already discussed (above) the fact that the normal/reversed magnetic polarity boundary at c. 4 2 2 m may not be the Chron C24r/C25n boundary. However, even if it is not, the biostratigraphic data reviewed above indicate quite clearly that the reversed polarity interval just above c. 422 m corresponds to the earliest part of Chron C24r. In any event, Chron C25n is present in Hole 549 and the thickness of the stratigraphic interval between the Chron C25n/C24r ~md the NP9/NP10 contact is essentially the same as that of the interval between the unconformity and the questionable Chron C24r/C25n boundary in Hole 550. Hole 549 could just as easily be substituted for the lower (pre-unconformity) interval. Of greater importance is the presence of an unconformity at c. 408 m heralded by the close juxtaposition of the FAD of Tribrachiatus bramlettei, the benthic foraminifieral extinction and the carbon isotope excursion - events which are separated by a discrete stratigraphic interval in continuous sections as in Hole 690B (Aubry et al. 1996). To determine the relative position of the

332

w . A . BERGGREN & M. P. AUBRY % C24r ISed- Rate Dissol. T in m I m/million]Interval Strat. I years (cumul.] Interv.

ZONES

NP

P

i

,rm r.)

RESTORED STRATI( IRAPHIC POSITION OI DATUM LEVEl

.,..,_

350 - -

3.~r NP11

;24n.2-3

P6b

TTT

--366~ _

--374--

_

P6a

_

0 0 >0r

400 - - o o _

~ NP10 ooooo

)oc

51.8

62.85

-4.7

.J_

P6a

O 0 +19 Ash, 54 Ma

ooooo

P6b

P5

NP5

T

Benthic Ext.

NP9

0 0 398

ooo-...~.. 0 0 0 0 0

0 0 -17 Ash, 545 Ma

ooooo

~408:~

450

NP11

_

i

NPIO

ooo

405

P5 ~

(422)

P5

(P4)

C25n-(441) C27n

O

Fig. 10. Revision of datum levels and restored thickness: DSDP Hole 550.

513C spike and the -17 and +19 ashes in Chron C24r in order to obtain a better estimate of the chronology of this latter and determination of the age of the 513C spike and the biochronology of the calcareous plankton events in Chron C24r, it is necessary to have a complete record of it in a deep sea core or outcrop (which we do not believe has been demonstrated in any published record to date) or to reconstruct a composite record. This has been attempted below and in Aubry et al. 1996). The problem with deriving an internally consistent chronology within Chron C24r and C24n is exacerbated by the fact that there is an unconformity at 408 m in Hole 550, at a level which separates mid-Zone NP9 from lower Zone NP10. A further, and more immediate, problem is deriving an internally consistent magnetobiochronology of calcareous plankton datum events in the Paleocene-Eocene boundary interval spanned by Chron C24r. Figure 11 illustrates (a part of) the problem and its construction and implications are discussed below.

Line A is drawn through the age v. depth calibration points in Hole 550 of magnetochron boundaries from Chron C25n/C24r to Subchron C23n.2n/C23n.lr using the ages estimates in the GPTS (Cande & Kent 1995; Tables 2 and 3). A relatively straight line over the c. 5 million years interval is seen with an inflection in the Chron C24n interval suggesting decreasing rates of sedimentation in the younger c. 2 million years part of the record. The fact that the NP9/NP 10 boundary at 408 m does not fall on the line reflects the fact that the NP9/NP10 (chrono)zonal boundary in the GPTS (CK92, 94) is located two-thirds of the way down in Chron C24r, thus differing slightly in position compared to Hole 550. Line B is constructed using the age v. depth calibration points on the -17 (400 m) and +19 (393 m) ashes and extrapolating down to the Chron C25n/C24r boundary at c. 423 m. The position of the -17 ash as dated by Obradovich (in Berggren et al. 1992) at 55.0 Ma is seen to lie on Line A at c. 400 m. The fact that the -17 and +19 ashes do not fall on Line A reflects

Table 2. Revised normal polariO, intervals (Chron C25n to C23n.2n; Cande & Kent 1995)

Chron/Subchron

Age (Ma)

Depth (m) in Hole 550

C25n

55.904-56.391

c. 423 (top)-?

C24n.3n C24n.2n C24n. in C23n.2n

52.903-53.347 52.757-52.801 52.364-52.663 51.047-51.743

C24n.3n(o)-NP9/10 boundary at 408 m

53.347-55.00

359.65-349.62 342.60-344.81 325.9-328.74/337.33 (or 333.03) 359.65-408

Remarks C25n truncated/concatenated with ?C27n in mid part. 55.0 Ma is calibration point at 408 m used by Cande & Kent (1995)

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

333

Table 3. Estimated sedimentation rates in DSDP Hole 550 Chron/Subchron

Time span (million years)

Cm/1000 years

C25n(y)-C23n.2n(y) C25n(y)-C23n.2n(o) C25n(y)--C24n. ln(y) C25n(y)-C24n.3n(y) C25n(y)--C24n.3n(y) C25n.3n(o)-408 m C24n.3n(o)-C24n. ln(y) C24n.3n C24n.3n(o)-C24n. ln(y) C27n.3n(y)-C23n.2n(y) C24n.3n(o)-C23n.2n(y) C24n.3n(o)-C23n.2n(o) Top ash series to 408 m

4.857 4.161 3.540 3.001 2.557 1.653 0.983 0.444 0.539 1.856 2.30 1.604 0.540

1.99 2.16 2.28 2.44 2.45 2.92 1.78 2.25 1.40 1.27 1.467 1.65 4.28

Top ash series to base ash series + 19 ash to C24n.3n(o)

0.467

4.28

0.653

5.107

-17 ash to C24n(o)

0.847

4.76

-17 to +19 ash

0.500

1.4

-17 to +19 ash

0.16

4.28

the imprecision in placing the calibration point (55.0 million years) in the chronology of Chron C24 owing to the fact that it is based on an estimated sedimentation rate, imprecision in the dating, and to a lesser extent, may be a function of the 'wiggle' inherent in a cubic spline which has been fit to calibration points spaced c. 10 million years apart. The unrealistic aspect of Line B is seen in the steep increase in sedimentation rate above the +19 ash required to satisfy the age of the Chron C24r/C24n boundary (Cande & Kent 1995). In the discussion below, we present a line of argumentation directed at resolving, at least for the present, the dilemma posed by the present GPTS (Cande & Kent 1995) in constructing a coherent and consistent chronology of events in the younger half of Chron C24r and its application to events associated with the Paleocene-Eocene boundary interval. Radioisotopic ages on the - 1 7 (54.5 Ma) and +19 (54.0 Ma) ashes in N W Europe (Swisher & Knox 1991) have played a key role in the development of a revised GPTS (Cande & Kent 1992, 1995). These ashes are situated 7 m apart, at 400 m and 393 m, respectively, in DSDP Hole 550, and a sedimentation rate (1.4 cm/1000 years) was used to estimate

Remarks Based on derived age of 54.46 Ma for top ash series using age model in item 8 (above) and 55.0 Ma calibration at 408 m level. Using assumptions above. Using age of 54.0 Ma on +19 ash and GPTS of Cande & Kent (1995) for C25/C24n boundary. Using age of 54.5 Ma for -17 ash and GPTS of Cande & Kent (1995) for C25r/C24n.3n(o) boundary. Using ages of 54.5 and 54.0 Ma on -17 and +19 ashes, respectively. Using assumptions for top ash series to 408 m.

the age of the (supposed) calcareous nannoplankton NP9/NP10 zonal boundary (based on the FAD of Tribrachatus contortus and L A D of Fasciculithus tympaniformis in Mtiller 1985) at c. 407 m. Our studies have shown (1) that: T. contortus is not present at 407 m, nor indeed, in the stratigraphic section i m m e d i a t e l y above. The T. contortus complex, in fact, does not appear until c. 380 m; (2) On the other hand, Tribrachiatus bramlettei, nominate taxon for the base of Zone NPI0, does appear at c. 408 m; and (3) that there is an unconformity at 408 m which separates lower Zone NP10 from mid-Zone NP9. The appearance of T. bramlettei in Hole 550 is thus a delayed entry and not a true FAD. The amount of stratigraphic section missing (and the c o r r e s p o n d i n g time interval not represented) at Hole 550 is virtually impossible to determine owing to the fact that we are unable to derive an internally consistent c h r o n o l o g y for C h r o n C24r. This is due, in turn, to problems with the manner in which the recently revised GPTS (Cande & Kent 1992, 1995) was constructed and with the calibration point used in Chron C24r. The age estimate of 55 Ma on the 408 m level in Hole 550 was used as one of nine calibration points to which a spline function was fit by Cande

334

w . A . BERGGREN & M. P. AUBRY

& Kent (1992, 1995) in the construction of the GPTS. The position of this calibration point within Chron C24r is obviously a critical element in determining the Chron C24r chronology and its constituent biostratigraphic datums. Its position was placed by Cande & Kent (1992, 1995) at Chron C24r.(0.66) based on Berggren et al. (1985). Our studies on Hole 550 have shown that the calibration point (55.0 Ma) is situated at an unconformity, and that significant dissolution has diluted the stratigraphic section in Chron C24r, particularly the lower part. If we attempt a restoration of the section to its original thickness (Fig. 11), the position of the unconformity lies somewhat higher in the expanded (but still incomplete) Chron C24r. Generally speaking the position of the (true) NP9/NP10 zonal boundary could lie somewhat higher in Chron C24r if sedimentation rates (which we cannot calculate because of the presence of an unconformity at c. 408 m) remained constant over the early part of Chron C24r (this point is discussed at greater length in Aubry et al. 1996). Table 2 lists the ages of boundaries between Chron C25n and Subchron C23n.2n from Cande & Kent (1995). Sedimentation rates in Hole 550 are then estimated (Table 3) for the stratigraphic sequence spanning all or parts of the corresponding nearly 5 million year interval. Observations on this approach are listed below. 1. At first glance sedimentation rates over the nearly 5 million year interval would appear to have varied from c. 1.4 cm/103 years to over 5 cm/ 103 years. These rates, however, are based on a variety of assumptions which we shall examine further below. 2. A rate of 2.44 cm/103 years is estimated for Chron C24r using Chron C25n(y) to Subchron C24n.3n(y). The problem is that this estimate runs across the unconformity at 408 m and does not account for the time not represented at this level. However, it is consistent with the GPTS (Cande

& Kent 1995) which was derived by using the calibration of 55.0 Ma at c. 408 m in deriving an age for the Chron C24r/C25n boundary in Cande & Kent (1995). 3. Sedimentation rates of c. 1.5 cm/103 years to 2.25 cm/103 years occurred over the Chron C24n to C23n interval. The fact that sedimentation rates for Subchron C24n.3n (2.25cm/103 years) and Subchron C24n.3n(y) to Subchron C23n.2n(y) (1.27 cm/103 years) were reduced by almost a factor of two attests to the gradual reduction in sedimentation rates over the later part of Chron C24r to C23n interval. 4. Sedimentation rates fluctuated significantly, if not radically, in Chron C24r if one uses both the polarity estimates at the lower and upper boundaries of the Chron and integrates the radioisotopic ages on the -17 and +19 ashes at face value. (a) A sedimentation rate of 1.4 cm/103 years is derived for the 7 m (and 0.5 million years) interval between the two ashes and estimated ages for the top (384.85 m) and base (404.60 m) of the 19.75 m thick ash series are 53.41 Ma and 54.82 Ma, respectively, for a span of 1.41 million years. (b) If we calculate a rate of sedimentation from the +19 (54.0 million years) ash to the Chron C24r/C24n boundary (to satisfy the constraints of the GPTS, Cande & Kent 1995), a rate of 5.1 cm/103 years for the upper c. 33 m of Chron C24r results. If the 54.5 Ma age on the -17 ash (400 m) and the 53.347 Ma age estimate on the Chron C24r/C24n boundary at 359.65 m (GPTS; Cande & Kent 1995) is used to calculate a sedimentation rate for the upper 40 m of Chron C24r, a rate of c. 3.5 cm/103 years is derived. These rates are counterintuitive to evidence from the lithostratigraphic record of the cores in the Chron C24rC23n interval of Hole 550. The sediment is a calcareous nannofossil ooze with significant dissolution in the interval between c. 400 to 425 m and

.

300 NPl1112 (b '~

325 A

~ 375

I~

-17 Ash P5/6

C24n'3n(o) NPlOI11 /

C24n.3n(y) J

~

l

024n.1 n(y) / ~

P6/7

C23n.2n(o ) ~

C23n.2n(y) ]

1 -

-

(+1g/Ash)

-

"NP9/10" (JD~)) l ~ ~ NP10/11 C25ny) ~ ~ ~ "~. P6aro 400-" 1 ~ ~ B ~ unconformity at ~408m ..,.,,T ~ ~ ~ ~ - - - "NP9/10" -17 Ash (CCS) +19 Ash (CCS) 42556.0.'g '.'a '.; '.'6 '.; '.: '.'3 '; '', 55.0 '9 ' 'a' '7' ', ''s' '4 ''3' ;' '~'5/.'0 '9' =s' ',' '6' 's' .'4' .'3' '2' ', 5~.0 '9' '8' '7' '6' '5 ' .', ' .'3 ' .'2 ' ', '5~.0.'9 ''8 ' .'7 ' .'6 ' .'5 ' : '.'3 ' '2 ' .', '5~.0

Time in million years

Fig. 11. Age versus depth framework of magnetostratigraphic boundaries and -17 and +19 ash beds in DSDP Hole 550. See text for discussion.

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE only minor dissolution above. Sedimentation rates would be expected to have increased (not decreased) over the 20 m interval characterized by the input of volcanogenic detrital material to the normal pelagic carbonates. Sedimentation rates of c. 3.5 to > 5 cm/103 years in a normal pelagic carbonate (later part of Chron C24r in Hole 550) are unrealistic unless a significant increase in productivity can be demonstrated. No such evidence is seen in the carbonate record (see de Graciansky, Poag et al. 1985, DSDP Leg 80, Hole 550, p. 300-303 (Barrel Sheets) and p. 335-338 (core photographs)). Such high rates are contradicted by the significantly lower (c. 2-1.5 cm/103 years) rates of sedimentation in the pelagic carbonates of Chron C24n-C23n interval above and which would appear to be extrapolatable downwards into the carbonate section in the younger part of Chron C24r. These rates also assume constant/similar rates of sedimentation in both the volcanogenic part of the section (c. 385-405 m) and the non volcanogenic, carbonate part (c. 385 to 360 m, and above). We can only conclude that the use of a constant sedimentation rate over the entire Chron C24r interval is an appropriate working assumption only if the radioisotopic ages on the ashes are not used in deriving calibration points for the calculations. However, this assumption falters also on other grounds, which are discussed below (see point 5). (c) These observations suggest to us that the use of the radioisotopic ages on the -17 and +19 ashes to derive an estimated age for the 408 m level in Hole 550 and its use, in turn, as a calibration point in the GPTS (Cande & Kent 1995) is fraught with danger. While the isotopic ages on the ashes, at least on the -17 ash, may be analytically precise, we would query their geological accuracy. The 55.0 Ma age on the -17 ash by Obradovich (in Berggren et al. 1992) v. the 54.5 Ma age on the same ash by Swisher & Knox (1991) reinforces this uncertainty. Further to the point, Obradovich (pers. comm. 1994) informs us that he has recently dated the +19 ash at 54.5 ___0.5 Ma, the large error (comparable to that in the 54.0___0.5 Ma age of Swisher & Knox) being due to the heterogeneous nature of the ash with visible reworked detrital feldspars common. He cautions against using the +19 ash for dating purpose and indicates that the -17 ash remains the most homogeneous and 'cleanest' ash for dating the ash series. (d) There is no solution to this conundrum at the present time because the 55 Ma age at c. 408 m is an extrapolated age estimate at an unconformity whose position in Chron C24r remains elusive. The next generation of GPTS will require the integration of additional radioisotopic ages on well calibrated levels within a continuous/

335

complete Chron C24r and preferably within uniform lithic (carbonate) facies. 5. An alternative approach to estimating the sedimentation rate(s) within the Chron C24n to Chron C24r in Hole 550 involves the following three steps. (a) We have seen above (Table 3) that sedimentation rates decreased from c. 2.5 cm/103 years in Chron C24r (integrated over its extent) to c. 1.5 cm/103 years in the younger part of Chron C24n. An estimate of c. 2.25 cm/103 years for Subchron C24n.3n is seen to have decreased to c. 1.4 cm/103 years for the Subchron C24n.3n(y) to C24n.ln(y) interval and an integrated value of c. 1.78 cm/103 years is estimated for the Chron C24n interval. While an unconformity (and brief hiatus) cannot be excluded between subchronozone C24n.3n and C24n.ln (in the absence of definitive evidence for the presence of subchronozone C24n.2n), it is more likely that the section is continuous in this interval and that the absence of the relatively short (0.054 million years) Subchron C24n.2n is due to the moderate sedimentation rates. We proceed by using the 1.78cm/103 years estimated for the Chron C24n interval to estimate datum events in the Chron C24n to Subchron C23n.2r interval. (b) We use a sedimentation rate estimate of 2.25 cm/103 years (in Subchron C24n.3n; Table 3) for datum events in the carbonate section between the Chron C24r/C24n boundary (359.65 m) and the top of the ash series (384.85 m). We justify this rate by noting a sedimentation rate change in the Chron C24n-C23n interval (point 5a, above). The rate for Chron C24n is likely to be more appropriate for purposes of downward extrapolation in the pelagic carbonates of Chron C24r than in the interval of Chron C24n and C23n in view of the upward diminishing sedimentation rates (Table 3). Inasmuch as the ash series was almost certainly deposited at a higher rate of sedimentation than the pelagic carbonates above, we believe it justified to use this extrapolation down to the top of the ash series. (c) Downward extrapolation of the 2.25 cm/ 103 years sedimentation rate in the carbonate section below c. 360 m yields an estimated age of 54.46 Ma for the top of the ash series (384.85 m). If we then use the calibration of 55.0 million years for the 408 m level, we derive an estimated sedimentation rate of c. 4.28 crn/103 years) for the ash series (and c. 3.5 m of carbonates below the base of the ash series between 404.60 and 408 m). This rate is used to calculate the ages of datum events within the ash series. As a matter of interest using this rate results in an estimated duration of 0. 467 million years for the 19.75 m thick ash series, age estimates of c. 54.65Ma and c. 54.81Ma for the +19

336

w . A . BERGGREN & M. E AUBRY

and -17 ash, respectively, and an interval of c. 0.160 million years between the two. Needless to say these estimates are predicated on the closed loop of having used the 55.0 Ma calibration at 408 m to anchor the lower point of this interpolation. Finally, the estimated duration of the ash series in Hole 550 (0.467 million years) should be compared with the estimate of c. 1.4 million years (based on the dates themselves; point 4 above). The latter estimate is over half the duration of Chron C24r (Cande & Kent 1995) and is clearly unrealistic considering the discussion above. In this connection, a recent study by Fenner (1994) on the diatom stratigraphy of the ash series in the Fur-Olst formations in the Harre borehole in the southern Limfjord area of northwest Denmark are particularly germaine to our study of Hole 550. Fenner (1994) cross-correlated the range of stratigraphically important diatom species in the Danish diatomaceous ash series with their range in an upper P a l e o c e n e - l o w e r Eocene stratigraphic section in ODP Hole 752A in the northern Indian Ocean that contains both siliceous and calcareous microfossils. Conclusions pertinent to this study include the following. (i) The -30 to +30 ash series of the Fur Formation (and perhaps the entire Fur and Olst formations) are correlative with the upper Disr diastypus Subzone (CP9a) and/or upper Zone NP10. (ii) A time span of c. 0.5 million years is estimated for the deposition of the ash series of the Fur Formation (presumably by reference to Berggren et al. 1985). It will be readily seen that from our studies of Hole 550 (Fig. 9) that Fenner's results (1994) are in

close agreement with those in this paper. We would note, however, that the FAD of Dicoaster diastypus postdates the FAD of Tribrachiatus bramlettei (rather than being simultaneous) so that the two zones are not exactly equivalent. We cannot be certain that the entire Danish ash series is represented in Hole 550, but the c. 20 m thick series between c. 385 m and 405 m belongs to lower to mid Zone NP10/CP9a. The ash series in Hole 550 is terminated (c. 385 m) c. 12 m below the FAD of T. contortus (Morphotype B; c. 373 m) with an estimated age difference of c. 0.5 million years, supporting an early-mid (rather than late) Biochron CP9a/NP10 age for that portion of the ashes present in Hole 550. Previous estimates on the duration of the ash series which have varied from c. 3 million years (Perch-Nielsen 1976), c. 1 million years (Heilman-Clausen 1982, by correlation of the -19 to +140 ashes to the Deflandrea oebisfeldensis (dinoflagellate) Acme Zone of the Apectodinium hyperacanthum Zone), to c. 60 000 years (Bonde 1974) who interpreted the ash layers as laminated (seasonal) varves) are seen to have significantly over- and underestimated the probable duration (c. 0.5 million years) of the ash series. 6. Depth data on calcareous nannofossil and planktonic foraminiferal datum events in Hole 550 are tabulated in Tables 4 and 5, respectively. Age estimates of these datum events are tabulated in Tables 6 and 7, respectively, using the tripartite sedimentation rates discussed above (Point 5). Where then does the Paleocene/Eocene boundary lie in the context of the GPTS, and what is its

Table 4. Calcareous nannoplankton datum events in DSDP Hole 550 spanning Paleocene-Eocene Zones NP9-NP12

Datum event

Sample interval Core/Section (m)

FAD Discoaster lodoensis LAD Tribrachiatus contortus (morphotype B) FAD Tribrachiatus orthostylus LAD Tribrachiatus bramlettei LAD Discoaster multiradiatus FAD Tribrachiatus contortus (morphotype B) LAD Tribrachiatus controtus (morphotype A) FAD Tribrachiatus contortus (morphotype A)

Depth (m)

Midpoint value) (m

27/cc-2811:46--47 29/cc-30/1:20-23

346.50-346.46 365.50-365.72

346.23 346.61

30ll: 62-64 30/1:109-111 30/5:22-24 30/5:66-68 30/4:61-63 30/4:99-101 30/5:113-115 30/6:7-9 31/3:6-8 31/3:39-41 31/5:88-90 32/1:17-19

366.13-366.60

366.36

371.73-372.17

371.95

372.13-372.49

372.31

372.64-373.08

372.86

378.07-378.40

378.23

382.89-384.68

383.28

Remarks

NP 11/12 = CP 9/10 NP 10/11 = CP 9a/b boundary

Data from Aubry in Aubry et al. (1996) and from Mtiller 1985, table 10, p. 592-593, for the interval above 365.50 m.

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

337

Table 5. Planktonic foraminiferal datum levels in DSDP Hole 550 spanning Paleocene-Eocene Zones P5-P7 Datum event

Sample interval Core/Section (m)

Depth (m)

FAD Morozovella aragonensis

27/3: 41-4527/4:42-45

340.42-341.50

340.96

LAD Morozovella lensiformis

28/1: 50-5327/5:42-46 28/1: 50-5327/5:42-45 30/1: 49-5329/6:51-54 30/3: 49-5330/1:49-53 "31/2: 65-6731/2:105-107 33/2:59-61 33/2:59-61 33/1 : 59-6133/2:59-61 *33/2: 51-5333/2:70-72

347.0-343.42

345.21

347.0-343.42

345.21

366.03-363.94

364.99

368.99-366.03

367.51

375.17-375.55

375.36

394.51-396.09

395.05

394.51-396.09

395.05

396.02-396.21

396.11

399.52-400.03

399.78

403.72--405.22

404.47

405.65-408.62

407.14

409.02-409.20

409.11

LAD Morozovella marginodentata LAD Subbotina velascoensis

LAD Morozovella aequa FAD Morozovella lensiformis FAD Morozovellaformosa gracilis FAD Igorina broedermanni

LAD Morozovella acuta FAD Morozovella marginodentata FAD Acarinina wilcoxensis

FAD Morozovella subbotinae

*33/4: 101-10333/5:2-5 34/1 : 62-6534/2:62-65 a. 34/2:62-6534/4:62-65 b.* 34/4: 101-10334/5:119-121

Midpoint value) (m

Remarks

Sample 27/4:42-45 cm = barren zone; taxon not present in 27/5:42-45 cm (1.5 m below); P 6/7 zonal boundary.

P6a/b zonal boundary.

_=_P5/6 zonal boundary if LAD M. acuta is taken as proxy for LAD M. velascoensis.

Entry probably delayed entry here due to dissolution.

* This paper. Data from Snyder & Waters 1985, fig. 6: 448, 449, except as noted. Chronology based on Berggren et al. (1995). Depth (in m) of datum levels given in parenthesis.

age? There are essentially four Paleocene/Eocene boundary levels currently recognized by marine (bio)stratigraphers (with a fifth, the 5J3C spike, looming in the wings): (a) at the base of the Harwich Formation (Oldhaven Beds/Hales Clay), SE England; (b) at the base of the Ieper Clay Formation (base of the type Ypresian Stage), Belgium; (c) at the P5/P6a zonal boundary (planktonic foraminifera); (d) at the NP9/NPI0 zonal boundary (calcareous nannoplankton) (Fig. 12). North American vertebrate palaeontologists customarily link their Paleocene/Eocene boundary with the Tiffanian/Wasatchian N A L M A boundary, whereas their continental European counterparts place it at the base of the Sparnacian 'Stage'. The base of the Wasatchian contains a 513C spike in mammalian tooth and soil carbonates thought to correlate with a similar 8J3C spike in soil carbonates in the lower Sparnacian of the Paris

Basin and with the carbon excursion in marine sediments. It would appear that the Paleocene/Eocene boundary (as currently recognized in N W Europe at the base of the Harwich Formation) is correlative with a level in the lower part of Zone NP10 and upper part of Zone P5, somewhat below the midpoint of Chron C24r, near (but stratigraphically below) the -17 ash layer, dated at 54.5 Ma (but estimated here at c. 54.81 Ma). This differs significantly from the correlation proposed in Aubry et al. (1988, p. 734-735) in which the Paleocene/Eocene boundary = base Oldhaven Beds was estimated to lie in the upper part of Zone NP10. This difference reflects considerable improvement in the upper P a l e o c e n e - l o w e r Eocene lithobiostratigraphic correlations over the last few years. The base of the Ypresian Stage (as stratotypified in Belgium) commences one fourth-order cycle higher than the

338

W.A. BERGGREN & M.E AUBRY

POLARITY

HISTORY 51

CALCAREOUS LITHOSTRATIPLANKTON GRAPHIC

ZONES P

NR

P7

c~ E z

,,

L~176

P/E BOUNDARY

UNITS

CRITERIA/LEVELS

~

NALMA

Basin

n2

~ o

$2-

~

.1

n2 53

54

0

P6b

n3

,~.

~_

~

13_

~

~l~f

z

--"~'1 ~-

~

'

I

~

'

~

"~ (~

,._ P6a

~ ~

Oldhaven ~ ' .

/-(1) Base Ypresian (~-54.6 Ma) ~-(2) P5/6 Boundary (=54.7 Ma) - - (3) Base Harwich Frn. (=-17Ash,~54.8 Ma) - - ( 4 ) NP9/10 Boundary (_--55.0Ma)

(D

P5

< z

a: <

o~

s Z 56'

'

o6~

P4

O_ CO -.";" "" . . . . ' :.: ~ . ~ . "-....'

oo a.. Z

-- (5) ~513C(~55.5 Ma) CLARKFORK--- (6) Base Sparnacian (_--55.8Ma) IAN

TIFFANIAN

~

Temporal uncertainty

57.

Fig. 12. Relative stratigraphic position and estimated chronology of varying criteria for denoting the Paleocene/Eoceneboundary. Note: Ma is used to mean 'million years ago'.

base of the Harwich Formation (Oldhaven/Hales Clay; J. Hardenbol pers. comm.) and has an estimated age of c. 54.6 Ma. The P5/P6a zonal boundary is seen to lie at c. 54.7 Ma (midway between the -17 and +19 ashes) in Hole 550 and is probably close to the level of the base of the (type) Ypresian of Belgium. The base of the Harwich Formation (close to the -17 ash) has an estimated age of 54.8Ma. The NP9/NP10 chronozonal boundary has an estimated age of 55 million years (based on simple extrapolation of accumulation rates based on the ages of the two ashes). However, it should be borne in mind that this level is situated at an unconformity in Hole 550, with a hiatus of undetermined duration. We estimate that the duration of the hiatus corresponding to the early part of Biochron NP 10 is considerably shorter than that representing the late part of Biochron NP9. Inasmuch as a definitive GSSP for the Paleocene/Eocene boundary awaits completion of ongoing studies by IGCP 308, a best estimate based on the data reviewed here suggests that, following the various criteria given above, the boundary lies within the stratigraphic interval delimited by the

P5/P6a zonal boundary and/or the -17 ash (above) and the NP9/NP10 zonal boundary (below), an interval estimated here to span < 1 million years (54.6 Ma to > 55 Ma). An age estimate of 54.7 Ma for the P5/P6 zonal boundary, 54.8 Ma for the base of the Harwich Formation and of 55 Ma for the NP9/10 boundary based on the data reviewed here, may serve as a framework for an ultimate determination of the position and age of the Paleocene/ Eocene boundary. We may include the 813C spike as an appropriate criterion for denoting/correlating the position of the Paleocene/Eocene boundary. We are unable to provide a satisfactory age estimate for this event because it lies at a level which is within the early part of Chron C24r and which is distorted in the current GPTS (Cande & Kent 1992, 1995). However, a provisional age of 55.5 Ma is estimated for the 613C minimum and the benthic foraminiferal extinction event (BFE) based on the relative chronology established for Hole 690B (Aubry et al. 1996). This value is internally consistent with the GPTS (Cande & Kent 1995). Correlation with the multiple oceanographic events discussed above (carbon isotope minimum,

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE oxygen isotope warming event, benthic foraminiferal extinction event, size reduction in wind blown aerosols) appears to have been resolved (Aubry et al. 1996). These events seem to line up approximately with a level in mid Zone NP9 and with a level within Zone P5 at several sites. Pervasive unconformities which juxtapose the NP9-NP10 and P5-P6a-b biostratigraphic events are believed to be responsible for the apparent diachroneity of some of these events.

Sparnacian problem and marineterrestrial correlation The ensuing discussion of marine-terrestrial correlations in NW Europe (Fig. 13) is placed within the framework of the recently revised Cenozoic magnetochronology of Cande & Kent (1992, 1995). The basic assumptions/methodology used in constructing this framework and the suggested correlation network between marine and continental stratigraphies that we propose are explained below. The basic magnetobiochronological framework is based on the revised magnetochronology of Cande & Kent (1992, 1995). This revision has resulted in the magnetochronological boundary estimates shown on p. 313. This revised magnetochronology is based on fitting a spline curve to a series of nine age-calibration points in the Cenozoic, including an age estimate for the NP9/NP10 zonal boundary of 55.0 Ma, based, in turn, on extrapolation from 4~ dates of c. 54.0 and 54.5 Ma on the +19 ash and -17 ash, respectively (Swisher & Knox 1991). It will be readily seen that the revised age estimates differ from those estimated in Berggren et al. (1985) by c. 2.5-3 million years. The standard calcareous microfossil stratigraphy is placed in a magnetochronological framework by using the tripartite sedimentation rate calculations in Hole 550 discussed above (Tables 4-7; Figs 3, 9). Dinoflagellate biostratigraphy is added based on previous (Berggren et al. 1985) and recent (De Coninck 1990; Ali et al. 1992) correlations to magnetostratigraphy. At this point it may be useful to review the basic methodology which has been used in developing a geochronological framework for the events associated with Chron C24 and the construction of Fig. 13: (a) The dates of 54.0 and 54.5 Ma on the +19 and -17 ashes in NW Europe (and occurring at 393 m and 400 m, respectively, in DSDP Hole 550) were used by Swisher & Knox (1991) to estimate the age (55 Ma) of the NP9/NP10 chronozonal boundary (which lies approximately the same distance (7-8 m) below the -17 ash as the distance

339

between the two ashes. This boundary is considered here to be located at an unconformity in Hole 550 but we see no way around the dilemma of estimating its true age and the value of 55.0 Ma is retained here for the present. (b) The age estimate of 55 Ma on the NP9/NP10 chronozonal boundary was used by Cande & Kent (1992, 1995) as a calibration point in their GPTS and the position estimated to lie approximately 0.33 of the way up in Chron C24r following the position estimated for this datum level in Berggren et al. (1985). This level is retained here, although we believe that the NP9/NP10 chronozonal boundary is actually located somewhat younger in Chron C24r based on the arguments presented in Aubry et al. (1996). (c) The NP9/10 zonal boundary is located at c. 408 m in Hole 550 which is c. 23% of the way up in (incomplete) Chron C24r (Figs 3, 9). Restoration of Chron C24r to its approximately 'true' thickness (taking into consideration the strong dissolution affecting the lower part of Chron C24r in particular) results in the relocation of the NP9/NP10 zonal boundary to c. 421 m and a total restored thickness of c. 81 m for chronozone C24r. This has the effect of situating the NP9/NP 10 zonal boundary c. 25% of the way up in chronozone C24r. Given the uncertainties in restoring the c. 62 m thick chronozone C24r to its true thickness, we see that the estimate made here in Hole 550 is consistent with placing the NP9/10 zonal boundary 0.33 of the way up in chronozone C24r (Berggren et al. 1985). (d) The age estimates on the -17 ash (54.8 Ma) and the +19 ash (54.65 Ma) based on a calculated rate of sedimentation (rather than their respective radioisotopic values of 54.5 and 54.0 Ma; Swisher & Knox 1991; see discussion above), are used to constrain the stratigraphic position of the base of the Harwich Formation (Oldhaven Beds = Hales Clay) and the base of the Wrabness Member, respectively. (e) An interval low in chronozone C24r, and spanning the chronozone C24r/C25n boundary, is used to constrain the base (Upnor Formation; formerly the Bottom Bed) of the Lambeth Group (formerly the Woolwich & Reading Beds), consistent with Ali (1994). A major feature of the regional stratigraphic correlation framework presented here is the location of the main lithostratigraphic units of the upper Paleocene-lower Eocene succession in NW European basins in a sequence stratigraphic framework (Fig. 13). This framework has been established by the collaborative efforts of members of IGCP 308 (Paleocene/Eocene Boundary Events in Space and Time) Working Group Members during the course of field excursions to the Hampshire-

340

W . A . BERGGREN & M . P . AUBRY

i i

~'i

NORTH SEA BASIN

MAGNETIC B I O S T R A T I G R A P H Y LONDONPOLARITY HAMPSHIRE HISTORY PLANKT. CALCAREOUSDINOFL. BASINS FORAM,NANNOPLANKTON

"

W6

P7

PARIS

BASIN

BELGIAN

Argile de

~~ E

-- ~b]es"E -

~

@

- - - -

52"

~

C23

z~

~1

w~

-~

Sab,ed'Ai,y

~,c2

_

~ B1

10

1~

E

8

o

Laon ~ =

W3

o=

C2

de , ~,

13 --

O

~

~ :~

~E ,~

- I _ J

....

SEISMICSTRAT. LFrHO6TRA.DINOFL. ASH SER. TIGRAPHYBIOSTR.PHI,SE.'

E

Laon

D~

BASIN

~

C1

~I

B1

.~ _c

~

<

--~

~

Balder

-

2b

Fm. 9 P6a

55"

~

7E

~

:~

P5 Wl CP8

~ ~;~iT~'~a~-~

,

"

8 Bois

56-

--

_

! i

CP7

NP8

c2~ P.

~ <

~ Sills

~,2;2,~

7

i

9 ~ ~ _L,

Beauly i E

,

a

>-.

~

2a

;or'~o"~ch -~ ~

GiBes

~J

~

57-

~ ~

Sele Fm.

l.~:'i &"~=

~

NP9

A'

Tulfeaude Lincenl

6 5 4

Forties

~~

Fm.

_ _ , %~.Unnamed

unit

?

!

CP6 NP7 CP5 ~N.PP6

~~ _ , p

PegwelMarls Slourm0uth

Z

>"''" ~a~,~,,

lu. ~lar

,

Marnes de

Getir~en

3

Andrew f' Fan ,o,phw,~

Fig. 13. Regional correlation framework of NW European and North Sea upper Paleocene-lower Eocene stratigraphic sucession. The chronological framework is that of the GPTS (Cande & Kent 1994). The key tie point of the -17 ash (and its relative placement within Chron C24r) is based on the direct age of 54.5 Ma (in Swisher & Knox 1991) rather than the estimate of c. 54.8 Ma based on interpolated sedimentation rate in Hole 550. See text for further discussion.

London (1989), Paris (1990) and Belgian Basins (1991) (J. Hardenbol, unpublished data). This has resulted in recognition and correlation of the main lithostratigraphic succession in these basins to the major sequences of Haq et al. (1987). The basic assumption in this presentation is the recognition of two major sequence boundaries (56.5 and 54.5 Ma in the chronology of Haq et al. 1987) which correspond, respectively, to the (Thanetian s.str.) Reculver Silts/(post-Thanetian s.str.) Upnor Formation, and the Woolwich & Reading/Harwich Formation (Oldhaven Beds) boundaries. Lithostratigraphic studies and correlations (King 1981, 1990b; Knox 1990) and magnetostratigraphic studies (Ali et al. 1992) have shown that the base of the Harwich Formation is within early Chron C24r. The -17 ash (dated at 54.5 Ma) occurs

in equivalents of the Hales Clay (Knox 1990) and serves to locate the base of the Oldhaven Beds in the revised magnetochronolgy of Cande & Kent (1992, 1995). The underlying Woolwich and Reading formations are within chronozone C24r and correlative with (lower) Zone NP9. There is a stratigraphic gap between these beds and the underlying Reculver Silts which spans the NP8/NP9 boundary and (except for a small area) chronozone C25n (which is generally missing in NW Europe). The Lambeth Group is shown as a three-fold succession of lithostratigraphic units separated by two (minor) sequence boundary unconformities and bracketed by the major 56.5 and 54.5 Ma sequence unconformities of Haq et al. (1987). In this study these unconformities have a slightly older age than in Haq et al. (1987) and estimates of the hiatus are 0.7 million years (56.6-55.9 Ma; duration of Chron

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

341

Table 6. Estimated ages of calcareous nannoplankton stratigraphic zones/markers spanning the Paleocene/Eocene boundary in DSDP Hole 550 Datum event

Age in Ma (Depth in m)

FAD Discoaster lodoensis

52.62 (346.23)

LAD Tribrachiatus contortus (morphotype B)

53.61 (365.61)

FAD Tribrachiatus orthostylus LAD Tribrachiatus bramlettei LAD Discoaster multiradiatus

53.64 (366.36) 53.89 (371.95) 53.91 (372.31)

FAD Tribrachiatus contortus (morphotype B) LAD Tribrachiatus controtus (morphotype A) FAD Tribrachiatus contortus (morphotype A)

53.93 (377.86) 54.17 (378.23) 54.37 (383.28)

Remarks

C24n. 1-2r NP 11/12 boundary Late C24r NP 10/11 boundary

In upper NP10 close to LAD Tr. bramlettei Repoprted in NP11 elsewhere

Ages calculated using sedimentation rates of 2.25 cm/103 years (Table 3, item 8). Chronology based on Cande & Kent (1995) and Berggren et al. (1995). Depth (in m) of datum levels given in parenthesis.

C25n at minimum = 0.5 million years) and c. 0.5 million years (55.5-55.0 Ma), respectively, which would appear to be justified by the significant tectonic and mineralogical provenance changes that occurred between these stratigraphic sequences (Knox 1990; Dupuis, pers. comm. 1993). The Lambeth Group (formerly the Woolwich & Reading Beds) is seen to have been deposited in c. 0.5 million years, equivalent to at least the early half of Biochron NP9, and to contain two (minor) sequence boundaries within it (Fig. 13). The important unconformity/hiatus between the

Lambeth Group and the Harwich Formation is probably correlative with the 54.5 Ma sequence boundary in Haq et al. (1987) and is seen to span the late half of Biochron NP9 and earliest Biochron NP 10 as in much of the deep sea record. A vertebrate fauna belonging to mammal Zone MP7 has been recorded from the Suffolk Pebble Bed associated with the Woolwich Shell Beds, but is almost certainly reworked from older levels (Hooker 1991), inasmuch as Zone MP8-9 faunas are known from stratigraphically lower (older) levels at the base of the Sparnacian in France (see

Table 7. Estimated age of planktonic foraminiferal zones~markers spanning the Paleocene/Eocene boundary in DSDP Hole 550 Datum event

Age in Ma (Depth in m) 1

FAD Morozovella aragonensis

52.302 (340.96)

LAD Morozovella marginodentata LAD Subbotina velascoensis LAD Morozovella aequa FAD Morozovella lensiformis

52.54 (345.21)

2

Remarks 3

C23r (early part) P 6/7 boundary 53.53 (364.99) 53.65 (367.51) 53.40 (375.36)

Late C24r P 6a/b boundary

FAD Morozovella formosa gracilis FAD Igorina broedermanni LAD Morozovella acuta

54.69 (395.05) 54.69 (395.05) 54.72 (396.11)

FAD Morozovella marginodentata FAD Acarinina wilcoxensis

54.80 (399.78) 54.91 (404.47)

Mid-C24r Proxy for P 5/6 boundary

Ages calculated using following sedimentation rates: column 1:1.78 crn/103 years (Table 3, item 7); column 2:2.25 cm/103 years (Table 3, item 8); column 3:4.28 crrdl03 years (Table 3, item 14). Chronology based on Cande & Kent (1995) and Berggren et al. (1995). Depth (in m) of datum levels given in parenthesis.

342

W.A. BERGGREN & M.E AUBRY

below). The Blackheath Beds Zone MP8-9 fauna occurs in a deep channel that cuts down into the Woolwich-Reading Beds and is ascribed here to the interval of time represented by the lowstand hiatus between the Woolwich Shell Beds and the Oldhaven Beds (probably extending from late Biochron NP9 to early Biochron NP10). Finally, Zone MP8-9 (with Hyracotherium leporinum and Platychoerops richardsonii) and questionable Zone MP10 vertebrate faunas have been recorded from the Harwich Formation (Division A 1 of King 1981) (= lower half of Zone NP10) and the London Clay Formation (Division B2) (Zone NPll), respectively (Hooker 1991). The Sparnacian consists of a heterogeneous group of terrestrial (paludal, lignitiferous clays and sands) and (sporadically) marginally marine sediments which outcrop in the Paris Basin. Originally defined as a chronostratigraphic unit (Dollfus 1880) between the Thanetian and Cuisian, it is currently more simply considered as a facies (Laurain et al. 1983) and has been regarded as upper Paleocene by some (e.g. Curry et al. 1978; Berggren et al. 1985) and lower Eocene by others (e.g. Pomerol 1973). The Argiles plastiques (bariolres et kaolinitiques) are clay-rich continental deposits on the southern border of the Paris Basin. They overlie the Conglom6rat de Meudon which yields a very rich primitive mammalian fauna (Russell et al. 1989). The Argiles plastiques bariol6es consist of mottled clays and nodular carbonates deposited in subaerial conditions in the western part of the Paris Basin. The younger Argiles plastiques kaolinitiques which consist of kaolinitic clays and sands, rich in organic matter, were deposited in an anoxic lake to the east (Thiry 1981). The Argiles ~ Lignites and the Fausses Glaises are marginal marine deposits in the northern part of the basin. They are organic rich and dinoflagellate stratigraphy suggest that they include at least two sequences (Chateauneuf & Gruas-Cavagnetto 1978). In the western part of the Paris Basin the Sparnacian is bracketed by the Sables de Bracheux (below = Thanetian s.1.) and the Sables de Laon (above = Ypresian). The Sparnacian is generally considered to be biostratigraphically equivalent to the Woolwich and Reading formations of SE England and to the Apectodinium hyperacanthum (W1) dinocyst Zone. Its base, taken as the Paleocene/Eocene boundary by most (European) vertebrate palaeontologists, is correlative with a level within Zone NP9, c. 0.5 million years older than the NP9/10 zonal boundary (used by some marine micropalaeontologists to delineate the Paleocene/Eocene boundary) and > 1 million years older than the base of the leper Clay (base Ypresian of Belgium, the Paleocene/Eocene boundary defined chronostratigraphically on a lithostrati-

graphic basis). A recent study (Russell et al. 1993) has shown that the Argiles plastiques just below the (type) Conglomrrat de Meudon (Paris) is of reversed polarity and thus low in chronozone C24r inasmuch as the ~513Cspike has been found c. 6 m above the base of the Sparnacian Argiles plastiques in the Limay Quarry in the western part of the Paris Basin (Sinha & Stott 1993a, b; Stott et al. 1996). The upper Paleocene-lower Eocene of the Paris Basin, and particularly the nature and extent of the Sparnacian Stage was reviewed at length by Berggren et al. 1985. Subsequent reviews by Cavelier & Pomerol (1986) and Pomerol (1989) have shed additional light on local/regional correlation problems in this region. Pomerol (1989) has delineated several unconformities/hiatuses in the Paris Basin which are seen to correspond to the sequence boundaries of Haq et al. (1987). Of particular importance is the recognition of a significant intra-Sparnacian hiatus (H5) and the reported presence of dinoflagellate Zones W2-5 in the Argiles d'Epernay (Laurain et al. 1983), indicating paradoxically, that while the (type) Sparnacian is the stratigraphic equivalent of the lower Cuisian s.1. (Sables de Laon = NP10-NPll zonal equivalent), it is clearly older than the type Cuisian (Sables de Cuise =Zone NP12; Aubry 1986), except in its uppermost part. To the southeast (Soissons) the Argile h lignites contains the Zone W1 dinoflagellate assemblage up to, and including, the marginally marine Sables de Sinceny. Thus, if this interpretation is correct, the Sparnacian is seen to become significantly younger to the east (Laurain et al. 1983; Laurain & Gurrin 1989), spanning nearly 4 million years. The Sparnacian plastic clays and lignites of the Paris Basin are interpreted here as a (predominantly) terrestrial (paludal) and (sporadically) marginal marine facies deposited during the late Paleocene, corresponding to, and reflecting, the large scale sequestering of organic carbon during Chron C24r observed in deep sea 513C isotope records. Our preliminary field investigations in the type area of the Sparnacian in the eastern part of the Paris Basin (in the spring of 1993) would appear to support the interpretation of Laurain & Gurrin (1989), and their suggested correlation of the Argiles d'Epernay with the lower Eocene in the eastern part of the Paris Basin is considered tentative at this stage pending independent verification of the dinoflagellate biostratigraphy (currently under study by H. Brinkhuis, Utrecht). It is curious that Cavelier & Pomerol (1986, p. 257), while indicating the diachrony/time transgressive nature of the Sparnacian facies in the Paris Basin, recommended restricting the term Sparnacian s.str, to the lower (pre-Cuisian) formations (Conglom6rat

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE de Meudon, Argile plastique, Argiles ~t lignites du Soissonnais s.str, and the Sables de Sinceny), while suggesting eliminating the supposedly upper (Sparnacian stratotype) Argiles ~ lignites d'Epernay = Sables de Laon = Zones W2-W3 = NP10-11) from the concept of Sparnacian s.str. The Cernay fauna (type reference level for the Zone MP6 vertebrate fauna is located stratigraphically above the Sables de Chfilons-sur-Vesle. The Sables de Ch~lons-sur-Vesle s.1. (including the Tuffeau du Moulin compens6), long thought to be the (eastern) lateral equivalent of the Sables du Tillet (= Sables de Bracheux s.1.) are in fact, at least in part, older than the latter. The lower part of this formation belongs to Zones NP6-NP7, and is correlative with the Marnes de Gelinden and the Stourmouth Clay and Pegwell Marl (Fig. 11; Aubry et al. unpublished data). In the light of this new finding, the correlations between the upper Paleocene formations in northwestern Europe can be revised as follows. (a) The Stourmouth Clay and Pegwell Marl, the Tuffeau du Moulin compens6 and Sables de Ch~lons-sur-Vesles and the Marnes de Gelinden are lateral equivalents that belong to Zones NP6 and NP7. (b) The Reculver Silts, Sables de Dieppe, Sables du Tillet, Sables de Grandglise and Tuffeau de Lincent belong to Zone NP8 and represent a major sequence. (c) The Upnor Formation (ex Bottom Bed), Sables de Criel, Sables de Bracheux s.st., and Sables de Bois Gilles, which all belong to Zone NP9, constitute one (or more) younger sequence(s). The age of the Conglom6rat de Cernay (Zone MP6) was thought to be constrained by Zone NP8 (Sables de Bracheux s.1. = Sables du Tillet) below, and the Argiles d'Epernay, above. However, since there is no evidence that the Sables de Chfilons-surVesles are the lateral equivalent of the Sables du Tillet, even in their upper part, the constraints on the age of Cernay are broader, and it may be that Zone MP6 is as old as Zone NPT. Cernay (Zone MP6) may be limited to the time represented by the major sequence boundary (in TA2 2.1/2.2; 56.5 million years in Haq et al. 1987), which corresponds to hiatus HP3 (Pomerol 1988) in the Paris Basin. This hiatus is estimated to have had a duration of at least 0.5 million years, i.e. the duration of Chron C25n, missing in most NW European successions. In the case of the eastern part of the Paris Basin, Cernay may actually represent the continental expression/equivalent of the highstand deposits of the upper part of cycle TA2 (2.1) or of an older deposit. The Conglom6rat de Meudon is usually placed in Zone MP7 but has recently been assigned to Zone MP8-9 (Hooker 1991) and said to contain faunal elements younger than Dormaal (reference level for Zone MP7; see below) yet older than known Zone

343

MP8-9 faunas ). It lies at the base of the Argiles plastiques (basal Sparnacian) which, in turn, rest unconformably on upper Cretaceous chalks. While it is possible that the Argile plastique is diachronous, significant diachroneity has not been demonstrated, and the Argile plastique/Sables de Bracheux boundary is drawn by IGCP 308 Working Group members at a minor sequence boundary (55.5 Ma; Haq et al. 1987) which corresponds to the widespread hiatus HP4 (Pomerol 1989) in the Paris Basin. Additional constraints on the age and stratigraphic position of the Meudon fauna are provided by the only slightly older Dormaal (Zone MP7) fauna of Belgium (see below) and the slightly older Cernay (Zone MP6) fauna of the Paris Basin. We estimate all these faunas to be separated by less than 1 million years The Meudon fauna heralds the major transition from archaic Paleocene to modern Eocene animals, involving the initial appearance of artiodactyls, perissodactyls, primates, chiropterid bats, (local) introduction of rodents and the concomitant reduction in the diversity of condylarths, plesiadapid 'primates' and multituberculates. The base of the Sparnacian (Conglom6rat de Meudon, Zone MP8-9) is traditionally taken by vertebrate palaeontologists as the base of the Eocene. In terms of the stratigraphic correlation framework suggested here it will be seen that this level lies within early Chron C24r = early Biochron NP9 = early Biochron P5 = early Biochron W1, and has an age estimate of c. 55.7 Ma, c. 0.7 million years older than the Paleocene/Eocene boundary as delineated by the NP9/NP10 zonal boundary and c. 1 million years older than the Paleocene/Eocene boundary as delineated by the base of the Harwich Formation (Oldhaven Beds = Hales Clay). Comprehensive lithostratigraphic and biostratigraphic studies on the Ypresian stratotype in Belgium (Dupuis et al. 1990, 1991) and the Knokke well (llRE/138) in NW Belgium (Laga & Vandenberghe 1990), together with magnetostratigraphic studies on lower Eocene outcrops in northern France and Belgium (Ali et al. 1992) now make it possible to close the (correlation) loop in NW Europe upper Paleocene-lower Eocene stratigraphies. Direct lithostratigraphic and foraminiferal (King 1981,1990a, b), dinoflagellate (Dupuis et al. 1990; De Coninck 1990) and calcareous nannofossil (Steurbaut 1990a, b) correlations between the Belgian and LondonHampshire Basins as well as the intermediate section(s) in Normandy in the vicinity of Varengeville are seen to be possible. Two major sequence boundaries are recognized in Belgium corresponding to the major unconformities postulated in the London-Hampshire Basins. The Ieper Clay Formation is essentially the stratigraphic

344

w.A. BERGGREN & M. P. AUBRY

equivalent of the London Clay Formation, although its upper part is the stratigraphic equivalent of the lower part of the Brackelsham Group (Wittering Formation) and the Cuisian s.s. It would appear possible to situate the few vertebrate faunas known from the lower Paleogene of Belgium in the stratigraphic framework established here (Fig. 13). A lone specimen of Arctocyonides has been retrieved in the Tuffeau de Lincent (= Zone NPS), which is stratigraphically older than the Cernay (Zone MP6) fauna of the Paris Basin. Dormaal (reference level for Zone MP7) presents something of a problem. Situated in the so-called 'terminal Paleocene continental Landenian' it overlies the Hoegaar-den-Racour Sand Member of the Tuffeau de Lincent. A major sequence boundary (56.5 million years in the chonology of Haq et al. 1987) separates the Tuffeau de Lincent (below) from the transgressive Formation de Landen (Bois Gilles = Zone NP9; above) and a minor sequence boundary (55.5 Ma in the chronology of Haq et al. 1987) separates the Bois Gilles from the Erquelinnes level (= Zone NP9) with a small vertebrate fauna including Coryphodon (Hooker 1991). This would suggest that Dormaal (Zone MP7) may be situated within the time interval (c. 0.5-0.8 million years) represented by the major hiatus separating the Tuffeau de LincentGrandglise and Bois Gilles corresponding to Chron C25n. It was suggested above that Cernay (Zone MP6) is also situated within this time interval which, if true, would imply that the major faunal change that occurs between Zones MP6 and MP7 faunas took place over a relatively brief (0.5 million years) interval of time in the late Paleocene, and c. 0.5 million years earlier than the Meudon (Zone MP8-9) fauna. In Fig. 13, Cernay (Zone MP6) and Dormaal (Zone MP7) faunas are placed near the lower and upper part, respectively, of the unconformity corresponding to the major sequence boundary which straddles chronozone C25n. Finally we explore briefly the relationship between the Paleocene/Eocene boundary drawn in terrestrial sections in NW Europe (base Sparnacian) and that drawn in western North America. The Paleocene/Eocene boundary in terrestrial sections of mid-continent North America is generally placed at the Clarkforkian/Wasatchian North American Land Mammal Age (NALMA) boundary (Gunnell et al. 1993) based on correlation with the base of the (European) Sparnacian or within the lower Wasatchian (Lucas & Williamson 1993) based on correlation with the base of the (marine) Ypresian. The Clarkforkian NALMA is characterized by the appearance of new mammalian immigrants of uncertain (?Asian) geographical origin such as rodents (Paramys) and tillodonts (Azygonyx),

together with condylarths (Phenacodus, Ectocion) and proprimates (Plesiadapis). The Wasatchian is characterized by the appearance and diversification of members of new mammalian orders as artiodactyls (Diacodexis), perissodactyls (Hyracotherium) and primates ( Cantius, Teilhardina). The close (temporal) correspondence of the Clarkforkian-Wasatchian transition with the lower Sparnacian is seen in the fact that the Meudon fauna of the Paris Basin heralds the transition from archaiac Paleocene to modern (Eocene) mammals; initial appearance of artiodactyls, perissodactyls, primates, chiropterid bats, (local) introduction of rodents and the concomitant reduction in the diversity of condylarths, plesiadapid 'primates' and multituberculates (Russell et al. 1989). The generally accepted correlation of Zone MP7 European mammalian faunas with the younger part of the Clarkforkian North American Land Mammal Age (NALMA) has had the effect of suggesting to most vertebrate palaeontologists that the two boundaries were slightly diachronous. Correlation between marine and terrestrial stratigraphies was suggested by Wing (1984) through the intermediary of the FAD of Platycarya pollen at the NP9/NP10 zonal boundary on the one hand, and in the early part of the Wasatchian NALMA, on the other, which, in view of the stratal chronology suggested here would serve to reduce this (apparent) diachroneity considerably. Indeed Wing et al. (1991) indicate that strata just below the basal Wasatchian, Wa0 (Gingerich 1989), are correlative with the NP9/NP10 zonal boundary with an estimated age of 55.7 Ma (based on dating of the -17 ash at 55.1 _ 0.3 Ma by Obradovich) which is seen to be remarkably close to the value of 55 Ma estimated here based on an age of 54.5 Ma on the -17 ash by Swisher & Knox (1991). One remaining fly in the ointment is the recognition that the 813C spike in the oceanic record at Southern Ocean ODP Site 690 (Kennett & Stott 1991), NE Atlantic DSDP Sites 549 and 550 (Sinha & Stott 1993a, b; Stott et al. 1996), Pacific Ocean (Shatsky Rise) DSDP Site 577 (Pak & Miller 1992), among others, occurs in Zone NP9 (see Aubry et al. 1996) and is believed to be stratigraphically correlative with the 813C record in palaeosol carbonates and tooth enamel in the terrestrial record at the base of the Wasatchian NALMA (Koch et al. 1992; Koch 1993), as well as in palaeosol carbonates and coexisting organic matter in the lower part of the Argiles plastiques bariolres (basal Sparnacian) in the Limay Quarry of the western part of the Paris Basin (Sinha & Stott 1993a, b; Stott et al. 1996), but not in Zone NP10. It would appear that the Clarkforkian NALMA is stratigraphically equivalent to the basal (pre-813C spike) part of the Sparnacian, whereas the 813C spike serves to link

LATE PALEOCENE--EARLYEOCENE MAGNETOBIOCHRONOLOGYOF NW EUROPE the Clarkforkian/Wasatchian boundary and the 6 m level in the lower Sparnacian in the Limay Quarry. The Clarkforkian NALMA is thus of relatively short duration (c. 0.5 million years) inasmuch as its base (the Tiffanian/Clarkforkian boundary) is within Chron C25n (Butler et al. 1981) with estimated bracketing ages of 55.904-56.391 Ma 1995). This implies that the (Berggren et al. Clarkforkian/Wasatchian boundary lies in Zone NP9 and not in Zone NP10 (Gunnell et al. 1993; see also Wing et al. 1991) and that the original correlation of the FAD of the P l a t y c a r y a pollen flora to the NP9/NP10 zonal boundary may have been due to an incorrect determination of that boundary in Atlantic Coastal Plain sections, or to unrecognized unconformities. This interpretation is reinforced by recent palaeomagnetic investigations on Wasatchian strata from the McCullough Peaks section in the northern Bighorn Basin, Wyoming (Clyde et al. 1994). The base of the Wasatchian (Wa-0) and Zones Wa-1 through the lower part of W-5 (Gingerich 1983, 1989, 1991) are shown to belong to Chron C24r, whereas the mid part of Zone W-5 through midZone W-6 (Lysitean Subage) is correlative with Subchron C24n.3n (terminology in Cande & Kent 1992), the upper Zone W-6 is correlative (in part) with Subchron C24n.2n and the lower part of Zone W-7 (Lostcabinian Subage) with Subchron C24n.lr. The remainder of the Wasatchian (Zone W-7) would then correspond to the > 3 million year interval from Chron C24n. lr to Chron C22n (Flynn 1986; or even longer if the Bridgerian fauna recorded by Flynn is referable to Chron C21n) and represents over half of the duration of the Wasatchian NALMA. The base of the Wasatchian is assigned an age of 55 Ma by Clyde et al. (1994) following Wing et al. (1991); we would estimate this level at 55.5 Ma following the arguments laid out above. In either case the stratigraphic position of the base Wasatchian is demonstrably within the lower half of chronozone C24r by correlation with the stable isotope record in marine sediments. It would appear that we now have definitive data for an accurate correlation between marine and terrestrial stratigraphies over the c. 4 million years interval from Chron C25n (late Thanetian = Clarkforkian) to C24n.ln (early Ypresian=midWasatchian). The location of the base of the Wasatchian (Wa-0) at a stratigraphic level correlative with mid-Zone NP9 (rather than NP10) would appear demonstrated as well (see also Tauxe et al. 1994). Hooker's (1991) suggestion that the Meudon fauna (which lies at the base of the Argile plastique = basal Sparnacian) is more appropriately associated with Zone MP8-9 than Zone MP7 remains the second fly in the ointment of a unified, integrated

345

mammalian correlation framework for the late Paleocene-early Eocene of NW Europe. A terrestrial boundary placed between Zone MP7/MP8-9 faunal levels (leaving aside the affinity of the Meudon fauna for the moment) would appear to coincide approximately - within the limits of stratochronological correlation suggested here owing to hiatuses in the NW European stratigraphic succession - with a level between mid Zone NP9 and the NP9/NP10 (marine) zonal boundary, which is, in turn, c. 0.5-1 million years older than the base of the Harwich Formation (Oldhaven Beds = HalesClay). The age of the Paleocene/Eocene boundary, as currently delineated by the lithostratigraphic base of the Harwich Formation (Hales Clay = Oldhaven Beds) is estimated at c. 54.8 Ma (early-mid part of Chron C24r) based on magnetobiochronological interpolations used here (in Hole 550) or 54.5 Ma based on recent radioisotopic dating of the -17 ash. A more correct estimate of age of the Paleocene/ Eocene boundary would link this boundary with the base of the (type) leper Clays in Belgium which serve as the reference for the Ypresian Stage of the lower Eocene. That level is estimated here to lie one sequence/cycle above the base of the Harwich Formation and, consequently, have an estimated age of 54.6 or 54.3 Ma. The boundary as drawn by (most) calcareous nannoplankton micropalaeontologists at the NP9/NP10 zonal boundary is located at c. 55 Ma, within the early to mid Chron C24r. Some geologists have recently suggested that the 813C spike can serve as a criterion for denoting (not defining) the Paleocene/Eocene boundary, inasmuch as it has now been observed in both marine and terrestrial stratigraphies. The importance of identifying the 613C spike in the basal Wasatchian in the Big Horn Basin of North America, the lower Sparnacian in the Paris Basin and in mid-Zones NP9 and P5 in the deep sea lies in the fact that it provides an unequivocal first order global correlation between disjunct marine and terrestrial stratigraphies at a correlative level in the early part of Chron C24r. That level is estimated here to have an age of c. 55.5 Ma (within the constraints of the GPTS; Cande & Kent 1995).

Correlation of NW European and North Sea stratigraphies During the Paleocene and early Eocene, the North Sea Basin was characterized by a series of deltaic (shelf) and submarine fan (slope) deposits overlying lower Paleocene (Danian) carbonates (Gradstein et al. 1994). The lower Paleogene sediments of the North Sea Basin, ranging in thickness

346

W.A. BERGGREN & M . E AUBRY

from less than 200 m in the southern part of the Central Graben to nearly 1500 m in the Witch Ground Graben to the northwest, closer to the source area of Scotland and the Shetland Platform, reflect the combined influence of regional/ differential subsidence of the North Sea Basin, regional uplift of the Scottish Highlands and the nascent rifting of the NE Atlantic. These deposits have been subdivided into 10 depositional sequences based on integrated seismic and biostratigraphy (Stewart 1987). In Fig. 13 we have attempted to translate the upper Paleocene stratigraphic succession of NW Europe into a temporal framework and provide estimates of the duration of the hiatuses corresponding to the unconformities separating the major lithostratigraphic units (Aubry 1983, 1985, 1986). Stewart (1987) noted that the depositional unconformities/hiatuses in the NW European marginal basins correspond to the regressive, lowstand systems tracts of the North Sea Basin, whereas the lithic units preserved on land correspond essentially to the transgressive, highstand systems tracts. If the influence of fluctuating sea level is the primary control on depositional histories (see Cloetingh et al. 1987 for a view that emphasizes the primacy of tectonics in the form of stress fields in the crust), it should be possible to effectuate an approximate correlation between the depositional histories of the North Sea sequences and those in the marginally situated NW European margins (Fig. 13). The main features include the following. 1. Sequence 7, separating the Montrose (below) and Moray (above) Groups, is characterized by predominantly lowstand submarine fan deposits (Forties Formation) and succeeding transgressive hemipelagic mudstones which were deposited during the regional hiatus in NW Europe i.e., between the Sables du Tillet and the Sables de Bracheux (HP3), between the Reculver Silts and the Upnor Formation, and between the Bois Gilles and Tuffeau de Lincent. Sequence 7 lies within the Apectodinium augustum (dinoflagellate) Zone whose base corresponds to the top of Alicocysta margarita (Powell 1988, p. 331, fig. 2) in Sequence 6 (below). The Deflandrea speciosa/Apectodinium hyperacanthum (W1) zonal boundary is generally correlated with the NP8/NP9 zonal boundary in NW Europe (Powell 1988, p. 33I, fig. 2). This boundary is seen to be unconformable, however, and it may be that the dinoflagellate zonal boundary actually does correspond to the NP8/NP9 chronozonal boundary which lies within Chron C25n, some 0.3-0.4 million years older. This is difficult to ascertain, however, because the A. augustum and succeeding C. dartmooria Zones are based on LADs, whereas the A. hyperacanthum Zone is based on a FAD (Powell 1988).

2. Sequence 8 consists of prograding (predominantly) deltaic mudstones equivalents belonging the upper part of the A. augustum Zone and is correlative with the transgressive Upnor Formation-Sables de Bracheux-Bois Gilles units of NW Europe which belong to lower Zone NP9. Delta plain lignites occur in the upper part of Sequence 8 and are likely correlative with the basal Sparnacian Argile plastique and Argile ~ lignites du Soissonais in the western Paris Basin. 3. Sequence 9 is bracketed at the base by the onlap of seismic reflectors on Sequence 8 delta plain lignitiferous sandstones and mudstones and at the top by a distinct seismic reflector corresponding to the top of pyroclastic phase 2c in the North Sea at a level equivalent to Zone N P l l (Knox & Morton 1988; Morton & Knox 1990). It corresponds to the ash-bearing lowstand submarine fans of the Sele Formation and the succeeding (transgressive) Balder Formation. The Sele Formation corresponds, in large part, to the major unconformity between the Woolwich and Reading formations and Oldhaven Beds and the equivalent unconformity at the base of the Ieper Clay Formation. In Fig. 13 the base of pyroclastic phase 2a is shown extending questionably into the terminal part of Zone NP9 because of the reported occurrence of volcanic ashes in sediments (Erquelinnes) belonging to this zone. 4. Sequence 10 is a highstand system tract consisting of (predominantly) reddish (below) and grey (above) mudstones equivalent to the (upper) London Clay Formation and lower Bracklesham (Wittering Formation) Group and the (upper) leper Clay Formation and Egem-Panisel Sands. The lower, reddish mudstones have their counterpart in the RCsnaes (red clay) Formation of Denmark and the upper part of the London Clay and leper formations where the brief incursion of a planktonic foraminiferal fauna (equivalent to Zone P6b) and a Zone NP11 calcareous nannoplankton allows correlation to Subchron C24n.3n (Cande & Kent 1995). It is clear that a sequence- and magnetobiostratigraphic approach provides a unifying correlation framework within which to delineate the disjunct depositional histories of marginal and deep marine basinal stratigraphies and allows the stratigrapher to move from rock to time.

Significance of East GreenlandHebridean province volcanism dates on revised time scale It is generally agreed that the Faeroe-East Greenland (Blosseville coast) basalts were extruded during an approximately 3 million years

LATE PALEOCENE--EARLY EOCENE MAGNETOB1OCHRONOLOGY OF NW EUROPE period associated with Chron C24r and that Phase 2a and 2b ashes are the expression of these eruptions in the North Sea and N W European basins. In the time scale of Berggren et al. 1985, Chron C24r was estimated to span the interval of c. 5 9 - 5 6 Ma. In a review of the Blosseville coast basalts Noble et al. (1988) estimated that the extrusion took place in the interval of 57-53 Ma. In a similar and contemporary review of igneous activity in the British Tertiary Igneous Province (BTIP) Mussett et al. (1988) observed that volcanic activity occurred within the interval of c. 63-52 Ma and, like Noble et al. (1988), they pointed out that the isotopic dates were not fully reconcilable with the magnetochronological estimates of any of the then current 'time scales' (Harland et al. 1982; Berggren et al. 1985; Odin & Curry 1985). Chron C24r is now seen to have a 2.557 million years duration from 53.347-55.904 Ma (Cande & Kent 1992, 1995) in line with estimates based on isotopic dates from Faeroe-East Greenland and BTIP. The estimate of Cande & Kent (1995) and Berggren et al. (1995) for the age of Chron C 2 6 n of 57.554-57.911 Ma is also consistent with other more recent radioisotopic ages of c. 58 Ma for Chron C26n (Parson & Morton 1988). In short the revised magnetochronology presented in Cande & Kent (1992, 1995) would appear to reconcile the

347

discrepancies between measured ages and palaeom a g n e t i c signature o f NE Atlantic early Paleogene volcanic activity and various published chronologies (including Harland et al. 1989). We acknowledge the generous cooperation and helpful discussions we have had with members of IGCP Project 308 over the past few years (among others, C. Cavelier, I. Cojan, R. M. Corfield, C. Dupuis, J .J. Flynn, T. Gibson, J. Hardenbol, J. J. Hooker, C. King, R. W. O'B. Knox, P. Laga, K. G. Miller, D. Pak, C. C. Swisher, M. Thiry and E. Thomas). We particularly acknowledge the advice and insight of J. Hardenbol in placing the NW European stratigraphic succession in a sequence stratigraphic framework; of L. Stott and A. Sinha in our mutual attempts to integrate stable isotope and calcareous plankton bio stratigraphy of various deep sea drilling sites through the Paleocene-Eocene boundary interval; and of Dennis V. Kent regarding implications of data presented here for Paleogene geochronology. They have played no small part in the formulation of our ideas and interpretation, although we accept full responsibility for the conclusions presented here. Helpful comments and critique of early versions of the manuscript of this paper by J. Hardenbol, R. W. O'B. Knox, K. G. Miller, R. Norris, D. Pak, B. Schmitz, E. Thomas and M. Woodburne have materially improved the final version. This paper is a contribution to IGCP Project 308: Paleocene/Eocene Boundary Events in Time and Space. This is Woods Hole Oceanographic Institution Contribution No. 8855 and ISEM Contribution No. 95045.

R e f e r e n c e s

ALI, J. R. 1994. Magnetostratigraphy of the upper Paleocene and lowermost Eocene of SE England. GFF, 116, 41-43. & JOLLEY,D. W. 1996. Chronostratigraphic framework for the Thanetian and lower Ypresian boundary deposits of southern England. This volume. - - , HAmWOOD,E.A. & KING, C. 1996. The 'Oldhaven magnetozone' on East Anglia: a revised interpretation. This volume. --, KING, C. & HAILWOOD, E. A. 1992. Magnetostratigraphic calibration of Ypresian depositional sequences on the southern North Sea Basin. In: HAILWOOD, E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. ANDERSEN, S. A. 1937. De vulskanske Askelag i Veijgennemskaeringen ved Olst og deres Udbredelse i Danmark. Danmarks Geologiske UndersOgelse, 59. APt'LEGATE, J. L. & WISE, S. W., JR. 1987. Eocene calcareous nannofossils, Deep Sea Drilling Project Site 605, upper continental rise off New Jersey, U.S.A. In: VANHINTE, J. E., WISE, S. W., JR. ET AL. Initial Reports of the Deep Sea Drilling Project, 93, 685--698. AuBRv, M.-P. 1983. Biostratigraphie du Pal~ogene Epicontinental de l'Europe du Nord-ouest. Etude Fond~e sur les Nannofossiles Calcaires. Documents des Laboratoires de G6ologie de Lyon, 89.

1985. Northwestern European Paleogene magnetostratigraphy, biostratigraphy and paleogeography: Calcareous nannofossil evidence. Geology, 13, 198-202. 1986. Paleogene calcareous nannoplankton biostratigraphy of northwestern Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 55, 267-334. - - - , BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD, K. D., OBRADOVICH,J. D. & PROTHERO, D. R. 1988. Paleogene geochronology: an integrated approach. Paleoceanography, 3 (6), 707-742. --, --, STOTT, L. & SIYr~A, A. 1996. The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology. This volume. , HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous nannofossil stratigraphy of the lower Palaeogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735. BERGGREN,W. A. 1971. Tertiary boundaries. In: FUNNELL, B. F. & RmDEL, W. R. (eds) Micropaleontology of the Oceans. Cambridge University Press, 693-809. - & MILLER, K. G. 1988. Paleogene tropical planktonic foraminiferal biostratigraphy and magnetobiochronology. Micropaleontology, 34 (4), 362-380. -

-

348

w . A . BERGGREN & M. E AUBRY

& SCHNITKER,D. 1983. Cenozoic marine environments In: The North Atlantic and NorwegianGreenland Sea. In: BOTT, M., SAXOV, O., TALWANI, M. & THIEDE,J. (eds) Structure and Development of the Greenland-Scotland Ridge. Plenum Publishing Corporation, 495-548. -. , KENT, D. V. & FLYNN, J. J. 1985. Palaeogene geochronology and chronostratigraphy. In: SNELLING, N. J. (ed.) Geochronology of the Geological Record. Geological Society, London, Memoir, 10, 141-195. --, OBRADOVICH, J. D. & SWISHER, C. C. III 1992. Toward a revised Paleogene geochronology. In: PROTHERO, O. R. & BERGGREN, W. A. (eds) Eocene-Oligocene Climatic and Biotic Evolution. Princeton University Press, Princeton, 29-45. --, --, SWISHER, C. C. I I I & AUBRY, M.-E 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A., KENT, D. V., AUBRY, M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation; Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Publication, 54. BLEIL, U. 1985. The magnetostratigraphy of northwest Pacific sediments, Deep Sea Drilling Project Leg 86. In: HEATH,G. R., BURCKLE,L. H., ET AL. Initial Reports of the Deep Sea Drilling Project, 86, 441-458. BLOW, W. H. 1979. The Cainozoic Globigerinida (3 vols). E. J. Brill, Leiden. BOGGILD, O. B. 1918. Den vulkaniske Aske i Moleret samt en Oversigt over Danmarks ~eldre Terti~erbj ~eregarter. Danmarks Geologiske UndersOgelse, Raekke H, 33, 1-142 (French summary, 145-155). BOLLI, H. M. 1957. The genera Globigerina and Globorotalia. In: The Paleocene-Lower Eocene Lizard Springs Formation of Trinidad, B.W.I. Bulletin of the US National Museum, 215, 61-81. Born)E, N. 1974. Palaeoenvironment as indicated by the "Mo-Clay Formation" (lowermost Eocene of Denmark). Tertiary Times, 2, 29-36 BUTLER, R. F., GINGERICH,P. D. & LrNDSAY,E. H. 1981. Magnetic polarity stratigraphy and biostratigraphy of Paleocene and lower Eocene continental deposits, Clark's Fork Basin, Wyoming. Journal of Geology, 89, 299-316. CANOE, S. C. & KENT, D. V. 1992. A new geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 97 (B10), 13 917-13 951. & -1995. Revised calibration of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100 (B4), 6093--6095. CANUDO, J. I. & MOLrNA, E. 1992. Planktic foraminifera turnover and biochronostratigraphy of the Paleocene-Eocene boundary at Zumaya section (northern Spain). Revista Sociedad Geologica de Espana, 5 (1-2), 145-157. CARO, Y., LUTERBACHER, H. P., PERCH-NIELSEN, K., PREMOLI SILVA,I., RIEDEL,W. R. & SANFILIPPO,A. --

1975. Zonations ~ l'aide de microfossiles p61agiques du Pal6oc~ne sup6rieur et del l'Eoc~ne inf&ieur. Bulletin de la Socigt~ g~ologique de France, 17, 125-147. CAVELIER, C. & POMEROL, C 1986. Stratigraphy of the Paleogene. Bulletin de la Soci~t~ G~ologique de France, 8 (II,2), 255-265. CHATEAUNEUF,J.-J. & GRUAS-CAVAGNETTO,C. 1978. Les zones de Wetzeliellacea (Dinophyceae) du Bassin de Paris. Bulletin du Bureau de Recherches Gdologiques et Minikres, 4 (2), 59-93. CIESELSKX, P. F., KRISTOFFERSEN, Y. ET AL. 1988. Proceedings of the Ocean Drilling Program, Initial Reports, 114. College Station, TX. CLOETINGH, S., LAMBECK, K. & MCQUEEN, H. 1987. Apparent sea-level fluctuations and a palaeostress field for the North Sea region. In: BROOKS, J. & GLENNIE, K. (eds) Petroleum Geology of North West Europe, Graham & Trotman, London, 49-57. CLYDE, W. C., STAMATAKOS,J. & GINGERICH,P. D. 1994. Chronology of the Wasatchian Land Mammal Age (early Eocene): Magnetotratigraphic results from the McCullough Peaks section, northern Bighorn Basin, Wyoming. Journal of Geology, 102, 367-377. CORFIELD, R. M. 1987. Patterns of evolution. In: Palaeocene and Eocene planktonic Foraminifera. In: HART, M (ed.) The Micropaleontology of Carbonate Environments. Ellis Horwood, Chichester, 93-110. ~-, CARTLIDGE,J. E., PREMOLI SILVA, I. & HOUSLEY, R. A. 1991. Oxygen and carbon isotope stratigraphy of the Palaeogene and Cretaceous limestones in the Bottacione Gorge and the Contessa Highway sections, Umbria, Italy. Terra Nova, 3, 414--422. Cox, E C., HAILWOOD,E. A., HARLAND,R., HUGUES,M. J., JOHNSTON, N. & KNOX, R. W. O'B. 1985. Palaeocene sedimentation and stratigraphy in Norfolk, England. Newsletters in Stratigraphy, 14, 169-185. CRUX, J. A. 1991. Calcareous nannofossils recovered by Leg 114 in the subantarctic South atlantic Ocean. In: CIESELSKI, P. F., KRISTOFFERSEN, Y. ET AL. Proceedings of the Ocean Drilling Program, Initial Reports, 114, 155-177. CURRY, D. 1965. The Palaeogene beds of SE England. Proceedings of the Geologists' Association, 77, 437-467. ~, ADAMS, C. G., BOULTER, M. C., DILLEY, E C., LAMES, E E., FUNNELL, B. M. & WELLS, M. K. 1978. A Correlation of Tertiary Rocks in the British Isles. Geological Society London, Special Report, 12, 1-72. DE CONrNCK, J. 1990. Ypresian organic-walled phytoplankton in the Belgian Basin and adjacent areas. Bulletin de la Societ( Beige de Gdologie, 97 (3-4), 287-319. DEEGAN, C. E. & SCULL, B. J. 1977. A standard lithostratigraphic nomenclature for the central and northern North Sea. Institute of Geophysical sciences report 77/25. DOLLFUS, G. E 1880. Essai sur l'extension des terrains tertiares darts le bassin anglo-parisien. Bulletin

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE

Socigt~ G~ologique du Normandie, 6 (1979), 584-605. DuPuIS, C., DE CONINCK, J., GUERNET, C. & ROCHE, E. 1990. Biostratigraphic data - ostracods and organic walled microfossils - of the Landen Formation and the base of the Ieper Formation. In: LAGA, P. 8Z VANDENBERGHE, N. (eds) The Knokke Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M~moires pour servir ~ l'Explication des Cartes G~ologiques et Mini&es de la Belgique, 29, 33-45. , & STEURBAUT, E. 1991. The Ypresian Stratotype. Bulletin de la Soci~t~ Belge de G~ologie, 97 (3/4). ELLISON, R. A., ALI, J. R., HINE, N. M. & JOLLEY, D. W. 1996. Recognition of Chron 25n in the upper Paleocene Upnor Formation of the London Basin, UK. This volume. , KNOX, R. W. O'B., JOLLEY,D. W. & liNe, C. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. FENNER, J. 1994. Diatoms of the Fur Formation, their taxonomy and biostratigraphic interpretation. Results from the Harre borehole, Denmark. Aarhus Geoscience, 1, 99-163. FLYNN, J. J. 1986. Correlation and geochronology of middle Eocene strata from the western United States. Palaeogeography, Palaeoclimatology, Palaeoecology, 55, 35-406. GL~GEPOCH, P. D. 1983. Paleocene-Eocene faunal zones and a preliminary analysis of Laramide structural deformation. In: The Clarks Fork Basin, Wyoming. 34th Annual Field Conference Guidebook. Wyoming Geological Association, 185-195. - 1989. New earliest Wasatchian mammalian fauna from the Eocene of northwestern Wyoming: composition and diversity In: A rarely sampled high-floodplain assemblage. University of Michigan Papers on Paleontology, 28, 1-97. 1991. Systematics and evolution of early Eocene Perissodactyla (Mammalia) In: The Clark Forks Basin, Wyoming. Contributions Museum Paleontology University of Michigan, 28, 181-213. GRAC1ANSKV,P. C. DE, POAC, C. W. ETAL. 1985. Site 550, Initial Reports of the Deep Sea Drilling Program, 80, 251-355. GRADSTEIN, E M., KAMINSKI, M. A., BERGGREN, W. A., KRISTANSEN,L. L. & D'IORO, M. A. 1994. Cenozoic biostratigraphy of the North Sea and Labrador shelf. Micropaleontology, 40 (supplt), 1-152. G ~ L L , G., BARTELS,W. S. & GINGERICH, P. D. 1993. Paleocene-Eocene boundary In: Continental North America: biostratigraphy and geochronology, northern Bighorn Basin, Wyoming. In: LUCAS, S. P. & ZEDEK, J. (eds.) Vertebrate Paleontology in New Mexico. New Mexico Museum of Natural History and Science, Bulletin, 2, 137-144. HANSEN, J. M. 1979. Age of the Mo-Clay Formation. Bulletin of the Geological Society of Denmark, 27, 89-91. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987.

349

Chronology of fluctuating sea levels since the Triassic. Science, 235, 1156-1167. HARLAND, W. B., ARMSTRONG,R. L., Cox, A. V., CRAIG, L. E., SMITH, A. G. r SMITH, D. G. 1989. A Geologic Time Scale. Cambridge University Press. --, Cox, A. V., LLEWELLYN,P. G., PICKTON, C. A. G., SMITH, A. G. & WALTERS,R. 1982. A Geologic Time Scale. Cambridge University Press. HEILMANN-CLAUSEN, C. 1982. The Paleocene-Eocene boundary in Denmark. Newsletters on Stratigraphy. 11 (2), 55-63. 1985. Dinoflagellate stratigraphy of the uppermost Danian to Ypresian in the Viborg 1 borehole, central JyUand, Denmark. Danmarks Geologiske Unders0gelse, Series A, 7. HOOKER, J. J. 1991. The sequence of mammals in the Thanetian and Ypresian of the London and Belgian Basins. Location of the Paleocene-Eocene boundary. Newsletters on Stratigraphy, 25 (2), 75-90. JOLLEY, D. W. 1996. The earliest Eocene sediments of eastern England: an ultra high resolution palynological correlation. This volume. KATZ, M. & MILLER, K. G. 1991. Early Paleogene benthic foraminiferal assemblages and stable isotopes in the Southern Ocean. In: CrESmLSra, P. E, KR1STOFFERSEN,Y. ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 114, 481-512. KEN~NErr, J. P. & STOTT, L. D. 1991. Abrupt deep-sea warming, paleoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229. lING, C. 1981. The Stratigraphy of the London Clay and Associated Deposits. Tertiary Research Special Paper, 6, Backhuys, Rotterdam. - 1990a. Eocene stratigraphy of the Knokke borehole (Belgium). In: LAGA, P. & VANDENBERGHE,N. (eds) The Knokke Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M~moires pour servir ~ l'Explication des Cartes G~ologiques et Mini~res de la Belgique, 29, 67-102. 1990b. Stratigraphy of the Ieper Formation and Argile de Flandres (early Eocene) in western Belgium and northern France. In: LAGA, P. & VANDENBERGHE, N. (eds) The Knokke Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M~moires pour servir gt l'Explication des Cartes G~ologiques et Minidres de la Belgique, 29, 349-372. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. - 1985. Stratigraphic significance of volcanic ash. In: Paleocene and Eocene sediments at Sites 549 and 550. In: DE GRACIANSKY,P. C, POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 845-850. 1990. Thanetian and early Ypresian chronostratigraphy in south-east England. Tertiary Research, 11 (2--4), 57--64.

350

W.A. BERGGREN i:~ M. P. AUBRY

1994. From regional stage to standard stage: Implications for the historical Paleogene stratotypes of NW Europe. GFF, 116, 56, 57. & MORTON,A. C. 1983. Stratigraphic distribution of early Palaeogene pyroclastic deposits in the North Sea. Proceedings of the Yorkshire Geological Society, 44, 355-363. -& -1988. The record of early Tertiary North Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. , HINE, N. M. & ALl, J. R. 1994. New information on the age and sequence stratigraphy of the type Thanetian of southeast England. Newsletters on Stratigraphy, 30 (1), 45-60. , MORIGI, A. N., ALI, J. R., HAILWOOD, E. A. & HALLAM, J. R. 1990. Early Paleogene stratigraphy of a cored borehole at Hales, Norfolk. Proceedings of the Geologists' Association, 101, 15-151. , MORTON, A. C. & HARLAND, R. 1981. Stratigraphical relationships of Palaeocene sands in the UK sector of the Central North Sea. In: ILLING, L. V. & HOBSON, G. D. (eds) Petroleum Geology of the Continental Shelf of North West Europe. Heyden & Sons, London, 267-281. KOCH, R L. 1993. Chemostratigraphic correlation of marine and continental strata across the Paleocene/Eocene boundary. Journal Vertebrate Paleontology, 13 (3) (Supplement) (Abstracts of 53d Annual meeting of Society of Vertebrate Paleontologists). , ZACHOS,J. C. & GINGERICH,P. D. 1992. Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary. Nature, 358, 319-322. KRASHENINNIKOV,V. 1979. Stratigraphy and planktonic foraminifers of Cenozoic deposits of the Bay of Biscay and Rockall Plateau, DSDP Leg 48. In: MONTADERT, L., ROBERTS, D. G. ET AL. Initial Reports of the Deep Sea Drilling Project, 48, 431-450. LAGA, P. & VANDENBERGHE,N. 1990. The Knokke Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. Mdmoires pour servir b l'Explication des Cartes Gdologiques et Minikres de la Belgique, 29. LANG, T. H. & WISE, S. W. 1987. Neogene and Paleocene-Maestrichtian calcareous nannofossil stratigraphy, Deep Sea Drilling Project Sites 604 and 605, upper continental rise of New Jersey: sedimentation rates, hiatuses, and correlations with seismic stratigraphy. In: VAN HINTE, J. E., WISE, S. W., JR. er al~. Initial Reports of the Deep Sea Drilling Project, 93, 661-683. LAURAIN, M. L. & GUERIN, H. 1989. Progrbs dans la connaissance du C6nozoique dans la rfgion r6moise. In: l14e CongrOs National de la Societd des Savants de Paris: Ggologie Bassin Parisien, 71-82. , BARTA, C., BOLIN, C., GUERNET, C., GRUASCAVAGNETTO,C., LOUIS, P., PERREAU,M., RIVELINE, J. & THIRY, M. 1983. Le sondage et la coupe du Mont Bernon a Epernay (Marne). Etude s6di-

-

-

mentologique et pal6ontologique du stratotype du Sparnacien et de la s6rie 6oc~ne. Gdologie de France, 3, 235-254. LUCAS, S. G. & WILLIAMON, T. E. 1993. Eocene vertebrates and late Laramide stratigraphy of New Mexico. In: LUCAS, S. G. & ZEDEK, J. (eds) Vertebrate Paleontology in New Mexico. New Mexico Museum of Natural History and Science Bulletin, 2, 145-158. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the Second Planktonic Conference. Tecnoscienza, Rome, 739-785. MILLER, K. G. & TUCHOLKE, B. E. 1983. Evidence for development of abyssal circulation south of the Greenland-Scotland Ridge. In: BOTT, M. H. E, SAXOV, S., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. Plenum Publishing Corporation, New York, 549-589. --, JANECEK, T. R., KATZ, M. E. & KEIL, D. J. 1987. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary. Paleoceanography, 2 (6), 741-761. MONECHI, S. 1985. Campanian to Pleistocene calcareous nannofossil stratigraphy from the northwest Pacific Ocean, Deep Sea Drilling Project Leg 86. In: HEATH, G. R., BURCKLE,L. H., ET AL. Initial Results of the Deep Sea Drilling Project, 86, 301-336. --, BLEIL, U. & BACKMAN, J. 1985. Magnetobiochronology of Late Cretaceous - Paleogene and late Cenozoic pelagic sedimentary sequences from the northwest Pacific (Deep Sea Drilling Project Leg 86, Site 577). In: HEATH,G. R., BURCKLE,L. H., ET AI.. 1985. Initial Results of the Deep Sea Drilling Project, 86, 787-797. MORTON, A. C. & KNOX, R. W. O'B. 1990. Geochemistry of late Palaeocene and early Eocene tephras from the North Sea Basin. Journal of the Geological Society, London, 147, 425-437. & PARSON, L. M. (eds) 1988. Early Tertiary volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39. --, BACKMAN,J. & HARLAND,R. 1983. A reasessment of the stratigraphy of DSDP Hole 117A, Rockall Plateau: implications for the Palaeocene-Eocene boundary in NW Europe. Newsletters on Stratigraphy, 12, 104-111. MOLLER, C. 1979. Calcareous nannofossils from the North Atlantic (Leg 48). In: MONTADERT, L., ROBERTS, D. G. ET AL. Initial Reports of the Deep Sea Drilling Project, 48, 589-639. 1985. Biostratigraphic and paleoenvironmental interpretation of the Goban Spur region based on a study of calcareous nannoplankton: In: DE GRACIANSKY, E C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 573-599. MUSSETT, A. E., DAGLEY,P. & SKELHORN,R. R. 1988. Time and duration of activity in the British Tertiary igneous province. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the North East Atlantic. Geological Society, London, Special Publication, 39, 337-348.

LATE PALEOCENE--EARLY EOCENE MAGNETOBIOCHRONOLOGY OF NW EUROPE NIELSEN, O. B. & HEILMANN-CLAUSEN, C. 1988. Palaeogene volcanism: the sedimentary record in Denmark. In: MORTON,A. C. & PARSON,L. M. (eds)

Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 395-405. NOBLE, R. H., MCINTYRE, R. i . & BROWN, P. E. 1988. Age constraints on Atlantic evolution: timining -~. of magmatlc activity along the E Greenland continental margin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the North East Atlantic. Geological Society, London, Special Publication, 39, 201-214. NOCCHI, M., AM1CI, E., & PREMOLI SILVA, I. 1991. Planktonic foraminiferal biostratigraphy and paleoenvironmental interpretation of Paleogene faunas from the Subantarctic transect, Leg 114. In: CIESIELSKI, P. F., KRISTOFFERSEN, Y., E T AL. Proceedings Ocean Drilling Program, Scientific Results, 114, 233-279. ODIN, G. S. & CURRY, D. 1985. Paleogene time-scale: radiometric versus magnetostratigraphic approach. Journal of the Geological Society, London, 142, 1179-1188. OKADA, H. & BUKRY, D. 1980. Supplementary modification and introduction of code numbers to the low-latitude coccolith biostratigraphic zonation. Marine Micropaleontology, 5, 321-325. PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene beznthic foraminiferal isotopes and assemblages: implications for deep-water circulation. Paleoceanography, 7 (4), 405-422. PARSON, L. M. & MORTON, A. C. 1988. Introduction. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, ix-xii. PERCH-NIELSEN, K. 1976. New silicoflagellates and a silicoflagellate zonation, ln: North European Palaeocene and Eocene diatomites. Bulletin of the Geological Society of Denmark, 25, 27-40. POMEROL, C. 1973. Ere C~nozoique (Tertiaire et Quaternaire). DOIN, Paris. 1989. Stratigraphy of the Palaeogene: hiatuses and transitions. Proceedings of the Geologists' Association, 100 (3), 313-324. POSPICHAL, J. J. & WISE, S. W. JR. 1990. Maestrichtian calcareous nannofossil biostratigraphy of Maude Rise ODP Leg 113 Sites 689 and 690, Weddell Sea. In: BARKER, P. F., KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 465-487. POWELL, A. J. 1988. A modified dinoflagellate cyst biozonation for latest Paleocene and earliest Eocene sediments from the central North Sea. Review of Palaeobotany and Palynology, 56, 327-334. PREMOLISILVA,I. & BOLLI, H. M. 1973. Late Cretaceous to Eocene planktonic foraminiera and stratigraphy of Leg 15 sites in the Caribbean Sea. ln: EDGARD, N. T., SAUNDERS, J. B., ET AL. Initial Reports of the Deep Sea Drilling Project, 15, 449-547 ROMEIN, A. J. T. 1979. Lineages in Early Paleogene Calcareous Nannoplankton. Utrecht Micropaleontological Bulletin, 22.

351

RUSSELL, D. E., BIGNOT, G., GALOYER,A., GUERNET,C., POMEROL, C., RIVELINE, J., SEN, S., THmY, M. & TOURENQ, J. 1993. De la Craie ~ l'Argile plastique: un affleurement remarkable h Meudon pros de Paris. Bulletin d'lnformation des Gdologues du Bassin de Paris, 30 (2), 3-9. --, THmY, M. & GALOYER, A. 1989. Le Conglom6rat de Meudon-hier et aujourd'hui. In: l14e Congrks National de la Socidtd des Savants de Paris 1989, Gdologie Bassin Parisien, 305-327. SAINT-MARC, P. 1987. Biostratigraphic and paleoenvironmental study of Paleocene benthic and planktonic foraminifers, Site 605, Deep Sea Drilling Project Leg 93. In: VAN HINTE, J., WISE, S. W., JR., ET AL. Initial Reports of the Deep Sea Drilling Project, 93, 539-547. SCHNtTKER, D. 1979. Cenozoic deep water benthic foraminifers, Bay of Biscay. In: MONTADERT, L., ROBERTS, D. G., ET AL. Initial Reports of the Deep Sea Drilling Project, 48, 377--414. SHACKLETON, N. J., HALL, M. A. & BOERSMA, A. 1984. Oxygen and carbon isotope data from Leg 74 foraminifers. In: MOORE, T. C. JR, RABINOW~TZ,E, et al. Initial Reports of the Deep Sea Drilling Project, 74, 599-612. --, HALL, M. A. & BLEIL, U. 1985. Carbon isotope stratigraphy, Site 577. In: HEATH,G. R., BURCKLE, L. H., ET AT. 1985. Initial Reports of the Deep Sea Drilling Project, 86, 503-511. SmSSER, W. G., WARD, D. J. & LORD, A. R. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage (Palaeocene) in the type area. Journal of Micropaleontology, 6, 85-102. SINHA, A. & STOTT, L. D. 1993a. Recognition of the Paleocene/Eocene boundary in the northwestern European stratotype sections on the basis of carbon isotope stratigraphy. Journal of Vertebrate Paleontology, 13 (3), Supplement, 57A (Abstracts of Papers, 53rd Annual Meeting of Society of Vertebrate Paleontology). & -1993b. Recognition of the Paleocene/ Eocene boundary carbon isotope excursion in the Paris Basin, France. In: Symposium on Correlation of the Early Paleogene in Northwest Europe, 1-2 December 1993, Programme and Abstracts. Geological Society, London. SNYDER, S. W. & WATERS, V. J. 1985. Cenozoic planktonic foraminiferal biostratigraphy of the the Goban Spur region, Deep Sea Drilling Project Leg 80: In: DE GRACIANSKY,P. C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 439-472. SPmss, V. 1990. Cenozoic magnetostratigraphy of Leg 113 drill sites, Maude Rise, Weddell Sea, Antarctica. In: BARKER,P. E, KENNETT,J. P., ET AL. 1990. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 261-315. STEURBAUT,E. 1988. New early and middle Eocene calcareous nannoplankton events and correlation in middle to high latitudes of the northern hemisphere. Newsletter on Stratigraphy, 18 (2), 99-115. 1990a. Calcareous nannoplankton assemblages from the Tertiary In: The Knokke borehole. In: LAGA, P. & VANDENBERGHE,N. (eds) The Knokke

352

w . A . BERGGREN •

Well (11E/138) with a description of the Den Haan (22W/276) and Oostduinkerke (35E/142) wells. M~moires pour servir & l'Explication des Cartes Gdologiques et Minikres de la Belgique, 29, 47-66. 1990b. Ypresian calcareous nannoplankton biostratigraphy and paleogeography of the Belgian Basin. In: Dupuis, C., DE CON~NCK, J. & STEURBAtrr, E. (eds) Bulletin de la Socidtg Beige de G~ologie, 97 (3-4), 251-285. - & NOLF, D. 1986. Revision of Ypresian stratigraphy of Belgium and northwestern France. Mededelingen van der Werkgroep voor Tertiare en Kwartair Geologie, 23 (4), 115-172. STEWART,I. J. 1987. A revised stratigraphic interpretation of the Early Palaeogene of the central North Sea. In: BROOKS, J & GLENNIE, K. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 557-576. STOTr, L. D. & KENNETT,J. P. 1990. Antarctic Paleogene planktonic foraminifer biostratigraphy: ODP Leg 113, Sites 689 and 690. In: BARKER,P. E, KENNETT, J. P., Er AL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 549-569. - - , - - , SHACKLETON,N. J. & CORFIELD,R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: Inferences from the stable isotopic composition of planktonic foraminifers, ODP Leg 113. In: BARKER, P. E, KENNETT, J. P. ET AL. Proceedings of the Ocean Drilling Program, Scientific" Reports, 113, 849-863. , SINHA, A., THIRY, M., AUBRY,M.-E & BERGGREN, W. A. 1996. Global ~13C changes across the Paleocene-Eocene boundary: criteria for terrestrialmarine correlations. This volume. SWISHER,C. C. III & KNOX, R. W. O'B. 1991. The age of the Paleocene/Ex~cene boundary: 4~ dating of the lower part of NP 10, North Sea Basin and Denmark. In: IGCP 308 (Paleocene/Eocene Boundary Events) International Annual Meeting and Field Conference, 2-6 December 1991, Brussels, Abstracts with Program. TAUXE, L., GEE, J., GALLET, Y., PICK, T. & BROWN, Z. 1994. Magnetostratigraphy of the Willwood Formation, Bighorn Basin, Wyoming: New

M. E AUBRY

constraints on the location of the Paleocene/Eocene boundary. Earth and Planetary Sciences Letter, 125, 159-172. THLRV,M. 1981. S6dimentation continentale et alt6rations associ6es: calcitisations, ferruginisations, et silicifications. Les argiles plastiques du bassin de Paris. Sciences Ggologiques, Mdmoire, 64. THOMAS, E. 1990. Late Cretaceous through Neogene deep-sea benthic foraminifers (Maude Rise, Weddell Sea, Antarctica). In: BARKER, P. E, KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 571-594. -& SHACr,LETON, N. 1996. The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies. This volume. --, BARERRA, E., HAMILTON, N., HUBER, B. T., KENNETT, J. P., O'CONNELL, S. B., POSPICHAL,J. J., SPIESS, V., STOTT, L. D., WEI, W. & WISE, S. W. JR. 1990. Upper Cretaceous-Paleogene stratigraphy of Sites 689 and 690, Maude Rise (Antarctica). In: BARKER, P. E, KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 901-914. TOUMARKINE, M. & LUTERBACHER, H. 1985. Paleocene and Eocene planktonic foraminifera. In: BOLLI, H. M., SAUNDERS, J. B. & PERCH-NIELSEN, K. (eds) Plankton Stratigraphy. Cambridge University Press, Cambridge, 87-154. TOWNSEND, H. A. 1985. The paleomagnetism of sediments acquired from the Goban Spur on Deep Sea Drilling Project Leg 80. In: DE GRACIANSKY,P. C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 389--414. HAILWOOD, E. A. 1985. Magnetostratigraphic correlation of Palaeogene sediments in: the Hampshire and London Basins, southern U.K. Journal of the Geological Society, London, 142, 1-27.

WING, S. L. 1984. A new basis for recognizing the Paleocene/Eocene boundary in Western Interior North America. Science, 226, 439-441. ~, BOWN, T. M & OBRADOVICH, J. D. 1991. Early Eocene biotic and climatic change in interior western North America. Geology, 19, 1189-1192.

The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology MARIE-PIERRE

A U B R Y l, W I L L I A M & ASHISH

A. B E R G G R E N 2, L O W E L L

STOTT 3

SINHA 3

1 Institut des Sciences de l'Evolution, Universit~ Montpellier II, 34095 MontpeIlier Cedex 5, France 2 Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA 3 Department of Earth Sciences, University of Southern California, Los Angeles, CA 90089-0740, USA Abstract: The late Paleocene carbon isotope excursion which was first identified at ODP Site

690, is currently regarded as the best means for an exact corrrelation at a specific point in space and time between deep sea and continental stratigraphies. Yet its position relative to biostratigraphic datums in several outcrop and deep sea sections is apparently inconsistent. Based on the integration of magnetostratigraphic, biostratigraphic (calcareous nannofossils and planktonic foraminifera) and carbon isotopic data from the North (DSDP Sites 549 and 550) and South (ODP Site 690) Atlantic Ocean, we show that in order to understand the temporal relationships and sequence of late Paleocene to early Eocene biotic and isotopic events, it is necessary to construct a composite reference section. We infer from this that in addition to the well documented 'classic' carbon isotope excursion which occurs in mid-Biozone NP9, a younger isotopic excursion, yet undocumented, occurs in lower Biozone NP10. The 55 Ma age estimate in Chron C24r (0.66) used as a calibration in the construction of the latest geomagnetic polarity time scale in DSDP Hole 550 lies at an unconformity (separating stratigraphic levels low in Zone NP9 from low in Zone NP10) with the result that the position of the NP9/NP10 zonal/chronal boundary is higher/younger in Chron 24r. Because the sections recovered from DSDP Sites 550 and 690 probably do not overlap over the upper NP9-1ower NPI0 zonal interval, and because the former section is unconformable at the NP9/NP10 zonal boundary, there is currently no means of establishing satisfactorily a numerical chronology for Chron C24r, which means that the age estimates for the upper Paleocene to lowermost Eocene biostratigraphic and isotopic events await documentation of bio-, magneto-, and isotopic stratigraphy in continuous stratigraphic section(s).

The latest Paleocene-earliest Eocene now clearly appears as a critical interval in earth history. Global warming (Stott & Kennett 1990; Kennett & Stott 1991), reduction in atmospheric circulation (Rea et al. 1990) and change in oceanic circulation associated with brief production of deep water at low latitude (Miller et al. 1987; Kennett & Stott 1991; Thomas 1990a, b 1993) concomitant with major turnovers in the bathyal and abyssal benthic foraminifera (Tjalsma & Lohman 1983; Thomas 1990a, b), and the terrestrial vertebrates (see review in Hooker 1991), are events which occurred abruptly and attest to the global nature of the mechanism(s) at work. One of the events that has received considerable attention in recent years is a distinct 3 to 4%o negative excursion in the isotope composition of dissolved inorganic carbon in the oceans. First

identified at ODP Site 690 on Maud Rise in the Weddell Sea, Antarctic Ocean (Stott & Kennett 1990; Kennett & Stott 1991), this excursion was subsequently identified at other deep sea sites (Pak & Miller 1992; Stott 1992; Thomas & Shackleton 1993; Sinha & Stott 1993), in land sections deposited at bathyal (Corfield et aI. 1991; Corfield & Cartlidge 1992) and neritic (B. Schmitz et al., unpublished data) depths, and in terrestrial sections (Koch et al. 1992; Sinha & Stott 1993). Aside from its significance with regard to oceanic productivity and circulation (Stott 1992), the carbon isotope excursion is receiving considerable attention because it is seen as the best, if not the only, means of correlation between marine and terrestrial deposits. In the deep sea, it is associated with the benthic foraminiferal extinction (Pak & Miller 1992; Thomas 1993; Thomas & Shackleton 1996).

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 353-380.

353

354

M.-P. AUBRY ET AL.

In the North American terrestrial record, it is associated with the base of the Wasatchian North American Land Mammal Age (Koch et al. 1992), and in the Paris Basin, it occurs close to the base of the Sparnacian Stage (sensu Cavelier & Pomerol 1986; Sinha & Stott 1993; Stott et al. 1996). Because of its global occurrence, the carbon isotope excursion may be selected by IGCP Project 308 as the criterion to characterize the Paleocene/Eocene boundary. This boundary is currently placed by most stratigraphers at a level near or at the planktonic foraminiferal P6a/P6b zonal boundary (in Berggren & Miller 1988= the P5/P6 zonal boundary in Berggren et al. 1995) and in mid calcareous nannofossil Zone NP10 (Aubry et al. 1988), at a level which approximates the base of the London Clay Formation (sensu Ellison et al. 1996) of the London Basin, and which is younger (by c. 0.5 million years) than the level where the carbon isotope excursion occurs. While the carbon isotope excursion is sufficiently distinctive and large to guarantee that the isotopic change which it reflects is globally synchronous, thereby ensuring that the stratigraphic levels where it occurs are correlative, its position relative to biostratigraphic datums is inconsistent between sections. For instance at ODP Site 690, it occurs slightly below the mid point of Zone NP9 (Fig. 1), whereas at DSDP Sites 549 and 550 in the North Atlantic Ocean, it occurs just below the NP9/NP10 zonal boundary (Figs 2, 5). Although it is always tempting to attribute such discrepancies to diachrony of the biostratigraphic datums, unrecognized unconformities could also account for them (Aubry 1995). Because DSDP Hole 550 contains two mineralogically distinct ash layers (+19 a n d - 1 7 ) which provide radioisotopic age control (Swisher & Knox 1991) for late Paleocene to early Eocene magnetochronology (Cande & Kent 1992, 1995; Berggren etal. 1995) the > 62 mthick upper Paleocene to lower Eocene (Chron C24r) section from this site is becoming an important reference section. Its carbon isotopic record together with that from adjacent Site 549 (Stott et al. 1996) may serve as a reference for chemostratigraphic correlations and, ultimately, for chemochronology. Diachrony has no relevance with regard to deriving a chronology from a continuous section. On the other hand, a chronology derived from a section which contains unrecognized stratigraphic gaps will undoubtedly be inaccurate. It is thus critical to examine the upper Paleocene to lower Eocene stratigraphic successions in deep sea sites where the carbon isotope excursion occurs in order to evaluate their temporal continuity and establish accurately the stratigraphic position of the excursion.

Upper Paleocene-lower Eocene elements of correlation Marine correlations around the Paleocene/Eocene boundary are among the most difficult in the Paleogene. One reason is that the boundary falls within a long magnetic reversal, Chron C24r, estimated to be 2.5 million years long in the magnetochronology of Berggren et al. (1985) and 2.73 ! and 2.557 million years long, respectively, in the revised magnetochronology of Cande & Kent (1992, 1995). Magnetobiostratigraphic correlations, which are of critical importance for other chronostratigraphic boundaries, are thus of limited usefulness for the characterization of the Paleocene/Eocene boundary. Another reason for the difficulty is that the calcareous microfossils that serve as zonal markers are often absent. In many sections, the boundary is associated with a dissolution event which affects more or less severely the planktonic foraminifera. As for the calcareous nannofossils, the absence of the markers is mostly attributed to ecological exclusion, in particular for the representatives of the critical genus Tribrachiatus.

Planktonic foraminiferal stratigraphy The stratigraphic interval representing the c. 2.5 million years span of Chron C24r is correlative with (redefined) (sub)tropical planktonic foraminiferal Zones P5 and P6 (partim) (Berggren, in Berggren et al. 1995). The P4/P5 zonal boundary (= highest occurrence (HO) of Globanomalina pseudomenardii) is associated with Chron C25n.., and the P5/P6 zonal boundary ( = H O tYo)f Morozovella velascoensis/acuta) with a level in mid-Chron C24r. Zone P6 is divided into two subzones; the P6a/b zonal boundary (= lowest occurrence (LO) of M. lensiformis/M, formosa formosa) is associated with the younger part of Chron C24r. The P6b/P7 zonal boundary (= LO of M. aragonensis) is associated with the basal part of Chron C23r. With regard to the zonal bitstratigraphy of Holes 549, 550 and 690 the problems are the following. (i) At Holes 549 and 550 dissolution has affected the preservation of planktonic foraminifera over critical parts of the stratigraphic record, particularly at Hole 550, resulting in delayed entries of certain taxa (M. subbotinae). In addition Gl. pseudomenardii is either absent (e.g. DSDP Site 550) or extremely rare (e.g. DSDP Site 549) and thus not applicable to local biostratigraphy. Morozovella acuta is not present at Site 549. (ii) at Hole 690 the taxa used in (sub)tropical biostratigraphies are not present, rendering inter-regional correlation difficult. The LO of

UPPER PALEOCENE-LOWER EOCENE STRATIGRAPHYAND CARBON ISOTOPE EXCURSION G1. australiformis, considered by Stott & Kennett (1990) to be correlative with the P6a/P6b subzonal boundary of Berggren & Miller (1988) (= P5/P6 zonal boundary of this paper) is seen to be correlative with a level within mid-Zone P5 since it is correlative with the 813C excursion peak which lies within mid Zone NP9. Calcareous nannofossil stratigraphy

Calcareous nannofossil Zones NP9 to N P l l correlate with Chron C24r interval. The base of Zone NP9, defined by the LO of Discoaster multiradiatus, is associated with the younger part of Chron C25n. The NP9/NP10 and NP10/NPI1 zonal boundaries fall within Chron C24r but the upper part of Zone NP11 correlates with Subchron C24n.3n, Subchron C24n.2r and the older part of Subchron C24n.l-2n (Berggren et al. 1985, 1995; Aubry et al. 1988). Although the base of the Eocene, as defined by the base of the Harwich Formation/Oldhaven Beds (or by the base of the Ieper Clay), is located within Zone NP10 (Knox 1984; Aubry et al. 1988), the NP9/NP10 zonal boundary is often used in marine sections to approximate the Paleocene/Eocene boundary. Zone NP10 is a hybrid between a concurrent range zone and an interval zone, defined by taxa that are part of the same lineage. The LO of Tribrachiatus bramlettei defines the base of the zone while the HO of T. contortus defines its top (Martini 1971). Romein (1979) has shown how T. bramlettei evolved into T. contortus which itself evolved into T. orthostylus in the later part of biochron NPI0. Tribrachiatus orthostylus is common in many lower Eocene (uppermost Zone NP10 to top of Zone NP12) sections but it is more abundant at mid and high latitudes. Tribrachiatus contortus seems to be less restricted in occurrence than T. bramlettei. In the absence of T. bramlettei, the HO of Fasciculithus tympaniformis (or Fasciculithus spp.) is used to approximate the NP9/NP10 zonal boundary. This practice may be hazardous. In several sections which yield the NP9/NP10 zonal boundary and where T. bramlettei occurs, there is a short overlap between the range of this species and that of F. tympaniformis. This is seen, for instance, in the Possagno section, Italy (ProtoDecima et al. 1975) and in the Nahal Avdat (Israel) and Caravaca (Spain) sections (Romein 1979). However, in other sections (such as DSDP Site 527: Backman 1986; DSDP Sites 549, 550 and ODP Site 690B: this paper) there is a sharp decrease in the abundance of E tympaniformis much below the top of Zone NP9, so that it is difficult to differentiate levels where the species is in situ or reworked. Although the three species of Tribrachiatus clearly constitute a lineage, their stratigraphic relationships remain some-

355

what unclear. In the Nahal Avdat section, the species first occur in a sequential fashion with substantial overlap between successive species so that T. contortus co-occurs either with T. bramlettei (in its lower range) or with T. orthostylus (in its upper range). On the other hand, in the Possagno section, the LOs of T. contortus and T. orthostylus are at the same level so that their lower ranges overlap with the upper range of T. bramlettei. Still in other sections (DSDP Sites 549, 550: de Graciansky et al. 1985a, b; DSDP Site 577: Heath et al. 1985) the HO of T. contortus is immediately below the LO of T. orthostylus (Mtiller 1985; Monechi 1985; Backman 1986). In ODP Hole 690B, the three species first occur in a sequential fashion. In this hole, however, the cooccurrence of T. contortus and T. orthostylus at 137.80 and 137.4 mbsf may not be reliable because of strong bioturbation (Barker et al. 1988) and because of proximity to an unconformity (see below). Diachrony is often held responsible for unexpected co-occurrence of taxa. The close juxtaposition of the HO of F. tympaniformis and the LO of T. contortus seen in some sections could be interpreted as reflecting diachrony of the former event, of the latter, or of both. However, in geographically distant sections such as the Possagno (Italy) and the Nahal Avdat (Israel) sections, similar relationships are seen between the LO of T. bramlettei, the HO of F. tympaniformis and the LO of T. contortus. At North Atlantic DSDP Sites 403 and 404 (Montadert & Roberts 1979), and 553 (Roberts et al. 1984) located further north than DSDP Site 550 and that did not penetrate into Zone NP9, T. bramlettei (without F. tympaniformis) occurs in sediments assigned to Zone NP10 (Mtiller 1979; Backman 1984). Thus we do not believe that latitudinal diachrony is responsible for different stratigraphic relationships between the three species, as for instance in DSDP Hole 550 (see below). The latest Paleocene-earliest Eocene was the warmest interval of the Cenozoic (Rea et al. 1990; Stott 1992), when tropical water masses extended to the high latitudes as shown by the biogeography of calcareous micro- and nannoplankton (Boersma et al. 1987; Aubry 1990). As minimal biogeographical differentiation occurred in the calcareous nannoplankton then, diachrony of calcareous nannofossil species in uppermost Paleocene and lowermost Eocene sediments is expected to have been minimal. It should be noted here that diatoms have a wide biogeographical distribution in upper Paleocene and lower Eocene sediments (Fenner 1994), further supporting our expectation. We thus accept the following relative chronology of events: HO of F. tympaniformis, LO of

356

M.-P. AUBRY ET AL.

T. bramlettei, LO of T. contortus, HO of T. bramlettei, LO of T. orthostylus, HO of T. contortus. We recognize, however, that, because of the lack of integrated magnetobiostratigraphic studies on sections representing continuous deposition during Chron C24r (if such sections exist), numerical age estimates of these events are premature at this time (see below). O D P Site 6 9 0 (65~

"S; 1 ~

A c. 60 m-thick upper Paleocene-lower Eocene section, extending from calcareous nannofossil Zone NP9 to Subzone NP14a was recovered from Hole 690B (Fig. 1) located on the southwestern flank of Maud Rise in 2914 m water depth (Barker et al. 1988). The section contains two lithostratigraphic units. Unit III between 128.1 and 137.8 mbsf, consists almost exclusively of calcareous biogenic particles, whereas Unit IV includes various amounts of terrigenous components. Different interpretations have been given of the upper Paleocene-lower Eocene section recovered from Hole 690B but all agree that the section yields all magnetic reversals between Chrons C22n and C25n, and in particular, that Chron C24n is well developed (compare Spiess 1990; Stott & Kennett 1990; Stott et al. 1990; Thomas et al. 1990). We disagree with these interpretations and, based on biostratigraphic evidence, we demonstrate that Chron C24n is not represented. We also suggest that two significant stratigraphic gaps occur, at c. 137.8 mbsf and at c. 139 mbsf. As pointed out by Stott & Kennett (1990), the low latitude planktonic foraminiferal zonations cannot be applied to the sediments recovered from Maud Rise because most low latitude index species are absent. Among them are the late Paleoceneearly Eocene large, keeled morozovellids. However, the Paleocene-lower Eocene calcareous nannofossil assemblages recovered from Hole 690B yield the marker species used in Martini's zonal scheme (1971), so that all boundaries from Zone NP14 to Zone NP9 can be delineated, allowing identification of the magnetozones recognized in the hole on the basis of direct biostratigraphic correlations and reference to the magnetobiochronological framework of Berggren et al. (1985) and Aubry et al. (1988). We have re-examined the calcareous nannofossil assemblages in Cores 23H-4 to 15H-1 (196.10 and 128.50 mbsf). We essentially agree with the biozonal subdivision of this interval by Pospichal & Wise (1990) although we disagree on minor points that are discussed below where appropriate. However, contrary to these authors, we stress the facts that (i) preservation decreases uphole and is

poor above Core 16H-4, 41-44 cm (142.70 mbsf) leading to uncertain specific determination; (ii) reworking is common, in particular of Tribrachiatus orthostylus and Fasciculithus tympaniformis, this latter species occurring throughout the section; (iii) discoasters are exceedingly rare at some levels, particularly in the lower and upper parts of the section. When combined, these three facts lead to difficulties in positioning zonal boundaries, and in delineating LOs and HOs of taxa, particularly of Discoaster species. We agree essentially with Spiess (1990) and Thomas et al. (1990) in identifying the normal polarity intervals that occur between 131.17 and 132.10 mbsf and between 132.85 and 133.10 mbsf (see Spiess 1990, appendix B) as Chons C22n and C23n, respectively. The former interval correlates with Subzone NP14a as indicated by the cooccurrence of Discoaster sublodoensis, D. lodoensis and D. kuepperi (see Aubry 1991). The latter interval correlates with Zone NP12 as indicated by the co-occurrence of D. lodoensis and Tribrachiatus orthostylus. Zone NP11 (= Subzone CP9b of Okada & Bukry (1980; subzone defined by Bukry (1975) on the same criteria as Martini's Zone N P l l ) is represented in Hole 690B as a thin interval between the HO of Tribrachiatus contortus (at 137.41 mbsf following Pospichal & Wise 1990) and the LO of D. lodoensis (between 134.51 and 134.41 mbsf, combining our data with those of Pospichal & Wise 1990). In Hole 690B, Zone NP11 corresponds to an interval with reversed polarity which may represent Subchron C24n.lr, Subchron C24n.2r, or Chron C24r (using the terminology of Cande & Kent (1992); see Berggren et al. (1985) and Aubry et al. 1988 for correlations). Zone NP10, defined as the biostratigraphic interval between the LO of Tribrachiatus bramlettei and the HO of T. contortus (Martini 1971) is well characterized in Hole 690B. We mostly agree with Pospichal & Wise (1990) as to the extension of Zone NP10 in Hole 690B, but we disagree with those authors as to the range of the three species in the genus Tribrachiatus and their abundance. According to Pospichal & Wise (1990, table 3), T. bramlettei is rare to common between Core 17H-2, 28-30 cm and 15H-7, 30-32 cm (149.29-137.41 mbsf), questionably present in Core 15H-6, 30-32 cm (135.91 mbsf), and rare in Core 15H-3, 30-32 cm (131.41 mbsf). In contrast, we observe that T. bramlettei is exceedingly rare, and we record rare to common Rhomboaster cuspis. The lowest occurrence of T. bramlettei is extremely difficult to delineate. Pospichal and Wise (1990, table 3) locate it between Core 17H-2, 28-30 cm and 17H-3, 28-30 cm (149.29 and 150.79 mbsf). Although we have observed specimens questionably assignable to T. bramlettei

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

357

Fig. 1. Stratigraphy of the upper Paleocene-lower Eocene section recovered from Hole 690B. Note the relationships between the carbon isotope excursion, the deep sea benthic foraminifera extinction and the selected calcareous nannofossil datums. See text for further explanation. (1) Lithology from de Graciansky et al. (1988); (2) magnetostratigraphy from Spiess (1990), based on very high density sampling; (3) our preferred magnetostratigraphic interpretation; (4) calcareous nannofossil biozonal subdivisions based on data from Pospichal & Wise (1990) and personal observation (MPA); (5) planktonic foraminifera biozonal subdivisions (Stott & Kennett 1990); (6) selected calcareous nannofossil and planktonic foraminifera datums (from Pospichal & Wise (1990), Stott & Kennett (1990) and personal observation (MPA)).

358

M.-P. AUBRY ET AL.

as low as Core 17H-3, 43-46 cm (150.94 mbsf), we have recorded the lowest typical specimens in Core 17H-l, 86-90 cm (148.40 mbsf). Accordingly, we prefer to delineate the NP9/NP10 zonal boundary between Core 17H-l, 86-90 cm and Core 17H-2, 40--42 cm (148.40 and 149.4 mbsf), but recognize that this represents a negligible difference with Pospichal & Wise (1990). We note that only early morphotypes of T. bramlettei occur in Hole 690B, suggesting the lowermost part of Zone NP10. Pospichal & Wise (1990, table 3) also report that the upper range of T. bramlettei overlaps with the range of T. contortus between at least Core 16H-1, 28-30 cm and Core 15H-7, 30-32 cm (138.09 and 137.41 mbsf) and with that of T. orthostylus between Core 15H-CC and Core 15H-7, 30-32 cm (137.80 and 137.41 mbsf). We did not examine samples from Core 15H-7 and -CC to avoid mixing problems around 137.80 mbsf (Barker et al. 1988), but we did not observe a range overlap of T. bramlettei and T. contortus in Core 16H-1 and -2. We place the HO of T. bramlettei in Core 16H-2, 41--43 cm (139.70mbsf) and the LO of T. contortus in Core 16H-I, 41-43 cm (138.22 mbsf). In addition to the lack of overlap in the range of the two taxa, we note a striking change in the composition of the assemblages between these two levels, in particular a decreased abundance of Zygrhablithus kerabyi, and the absence of Discoaster multiradiatus above 139.70 mbsf. Although exceedingly rare and poorly preserved, only Morphotype B of T. contortus (see discussion below) has been found in Hole 690B (see also Pospichal & Wise 1990, pl. 6, fig. 7). The mud-bearing nannofossil oozes that we assign to Zone NP10 yield primarily a strong normal polarity (Spiess 1990, fig. 9 and appendix B). Spiess (1990) interprets the normal polarity intervals between 138 and 140 mbsf and 144 and 155mbsf as Subchrons C24n.ln, C24n.2n and C24n.3n, respectively, and he is followed in this interpretation by Stott & Kennett (1990), Stott et al. (1990) and Thomas et al. (1990). Contrary to Spiess's claim, calcareous nannofossil stratigraphy does not support this interpretation. Biochron NP11 is correlative with Chron C24n (partim) which, thus, is not represented in Hole 690B (see above). Although the magnetic record is straightforward and clean, the normal polarity intervals which occur between 137.60 and 154.50 mbsf can only be interpreted as normal overprinting within Chron C24r, without global significance. Chron C24r thus extends from at least 137.60 to 185.70 mbsf. It may also extend up to 131.41 mbsf. The Chron C24r/C25n reversal is clearly delineated between 185.25 and 185.70 mbsf and is associated with the LO of Discoaster multiradiatus (between 185.70 and 185.90mbsf). However, we

caution that the extreme scarcity of discoasters in Core 22H-1, 50-52 cm (185.70 mbsf) and below does not allow a firm delineation of the NP8/NP9 zonal boundary. We thus interpret the upper Paleocene-lower Eocene section recovered from Hole 690B to consist of a discontinuous lower Eocene interval (between 128.1 and c. 137.8mbsf; Subzone NP14a to Zone NP11) and an expanded uppermost Paleocene-lowermost Eocene interval (c. 137.8c. 185.80 mbsf; Zones NP10 and NP9). At least four major stratigraphic gaps occur in the lower Eocene interval. The youngest occurs between 131.41 and 132.91 mbsf. It is indicated by the superposition of Zone NP12 and Subzone NP14a. The hiatus is at least 1 million years. Detailed calcareous nannofossil stratigraphy would be necessary to determine precisely the position of the unconformity and to confirm that the normal polarity interval between 131.17 and 132.10 mbsf corresponds to Chron C22n (and does not result from the concatenation of Chrons C22n and C23n.ln). Thomas et al. (1990) suggest that the presence of a thin debris flow with an erosional contact at c. 133 mbsf, and Spiess (1990, fig. 10) indicates that the younger part of Chron C23n is not represented in Hole 690B. A second unconformity occurs at the N P l l / NP12 zonal contact and is indicated by the absence of normal polarity sediments representative of Chron C24n. This unconformity occurs between 134.41 and 134.51 mbsf. It likely corresponds to heavy bioturbation at c. 134.5 mbsf (Barker et al. 1988). The hiatus represents over 1 million years. The third unconformity corresponds to the lithological change that occurs at 137.8 mbsf, from exclusively calcareous oozes to clay-bearing nannofossil oozes. It is biostratigraphically characterized by the close juxtaposition of the HO of T. contortus and the LO of T. orthostylus. Whether there is a natural overlap in the range of the two species is unclear. According to Pospichal & Wise (1990, table 3) there is a 40 cm overlap in the range of the two taxa. However, strong bioturbation at the lithological boundary and above (Barker et al. 1988) may have caused mixing in Core 15H-7 and -CC. The fourth unconformity is located between 138.22 and 139.70 mbsf, and is inferred from the juxtaposition of Morphotype B of T. contortus (at 139.70mbsf) and early morphotypes of T. bramlettei (at 138.22mbsf). This is well supported also by the change in assemblages between these two levels (see above). This indicates that the lowermost part of Zone NP10 underlies the mid to upper part of the zone (see below). We interpret the lower part of the section (between c. 139 and 185.80 mbsf) as being essen-

UPPER PALEOCENE--LOWEREOCENE STRATIGRAPHYAND CARBON ISOTOPE EXCURSION tially continuous, although we have little means to verify this assumption. The calcareous nannofossil assemblages are of limited diversity, and poor preservation, scarcity (e.g. Discoaster diastypus), or absence (e.g. Cruciplacolithus eodelus) of stratigraphically significant species do not allow objective evaluation of the section. However, we note the sequential LOs of Zygrhablitus kerabyi (in Core 20H-2, 40-42 cm; 149.20 mbsf), Pontosphaera plana (in Core 17H-6, 4 3 - 4 5 c m ; 155.45 mbsf) and Rhomboaster cuspis (in Core 17H-4, 40-41 cm; 152.40 mbsf; although questionable (poorly preserved) specimens may occur as low as Core 17H-6, 87-90 cm; 165.90 mbsf). The best indication that the section is continuous across the NP9/NP10 zonal boundary is the occurrence of early morphotypes of T. bramlettei (see Pospichal & Wise 1990, pl.6, fig. 6) that are compact and suggestive of a direct evolution from R. cuspis. As reported by Pospichal & Wise (1990), the relationship between the LO of T. bramlettei and the HO of Fasciculithus tympaniformis cannot be firmly established in Hole 690B. The abundance of this latter species varies greatly from common (e.g. in Core 18H-6, 40-42 cm; 163.60 mbsf) to very rare (e.g. in Core 18H-3, 32-35cm; 160.55 mbsf; 17H-4, 41-44 cm; 152.40 mbsf) in the interval assigned to Zone NP9, and it was found at all levels examined above c. 149mbsf (NP9/NP10 zonal boundary). Using the stratigraphic framework established above, and considering that Zone NP9 apparently contains no significant stratigraphic gap(s), the stratigraphic position of the carbon isotope excursion and of the deep sea benthic foraminiferal extinction can be precisely determined. Zone NP9 is c. 36.8 m-thick in Hole 690B (c. 149185.80mbsf). The sharp benthic foraminiferal extinction occurs between 170.63 and 170.67 mbsf (Thomas & Shackleton 1996) and the minimum 813C values occur at 170.26 mbsf (Stott et al. 1990; Kennett & Stott 1991). Thus, the deep sea events occur in the lower part of Zone NP9 (42.2% at Hole 690B). We note in passing that the deep sea Paleocene/Eocene boundary events occur at a stratigraphic position that is well below the currently accepted position of the Paleocene/ Eocene boundary (Aubry et al. 1988), which implies that these events are older than 57.3 Ma (in the chronology of Berggren et al. 1985) as determined through extrapolation using the incorrectly identified Chron C24n/C24r magnetic reversal in Hole 690B (Kennett & Stott 1991).

D S D P Site 550 (48~

13~

DSDP Site 550 (Fig. 2) is located on the Porcupine Abyssal Plain 10 km southwest of the seaward edge

359

of the Goban Spur, NE Atlantic Ocean. Hole 550, drilled in 4432 m water depth, recovered a 112 mthick section of upper Paleocene-lower Eocene (calcareous nannofossil Zone NP9 to Subzone NPI4a) marly nannofossil chalk (lithological Subunit 2a, 310.34-408 mbsf) and siliceous marly nannofossil chalk and mudstone (Subunit 2b, 408426.5 mbsf) (de Graciansky et al. 1985b). The stratigraphic interpretation of the upper part of the section (above 356 mbsf) is discussed in Aubry (1995). It has no relevance to our discussion because in Hole 550, the upper Paleocene-lower Eocene (Zones NP9 to N P l l ) magnetobiostratigraphic correlations are essentially straightforward (Fig. 2). Subchron C24n.3n is well represented between 349.86 and 359.27 mbsf, and Subchron C24n.3r extends at least down to 419.51 mbsf (Townsend 1985; Snyder et al. 1985). It is uncertain whether the magnetic reversal between 419.51 and 425.5 mbsf (Townsend 1985) corresponds to the Chrons C24r/C25n boundary as indicated by Snyder et al. (1985). The lowest reported occurrence of D. multiradiatus is at 424.85 mbsf (MOiler 1985) in an (6.5 m-thick) unsampled interval of unknown polarity (Fig. 2). In addition, the sharp lithological contact at 426.5 mbsf between lithological Units 2b and 3a (de Graciansky et al. 1985b) occurs within Zone NP5 (see MOiler 1985), not at the NP5/NP9 contact. Thus, the upper c. 50 cm of the normal polarity interval that extends between 425.5 and 434.04 mbsf may or may not represent Chron C25n. In Hole 549 (Fig. 3), a rhyolitic ash layer occurs at 352.66 mbsf in the upper part of Chronozone C25n. This layer does not occur in Hole 550 (Knox 1985). Thus, we strongly suspect that the upper surface of the unconformity between Zones NP9 and NP5 in Hole 550 is younger than the Chron C24r/C25n reversal, and that Chron C25n is not represented in Hole 550. Because of strong dissolution, no biostratigraphically diagnostic planktonic foraminifera are preserved between 409 and 425 mbsf (Snyder & Waters 1985; Berggren & Aubry 1996). This interval is assigned to calcareous nannofossil Zone NP9 by MOiler (1985) who, in the absence of Tribrachiatus bramlettei, used the HO of Fasciculithus tympaniformis to approximate the NP9/NP10 zonal boundary at c. 408 mbsf. This author (1985, table 10) reported the LO of Tribrachiatus contortus low in Zone NP10 (at 405.6 mbsf), i.e. 2.4 m above the NP9/NPI0 zonal boundary, which implied that the section between 408 and 365.96 mbsf (at which level M~iller reports the HO of T. contortus) belongs to the mid and upper part of Zone NP10. We have re-examined the calcareous nannofossil assemblages in the interval between Core 34-7 13-15 cm and Core 30-1, 21-23 cm (412.63 and

360

M.-P. AUBRY E T AL.

Fig. 2. Stratigraphy of the upper Paleocene-lower Eocene section recovered from DSDP Hole 550. Note the relationships between the carbon isotope excursion, the deep sea benthic foraminifera extinction and the selected calcareous nannofossil and planktonic foraminiferal datums. See text for further explanation. (1) Lithology from de Graciansky et al. (1985); (2) magnetostratigraphy from Townsend (1985): tick marks indicate position of samples analysed; (3) magnetostratigraphic interpretation; (4) calcareous nannofossil biozonal subdivisions (from Mtiller 1985, and personal observations, MPA); (5) planktonic foraminifera biozonal subdivisions (see Berggren & Aubry 1990); (6) selected calcareous nannofossil and planktonic foraminifera datums, based on MUller (1985), Snyder & Waters (1985) and personal observations (MPA and WAB).

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHYAND CARBON ISOTOPE EXCURSION

361

Fig. 3. (a-i) Tribrachiatus contortus Morphotype A. Note the flat morphology of the specimens. This form is reminiscent of early morphotypes of Tribrachiatus orthostylus but with long bifurcations. (j, k) Tribrachiatus bramlettei. (l) Rhomboaster cuspis. Sample 550-31-5, 17-19 (381.18 mbsf).

365.72 mbsf), using all samples which served for stable isotope analysis (Stott et al. 1996, table 2). We agree with MUller's interpretation (1985) of the section below 408 mbsf but we disagree with her interpretation of the section above this level. We locate the NP9/NP10 zonal boundary at 408 mbsf between Core 34-4, 2-4 cm and Core 34-3, 125127 cm (408.02 and 407.75 mbsf) based on the LO of Tribrachiatus bramlettei at this latter level. The HO of E tympaniformis in the section immediately precedes the LO of T. bramlettei. Zone NP10 extends up to at least Core 30-1, 21-23 cm (365.72 mbsf). This is in agreement with Mtiller (1985), who placed the NP10/NP11 zonal boundary between 365.96 and 365.30 mbsf. Tribrachiatus bramlettei ranges from Core 34-3, 125-127 to Core 30-5, 66-68 (407.75-372.17 mbsf). It is scarce to common at most levels except between Core 34-3, 47-49 and Core 34-1, 146-148 (406.98-404.97 mbsf) where it is rare to exceedingly rare.

In Hole 550, Tribrachiatus contortus is represented by two morphotypes (which probably correspond to two distinct species: MPA, work in progress) with a disjunct range. Morphotype A (Fig. 3a-i) ranges between Core 31-5, 88-90 cm and Core 31-3, 39-41 (381.9-378.4 mbsf). Morphotype B (Fig. 4a-h) occurs between Core 30-5 113-115 and Core 30-1, 21-23 (372.65365.72 mbsf). Transitional forms between T. bramlettei and T. contortus Morphotype B occur in Core 30-6, 7-9 (373.09 mbsf), while transitional forms between the latter and T. orthostylus occur in Core 30-1 141-143 cm and Core 30-1 109-111 cm (366.91-366.6 mbsf). In contrast, intermediate forms between T. contortus Morphotype A and T. bramlettei have not been observed. Morphotype A of T. contortus is flatter than Morphotype B, which has strongly twisted tips. Both morphotypes have been distinguished in the course of the study of other sections in the Atlantic Ocean and the middle East (MPA, unpublished data). Previous

362

M.-P. AUBRY ET AL.

Fig. 4. (a-h) Tribrachiatus contortus Morphotype B. Note the twisted tips of the arms which lie in different planes. Sample 550-30-4, 138-140 cm (371.40 mbsf). (i-I) Tribrachiatus bramlettei (Figs 10-12: same specimen focused at different levels.) Sample 550-34-3, 125-127 cm (407.75 mbsf).

lack of distinction between these two forms results in difficulties in comparing the succession of biostratigraphic events in Hole 550 with that in other sections. We believe that the evolutionary sequence

hole, the range of T. contortus is entirely within that of T. bramlettei (Morton et al. 1983, table 1). In Hole 550, we note that the HO of Morphotype A is close to the LO of Morozovella

T. b r a m l e t t e i -

lensiformis.

T. contortus

-

T. orthostylus

described by Romein (1979) from the Nahal Avdat section over a narrow interval in the upper part of Zone NPI0, is that observed in Hole 550 between 373 and 366.6 mbsf. This evolutionary sequence involves Morphotype B of T. contortus. It is likely that Backman (1984, table 4) encountered Morphotype B in his study of the lower Eocene in Hole 553A (Roberts et al. 1984) where a short overlap between T. nunnii (vel T. bramlettei) and T. contortus occurs in Core 22-1. Similarly, the very short overlap between T. contortus and T. orthostylus shown by Backman (1986, fig. 1) in DSDP Hole 577 (Heath et aL 1985) suggests that only Morphotype B of T. contortus was encountered in the hole. Instead, it is likely that Morphotype A alone occurs in the lower Eocene in DSDP Hole 117 (Laughton et al. 1972). In this

Because of poor preservation, it is difficult to delineate precisely the benthic foraminiferal extinction in Hole 550. It occurs between 413.62 and 408.65 mbsf. The 513C minimum occurs between 407.5 and 409 mbsf. Assuming that there is stratigraphic continuity over the NP9-NP10 zonal interval in Hole 550, it implies that (i) the benthic foraminifera extinction occurs just below the NP9/NP10 zonal boundary and (ii) the carbon isotope excursion straddles it. We observe in Hole 550 that the LO of Rhomboaster cuspis (Core 34-4, 89-91 cm, at 408.90 mbsf), a species which first occurs within Zone NP9 is closely associated with the HO of Fasciculithus tympaniformis. In addition, this latter biostratigraphic event is close to the LO of Tribrachiatus bramlettei. Finally, we note that the

UPPER PALEOCENE--LOWEREOCENE STRATIGRAPHYAND CARBON ISOTOPE EXCURSION NP9/NP10 zonal boundary is associated with a sharp lithological change at 408 mbsf.

D S D P Site 549 (49~

13~

DSDP Site 549 is located on the Goban Spur, Irish Continental Margin, NE Atlantic, c. 50 km N-NE of Site 550, near the Pendragon escarpment. Hole 549, drilled in 2533 m water depth, recovered 76.5 m (353 to 277.5 mbsf) of upper Paleocenelower Eocene (calcareous nannofossil Zone NP9 to Subzone NP14a) brown and grey nannofossil chalks which, on the basis of colour and silica content, are subdivided into four lithological subunits. The boundaries occur at 322 mbsf between Subunits 3a and 3b, at 335 mbsf between Subunits 3b and 3c and at 350.5 mbsf between Subunits 3c and 3d. The boundary between Subunits 3b and 3c is the most distinct (de Graciansky et al. 1985a). The stratigraphic interpretation of the upper part of the lower Eocene interval (above 303 mbsf) is discussed in Aubry (1995). It is not relevant to the present discussion because the magnetobiostratigraphic correlations for the upper Paleocene-lower lower Eocene interval in this hole are straightforward. Subchron C24n.3n is well represented in the upper part of Zone NP11, and the normal polarity interval between ?351.53 and 353.81 mbsf, associated with the NP8/NP9 zonal boundary (Mtiller 1985), represents Chron C25n (Townsend 1985; Snyder et al. 1985). Dissolution in the NP9 zonal interval is less intense than in Hole 550 and it affects a thinner interval. Yet, the marker species among the planktonic foraminifera are absent below 323 mbsf, and Zones P6a, P5 and P4 are only approximated (Fig. 5). Mtiller (1985) reported calcareous nannofossil assemblages of high diversity in the upper Paleocene recovered from Hole 549. She delineated the NP9/NP10 zonal boundary between Core 16-3, 105-106cm and Core 16-3, 50-51 (335.55 and 335 mbsf) and reported the HO of Fasciculithus tympaniformis and the LO of Tribrachiatus contortus in the former and latter sample, respectively. We have re-examined the calcareous nannofossil assemblages in the interval from Core 17-7, 24-26 cm to Core 16-1, 22-24 cm (350.25 to 331.70mbsf), using all samples that served for stable isotope analysis (see Stott et al. 1996, table 3). Based on this, we delineate more precisely the NP9/NP10 zonal boundary between Core 16-3 100-102 cm and 16-3, 78-80 cm (335.50 and 335.30 mbsf). The former sample yields Rhomboaster cuspis, Fasciculithus tympaniformis (few) and no representatives of the genus Tribrachiatus. In contrast, the latter sample yields Tribrachiatus bramlettei, Rhomboaster cuspis, no

363

Fasciculithus tympaniformis and no T. contortus. Tribrachiatus bramlettei (few) was encountered only at this level. Tribrachiatus contortus (Morphotype B, except at 335.16mbsf where both morphotypes occur) is common from Core 16-3, 66-68 cm to Core 16-1, 22-24 cm (335.16331.70 mbsf), and according to Mtiller (1985) it has its HO in Core 15-CC (c. 331.40 mbsf). The lithological boundary between Subunits 3b and 3c, delineated on the basis of colour and composition (de Graciansky et al. 1985a) has been associated with the NP9/NP10 zonal boundary (Knox 1985). Our detailed biostratigraphic analysis does not support this association. If located at 335 mbsf, the contact between Subunits 3b and 3c falls within Zone NP10. We note, however, that this lithological contact is not sharp (de Graciansky et al. 1985a, barrel sheets, p. 172 and core photographs p. 172). There is a c. 40 cm interval with fine banding, colour changes, and a sharp colour contact is reported at c. 80 cm in Core 16-3. We suggest that this colour contact corresponds to the NP9/NP10 zonal boundary, represents an important stratigraphic gap, and may be a better choice for the lithological boundary between Subunits 3b and 3c. The colour changes between 334.90 and 335.30 mbsf likely reflect sedimentary disturbances, and we suggest that the interval between the HO of F. tympaniformis (335.50 mbsf) and the LO of T. contortus (335.16 mbsf) is stratigraphically disturbed and comprises several stratigraphic gaps. Stratigraphic discontinuity near the NP9/NP10 zonal boundary has been recognized by Knox (1985) who observed that the upper Paleocenelower Eocene bentonitic ashes which occur in the NP 10 zonal interval in Hole 550 are absent in Hole 549. The juxtaposed LOs of MorozoveUa subbotinae, M. marginodentata and Acarinina wilcoxensis at 335 mbsf (or between 335 and 336.60 mbsf; Snyder & Waters 1985), events which are spread over a 10 m interval of Zone NP10 in Hole 550, are a clear expression of a stratigraphic discontinuity. The NP9/NP10 zonal contact is unconformable, as marked by the juxtaposition of the HO of Fasciculithus tympaniformis, the LO of Discoaster diastypus and that of late morphotypes of Tribrachiatus bramlettei. In addition, we suggest that a stratigraphic gap separates samples 16-3, 78-80 cm (335.30 mbsf) and 16-3, 66-68 cm (335.16 mbsf), which both belong to Zone NP10. The lower level yields T. bramlettei alone, whereas the upper one yields T. contortus solely. In Hole 550, the ranges of the two species overlap (see above). In addition, the presence of Morphotype A of T. contortus in an assemblage largely dominated by Morphotype B of this species in Core 16-3, 66-68 cm (335.16 mbsf) is seen as further support of an unconformity immediately below this level.

364

M.-P. AUBRY E T AL.

Fig. 5. Stratigraphy of the upper Paleocene-lower Eocene section recovered from Hole 549. Note the relationships between the carbon isotope excursion, the deep sea benthic foraminifera extinction and the selected calcareous nannofossil datums. See text for further explanation. Note the two unconformities at c. 322.20 and 335.40 mbsf and the restricted occurrence of T. bramlettei. (1) Lithology from de Graciansky et al. (1985); (2) magnetostratigraphy from Townsend (1985): tick marks indicate position of samples analysed; (3) magnetostratigraphic interpretation; (4) calcareous nannofossil biozonal subdivisions (Mfiller 1985, and personal observation (MPA)); (5) planktonic foraminifera biozonal subdivisions (see Berggren & Aubry 1996); (6) selected calcareous nannofossil and planktonic foraminifera datums, based on Miiller (1985), Snyder & Waters (1985) and personal observations (MPA and WAB).

UPPER PALEOCENE-LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION Despite the definitive stratigraphic discontinuities at c. 335.40mbsf and above c. 335.20 mbsf, the extinction event in the benthic foraminifera and the minimum ~513Cvalues in Hole 549 are clearly associated with the NP9/NP10 zonal boundary. Both occur 2.60 m below the boundary.

Correlation between Holes 690B, 550 and 549 Comparison between the upper Paleocene-lower Eocene sections recovered at ODP Site 690 and DSDP Sites 549 and 550 shows that the NP9-NP10 zonal interval consists of three distinct stratigraphic intervals. The first (lowest) extends from the base of Zone NP9 to the carbon isotope excursion/benthic foraminifera extinction (CIE/ BFE), and is similar in the three holes. The second, which extends from the CIE/BFE to the LO of Tribrachiatus bramlettei, and the third, which corresponds mainly to the range of T. bramlettei, occur in all three holes but vary considerably in thickness between holes. The similarity between the lower interval in the three holes is striking. In Holes 690B and 549, where Chron C25n is represented, its thickness is 15.5 and 14.5 m, respectively. In Hole 550 where the lower part of Zone NP9 may be truncated, the interval is estimated to be 11-15 m thick. The late Paleocene sedimentation rates at the three sites are unknown; intensive dissolution probably results in reduced thickness in Hole 550 and, to a lesser extent, in Hole 549. Yet, it is likely that the intervals comprised between the base of Zone NP9 and the CIE/BFE in the three holes are essentially correlative and represent the early part of Chron C24r and of Biochron NP9. The second interval, between the CIE/BFE and the LO of T. bramlettei is 21.26 m thick in Hole 690B, but 2.6 m-thick in Hole 549 and non-existent to negligible in Hole 550 (the BFE in this hole is poorly delineated because of intense dissolution; see above). Based on the LO of Tribrachiatus bramlettei in the three holes, the levels where the NP9/NP10 zonal boundary is delineated in Holes 550 (408 mbsf) and 549 (c. 335.40 mbsf) should be essentially correlative with level 149 mbsf in Hole 690B. Yet, this is not true in Hole 549 where an unconformity occurs at 335.40 mbsf (the NP9/NP10 zonal boundary, Fig. 5). The absence of the lower part of Zone NP10 in this hole has been demonstrated previously (Knox 1985; see above). We now conclude from the proximity of the CIE/BFE to the NP9/NP10 zonal boundary in Hole 549 that the stratigraphic gap at 335.40 mbsf also includes the upper part of Zone NP9 (Fig. 6). This is well supported by isotope stratigraphy (see

365

below). The situation is similar in Hole 550. Zone NP10 is more complete in this section than in Hole 549 (Knox 1985; see also above and below), but the proximity of the CIE/BFE to the LO of T. bramlettei in this hole indicates that the upper part of Zone NP9 is missing. The unconformity in Hole 550 is at 408 mbsf and it corresponds to the contact between lithological Subunits 2a and 2b. In Hole 549, a lithological boundary is also closely associated with the unconformity at 335.40 mbsf. However, whereas Subunits 2b and 3c, respectively, in Holes 550 and 549, represent the same stratigraphic interval, Subunit 2a in Hole 550 is older than Subunit 3b in Hole 549. The third interval, which essentially corresponds to the range of T. bramlettei in the three holes, is over 35 m thick in Hole 550, c. 10 m thick in Hole 690B and no more than 34 cm in Hole 549. The difference in thickness between Holes 550 and 549 reflects the larger stratigraphic gap associated with the NP9/NP10 zonal boundary in Hole 549. We have shown above that an unconformity occurs at c. 139mbsf in Hole 690B below the contact between lithological Units III and IVa. Figure 6 attempts to correlate the three holes. It is extremely difficult to correlate the NP10-NP11 zonal intervals of Holes 550 and 549 because of the inconsistencies between biostratigraphic ranges and between them and magnetostratigraphy (compare Figs 2 and 5). The 19 m-thick interval (385.5 to 404.5 mbsf) with bentonite layers in Hole 550 is not represented in Hole 549 (Knox 1985). Thus the stratigraphic gap in Hole 549 includes at least the 22 m of sediments that lie immediately above the unconformity at 408 mbsf in Hole 550. Different early Eocene sedimentation rates at Sites 550 and 549 likely account for the difference in thickness of the upper part of Zone NP10 and Zone N P l l in the two holes. Knox (1985) reports the presence of disseminated ash particles in Core 15-5 and 16-1 similar to those associated with the bentonites in Hole 550, and scattered specimens of Stensioina beccariiformis occur in this interval (Reynolds 1992: unpublished M.Sc. thesis, Univ. Maine, Orono). This indicates reworking in Hole 549. Most of the c. 31 m-thick section in Hole 690B from a level just above the carbon isotope excursion to the unconformity at c. 139 mbsf, is not represented in Hole 549 (gap (2) in Fig. 6). It is also likely that the c. 28 m-thick section between the level of the carbon isotope excursion and the HO of T. bramlettei in Hole 690B corresponds to a stratigraphic gap in Hole 550 (gap (3) in Fig. 6). As pointed out above, the specimens of T. bramlettei that occur in Hole 690B are very compact whereas those which occur immediately above the unconformity in Hole 550 have longer and more

366

M.-P. AUBRY ET AL.

Fig. 6. Correlation between the upper Paleocene and lower Eocene sections recovered from ODP Site 690 and DSDP Sites 549 and 550. See text for further explanation. D.m.; Discoaster multiradiatus; T.b.; Tribrachiatus bramlettei; T.c.; T. contortus; T.o.; T. orthostylus; Et.; Fasciculithus tympaniformis; M.s., Morozovella subbotinae; A.w., Acarinina wilcoxensis; M.m., Morozovella marginodentata; M.I., M. lensiformis; BFE, deep sea benthic foraminifera extinction; CIE, carbon isotope excursion.

slender arms (Fig. 4i-1). Unless it can be demonstrated that these are e n v i r o n m e n t a l l y related differences between southern high and northern mid-latitude ecomorphotypes, it is reasonable to interpret t h e m as c o r r e s p o n d i n g to different

evolutionary stages, with earliest evolutionary morphotypes in Hole 690B and more evolved morphotypes in Hole 550. This is in agreement with our interpretation of stratigraphic continuity across the N P 9 - N P 1 0 zonal boundary in Hole

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

690B (but we acknowledge a certain amount of circular reasoning) and the fact that the NP9/NP10 zonal contact is unconformable in Hole 550. Thus we consider that there is no overlap of the intervals assigned to Zone NP10 in Holes 550 and 690B. Finally, a part of the section present in Hole 550 between the LO of T. bramlettei and an undetermined level in Chron C24r (but within the range of T. contortus Morphotype B) is absent in Hole 690B where it corresponds to the unconformity at c. 139 mbsf (gap (4) in Fig. 6). We have established that Holes 690B, 550 and 549 yield discontinuous, but complementary upper Paleocene-lower Eocene (Chron C24r; Zones NP9-NP11) stratigraphic records. The section between the carbon isotope excursion in Hole 690B and the LO of T. bramlettei is almost non-existent in Hole 550 and very thin in Hole 549. On the contrary, the greater part of Zone NP10 is absent in Hole 690B. A numerical chronology is necessary for a temporal interpretation of the sections, in particular to estimate the duration of the hiatuses associated with the unconformities in each. Because the Hole 550 section was considered continuous across the NP9/NP10 boundary, the age estimate of this latter was used to constrain magnetochronology between it and the Cretaceous/Paleogene boundary in the GPTS of Cande & Kent (1992, 1995). The -17 and +19 ash layers, isotopically dated in Danish sections (Swisher & Knox 1991), occur in Hole 550. Extrapolation through these allowed determination of the age of the NP9/NP10 zonal boundary (and, incorrectly, the FAD of T. contortus), which, in turn served as a tie point in the magnetochronology of Cande & Kent (1992, 1995). Following our biostratigraphic revision of Hole 550, we point out that the biostratigraphic position of the bentonites layers (North Sea tephra phases 2a and 2b, Knox & Morton 1988; see Berggren & Aubry 1996) in this hole is in the lower to mid part of Zone NP10, below (c. 3 m) the LO of T. contortus Morphotype A and much below (13 m) that of T. contortus Morphotype B (and not within the lower range of T. contortus following Mtiller 1985). As we have now established the presence of a stratigraphic gap at the NP9/NP10 zonal boundary in Hole 550, we recognize that the age of the NP9/NP 10 zonal boundary as derived from this site is erroneous. Thus we have at present no reliable chronological framework for the earliest Eocene and the Paleocene (see below). The temporal interpretation attempted here (Fig. 7) is based on a conversion depth-time primarily based on the reconstructed record of magnetozone C24r at Site 550. The methodology followed is explained below. We emphasize that the temporal framework

367

used here is highly speculative because our reconstruction is based on the unverifiable assumption that the sedimentation rates were essentially the same at Sites 550 and 690. The objective of Fig. 7 is to show the relationships between the upper Paleocene-lower Eocene records in Hole 690B, 550 and 549, and the overlap between the hiatuses in the three sections.

Implications for isotope stratigraphy Detailed isotope stratigraphies based on the variations in isotopic composition of the tests of the deep-dwelling planktonic foraminiferal species Subbotina patagonica have been obtained for Holes 690B (Fig. 1; Stott & Kennett 1990, Stott et al. 1996), 550 and 549 (Figs 2, 5; Stott et al. 1996). In Hole 690B, the interval between c. 165 and 147.5 mbsf is characterized by relatively constant 813C values, which vary between 1.5 and 1.2%o. From 147.5 to 137.8 mbsf, a decrease occurs from 1.2%~ at 147 mbsf, to 0.6%~ at 138 mbsf. Despite the appearance, the low isotopic values (0.7 to 0.6%~ between 137.8 and c. 134 mbsf cannot be part of the decrease because of the unconformity at 137.8 mbsf (Fig. 1). In Hole 550, the interval between 406 and 380 mbsf also yields relatively constant values which vary between 1.5 and 1.0%o. A decrease occurs between 380 and 370 mbsf, from 1.6 to c. 0.6%0. The isotopic record in the interval between 165 and 137.8 mbsf in Hole 690B and between 406 and 370 mbsf in Hole 550 yields similar trends and values, which suggests that the two stratigraphic intervals are correlative (Fig. 8). This, however, is contradicted by their biostratigraphic age as discussed above. Relying upon biostratigraphy, the decrease occurs in the lowermost part of Zone NP10 in Hole 690B, whereas in Hole 550 it occurs in the upper part of this zone. Also, the interval characterized by isotopic values ranging between 1.5 and 1.0%c, which underlies the decrease in both holes, is assigned to the upperpart of Zone NP9 in Hole 690B, to lower Zone NP10 in Hole 550. There are two alternative interpretations to resolve the discrepancy between the isotopic and the biostratigraphic information. If the intervals between 165 and 137.8 mbsf in Hole 690B and between 406 and 375 mbsf in Hole 550 are correlative, which would account for the similarity in the isotope records, all the calcareous nannofossil stratigraphic events in the upper NP9 and NP10 zonal interval are diachronous (Fig. 8). In particular, this implies that the NP9/NP10 zonal boundary cannot be used for correlation to extreme

368

M.-P. AUBRY ET AL.

CHRONOLOGY MAGNETIC HISTORY

TEMPORAL INTERPRETATION

BIOCHRONOLOGY 1 I2 I

ODP HOLE 690B

DATUMS

DSDP 'HOLE 549

DSDP HOLE 550

1

to 13m

o

m

l

il

=_ BFE

~

J

CIE

--

<

_

_

~

IT

-

-

-

-~-

--

....

-

Z

----

1

Fig. 7. Temporal interpretation of upper Paleocene-lower Eocene deep sea sections. The relative chronology of events is arbitrary because no section continuous through Chron C24r has yet been identified. The temporal duration of the two thin stratigraphic intervals, one which belongs to the lower part of Zone NP10, bracketed by unconformities at c. 335.22 and 335.40 mbsf, in Hole 549, the other which belongs to upper Zone NP10, bracketed by unconformities at c. 139 and 137.8 mbsf, are exaggerated for clarity.

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

~A

"7 7

O

tr)

m~ O

:=2

O

gr~ oN

0 0

~

0

N

~z

~.~

90,-~ . 0,...~

N.=

~ .m

m

369

370

M.-P. AUBRY ET AL.

high latitudes. Furthermore, it implies that there is no unconformity at 408 mbsf in DSDP Hole 550. On the other hand, if the biostratigraphic/temporal reconstructions of the NP9-NP 10 zonal interval are correct, as discussed above and shown in Figs 6 and 7, the isotopic records in both holes, although similar, are not correlative. Indeed, we suggest that they are complementary (Fig. 9). We base this on several lines of evidence: 1. If we favour a correlation based on isotopic evidence, the resulting pattern of diachrony (Fig. 8) is difficult to explain. It would indicate that Fasciculithus tympaniformis persisted longer at high latitude Site 690 than at mid latitude Site 550. In contrast, the LO of Tribrachiatus bramlettei would be delayed at Site 690 compared to Site 550, and so would be the LO of R h o m b o a s t e r cuspis. On another hand, the LO of T. contortus Morphotype B would be synchronous at mid and high latitudes. Yet, Morphotype A of T. contortus would not occur at high latitudes. 2. Many deep sea sites are discontinuous across the NP9/NP10 zonal boundary. In virtually all the deep sea sites from which the carbon isotope excursion has been reported, the lower part of Zone NP9 is unconformable with the mid or upper part of Zone NP10 (Berggren & Aubry 1996). In this respect, ODP Hole 690B represents an exceptional record. 3. The most decisive argument is provided by the isotopic record in Hole 549 (Fig. 9). The carbon isotope excursion was identified at 348 mbsf in this hole. The interval represented between the excursion and the unconformity at 335.40 mbsf belongs to Zone NP9, and is correlatable with an interval of undetermined thickness (because of unknown sedimentation rates) immediately above the carbon isotope excursion in Hole 690B. This 2.60 m thick interval of Zone NP9 in Hole 549 yields a significant, though truncated, isotopic record, characterized by relatively constant values which vary between 1.5 and 1.2%o. This isotopic record is similar to those which characterize the interval between 165 and 147.5 mbsf in Hole 690B and between 406 and 380 mbsf in Hole 550. There is no doubt that, despite their similarity, these isotopic records in Holes 550 and 549 are not correlative. Thus, in the North Atlantic, two distinct intervals with isotopic values ranging between 1.5 and 1.2%o occur, one in Zone NP9 immediately above the carbon isotope excursion, the other in lower to mid Zone NP10. We now confidently deduce from this that the interval comprised between 165 and c. 139 mbsf in Hole 690B is not correlative with the interval between 406 and 370 mbsf in Hole 550. Instead, both records are complementary (Fig. 9). We note, however, that together they do not provide a continuous record.

Isotopic values below the unconformity at 139 mbsf in Hole 690B are of c. 0.8%o. Above the unconformity at 408 mbsf in Hole 550, they are of 1.3 to 1.1%o. This implies the occurrence of an unidentified shift or excursion between the two records, in the lower part of Zone NP10 (see also Stott et al. 1996). This also implies that there is no stratigraphic, and thus temporal, overlap between the NPI0 zonal interval recovered from Holes 690B and 550 (Figs 6, 7, 9). We remark in passing that the isotopic values (0.6 to 0.8%0 above the unconformity at 335.20 mbsf in Hole 549 are similar to those (0.6 to 0.8%o) between 372 and 370 mbsf in Hole 550, which helps correlation between the two holes (Figs 6, 9). We also note that the isotopic values (c. 0.6%0 between 137.5 and 139 mbsf in Hole 690B compare well with those between 370-372 mbsf in Hole 550, resulting in perfect agreement with our assignment of this level to the upper part of Zone NP 10 and our delineation of an unconformity at c. 139 mbsf. The isotopic data support our biostratigraphic interpretation of the upper Paleocene-lower Eocene interval recovered in Holes 550, 549 and 690B. The challenge is now to identify this second decrease/ excursion in an appropriate section. If an excursion occurs in lower Zone NP10, an additional challenge is to correlate properly the marine and the terrestrial records based on carbon isotope stratigraphy (Koch et al. 1992; Stott et al. 1996).

Implications for geochronology Numerous events occur in Chron C24r, which can serve for stratigraphic correlation and/or climatic-oceanographic reconstructions. Whereas their order of succession is quite well known at this point, their relative and numerical chronology are poorly established. We explore below different paths used in estimating these chronologies, and explain the difficulties which hamper establishing a satisfactory chronology at this time. We have seen above that there are multiple unconformities associated with the Chron C24r interval at DSDP Sites 549, 550 and ODP Site 690, and that the three sections, when properly correlated, constitute an almost complete record of Chron C24r. Sites 550 and 690 are the most important constituents in this reconstruction, and both constitute unmatched reference sections, ODP Site 690B for its continuous NP9-NP10 zonal interval, DSDP Site 550 for its unprecedented documentation of Zone NP10, in particular of the Tribrachiatus lineage. Hole 690B provides an almost complete record of the lower part of Chron C24r. We believe that Zone NP9 is (almost) completely represented at this site and that the section is continuous across the NP9/NP10 zonal boundary. However, only the lowermost part of

371

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

=~.,~

~o~ tt3

9

O

,.--,

O

~" ~

,.~

,,-k O

o ~

~ >

~

o

~..,~

o'3 o-~

~~

,...~ t,.r

o.~ 9o ~ ~J .=.25

r,.)~

372

M.-P. AUBRYET AL.

Zone NP10 is represented in Hole 690B. In contrast, Hole 550 provides an excellent record of most of Zone NP10, and we believe that the section is continuous across the NP 10/NP11 zonal boundary. Yet, the lowermost part of Zone NP10 is not represented in this section, and only the lower part of Zone NP9 is present. Hole 549 plays a smaller role in the construction of a composite section of Chron C24r, but it provides a more satisfactory record of the Chron C25n/C24r boundary than the other two holes. Also, it should be borne in mind that our correlation of the 813C isotopic records on Subbotina patagonica in Holes 550 and 690B hinges on the integrated biostratigraphic and isotopic records in Hole 549. Based on both isotope and calcareous nannofossil stratigraphies, we do not believe that there is an overlap of Holes 550 and 690B in the lower part of Zone NP10. We have no means of assessing the importance of the gap until additional isotopic studies become available or until rates of evolution in Tribrachiatus bramletti become established. At this time, for the sake of simplicity, we assume that the gap is extremely small and less than 0.1 million years. Relative c h r o n o l o g y o f events in C h r o n C 2 4 r

To establish a relative chronology within Chron C24r, we now need to adjust the two sections 550 and 690B which serve for the composite reconstruction to a common vernier. This requires conversion of the stratigraphic information into a relative chronology (distance between events) taking into account differences in sedimentation rates. Relative c h r o n o l o g y in the N P 9 z o n a l interval

Relative chronology in this interval is primarily provided by Hole 690B. Zone NP9 is 36.90 m-thick

in this hole. Compared to other records of the correlation of the LO of Discoaster multiradiatus and Chron C25n, the latter LO may be delayed; we have indicated the rarity of discoasters below 184.70 mbsf. For the sake of simplicity, we ignore this uncertainty which appears to be very minor compared to those encountered further in this discussion. Reference to Barker et al. (1988, p. 209, fig. 29) indicates that the calcium carbonate content varies generally between 75% and 90% over the stratigraphic interval of 125-200m, suggesting that sedimentation rates remained relatively stable and uniform. However, Thomas & Shackleton (1996) propose a local increase in productivity around Site 690, immediately following the benthic foraminifera extinction (e.g. above c. 170 mbsf) in order to reconcile the increase of infaunal species indicative of higher productivity with the carbon isotopic data which show decreased surface to deep water gradients suggestive of lowered productivity. However, there is no supporting evidence of increased productivity and sedimentation rates at Hole 690B. Thus assuming that the sedimentation rates were constant, the CIE occurs at 42.11% and the BFE at 41.52% of the way up in Zone NP9. In addition, the FAD of G. australiformis is at 41.07%, the decrease in abundance of Fasciculithus tympaniformis (from few to very rare) is at 63.68% and the (tentative) FAD of Rhomboaster cuspis is at 81.62% of the way up (Table 1). R e l a t i v e c h r o n o l o g y in the N P I O z o n a l interval

Relative chronology in this interval is provided by Hole 550 which constitutes an excellent continuous sedimentary record of the lower Eocene from low Zone NP10 to Zone NP13 (above and Aubry 1995).

Table 1. Relative chronology of events in early Chron C24r (essentially correlative with Biochron NP9) as deduced from the stratigraphic succession of events in ODP Hole 690B Datum FAD T. bramlettei < E tympaniformis FAD R. cuspis Carbon excursion Benthic extinction FAD G. australiformis Chron C24r/C25n FAD D. multiradiatus Chron C25n/C25r

Samples 17-1, 86-90/17-2, 40-42 18-4, 39--42/18-4, 88-92 17-6, 43-45/17-6, 87-90 21-CC, 4--6/22-1, 49-51 22-1, 50-52/22-1, 68-72 23-3, 148-150/23-4, 48-50

FAD of R. cuspis is tentative (see text).

Depth interval (mbsf) 148.40-149.40 162.10-162.50 155.45-155.90 170.31-170.65 185.25-185.70 185.25-185.70 195.69-196.19

Mean depth (mbsf)

Chronology (%)

148.90 162.30 155.68 170.26 170.48 170.64 185.47 185.90 195.94

100 63.68 30.12 42.11 41.52 41.08 0 -

UPPER PALEOCENE--LOWEREOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION However, for the upper (late) part of Chron C24r in Hole 550, comprised between 408 m (the unconformable NP9/NP10 zonal contact) and 359.65 mbsf (the Subchron C24n.3n/Chron C24r boundary) the relative chronology of events cannot be established in the straightforward depth-relative age conversion that we have applied for the lower (early) part of Chron C24r. The NP10 zonal interval in Hole 550 includes a bentonite-rich interval between 384.85 and 404.60 mbsf. Sedimentation rates in this interval are expected to be higher than in the overlying calcareous nannofossil oozes. Also, the interval between 408 and 400 m has been strongly affected by dissolution with consequently lower (apparent) sedimentation rates. To explore the variations in sedimentation rates in Hole 550, we benefit from at least 5 tie points. The -17 and +19 ashes constitute two dated levels with respective ages of 54.5 and 54.0 Ma (Swisher & Knox 1991). The 408 mbsf level yields an estimated age of 55 Ma, derived through extrapolation of the sedimentation rates between the two ashes (Swisher & Knox 1991). The two other tie points are the magnetic reversals at 359.65 and c. 422 mbsf which bound Chron C24r, with respective ages of 53.347 and 55.904 Ma in the GPTS of Cande & Kent (1995). It should be borne in mind that the age of these reversals is dependent upon the age of 55 Ma for level 408 mbsf, so that if all correlations have been established correctly, internal consistency between the GPTS and the stratigraphic record in Hole 550 for Chron C24r should be maintained. Additional tie points are found at younger levels and correspond to the magnetic reversal boundaries of Chron C24n (Table 2). An average sedimentation rate of 2.92 cm/103 years is determined for the upper part of Chron C24r (between 408 mbsf with an age of 55 Ma and

373

359.65 m with an age of 53.347Ma). This compares with an average rate of sedimentation of 2.44 cm/103 years for the whole of Chron C24r (using the Chron C24r boundaries as tie points). This approach would be normally improper because of the unconformity at 408 mbsf, but it is acceptable here because this level was not seen as unconformable when it was selected to provide an age control in the GPTS of Cande & Kent (1992, 1995). A rate of 2.92 cm/103 years is, however, twice the rate of 1.4 cm/103 years calculated for the 7 m interval between the -17 ash (at 400 m) and the +19 ash (at 393 m) with ages of 54.5 and 54.0 Ma, respectively (Swisher & Knox 1991). This implies that the bentonite rich interval between c. 385 and 405 mbsf in Hole 550 was deposited at much lower rates than the overlying calcareous nannofossil oozes between c. 385 and 360 m. A sedimentation rate in excess of 5 cm/103 years is obtained for the oozes between c. 385 and 360 mbsf using the +19 ash and the Subchron C24n.3n/Chron C24r boundary as tie points. This is counter-intuitive in as much as the contribution of volcanogenic detrital material to an otherwise calcareous nannofossil ooze would be expected to increase normal carbonate sedimentation rates significantly. Indeed, we calculate sedimentation rates of 1.78 cm/103 years for the interval representing Chron C24n (Subchron C24n. In to Suchron C24n.3n: c. 342-360 mbsf). Although dissolution has occurred in the bentonite rich interval, it is insufficient to explain the low sedimentation rates that we calculate for it. One of us (AS) has estimated that c. 12.5% of the 40 m-thick stratigraphic interval between the -17 ash and the Subchron C24n.3n/Chron C24r boundary, and c. 60% of the 8 m interval between the -17 ash and the NP9/NP10 zonal boundary (408 mbsf) have been dissolved. This leads to a

Table 2. Position of magnetic reversals in the upper Paleocene-lower Eocene section (Chron C25n to Chron C24n) in DSDP Hole 550 Chron/Subchron

Time interval (million years)

Duration (million years)

Depth interval (mbsf)

C24n. in C24n. lr C24n.2n C24n.2r C24n.3n C24r C25n

52.364-52.663 52.663-52.757 52.757-52.801 52.801-52.903 52.903-53.347 53.347-55.904 55.904-56.331

0.299 0.094 0.044 0.102 0.444 2.557 0.487

342.06-344.81 344.81-349.62 344.81-349.62 344.81-349.62 349.62-359.65 359.65-c. 423 c. 423-?

The magnetochronology is that of Cande & Kent (1995). Note that Subchron C24n.ln to Subchron C24n.2n, Subchron C24n.2r, and Suchron C24n.3n in the terminology of Cande & Kent (1992, 1995) correspond to late Subchron 24n (=Subchron C24AN), Subchron C24Ar, and Suchron C24Bn, respectively, in the terminology of Berggren et al. (1985).

374

M.-P. AUBRY ET AL.

restored thickness of 57.80 m for the upper part (408-359.68 mbsf) of Chron C24r in Hole 550. Using these figures, the corrected sedimentation rate for the (now 7.84 m thick) interval between the two ashes is 1.56 cm/103 years (versus 1.4 cm/ 103years). This represents a minor correction which has little significance regarding the compared rates of deposition for the bentonite rich interval and the overlying carbonate oozes. On the other hand, the thickness correction between the -17 ash and the NP9/NP10 zonal boundary is important because it implies a greater age than 55 Ma for the 408 m level if the sedimentation rates for the +19 ash to -17 ash interval is used. Using a corrected rate of 1.56 cm/103 years, the age is 55.32 Ma. Using the non-corrected rate of 1.4 cm/103 years, the age is 55.41 Ma. Yet, we cannot use a revised age for level 408 mbsf because, as it was believed to correspond to the NP9/NPI0 chronozonal boundary, its age of 55 Ma was used as a tie point by Cande & Kent (1992, 1995) for deriving an early Paleogene magnetochronology. The age of the Subchron C24n.3n/Chron C24r boundary (and of all reversal boundaries between Chron C29 and Chron C21) is (are) fully dependent upon the 55 million years age of level 408 mbsf in Hole 550. Thus we cannot use new revised data (age and thickness) to estimate the average sedimentation rates for the upper part of Chron C24r. We recognize (i) that the section may have been more expanded than represented now, and (ii) that the NP9/NP10 chronozonal boundary may be 0.34).4 million years older than previously estimated. Yet, we cannot apply these new estimates because the age of 55 Ma for level 408 mbsf constitutes a fixed point in the GPTS of Cande & Kent (1992, 1995). This is a deadlock situation, which, as we will show below, is of greater gravity for establishing a numerical chronology for Chron C24r. Unable to resolve the problem until the position of the NP9/NP10 chronozonal boundary in Chron C24r is correctly established, we estimate the relative chronology of the events located above the bentonite bearing nannofossil oozes (above 384.85 mbsf) based on a sedimentation rate of 2.25 cm/103 years (the sedimentation rate for Subchron C24n.3n). We estimate the relative chronology of the events within the bentonite rich interval using a sedimentation rate of 4.28 cm/ 103 years (the rate determined by interpolating between the derived age of 54.46 Ma for the top of the ash-series using the 2.25 cm/103 years in the carbonate section above and 55.0 million years calibration at the 408 mbsf level; see Berggren & Aubry (1996) for additional discussion on these and related points). We emphasize that these two values are workable estimates in a 'desperate' attempt to derive a relative chronology for Chron C24r.

Comparison of the sedimentation rates calculated on the basis of magnetochronology for different segments of Chron C24r and the Subchron C24n.3n to Chron C23n interval indicates that there is a decrease in rates around or just above Subchron C24n.3n (Berggren & Aubry 1996). The rate for Subchron C24n.ln to Subchron C24n.3n is only 1.78 cm/103 years compared to the rate of 2.25 for Subchron C24n.3n that we have chosen to apply to the upper 25.20 m of Chron C24r. The rate of 4.28 cm/103 years compares poorly with the rate of 1.4 cm/103 years derived from the radioisotopic ages on the ash layers. We do not underestimate the fact that the age of 55 Ma for level 408 mbsf is tainted with uncertainty, particularly owing to the fact that intense dissolution occurred in the lower 8 m (400-408 mbsf) of the section. Studies of more complete sections in the near future are the only way we will establish a satisfactory chronology for the upper (younger) part of Chron C24r. If we accept the sedimentation rates of 2.25 and 4.28 crn/103 years for the upper part of Chron C24r, the relative chronology of the palaeontological and isotopic events are those given in Table 3.

Relative chronology for Chron C24r If the average rates of sedimentation were comparable in the NP10 zonal interval in Hole 550 and the NP9-NP10 zonal interval in Hole 690B, a direct conversion depth-relative time based on the total thickness of the NP9-NP10 zonal interval in the composite section of > 89 m (considering that there is probably no overlap of the NP10 zonal intervals in the two sections) would be possible. If that were correct, the NP9/NP10 zonal boundary would lie c. 41% up from the base of Zone NP9. The CIE would then lie at 17.46% and the BFE at 17.20% up from the base of Zone NP9. If we add 6 m to represent the uppermost part of Chron C24r (which correlates with Zone NPll), the NP9/NP10 zonal boundary lies approximately 38.75% up from the base of Chron C24r. We have little means to constrain the sedimentation rates for the NP9-1owermost NP10 zonal interval in Hole 690B. The age of 55 Ma for the NP9/NP10 zonal boundary is not applicable because level 149 mbsf in Hole 690B is not correlative with level 408 mbsf in Hole 550. The only tie points that can be used in Hole 690B are the boundaries of Chron C25n. The normal polarity interval between 185.47 and 195.94mbsf is interpreted by Spiess (1990) to represent Chron C25n. Considering that Chron C25n is 0.487 million years long in Cande & Kent (1995), the corresponding sedimentation rate is 2.14cm/ 103 years. This is a relatively high sedimentation rate for calcareous oozes (but comparable to the

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

375

Table 3. Relative chronology of events in late Chron C24r (essentially correlative with Biochron NPIO) as deduced from the stratigraphic succession of events in DSDP Hole 550 Event

Subchron C24n. l-2n/Chron C23r Subchron C24n. 1-2n/C24n.2r Subchron C24n.2r/C24n.3n Subchron C24n.3n/chron C24r LAD S. velascoensis LAD T. contortus (Morph B) FAD T. orthostylus LAD M. aequa LAD T. bramlettei FAD T. contortus (Morph B) FAD M. lensiformis LAD T. contortus (Morph A) FAD T. contortus (Morph A) FAD M. formosa gracilis LAD M. acuta FAD A. wilcoxensis FAD T. bramlettei

Core interval

27-3, 146-149/27-4, 117-120 27-6, 12-15/27-6, 102-105 28-2, 138-140/28-3, 36-38 29-3, 27-30/29-3, 103-105 30-1, 49-53/29-6, 51-54 29-CC/30-1, 20-23 30-1 62-64/30-1,109-111 30-3 49-53/30-1,49-53 30-5 22-24/30-5,66-68 30-5 113-! 15/30-6, 7-9 31-2 65-67/31-2, 105-107 31-3 6-8/31-3, 39-41 31.5. 88-90/32-1,17-19 33-1 59-61/33-2,59-61 33-2 51-53/33-2,70-72 34-1 62-65/34-2,62-65

rate we have estimated for the post bentonite carbonate interval of Zone NP10 and N P l l in late Chron C24r in Hole 550), but which is compatible with the presence of fine grained terrigenous material in the section (Barker et al. 1988). We are concerned with the possibility that the normal polarity interval below 191.49 mbsf may not represent Chron C25n. Although well preserved, the calcareous nannofossil assemblages between 188.49 and 200.10 mbsf yield no species indicative of biostratigraphic levels younger than Zone NP5 (in particular no discoasters and no species of Heliolithus; Pospichal & Wise 1990, table 4; MPA, pers. observ.). However, unable at present to demonstrate stratigraphic disturbance around Core 23H in Hole 690B (although we note the alternation of lithologies, see Barker et al. 1988, core photograph p. 265), we tentatively accept a sedimentation rate of 2.14 cm/103 years for the NP9-NP10 zonal interval in Hole 690B, aware of the possibility of an increase in the rates immediately following the benthic foraminifera extinction (Thomas & Shackleton 1996). We thus have established a relative chronology for two segments of the composite section, but based on different sedimentation rates in each. The (average) sedimentation rate of 2.14 cm/103 years for the NP9 segment is 26.17% lower than the average rate of 2.92 cm/103 years for the NP10 segment. Thus if anything the NP9/NP10 boundary lies higher in Chron C24r than the estimate of 38.75% obtained from direct conversion thicknessduration. Since we cannot assess the importance of

Depth interval (mbsf)

Mean depth (mbsf)

Chronology (%)

341.46-342.67 344.60-345.02 349.38-349.86 359.27-360.03 366.03-363.94 365.50-365.72 366.13-366.60 366.03-368.99 371.73-372.17 372.64-373.08 375.17-375.5 378.07-378.40 381.89-384.68 394.51-396.09 396.02-396.21 403.72-405.22

342.06 344.81 349.62 359.65 364.99 365.61 366.36 376.51 371.95 372.86 375.36 378.23 383.28 395.05 396.11 404.47 408

0 11.07 15.91 17.72 18.33 32.84 35.26 37.98 49.78 61.88 81.24 83.06 94.55 100

-

the stratigraphic gap between Holes 550 and 690B, and considering the uncertainties on the contruction of the composite section, particularly on the rates at which the two component sections were deposited, we arbitrarily propose that the NP9/NP10 zonal boundary lies 40% up in Chron C24r. We emphasize that our study does not demonstrate that this is the position of the NP9/NP10 zonal boundary in Chron C24r, which differs from both Berggren et al. (1985) and Aubry et al. (1988), who placed it one-third and two-thirds, respectively, of the way up in Chron C24r. If the sedimentation rates that we have used for Holes 550 and 690B are essentially correct, our interpretation of the relative chronology over Chron C24r suggest that the NP9/NP10 zonal boundary lies closer to mid-way in Chron 24r than thought earlier. Obviously future work on P a l e o c e n e - E o c e n e sections must focus on verifying/improving the relative chronology that we suggest here. N u m e r i c a l c h r o n o l o g y o f e v e n t s in Chron C24r

Based on the relative chronology that we have tentatively established for Chron C24r, it should be a rather simple exercise to establish a numerical chronology for it based on age estimates for the Subchron C24n.3n/Chron C24r and Chron C24r/C25n boundaries in Cande & Kent (1995). Yet, incorrect assumptions in the GPTS of Cande & Kent (1992) prevent us from achieving our goal.

376

M.-P. AUBRYET AL.

An age estimate of 55 Ma for the NP9/NP10 zonal boundary constitutes one of the nine points that Cande & Kent (1992) used to calibrate their composite polarity sequence for the Late Cretaceous and Cenozoic. Swisher & Knox (1991) obtained laser total fusion 4~ ages on two characteristic volcanic ashes from Denmark. The -17 ash yielded a weighted mean age of 54.51 _+0.05 Ma, the +19 ash yielded a weighted mean age of 54.0 _ 0.53 Ma (Berggren et al. 1995). These ashes were identified in DSDP Hole 550 where they occur at c. 400 and 393 mbsf, respectively, i.e. 7 and 13 m above the NP9/NP10 zonal boundary following Mtiller (1985). The NP9/NP10 zonal boundary is used by many workers to characterize the Paleocene/Eocene boundary; thus, the occurrence of two dated ashes in a section which apparently contains the Paleocene/Eocene boundary was seen as providing the opportunity of estimating first hand its age. Extrapolating a calculated sedimentation rate of 1.4 cm/103 years between the two ashes, Swisher & Knox (1991) estimated an age of 55 Ma for the Paleocene/ Eocene boundary (as denoted by the NP9/NP10 zonal boundary), an age that Cande & Kent (1992) retained for the construction of their GPTS. Following placement of the Paleocene/Eocene boundary at c. two-thirds down in Chron C24r by Berggren et al. (1985), Cande & Kent located their calibration point at 0.66 in Chron C24r (= 1221.20km from the ridge axis in the South Atlantic). We now recognize several difficulties with the procedures followed by Cande & Kent (1992). First, as a minor point that we mention briefly, the age of the two ashes is controversial. Obradovich (in Berggren et al. 1992) obtained a bulk incremental heating 4~ plateau age of 55.07 _+ 0.16 Ma on the -17 ash. We have questioned above the low sedimentation rates implied by the age of the two ashes for the bentonite rich interval in DSDP Site 550. Underestimated sedimentation rates would have substantial consequences for the estimated age of the 408 mbsf level in Hole 550 taken by Swisher & Knox (1991) to correspond to the Paleocene/Eocene boundary. Of relevance here is the rather large uncertainty (___0.53 million years) on the +19 ash. Further problems relative to the age of the ashes but that are not directly pertinent to a numerical chronology in Chron C24r are discussed in Berggren & Aubry (1996). The main difficulty with the procedure followed by Cande & Kent (1992) resides in the fact that it was not recognized until now that the NP9/NP10 zonal boundary at 408 mbsf in Hole 550 is, in fact, an unconformable contact. The age estimate of 55 million years by Swisher & Knox (1991) is indeed the age estimate of the upper surface of the un-

conformity at 408 mbsf. It dates a level within lower Zone NP10 but of uncertain stratigraphic position with regard to the NP9/NP10 chronozonal boundary. We have seen that the lower surface of the unconformity is located low in Zone NP9, approximately coincident with the carbon isotope excursion that we have placed at c. 42% up from the base of Zone NP9. As a consequence, positioning of the NP9/NP10 zonal boundary approximately half way down in Chron C24r results in the Paleocene part of Chron C24r (which essentially correlates with Biochron NP9) being considerably misrepresented in the GPTS of Cande & Kent (1992, 1995). The Paleocene part of Chron C24r is represented by a 36.47 m thick interval in Hole 690B. Using the 2.14 cm/103 years that we have determined for the stratigraphic interval representing Chron C25n in the hole (see above), the duration of the Paleocene part of Chron C24r is 1.7 million years rather than 0.904 million years in Cande & Kent (1995). We recognize that this duration may be shorter if there is a local increase in productivity immediately following the benthic foraminiferal extinction, as suggested by Thomas & Shackleton (1996), but we note that the interval of Chron C24r below the carbon isotopic excursion alone represents 0.740 million years. Unfortunately, we cannot propose a corrected estimate for the position of the NP9/NP10 biochronal boundary in Chron C24r because that data required to do so (i.e. the age of the magnetic reversals, the sedimentation rates that we are using, and the duration of the Eocene part of Chron C24r) are contigent upon the use of the 55 Ma calibration point in Cande & Kent (1995). The rather frustrating consequence of this is that we cannot properly calibrate datums which occurred in the late part of Biochron NP9. It is obvious that an age of 55 Ma for the carbon isotope excursion (which we would estimate if Hole 550 was thought to be continuous across the NP9/NP10 boundary) is totally unrealistic. Mislocation of the NP9/NPI0 chronozonal boundary within Chron C24r results in compressing the succession of events which occurred during Biochron NP9 while, on the contrary, stretching the succession of events which occurred during Biochron NP10 (and NPll). However, we cannot relocate the NP9/NP10 zonal boundary in Chron C24r in the GPTS of Cande & Kent (1992, 1995) because it carries the age calibration of 55 Ma, upon which the numerical age of the boundaries of Chron C24r are dependent. In summary, what is to be done? The GPTS of Cande & Kent (1992, 1995) will necessitate revision based on repositioning of their '55 Ma' tie point and decoupling it from the Paleocene/Eocene (and NP9/NP10 zonal) boundary. This is needed independently of any confirmation or revision of

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION

the age of the early Eocene ashes dated in northwestern Europe. We recommend that the age of the ashes themselves be used rather than an estimated age derived from them. This revision implies modification of the magnetochronology between the next two calibration points (in Chron C29r and C21n) in Cande & Kent (1995), with decreasing age differences towards the next tie-points, and greater age differences in the Paleocene than in the Eocene. The current values in the GPTS on Chron C24r allow the establishment of a numerical magnetobiochronological framework for the late part of Chron C24r (between 53.347 and 55 Ma; Table 4). However, at present it is impossible to construct a firm numerical chronology for the early part of Chron C24r (between 55 and 55.904 Ma) because more than half of this time interval is not accounted for in the current GPTS (Cande & Kent 1995). Yet, without a sound GPTS, it is impossible to measure the rates at which changes proceeded at a time as critical in earth history as around the Paleocene/Eocene boundary. Until a revision to the GPTS is available, the only (admittedly unsatisfactory) solution is to construct the numerical chronology of events in early Chron 24r into the 0.904 million years allowed for the early part of Chron C24r (Table 4). The deadlock situation at which we have arrived in using the GPTS in geological history stems largely from the fact that Cande & Kent (1995)

377

used a calibration point which was also thought to correspond to an epoch boundary (and additionally a biozonal boundary). This has created a closed system from which there is no escape. In view of the tremendous progress currently achieved in understanding the relationship between the stratigraphic record and geological time, it would seem highly suitable that calibration points in the GPTS be chosen independant from epoch boundaries.

Summary Regional stable isotope (C, O) correlations consist of pattern matchings often within a weakly constrained biostratigraphic framework. In this exercise we have demonstrated the need for close sample spacing (similar to that commonly done in upper Neogene, Pliocene-Pleistocene) in studies in the Paleogene, as well as rigourous, integrated analysis of calcareous nannofossils and planktonic foraminifera to provide a relatively precise biostratigraphic framework within which stable isotopic curves can be placed. We have further shown that multiple unconformities with variable hiatuses occur at three deep sea sites in the North and South Atlantic Ocean. Simple pattern matching (observed in different biostratigraphic zones) would have led to a false interpretation of the carbon isotope record. Once a composite standard

Table 4. Numerical chronology of events in Chron C24r Chron C24r Late Chron C24r (53.347-55 million years)

NP9/NP 10 biochronal boundary Early Chron C24r (55-55.904 million years)

Datum events

Estimated ages (Ma)

LAD Tribrachiatus contortus (morphotype B) FAD T. orthostylus LAD T. bramlettei FAD T. contortus (morphotype B) LAD T. contortus (morphotype A) FAD T. contortus (morphotype A) FAD T. bramlettei FAD Rhomboaster cuspis < abundance (F to VR) Fasciculithus tympaniformis carbon isotope excursion benthic foraminifera extinction FAD G. australiformis

53.61 53.64 53.89 53.93 54.17 54.37 55 55.16 55.33 55.52 55.52 55.53

Age estimates for events in the early part of Chron C24r are provisional and established within the constraints of the current GPTS (Cande & Kent 1992, 1995) which allows only 0.904 million years for the interval comprised between the Chron C25n/C24r boundary and the NP9/NP10 biochronal boundary. Age (55 Ma) of T. bramlettei, estimated through extrapolation of sedimentation rates between -17 and +19 ashes, corresponds to the Chron C24.(0.66) calibration point in Cande & Kent (1992, 1995). The age of this datum is one of the constraints of the GPTS and is fixed. FAD of Rhomboaster cuspis is is tentative here because of preservation. There is a worldwide decrease in abundance of Fasciculithus tympaniformis during Biochron NP9, but its timing at different latitudes remains to be established.

378

M.-P. AUBRY ET AL.

was constructed within a biostratigraphic framework it became apparent that there are two, rather than one, decreases/excursions in the carbon isotope record in Chron C24. The chronology of these two events has been rendered difficult to determine/estimate because of the recognition that the NP9/NP10 zonal boundary at DSDP Site 550, chosen as one of the numerical calibration points in the recently revised Cenozoic geochronology of Cande & Kent (1992, 1995), is actually situated at an unconformity. This results in the impossiblity to derive a sound chronological framework for the early part of Chron C24r. This study shows the importance of detailed studies focusing on specific stratigraphic-temporal intervals aiming at defining GSSPs for epoch boundaries, and the need for rigorous interpretation of the stratigraphic record in geochronology as in geological history. Because they encourage the integration of data from very different disciplines, IGCP Projects provide an unmatched opportunity to m a k e f u n d a m e n t a l progress in geological sciences based on a continued and growing interest in the science of stratigraphy.

Postscript

Tribrachiatus contortus Morph. A is now described under the name Tribrachiatus digitalis Aubry, 1995 (Israel Journal o f Earth Sciences, in press). Its holotype is from level DSDP Site 550, Core 34, Section 3, 125-127 cm, and is illustrated in Fig. 3b in this paper.

We thank our colleagues in IGCP Project 308 (Paleocene/Eocene Boundary Events in Space and Time) for their continued support and collaboration. In particular stimulating discussions of an early version of this paper with R. M. Corfield, J. Hardenbol, D. V., Kent, R. W. O'B. Knox, K. G. Miller, R. Norris, D. Pak, B. Schmitz and E. Thomas have helped clarify the complex problems associated with determining the temporal and spatial sequence of events associated with the P/E boundary. MPA kindly thanks A. Boerma and E. Thomas for sharing with her samples from Hole 690B. We thank D. V. Kent, K. G. Miller, E. Thomas and B. Schmitz for reviewing the manuscript. This is a contribution towards IGCP Project 308. This is ISEM Contribution no. 95044 and Woods Hole Contribution no. 8856.

References

AUBRY, M.-R 1990. Late Paleogene calcareous nannoplankton evolution: A tale of climatic deterioration. In: PROTHERO, D. R. & BERGGREN, W. A. (eds) Eocene-Oligocene Climatic and Biotic Evolution. Princeton University Press, 272-309. -1991. Sequence stratigraphy: Eustasy or tectonic imprint? Journal of Geophysical Research, 96(B4), 6641-6679. 1995. From chronology to stratigraphy: assessing the continuity of the stratigraphic record. In: BERGGREN, W. A., KENT, D. V., AUBRY, M.-P. & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlations: Framework for an Historical Geology. Society for Economic Paleontologists and Mineralogists, Special Publication, 54. --, BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD,K. D., OBRADOVICH,J. D. & PROTHERO, D. R. 1988. Paleogene geochronology: An integrated approach. Paleoceanography, 6(3), 707-742. BACKMAN, J. 1984. Cenozoic calcareous nannofossil biostratigraphy from the northeastern Atlantic Ocean - Deep Sea Drilling Project Leg 81. In: ROBERTS, D. G., SCHNITKER, D., ET AL. Initial Reports of the Deep Sea Drilling Project, 81, 403-428. -1986. Late Paleocene to middle Eocene calcareous nannofossil biochronology from the Shatsky Rise, Walvis Ridge and Italy. Palaeogeography, Palaeoclimatology, Palaeoecology, 57, 43-59.

BARKER, P. E, KENNETT, J.-E, ET AL. 1988. Site 690. Proceedings of the Ocean Drilling Program, Initial Reports, 113, 183-292. BERGGREN, W. A. & AUBRY, M.-E 1996. A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network. This volume. & MILLER, K. G. 1988. Paleogene tropical planktonic foraminiferal biostratigraphy and magnetochronology. Micropaleontology, 34(4), 362-380. --, KENT, D. V. & FLYNN, J. J. 1985, Paleogene geochronology and chronostratigraphy. In: SNELLIN6, N. J. (ed.) The Chronology of the Geological Record, Geological Society, London, Memoir, 10, 141-195. --, --, SWISHER, C. C. III & AUBRY, M.-R 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A., KENT, D. V., AUBRY, M.-R & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society for Economic Paleontologists and Mineralogists, Special Publication, 54. BOERSMA, A., PREMOLI-SILVA, I. • SHACKLETON, N. 1987. Atlantic Eocene planktonic foraminiferal paleohydrologic indicators and stable isotope stratigraphy. Paleoceanography, 2, 287-331. BUKRY, D. 1975. Coccolith and silicoflagellate stratigraphy, northwestern Pacific Ocean, Deep Sea Drilling Project Leg 32. In: LARSON, R. L.,

UPPER PALEOCENE--LOWER EOCENE STRATIGRAPHY AND CARBON ISOTOPE EXCURSION MOBERLY, R., ET AL. Initial Reports of the Deep Sea Drilling Project, 32, 677-701. CANDE, S. C. & KENT, D. V. 1992. A new geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 97(B 10), 13917-13951. & -1995. Revised calibration of the geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100 (B4), 6093-6095. CAVELIER, C. & POMEROL, C. 1986. Stratigraphy of the Paleogene. Bulletin de la Socidtd Gdologique de France, 8 (II, 2), 255-265. CO~IELD, R. M. & CARTLIDCE,J. E. 1992. Oceanographic and climatic implications of the Palaeocene carbon isotope maximum. Terra Nova, 4, 443-445. 9- - , PREMOLI SILVA, I. & HOUSEY, R. A. 1991. Oxygen and carbon isotope stratigraphy of the Palaeogene and Cretaceous limestones in the Bottaccione Gorge and the Contessa Highway sections, Umbrian, Italy. Terra Nova, 3, 414-422. ELLISON, R. A., JOLLEY, D. W., KING, C. & KNOX, R. W. O'B. 1994. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. DE GRACIANSKY,P. C., POAG, C. W., ET AL. 1985a. Site 549. In: Initial Reports of the Deep Sea Drilling Project, 80, 123-250. , ET AL. 1985b. Site 550. In: Initial Reports of the Deep Sea Drilling Project, 80, 251-355. FENNER, J. 1994. Diatoms of the Fur Formation, their taxonomy and biostratigraphic interpretation Results from the Harre borehole, Denmark. Aarhus Geoscience, 1, 99-163. HEATH, G. R., BURCKLE,L. H., ETAL. 1985. Initial Reports of the Deep Sea Drilling Project, 86. HOOKER, J. J. 1991. The sequence of mammals in the Thanetian and Ypresian of the London and Belgium basins. Location of the Paleocene-Eocene boundary. Newsletters on Stratigraphy, 25(2), 75-90. KENNETT, J. E & STO'rr, L. D. 1991. Abrupt deep-sea warming, paleoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Paleocene/Eocene boundary beds of northwestern Europe: An indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. - 1985. Stratigraphic significance of volcanic ash in paleocene and Eocene sediments at Sites 549 and 550. In: DE GRACIANSKY,P. C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 8 0 , 845-850. -& MORTON,A. C. 1988. The record of early Tertiary N Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary volcanism and the opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419.

379

KOCH, R L., ZACrtOS, J. C. & GINGERICH, R D. 1992. Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary. Nature, 358, 319322. LAUGHTON, A. S., BERGGREN, W. A., ET AL. 1972. Initial Reports of the Deep Sea Drilling Project, 12. MARTINI, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In: FARINACCI, A. (ed.) Proceedings of the Second Planktonic Conference, Roma 1970, 739-785. MILLER, K. G., JANECEK,T. R., KATZ,M. E. & KEIL, D. K. 1987. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary. Paleoceanography, 2(6), 741-761. MONECHI, S. 1985. Campanian to Pleistocene calcareous nannofossil stratigraphy from the northwest Pacific Ocean, Deep Sea Drilling Project Leg 86. In: HEATH, G. R., BURCKLE,L. H., Er AL. Initial Reports of the Deep Sea Drilling Project, 86, 301-336. MONTADERT, L. & ROBERTS, D. G. 1979. Sites 403 and 404. In: Initial Reports of the Deep Sea Drilling Project, 48, 165-209. MORTON, A. C., BAC~AN, J. & HARLAND, R. 1983. A reassessment of the stratigraphy of DSDP Hole 117A, Rockall Plateau: Implications for the Paleocene-Eocene boundary in N.W. Europe. Newsletters on Stratigraphy, 12(2), 104-111. MOLLER, C. 1979. Calcareous nannofossils from the North Atlantic (Leg 48). In: MONTADERT, L., ROBERTS, D. G., Er AL. Initial Reports of the Deep Sea Drilling Project, 48, 589-639. 1985. Biostratigraphic and paleoenvironmental interpretation of the Goban Spur Region based on a study of calcareous nannoplankton. In: DE GRACIANSKY,P. C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 573-599. OKADA, H. • BUKRY, D. 1980. Supplementary modification and introduction of code numbers to the low-latitude coccolith biostratigraphic zonation. Marine Micropaleontology, 5, 321-325. PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene benthic foraminiferal isotopes and assemblages: Implications for deepwater circulation. Paleoceanography, 7(4), 405-422. POSPICHAL,J. J. & WISE, S. W. 1990. Paleocene to middle Eocene calcareous nannofossils of ODP Sites 689 and 960, Maud Rise, Weddell Sea. In: BARKER, P. E, KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Project, Initial Reports, 113, 613-666. PROTODECIMA, E, ROTH, P. H. & TODESCO, I. 1975. Nannoplancton calcareo del Paleocene e dell'Eocene della Sezione di Possagno. Schweizerische Palgiontologische Abhandlungen, 97, 35-161. REA, D. K., ZACHOS, J. C., OWEN, R. M. & GINGERICH, R. D. 1990. Global change at the Paleocene/Eocene boundary: Climatic and evolutionary consequences of tectonic events. Palaeogeography, Palaeoclimatology, Palaeoecology, 79, 117-128. ROBERTS, D.G., SCHNITKER, D., ET AL. 1984. Initial Reports of the Deep Sea Drilling Project, 81.

380

M.-P. AUBRY ET AL.

ROMEIN, A. T. J. 1979. Lineages in early Paleogene calcareous nannoplankton. Utrecht Micropaleontological Bulletin, 22, 1-231. SINHA, A. & STOTT, U D. 1993. Recognition of the Paleocene/Eocene boundary carbon isotope excursion in the Paris Basin, France. (Abstract) In: Symposium on the Correlation of the Early Paleogene in Northwestern Europe, 1-2 December 1993. Geological Society, London. SNYDER, S. W. & WATERS, V. J. 1985. Cenozoic planktonic foraminiferaf biostratigraphy of the Goban Spur region, Deep Sea Drilling Project Leg 80. In: DE GRACIANSKY,P. C., POAG, C. W., ETAL. Initial Reports of the Deep Sea Drilling Project, 80, 439-472. --, MULLER, C., SIGAL, J., TOWNSEND, H. & POAG, C. W. 1985. Biostratigraphic, peloenvirnomental, and paleomagnetic synthesis of the Goban Spur region, Deep Sea Drilling Project Leg 80. In: DE GRACIANSKY, P. C., POAG, C. W., ET AL. lnitial Reports of the Deep Sea Drilling Project, 80, 1169--1186. SPIESS, V. 1990. Cenozoic magnetostratigraphy of Leg 113 drill sites, Maud Rise, Weddell Sea, Antarctica. In: BARKER,P. E, KENNETT,J. P., ETAL. Proceedings of the Ocean Drilling Project, Initial Reports, 113, 261-315. STOTT, L. D. 1992. Higher temperatures and lower oceanic pCO2: A climate enigma at the end of the Paleocene epoch. Paleoceanography, 7(4), 395-404. & KENNETT,J. P. 1990. Antarctic Paleogene planktonic foraminifer biostratigraphy: ODP Leg 113, Sites 689-690. In: BARKER, P. F., KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Initial Reports, 113, 549-569. - - , SHACKLETON,N. J. & CORFIELD,R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: Inferences from the stable isotopic composition of planktonic foraminifers, ODP Leg 113. In: BARKER, P. F., KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Initial Reports, 113, 849-863. , SINHA, A., THIRY, M., AUBRY,M.-P. & BERGGREN, W. A. 1996. Global ~13C changes across the Paleocene-Eocene boundary: criteria for terrestrialmarine correlations. This volume. SWISHER, C. C. III& KNOX, R. W. O'B. 1991. The age of the Paleocene/Eocene boundary: 4~ dating -

-

of the lower part of NP10, North Sea Basin and Denmark. In: IGCP Project 308 (Paleocene/Eocene boundary events), International Annual Meeting and Field Conference, Brussels, 2-6 December 1991, Abstracts with Program. 16. THOMAS, E. 1990a. Late Cretaceous--early Eocene mass extinctions in the deep sea. In: SHARPTON,V. L. & WARD, P. D (eds) Global Catastrophes in Earth History. Geological Society of America, Special Paper, 247, 481-495. 1990b. Late Cretaceous through Neogene deep-sea benthic foraminifers (Maud Rise, Weddell Sea, Antarctica). In: BARKER,P. E, KENNETT,J. P., ETAL. Proceedings of the Ocean Drilling Program, Initial Reports, 113, 571-594. 1993. Cenozoic deep sea circulation: evidence from deep sea benthic foraminifera. In: KEr,~Err, J. E & WARNKE, D. (eds) The Antarctic Paleoenvironment: A Perspective on Global Change. American Geophysical Union Antarctic Research Series, 56, 141-165. -• SHACKLETON,N. J. 1993. The Paleocene benthic foraminiferal extinction: timing, duration and association with stable isotope anomalies. (Abstract) In: Symposium on the Correlation of the Early Paleogene in Northwestern Europe, Geological Society of London (1-2 December 1993). Geological Society, London. & 1996. The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies. This volume. --, BARRERA,E., HAMILTON,N., HUBER, T., KENNETT, J. P., O'CONNELL, S. B., POSPICHAL,J. J., SPIESS, V., STOTT, L., WEI, W. & WISE, S. W. JR. !990. Upper Cretaceous-Paleogene stratigraphy of Sites 689 and 690, Maud Rise (Antarctica). In: BARKER, E E, Kennett, J. E, ET AL. Proceedings of the Ocean Drilling Program, Initial Reports, 113, 901-914. TJALSMA, R. C. & LOHMAN, G. P. 1983. PaleoceneEocene bathyal and abyssal benthic foraminifera from the Atlantic Ocean. Micropaleontology, Special Publication, 4, 1-90. TOWNSEND, H. A. 1985. The paleomagnetism of sediments acquired from the Goban Spur on Deep Sea Drilling Project 80. In: DE GRACIANSKY,P. C., POAG, C. W., ETAL Initial Reports of the Deep Sea Drilling Project, 80, 389-421.

-

-

Global ~13C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine correlations LOWELL

D. S T O T T 1, A S H I S H

MARIE-PIERRE

S I N H A 1, M E D A R D

AUBRY 3 & WILLIAM

T H I R Y 2,

A. B E R G G R E N 4

1 Department of Earth Sciences, University of Southern California, Los Angeles, CA 90089-0740, USA 2 Ecole des Mines de Paris, Centre d'Information Gdologique, 35 rue Saint-Honord 77305 Fontainbleau cedex, France 3 Institut des Sciences de l'Evolution, Universitd Montpellier II, 34095 Montpellier, Cedex 5, France 4 Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA Abstract: The early Cenozoic marine carbon isotopic record is marked by a long-term shift from high 813C values in the late Paleocene to values that are 2 to 3 lower in the early Eocene. The shift is recorded in fossil carbonates from each ocean basin and represents a large change in the distribution of 12C between the ocean and other carbon reservoirs. Superimposed upon this long-term shift are several distinct carbon isotopic negative excursions that are also recorded globally. These carbon isotopic 'events' near the Paleocene-Eocene boundary provide stratigraphic information that can facilitate intersite correlations between marine and non-marine sequences. Here we present a detailed marine carbon isotopic stratigraphy across the Paleocene-Eocene boundary that is constrained by calcareous nannofossil and planktonic foraminifera biostratigraphy and magnetostratigraphy. We show that several distinct carbon isotopic changes are recorded in uppermost Paleocene and lowermost Eocene marine biogenic carbonate sediments. At least one of these isotopic changes in the ocean's carbon isotopic composition was transmitted to terrestrial carbon reservoirs, including plant biomass via atmospheric CO 2. As a consequence of this exchange of 12C between the ocean and terrestrial carbon reservoirs, it is possible to use carbon isotope stratigraphy to correlate the uppermost Paleocene and lowermost Eocene non-fossiliferous terrestrial sediments of the Paris Basin with marine sequences.

The transition from the Paleocene to the Eocene epoch was one of the most important environmental and biological transitions in earth history. It was at this time that many archaic terrestrial mammals became extinct and modern groups began to diversify (Hooker 1991). The biological changes that occurred during the transition between the Paleocene and Eocene were accompanied by dramatic warming of high latitude sea surface temperatures and deep waters throughout the oceans (Shackleton et al. 1984; Miller et al. 1987; Kennett & Stott 1990; Rea et al. 1990; Stott et al. 1990; Zachos et al. 1993; Pak & Miller 1992; Stott 1992). These changes also accompanied a longterm decrease in the carbon isotopic composition of the oceans (Fig. 1). Superimposed on the long-term climate, biotic and ocean chemistry changes was a short-term (c. 100 000 years) extreme warming of high latitude sea surface temperatures near the end of the Paleocene epoch and a rapid 3 negative

excursion in the carbon isotopic composition of dissolved inorganic carbon in the oceans (Fig. 2). These rapid oceanic changes coincided with an abrupt mass extinction of approximately 50% of deep sea benthic foraminifera (Tjalsma & Lohmann 1983; Thomas 1989, 1990; Pak & Miller 1992; Thomas & Shackleton 1993; Kennett & Stott 1995). These geochemical and biotic changes occurred within calcareous nannofossil Zone NP9 and provide a series of stratigraphic datums that can facilitate correlations between marine and nonmarine sections. For example, the large c. 3 carbon isotope change in the isotopic composition of oceanic CO 2 that occurred in association with the benthic foraminifera extinction would have been associated with a change of similar magnitude and direction in the isotopic composition of atmospheric CO 2. Plant biomass that utilized atmospheric CO 2 for photosynthesis should also record an isotopic change of similar magnitude. However, this excursion is not unique. Using

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 381-399.

381

382

L . D . STOTT ET AL.

Plankton Magnetic Zonation Polarity Nanno Foram Chron Hist.

({~ 13 C I : ~ B -1.0

0,0

2.0

1,0

,

I

MP1t P7Ic~alm

~P11 '1c24 qPlOlP6a1 ~1PlllP6bI /

,,

3.0

I

'

i

I

[]

i

I

9O

~

Long.Term ,-,

O

oilJ o L ' ~ ~ ( ~ l t l ~ ~

1C24

qP1OI

BenthicExtincti~ O

, o-~

V-

O0 00

O,O..,,.v

9 DSDPSite 527

[] ODP Sites689/690 o

% Fig. 1. Paleocene to middle Eocene biostratigraphy, magnetostratigraphy and carbon isotope stratigraphy. The carbon isotopic record is based upon Atlantic benthic foraminifera NuttaUides (circles) and Cibicidoides (squares). Planktonic ~I3c values would be slightly more positive. The isotopic data is from deep sea sequences located in the Antarctic (Kennett & Stott 1990) and the South Atlantic (Shackleton et al. 1984). Note the marked increase in 513C from the C27n to the top of C25r. Note also the decrease in isotope values across the Paleocene-Eocene boundary.

detailed biostratigraphy to constrain the relative timing of carbon isotopic changes recorded at various oceanic sites we attempt to show that there must have been more than one isotopic excursion superimposed on the longer-term isotopic change outlined earlier. In order to apply isotope stratigraphy to marine-terrestrial correlations a continuous composite isotopic record for marine sections must first be developed. This paper represents our initial attempt to develop such a standardized composite record that can be used to

correlate between marine and terrestrial sections of northwestern Europe. In this paper we describe the systematic carbon isotopic changes across the P a l e o c e n e - E o c e n e boundary in three deep sea sites, two located within the Bay of Biscay (DSDP Sites 549 and 550) and one in the Antarctic (ODP Site 690). We use biostratigraphy and magnetostratigraphy to constrain the timing of the distinct carbon isotope changes for the Paleocene and Eocene. The carbon isotope stratigraphy from deep sea sections is used

GLOBAL PALEOCENE--EOCENE CARBON ISOTOPE CHANGES

383

1~13 CpDB

166

-2.0

-1.0

. . . .

I

0.0 . . . .

I

1.0 . . . .

I

2.0 . . . .

I

3.0 . . . .

t

4.0 . . . .

167

168 169

Depth (mbsf)

Benthic

170

Extinction

171 172 173174

1~18OpDB -2.5 166

. . . .

-2.0

-1.5

I

I

. . . .

-1.0 . . . .

t

. . . .

-0.5

0.0

I

I

. . . .

0.5 I..'..

. . . .

167

168 169

Depth 170 (mbsf)

,Benthic

Extinct ior

171 172 173 174

c .... 23

~ .... 21

~ .... 19

t .... 17

,'" 15

, 13

11

~

Fig. 2. The carbon and oxygen isotopic excursion in Antarctic ODP Site 690 that was associated with the mass extinction of deep sea benthic foraminifera at the end of the Paleocene (from Kennett & Stott 1991). Planktonic (Acarinina praepentacamerata, > 250 mm; Subbotina patagonica > 250 ram) and benthic (Nuttallides truempyi) are plotted together. The isotopic change was similar change in both planktonic and benthic values. During the excursion the surface (planktonic) to bottom-water (benthic) gradient was eliminated.

384

L.D. STOTT ET AL.

to develop a conceptual stratigraphic model for the terrestrial isotope record. This model predicts what the isotopic changes would have been in terrestrial sections across the Paleocene-Eocene boundary. We compare isotopic values from a Sparnacian sequence from the Paris Basin to this chemostratigraphic model to illustrate that the systematic carbon isotopic changes observed in marine carbonates are also preserved in terrestrial deposits. These systematic isotopic changes include a distinct carbon isotope excursion similar in magnitude and direction to that seen near the Paleocene-Eocene boundary in marine sediments. This excursion is used to correlate the PaleoceneEocene transition between these terrestrial sequences marine records.

Methods Organic carbon Organic carbon occluded within the carbonate nodules from the Paris Basin continental deposits was extracted for isotopic measurement in the following way. Cleaned nodules were crushed and placed into 250 ml precombusted Pyrex beakers. A dilute HC1 solution (10%) was added to each beaker (c. 50 ml). The beakers were then covered for 24 hours. When the samples were decarbonated, the beaker was poured over precombusted quartz fibre filters. The quartz fibre was placed into precombusted 9 mm quartz or Vycor tubes and placed on a high vacuum line and lyophilized. When dry, CuO, Cu and a small piece of Ag was added to each tube, re-evacuated and sealed. Each sample was combusted at 850~ for three hours. CO 2 was extracted cryogenically on a high vacuum line and sealed in 6 m m break seals until isotopic analysis was conducted on a VG Prism mass spectrometer. A total of 26 samples was analysed. Of these, six were duplicates. Reproducibility of the six duplicates is _+0.29 (1~). Additionally, ten organic standards (USC cellulose) were run to monitor analytical precision. Reproducibility of these ten samples is _+0.22 (lc~). Fossil c a r b o n a t e s All deep sea samples were dried, weighed and briefly soaked in sodium hexametaphosphate solution and then washed with tap water over a > 63 mm screen. The > 63 mm fraction was dried at 50~ and reweighed. Cleaned shells (c. 15-20) of size-specific planktonic foraminiferal species (e.g. Morozovella subbotinae and Subbotina patagonica) were isolated under a binocular microscope. The samples were briefly sonified in

methanol to remove adhering particles and dried at 50~ before being analysed in an automated carbonate preparation device connected to a VG Prism isotope ratio mass spectrometer. The isotopic results are presented in Tables 1 to 4.

The Paleocene-Eocene boundary problem in the Paris Basin The clay formations that comprise the stratigraphic succession across the Paleocene-Eocene boundary in the Paris Basin are generally designated the 'Sparnacian' (Dollfus 1880). The Sparnacian was once considered to be a chronostratigraphic unit (Stage). It is now more simply regarded as a facies (Laurain et al. 1983). The paralic Sparnacian deposits lie stratigraphically between marine deposits that are within mid-late calcareous nannofossil Zone NP9

Table 1. Isotopic composition of co-existing paleosol organic matter and pedogenic carbonate from the Paris Basin (Limay) (isotopic results expressed relative to the PDB standard) Depth (metres) 0.50 2.00 2.40 2.65 3.00 3.00 3.30 4.00 4.50 6.00 6.10 6.10 6.20 6.20 6.60 6.85 7.00 7.00 7.20 7.50 7.80 8.80 9.70 10.30

Sample

3229 3232 1 2 3 3* 3240 4 5 7 3156 3156* 10 10" 11 12 3160 3160* 22 14 15 16 17 18

11.10

19

11.50

20

* Duplicates

13C%e Organic

-23.2 -23.2 -23.4 -23.2 -24.6 -24.7 -24.1 -23.6 -23.6 -21.3 -22.6 -22.4 -22.0 -21.6 -21.2 -21.7 -22.1 -22.2 -26.9 -23.3 -22.1 -23.6 -23.1 -23.5 -23.0 -24.6

385

GLOBAL PALEOCENE--EOCENE CARBON ISOTOPE CHANGES Table 2. Carbon and oxygen isotopic results of planktonic foraminifera from DSDP Site 549 (isotopic results presented in per mil notation relative to the PDB standard) Core-section-cm

Depth (mbsf)

813C 8180 M. subbotinae

813C 8180 S. patagonica

16-1-22-24 16-1-71.5-73.5 16-1-102.5-104.5 16-1-136-138 16-2-42-44 16-2-72-74 16-2-102.5-104.5 16-2-137-139 16-3-13-15 16-3-23-25 16-3-34-36 ' 16-3-48-49 16-3-58-60 16-3-66~i8 16-3-78-80 16-3-100-102 16-3-117-119 16-3-144-146

331.73 332.21 332.52 332.86 333.43 333.73 334.02 334.38 334.64 334.74 334.85 334.99 335.08 335.17 335.29 335.50 335.68 335.95 336.09 336.24 336.41 336.73 336.85 336.98 337.11 337.67 337.73 337.83 338.05 338.05 338.21 338.34 338.42 338.47 338.70 338.98 339.15 339.68 339.97 340.40 340.60 340.72 340.95

3.21 2.94 2.37

-2.59 -2.51 -2.50

3.03 3.01 2.75 3.20 2.75 3.41

-2.34 -2.21 -2.31 -2.40 -2.30 -2.60

0.82 0.94 0.64 0.77 0.71 0.78 0.84 0.69

-1.36 -1.40 -1.18 -1.21 -1.21 -1.33 - 1.46 -1.12

2.98 2.89 3.01 2.95 3.12 3.82 3.52 3.91 3.52 3.65 3.78 3.76 3.83 3.39 3.67 3.42 3.28 2.82 2.99 2.35 3.03 2.59 1.69 0.22

-2.35 -2.28 -2.64 -2.05 -2.21 -2.01 -1.66 -2.12 -2.29 -2.31 -2.00 - 1.92 -2.09 -2.02 -2.28 -2.35 -2.12 -2.43 -2.30 -2.93 -2.45 -2.30 -3.29 -1.86

0.04 -0.40

-1.55 -2.23

4.31 4.71

-2.95 -2.34

3.82 3.88 3.49

-2.66 -2.00 -2.17

1.76 2.23 1.64

-1.20

16-4-08-10 16-4-23.5-25.5 16-4-40-42 16-4-72-74 16-4-84-86 16-4-97-99 16-4-110-112 16-5-16-18 16-5-22-24 16-5-32-34 16-5-54-56 16-5-54-56 16-5-70-72 16-5-83-85 16-5-91-93 16-5-103-105 16-5-119-121

16-5-147-149 16-6-14-16 16-6-67-69 16-6-96-98

16-6-139-141 16-7-009-011 16-7-21-23 16-7-44-46

(Sables de Bracheux: Aubry 1983, 1985; Woolwich & Reading Bottom Bed: Siesser et al. 1987) below, and mid Zone N P l l (Aubry 1983, 1985) above. Based on indirect magnetobiostratigraphic correlations, the Sparnacian beds were deposited during Chron 24R (Aubry 1983, 1985; Aubry et al. 1986). However, the precise stratigraphic position of the various Sparnacian f o r m a t i o n s is difficult to establish because the beds do not contain stratigraphic e l e m e n t s that allow for interregional correlations.

Carbon isotope stratigraphy DSDP Sites 549 and 550 are located on the Goban Spur in the northeast Atlantic, adjacent to the classical Paleocene and Eocene sections of the Paris and L o n d o n basins. At the present time these two deep sea sites, together with O D P Site 690, provide the most detailed chemostratigraphic and biostratigraphic records available for the Paleocene-Eocene boundary interval (Aubry et al. 1996; Berggren & Aubry 1996). D S D P Sites 550

Table 3. Carbon and oxygen isotopic results of planktonic foraminifera from DSDP Site 550 (isotopic results are presented in per mil notation relative to the PDB standard) Core-section-cm

Depth (mbsf)

813C

8180

M. subbotinae 30-1-21-23 30-1-62-64 30-1 - 109-111 30-1-141-143 30-2-28-30 30-2-63-65 30-2-10-104 30-2- 136-138 30-3-39-41 30-3-74-76 30-3-99-101 30-3-138-140 30-4-99-101 30-4-99-103 30-4-138-140 30-5-22-24 30-5-66-68 30-5-99-101 30-5-113-115 30-6-007-009 30-6-42-44 30-6-82-84 30-6-120-122 31-1-15-17 31 -1-47-49 31 - 1-91-93 31 - 1- 120-122 31-2-19-21 31-2-65~57 31-2-105-107 31-3-39-41 31-3-91-93 31-3-127-129 31-4-006-008 31-4-42--44 31-4-75-77 31-4-134-136 31-5-17-19 31-5-48-50 31-5-88-90 31-5-141-143 32-1-17-19 32-1-47--49 32-1-87-89 32-1-118-120 32-2-51-53 32-2-81-83 32-2-131-133 32-3-011-013 32-3-47-49 32-3-88-90 32-3-132-134 32-4-008-010 32-4-49-51 32-4-88-90 32-4-137-139 32-5-007-009 32-5-14-16 32-5-61-63 32-5-95-97 32-5-139-141 32-6-033-157 32-6-41--43 32-6-61-63 32-6-106-108 32-6-145-147 33-1-003-005 33-1-30-32

365.72 366.13 366.60 366.92 367.29 367.64 368,03 368.37 368.90 369.25 369.50 369.89 371.00 371.03 371.39 371.73 372.17 372.50 372.64 373.08 373.43 373.83 374.21 375.16 375.48 375.92 376.21 376.70 377.17 377.56 378.40 378.92 379,28 379.57 379.93 380.26 380.85 381.18 381.49 381.89 382.42 384.68 384.98 385.49 385.69 386.52 386.82 387.32 387.62 387.98 388.49 388.83 389.09 389.50 389.89 390.38 390.59 390.65 391.12 391.46 391.90 392.11 392.42 392.62 393.07 393.46 393.95 394.22

2.20 2.96 2.99 2.93 2.62 3.24 2.98 3.50 2.91 3.15 3.37

-1.95 -1.84 -1.76 -2.07 - 1.97 -2.27 -1.94 -2.11 -2.07

2.57 3.41 2.71

-1.80 -2.32 -1.83

3.10 2.94

-2.32 -1.96

3.05

-2.09

3.00 3.35 2.94 3.19 3.35 2.74

-2.36 -1.82 -1.64

3.68 3.13 3.36

2.86

813C

8180

S. patagonica

-1.48

0.78 0.73

-1.05 -1.18

0.56 0.65 0.73 0.84 0.77 0.58

-1.19 -1.43 -1.44 -1.18 -1.24 -1.68

-2.14

0.72

-1.52

-2.27

0.87 0.80 0.65 0.91 1.08 0,91

-1.6O -1.84 -1.59 -1.48 -1.37 -1.22

-2.25 -2.18 -1.98

1.28 1.56 1.30

-1.36 -1.59 -1.44

-1.91

1.24 1.01

-1.33 -1.28

1.22 0.99

-1.41 -1.12

1.05

-1.56

1.01

-1.41

1.30 1.43 1.25 1.29

-1.12 -1.37 -1.37 -1.24

1.53

-1.59

1.40 1.33 1.34 1.26

-1.56 -1.51 -1.63 -1.50

3.0l 2.77 3.13

-1.97 -l.88 -2.01

3.20

-2.09

3.35 3.51 3.06 3.35

- 1.93 -1.91 -2.41 -2.13

1.35

-1.91

3.55 1.14 3.13 3.40 4.04 3.64

-2.31 -1.56 -1.99 -2.00 -2.55 -2.13

Table 3. Continued Core-section-cm

Depth (mbsf)

ill3C

fi180

M. subbotinae 33-1-54-56 33-1-70-72 33-1-92-94 33-1-109-111 33-1-130-132 33-1-148-150 33-2-010-012 33-2-32-34 33-2-51-53 33-2-70-72 33-2-92-94 33-2-117-119 33-2-139-141 33-3-008-010 33-3-29-31 33-3-50-52 33-3-74-76 33-3-89-91 33-3-109-111 33-3-129-131 33-3-147-149 33-4-010--012 33-4-31-33 33-4-55-57 33-4-80-82 33-4-101-103 33-4-118-120 33-5-002~004 33-5-27-29 33-5-50-52 33-5-70-72 33-5-110-112 34-1-13-15 34-1-54-56 34-1-87-89 34-1-146-148 34-2-004-006 34-2-31-33 34-2-50-52 34-2-71-73 34-2-92-94 34-2-103-105 34-2-114-116 34-2-125-127 34-2-142-144 34-3-003-005 34-3-19-21 34-3-27-29 34-3-47--49 34-3-62-64 34-3-70-72 34-3-83-85 34-3-97-99 34-3-100-102 34-3-109-111 34-3-125-127 34-4-004-006 34-4-14-16 34-4-23-25 34-4-29-31 34-4-33-35 34-4-38-40 34-4-45-47 34-4-54-56 34-4-62-64 33-4-76-78 34-4-89-91 34-4-101-103 34-4-119-121 34-4-138-140

394.46 394.62 394.84 395.01 395.31 395.49 395.61 395.83 396.02 396.21 396.43 396.68 396.90 397.09 397.30 397.51 397.75 397.90 398.10 398.30 398.48 398.61 398.82 399.06 399.41 399.52 399.69 400.03 400.28 400.53 400.71 401.11 403.64 404.05 404.38 404.97 405.05 405.32 405.51 405.72 405.93 406.04 406.15 406.26 406.43 406.54 406.70 406.78 406.98 407.13 407.21 407.34 407.48 407.51 407.60 407.76 408.05 408.15 408.24 408.30 408.34 408.39 408.46 408.55 408.63 408.77 408.90 409.02 409.20 409.39

613C

6180

S. patagonica 1.36 1.24 1.44

-1.56 -1.50 -1.74

1.29 1.17

-1.49 -1.43

1.46

-1.88

0.95 1.04

-1.72 -1.67

-2.26 -2.35 -2.31 -2.20 -2.44 -2.58 -2.39 -2.29

1.40

-1.47

1.43

-1.79

1.38 1.31 1.21 1.20

-1.55 -1.64 -1.66 -1.84

4.05 3.84

-2.28 -2.43

1.24

-1.46

3.75

-2.04

3.88 3.84 3.90 3.98 3.85

-1.95 -2.01 -2.48 -2.12 -2.35

1.31 1.08 1.50 1.50 1.31 1.12 1.06

-1.44 -1.84 -1.55 -1.55 -1.64 -1.68 -1.54

3.81

-2.35

3.80 3.96

-2.05 -2.28

1.18 1.32

-1.25 -1.71

3.80 3.83 3.65 3.80 4.04 3.53 3.96 3.86 3.43 3.68 3.32

-2.57 -2.21 -2.06 -1.86 -2.23 -1.98 -~2.29 -2.44 -2.23 -1.94

4.04 4.10 3.39 3.36 3.21 3.30 3.23

-2.17 -2.30 -2.49 -2.45 -2.49 -2.52 -2.34

3.00 3.00 3.11 3.11 3.10 2.54 2.18

-2.16 -2.95 -2.39 -2.69 -2.54 -2.31 -2.52

0.54 0.52

-1.01 -1.94

3.38 3.51 3.37 3.93 3.87

-2.09 -2.20 -2.11 -2.48 -2.53

3.31 2.76 3.70 3.59 3.25 3.14 3.60

-2.17 -2.07 -2.10 -2.23 -2.34 -2.45 -2.35

3.47 3.94 3.88 3.82 3.95 3.57 3.77 4.01

388

L . D . STOTT ET AL.

Table 4. Carbon and oxygen isotopic results of planktonic foraminifera from ODP Site 690B (Kennett & Stott 1991; isotopic results presented in per mil notation relative to the PDB standard)

Core-section-cm

Depth (mbsf)

fi13C

~180

S. patagonica 19-1-15-19 19-1-27-30 19-1-45--49 19-1-55-59 19-1-64-68 19-1-73-77 19-1-86-88 19-1-109-112 19-1-135-139 19-I-147-150 19-2-6-9 19-2-15-19 19-2-45-49 19-2-58-62 19-2-73-77 19-2-86-89 l 9-2-94-99 19-2-109-112 19-2-115-119 19-2-125-129 19-2-137-141 19-2-145-149 19-3-4-5 19-3-15-18 19-3-26-30 19-3-45-47 19-3-55-58 19-3-64-68 19-3-72-76 19-3-86-90 19-3-94-97 19-3-109-112 19-3-115-119 19-3-127-130 19-3-136-140 19-3-148-150 19-4-5-9 19-4-15-19 19-4-26-28 19-4-45-49 19-4-52-55 19-4-65-68 19-4-74-77 19-4-85-89 19-4-93-99 19-4-109-112 19-5-5-9 19-5-15-19 19-5-26-28 19-5-44--47 19-5-56-59 19-5-64-68 19-5-73-77 19-5-86-89 19-5-94-96 19-5-109-112 20-2-110-112 23-6-36--40

167.05 167.17 167.35 167.45 167.54 167.63 167.76 167.99 168.25 168.37 168.46 168.55 168.85 168.98 169.13 169.26 169.34 169.49 169.55 169.65 169.77 169.85 169.94 170.05 170.16 170.35 170.45 170.54 170.62 170.76 170.84 170.99 171.05 171.17 171.26 171.38 171.45 171.55 171.66 171.85 171.92 172.05 172.14 172.25 172.33 172.49 172.95 173.05 173.16 173.34 173.46 173.54 173.63 173.76 173.84 173.99 176.8 199.06

-0.704 -0.852 -0.543 -0.818 -1.16 -1.191 -0.95 -1.004 -1.208 -1.465 0.117 0.173 0.299

0.360 0.037 -0.083 -0.118 -0.657 -0.757 -0.697 -0.356 -0.597 -0.451 -0.420 1.578 1.725 1.706

0.331 -0.01 0.001 0.27 -0.064 0.15 0.241 0.196 0.052 0.068 0.302 0.145 0.121

1.376 1.296 1.399 1.136 1.616 1.619 1.552 1.600 1.830 1.556 1.604 1.601 1.510

0.081 0.264 0.062 0.224 0.273 0.177 0.095 0.149 0.06

1.674 1.689 1.699 1.764 1.608 1.573 1.806 1.771 1.766

0.215 0.300

2.280 3.000

~13C

~180

M. praepentcamerata

~13C

~180

N. truempyi -0.2 -0.i12 -0.316 -0.161 -0.235 -0.308 -0.304 -0.36 0.026 -0.414 -0.315 -0.124 -0.464 -0.37 -0.052 0.582 -0.439 -0.83

0.561 0.626 0.470 0.533 0.551 0.409 0.411 0.330 0.518 0.246 0.318 0.337 0.114 0.100 -0.198 0.052 -0.470

-0.989 -1.063 -1.161 -1.316 -1.399

-0.853 -0.902 -0.999 -1.163 -1.150

1.565 1.558 1.453 0.930

-1.407

1.710

-1.657 -1.88

1.131 -0.t54

-2.159 -1.831 -1.667 -l.065 -1.11 -1.109 -0.26 -0.156

1.072 0.477 0.098 -1.052 -0.706 0.889 3.179 3.402

-0.600

3.476

0.151 0.270 0.078 -0.130

-0.581

2.995

0.262

1.230

-0.872 -0.751 -0.68 -0.329 -0.400 -0.333 -0.218 0.079 -0.642 -0.583

3.004 2.919 3.249 3.214 3.093 3.069 3.729 3.044 3.154 3.583

0.018 -0.378 0.043 0.206 0.073

1.134 0.803 1.580 1.621

0.066

1.573

-0.449 -0.574 -0.524 -0.285 -0.442 -0.581 -0.612 --0.345 -0.538

3.446 3.176 3.661 3.623 3.201 3.720 3.318 3.978 3.687

-0.100 -0.026 0.188

1.200 1.439 1.724

0.338 0.226 -0.034 0.105 0.086 0.018 -0.220

1.717 1.599 1.260 1.424 1.611 1.529 1.370

GLOBAL PALEOCENE--EOCENE CARBON ISOTOPE CHANGES and 549 are important because the low-latitude planktonic foraminiferal and calcareous nannofossil zonations can be recognized (Mtiller 1981; Snyder & Waters 1981). Furthermore, the sites are characterized by relatively high sedimentation rates across the Paleocene-Eocene boundary which affords an opportunity to sample at higher resolution to resolve more precisely correlations between isotopic and biotic changes. Site 550 also contains a series of ash layers, components of which have been found in the basal part of the London Clay Formation (now Harwich Formation: Ellison et al. 1992) of the London Basin (Knox 1984; Berggren & Aubry 1996). Knox (1984) showed that a series of over 40 ashes (correlated with the Phase 2 ashes of the North Sea area) are restricted to Zone NP10. The well characterized -17 and +19 ash layers that are recognized in the London Basin and the North Sea region also occur in Site 550. Both ash layers occur within the lower third of Zone NP10 and the lower half of the C24R, providing additional stratigraphic constraints on terrestrial-marine correlations across the Paleocene-Eocene boundary. The carbon isotopic stratigraphies for DSDP Sites 549 and 550 (Fig. 3) are combined here with the isotopic record from ODP Site 690 (Kennett & Stott 1991) following the stratigraphic interpretation for the upper Paleocene and lower Eocene by Aubry et al. (1996). These records are used to develop a composite carbon isotope record across the Paleocene-Eocene boundary and to estimate carbon isotope changes in terrestrial carbon reservoirs. We have chosen to use the planktonic foraminifera Subbotina patagonica for isotopic correlations. This species is inferred to be a deepdwelling planktonic foraminifera based upon its carbon and oxygen isotopic composition relative to other planktonic species (Stott et al. 1990; Corfield 1993). The reasons we chose this species to use in correlations are listed below.

Surface-productivity effect. The isotopic composition of surface-dwelling planktonic foraminifera varies in the ocean in response to differing productivity levels. This is because lZc is preferentially removed from surface waters during photosynthesis, leaving surface waters enriched in 13C. Consequently, the isotopic composition of surface waters in the modern ocean can vary by as much as 1.0 (Kroopnick 1985). These differences are recorded by surface-dwelling foraminifera. Such differences may have nothing to do with the ocean-wide changes that were occuring in the Paleocene-Eocene transition. The 12C that is extracted in surface waters is retumed to the water column by oxidation of

389

organic matter. In the modem ocean much of this return occurs in the upper 100-200 m. Using a deep-dwelling species such as S. patagonica that inhabited the these subsurface waters mitigates much of the surface water productivity effect. This is illustrated in Fig. 3 which shows the vertical profiles of dissolved inorganic carbon (DIC) in the modern ocean at GEOSEC stations in each of oceans. The inferred depth habitat of S. patagonica is indicated. The depth range occupied by S. patagonica in the ancient ocean is inferred to be where the isotopic compositon of the water column is the most similar between sites. This is the base of the nutricline where nutrients and carbon extracted from the surface waters by photosynthesis have largely been returned to the water column. We emphasize that this argument for use of S. patatgonica in isotopic correlations does not imply that the isotopic composition of S. patagonica will be exactly the same from site to site. It only means that we expect the pattern of isotopic variability across the Paleocene-Eocene boundary interval as recorded by S. patagonica will be more similar between sites than isotopic records of a surface-dwelling species. We also recognize that comparison of planktonic isotopic records must also take into account the possibility that upwelling influenced the isotopic composition of deep and shallow-dwelling species. We would not expect to find similar isotopic records between sites if one or more sites was influenced by upwelling since this results in relatively lower ~513C values in the DIC pool.

Biological effect. A number of modem surfacedwelling planktonic foraminifera contain photosymbionts which can affect the isotopic composition of the foraminiferal calcite (e.g. Spero et al. 1991). The deep-dwelling species do not contain photosymbionts. Depth effect. Subbotina patagonica is a ubiquitous species in the early Paleogene oceans. It has been observed in all of the oceans and occurs in marginal settings up to at least bathyal depths. Benthic foraminiferal species do not occur everywhere in the ocean. Nuttallides truempyi, for example, inhabited a wide range of depths but was restricted to the deep ocean. Even though the change in isotopic composition of deep waters is relatively small over a wide depth range a comparison of different sites may involve very different water depths and, perhaps, different water masses. The isotopic evidence from Maud Rise (Kennett & Stott 1990) suggests that even at closely adjacent sites there may be signficant depth-dependent changes in the water mass chemistry that can be attributed to different water masses.

390

L . D . STOTT ET AL.

GEOSEC 6 CPD8 -0.50 0

200

400

0.50

0.0

1.0

1.5

2.0

2.5

I-

i

N E

Inferred Depth of S. patagonica

600 depth(m) 800

1000

GEOSEC Station 1200

1400

[ ] 430 Antarctic --~--431 Antarctic --[~--448 N. Indian --~--441 S, eq. Indian ~251 S.W. eq. Pacific [~ 37 N.W. eq. Atlantic --0--26 N. Atlantic

1600

Fig. 3. Carbon isotopic composition of dissolved inorganic carbon in the modem ocean as measured at GEOSEC stations in each ocean basin. The inferred depth habitat of S. patagonica with respect to the modem 513C distribution in the water column is shown on the left. Note the differences between sites near the surface. This reflects the variable influences of productivity and thermodynamic exchange effects. Note how each of the sites, including those in the Antarctic are similar around 100 m water depth but diverge below. The divergence below reflects the 'aging effect'.

Basin-basin effect.

Using a deep-dwelling planktonic species such as S. patagonica mimimizes the water mass 'age' effect associated with benthic foraminifera. There is as much as 1.0 difference in the isotopic composition of deep waters between ocean basins today. This reflects the progressive incorporation of 12C into the dissolved CO 2 pool during its transit through the ocean basins (Kroopnick 1985). There may have have been significant changes in deep water circulation during the latest Paleocene and earliest Eocene that involved changes in sources of deep water (Kennett & Stott 1990; Pak & Miller 1992). Consequently, the isotopic compositon of deep waters at a particular site may have changed. This may have produced contrasting patterns of isotopic change between deep sea sites, depending on their proximity to the old and to the new deep water source. We believe that by using a deep-dwelling plank-

tonic foraminiferal species for carbon isotopic correlation we have minimized the large isotopic variability that might be expected between benthic foraminiferal records from sites located at different depths and in different ocean basins. We have also avoided using planktonic species for correlation that inhabited the surface waters in order to minimize the effects of variable productivity. However, as discussed below we have utilized surfacedwelling species to reconstruct atmospheric ~13C values. The reason for their use and the potential errors that this method introduces are discussed below. In Figs 4 and 5 the carbon isotope stratigraphies are shown for each of the sites. We have labelled portions of the isotope records in each site to facilitate the following discussion (Fig. 5). The carbon isotope stratigraphies for each of the sites look similar. Each contains a portion of a large negative excursion near the base of the record

GLOBAL PALEOCENE--EOCENE CARBON ISOTOPE CHANGES

.

.

.

.

i

.

.

.

.

r

9

391

9

i:i :::

'ill "~'

.~.~

"0"

:i:

':

,9: - ,

/% ,~

;:: 9 .

.~'

-~ ~9

s

",' 9

o

.

~

.~

.:. .., :::

~== (tsquj) q ldeQ

~

,

i!!

~ =

:i:

~: ~ i

:.':,

, II,

~

o

~

I!SSO t o u u ~ N snoeJ~OleO

..~

eJej!U!LUeJOd o!uo ~plUeld fJeAooet:l eJoo

~

~

'

'

"

'

"

,

'

..

9 i"

,

.

.

.

.

9

: : : : : ~ :

,

iiil/~

~

,

i!.~iiii

"

~

:it i=!: 9~..1~..

I

:1. 9 .~_.

"'~

9

~

~ .~

~ . ~

~

~ ' ~

1~

~.

:

:':':':

,

O

,

~ !:!.~:i ,

c,~

9

:::~:: -~

r~

,.

0

~

-I.:.: ....

.~

~.::::~

,~

,:,:. 9 9 ,,a.d

I

-,

i:i::,~ ,

.:.:.~

I

. . ,

, , , ,

~,

,I,

: . " 9"',"

,

,

,

,

,

,

:i.:.~

{~sqw) q IdeQ

l!$SOlOUUeN

I

Sno~J~31~

0 LdN

e~ej!u![ueJod o!uo ~,)lUeld

qgcl

I i ~

6dN Sd

0-

I

~t_~

~o^ooet:l

.~=

eJoo

-9

..~

. . ~

392

L. D . S T O T T ET AL.

eJe)!u!~eJo~ o!uo~u~d

leO!doJi

l!sso~ouue N snoeJeoeO

J~l

~ I =~

n

n

J

0 LdN

I

....., tI,D

,.: ....

I..

I

in

o u~

Q.

9 CO 13

I

9 ~t ~ ,

I

~

,

I

,

,

,

I

un

oo T--

'

'

'

9

'

I

9

I

~j~l

m 13

o

~

-

go

d

-

~

0

0

o

-

o 9

E

A ~

~o ._.

~o

-

-

"

~

.,,..-

..,-

"

"-

I!SSO~OUUeN snoeJeoleO eJej!u!uueJoB o~uol~ue~ o~;oJe~uv

0

I OLaN 1

P~IN

]

~/~sx/%~%~%/%/%/%s%s% /~/xsx/x/~/x/x/~/~sx

r

d ~o

.

.

.

.

.

.

s

s / / / / / p s z ~ s i%~1%~/~r162

.

i

,

u')

o0

...... 0,I CO

IrSSOIOUU~N s n o e J e O l ~O eJe;!u!weJ0-i o!uo]~lUeld

leo!doJj.

,lii!l......

l~....

'~"

CO

0LdN q9d

i I I

CO

6dN ~d

~

i.

.cJ 0

GLOBAL PALEOCENE-EOCENE CARBON ISOTOPE CHANGES

(label A) and an interval immediately above this where the values remain essentially constant (label B in Sites 549 and 690 and label D in Site 550). Each record also contains a decrease in carbon isotope values near the top of the analysed interval (label C in Site 690 and label E in Site 550). However, when the calcareous nannofossil biostratigraphies are included in this comparison it is clear that carbon isotope changes are not of equivalent age. The intervals labelled A to E represent different portions of the isotope stratigraphy of the Paleocene-Eocene transition. At sites 549 and 690, the distinct carbon isotope excursion (A) associated with the benthic foraminiferal excursion occurs within calcareous nannofossil Zone NP9. At Site 550, however, the excursion (A) and extinction occur within a zone of strong dissolution (Fig. 4). The calcareous nannofossil biostratigraphy indicates a hiatus at this level, marking the NP9-NPI0 zonal boundary (Aubry et al. 1996). Consequently, the interval immediately above the excursion in Holes 549 and 690B, which is characterized by essentially constant values of about 1.2%o (label B) is not the same interval above the excursion in Site 550 (label D) (Fig. 5). At Site 550, interval (D) is younger, occurring within Zone NP10. The hiatus at 408 m in Site 550 encompasses the upper portion of Zone NP9 and the lower portion of Zone NP10. At Site 690, the NP9-NP10 zonal boundary marks the base of a distinct negative shift in carbon isotope values (label C). At the base of this shift at 147.5 m the carbon isotope values of Subbotina patagonica are between 1.2 and 1.4 (Fig. 5). The values become progressively lower, reaching values of about 0.6 at 137.8 m where the calcareous nannofossil biostratigraphy indicates the presence of another hiatus (Fig. 5). At Site 550 an isotopic shift of similar magnitude and direction is observed between 380 m and 370 m (Aubry et al. 1996). Here too the values shift from between 1.2 to 1.4 to values of between 0.6 and 0.8 (Fig. 5). However, the calcareous nannofossil biostratigraphy indicates that this portion of Site 550 is within the upper Zone NP10 rather than the lower Zone NP10 (Fig. 5). At Site 549 a similar shift occurs between 336 m and 335 m. At this site a hiatus occurs at 335 m coincident with the NP9-NP10 zonal boundary. Therefore, the values above this hiatus appear to represent values typical of the upper Zone NP10 as seen at Site 550 between 380 m and 370 m (Fig. 5). A schematic representation of the carbon isotope stratigraphy across the Paleocene-Eocene boundary is shown in Fig. 6. On the basis of results presented here the Paleocene-Eocene boundary transition can be characterized by a series of distinct carbon isotope changes. There is a distinct

393

negative excursion that occurs within NP9 that is associated with the benthic foraminifera extinction. There is a second negative excursion that occurs in lower NP10. The upper portion of this second excursion is not represented in the records presented here because of a hiatus. We infer a smooth and linear trend between the most negative point in the excursion below the hiatus and the values directly above the hiatus. Each of these two excursions is superimposed upon the longer-term shift in carbon isotope values that encompassed much of the late Paleocene and early Eocene (Fig. 1). This is a different picture of the carbon isotope stratigraphy across the Paleocene-Eocene boundary from that previously presented, wherein only one major isotopic excursion had been well documented. Although several previous studies have indicated the presence of more than one isotopic excursion (Bah'era & Huber 1991; Seto et al. 1991; Lu & Keller 1993), the records used to describe them were too incomplete to derive conclusive evidence that they represent the same isotopic changes as described here. A picture of multiple isotope events or changes becomes apparent only when a relatively high resolution carbon isotope record from several sites is constrained with detailed biostratigraphy. It is important to recognize that although the duration of each of these isotopic changes differs, if a site contains hiatuses or changes in sedimentation rate across one of these isotopic changes it is possible to have similar appearing isotopic curves from stratigraphically different intervals.

The exchange of 12C between the ocean and terrestrial carbon reservoirs Ocean-atmosphere 13C02 exchange The large carbon isotope changes in the oceans across the Paleocene-Eocene boundary must have involved the exchange of 12C and 13C between the larger marine and the smaller terrestrial and atmospheric carbon reservoirs. This is because the ocean and atmosphere CO 2 reservoirs tend to maintain isotopic equilibrium on short time scales (mixing time of the ocean). The fractionation between the ocean and atmospheric CO 2 reservoirs is about -9.0. This means that if the surface ocean ZCO 2 ~13C is +2.0, the equivalent atmospheric CO 2 ~13C should be close to -7.0 (Fig. 7). Therefore, the large negative carbon isotopic changes observed in marine carbonates near the Paleocene-Eocene boundary would have been transmitted to terrestrial carbon reservoirs that were in contact with the atmosphere at that time. Koch et al. (1992) were the first to apply this reasoning to

394

L.D. STOTT ETAL.

u'~,

O~

O

9

z

I1

0,1

u"=-

~LL

AF~3b

> 0

~

~

.'=--

0.6

I'=

13 C

0.8

1

I

I

PDB

1.2 I

1.4 I

1.6 I

I:~b P6a NP10

AP6a P5

NP9

AP5 AP4

Fig. 6. Schematic composite curve of carbon isotopic changes across the Paleocene-Eocene boundary recorded by the plankonic foraminifer Subbotina patagonica. This species is plotted because it occurs in all of the sites. It is inferred to have been a deep dweller based upon its oxygen and carbon isotopic values. The portion of interval C indicated with ? is inferred, since this does not appear in any of the three records described here. This inference simply connects the two adjacent portions of the curve.

carbon isotopic studies of the terrestrial sequences of the Big Horn Basin, Wyoming. They showed how the long-term change in the isotopic composition of the ocean was also recorded by soil carbonates and mammal tooth enamel and how these changes could be used to correlate marine and terrestrial sections. Terrestrial plants, and freshwater carbonates are other potential carbon reservoirs that could also preserve the large isotopic signals across the Paleocene-Eocene boundary since their carbon is derived directly or indirectly from atmospheric CO 2. A conceptual model of carbon isotopic stratigraphy for plant-derived organic carbon across the Paleocene-Eocene boundary, similar to that developed by Koch et al. (1992) is developed here. In our model the available carbon isotope records from marine sections at DSDP Site 550, 549 and ODP Site 690B that have been correlated with calcareous nannofossil biostratigraphy and

magnetostratigraphy (Aubry et al. 1996) are used to estimate what the 513C of atmospheric CO 2 would have been across the Paleocene-Eocene boundary (calcarous nannofossil Zones NP9 to NPll). The carbon isotopic composition of the palaeoatmosphere and terrestrial plant organic carbon is estimated using the ~513C of surface-dwelling planktonic foraminiferal calcite and assuming a fractionation of lZc by the C3 photosynthetic pathway of-19.0%o -2.0%~ (Bender 1971). The magnitude of the photosynthetic fractionation differs between plants that utilize other photosynthetic pathways (i.e. C4 and CAM). However, during the Paleocene and early Eocene the floral communities did not contain elements with these other photosynthetic pathways (C4 and CAM). Therefore, the model need not contend with possible mixtures of C4 and C3 plants. Furthermore, Ceding (1992) showed that the isotopic composition of organic carbon buried in soils

GLOBAL PALEOCENE-EOCENE CARBON ISOTOPE CHANGES

Atmospheric CO2 (513C -- -7

-1

1 i

t

3 BIOMASS 813C = -

26

' A = -9

Surface MarineCarbonate ba3C = + 2

f Soilorganicmatter l

3C = -26

~"~+14.5 Soil Carbonate 813c = - n . s Fig. 7. Conceptual model of carbon isotope fractionations between marine, atmosphere and continental carbon reservoirs. Absolute values for each reservoir represent modern values characteristic of C3 ecosystems. Note that atmospheric CO 2 in equilibrium with the oceanic carbon reservoir is depleted in 13C by -9. A 4 change in the ocean as occurred at the end of the Paleocene would be transferred to the soil carbonate via the atmosphere and vegetation.

retains the isotopic composition of the original plant community from which the carbon was derived. The plant-derived organic carbon buried in the pedogenic sediments of the Paris Basin should only record isotopic variability related to changes in the isotopic composition of the atmosphere, with a predictable offset due to C3 photosynthesis. Even if the fractionation of 13C by C3 plants was, say, -21 rather than -19.0, we would still expect to see the temporal pattern of isotopic change from the mid Paleocene to early Eocene.

A terrestrial chemostratigraphic model With the considerations outlined above a conceptual model for ~13C changes in terrestrial C3 biomass in the latest Paleocene is shown in Fig. 7. This model uses the isotopic composition of the planktonic foraminifer Morozovella subbotinae from DSDP Sites 550 and 549 and Acarinina praepentacamerata from ODP Site 690 to estimate the 813C of atmospheric CO2. These species were chosen because the carbon and oxygen isotopic compositions of their shells suggest they inhabited near-surface waters. We assume that calcite precipitated in isotopic equilibrium with ECO 2 in the

395

surface ocean and that the surface ocean and the atmosphere were in approximate isotopic equilibrium. Additionally, we assume a constant isotopic fractionation between the surface ocean, the atmosphere and soil CO 2. This latter assumption is reasonable since changes in temperature and other hydrologic changes would have a relatively minor influence on the fractionation of 13C between CO 2 phases relative to the size of the isotopic shifts between the Paleocene and Eocene (Mook 1974). Perhaps the largest uncertainty in our model is the assumed constant -19 fractionation of 13C by C3 plants. The fractionation factor may have varied by one or two per mil, an effect that will produce an offset between the model and actual values. The effect of changing pCO 2 of the metabolic fractionation of 13C by plants is not well constrained. However, such changes should be small at pCO 2 levels at or above present day levels (Bender 1971). Furthermore, any small changes in fractionation would be superimposed on the longterm temporal changes associated with changes in the isotopic composition of the ocean. By analysing a large number of samples we hope to average-out shorter-term variations. A sustained change in the fractionation factor would, however, be seen as a systematic offset between the model values and the measured values. The temporal pattern of 813C change during the Paleocene-Eocene transition would still be evident. Another source of uncertainty in our model is the use of planktonic foraminiferal calcite from DSDP Sites 550 and 549 to reconstruct equilibrium isotopic values of dissolved CO 2 in the surface ocean. Because the planktonic foraminifera used here to estimate the isotopic composition of surface water are extinct we cannot be sure to what extent their calcite records isotopic equilibrium values. We recognize that many modern foraminifera display offsets between predicted equilibrium isotopic compositions and measured values. This is typically a relatively small offset in many modern species. We assume that the any offset from isotopic equilibrium among the Paleocene and Eocene species is similarly small and that the error associated with this assumption will produce a small and nearly constant offset between our model predicted values and actual values. The model predicts that the carbon isotopic composition of C3 plant organic carbon should have been approximately -24.0 in the latest Paleocene and become progressively more negative across the Paleocene-Eocene boundary, reaching values of (c. -25.0 to -26.0) in the earliest Eocene (Fig. 7). The isotopic excursions that occur within Zones NP9 and NP10 would be superimposed upon this long-term trend.

396

L . D . STOTT E T A L .

'"~

o

"F~

oS

N ~=N -~ ~ ' ~

N~o = ~ o 9

o

GLOBAL PALEOCENE-EOCENE CARBON ISOTOPE CHANGES

Preliminary carbon isotopic results from the Paris Basin

397

Conclusion

The initial results of our isotopic measurements of organic carbon from the Limay section is shown in Fig. 8. A comparison of the carbon isotope results from the Limay section matches the predicted pattern of isotopic change across the PaleoceneEocene boundary. All of the data falls within the uncertainty of the model predicted values across the Paleocene-Eocene boundary. In fact, the distinct carbon isotope excursion near the PaleoceneEocene boundary in the marine records is clearly evident in the Limay section (Fig. 8). Based upon a visual comparison between the model curve and Limay results we believe this portion of the Limay section encompasses the latest Paleocene (approximately equivalent to middle to upper NP9 Zone). We expect that our ongoing work in this basin will more fully resolve the isotopic record through the Sparnacian and allow a more definitive age determination. These results are compelling support for the use of carbon isotope stratigraphy as a unifying proxy for locating the Paleocene-Eocene boundary in terrestrial and marine sections. We believe that the position of the 813C excursion may provide the best criterion for pin-pointing the epoch boundary in non-fossiliferous sections. These results are therefore important to the goals of the IGCP 308 which is attempting to identify a standard section for the Paleocene-Eocene boundary and a criterion by which stratigraphers can correlate to it.

A detailed isotopic record from marine biogenic carbonates has been developed across the Paleocene-Eocene boundary. These marine records exhibit several distinct isotopic patterns, including the large negative excursion within Zone NP9 and P5, that can be recognized globally. The systematic isotopic patterns seen in ODP Site 690 and in DSDP Sites 550 and 549 serve as a reference for correlation to other sections that may not contain the requisite faunal or floral fossil elements needed for biostratigraphic correlation but can provide carbon isotope information. The isotopic changes recorded in the deep sea were transferred to the terrestrial carbon reservoirs via the atmosphere. A carbon isotope stratigraphy developed from terrestrial organic carbon extracted from the Limay section of the Paris Basin documents a distinct excursion and isotopic patterns that closely match changes in the marine isotopic record within Zone NP9. The recognition of these isotopic changes in the Limay section of the Paris Basin illustrates how carbon isotope stratigraphy can be used for marine-terrestrial correlations. Carbon isotope stratigraphy may provide the most appropriate criterion for locating the position of the Paleocene-Eocene epoch boundary. This work was supported by a grant from the NSF EAR9219093 to LDS. This paper represents a contribution to IGCP Projects 308 and 317. The comments and suggestions of D. Pak, K. Miller, E. Thomas, and J. Zachos are greatly appreciated. This is Woods Hole Oceanographic Contribution No. 8854 and ISEM Contribution No. 95048.

References AUBRY,M.-P. 1983. Biostratigraphie du Paldogbne dpicontinental de l'Europe du nord-ouest. Etude fond~e sur les nannofossiles calcaires. Documents des Laboratoires de Gdologie, Lyon, 89. 1985. Paleogene calcareous nannoplankton bitstratigraphy of northwestern Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 55, 267-334. --, HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous nannofossil stratigraphy of lower Paleogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735. ~, BERC~REN,W. A., STOTT,L. & SINHA,A. 1996. The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology. This volume. BARRERA,E. & HUBER,B. T. 1991. Paleogene and early Neogene oceanography of the southern Indian Ocean: Leg 119 foraminifer stable isotope results.

In: BARRON, J. A., ~T AL. Proceedings of the

Ocean Drilling Program, Scientific Results, 119, 693-717. BENDER, M. M. 1971. Variation in the 13C/12C ratios of plants in relation to the pathway of photosynthetic carbon dioxide fixation. Phytochemistry, 10, 12391244. BERGGREN, W. A. & AUBRY, M.-P. 1996. A late Paleocene-early Eocene NW European and North Sea magnetobiochronological correlation network. This volume. CERHNG, T. E. 1992. Use of Carbon isotopes in paleosols as an indicator of the pCO2 of the paleoatmosphere. Global Biogeochemical Cycles, 6, 307-314. CORFIELD,R. M. 1993. Depth habitats and the Palaeocene radiation of the planktonic foraminifera monitored using oxygen and carbon isotopes. In: LEES, D. & EDWARDS, D. (eds) Evolutionary Patterns and Processes. Linnaean Society, London, 59-70. DOLLFUS, G. F. 1880. Essai sur l'extension des terrains tertiairs dans le Bassin Anglo-Parisien. Bulletin

398

L.D. STOTT ETAL.

de la Socigt~ G~ologique de Normandie, 6(1879), 584-605. ELLISON, R. A., KNox, R. W. O'B., JOLLEY, D. W. & KINr, C. 1992. A revision of the lithostratigraphical classification of the early Palaeogene strata in the London Basin and East Anglia. Proceedings of the Geologists' Association, 105, 187-197. HOOKER, J. J. 1991. The sequence of mammals in the Thanetian and Ypresian of the London and Belgium Basins: location of the Palaeocene-Eocene boundary. Newsletters on Stratigraphy, 25, 75-90. KENNEW, J. P. & STOT'r, L. D. 1990. Proteus and ProtoOceanus: ancestral Paleogene oceans as revealed from Antarctic stable isotopic results. In: BARKER, R E & KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 865-880. &-1991. Terminal Paleocene deep-sea benthic crisis: sharp deep sea warming and paleoceanographic changes in Antarctica. Nature, 353, 225-229. & -1995. Terminal Paleocene mass extinction in the deep sea: association with global warming. In: STANLEY,S., KENNETT,J. R & KNOLL, A. (eds) The Effects of Global Change on Life. National Research Council, National Academy Press. KNox, R. W. O'B. 1984. Nannoplankton zonation and the Palaeocene/Eocene boundary beds of NW Europe: an indirect correlation by means of volcanic ash layers, Journal of the Geological Society, London, 141, 993-999. KOCH, P. L., ZACHOS, J. C. & GINGERICH, P. D. 1992. Coupled isotopic change in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary. Nature, 358, 319-322. KROOPNICrr P. M. 1985. The distribution of C-13 of total CO 2 in the world oceans. Deep Sea Research, Part A, 32, 57-84. LAURAIN, M., Barta, L., Bolin, C., Guemier, C., GruasCavagnetto, C., Louis, P., Riveline, J. & Thiry, M. 1983. Le sondage et la coupe du Mont Bernon Epernay (Marne). Etude srdimentologique et palrontologique du stratotype du Sparnacien de la srrie 6oc~ne. G~ologie de la France, 3, 235-253. Lu, G. & KELLER, G. 1993. Climatic and oceanographic events across the Paleocene-Eocene transition in the Antarctic Indian Ocean: inference from planktic foraminifera. Marine Micropaleontology, 21, 101-142. MILLER, K. G., JANECEK,T. R., KATZ, M. E. & KEIL, D. J. 1987. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary, Paleoceanography, 2, 741-761. MooK, W. G. 1974. Carbon isotope fractionation between dissolved bicarbonate and gaseous carbon dioxide. Earth and Planetary Science Letters, 22, 169176. Mt3LLEa, C. 1981. Biostratigraphic and paleoenvironmental interpretation of the Goban Spur region based upon a study of calcareous nannoplankton. In: GRACIANSKY,P. C. DE, POAG, C. W., ETAL. Initial Reports of the Deep Sea Drilling Project, 80, 573-597. -

-

-

-

PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene benthic foraminiferal isotopes and assemblages: implications for deep-water circulation. Paleoceanography, 7, 405-422. REA, D. K., ZACHOS, J. C., OWEN, R. M. & GINGERICH, P. D. 1990. Global change at the Paleocene-Eocene boundary. Climatic and evolutionary consequences of tectonic events. Palaeogeography, Palaeoclimatology, Palaeoecology, 79, 117-128. SETO, K., NOMURA,R. & NIITSUMA,N. 1991. Data report: oxygen and carbon isotope records of the upper Maestrichtian to lower Eocene benthic foraminifers at Site 752 in the eastern Indian Ocean. In: PEmCE, J. W., WEtSSEL, J., Er AL. Proceedings of the Ocean Drilling Program, Scientific Results, 121, 885-889. SHACKLETON, N. J., HALL, M. A. & BOERSMA, A. 1984. Oxygen and carbon isotope data from Leg 74 foraminifers. In: Initial Reports of the Deep Sea Drilling Project, 74, 599-612. SIESSER, W., LORD, A. R. & WARD, D. 1987. Calcareous nannoplankton biozonation of the Thanetian Stage in the type area. Journal of Micropaleontology, 6, 85-102. SNYDER, S. W. t~ WATERS, V. J. 1981. Cenozoic planktonic foraminiferal biostratigraphy of the Goban Spur region, Deep Sea Drilling Project Leg 80. In: DE GRACIANSKV,P. C., POAG, C. W., ET AL. Initial Reports of the Deep Sea Drilling Project 80, 439-472. SPERO, H. J., LERCHE, I. & WILUAMS, D. E 1991. Opeing the carbon isotope "vital effect" black box. 2, Quantitative model for interpreting foraminiferal carbon isotope data. Paleoceanography, 6, 639-655. STOTT, L. D. 1992. Higher temperatures and lower oceanic pCO2: A climate enigma at the end of the Paleocene epoch. Paleoceanography, 7, 395-404. --, KENNETX, J. E, SHACKLETON, N. J. ~r CORFIELD, R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: Inferences from the stable isotopic composition of planktonic foraminifers. In: BARKER,P. E, KENNETT,J. R, ETAL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 849-864 THOMAS, E. 1989. Development of Cenozoic deep-sea benthic foraminiferal faunas in Antarctic waters. In: CRAME, J. A. (ed.) Origins and Evolution of Antarctic Biota. Geological Society, London, Special Publication, 47, 283-296. 1990. Late Cretaceous-early Eocene mass extinctions in the deep sea. In: SHARPTON,V. L. & WARD, P. D. (eds) Global Catastrophes in Earth History Geological Society of America, Special Publication, 247, 481-495. & Shackleton, N. J. 1993. The Paleocene benthic foraminiferal extinction: timing, duration and association with stable isotope anomalies. (Abstract) Correlation of the early Paleogene in Northwest Europe. Geological Society, London. TJALSMA, R. C. & LOHMANN, G. P. 1983. PaleoceneEocene Bathyal and Abyssal Benthic Foramnifera from the Atlantic Ocean. Micropaleontology, Special Publication, 4. -

-

GLOBAL PALEOCENE-EOCENE CARBON ISOTOPE CHANGES ZACHOS, J. C., REA, D. K., SETO, K., NOMURA, R. & NIITSUMA, N. 1993. Paleogene and early Neogene deepwater paleoceanography of the Indian Ocean as determined from benthic foraminifera stable isotope

399

records. In: DUNCAN,R. A., Er AL. (eds) The Indian Ocean: A synthesis of results from the Ocean Drilling Program. Geophysical Monograph Series, American Geophysical Union, Washington, DC.

The Paleocene-Eocene benthic foraminiferal extinction and stable isotope anomalies E. T H O M A S

1' 2 & N. J. S H A C K L E T O N

3

1 Department of Earth Sciences, Downing Street, University of Cambridge, Cambridge CB2 3EQ, UK Present address: Center for the Study of Global Change, Department of Geology and Geophysics, PO Box 208109, Yale University, New Haven CT 06520-8109, USA 2 Department of Earth and Environmental Sciences, Wesleyan University, Middletown, CT 06459-1309, USA 3 Godwin Laboratory, Subdepartment of Quaternary Research, University of Cambridge, Cambridge CB2 3RS, UK Abstract: In the late Paleocene to early Eocene, deep sea benthic foraminifera suffered their only global extinction of the last 75 million years and diversity decreased worldwide by 30-50% in a few thousand years. At Maud Rise (Weddell Sea, Antarctica; Sites 689 and 690, palaeodepths 1100 m and 1900 m) and Walvis Ridge (Southeastern Atlantic, Sites 525 and 527, palaeodepths 1600 m and 3400 m) post-extinction faunas were low-diversity and high-dominance, but the dominant species differed by geographical location. At Maud Rise, post-extinction faunas were dominated by small, biserial and triserial species, while the large, thick-walled, long-lived deep sea species Nuttallidestruempyiwas absent. At Walvis Ridge, by contrast, they were dominated by long-lived species such as N. truempyi,with common to abundant small abyssaminid species. The faunal dominance patterns at the two locations thus suggest different post-extinction seafloor environments: increased flux of organic matter and possibly decreased oxygen levels at Maud Rise, decreased flux at Walvis Ridge. The species-richness remained very low for about 50 000 years, then gradually increased. The extinction was synchronous with a large, negative, short-term excursion of carbon and oxygen isotopes in planktonic and benthic foraminifera and bulk carbonate. The isotope excursions reached peak negative values in a few thousand years and values returned to preexcursion levels in about 50 000 years. The carbon isotope excursion was about -2%0 for benthic foraminifera at Walvis Ridge and Maud Rise, and about -4%0 for planktonic foraminifera at Maud Rise. At the latter sites vertical gradients thus decreased, possibly at least partially as a result of upwelling. The oxygen isotope excursion was about -1.5%o for benthic foraminifera at Walvis Ridge and Maud Rise, -1%o for planktonic foraminifera at Maud Rise. The rapid oxygen isotope excursion at a time when polar ice-sheets were absent or insignificant can be explained by an increase in temperature by 4-6~ of high latitude surface waters and deep waters world wide. The deep ocean temperature increase could have been caused by wanning of surface waters at high latitudes and continued formation of the deep waters at these locations, or by a switch from dominant formation of deep waters at high latitudes to formation at lower latitudes. Benthic foraminiferal post-extinction biogeographical patterns favour the latter explanation. The short-term carbon isotope excursion occurred in deep and surface waters, and in soil concretions and mammal teeth in the continental record. It is associated with increased CaCO 3dissolution over a wide depth range in the oceans, suggesting that a rapid transfer of isotopically light carbon from lithosphere or biosphere into the ocean-atmosphere system may have been involved. The rapidity of the initiation of the excursion (a few thousand years) and its short duration (50 000 years) suggest that such a transfer was probably not caused by changes in the ratio of organic carbon to carbonate deposition or erosion. Transfer of carbon from the terrestrial biosphere was probably not the cause, because it would require a much larger biosphere destruction than at the end of the Cretaceous, in conflict with the fossil record. It is difficult to explain the large shift by rapid emission into the atmosphere of volcanogenic CO2, although huge subaerial plateau basalt eruptions occurred at the time in the northern Atlantic. Probably a complex combination of processes and feedback was involved, including volcanogenic emission of CO 2, changing circulation patterns, changing productivity in the oceans and possibly on land, and changes in the relative size of the oceanic and atmospheric carbon reservoirs.

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlationof the EarlyPaleogenein NorthwestEurope, Geological Society Special Publication No. 101, pp. 401--441.

401

402

E. THOMAS & N. J. SHACKLETON

During the late Paleocene and early Eocene important changes occurred in global climate, in plate tectonic processes and in the global carbon cycle. The discussion of these events has been complicated and confused by two facts: that there is no unequivocal definition of the 'Paleocene-Eocene boundary' because these words mean different things to different people (e.g. Berggren & Aubry 1996); and that different events occurred at different time scales, from millions of years to thousands of years. We will first discuss longer timescale events, occurring over several millions of years in palaeomagnetic chrons C25 and C24, at about 55-60Ma in the geomagnetic polarity time scale of Berggren et al. (1985), and 52.5-57.5 Ma in that of Cande & Kent (1995). The beginning of this interval has generally been placed in the Paleocene and the end in the Eocene, but the location of the boundary has varied. This transitional Paleocene-Eocene period witnessed a long-time warming of the deep oceans beginning in the earliest Paleocene, and a long-term decrease of carbon isotopic values in tests of benthic and planktonic foraminifera and bulk carbonate, beginning in early Chron C25 (Shackleton 1986, 1987; Zachos et al. 1993). Highlatitude areas experienced the highest temperatures of the Cenozoic as indicated by the presence of warm-water pelagic marine organisms (e.g. Haq et al. 1977; Premoli-Silva & Boersma 1984; Boersma et al. 1987; Stott & Kennett 1990; Aubry 1992; Berggren 1992; Ottens & Nederbragt 1992). Crocodiles, turtles and other thermophilic biota occurred at high northern latitudes (Estes & Hutchison 1980; McKenna 1980; Gingerich 1983; Markwick 1994) in the presence of vegetation and soil-types indicating warm climates (Kemp 1978; Nilsen & Kerr 1978; Wolfe 1978; Wolfe & Poore 1982; Schmidt 1991). Palynological data from the North Sea indicate peak warmth (Schroeder 1992). Clay mineral associations in oceanic sediments indicate high humidity and intense chemical weathering as indicated by high abundances of kaolinite (Robert & Maillot 1990; Robert & Chamley 1991; Robert & Kennett 1992, 1994); similar peaks in kaolinite abundance have been observed in sediments from the New Jersey margin (Gibson et al. 1993) and the North Sea region (Knox, written comm. 1993). Oxygen isotopic measurements show high temperatures and shallow temperature gradients from low to high latitudes (Shackleton & Boersma 1981; Oberh~insli & Hsti 1986; Stott et al. 1990; Barrera & Huber 1991; Seto et al. 1991; Stott 1992; Zachos et al. 1994; Bralower et al. 1995a, b). Dust concentrations in oceanic sediments reached very low levels in the upper part of Chron C24r (Janecek & Rea 1983;

Miller et al. 1987b; Rea et al. 1990; Hovan & Rea 1992; Rea 1994), suggesting very low wind strength. During this warm period terrestrial and shallowwater marine organisms did not suffer major net extinctions (e.g. Raup & Sepkoski 1986), but it was a time of major origination of species and high diversity on land (European mammals and flora: Hooker 1991; Collinson 1983; North American mammals: Butler et al. 1981, 1987; Wing 1984; Rea et al. 1990; Wing et al. 1991) and in the surface oceans (e.g. planktonic foraminifera: Kennett 1978; Boersma et al. 1987; Boersma & Premoli-Silva 1991; Corfield & Shackleton 1988; Berggren 1992; Corfield 1993; calcareous nannofossils: Romein 1979; Aubry 1992; dinoflagellates, Oberh~nsli & Hsfi 1986; McGowran 1991). An explanation of the several million year-long warm period has commonly been sought in elevated levels of atmospheric CO 2, caused by plate tectonic related processes (e.g. Williams 1986; McGowran 1991; Rea et al. 1990). During Chrons C24-C25 there was a worldwide plate-tectonic reorganization, involving the slow-down of the northward motion of the Indian subcontinent because of its collision with Asia (Klootwijk et al. 1991; Beck et al. 1995a). According to some researchers high-temperature metamorphism started in the Himalayas (Tonarini et al. 1993; Smith et al. 1994; Beck et al. 1995a) and delivered large amounts of CO 2 to the atmosphere by decarbonation (Touret 1992; Kerrick & Caldeira 1993, 1994), but this mechanism of CO 2 delivery has been doubted (Selverstone & Gutzler 1993). Increasing levels of CO 2 in the atmosphere resulting from the India-Asia collision could alternatively have been generated by erosion of sediments rich in organic matter (Beck et al. 1995b). In addition, subduction changed in direction in the North Pacific (Goldfarb et al. 1991). Continental break-up started in the North Atlantic (Roberts et al. 1984; Eldholm 1990; Larsen et al. 1992), accompanied by violent, partially subaerial plateau basalt eruptive activity (White 1989; White & MacKenzie 1989; Eldholm & Thomas 1993; Kaiho & Saito 1994). High hydrothermal activity along Pacific oceanic ridges may have contributed to increased atmospheric CO 2 levels (Owen & Rea 1985; Olivarez & Owen 1989; Kyte et al. 1993), although it is doubted whether increased spreading activity will cause a net increase in atmospheric pCO 2 (Staudigel et al. 1989; Varekamp et al. 1992). High concentrations of other greenhouse gases such as methane have also been invoked (Sloan et al. 1992, 1995). High atmospheric pCO 2 can not fully explain the global climate, however: how could high latitudes have warmed while the equatorial regions were not

PALEOCENE--EOCENE BENTHIC FORAMINIFERALEXTINCTION warmer than today (Shackleton & Boersma 1981; Zachos et al. 1994; Sloan et al. 1995)? The oceans could have transported more heat, involving a larger part of the ocean waters (Barton 1987; Sloan & Barron 1992; Barron & Peterson 1991), but it is not clear that even involvement of the whole ocean in heat transport can succeed in keeping the high latitudes at the temperatures of 15-18~ suggested by oxygen isotope studies (Crowley 1991; Walker & Sloan 1992; Sloan et al. 1995). The theory of increased oceanic heat transport appeared attractive because it had been long theorized that deep and intermediate waters in the oceans could have formed as high salinity, high temperature waters by evaporation in subtropical latitudes in the absence of very cold polar areas (Chamberlin 1906; Brass et al. 1982; Hay 1989). Surface oceans at low latitudes show the cooling expected if heat transport was more active than today (Zachos et al. 1993, 1994; Bralower et al. 1995b). The mechanisms for enhanced oceanic heat transport, however, remain undefined (Sloan et al. 1995). Evidence for the presence of such warm saline bottom water (WSBW) is not unequivocal. Some investigators suggest that WSBW was the dominant water mass over much of the Cenozoic (Matthews & Poore 1980; Prentice & Matthews 1988), or during at least the early Paleogene (Kennett & Stott 1990). Some models, however, predict that the deep oceans will turn anoxic over large regions when deep waters form largely at subtropical latitudes (Herbert & Sarmiento 1991), and this did not happen during the Cenozoic (e.g. Thomas 1992). Some investigators concluded that much of the oceans' intermediate and deep waters was formed at high southern latitudes during most of the Maastrichtian and Cenozoic (e.g. Barrera et al. 1987; Miller et al. 1987a; Katz & Miller 1991; Thomas 1992; Zachos et al. 1992, 1993). Major changes in the deep-water environments of the oceans would be expected to result from a reversal in deep water circulation (Kennett & Stott 1991), and these should be reflected in the composition of deep-sea benthic foraminiferal faunas. Paleocene deep-sea benthic foraminiferal faunas closely resemble Late Cretaceous faunas (Cushman 1946), and the major break in deep-sea benthic foraminiferal faunas occurred somewhere in Chron 24, i.e. between Paleocene and Eocene if seen at coarse time-scales (e.g. Beckmann 1960; von Hillebrandt 1962; Braga et al. 1975; Schnitker 1979; Tjalsma & Lohmann 1983; Boersma 1984b; Thomas 1990b; Bolli et al. 1994). In Chron 24r (corresponding to planktonic foraminiferal zones P5-P6, and calcareous nannofossil zone NP9) major faunal change occurred in bathyal and abyssal faunas in all the world's oceans (Miller et al. 1987b; Boltovskoy & Boltovskoy 1988, 1989;

403

Berggren & Miller 1989; Thomas 1989, 1990a, b, 1992; Katz & Miller 1991; Mackensen & Berggren 1992; Reynolds 1992, unpublished MSc thesis, Univ. of Maine; Pak & Miller 1992, 1995; Nomura 1991; Kaiho 1988, 1991, 1994a, b; Miller et al. 1992; Kaiho et al. 1993; Bolli et al. 1994. Benthic foraminifera underwent coeval extinction in neritic to upper-middle bathyal environments in land sections from Israel to Egypt, North Africa and Spain (Molina et al. 1992; Speijer 1994), in the North Sea shelf seas (King 1989; Charnock & Jones 1990), in New Jersey (USA, Gibson et al. 1993), and along the western Pacific margin from Japan (Kaiho 1988) to New Zealand (Hornibrook et al. 1989; Kaiho et al. 1993). At these depths the faunal change was associated with low oxygen conditions, as indicated by the presence of black or dark grey, commonly laminated sediments (North Sea, Japan, New Zealand, New Jersey, Spain, Israel, Egypt). Many authors have explained these conditions as resulting from local lack of circulation (e.g. for the North Sea Basin; Charnock & Jones 1990), but such local effects occurred during worldwide low oxygen conditions, which may have been at least partially responsible. Some authors suggested that relatively low oxygen conditions also occurred, at least locally, in the world's deep oceans (Thomas 1990b, 1992; Kaiho 1991). Most authors implied at least some form of change in deep-water circulation in the benthic foraminiferal extinction, which was seen as caused by changes in temperature as well as nutrient and oxygen content of the deep waters (Miller et al. 1987b; Thomas 1989; Katz & Miller 1991; Nomura 1991). Others suggested that oceanic productivity decreased drastically (e.g. Moore et al. 1984; Shackleton 1987; Shackleton et al. 1985; Coffield & Shackleton 1988; Rea et al. 1990; Corfield & Cartlidge 1992a, b; Corfield 1993), with possible effects on the benthic faunas. Yet others suggested that at least in some areas productivity increased (Thomas 1992; Speijer 1994). Only recently, however, has it become clear that the deep sea benthic extinction in the latest Paleocene was a unique event in its global extent and rapidity (Thomas 1989; Kennett & Stott 1991; Pak & Miller 1992; Kaiho 1994b; Robert & Kennett 1994). In the next paragraphs we will discuss events which happened on timescales of thousands to ten thousands and not millions of years, at some time early in the reversed part of chron C24, in nannofossil zone NP9, and in planktonic foraminiferal zone P5 (Berggren & Aubry 1996; Aubry et al. 1996). It had seemed reasonable to suppose that the benthic foraminiferal extinction was not synchronous world wide (Miller et al.

404

E. THOMAS 8~ N. J. SHACKLETON tope excursion was very short-lived, there may have been hiatuses at Site 577 (Berggren et al. 1995), and sampling by Pak & Miller (1992) was not so detailed as to ensure capture of the most extreme values. The explanation of these short-term changes in climate and in the carbon cycle is being actively debated. Oxygen isotope data suggest rapid warming (over less than 5000 years) of the deep ocean waters at high and low latitudes by 4-6~ but essentially no warming of surface waters in the tropical Pacific (Zachos et al. 1994; Bralower et al. 1995a, b), so that latitudinal temperature gradients of surface water were very low for about 50 000 years. Such a rapid, deep ocean-wide temperature change has been explained by a change in dominant oceanic circulation pattern, from dominant production of deep to intermediate waters at high latitudes to dominant production in subtropical regions (e.g. Kennett & Stott 1991; Thomas 1992). Major questions remain as to the exact nature of the upheaval in the carbon cycle, the feedback processes resulting from atmosphere-ocean interactions, and whether deep to intermediate waters did indeed form dominantly at subtropical latitudes, and for how long? In this paper, we present new benthic faunal and isotope data at high resolution from four sites at different depths in the southeastern Atlantic Ocean and the Weddell Sea (Fig. 1), to investigate the magnitude of the isotopic excursions in different areas and at different water depths. We use the new, detailed data to discuss

1987b) because a mechanism for globally synchronous extinction in such a large part of the earth's environment as the deep ocean is difficult to envisage. The observation, however, that the extinction was coeval with a very large, short-term negative excursion in the benthic as well as planktonic dl3C and dlsO records at ODP Site 690 (Kennett & Stott 1991) suggested that the extinction might have been caused by rapid warming at high latitudes, causing large-scale changes in oceanic circulation and upheaval in the global carbon cycle. Additional research demonstrated that the short-term, extremely negative shift in the carbon isotope values was, as expected by Kennett & Stott (1991), a more than local phenomenon: the event has been recognized in the Bay of Biscay (Pak & Miller 1992; Stott et al. 1996), Pacific Ocean (Pak & Miller 1992, 1995; Bralower et al. 1995a, b), and Indian Ocean (Thomas et al. 1992; Lu & Keller 1993). A large negative anomaly in carbon isotopes was also recognized in enamel of land-herbivore teeth and in carbonate concretions, clearly demonstrating that the atmosphere as well as the ocean was involved (Koch et al. 1992; Stott et al. 1996). The globally averaged magnitude of the shortterm shift is in question: it is about -2%0 in the deep Antarctic (Kennett & Stott 1991; this paper) and the southern Indian Ocean (Lu & Keller 1993), but only about -1%o in the Pacific and the Bay of Biscay (Pak & Miller 1992). The smaller values may not reflect a global average because the iso-

90 60

"

30

~

9" 5 2 7

.

.

~,

9 -'-~',

~

90

Fig. 1. Palaeogeographic map of the continents in the late Paleocene, after Zachos et al. (1994). Sites from which data are presented in this study are indicated by *; other sites mentioned in the text by +.

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

the bathymetric and biogeographical patterns of the benthic foraminiferal extinction, and re-evaluate the existing database.

Material and methods Sites a n d stratigraphy: M a u d Rise

Sites 689 (64~ 03~ present water depth 2080 m) and 690 (65~ 1~ present water depth 2914 m) were drilled on Maud Rise at the eastern end of the Weddell Sea (Barker et al. 1988, fig. 1). Site 689 is on the northeastern side of the ridge near its crest, Site 690 is 116 km to the southwest on its southwestern flank. Depth estimates for the sites at the end of the Paleocene, based on faunal contents, agreed well with backtracking estimates, giving about 1100 m for Site 689, about 1900 m for Site 690 (Thomas 1990, 1992). At both sites sediments of Maastrichtian to Pleistocene age were recovered. Paleogene biomagnetostratigraphy was reviewed by Thomas et al. (1990). Upper Paleocene to lower Eocene sediments consist of chalks and calcareous oozes, with admixture of fine-grained terrigenous matter at Site 690. Recovery was good at Site 690, less so at 689. Core deformation was minimal, but the biostratigraphy was difficult to interpret because the high latitude of the sites caused the absence of many marker species of planktonic foraminifera, and opened the possibility of diachroneity for nannofossil markers (Pospichal & Wise 1990; Aubry et al. 1996). Major differences of opinion in the interpretation of the records centred on the lowermost Eocene. Stott & Kennett (1990) maintained that the section was basically complete over the upper Paleocene-lowermost Eocene, whereas SpieB (1990), Thomas et al. (1990) and Pospichal & Wise (1990) thought there was at least one, about 2 million year-long hiatus in Chrons 22 and 23. Aubry et al. (1996) concluded that an additional hiatus occurs in Chron C24n. We do not agree with Kennett & Stott (1991) that the records at Site 690 were undisturbed by bioturbation. In our opinion, the sediments show a different ichnofossil assemblage in the interval just after the extinction event, but no lamination. A few corroded specimens of Gavelinella beccariiformis are present two samples above the extinction, and presumably reworked. Therefore we think that the exact sequence of events in samples spaced only by a few centimetres is difficult to determine. At Site 689 the upper Paleocene-lower Eocene succession was very incomplete, but a section of sediment extending from in the lower part of Core 689B-22X to the poorly recovered Core 689B-24X contains the benthic foraminiferal extinction and the CP7/CP8 nannofossil zonal boundary (Thomas

405

1990; Pospichal & Wise 1990; Thomas et al. 1990). The section is bounded by hiatuses and sedimentation rates can not be determined with precision. Benthic foraminifera were studied by Thomas (1990a); in this paper we present additional data at higher resolution for Sites 689 and 690 (Appendices 1, 2). For Site 690 our data are from Cores 690B-16 to 690B-22, encompassing sediment deposited in Chron C24r according to Aubry et al. (1996). Our data from Hole 689B are from the interval described above (Fig. 2). Carbon and oxygen isotope data of benthic foraminifera were collected by Kennett & Stott (1990, 1991), for planktonic foraminifera by Stott et al. (1990) and Corfield & Cartlidge (1992b), and for bulk carbonate by Shackleton & Hall (1990). We present additional data at high resolution for both sites (Appendix 5). Sites a n d stratigraphy: Walvis R i d g e

Sites 525 (29.~ 02~ present water depth 2467 m) and 527 (28~ 01~ present water depth 4428 m) were drilled on the summit area and the lowermost western slopes of the Walvis Ridge, respectively (Moore et al. 1984, fig. 1). At both sites, Paleocene to Eocene calcareous oozes and chalks were recovered (Shackleton et al. 1984a). Palaeodepths for the end of the Paleocene as derived by backtracking (1600 m for Site 525, 3400 m for Site 527; Moore et al. 1984) are in agreement with benthic faunal data (Boersma 1984b; this paper). Carbon and oxygen isotope data from foraminifera as well as bulk carbonate were determined (Shackleton et al. 1984b; Shackleton & Hall 1984). Nannofossil stratigraphy was described at fairly low resolution by Manivit (1984) and in more detail by Backman (1986a, b). Planktonic and benthic foraminiferal biostratigraphy at low resolution was described by Boersma (1984a, b). In this paper, we present additional benthic faunal data and isotope data on benthic foraminifera and bulk carbonate over the interval of the benthic extinction. At both sites, despite the difference in palaeodepths, the extinction occurred within the lower few centimetres, but above the base of, an interval of dissolution of 40-50 cm thick, with a sharp lower boundary and a gradual upper boundary (Table 1). In this interval planktonic foraminifera were fragmented, but benthics showed good preservation. We took samples from the cores in which the extinction occurred only. N u m e r i c a l ages

Numerical ages for Paleocene sediments are in the process of being revised (e.g. Odin & Luterbacher

406

E. THOMAS & N. J. SHACKLETON <5180, %0 3.0 30''

0.0 -1.0 9.~;:2.0. . . . . . . .1.0 ............

-2.0

35

20 30 40 30 . . . . :'. . . . . I . . . . . . ;..:1: =. 35

40

P9

40

* o***

" **- ** "" =~

"

.-. ! : . . .': ~.

~3 50

60 70 .............

80

35t

.*

so

9

::;:.:

9~.

401 ]

*.:*

45

-1.0 0.0 1.0 2.0 3.0 301 . . . . . ;.; . . . . . . . . . . . . . . .

. "'-.-':"

e

45"]

~'"

1

1

9 ~

=

.~

so

i

SSl

:,

-

"

";:~-'3~

55 "]

,o]

.

60

9 ,',':.~, *

60t

~'~.

o

~

65

,

,

.~

65t .

t

t **'2 '~ %* . . ;.I.

ft.

9 ....,:~."

~..,.'~ COld

o

.5"%.

.

]

~ 55 ~

50 "" ,

::.;I":"" .:

m 45

(~130, %0

of species

number

warm

'~1 75

I ~. 9 ..lie =t.

~

o*

9

70-1

""

,~

~*~

b. 6

~ o

75'

Fig. 2. Benthicforaminiferal carbon and oxygen isotope data (Kennett & Stott 1990; this paper) and species richness of benthic foraminifera (Thomas 1990; this paper) over the Upper Cretaceous through middle Oligocene at Maud Rise, plotted versus numerical age on the Berggren et al. (1985) and Aubry et al. (1988) timescale. Note the extreme nature of the carbon isotope event as compared to the events at the K/T boundary (66.4 Ma), and its location toward the end of the long-term shift to more negative values, from about 61-58 Ma.

1992) and major changes are proposed from the Berggren et al. (1985) and Aubry et al. (1988) geomagnetic polarity timescale. The most recent compilations show Paleocene-Eocene boundary ages several million years less than in these two papers (Wing et al. 1991; Cande & Kent 1992, 1995; Jenkins & Luterbacher 1992; Odin & Luterbacher 1992; Berggren et al. 1995; Berggren & Aubry, 1996; Aubry et al. 1996). We are concerned with events during Chron C24r, which has changed little in length between different time scales (around 2.5 million years), although its numerical age has changed from 56.1458.64Ma in Berggren et aL (1985) to 53.25055.981 Ma (Cande & Kent 1992) to 53.34755.904 Ma (Cande & Kent 1994; see also Berggren & Aubry 1996). Modifications have also been proposed for the planktonic foraminiferal zonation in the PaleoceneEocene boundary interval (Berggren et al. 1995). In the new zonation Zones P5 and P6a of Berggren & Miller (1988) are combined in Zone P5 (the M o r o z o v e l l a v e l a s c o e n s i s Partial Range Zone; Berggren et al. 1995). This revised zonation supports our statement that the benthic foraminiferal extinction was synchronous worldwide. At Sites 525 and 527 the benthic extinction occurred before the last appearance of M o r o z o v e l l a v e l a s c o e n s i s and before the first appearance of M o r o z o v e l l a s u b b o t i n a e (Boersma 1984a), i.e. in planktonic foraminiferal Zone P5 as modified by Berggren et al. (1995). In the planktonic

Table 1. Dissolution intervals and benthic foraminiferal extinction, Walvis Ridge sites Depth (mbsf) Hole 525A Top dark layer Benthic extinction Bottom dark layer

Hole 527

391.87-392.01 200.37-200.56 392.41-392.57 200.89-201.07 392.57 201.08

foraminiferal zonation of Berggren & Miller (1988) the extinction occurred in Zone P5 at Sites 525 and 527, but in P6a at equatorial Pacific Site 577 (Miller et al. 1987b; Pak & Miller 1992) and in Mediterranean land sections (Speijer 1994). We agree with Pak & Miller (1992) and Aubry et al. (1996) that the benthic extinction occurs lower in palaeomagnetic Chron C24r than estimated by assuming constant sedimentation rates in Chron 24r (Kennett & Stott 1991; Thomas 1992), and that its age was close to 58 Ma in the Berggren et al. (1985) time scale (Eldholm & Thomas 1993). In the time scale of Berggren et al. (1995) the extinction is close to 55.5 Ma as proposed by Aubry et al. (1996). The isotope excursions and the benthic faunal extinction were very short events (less than 105 years; Kennett & Stott 1991; Robert & Kennett 1994; Aubry et al. 1996). We cannot assume that sedimentation rates were constant over the interval between age markers where the event occurred at

PALEOCENE--EOCENE BENTHIC FORAMINIFERALEXTINCTION Sites 525 and 527, because of the presence of a dissolution interval, suggesting decreased sedimentation rates for a short interval starting just before the extinction. Our measurements of CaCO 3 content gave minimum values of 35% at Site 525, 20% at Site 527. At Site 690, where sedimentation rates over the interval between the top of Chron 25 and the nannofossil zone NP9/NP10 boundary were the highest of all sites, CaCO3-values also fluctuate, with minimum values of about 65% just below the benthic extinction (O'Connell 1990; fig. 6 in Thomas 1992). We used the geomagnetic polarity timescale of Berggren et al. (1995) and the biostratigraphic interpretation of Site 690 of Aubry et al. (1996), giving ages of 55.5 Ma to the benthic extinction event, and 55 Ma to the NP9/NP10 boundary (Berggren & Aubry 1996). We extrapolated age at constant sedimentation rates between 55 and 55.5 Ma, and between 55.5 Ma and the top of Chron 25, using numerical ages as derived from the Berggren et al. (1995) timescale for all age marker levels given in Thomas et al. (1990) for the Maud Rise sites, in Shackleton et al. (1984b) for the Walvis Ridge sites. Then we revised the numerical ages for Site 689 and the Walvis Ridge samples close to the extinction to reflect the supposedly lower sedimentation rates of the intervals with low values of CaCO 3. To do this we used the carbon isotope record of Nuttallides truempyi at Site 690 as standard, because sedimentation rates were highest at this site. For the interval between the boundary between NP9 and NP10 and the benthic foraminiferal extinction we revised ages until the carbon isotope curves at Sites 525, 527 and 690 agreed best in general shape (Fig. 3); we then did the same for the Lenticulina spp. isotope record at Site 689. The re-calculation of the ages using the shape of the carbon isotope curve resulted in lower calculated sedimentation rates for the CaCO3-poor sediment interval at the sites. All numerical ages are listed in the appendices.

407

o

# o

d

.

.

.

.

i

.

.

.

.

,

.

.

.

.

i

.

.

.

.

,

.

.

.

.

Age, Ma

---o--"7,

525

9

527

689

= 0 4,

.

.

.

(Lenticulina)

690 .

,

. . . .

,~ , , , , ,

.

.

,

.

.

.

.

,

. . . .

Age, Ma

Fig. 3. Benthic foraminiferal carbon and oxygen isotope excursions at Sites 689, 690, 525 and 527; plotted are uncorrected values for N. truempyi for Sites 690, 525 and 527, for Lenticulina spp. for Site 689, using the geomagnetic polarity timescale of Cande & Kent (1995) and Berggren et al. (1996). Ages of Site 690 after Aubry et al. (1996); ages for the other sites derived by fitting the carbon isotope curves as described in the text.

Sample preparation

Samples for benthic foraminiferal faunal and isotope analysis were dried overnight at 50~ and weighed, then soaked overnight in distilled water. Most samples disaggregated readily and could be washed over a 63 gm screen. For isotope analysis of planktonic foraminifera, samples were split and foraminifera picked from two size fractions: size e (212-250gm) and size d (300-250gm) as is standard use in the Cambridge and Oxford isotope laboratories; there were no specimens larger than 300 gm in the Maud Rise samples. We analysed 20-40 specimens of Acarinina mckannai. In a few

samples 'Morozovella' convexa was present in large enough numbers for analysis. Benthic foraminifera for isotope analysis were picked from the > 125 ~tm size fraction, and we analysed 12-20 specimens. Samples for isotope analysis of the small species Tappanina selmensis consisted of 60-80 specimens. Lenticulina spp. commonly show large fluctuations in isotope values; we excluded from analysis specimens larger than 500 ktm, which showed erroneous values, possibly because they resulted from reworking. All specimens were ultrasonicated to remove adhering

408

E. THOMAS • N. J. SHACKLETON

microfossils, then dried at 50~ Foraminifera were transferred to reaction vessels and roasted at 400~ Bulk samples, weighing a few milligrams, were taken from samples from Cores 525A-32 and 527-24, dried and vacuum roasted at 400~ to remove organic contaminants. The samples were then reacted with 100% orthophosphoric acid at 90~ using a VG Isotech Isocarb common acid bath system. The evolved carbon dioxide was analyzed in a VG isotech SIRA Series II mass spectrometer. The results were calibrated to PDB by repeated analysis of a carbonate standard. Analytical accuracy is better than 0.08%0 for both 5180 and ~513C. Benthic foraminifera for faunal analysis were picked from the > 63 lam size fraction, following Thomas (1990a). All specimens counted were picked and mounted in cardboard slides. All samples contained sufficient specimens for analysis (> 250), and counts are shown in the appendices. Taxonomy is as in Thomas (1990a) and largely follows Van Morkhoven et al. (1986). Thomas (1990a) misidentified Neoeponides hillebrandti as Neoeponides lunata and the reverse; that mistake has been rectified in the appendices of this paper. All species richness numbers were recalculated to 100 specimens using rarefaction (Sanders 1968).

Results Faunas We define the level of extinction as the interval where species richness declined most rapidly. Preextinction faunas at Walvis Ridge and Maud Rise are similar, with almost all species present at all 4 sites (Figs 4, 5, 6, 7), although at varying relative abundances. These faunas contain many cosmopolitan, Late Cretaceous through Paleocene taxa with a large depth range, such as Gavelinella beccariiformis, Gavelinella hyphalus, Neoeponides hillebrandti, Neoeponides lunata, Pullenia coryelli, Bolivinoides delicatulus, Neoflabellina semireticulata and agglutinated taxa such as Tritaxia paleocenica, Tritaxia havanensis and Dorothia oxycona. The agglutinated taxa are more common at Site 527, the deepest site studied. Aragonia velascoensis is rare or absent at Maud Rise, and rare at Site 525 (Fig. 6). Stilostomella spp. are more common at Maud Rise than at Walvis Ridge. Preextinction faunas have a very high species richness, with many rare uniserial lagenid species and unilocular taxa. Many common, long-lived species occur, such as Cibicidoides pseudoperlucidus, Oridorsalis umbonatus, Nonion havanense, Nonionella robusta, Anomalina spissiformis and Anomalinoides semicribrata. Overall, bi- and triserial species were less common at Walvis Ridge

than at Maud Rise. Pre-extinction, large peaks in relative abundance of species such as Bulimina thanetensis and S. brevispinosa are not observed at Walvis Ridge. Rectobulimina carpentierae occurs only at the two deeper sites, 527 and 690, with strongly fluctuating relative abundances at the latter site, and Bulimina thanetensis was more common at these sites. Gavelinella beccariiformis was more common at the shallower sites with highest relative abundances at Site 689, in agreement with Katz & Miller (1991) who consider this species typical for bathyal sites at high latitudes. Overall, the pre-extinction faunas were remarkably similar in species composition given the large differences in depth and geographical location, in agreement with Tjalsma & Lohmann (1983) and Kaiho (1988, 1991). This uniformity ended with the extinction. At Site 690, many of the typical Paleocene species had their last appearance in the highest sample with high diversity, 690B-19H-3, 72-74 cm. In sample 690B-19H-3, 66-68cm, the diversity decreased from 49 to 26 species per 100 specimens, and bitriserial species were much more abundant, although in our age model these samples differ in age by only 1000 years. Faunal abundance patterns thus changed at the same level where a high number of last appearances occurred. At Site 689, however, the last appearance of G. beccariiformis and other Paleocene species occurred in sample 689B-23X-1, 80-82 cm, whereas the major drop in species richness from 60 to 42 species per 100 specimens occurred between 23X-1, 80-82 cm and 23X-1, 87-89 cm. Samples 23X-1, 80-82 and 23X-1, 87-89 both contain common T. selmensis and A. aragonensis, typical species for the postextinction period, but the species richness is high in these samples and G. beccariiformis is present. These appearances might result from bioturbation or core disturbance. At both Maud Rise sites the extinction occurred after a decrease in carbonate content of the sediments to about 65% (O'Connell 1990; Thomas 1992), and after the first appearance of the keeled, warm-water planktonic species 'Morozovella' convexa at high southern latitudes (Stott & Kennett 1990; this paper). At Walvis Ridge a sharp drop in benthic foraminiferal species richness from 52 to 34 species per 100 specimens occurred between samples 525A-32-6, 145-147cm and 525A-32-6, 130132cm. At Site 527 species richness decreased from 53 to 24 species per 100 specimens between 527-24-2, 56-58 cm and 527-24-2, 38-40 cm. The level of most rapid decrease in species richness is within the lower part of a dark, low-carbonate layer (Table 1): the benthic extinction thus occurred after the onset of increased CaCO 3 dissolution. The preservation of benthic foraminifera in this layer of

PALEOCENE-EOCENE BENTHIC FORAMINIFERAL EXTINCTION

Fig. 4. Benthic foraminiferal relative abundances c-f most taxa at Site 689; see Appendix 1 for counts.

409

410

E. THOMAS •

N. J. SHACKLETON

Fig. 5. Benthic foraminiferal relative abundances of the taxa, Site 690; see Appendix for counts.

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

411

Fig. 6. Benthic foraminiferal relative abundances of the most common taxa at Site 525; see Appendix 2 for counts.

412

E. THOMAS • N. J. SHACKLETON

Fig. 7. Benthic forarniniferalrelative abundances of the most taxa, Site 527; see Appendix 4 for counts.

low CaCO3-content and increased fragmentation of planktonic foraminifera is good. The samples in the interval with low diversity in the first 50 000 years after the extinction differ in faunal composition at the sites, with most differences between the Walvis Ridge and the Maud Rise

faunas, lesser differences between the sites at different depths in each area. These faunas occur during the interval with very low ~513C isotopic values at all 4 sites (see below), and thus can be considered coeval. At Site 690, the post-extinction faunas had high relative abundances of Eouvigerina

PALEOCENE--EOCENE BENTHIC FORAMINIFERALEXTINCTION sp., and at both Maud Rise sites Tappanina selmensis, Bolivinoides cf. decorata, Bulimina ovula, Bulimina simplex and Bulimina trinitatensis had high, but strongly fluctuating relative abundances. N. truempyi was rare or absent just after the extinction. Nuttallides umbonifera first occurred just after the extinction at Maud Rise, but is not present in the samples studied from Walvis Ridge. At the Maud Rise sites abyssaminid species (small, thin-walled, spiral species; Schnitker 1979; Tjalsma & Lohmann 1983) were more common after the extinction. At Walvis Ridge there was a much stronger increase in the relative abundance of these species, with the highest at the deeper Site 527, in agreement with Tjalsma & Lohmann (1983) and Katz & Miller (1991). In contrast to the patterns at Maud Rise, the post-extinction faunas at Walvis Ridge had very high relative abundances of Nuttallides truempyi, as described by Tjalsma & Lohmann (1983) and Katz & Miller (1991) for other Atlantic sites. A distinct drop in relative abundance of biserial and triserial species occurred at Site 527. Slightly higher in the section, however, the biserial species Tappanina selmensis and Aragonia aragonenis increased in relative abundance, with highest abundances at the shallower site. Tappanina selmensis never reached such a high abundance as at the Maud Rise sites (Figs 4, 5, 6, 7). Many rare species (such as lenticulinids, uniserial lagenids, unilocular taxa) were absent at all sites in the post-extinction faunas, at least partially causing the low species richness. Many of these species, however, reappeared higher in the section, and their absence may thus be apparent because of the decreased evenness of the faunas (Signor & Lipps 1982). At counts of about 300 specimens a lower number of species would be observed in an assemblage with higher dominance (Gage & Tyler 1991). Comparison of the faunas (Fig. 8) showed that at all sites the observed species richness dropped precipitously, with lowest values reached at the deepest Maud Rise site. The species richness appears to recover more quickly at the Walvis Ridge sites, especially at Site 525, if ages are determined according to Fig. 3 (fitting of the 813C curve of N. truempyi). Interesting is the strong increase in relative abundance of N. truempyi at the Walvis Ridge Sites, coeval with its strong decline at both Maud Rise sites. Abyssaminid species increased in relative abundance after the extinction at all sites, but much more so at the Walvis Ridge sites, and the shallower Walvis Ridge site has higher relative abundances of these species than the deeper Maud Rise site. Biserial and triserial species increased in relative abundance just after the extinction at both Maud Rise sites, but decreased

413

at the deeper Walvis Ridge site and showed little change at Site 525. A typical component of post-extinction faunas in all oceans are the species T. selmensis and A. aragonensis (Boersma 1984b), although their relative abundances vary from site to site. T. selmensis first appeared in the Late Cretaceous, but occurred at low abundances until the extinction event. A. aragonensis had its first appearance at the time of the extinction (Tjalsma & Lohmann 1983; Van Morkhoven et al. 1986; Bolli et al. 1994). G. beccariiformis last appeared at all sites at about the same time, although the species started a decline in relative abundance at Site 689 several hundred thousand years earlier (Fig. 4). Overall, however, this pattern does not agree with the statement by Tjalsma & Lohmann (1983) that the G. beccariiformis faunas became gradually restricted to shallower depths before becoming extinct. B. thanetensis became extinct in the late Paleocene at all sites, but always a few samples higher than G. beccariiformis.

Benthic foraminiferal isotopes We measured stable isotopes in several species of deep sea benthic foraminifera from the Maud Rise sites (Figs 9, 10). We wanted to derive an isotopic record of the extinction event that is as complete as possible, and N. truempyi and Cibicidoides spp. are not present in large enough numbers for isotope analysis just after the extinction. We also wanted to determine whether species that are presumed to be infaunal or epifaunal because of their morphology (Corliss & Chen 1988; Rosoff & Corliss 1992; Thomas 1990) show a different carbon isotopic signature, as observed in Recent and Neogene taxa (Woodruff & Savin 1985; Zahn et al. 1986; Altenbach & Sarnthein 1989; McCorkle et al. 1990). Infaunal species precipitate their tests in contact with pore waters, not sea water, and can thus be expected to have lower 813C values. Most species thought to be infaunal because of their morphology had indeed lower 813C values than species thought to be epifaunal, with the exception of Oridorsalis umbonatus. This longlived, cosmopolitan, trochospiral species plots in the field of infaunal species (Fig. 10). Its carbon isotope data should thus be interpreted carefully, especially because Rathburn & Corliss (1994) reported that Recent representatives of this taxon live infaunally. The benthic species show considerable scatter in their isotopic values, and there are differences between the patterns at Sites 689 and 690. At Site 690, Lenticulina spp., a presumed infaunal group, shows clear separation in 813C values from the epifaunal species, and has consistently lower values

414

E. THOMAS & N. J. SHACKLETON

%..., ,..-0 0

~E ~/)

e.-

0

0

,,~..

oo

r

O

o

~2

o

o

o

.

.

.

.

i

;

9

9

,

i

.

.

.

.

o

,=,

~

. . . .

i

. . . .

o

= ~

689 690 525 527

. ,,..~

/I

N

~9 o0,1' .~ E~

Age, Ma

Age, Ma

~ ~

PALEOCENE--EOCENE BENTHIC FORAMINIFERALEXTINCTION

415

03

r

~o

u

o

9

.

.

!

9

,

.

|

9

.

.

o o od

m.. "7'

<1

o

od

od

r

!-

o

"7,

o "

o d

A

o

o

A o

)

u'2.

c~

. . . .

i

. . . .

i

. . . .

v ,

,

9 ,~

. . . .

~ ~ . . . . . . o o c~

;~

. . . .

&

. . . .

~,

. . . .

o

Depth, mbsf

Depth, mbsf o Cibicidoides

spp.

9 Bulimina s p p , 9 T. selmensis t= O. umbonatus

~

A

N. truempyi

o

Gavetinetta Lenticulina

spp. spp.

Fig. 9. Oxygen and carbon isotope value for selected benthic foraminiferal species, Sites 689 and 690. Open symbols indicate species considered infaunal because of their morphological resemblance to Recent infaunal taxa (Corliss & Chen 1988; Kaiho 1988, 1991, 1994a; Thomas 1990a); closed symbols show epifaunal species.

than N. truempyi (Fig. 9). This is not so at Site 689, and Lenticulina spp. values are commonly higher than those of N. truempyi. None of the other species in the same samples shows a similar pattern, so that this difference between sites is probably not due to difference in diagenesis. It has been proposed that greater separation in carbon isotopic values of infaunal and epifaunal species results from a higher flux of organic matter to the sea floor (Woodruff & Savin 1985; Zahn et al. 1986). With the exception of the interval of anomalously low 813C values, the carbon isotopic values of Lenticulina spp. are overall lower at Site 690 than at Site 689, thus suggesting a lesser flux of organic matter at the latter site. This is unexpected, because at shallower sites a greater flux of organic matter arrives at the seafloor than at deeper sites at equal productivity (e.g. Berger et al. 1994). Higher fluxes of organic matter to the deeper site thus suggest higher productivity at that site. Higher sedimentation rates at Site 690 might be seen as supportive of higher productivity at that site, but

might as well have been caused by admixture of fine terrigenous material (Barker et al. 1988). We can not at present explain why there should have been higher productivity at Site 690, because the sites are very close. Possibly, the pattern of deep-water circulation around Maud Rise caused upwelling at Site 690, on the southwestern flank, but not at Site 689, closer to the top. At present, waters well up along the southwestern flank of the rise (Comiso & Gordon 1982). In the presence of large stratigraphic variations in isotopic values, bioturbation may disturb the signal easily, and we need better age control at Site 689 before we can decide whether the difference in benthic infaunal carbon isotope values at the sites reflects a real environmental difference. Lenticulina spp. constitute the only group present at both Maud Rise sites over the extinction interval. Comparison of the isotope data of this taxon at both sites (Fig. 9) shows that the excursion had a similar magnitude in the bottom waters of the two sites. Oxygen isotope values at both sites

416

E. T H O M A S 8~; N. J. S H A C K L E T O N

were not significantly different at both sites before, during or after the excursion.

3.0

A&

2.0

Isotope excursions

Extreme excursions in carbon and oxygen isotopic values are present in benthic and planktonic foraminiferal records from the Maud Rise sites, and in benthic and bulk records from the Walvis Ridge sites (Figs 3, 11,12,13). Our data for Site 690 agree with those of Kennett & Stott (1990, 1991) and Stott et al. (1996). The benthic foraminiferal extinction was synchronous with a very large, negative, short-term excursion of carbon and oxygen isotopes in planktonic and benthic foraminifera and bulk carbonate (Shackleton & Hall 1990). The isotope excursions reached peak negative values in a few thousand years, and values returned to pre-excursion levels more gradually, in about 50 000 years as estimated at Site 690. We used the isotope excursion to construct the timescale (see above; Fig. 3), and therefore we assumed that the duration was the same at all 4 sites. The carbon isotope excursion was about -2%~ for benthic foraminifera at Walvis Ridge and /Vlaud Rise, about -4%~ for planktonic foraminifera at Maud Rise, leading to decreased vertical gradients. Bulk carbon isotope values decreased by about 2%o at the Walvis Ridge sites. The oxygen isotope excursion was about -1.5%o for benthic foraminifera at Walvis R i d g e ( 1 6 0 0 - 3 4 0 0 m palaeodepth) and Maud Rise ( l 1 0 0 - 1 9 0 0 m palaeodepth), about - 1%~ for planktonic foraminifera at Maud Rise. Tfiere are no significant differences in the oxygen isotope values of benthic foraminifera at Sites 689, 690, 525 and 527 over the full interval studied, suggesting little stratification of the water masses at the sites before, during and after the isotope excursions. This observation suggests that intermediate to deep watermasses were poorly stratified, from whatever source they were derived. Comparison of the planktonic foraminiferal data from Sites 689 and 690 (Fig. 13) shows that oxygen isotopic values were very close throughout the studied sequence, as expected for sites in such close proximity. Possibly the carbon isotope values of planktonic foraminifera were more positive at Site 690, in support of the hypothesis of higher productivity at that site, as outlined above.

Discussion of faunas, isotopes and palaeoenvironments An investigation of the biogeography of the benthic foraminiferal faunas after the extinction is difficult because many stratigraphical sections have

oa

o o

2.5 o

A ~t~XAai~

Z~

1.5

~ , 0 +.

o 1.0 O- 0.5 ~

-13.5

§

,, ,"

0.0 Z~

A

~e

Z~

k

-1.0 -1.5 -2.0 , -1.5

.

.

-1.0

.

.

.

.

.

.

.

-0.5

.

.

0.0

.

i

0.5

.

9

'."."

1.0

~180, %o o

Cibicidoides spp.

9 Bulimina

spp.

9 Lenticulina

spp.

9 T. selmensis +

O. umbonatus N. tmernpyi

o

G beccariiformis

Fig. 10. Scatter plot of ~13C versus ~180 values of all benthic foraminiferal data for Sites 689 and 690. Presumed infaunal species (Bulimina spp., T. selmensis, Lenticulina spp.) have closed symbols, presumed epifaunal species ( Cibicidoides spp., N. truempyi, Gavelinella spp.) have open symbols. O. umbonatus, which has an epifaunal morphology according to Corliss & Chen (1988) and Kaiho (1988, 1991) is indicated by +, to show its presence within the field of infaunal species.

hiatuses (Aubry et al. 1996). We can therefore only be confident about our time correlations if we compare faunas from the interval of the large, short-lived, negative ~St3C-excursion. The benthic foraminiferal extinction shows biogeographical variation between the Walvis Ridge and the Maud Rise sites, which were in the same ocean basin and at similar longitude, separated by about 30 ~ of latitude. The pre-extinction faunas were largely cosmopolitan, with relatively minor differences with location as well as depth (e.g. Kaiho 1988, 1991), but post-extinction faunas show very wide differences in composition. At our sites biogeographical differences were more important than depth differences. At Maud Rise, the extinction was followed by an increase in relative abundance of species presumed to live infaunally (Corliss & Chen 1988). Such an increase most probably reflects an increase of the flux of particulate organic matter to the sea floor (e.g. Lutze & Coulbourn 1984; Pedersen et al. 1988; Hermelin & Shimmield 1990; Hermelin

417

PALEOCENE-EOCENE BENTHIC FORAMINIFERAL EXTINCTION 1~130, %0 -2

55.1j

-1

0

1

(~180, %0 2

3

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

4 oo

1

5

0.5

-0.5

~o ~,- ; ~, oo

55.51

L

ooo

]

.oo oo o o o oo

~

.~

o

55.7 ~ 55 8 4.

~

55

o

ooo

55.5-J

o

~

0~

~

7 -1

o o

~

o

oo

8

'

"

. . . .

0 1 2 ' ................... ,

. . . .

,

. . . .

,

3

. . . .

,

. . . .

4 5 ', . . . . . . . .

,

0.5 55.1

. . . .

55.2

55.3

55.3

55.40 55.5

~

o

A. mckannai, e

9

M.convex8

A. r~.,~nai,,~

o

~180, %0

55.2

3>

o

o

~

~130, %0

,

~,"

~,.~oOo

' ~ . .

55

-1 -2 55.1 ' . . . . . . . . .

-2.5

. . . . . . . .

o

L

554

t~

-1.5

55.1 . . . . . . . . . .

....

'~ oo

55.4-

o

55.5 -

-0.5 ~ . . . .

-1.5 ' . . . .

-2.5 '

~ 1 7'~6

55.6

55.6 -

55.7

55.7

55.8

55.8

/

P o

-

-.= A

enticulina

0

M. convexa A. rnckannai, d

o

A. rnckannai, e

i

Fig. 11. Carbon and oxygen isotope data for benthic and planktonic foraminifera, Sites 689 and 690. Included are data from Stott et al. (1990), Kennett & Stott (1990), and our data.

1992; Rathburn & Corliss 1994), rather than a decrease in dissolved oxygen. This faunal evidence appears to conflict with the carbon isotope data, which show decreased surface-to-deep gradients, suggesting lowered productivity just after the extinction (Kennett & Stott 1991; Fig. ll). We can try to reconcile faunal and isotope data by speculating that the decreased gradients resulted from upwelling of deep, nutrient-rich waters as a result of a changes in deep-water circulation. The upwelling could have caused increased productivity and isotopically lighter total dissolved carbonate in the surface waters. This mechanism can not explain the short-term carbon isotope

excursion, and can be only a part of the explanation for the depressed gradients at Sites 689 and 690. We do not suggest that productivity increased worldwide, only that it increased at Maud Rise, and possibly at other sites at high latitudes, or on continental margins, or close to the equator. We suggest that all these sites may have been characterized by increased deposition of organic matter (thus probably increased surface productivity) after the extinction. In contrast, benthic foraminifera data suggest that the flux of organic matter to the seafloor decreased after the extinction at the Walvis Ridge sites, in line with suggestions that productivity

418

E. THOMAS & N. J. SHACKLETON (~130, %o -2 .1 55.3 t .........

o

~3180, %0

1

, ...............,

2 ,

55.4

C

0.0 .

55.4

55.5

~

0.5

3

.

.

.

i

-0: .

.

.

.

i

-1.0 .

.

.

.

i

-1.5 .

.

.

.

|

-2.0 .

.

.

.

i

p

55.5"

9

55.6 -

55.6

=-

525, N. truempyi

9

527. N. truempyi I

[ --.-D---- 527,bulk

J

55.7

55.7

Fig. 12. Carbon and oxygen isotope data for benthic foraminifera and bulk carbonate, Sites 525 and 527. Included are data from Shackleton et al. (1984a), Shackleton & Hall (1984), and this paper.

decreased globally at the time (e.g. Corfield & Shackleton 1988; Corfield 1993). Thomas (1990) speculated that this difference might have been apparent only, and the result of the fact that many researchers study such a large size fraction (> 125 ~tm) that the abundant presence of the small, biserial species would go undetected. Data presented in this study, however, clearly show that the difference is real, not the result of difference in size-fraction studied, nor of the presence of hiatuses, since these different faunas coexisted with the short-lived carbon isotope excursion. Studies from many locations in the Atlantic Ocean have described post-extinction faunas resembling those at the Walvis Ridge sites, with common to abundant abyssaminid species and N. truempyi (Clark & Wright 1984; Tjalsma & Lohmann 1984; Katz & Miller 1991; Mtiller-Merz & Oberh~insli 1991; Oberh~asli et al. 1991; Pak & Miller 1992); similar faunas have also been described from land sections in Spain (Zumaya: G. Keller pers. comm. 1994) and Egypt (Speijer 1994). We cannot be certain that these faunas are indeed coeval with the immediate post-extinction faunas because of the lack of isotopic data or lowresolution sampling. At Site 549 in the Bay of Biscay (Reynolds 1992 MSc Thesis, op. cit.) biserial and triserial species increased in relative abundance after the extinction, as at Maud Rise. In the land section at Caravaca (Spain), a similar faunal pattern was observed (G. Keller pers. comm. 1994). Faunas from some sites in the Indian Ocean show a pattern similar to that at the Atlantic sites,

but we do not know whether we see the immediate post-extinction faunas because of lack of isotope data, poor recovery, or low time resolution. At Site 762 (palaeodepth 1000--1500 m) the postextinction faunas were dominated by N. truempyi, Cibicidoides spp., and abyssaminid species; T. selmensis and A. aragonensis were present (Thomas, unpub, data; R. Nomura pers. comm. 1994). At Site 747 (Kerguelen Plateau, Indian Ocean; palaeodepth 2000-3000 m) post-extinction faunas were dominated by N. truempyi , but at the shallower Site 748 (palaeodepth 600-2000 m) Stilostomella spp. and Lenticulina spp. dominated (Mackensen & Berggren 1992). At Site 752 (palaeodepth 500-1000 m), post-extinction faunas had common Anomalinoides capitatus/danicus, N. truempyi and Cibicidoides spp. (Nomura 1991; R. Nomura pets. comm. 1994). At Site 738 (Indian Ocean, Kerguelen Plateau; Fig. 1) recovery was poor, but the short-term isotope excursion occurred in a recovered interval and is similar in magnitude to that at Maud Rise (Lu & Keller 1993). The benthic foraminiferal extinction occurred in the lower part of an interval of dissolution as at Walvis Ridge, and post-extinction faunas coeval with the carbon isotope negative peak had high relative abundances of Tappanina selmensis and small buliminids (Thomas, unpub, data) as at Maud Rise. Later faunas became dominated by N. truempyi (R. Nomura pers. comm. 1994). At equatorial Pacific Site 865 (palaeodepth about 1200, palaeolatitude 2~ Bralower et al. 1995a) post-extinction faunas coeval with the short-lived 813C excursion lack N. truempyi, have

419

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

~130, %o -2 55.1

! ....

-1 ' ....

~180, %0

0

1

2

3

4

' ....

' ....

' ....

' ....

' .....

55.2

~o

old O 9 o0 D O 9

55.4

55.5

55.1

o 9D

55.3

r

0,5

9 ePO 9 OdD 9 c~

9 oO

9 55.7

o

.......

ql

55.3

O 9

~

. . . .

9 00 e~

55.4'

55.5

-2.!

-1.5

'

55.2

oO 55.6

. . . .

9 >

o 9 O~149176

'

-0.5

o 'b o ~ 1 7 6 1 7 6 1 4 9 9 uP 9 oOe ~ecOo

9 9 9

55.8

55.6

ooO

DO 55.7

9

g

55.8

Fig. 13. Carbon and oxygen isotope data for planktonic foraminifera, Sites 689 and 690. Open symbols are data from Site 690, closed symbols are data from Site 689. Circles, Acarinina mckannai; triangles, 'Morozovella' convexa. Included are data from Stott et al. (1990), Kennett & Stott (1990), and this paper.

common Cibicidoides spp. and Bulimina semicostata, and high relative abundances of bi- and triserial species (Thomas, unpub, data). At Pacific Site 577 (palaeodepth 1500m) post-extinction faunas show increased relative abundance of Bulimina semicostata as well as N. truempyi (Miller et al. 1987b; Pak & Miller 1992), in contrast to the faunal pattern at Site 865, but there may be a hiatus in the immediate post-extinction interval (Corfield & Cartlidge 1992b; Berggren et al. 1995). The benthic extinction thus was followed by a period of varying faunas at different locations. At all locations T. selmensis and A. aragonensis are typical for the post-extinction period at bathyal to abyssal depths. These species might probably be seen as opportunists, quickly filling habitats emptied during the extinction. Species that became extinct included common species thought to have lived epifaunally (e.g. Neoeponides spp. and Gavelinella beccariiformis), many agglutinant species (e.g. Dorothia oxycona, Tritaxia paleocenica), but also infaunal species such as B. thanetensis and B. delicatulus. At some locations infaunal morphotypes increased in relative abundance for a short period (50 000 years), but at many sites in the Atlantic, Pacific and Indian Oceans epifaunal morphotypes increased in abundance, and infaunal morphotypes were replaced by other infaunal morphotypes (e.g. Bulimina velascoensis and Aragonia velascoensis before the extinction, Bulimina semicostata and Aragonia aragonensis after the extinction at Pacific sites; Miller et al. 1987b; E. Thomas, unpub, data).

It thus appears to be too simplistic to relate the benthic foraminiferal extinction only to increased temperatures with concomitant decreased oxygen levels, although these probably were involved (Thomas 1990; Kaiho 1991). A strong increase in volume of deep to intermediate waters derived from subtropical latitudes rather than from high southern latitudes (as invoked by many authors, including Thomas 1989, 1990, 1992; Kennett & Stott 1990, 1991; Lu & Keller 1993; Zachos et al. 1993) could have caused the observed patterns partially by direct and indirect effects on the deep sea benthic faunas: patterns of aging of intermediate and deep waters, and thus their nutrient and oxygen levels, would change as well as temperature. Probably as important for deep sea faunas, a change in direction of deep water circulation might influence the locations where deep, nutrient-rich waters welled up to the surface, and thus where high surface productivity occurred. If ~513C values of deep sea benthic foraminifera during the maximum-excursion were well known from various localities, one might attempt to reconstruct the direction of deep water circulation (e.g. Pak & Miller 1992; Zachos et al. 1993), lighter values indicating the presence of 'older' deep to intermediate water masses. Lu & Keller (1993) thus argue that the fact that the benthic carbon isotopic excursion was of smaller magnitude in the Pacific (Pak & Miller 1992) than at Maud Rise (Kennett & Stott 1991) indicates deepwater circulation from Pacific low latitudes to the Antarctic. We do not think that this argument is valid, however, because of the presence of hiatuses

420

E. THOMAS 8s N. J. SHACKLETON

at Site 577 (Corfield & Cartlidge 1992b; Berggren et al. 1996), so that we can not be certain that the

observed values reflect the maximum excursion. New isotope data from equatorial Pacific Site 865 suggest that deep waters might have been forming at that location just after the extinction (Bralower et al. 1995b). It is not clear whether a change in deep-water circulation could be related to decreased levels of dissolved oxygen in shallower marine areas on continental margins (e.g. North Sea Basin: King 1989; Charnock & Jones 1990; New Jersey continental margin: Gibson et al. 1992; southwest Pacific: Moore 1988; Kaiho 1994b; the southern edge of Tethys: Speijer 1994). There might be a link through locally increased productivity (as suggested by Speijer 1994), or decreased levels of dissolved oxygen might have resulted from the high temperatures. The rapid isotope excursions occurred against the background of longer term events. In the middle Paleocene, bulk carbonate ~13C values reached a maximum for the Cenozoic (e.g. Shackleton 1986, 1987), which was not higher than more long-term values during the Late Cretaceous (Corfield et al. 1991; Corfield & Cartlidge 1992b). This period of high 613C values could be seen as reflecting return to higher oceanic productivity (Shackleton et al. 1985; Shackleton 1986, 1987) after the biosphere recovered from the end-Cretaceous extinction, especially since oceanic ~13Cgradients were very high at the time (Shackleton et al. 1985). This explanation works only if the organic carbon produced was buried in the lithosphere, because ~513C changes that persist over periods much longer than the residence time of carbon in the ocean can arise only as a result of long-term changes in the global rate of storage of carbon in organic matter rather than in carbonates (Berger & Vincent 1986; Shackleton 1986, 1987; Corfield & Cartlidge 1992b). It is not obvious where the organic matter required to be deposited in the Paleocene was stored, but extensive organic matter-rich deposits of that age occur in southwestern Pacific (Moore 1988), the North Sea region, and possibly oil shales in the continental United States. Carbon isotopic values decreased over a few million years from the middle Paleocene until the early Eocene (Shackleton 1987; Corfield & Cartlidge 1992b). The long-term decreasing values of 813C have been tentatively explained by increased erosion of organic carbon rich sediments from the Himalyan orogen (Beck et al. 1995b). This long-term decrease occurred at a time when oxygen isotope data suggest increasing temperatures of global deep waters as well as surface waters at high latitudes (Shackleton 1986; Corfield & Cartlidge 1992b; Zachos et al. 1994). The high

temperatures have been explained by higher rates of CO 2 emissions from hydrothermal systems (Owen & Rea 1985; Rea et al. 1990) or from plateau basalts (Eldholm & Thomas 1993), especially because preliminary results from studies of chemical proxies for palaeo-CO2 levels suggest that early Eocene pCO 2 levels were 2 to 6 times present values (Freeman & Hayes 1992; Cerling 1991). What caused the rapid isotope excursions in the latest Paleocene to be superimposed on these longterm trends? Specifically, did sudden events (e.g. emission of large amounts of volcanic CO2) trigger the rapid isotope excursions, or did gradual changes cross a threshold-value (Zachos et al. 1994)? Planktonic foraminiferal data from Sites 689 and 690 (Stott e t a l . 1990, 1996; this paper) and 738 (Lu & Keller 1993) clearly demonstrate that lowlatitude, keeled planktonic foraminifera reached high latitudes (>60~ before the benthic foraminiferal extinction; extensive warming of high latitude surface waters thus preceded the isotope anomalies. Note that ~5~80 values of keeled planktonics at Site 690 are extremely low, while 613C values of the same specimens have not reached peak values, suggesting that temperature increase slightly pre-dated the 613C excursion (Fig. 11). The rapid oxygen isotope excursion at a time when polar ice-sheets were absent or insignificant should probably be explained by a rapid increase in temperature of high latitude surface waters and deep waters worldwide, by 4--6~ Such an ocean-wide increase in temperature would have the effect of raising seaievel by about 5-6 m only (Varekamp, pers. comm. 1994), which would not be expected to be noticeable in benthic foraminiferal faunas. Speijer (1994) did not observe a clear sea level signal in neritic sequences in the eastern Mediterranean. The large, short-term carbon isotope excursion occurred in deep as well as surface waters and in the continental record (Kennett & Stott 1990, 1991; Pak & Miller 1992; Lu & Keller 1993; Stott et al. 1996), and was associated with increased CaCO 3dissolution over a wide depth and geographical range in the deep oceans (Bay of Biscay: Reynolds 1992 MSc Thesis, op. cit.; Stott et aL 1996; Site 738: Lu & Keller 1993; Walvis Ridge: this paper). A large, negative shift in 513C composition of total dissolved inorganic carbon as well as increased dissolution and shallowing of the calcium carbonate compensation depth in the oceans could result from a transfer of isotopically light material from the lithosphere or biosphere into the oceanatmosphere system (Shackleton 1977; Broecker & Peng 1984). The shape of the excursion (rapid start, more gradual return to almost pre-excursion values) is as modelled for disturbance of the ocean-

PALEOCENE--EOCENE BENTHIC FORAMINIFERALEXTINCTION atmosphere sysl/em by transfer of large amounts of isotopically light carbon into the atmosphere (e.g. Kasting & Walker 1993). The rapidity of the initiation of the excursion (a few thousand years) and its short duration suggest that this transfer is not likely to have been caused by changes in ratio of organic carbon to carbonate deposition or erosion (Shackleton 1987; Corfield & Cartlidge 1992b; Walker 1993). One can estimate how much carbon from the terrestrial biosphere must be transferred to the oceans to explain the short-term carbon shift (Shackleton 1977), if one makes assumptions regarding the size of oceanic and atmospheric reservoirs, and the average isotopic composition of plant material, and if one knew the average, global size of the carbon isotope excursion. Estimates of the average isotopic composition of plant material can be made, assuming that only C3 plants were around in the late Paleocene (Ceding 1993). The amplitude of the carbon isotope excursion is not well constrained, because the event was short and many sections have low sedimentation rates and hiatuses (Aubry et al. 1996); a value between 1 and 2%~, possibly close to 2%~ seems most probable from present observations (Kennett & Stott 1991; Pak & Miller 1992; Stott et al. 1996; Bralower et al. 1995b). The sizes of atmospheric and oceanic reservoirs are less easy to estimate: the atmospheric reservoir may well have been several times larger than today, which would have led to a larger oceanic reservoir. On the other hand, higher deep ocean temperatures may have caused a decrease in size of the oceanic reservoir. In addition, the size of the oceanic reservoir may have been influenced by the deep sea ventilation rate (Broecker & Takahashi 1984; Thierstein 1989), leading to decreased oceanic reservoir sizes at decreased ventilation rates. Assumption of various values for these parameters, however, indicates that the excursion was so large that transfer of terrestrial biosphere is very unlikely as the cause: such a transfer would require destruction of continental biosphere much larger than that during the end-Cretaceous extinction, which did not occur according to the fossil record. It is difficult to explain fully such a large shift by rapid emission into the atmosphere of volcanogenic CO 2, derived from the subaerial plateau basalt eruptions of the North Atlantic Volcanic Province (White 1989; White & MacKenzie 1989). The minimum isotopic composition of mid-oceanic ridge volcanic CO 2 is only of the order of -6 to -7%~ (Eldholm & Thomas 1993), so that very large volumes would be necessary. Calculations of the size of plateau-basalt eruptions and the eruption rate required to cause the observed carbon isotope shift have the same problems as the estimates of

421

biomass required. Average rates of basalt eruption during Chron 24r (several times 1013g CO 2 per year; see Eldholm & Thomas 1993) were certainly too low to cause the excursion. Maximum eruption rates, however, could have been much higher than that (e.g. White & MacKenzie 1989), and an eruption rate of about 100 times the average rate, persisting over a few thousand years, could have been sufficient (Eldholm & Thomas 1993). The timing of maximum volcanic activity might be seen as problematic for this tentative explanation. The major ash falls, including the most explosive phase of volcanism (ashes -17, +19), clearly postdated the benthic foraminiferal extinction at the Bay of Biscay sites (Knox & Morton 1983, 1988; Backman et al. 1984; Aubry et al. 1986; Knox 1990; Eldholm & Thomas 1993; Berggren & Aubry 1996). We would, however, not necessarily expect high rates of CO2-effusion resulting from the most explosive volcanic activity, because basaltic (thus less explosive) magmas have the highest concentrations of CO 2 (Gerlach & Taylor 1990). Ashes and basalt flows are present in sediments from nannofossil zone NP9 and the dinoflagellate Apectodinium hyperacanthum zone (Boulter & Manum 1989; Knox 1990; Ali et al. 1992), and Berggren & Aubry (1996) tentatively place the initiation of phase 2 volcanic activity in the lower part of zone NP9. The extinction may be very close in age to the change of volcanic activity from western to eastern Greenland. This change may have triggered the break-up of the continent along the East Greenland margin with ensuing volcanism over very large areas adjacent to the line of break-up, which occurred close the chron C25n/C24r boundary (Larsen et al. 1992). We can speculate about the following scenario: high latitude surface ocean temperatures increased, possibly as a result of increased atmospheric pCO 2. These high temperatures may have been increased even more by high CO 2 levels from massive plateau basalt eruptions at the initiation of continental break-up, and they caused a decrease in density of surface waters at high latitudes so that a low density lid formed. Therefore formation of deep to intermediate waters at these latitudes decreased in volume, so that waters sinking at subtropical latitudes could increase in volume. The circulation changes caused changes in the location where deep, nutrient-rich waters could well up to the surface, thus influencing productivity of planktonic organisms, and extinction of deep sea benthic foraminifera. The latter extinctions were thus caused by a combination of increased temperature and concomitant decrease in dissolved oxygen concentration, as well as changes in local productivity - either increased or decreased productivity, depending upon location.

422

E. THOMAS t~ N. J. SHACKLETON

In this scenario, dominant deep water formation at subtropical latitudes was probably a short-term phenomenon, lasting not much longer than 50 000 to maybe 100 000 years, because of the selflimiting effect of deep water circulation from lower to higher latitudes (Pak & Miller 1992; Lu & Keller 1993). Such a circulation pattern would increase heat transport to high latitudes, thus further decreasing thermal gradients; but very low gradients would mean decreased speed of atmospheric as well as oceanic currents (as observed in the decreased intensity of dust transport at the time of the benthic extinction; Miller et al. 1987; Rea et al. 1990; Hovan & Rea 1992). This decreased transport would then cause cooling of high latitudes (Walker & Sloan 1992; Sloan et al. 1995), so that deep to intermediate waters could again form there. There are problems with this simplistic and speculative scenario. We do not think that volcanic CO 2 input can cause the short-term negative 813C anomaly by itself. It is not clear whether the total dissolved CO 2 content of the oceans would increase or decrease under this scenario: increased temperatures would lead to a decreased reservoir size, but increased atmospheric pCO 2 would lead to an increased oceanic reservoir (Broecker & Takahashi 1984; Walker 1993). Degassing of the ocean when it was warming up would be expected to lead to increased oceanic 813C values, contrary to observations. The increased CaCO 3 dissolution suggests increasing dissolved carbonate levels in the oceans (Broecker & Peng 1984; Kasting & Walker 1993). Stott (1992) suggested, on the basis of changes in isotopic composition of organic material in planktonic foraminifera, that atmospheric pCO 2 decreased sharply during the isotope anomalies, possibly as a result of oceanic CO 2 uptake. But data on carbon isotopic composition of organic marine material cannot simply be interpreted as resulting from changes in atmospheric pCO2, especially for times when other oceanic parameters (temperature, possible alkalinity) were changing (Goericke & Fry 1994; Hinga et al. 1994). In addition, we do not know the full effects of the speculative scenario on oceanic productivity and we do not know the average, global effects on productivity changes that vary from place to place. At the present time, we do not understand the extreme carbon isotope anomalies in the late Paleocene to early Eocene ocean-atmosphere system. More data on the anomalies, including a better understanding of their maximum extent and geographical variability, may lead to a better under-

standing of this major disturbance of the global carbon cycle, and thus of the functioning of the carbon cycle itself.

Conclusions 1. The deep sea benthic foraminiferal extinction in the latest Paleocene was coeval with short-term, extreme, negative excursions of 513C in planktonic and benthic foraminifera and bulk carbonate, in the South Atlantic and the Weddell Sea, over a range of palaeodepths from 1100-3400 m. 2. The benthic carbon isotope excursion had the same magnitude in both areas, about -2%o. The excursion was a very rapid (beginning in a few thousand years, duration about 50 000 years), and detailed sampling is necessary to evaluate differences in its magnitude in different areas. 3. We do not have enough geographical coverage of the minimum benthic 513C values to use these for reconstruction of deep water circulation. 4. Oxygen isotope data indicate rapid warming of high latitude surface waters as well as global deep waters. It is not clear whether this warming was caused by a change in dominant source region of deep to intermediate waters from high to subtropical latitudes, or by increased surface water temperatures at high latitudes. 5. The biogeography of benthic foraminiferal assemblages after the extinction suggests that a change in dominant source area of deep water formation was involved. 6. The short-term carbon isotope excursion was so large that it probably could not have been caused by transfer of terrestrial biomass into the oceanatmosphere system; it was so rapid that it probably could not have been caused by a change in deposition or erosion rates of carbon in carbonate as compared to carbonate in organic matter. 7. We cannot explain the short-term negative carbon isotope excursion, which probably is a complex signal, having global as well as local (productivity) components. Samples were supplied by the Ocean Drilling Project and we thank the ODP curatorial staff at the Gulf Coast and East Coast Core Repositories for their assistance. We thank Joop Varekamp, Mimi Katz, Edith Mtiller-Merz, Hedi Oberhansli, Jim Zachos and Lisa Sloan for discussion of the Paleocene-Eocene event. This paper improved considerably thanks to the reviews by Marie-Pierre Aubry, Bill Berggren, Richard Corfield, Mimi Katz and Birger Schmitz. This paper is a contribution to IUGS-IGCP Project 308 and is University of Cambridge Department of Earth Sciences contribution no. 3800.

423

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION R e f e r e n c e s

ALI, A. E., KING, C. & HAILWOOD, E. A. 1992. Magnetostratigraphic calibration of early Eocene depositional sequences in the southern North Sea Basin. In: HAILWOOD, E. A. & KIDD, R. B. (eds) High Resolution Stratigraphy. Geological Society, London, Special Publication, 70, 99-125. ALTENBACH, A. V. & SARNTHEIN,M. 1989. Productivity record in benthic foraminifera. In: BERGER, W. H., SMETACEK, V. S. & WEFER, G. (eds) Productivity of the Ocean: Present and Past. Life Science Research Reports, 44, 255-269. AUBRY, M.-P. 1992. Paleogene nannofossils from the Kerguelen Plateau, Leg 120. Proceedings of the Ocean Drilling Program, Scientific Results, 104, 471-491. , HAILWOOD, E. A. & TOWNSEND, H. A. 1986. Magnetic and calcareous nannofossil stratigraphy of lower Paleogene formations of the Hampshire and London Basins. Journal of the Geological Society, London, 143, 729-735 , BERGGREN, W. A., KENT, D. V., FLYNN, J. J., KLITGORD, K. D., OBRADOVICH,J. D. & PROTHERO, D. R. 1988. Paleogene chronology: an integrated approach. Paleoceanography, 3, 707-742. , --, STOTT, L. D. & SINHA, A. 1996. The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology. This

N. A., LAWRENCE, R. D. & KHAN, A. 1995a. Stratigraphic evidence for an early collision between NW India and Asia. Nature, 373, 55-58. --, OLSON, T. L. & KHAN, A. M. 1995b. Organic carbon exhumation and global warming during the early Himalayan collision. Geology, 23, 387-390. BECKMANN, J.-P. 1960. Distribution of benthic foraminifera at the Cretaceous-Tertiary boundary of Trinidad (West Indies). In: Report 21st session

International Geological Copenhagen. 5, 57-69.

Constraints on the Ocean's Role in Global Change. Springer-Verlag, New York, 385--411. BERGGREN, W. A. 1992. Paleogene planktonic foraminiferal magnetobiostratigraphy of the southern Kerguelen Plateau (Sites 747-749).

Proceedings of the Ocean Drilling Program, Scientific Results, 120, 551-568.

volume.

Palaeoclimatology, Palaeoecology,

57, 43-59. - 1986b. Accumulation patterns of Tertiary calcareous nannofossils around extinctions. Geologische Rundschau, 75, 185-196. --, MORTON, A. C., ROBERTS, D. G., BROWN, S., KRUMSIECK, K. & MACINTYRE, R. M. 1984. Geochronology of the lower Eocene and upper Paleocene sequences of Leg 81. Initial Reports of the Deep Sea Drilling Project, 81, 877-882. BARKER, P. F., KENNETT, J. P. ETAL. 1988. Proceedings of the Ocean Drilling Program, Initial Reports, 113. BARRERA, E. & HUBER, B. 1991. Paleogene and early Neogene oceanography of the southern Indian Ocean: Leg 119 foraminifer stable isotope results.

Proceedings of the Ocean Drilling Program, Scientific Results, 119, 693-717. --,

--, SAVtN, S. M. & WEBB, R N. 1987. Antarctic marine temperatures: late Campanian through early Paleocene. Paleoceanography, 2, 21--48. BARRON, E. J. 1987. Eocene equator-to-pole surface ocean temperatures: a significant climate problem? Paleoceanography, 2, 729-740. - & PETERSON, W. H. 1991. The Cenozoic ocean circulation based on ocean General Circulation Model results. Palaeogeography, Palaeoclimatology, Palaeoecology, 83, 1-18. BECK, R. A., BURBANK, D. W., SERCOMBE,W. J., RILEY, G. W., BARNDT,J. K., BERRY,J. R., AFZAL, J., KHAN, A., JURGEN, H., METJE, J., CHEEMA, A., SHAFIQUE,

Norden,

BERGER,W. H. & VINCENT,E. 1986. Deep-sea carbonates: reading the carbon isotope signal. Geologische Rundschau, 75, 249-269. --, HERGUERA,J. C., LANGE, C. B. & SCHNEIDER, R. 1994. Paleoproductivity: flux proxies versus nutrient proxies and other problems concerning the Quaternary productivity record. In: ZAHN, R., PEDERSEN, T. F., KAMINSKI, M. A. & LABEYRIE, L. (eds) Carbon Cycling in the Glacial Ocean:

BACKMAN, J. 1986a.Late Paleocene to middle Eocene calcareous nannofossil biochronology from the Shatsky Rise, Walvis Ridge and Italy. Palaeo-

geography,

Congress,

--,

--,

& AUBRY, M.-P. 1996. A late Paleocene-early Eocene NW European and North Sea magnetobiostratigraphic correlation network. This volume. & MILLER, K. G. 1988. Paleogene planktonic foraminiferal biostratigraphy and magnetobiochronology. Micropalaeontology, 34, 362-380. & 1989. Cenozoic bathyal and abyssal calcareous benthic foraminiferal zonation. Micropaleontology, 35, 308-320. KENT, D. V., FLYNN, J. J. & VAN COUVERING,J. A. 1985. Cenozoic geochronology. Geological Society of America, Bulletin, 96, 1407-1418. --, SWISHER, C. C. I I I & AUBRY, M.-P 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. a., KENT, D. V., AUBRY, M.-P & HARDENBOL, J. (eds)

Geochronology, Time Scales and Stratigraphic Correlations: Framework for an Historical Geology. SEPM Special Publication, 54. BOERSMA, A. 1984a. Cretaceous-Tertiary planktonic foraminifera from the south eastern Atlantic, Walvis Ridge area, Deep Sea Drilling Project Leg 74.

Initial Reports of the Deep Sea Drilling Project,

--,

74, 501-524. 1984b. Oligocene and other Tertiary benthic foraminifera from a depth traverse down Walvis Ridge, Deep Sea Drilling Project Leg 74. Initial Reports of the Deep Sea Drilling Project, 75, 1273-1300. & PREMOLI-SILVA,I. 1991. Distribution of Paleogene planktonic foraminifera - analogies with the Recent? Palaeogeography, Palaeoclimatology, Palaeoecology, 83, 29-48. -& SHACKLETON,N. J. 1987. Atlantic Eocene

424

E. THOMAS • N. J. SHACKLETON

planktonic foraminiferal paleohydrographic indicators and stable isotope stratigraphy. Paleoceanography, 2, 287-331. BOLU, H. M., BECKMANN,J.-P. & SAUNDERS,J. B. 1994.

Benthic Foraminiferal Biostratigraphy of the South Caribbean Region. Cambridge University Press. BOLTOVSKOY,E. & BOLTOVSKOV,D. 1988. Cenozoic deepsea benthic foraminifera: faunal turnovers and paleobiographic differences. Revue de Micropaldontologie, 31, 67-84. & 1989. Paleocene-Pleistocene benthic foraminiferal evidence of major paleoceanographic events in the eastern South Atlantic (DSDP Site 525, Walvis Ridge). Marine Micropaleontology, 14, 283-316. BOULTER, M. C. & MANUM, S. B. 1989. The Brito-Arctic Igneous Province Flora around the Paleocene/ Eocene boundary. Proceedings of the Ocean Drilling Program, Scientific Results, 104, 663-680. BRAGA, G., DEBIASE,R., GRUNIG, A. & PROTO-DECIMA,F. 1975. Foraminiferi bentonici del Paleocene e dell'Eocene della Sezione Posagno. Schweizerische Paliiontologische Abhandlungen, 97, 85-111. BRALOWER, T. J., PARROW, M., THOMAS, E. & ZACHOS, J. C. 1995a. Stable isotope stratigraphy of the Paleogene pelagic cap at Site 865, Allison Guyot.

Proceedings of the Ocean Drilling Program, Scientific Results, 143, 581-586. , ZACHOS, J. C., THOMAS, E., PARROW, M., PAULL, C. K., KELLY, D. C., PREMOLI-SILVA,1., SLITER, W. V. & LOHMANN, K. C. 1995b. The Paleogene Paleoceanography of the equatorial Pacific Ocean: stable isotopes recorded at ODP Site 865, Allison Guyot. Paleoceanography, 10, 841-865. BRASS, G. W., SOUTHAM,J. R. & PETERSON,W. H. 1982. Warm saline bottom water in the ancient ocean. Nature, 296, 620-623. BROECKER, W . S . & PENG, T.-H. 1984. The climatechemistry connection. American Geophysical Union Geophysical Monographs, 29, 327-336. - & TAKAHASHI, T. 1984. Is there a tie between atmospheric CO 2 content and ocean circulation?

American Geophysical Union, Geophysical Monographs, 29, 314--326. BUTLER, R. E, GINGERICH, P. & LINDSAY, E. H. 1981. Magnetic polarity stratigraphy and biostratigraphy of Paleocene and lower Eocene continental deposits, Clark's Fork Basin, Wyoming. Journal of Geology, 89, 299-316. , KRAUSE, D. & GINGERICH, P. 1987. Magnetic polarity stratigraphy and biostratigraphy of middlelate Paleocene continental deposits of south-central Montana. Journal of Geology, 95, 647-657. CANDE, S. C. & KENT, D. V. 1992. A new geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 97 (B10), 13917-13951. - & 1995. Revised calibration of the geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research, 100, 6093-6095. CHAMBERLIN, T. C. 1906. On a possible reversal of deepsea circulation and its influence on geologic climates. Journal of Geology, 14, 363-373.

CERLING, Z. E. 1991. Carbon dioxide in the atmosphere: evidence from Cenozoic and Mesozoic paleosols.

American Journal of Science, 291,377-400. CHARNOCK, M. A. & JONES, R. W. 1990. Agglutinated foraminifera from the Palaeogene of the North Sea. In: HEMLEBEN,C., KAMINSK1,M. A., KUHNT, W. & SCOTT, D. B. (eds) Paleoecology, Biostratigraphy,

Paleoceanography and Taxonomy of Agglutinated Foraminifera. Nato Advanced Study Institute Series, Series C, 139-243. CLARK, M. W. & WRIGHT, R. W. 1984. Paleogene benthic foraminifera from the cape and Angola Basins, South Atlantic Ocean. Initial Reports of the Deep Sea Drilling Project, 73, 459-480. COLLINSON,M. E. 1983. Fossil Plants of the London Clay. Palaeontological Association, London, Field Guides to Fossils, 1. COM1SO, J. C. & GORDON, J. A. L. 1987. Recurring polynyas over the Cosmonaut Sea and the Maud Rise. Journal of Geophysical Research, 92, 28192833. CORFmLD, R. M. 1993. Depth habitats and the Palaeocene radiation of the planktonic foraminifera monitored using oxygen and carbon isotopes. In: LEES, D. & EDWARDS, D. (eds) Evolutionary Patterns and Processes. Linnean Society, London, 59-70. -& CARTLIDGE, J. E. 1992a. Isotopic evidence for the depth stratification of fossil and recent Globigerinina; a review. Historical Biology, 5, 37-63. & 1992b. Oceanographic and climatic implications of the Palaeocene carbon isotope maximum. Terra Nova, 4, 443-445. - & SHACg~LETON,N. J. 1988. Productivity change as a control on planktonic foraminiferal evolution after the Cretaceous/Tertiary boundary. Historical Biology, 1, 323-343. , CARTLIDGE, J. E., PREMOLI-SILVA, I. & HOUSLEY, R. A. 1991. Oxygen and carbon isotope stratigraphy of the Palaeogene and Cretaceous limestones in the Bottaccione Gorge and the Contessa Highway sections, Umbria, Italy. Terra Nova, 3, 414-422. CORLISS, B. H. & CHEN, C. 1988. Morphotype patterns of Norwegian deep-sea benthic foraminifera and ecological implications. Geology, 16, 716-719. CROWLEY, T. J. 1991. Past CO 2 changes and tropical sea surface temperatures. Paleoceanography, 6, 387394. CUSHMAN, J. A. 1946. Upper Cretaceousforaminifera of

the Gulf Coastal Region of the United States and adjacent areas. United States Geological Survey Professional Paper, 206. ELDHOLM, O. 1990. Paleogene North Atlantic magmatictectonic events: environmental implications. Memorie della Societg~ Geologica Italiana, 44, 13-28. & THOMAS, E. 1993. Environmental impact of volcanic margin formation. Earth and Planetary Science Letters, 117, 319-329. ESTES, R. & HtrrCHISON, J. H. 1980. Eocene lower vertebrates from Ellesmere island, Canadian Arctic archipelago. Palaeogeography, Palaeoclimatology, Palaeoecology, 30, 325-347. FREEMAN, K. H. & HAYES, J. M. 1992. Fractionation of

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION carbon isotopes by phytoplankton and estimates of ancient CO 2 levels. Global Biogeochemical Cycles, 6, 185-198. GAGE, J. D. & TYLER, R A. 1991. Deep-sea Biology:

a Natural History of Organisms at the Deep-sea Floor. Cambridge University Press. GERLACH, T. M., & TAYLOR,B. E. 1990. Carbon isotope constraints on the degassing of carbon dioxide from Kilauea Volcano. Geochimica et Cosmochimica Acta, 54, 2051-2058. GIBSON, T. G., BYBELL, L. M. & OWENS, J. P. 1993. Latest Paleocene lithologic and biotic events in neritic deposits of southwestern New Jersey. Paleoceanography, 8, 495-514. GINGERICH, P. D. 1983. Paleocene-Eocene faunal zones and a preliminary analysis of Laramide structural deformation of in the Clark's Fork Basin, Wyoming. Wyoming Geological Association Guidebook, 34, 185-195. GOERICKE, R. & FRY, B. 1994. Variations in marine plankton 813C with latitude, temperature and dissolved CO 2 in the world ocean. Global Biogeochemical Cycles, 8, 85-90. GOLDFARB, R. J., SNEE, L. W., MILLER, L. D. & NEWBERRY, R. J. 1991. Rapid dewatering of the crust deduced from ages of mesothermal gold deposits. Nature, 354, 296-298. HAQ, B. U., PREMOLI-SILVA,I. & LOHMANN, G. P. 1977. Calcareous plankton biogeographic evidence for major climatic fluctuations in the early Cenozoic Atlantic Ocean. Journal of Geophysical Research, 82, 3861-3876. HAY, W. W. 1989. Paleoceanography: a review for Geological Society of America Centennial. Geological Society of America, Bulletin, 100, 19341956. HERBERT, T. D. & SARMIENTO,J. L. 1991. Ocean nutrient distribution and oxygenation: limits on the formation of warm saline bottom water over the past 91 m.y. Geology, 19, 702-705. HERMELIN, J. O. R. 1992. Variations in the benthic foraminiferal fauna of the Arabian Sea: a response to changes in upwelling intensity? Geological Society of London, Special Publication, 64, 151-166 -& SHIMMIELD,G. B. 1990. The importance of the oxygen minimum zone and sediment geochemistry in the Northwest Indian Ocean. Marine Geology, 91, 1-29. HINGA, K. R., ARTHUR, M. A., PILSON, M. E. Q. & WHITAKER, D. 1994. Carbon isotope fractionation by marine phytoplankton in culture: the effects of CO 2 concentration, pH, temperature, and species. Global Biogeochemical Cycles, 8, 91-102. HOOKER, J. J. 1991. The sequence of mammals in the Thanetian and Ypresian of the London and Belgium Basins: location of the Palaeocene-Eocene boundary. Newsletters on Stratigraphy, 25, 75-90. HORNIBROOK, N. DE B., BRAZIER,R. C. & STRONG, C. P. 1989. Manual of New Zealand Permian to Pleistocene Foraminiferal Biostratigraphy. New

Zealand Geological Bulletin, 56.

Survey, Paleontological

HOVAN, S. A. & REA, D. K. 1992. Paleocene/Eocene

425

boundary changes in atmospheric and oceanic circulation: a southern hemisphere record. Geology, 2 0 , 15-18. JANECEK, T. R. & REA, D. K. 1983. Eolian deposition in the northeast Pacific Ocean: Cenozoic history of atmospheric deposition. Geological Society of America, Bulletin, 94, 730-738. JENKINS, D. G. & LUTERBACHER, H. 1992. Paleogene Stages and their boundaries (introductory remarks).

Neues Jahrbuch, Geologische und Palgiontologische Abhandlungen, 186, 1-5. KAIHO, K. 1988. Uppermost Cretaceous to Paleogene bathyal benthic foraminifera1 biostratigraphy of Japan and New Zealand; latest Paleocene-middle Eocene benthic foraminiferal species turnover. Revue de Pal~obiologie, Special Volume, 2, 553559. 1991. Global changes of Paleogene aerobic/ anaerobic benthic foraminifera and deep-sea circulation. Palaeogeography, Palaeoclimatology, Palaeoecology, 83, 65-86. 1994a. Benthic foraminiferal dissolved-oxygen index and dissolved oxygen levels in the modern ocean. Geology, 22, 719-722 1994b. Planktonic and benthic foraminiferal extinction events during the last 100 m.y.

Palaeogeography, Palaeoclimatology, ecology, 111, 45-71.

Palaeo-

& SAITO, S. 1994. Oceanic crust production and climate during the last 100 Myr. Terra Nova, 6, 376-384. , MOR~AYS, H. E. G. & OKADA, H. 1993. Faunal turnover of intermediate-water benthic foraminifera during the Paleogene in new Zealand. Marine Micropaleontology, 21, 51-86. KAST1NG, J. E D. & WALKER, J. C. G. 1993. Long-term effects of fossil-fuel burning and deforestation on levels of atmospheric CO 2. In: OREMLAND, R. S. (ed.) Biogeochemistry of Global Change: Radiatively Active Trace Gases. Springer-Verlag, New York, 151-165. KATZ, M. R. & MILLER, K. G. 1991. Early Paleogene benthic foraminiferal assemblage and stable isotope composition in the Southern Ocean, ODP Leg 114. - -

Proceedings of the Ocean Drilling Program, Scientific Results, 4, 481-513. KEMP, E. M. 1978. Tertiary climatic evolution and vegetation history in the Southeast Indian Ocean region. Palaeogeography, Palaeoclimatology, Palaeoecology, 24, 169-208. KENNETT, J. P. 1978. The development of planktonic biogeography in the Southern Ocean during the Cenozoic. Marine Micropaleontology, 3, 301-345. -& STOTT, L. D. 1990. Proteus and Proto-Oceanus: ancestral Paleogene oceans as revealed from Antarctic stable isotopic results. Proceedings of

the Ocean Drilling Program, Scientific Results, 113, 865-880. & -1991. Terminal Paleocene deep-sea benthic crisis: sharp deep sea warming and paleoceanographic changes in Antarctica. Nature, 353, 225229. KERRICK,D. M. • CALDEIRA,K. 1993. Paleoatmosphefic consequences of CO 2 released during early

--

426

-

E. THOMAS • N. J. SHACKLETON

Cenozoic regional metamorphism in the Tethyan orogen. Chemical Geology, 108, 201-230. & -1994. Metamorphic CO 2 degassing and early Cenozoic palaeoclimate. Geological Society of America Today, 4, 57; 62-65. KING, C. 1989. Cenozoic of the North Sea. In: JENKINSD. G. & MURRAY, J. W. (eds) Atlas of Fossil Foraminifera (2nd edition). British Micropalaeontological Society Series, Ellis Horwood, London, 418-489. KLOOTWIJK, C. T., GEE, J. S., PEIRCE, J. W. & SMITH, G. M. 1991. Constraints on the India-Asia convergence: paleomagnetic results from Ninetyeast Ridge. Proceedings of the Ocean Drilling Program, Scientific Results, 121, 777-882. KNOX, R. W. O'B. 1990. Thanetian and early Ypresian chronostratigraphy in south-east England. Tertiary Research, ll, 57-64. & MORTON,A. C. 1983. Stratigraphical distribution of early Palaeogene pyroclastic deposits in the North Sea Basin. Proceedings of the Yorkshire Geological Society, 44, 355-363. -& -1988. The record of early Tertiary North Atlantic Volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON,L. M. (eds) -

Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-4 19. KOCH, P. L., ZACHOS, J. C. & GINGERICH, P. D. 1992. Coupled isotopic change in marine and continental carbon reservoirs at the Palaeocene/Eocene boundary. Nature, 358, 319-322. KYTE, E T., LEINEN, M., ROSS HEATH, G. & ZHOU, L. 1993, Elemental geochemistry of Core LL44-GPC3 and a model for the Cenozoic sedimentation history of the central North Pacific. Geochimica Cosmochimica Acta, 57, 1719-1740. LARSEN, L. M., PEDERSEN, A. K., PEDERSEN, G. K. & FhASECKI, S. 1992. Timing and duration of Early Tertiary volcanism in the North Atlantic: new evidence from West Greenland. In: STOREY, B. C., ALABASTER, T. & PANKHURST, R. J. (eds)

Magmatism and the Causes of Continental Break-up. Geological Society, London, Special Publication, 68, 321-333. Lu, G. & KELLER, G. 1993. Climatic and oceanographic events across the Paleocene-Eocene transition in the Antarctic Indian Ocean: inference from planktic foraminifera. Marine Micropaleontology, 21, 101142. LUTZE, G. E & COULBOURN,W. T. 1984. Recent benthic foraminifera from the continental margin of Northwest Africa: community structure and distribution. Marine Micropaleontology, 8, 361-401. MCCORKLE, D. C., KEIGWIN, L.D. JR, CORLISS, B.H. & EMERSON, S. R. 1990. The influence of microhabitats on the carbon isotope composition of deep-sea benthic foraminifera. Palaeoceanography, 5, 161-185. McGOWRAN, B. 1991. Evolution and Environment in the early Paleogene. Memoirs of the Geological Society of India, 20, 21-53. MCKENNA, M. C. 1980. Eocene paleolatitude, climate, and mammals of Ellesmere Island. Palaeo-

geography,

Palaeoclimatology, Palaeoecology,

30, 349-362. MILLER, K. G., FAIRBANKS, R. G. & MOUNTAIN, G. S. 1987a. Tertiary isotope synthesis, sea level history, and continental margin erosion. Paleoceanography, 2, 1-20, - - - , JANECEK,T. R., KATZ, M. R. & KEIL, D. J. 1987b. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary. Paleoceanography, 2, 741-761. --, KATZ, M. E. & BERGGREN, W. A. 1992. Cenozoic deep-sea benthic foraminifera: a tale of three turnovers. In: BENTHOS '90, Sendal, Tokai University Press, 67-75. MACKENSEN, A. & BERGGREN, W. A. 1992. Paleogene benthic foraminifers from the Southern Indian Ocean (Kerguelen Plateau): biostratigraphy and paleoecology. Proceedings of the Ocean Drilling Program, Scientific Results, 120, 603-630. MANIVIT, H. 1984. Paleogene and Upper Cretaceous calcareous nannofossils from Deep Sea Drilling Project Leg 74. Initial Reports of the Deep Sea Drilling Project, 74, 475-500. MARKWICK, P.J. 1994. Equability, continentality, and Tertiary climate: the crocodilian perspective. Geology, 22, 613-616. MATTHEWS, R. K. • POORE, R. Z. 1980. Tertiary 180 record and glacioeustatic sea-level fluctuations. Geology, 8, 501-504. MOL1NA,E., CANUDO,J. I., GUERNET,C., McDOUGALL,K., ORTIZ, N., PASCUAL, J. 0., PARES, J. 0., SAMSO, J. M., SERRA-KIEL, J. & TOSQUELLA,J. 1992. The stratotypic Ilerdien revisited: integrated stratigraphy across the Paleocene/Eocene boundary. Revue de Micropalgontologie, 35, 143-156. MOORE, P. R. 1988. Stratigraphy, composition and environment of deposition of the Whangai Formation and associated Late CretaceousPaleocene rocks, eastern North Island, New Zealand. New Zealand Geological Survey, Bulletin,

100. MOORE, T. C. JR, RABINOWITZ,P. D., BORELLA, P. E., SHACKLETON, N. J. & BOERSMA, A. 1984. History of the Walvis Ridge. Initial Reports of the Deep Sea Drilling Project, 74, 873-894. MCLLER-MERZ, E. & OBERHANSLI, n. 1991. Eocene bathyal and abyssal benthic foraminifera from a South Atlantic transect at 20~176 Palaeo-

geography,

Palaeoclimatology, Palaeoecology,

83, 117-172. NILSEN, Z. H. & KERR, D. R. 1978. Paleoclimatic and paleogeographic implications of a lower Tertiary laterite (latosol) on the Iceland-Faroe Ridge, North Atlantic region. Geological Magazine, 115, 153236. NOMURA, R. 1991, Paleoceanography of upper Maestrichtian to Eocene benthic foraminifera1 assemblages at ODP Sites 752, 753 and 754, eastern Indian Ocean. Proceedings of the Ocean Drilling Program, Scientific Results, 121, 3-30. OBERHANSH, H. & Hst), K. J. 1986. Paleocene-Eocene Paleoceanography. In: Mesozoic and Cenozoic Oceans. Geodynamics Series, 15, 85-100. - - , Mt)LLER-MERZ,E. & OBERHANSLI,R. 1991. Eocene

427

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION Paleoceanographic Evolution at 20~176 in the Atlantic Ocean. Palaeogeography, Palaeoclimatology, Palaeoecology, 83, 173-216. O'CONNELL, S. 1990. Variations in Upper Cretaceous and Cenozoic calcium carbonate percentages, Maud Rise, Weddell Sea, Antarctica. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 971-984. ODIN, G. S. & LUXERBACHER,H. 1992. The age of the Paleogene Stage Boundaries. Neues Jahrbuch,

Geologische und Paltieontologische Abhandlungen, 186, 21-48. OTTENS, J. J. 8`: NEDERBRAGT, A. J. 1992. Planktic foraminiferal diversity as indicator of ocean environments. Marine Micropaleontology, 19, 13-28. OLlVA~EZ, A. M. & OWEN, R. M. 1989. Plate tectonic reorganizations: implications regarding the formation of hydrothermal ore deposits. Marine Mining, 14, 123-138. OWEN, R. M. & REA, U. K. 1985. Sea floor hydrothermal activity links climate to tectonics: the Eocene CO 2 greenhouse. Science, 227, 166-169. PAR, D. K. & MmLER, K. G. 1992. Late Paleocene to early Eocene benthic foraminiferal stable isotopes and assemblages: implications for deepwater circulation. Palaeoceanography, 7, 405-422. 8`: MILLER, K.G. 1995. Isotopic and faunal record of Paleogene deep-water transitions, Leg 145. -

Eocene warm climates, as inferred from clay mineral variations in oceanic sediments.

Palaeogeography, Palaeoclimatology, Palaeoecology (Global and Planetary Change), 89, 315-332. & KENNETT, J. P. 1992. Paleocene and Eocene kaolinite distribution in the South Atlantic and Southern Ocean: Antarctic climatic and paleoceanographic implications. Marine Geology, 103, 9-110. 1994. Antarctic subtropical humid episode at the Paleocene-Eocene boundary: clay mineral evidence. Geology, 22, 211-214. 8`: MAILLOT, H. 1990. Paleoenvironments in the Weddell Sea area and Antarctic climates as deduced from clay mineral associations and geochemical data, ODP Leg 113. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 51-70. ROBERTS, D. G., MORTON, A. C. 8`: BACKMAN, J. 1984. Late Paleocene-Eocene volcanic events in the northern North Atlantic Ocean. Initial Reports of the Deep Sea Drilling Project, 81,913-923. ROMEIN, A. J. T. 1979. Lineages in early Paleogene calcareous nannoplankton. Utrecht Micropaleontology Bulletin, 22. Roso~, D. B. & CORLISS, B. H. 1992. An analysis of Recent deep-sea benthic foraminiferal morphotypes from the Norwegian and Greenland Seas. &

PEDERSEN, T. E, PICKERING, M., VOGEL, J. S., SOUTHON, J. N. 8`: NELSON, D. E. 1988. The response of benthic foraminifera to productivity cycles in the eastern equatorial Pacific: faunal and geochemical constraints on glacial bottom water oxygen levels. Palaeoceanography, 3, 157-168. POSPrCHAL, J.J. 8`: W~SE, S.W. JR. 1990. Paleocene to middle Eocene calcareous nannofossils of ODP Sites 689 and 690, Maud Rise, Weddell Sea.

Proceedings of the Ocean Drilling Program, Scientific Results, 113, 613-638. PREMOLI-SILVA,I. 8`: BOERSMA, A. 1984. Atlantic Eocene planktonic foraminiferal historical biogeographic and paleohydrologic indices. Palaeogeography, Palaeoclimatology, Palaeoecology, 67, 315-356. PRENTICE, M. L. 8`: MATTHEWS, R. K. 1988. Cenozoic ice-volume history: development of a composite oxygen isotope record. Geology, 16, 963-966. RATHBURN,A. E. & CORLISS, B. H. 1994. The ecology of living (stained) deep-sea benthic foraminifera from the Sulu Sea. Paleoceanography, 9, 87-150. RAte, D. M. & SEPKOSKI, J. J. 1986. Periodic extinction of families and genera. Science, 231,833-836. REA, D. K. 1994. The paleoclimate record proyided by Eolian deposition in the deep sea: the geologic history of wind. Reviews in Geophysics, 32, 159-195. - - , ZACHOS,J. C., OWEN, R. M. & GINGERICH,D. 1990. Global change at the Paleocene-Eocene boundary: climatic and evolutionary consequences of tectonic events. Palaeogeography, Palaeoclimatology, Palaeoecology, 79, 117-128. ROBERT, C. 8`: CHAMLEY,n. 1991. Development of early

-

Palaeogeography, Palaeoclimatology, ecology, 91, 13-20.

-

Proceedings of the Ocean Drilling Program, Scientific Results, 145.

-

Palaeo-

SANDERS, H. L. 1968. Marine benthic diversity: a comparative study. American Naturalist, 102, 243-282. SCHMIDT, K.-H. 1991. Tertiary palaeoclimatic history of the southeastern Colorado Plateau. Palaeo-

geography, Palaeoclimatology, Palaeoecology, 86, 283-296. SCHNITKER, D. 1979. Cenozoic deep water benthic foraminifers, Bay of Biscay. Initial Reports of the Deep Sea Drilling Project, 48, 377-414. SCrtROEDER, T. 1992. A palynological zonation for the Paleocene of the North Sea Basin. Journal of Micropalaeontology, 11, 113-125. SELVERSTONE, J. & GUTZLER, D. S. 1993. Post-125 Ma carbon storage associated with continent-continent collision. Geology, 21, 885-888. SETO, K., NOMURA,R. & NIITSUMA,N. 1991. Data report: oxygen and carbon isotope records of the upper Maestrichtian to lower Eocene benthic foraminifers at Site 752 in the Eastern Indian Ocean. Proceedings

of the Ocean Drilling Program, Scientific Results, 121,885-889. SHACKLETON, N. J. 1977. Carbon-13 in Uvigerina: tropical rainforest history and the equatorial Pacific carbonate dissolution cycles. In: ANDERSEN,N. R. & MALAHO~, A. (eds) The Fate of Fossil Fuel COe in the Oceans. Plenum Press, New York, 401--447. 1986. Paleogene stable isotope events. Palaeo-

geography, Palaeoclimatology --

-

-

PalaeoecoIogy,

57, 91-102. 1987. The carbon isotope history of the Cenozoic. In: BROOKS, J. & FLEET, A. J. (eds) Petroleum Source Rocks. Blackwell, 427--438. 8,: BOERSMA, A. 1981. The climate of the Eocene ocean. Journal of the Geological Society of London, 138, 153-157.

428 -

-

E. THOMAS • N. J. SHACKLETON

& HALL, M. A. 1984. Carbon isotope data from Leg 74 sediments. Initial Reports of the Deep Sea Drilling Project, 74, 613-619. & -1990. Carbon isotope stratigraphy of bulk sediments, ODP Sites 689 and 690, Maud Rise, Antarctica. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 985-989. , COREIELD, R. M. & HALL, M. A. 1985. Stable isotope data and the ontogeny of Palaeocene planktonic foraminifera. Journal of Foraminiferal Research, 15, 321-336. , HALL, M. A. & BELIE, U. 1985. Carbon isotope stratigraphy, Site 577. Initial Reports of the Deep Sea Drilling Project, 86, 503-511. --, -& BOERSMA, A. 1984a. Oxygen and carbon isotope data from Leg 74 foraminifers. Initial Reports of the Deep Sea Drilling Project, 74, 599612. & members of the Scientific Party. 1984b. Accumulation rates in Leg 74 sediments. Initial Reports of the Deep Sea Drilling Project, 74, 621-637. SIGNOR, P. & LIPPS, J. 1982. Sampling bias, gradual -

-

extinction patterns, and catastrophes in the fossil record. In: Geological Society of America, Special Publication, 193, 291-296. SLOAN, L. t~ BARRON,E. J. 1992. A comparison of Eocene climate model results to quantified paleoclimatic interpretations. Palaeogeography, Palaeoclimatology, Palaeoecology, 93, 183-202. --, REA, D. K. & ZACHOS, J. C. 1992. Possible methane-induced polar warming in the early Eocene. Nature, 357, 320-322. , WALKER, J. C. G. & MOORE T. C. JR. 1995. The possible role of oceanic heat transport in early Eocene climate. Paleoceanography, 10, 347-356. SMITH, H. A., CHAMBERLAIN,C. P. & ZEITLER,P. K. 1994. Timing and duration of Himalyan metamorphism within the Indian Plate, Northwest Himalaya, Pakistan. Journal of Geology, 102, 493-508. SPEIJER, R. P. 1994, Extinction and recovery patterns in benthic foraminiferal paleocommunities across the Cretaceous/Paleogene and Paleocene/Eocene boundaries. Geologica Ultraiectina, 124. SPIEB, V. 1990. Cenozoic magnetostratigraphy of Leg 113 drill sites, Maud Rise, Weddell Sea, Antarctica.

Proceedings of the Ocean Drilling Program, Scientific Results, 113, 261-315. STAUDIGEL,H., HART, S. R., SCHMINCKE,H.-U. & SMITH, B. M. 1989. Cretaceous ocean crust at ODP Sites 417 and 418: carbon uptake from weathering versus loss by magmatic outgassing. Geochimica et Cosmochimica Acta, 53, 3091-3094. STOTT, L. D. 1992. Higher temperatures and lower oceanic pCO2: a climate enigma at the end of the Paleocene epoch. Paleoceanography, 7, 395-404. & Kennett, J. P. 1990. Antarctic Paleogene planktonic foraminiferal biostratigraphy: ODP Leg 113, Sites 689 and 690. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 549-570 - - , SHACKLETON,N. J. & CORFIELD,R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: inferences from the stable isotopic

composition of planktonic foraminifera, ODP Leg 113. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 849-864. --, SINHA, A., THIRY, M., AUBRY,M.-P. & BERGGREN, W. A. 1996. Global 813C changes across the Paleocene/Eocene boundary: criteria for terrestrialmarine correlations. This volume. THIERSTEIN, H. R. 1989. Inventory of Paleoproductivity Records: the Mid-Cretaceous enigma. In: BERGER, W. H., SMETACEK V. S. • WEFER, G. (eds) Productivity of the Ocean: Present and Past. Life Science Reports, 44, 355-375. THOMAS, E. 1989. Development of Cenozoic deep-sea benthic foraminiferal faunas in Antarctic waters. In" CRAME, J. A. (ed.) Origins and Evolution of Antarctic Biota. Geological Society of London, Special Publication, 47, 283-296. 1990a. Late Cretaceous through Neogene deep-sea benthic foraminifers, Maud Rise, Weddell Sea, Antarctica. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 571-594. 1990b. Late Cretaceous-early Eocene mass extinctions in the deep sea. In: Sharpton, V. L. & Ward, E D. (eds) Global Catastrophes in Earth History Geological Society of America, Special Publication, 247, 481-495. 1992. Cenozoic deep-sea circulation: evidence from deep-sea benthic foraminifera. American -

-

Geophysical Union, Antarctic Research Series, 56, 141-165. , BARRERA, E., HAMILTON, N., HUBER, B. Z., KENNETT,J. P., O'CONNELL, S., SPIEB, V., WEI, W. & WISE, S. W. JR 1990. Upper Cretaceous-Paleogene stratigraphy of Sites 689 and 690, Maud Rise (Antarctica). Proceedings of the Ocean Drilling Program, Scientific Results, 113, 901-914. --, SHACKLETON,N. J. & HALL, M. A. 1992. Carbon isotope stratigraphy of Paleogene bulk sediments, Hole 762C (Exmouth Plateau, eastern Indian Ocean). Proceedings of the Ocean Drilling Program, Scientific Results, 122, 897-901. TJALSMA, R. C. t~ LOHMANN, G. P. 1983. Paleocene-

Eocene bathyal and abyssal benthic foraminifera from the Atlantic Ocean. Micropaleontology Special Publication, 4. TONARINI, S., VILLA, A. M., OBERLI, E, MEIER, M., SPENCER, D. A., POGNANTE, U. & RAMSAY, D. A. 1993. Eocene age of eclogite metamorphism in Pakistan Himalaya: implications for India-Eurasia collision. Terra Nova, 5, 13-20. TOURET, J. L. R. 1992. CO 2 transfer between the upper mantle and the atmosphere: temporary storage in the lower continental crust. Terra Nova, 4, 87-98. VAN MORKHOVEN,F. P. C., BERGGREN,W. A. & EDWARDS, A. S. 1986. Cenozoic cosmopolitan deep-water benthic foraminifera. Elf-Aquitaine, Pau, France. VON HmLEBRANDT, A. 1962. Das Pal/ioz/ine und seine Foraminiferenfauna im Becken von Reichenhall und Salzburg. Bayerische Akademie der Wissen-

schaften, Mathematisch-Naturwissenschaftliche Klasse, Abhandlungen, Neue Folge, 108. VAREKAMP, J. C., KREULEN, R., POORTER, R. P. E. & VAN BERGEN, M. J. 1992. Carbon sources in arc volcanism, with implications for the carbon cycle.

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

Terra Nova, 4, 363-373. WALKER,J. C. G. 1993. Biogeochemical Cycles of carbon on a Hierarchy of Time Scales. In: OREMLAND,R. S. (ed.) Biogeochemistry of Global Change: Radiatively Active Trace Gases. Springer-Verlag, New York 3-28. & SLOAN,L. 1992. Something is wrong with climate theory. Geotimes, June, 16-18. WHITE, R. S. 1989. Igneous outbursts and mass extinctions. EOS, Transactions American Geophysical Union, 70, 1480, 1490-1491. 8r MACKENZIE, D. 1989. Magmatism at riftzones: the generation of volcanic continental margins and flood basalts. Journal of Geophysical Research, 94, 7685-7729. WILLIAMS,C. A. 1986. An oceanwide view of Palaeogene plate tectonics. Palaeogeography, Palaeoclimatology, Palaeoecology, 57, 3-25. WING, S. L. 1984. A new basis for recognizing the Paleocene/Eocene boundary in western interior North America. Science, 226, 439-441. --, BOWN, T. M. & OBRADOVICH, J. D. 1991. Early Eocene biotic and climatic change in interior western America. Geology, 19, 1t89-1192. WOLFE, J. A. 1978. A paleobotanical interpretation of Tertiary climates in the northern hemisphere. American Scientist, 66, 694-703. -

-

-

-

429

& POORE, R. Z. 1982. Tertiary marine and nonmarine climatic trends. In: Climate in Earth History. National Academy Press, for the American Geophysical Union, 154-148. WOODRUFF, E & SAVIN, S. M. 1985. 513C values of Miocene Pacific benthic foraminifera: correlations with sea level and biological productivity. Geology, 13, 119-122. ZACHOS,J., REA, D., SETO, K., NOMURA,R. & NIITSUMA, N. 1992. Paleogene and early Neogene deep water paleoceanography of the Indian Ocean as determined from benthic foraminifer stable carbon and oxygen isotope records. American Geophysical Union, Geophysical Monographs, 70, 351-386. --, LOHMANN,K. C., WALKER,J. C. G. & WISE, S. W. 1993. Abrupt climate change and transient climates during the Paleogene: a marine perspective. Journal of Geology, 101, 191-213. --, STOTT, L. D. 8r LOHMANN, K. C. 1994. Evolution of early Cenozoic temperatures. Paleoceanography, 9, 353-387. ZAHN, R., WINN, K. & SARNTHEIN, M. 1986. Benthic foraminifera! fi~3C and accumulation of organic carbon: Uvigerina peregrina group and -

-

Cibicidoides wuellerstorfi. Paleoceanography, 1, 27-42.

430

E. THOMAS & N. J. SHACKLETON

Appendix 1. Faunal data, Site 689

~

~

E

22H-4. 108-116

203.081 55.401[ 42[ 28~-319[-

22H-5.113-118

~-(~[

55.450| 28

~---I-~o~.87~_~o~-~I

2213181

31132el

"

= ,

-~

2]-

2t |

I

l

~-Lb-4-~e/20S.151

55.4991 ~_~_I e l ~ 7 _ I - - - ~ -

~

~X-4~4~-42

55.7001 46I 2913271

I

T212.11!

I

~! ~o ~i ~o

[

"

I

;

~

I

I

I

E

~

~

~

-

I

~

81 1

~

~

T

~

15

~

T

'

l

]

-

1

]

~ 2'~-~i5~-~7~i ~ t ~12~71

-

23

~

5]

I

i

t

+1

-: - ]

I

:

22 - 1

~

1

~1

91

I

~

~ 31

81

13!

~ - ~ - ~ - - - . ~ - - ~ - ~ - ~ - - - - 1 2,x,cc

C 43~

321

!

21

I

,I

I

I

+ ~- -2l

~i

i

i

I

-~-~-

I

2'

I

! -l--

I

i

i

i

I

i

i

E

j

s

,

,,

sample

.

22H-4, 108-116 203.08 55.401 22H-5. 40- 42 , 203.9~ 5"~5.42~ 22H-5,113-118 I 204.63 55.450 22H-6, 8-12 1- 205.10 55.465 22H,CC ~ 55.475 23X-1, 40- 42 ; 207.61 55.490 23X-1, 80- 82 ! 20801 ~ . . ~ 23X-1,87- 89 ~ 208.08 55A98 ~ 9 6 7 208.1~5 55.499 ~2-3X-1,106-108 208.27 55.500 23X-1:i16-119 208.38 55.501 23X-2,40-42 209.11 55.510 23X-3, 42- 44 210.61 55.621 :23X-4.40-42_ ~ 1 2 . 1 1 55,7~

42 28 319 6 32 2-4 373 10 28 22 318 8 33 22 323 8 27 20 353 21 43 31 32(3 9 42 29 313 3 60 37 394 10 6~8 40 357 12 ',~287 17 5(3 39 310 18 68 51 304 3 1 ~ 52~--~1319 1E 461 29!327--71 1

5 5 3 4 5

,

1,,

1 8 5 2 5 2

8 12 9 g 10 t6 110 1 5

1

~

56.~ I ~eI 3 8 1 ~ 1 ~ - ~ -

~

21

4

- 1]

~4X,CC

~ ~

56.3731 33 i 241237

1

i

25

,

19

9 1 12 _ 4 14 41 26 11 28 7 2 4 35 2 4 14 1 2 i0'-4

~ : ~

22

,5

4 1 5 1 23 13 4 6 2 21

2

4

1 , 2 1

,

1

1

7 ..... e~-

2 1 5 5

5 4

1 4

4

3

9

13 3 2

5

1

5 3 3 1 1

2 2 2

e

!

!

,

121

--1[- 25 9 7 14 26 5 12 4 22 3 10 1! 12 5 3 2 14 4 6 3 6 3 3 3 7

1

3

---iI

2

1

1 2

~

it 5 27 28 41 ,~ 12- ~ 11 i 4~ 14~

111

431

PALEOCENE-EOCENE BENTHIC FORAMINIFERAL EXTINCTION

+o++-+++- +++!" .+

+=>,

~

'~ + .+

9

+

=

.+

.+|

:+ ++ + + -+

-

;+ 9 +++++++0 " + =I+ +.+~+ , + , ~ + o , .~ 2

-

m+=,

=,

-+ o

+

' ++!++~ i ~

++,.+,..,t+ +++ +,.0

22'

!

1

2

1

3

1

......

~

1 11 --

7

8

7

3

'~_

4

g

-

I 7

1

1

, 1

1 '-

7

~

m!

8 "

44

51

49 20 4912

5

!

1

i

+

3[ i - ++

!

.......

~

__

.--~

+

-

I

2 1

~5

1

7

...... 1

"

'

i

++o+++.

1

1 7

3

_6 ~24

--

2

~

1 li

1 2

2 5

1

1; '

i

=o+ ++

i+

~ ,=,

I

+ ! ~..,

~ +..

.

~

+

+~

=,

I

....

~

f

-2 9

i2

3

I

]

i

1

23

~ ~

--~

1

~--

! ~

i

_. ~.._~2~

26

5

12

98

3

11

72

41

1

2 5

17 1

2

5l

47 79 1 72

5 16

38

2

~I ....... t i~ ~ 8 3

55

__

1

-i

-

li

1

ii 4~

3

"

1

1

i

4

4

l

6

15 t 73

I

3 3 4

f

--~-

I

1 14 6

I ~

~0 'Ti :::L?

.........

3

--1

15 43 3 2 4 3 2

32 95

3

20

4

28

19

28!

15' 5

]:

3 2:

--+ ....

I~

6

-~1-

2

1 8

9 15 9 16 20 20 26 201 14

-

6 10 B

1

1

"" 1

2

;

9 .....

12

,

E. T H O M A S & N. J. S H A C K L E T O N

432

Appendix 2. Faunal data, Site 6690 i

i I

,

,

+

,i

,

J

,

l

-

sample 41-43 16H-2, 41- 43

16H-1,

~

.~i-~

E{ E ~ '~ 138.21 , ~ - - 5 6 ~ 42 - -T71~139 54.729 _._57~-+

1 ~ : 5 4 ! ~____.~ 142.72_~ 1BH-5, 41- 43 16H-fl, 41- 43 16HCC 17H-1,40-42

144.21 145.71 4~,,50 I 147.91

54.843 58' 54.881 49 54.926 35 54.937 56

38 34 27 38

!

_<__~ -- 7 11 1

i

'

,II i,,

+ E:

i,

'

i

.~

~

~

.c_

.~

~_

=

9

....

~' ~

3

8=

3

.___ 6

3]

___

4.1 .

---

6~

3~ 17H-3.40- 42 ! 150.01 55.022 46 34 1 7 H - 4 , 4 i : ~ , . . . . 1-152..42. 55.059 56 35 ~YR-~: ~ : - ~ .i I ~ 7 ~ T 55.095 51 33 i7H:--6,4~-42 I 'i~ 55.132 56 37 JTH~ ~_157.20 55.175~~ 41 32 ~ ~~:~ 553y-~-~ t-~ 18H-2, 40- 42 i 15~.11~55.221 42 28 18H-3,40-42 160.61 55.257 54 37 1 8 H ~ 40- 42 162.1 ~ -. ~ , ,

1ell-5. 40- ,.2

,~.~I ~ - ~ _ _

F ~.e

18H-6, 40- 42

16,5.11 ~ , 3 ~ - , 3 ~ 3 0 ~ - 2 2

,,,.cc +.1.,o.,,~

1 ~ ~ 1,,7~1

19H-1, 74- 76 19H-1,114-120 19H-2, 40- 42

167.65L~-8~ 168,07~~ 168.81 9 ~

.

.

.

.

.

.

.

. 3

..... 2--i 3

--'~

--~i

~ I~1~ 23

, 169.58~.474

hgH-~. 51- ~ 19H-3, 60- 62 19H'3' 66" ~4 i9H:3-7~: 19H-3, 74- 76 19H-3,115-121

~,~.~ 55.4L.5 4 . 0 ; ~ ~-70:,51~-7" ~--~ t " ~ ~ 55.499 37 29 55.5oo s8 40 i ' ~ . 6 5 ] 55.501 63 43 T~-71.~. 55.511 71 44 1 _ 7 ~ 7 6 ,

19H-5, 74- 76 19H,CC 20H-1, 40- 42 20H-2, 40- 42 2'0H-3, 40-42 20H-4. 41-43 20H,CC _21H21, ~ - 42_ 21H-2, 40- 42 21H-3, 40- 42 21H-4, 5- 7 21H,CC 2 2 ~ 40- 42 42- 44 22H-3, 42- 44 22H,CC

~ 6 55.584 0 174~.~ 55.601 51 5 55.612 2 _~ 176.21 I 55.654 73 ~ 55.696 65 -- 179,22f 55.739 61 ..... 1-~.28_~ 5 5 ~ 52. , 180,71~. 55.780 -61~_182.2t r 55.822 67 ~~ 55.864 62 [ 18486~ 55.896 55 i 185"201 55,900 54 ~185.7!~ 55.915 61 i 187~3i 55.38:, 58 ~ 56.070 01 I 191"151 56.163, 57,

~

_~ - ~ - ~ _ , - ~

1gH-2,118-124-

19H-4, 74- 76

19

1

"- 1"

-~---

-~-

__

42 37

.

.

.

.

.

.

.

.

.

.

2

/

37

!

_~__

i--~

~

~

1

I. i .

41 38}~-.

35] a30 1

. .

. .

. .

.

. . 1

-~T---~-

---

2i

I

1

5~-------1 ,~ 8i - - - - 1 3i 15 2T-i

261 11[

11 1

---

__

433

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

.

.

!o i

o

~176

'~'~i

--

~ i~~ ~

15

1

......

' 3

7

1

12

3r _

~-

7

7 3

'

~- 2 =

=

1

1

10

~

~,

2

8

8'

1"-

t

1

~ .

,~_~_~

~ !. . . . .

1 ,-

,_. :1-1

1

1

3

~i

L--1

3

d

J

J -

! - - ~--- r-

27

1

8

18

--~'~- 25

4

i

83 22

i

4

t

11!

,

1

~-,

1

5

2

1

10

1

2

6

2

1

7!

2

~8

2--

2

,

2-

18

, 1 2

t

. 8

614

.

.

I

1

. . . .

...... 4

~--~

1

~i

i

_i~

'-----1

ii~- :i

.

1"--2 ~

t

-

'L

~

~-i..i

:I"-2

'

~ ..... --'~-

.... , , , ~ - ' - ~ - - - 1 - - ~

1

1

~3 I

3

r~-~

1

13

,

~ ---

*2

3

1

,

"--

--

.....

11 1

~

'

I

I

2--~--~

E. THOMAS & N. J. SHACKLETON

434

Appendix

2.

continued

e

,t

,

. =

. ; . , , . . . . . . .

16H-1.41-43

1 ~ : ~ , 41- 43 16H-3.41-43 16H-4,41-43 16H-5, 41- 43 16H-8. 41- 43

17H-1,40-42 117H-2,40-45 17"-3,40-42 17H-5,40-42 17H:CC"

--

18H-2.40- 42 18H-3, 40- 42 ~8H-4,~_-~ _18H-5,40-42

18H-~,~-42 18H,CC

-19H~4--~0

~I-38. 54.690 "~ 1 3 ~ 711 54.7~ ~ , ~:767 ~42.72154.80.5 ~ _~144.21 1 4 4 " 2 154.843 54_~ 1 I

~

56 42318' 57' 3~ 328' ' ~,---~ ~ 52 3 7 3 1 4 5 8 ' 3 8 321' '

55.095 ~

~"

'

30 2-8 12 . . . 4-- .... ~ -~ __' ~ . " 15" ~ -1 5 | 12 3 3 9 2' ' f2" ' 15'--2~3-~-3 "~

9

15--~

159.11 55.221 42 2 8 3 2 4 160.81 55.257 ~ 3--7-~,2"

7 8-

15 5 3 1-6" 2" u ~,~ ~ -1 ~ ~ 10 1 7 6 2 4 3 20

~..._55.34~ ~ 1 ~ . ~ 55..,~ ~ T~8.071 ~

-~

2,?.~ ~3 289

-~

~-~-H:~2~

3-e~:,~ ~~ ~ _ ~

-19H-3, 80- 62 - -

1-~-55..~7-33

~

-170.575 5 . 4 ~ ~ -

~

32z

~

-~

----

2

--

1~

- - - ---5 . . . . . . .

-

~

--

2-7

-~

....

~

~H~.-~--~,~ 19H-3, 74-76 1 1 ~ 1

31 34' ~ 18 22

t

~

~--

~

10 3 J ~'---~- 9"~6" ....... ~ ----~-3 ~ 3 3

1

T~-

l

10 ~Y ~

3

14

2

i x~

-

L ~ 8 --[-i~

~ --'~1~--

!

!] 1

.

.

.

'

~

~

2

--

~

3~

4

1

>

~

'

1-=15

:

18

2

- t 7 2 . 1 6 55.542- ~'8 4 7 1 3 1 2 1 ] " ~ 173.31 55.573 66 48302' I 91

22H-2. ~ - 44

~7.~,.-. 5 5 ~ ~ . ~

22H-3, 42- 44

~88.~-58.o7o~,~3~

22H.cc

-~ -191.~ ~- - - - ~~ ; -~ i i

1" '

51 35 273 6 5,?: 37 298 7 73 48 342 t~ 1 65 42307 6 61. 37 343. 5. 1 52 34 308 12 81 41 297 8 87 38 334 8 3 62 44 316 23 5 55 3 6 ~ " ~ 15 54 36 2 6 1 ~ ~ 1,5 6 1 1 3 8 3 2 3 ~- - 7~ - - ~

~2~~

----1

t

-'~

1 ....

1~--

y ~

.....

~-

3-1-5

i ........

3

4

~26

55.50or eel 48 ~ e --- ~ - 1 - - - ~ - - ~ ] 170~65 55.501' 63 r 43 322' -171.05 55.511 71' 44 302' 11 13' 1-'

174.30 55.601 174.71 55.812 176.21 55.854 177.71 55.898 . 179.22 55.739 . 180.28 55,768 180.71 55.780 182.2--1 5 5 ~ L ~ 183.71 55.864 184.8-------'s 55.896 185.2~ 55.900 1185.71 55.915

18 23

t ---~----11 ~ ' - ~ . . . . ~

'"

-

~.~

19H,CC 20H-1.40- 42 20H-2.40-42 20H-3,40-42 20H-4, 41- 43 20H,CC 21H-1,40- 42 21H-2, 40- 42 21H-3, 40- 42 21H-4, 5 - 7 21H.CC 22H-I, 40-42

-

- - 3

.....

4

1 3 1 3 8 5 8 4 3 9

1

~ ~0 120,

~ 3 13i' ! 4' 51

4 2 2 4

t ,

i 7

I -4; ~IY 18,

I

I 3 , 2

9 8 9 4 6 7 7 5 2 8 10 3

li' 4" ~ J

1

1 2 3 1 I

4 3 2 - -

3 8 3 3 2

8 2 3 ~ 2 7 8, 1,

--~E

4

--~-~-~ [19-H:.4,7~76 119H-5,40-42

8

. -

---f . . . . . .

~-5 ~

~ ~

-

~

'=

~'~'~-~----4--3

21 11

--

-

,

o

5 2 3 28 --, - ~ _ --e-~ -~------~ ~ ..... ~ _ ~-9--'7~-~_ 1 28- :_- - - - - - - - - - : ~_: I ' r 13 8 8 2 14 ~ --, ' ' ' 1 1 ' - - 5 - - ~ .... ~ - ~ - ~

1-8"-

~7:26 55.175 41 32 278

6

.~

=

-9 ~ ~ I--2+9~-1~3~'~-~-~-~--I -~ ~ ~ ~---8~~-~22' ~ . ~ ' ~ - - ~ _ i ~ -- 1 ~ -~

~ ~ 8 _ I ~ - ~ ' i 6 ' - - ~ 8 j

/ !~.1~55.294_~7 29342 1~_._~.6155.330 40 28335

"

7.

...... 19 1~8 7' ~ ~ ~ ~ --" ---'---- 1--~L33- I ~ 1 { ' - 6

= . ,

= = . -

7

-~.91 54.937 58" 3 8 r ~ ~ -1~4-9:4154.974' 50 ~ 32.~337 I 150.91 55.022:48' 34 j 3 3 0 1 ~ ~

.

,

,

i

-~i

3 ' [ 2' i 3" 1

2~ 3 3 .... ~-~11 3 3 4

4

L 6L I u ~ 13 20 14 12 27 21 18 21 33 16 7 11 14 22 15

71

4" 8 14 24 3 5 6 14 1 10 12 1 5 - 3 11 5 4 3 3 9 4 2 6 1 8

15

"..... ~- 1 1

1-3

i

i 11

13

4

7

9 7

12

3,

I TO

1

! =3 ..... ~:--9

j 6 ~-~

.........

12 - - - - - ~ 9

~ ,

1

: :

~ - ~ .......... ,~-

11 8

1

7 2

I 3

2

-2~

~

2 4

9 8

1

9

9

2

5

~,

I

~--~ "-9

~, -6 ,

'

~

i

6

!

-1

435

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

:

I

"

9

'

i

= 9~-~,I

e

9

~

~

~

~

,_~ 12

i

'

IJ

:

~

2

i - -- ? --1~

....

~

47 20

--+"

~

4

_2 _ ~ _ _ _ 2

4j

2 ++i . . . . . . .

P

. 5

~__._

~----+~ _ _ ~ 3+ _

1~

__~-_

[ -]+[---~-/ -,--

.....

_ ,L_

!

:

[

--++---4---++~-

,

.

1

13

1

62

2

1 114

1

!

~ -

-"++--+--I

Ii

46 94'

.

.

.

.

.

.

.

....

~--.-~L - i

.

.

.

.

.

.

. .

'

.

.

-~

- - -, r - - - - ~ - ~

.

_~__+_+~_

- + - : ++-+~ +._+

___.,

. . . .

u

:

,

24

4 41

..... + - - ~ - - -

+

3{+1-1 , - + - + --t--,4----

~

......

.....

'_J~:

q

I"-~V-~- 3--gt~l

1 1 1 + + I

+~._..._.,-~.._

~-.----4

11

i

3:

-; -+i 4 -

i

I

+-~; .... ~ -

-.-

~

21

i

i

~

~

~

_

_

~____+

_.__:+____, . . . . .

++ _ _ ~ _ _

ff :----+Z ~ + ++ ..J+_~ ,"

+ + . ~

F-2

+~-~-1

:-+ .... -+

--~---~'i -

i

=

+ 2 ~

44 i

, -

+ t3,

o!

~-+-1- . . . . . . .

-:_~:L+-/+ 9-~,--- 9.-r-- ...... __+_ , + _ ~ _ + +3~~ 5, ....... 4!

+

~ ,+3

'

a.~ - - - - ~ :

!

.....

--4-------~

......

-+-::~+4~-

.

! -

19'

4

i

=

4

--

,

-'~

I

.~ r

~

,

,

1

436

E. THOMAS •

N. J. SHACKLETON

Appendix 3. Faunal data, Site 525 i

[

9 ~.

i! sample 3~-4, 75,. 77 32-4,125-127 32-5, 25- 27 32-5, 48- 50 32-5. 77- 79

I

.-~,~,,

I

'

I

'

-o~

~,

=~

,=E

ii -'-~ 389.36 389.86 39OO8 390.38

55,38~ 55.394 55.404 55,412

3 47! 33 3221 42 31 312: 47 31 329!

32-5.',.~-11~ 32-8, 2,5. 27

! 3so.7e 55,419 55.427 39 34 32 25 33011 309 ! i ~el.33 55.441 52 40J 350

32-6, 5O- 52 32-6, 76- 78 32-6, 90- 92 32-6,113-115 33-6,130-132 32-6,145-147 32-7, 22- 24 32-7, 53-55

39101 ~ 391 87 ,! 392.01 392.24 392.41 , 382.56 392.81 3~3.11 I

1 4 10 5 ---5 2~9 14 17

55,447 43 = 3~. ~ 4 ] 55.454 49' 35 326 ! 55.457 53 40 313 55.475 54 3~ 343 55.491 34 27 328: 55,501 60 40 317 55.519 81 56 352 55.540 73 5 2 3 0 9

'

2 4 6 9 9 42 34 29 6 21 11

--3

1 7 _.. 4

1 7 44! 13 I 37 2 4

I I J

2 4 3

1 2 1

.

1

1 2 2 ~ 1 2 5 4 1

9 e 5

8

3 4 13 3 15 10 13 ~ 31 2,

2

5

2

I Ij I

1

i

3 1 3

26 7 18 14 16 17 17 13 13 13 19 3 9 20 13

66 51 6 56 86 59 30 27 1-8 11 13 14 16 6 3

1 11 7 4 5

11 t2~

1

4i 5t 1 10

1

17 11 8 3 8 1

1 1 3

I ~

i

sample 32-4, 75-

32-4.12~127 ! 389.38 55394 32-5, 25- 27 32-5,48- 50 32-5, 77- 79 32-6, 25- 27 32-6,50-52 32-6, 76- 78 32-6, 90- 92 32-6,113-115 32-6.130-132 32-6,145-147 32-7, 22- 24 32~ 53-55

389.86 390.08 390.38 390.76 391.38 391.61 391.87 392.01 392.24 392.41 392.56 392.81 393.11

55.404 55A12 55.419 55.427 55.441 55.44/ 55.454 55.457 5.5.47E 55.491 55.501 55.519 55.544]

= 32322 31 31 32 25 40 32 35 40 39 27 40 56 52

312 329 330 309 350 324 326 313 343 328 317 352 309

,.

1~ 11 17 19 12 14 18 13 12 10 10 4 12 14

,

1 11

2 1

8 8 6 15 7 14 12 15 10 4 e 5 4 4 1[- 6 i

= ,

4 2 1

1

= 6 5 6 1 4 4 6 8 7 5 8 5'

2 2 3 2 2 6 1

o

o,i~

2~

,,=

-~i

27 18 20 50 56 38 65 40 45 108 18i 8[ 8}

8 12 15 7 1 6 10 9 8 17 48 9 10

2 1 8

2 5 2 2

4 18

5 2 4

1 2 4 2 2 1 4

.

~

.

=1 20 1 24 1 28 12 2 1 27 22 2 23 1 23 3 35 1 22 1 4 14 1 7 20 [11 8

1

t I

437

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

:+ ++:+

+ _,

+ +'++-++/+,~+, + ++, +:+ +

,- ~ ~ +

:+

m

+++,++ + o++ , + +, +, :: +++ '+ . +,+, ++ i ! +"

+++ .+.-+" .+::~' + 1 7+++++:+ ~ . ~+' + ++~+~-+ '6 j. t + ++" :+ +:++,-+++ .o +, !++ +

-.+

"r 9 ."r

~-~---

3 5 3 6 9 6 7 8 3 4 2

~

1

2

3 11.~ 61 161, ~24-11 17 17

9 13 18 18 ~

---

l 1i

1

t T. . . .

1

!

1 1

1

7 7

+

--

.,++

-

-~

2

1

.

,

.,,

,

.

4 1 11 4 1t 11 22 7 8 1 3

17

.

,

1 1 3 1 1

1 1 3

1

t,

.

,:'+

~

, I

E

~

|

u

_

2

~

. 24

_

2

-

3

1 1 1

3 1 I 4 1

3 1

1

-!!

11

3

9

2

-)0 ,5[

) ~

,41

/-~

7

1

1 2 1 5 2 2

,_. + ,

~ 'i+ .: ,, ..,. ,

-

4

1 2

++:8~_+

+ ~

4 1 3

1 2 1

+

"+ .~_

1 2

3

11 =

~

--4

- [ ~o I 14

.,++

438

E. T H O M A S & N. J. SHACKLETON

Appendix 4. Faunal data, Site 527

.

.

.

i' :

.

~m~ _

24.1, 62- 64 24-1, 85- 89 24-1,102-104 24.1,136-138 24.2, 5- 7 24.2, 20-22 24.2, 38- 40 24-2, 56-58 !24.2, 68- 70 r24-2,80-82 ~4-2, 88- 90 24-2,115-117 24-3, 28- 30 24-3,48-50 24-3, 75- 77

199.63 55.504 199.86 55.424 ~- 200.03 55.433 ~ 200.37 55.457 I 200.58 55,470 200.71 55.480 200.89 55.493 201.07 55.507 201.19 55.519 201.31 55529 201.39 55.538 201.86] 55.561 ~ 55,629 202.49 55.646 202.76:55.662 /

E sample 24-1, 62- 64 24.1,85-8g 24.1,102-104 24-1,136-138 24-2, 5- 7 24.2, 20-22 24-2, 38- 40 24-2.56-58 24.2, 68- 70 24-2,80-82 24-2.88- 90 24-2,115-117 24.3, 28- 30 24.3,48-50 24-3, 75- 77

19g,63 19g.8e 200.03 200137 200.58 200.71 200,89 201,07 201.19 201.31 201.39 201,66 202,29 202.49 202.76

31 37 39 32 31 19 24 53 62 571 63 60 55 60 62

24 320 28 313 27 324 27 304 22 315 15 342 20 216 37 '349 41 3 ~ 40!320 42 321 ~ 42 312 41 3,32~ 42 316 42 317

20 13 . 20 30 50 59 64 25 ' 10 18

---;-

1

i~ i i

1

2

i

31 37 39 32 31 19 24 53 62 57 83 80 55 60 62

24 320i 283131 27 324 27 3 0 4 - - - - 22 315 15 342 20 216 37 349 41 344 40 320 42 321 42 3 1 2 - 41 332 42 316 42 317

7 3 4

23 4 6 1 2

2

~ 11

8

!6

18 19 1 3 2

6 15 2 1

t

=

a

1

26 2o 12 3 17 3 3 I t 19

~ ~ i~415 20

1

~, 2

6

15 14 29 7 1 4

12 9 7

t e 2s ~--e

4

1 t

t

12 29 33

1

3 1

1

1 2 1 1

2 1

11 2

3 6

2 1 1

:

!-53

1

1

J I "i, i ! "~176 i

3 4

3 6 5 7

2

~,i--e

I ' ! 55.504 55.424 55.4,33 55.457 55.470 55.480 55.493 55.507 55.519 55.529 55.538 55,581 55.629 55,646 55.662

20 21

12

7e

2

_

3 36 4 12 5

10 12 7 8

5 8 4

5 12 3

15 8 65 15 5 61 1 8 12 63 6 13 54 11 2 51 5 26 67 3 7 50 13 33 1 8 37 5 35 4 28 5 16 6 28 9 14 11 28

181 14 _._ 7 13 3 il

5 8 30 14 1 6 9 7 4 3 7 8 4 15 6 16

2

2 2

8

10 -1-7 4 18 2--~4 2 6 3 5 3 10 5 1 5 8

39 22 38 23 26

2

2 1 2

1 11 7 10 15 9 17 7

6 9 7 . . . . . . 6 5 __ 6 8 1 10 !

5 5 2 10 5 4~ 2

439

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

i~,

---L

i

.i i ! :

~

=

0i .ii; i 00i0 :

24 22 16 28 24 21

i

2t

3 5

1 -

5 2

1

1

1

i 1

7

4 3

6 2

3

22

3

t l

O,.

0,.

O,.

0.,

02

1

,

;

1

2 2

1

2 8 1 6

1 1

4 3

1

2

1

! 1

.~-~! .~1=, .-=...=~

1 2

1

4

~:

~

~

11 19 3 161

1

.=,1=

4 61 3 4 4 11

~

1 8 1 12

i ii '.r 9

19 35 23 33 20 1

2

"~

8,

i

20 14 12 11 4 8 34 1 2O 4

.~ ~

5 3 2 1

2

24 17 27 9 11 32

| ~! _

4 17 23 24 12

1

2

5 l

4

52 !

45

2

~

I I 3 37 25 21 49

I

1

1

~~

~-~ 1 2 11 i, 6 - - ,

......

_

2 1 4 3 12 4 5 5

4

6 19 13 16 14 9 13

1 1 1 4 2 23

1 4

9 9 4 3 4

!

1 5 4 4

1 1

1 1

15

2

-

21

3

51

8~.

42 ~33 6

.i_~., lo 11 1:@ ~

i

1

....

12

4 is --~ ~o

1

1

1

6[ 14

j__ el_! ~ [ 8[ 19

-

440

E. THOMAS •

~)cLQellep ')llnq O~q mlep ')llnq

O;,Lq (e-Ha) exe^uoo lq

3~t~ !euumlom "V

OeL~ e feuumlotu "V 0 ~ P !euueNotu V

OCt~ '.S~UJOp.pe~q OgL~'s.~o~H~ooeq "~) O~t~ 'tAdtuerul "N Om~ '.V~u~ml 'N

O~L~'snlmaoqu~! "0

~o~,,,.,~ Ogt9 "aas ee.P~Ol~,~

u~

<

~IdUJBS

N. J. SHACKLETON

PALEOCENE--EOCENE BENTHIC FORAMINIFERAL EXTINCTION

441

Deep water circulation in the Paleocene Ocean R I C H A R D M. CORFIELD 1 & RICHARD D. N O R R I S 2 1 Department of Earth Sciences, University of Oxford, Parks Road, Oxford OX1 3PR, UK 2 Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA

Abstract: We have compiled deep water benthic fil3c data from the Paleocene portions of several DSDP and ODP holes and present it using the new timescale of Berggren et al. (1995). Our data show that the north Atlantic hole DSDP 384 was the most positive site for fil3c in the late Cretaceous and the earliest Paleocene, suggesting that the sub-tropical north Atlantic was an important locus of deep water production during these intervals. Salinity and temperature comparisons do not support unequivocal deep water production by halothermal means in this region so we prefer to avoid the term Warm Saline Deep Water (WSDW) and employ instead the more neutral term 'palaeo-North Atlantic Deep Water' (palaeo-NADW). During K/T boundary time, the southern ocean apparently became the major producer of deep waters. Based on ~t3C comparisons both the North Atlantic and Southern Ocean were deep water producers during the early Paleocene to the late Paleocene interval. In the latest Paleocene (during the 'Paleocene carbon isotope maximum') Southern Ocean ~13C was most positive, supporting a Southern Ocean deep water source. The earliest Eocene ocean was characterized by deep water production in the high southern latitudes with well developed interbasinal fil3c gradients. ~180 data show an overall decrease from the late Cretaceous into the Early Eocene interrupted by an increase between 64 and 57 Ma. This is interpreted as an overall warming trend with a superimposed, previously undocumented, cooling phase in the early to late Paleocene.

The Paleocene was a time of global warmth and reduced latitudinal temperature gradients compared with the present day (e.g. Zachos et al. 1993, 1994). It also experienced elevated levels of atmospheric CO 2 (compared to the present day) that may have been at least partly responsible for its high global temperatures (Berner 1990; Arthur et al. 1991; Freeman & Hayes 1992). The relative warmth of the high latitudes has been used to infer that the formation of deep waters has occurred at least episodically at low latitudes in the Tethyan region (Kennett & Stott 1990) rather than the high latitudes as was the case for the Neogene and much of the later Paleogene (Miller et al. 1987; Woodruff & Savin 1989; Wright et al. 1991, 1992; Pak & Miller 1992). It is still unclear, however, whether Tethyan deep water sources occurred and if so how important they were to deep water circulation in the late Cretaceous and early Paleogene. Existing stable isotope data suggest that deep waters originated primarily in the Southem Ocean (e.g. Barrera et al. 1987; Corfield & Cartlidge 1992; Pak & Miller 1992). However, the data on which this inference is based are derived mostly from high southem latitude sites and may not be representative of the ocean as a whole. Equatorial data have, until now,

been available only from DSDP Hole 577. Hence until now, we have been unable to establish accurately the strength of latitudinal or interocean differences in carbon isotopic gradients during much of the Paleocene. In particular has been the need for isotope records from higher northern latitudes and additional equatorial sites of comparable palaeodepth to monitor comprehensively the movements of deep waters during the Paleocene.

Carbon isotopes as monitors of deep water circulation Carbon isotope measurements in benthic foraminifera can, in principle, be used to trace deep water circulation. The basis of this technique is that the ~13C composition of benthonic foraminifera must reflect the isotopic composition of surrounding water at the time of calcification provided that the influence of vital effects (i.e. species-specific non-equilibrium fractionation) are trivial, or otherwise understood, and appropriately adjusted for. Since deep waters become isotopically more negative with time after leaving the surface in the area of their formation due to the rain of organic

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 443-456.

443

444

R.M. CORFIELD t~ R. D. NORRIS

material that progressively adds t2C to the water mass, comparison of sites of similar palaeodepth in different areas can be used to suggest the location of deep water sources (e.g. Curry & Lohmann 1982, 1983; Duplessy et aL 1984; Shackleton et al. 1984b; Miller & Fairbanks 1985; Woodruff & Savin 1989) and the pathway taken by deep waters as they journey from their source areas. This technique has been applied recently to Paleogene sediments particularly because of suggestions (e.g. Chamberlin 1906; Brass et al. 1982; Bralower & Thierstein 1984; Kennett & Stott 1990, 1991; Zachos et al. 1993) that deep waters may have been formed by evaporative processes at low latitudes inducing salinity driven (halothermal) circulation, as opposed to the temperature driven (thermohaline) circulation characteristic of the late Neogene.

Cenozoic deep water circulation During most of the Neogene, deep waters have been formed at high latitudes (Woodruff & Savin 1989; Wright et al. 1991, 1992). Today, both northern component waters (NCW) and southern ocean waters (SOW) are formed by cooling of surface waters which overturn in the winter and sink along isopycnals. These waters initially have a positive 813C because they are formed from surface waters which are nutrient-depleted. This lack of nutrients limits the rate of organic carbon production and consequently the fractionation of carbon isotopes, thereby preventing significant buildup of 12C in dissolved CO 2. In particular carbon isotopic analyses and studies of benthic foraminifera suggest that deep waters have been formed in the Southern Ocean throughout the Cenozoic (e.g. Barrera et al. 1987; Miller et al. 1987; Wright et al. 1991, 1992; Thomas 1992). During the Holocene, SOW has flowed into the South Atlantic, Indian and Pacific Oceans where it either ages, as in the Pacific, or is mixed with NCW, as in the Atlantic and Indian Oceans. Deep waters have also been formed in both the North Atlantic and the Mediterranean. Significant NCW probably did not form until about 19 million years when deep Atlantic 813C first became enriched relative to 8~3C in the deep Pacific (Wright et al. 1991). Prior to this time, the only significant source of deep or intermediate waters other than SOW was probably the Mediterranean and shallow epicontinental seas of Tethys (Brass et al. 1982; Kennett & Stott 1990). To date, the evidence for the sources of Paleocene deep waters has been very incomplete and has suggested that the most 13C enriched waters were generally found in the vicinity of Antarctica (Corfield & Cartlidge 1992; Pak & Miller 1992;

Thomas 1992). This in turn implied a predominantly Southern Ocean source for deep waters during this interval. When addressing the question of whether Paleocene deep waters were formed in Tethys, comparisons with the Holocene are informative. Modern Mediterranean deep waters have a 813C composition of approximately 1.2%o. This is significantly lower than their source waters in the Atlantic (with 813C near 2.0%0 (Oppo & Fairbanks 1987)). However, Mediterranean outflow is still highly enriched in 813C relative to the 'old', nutrient enriched deep waters of the Pacific characterized by 813C near 0.0%0. It therefore follows, if our analogy of the Mediterranean with Tethys is sound, that, in the absence of isotopically positive deep waters (NCW) produced in the Norwegian Sea, significant inputs of Tethyan waters into the deep ocean should have increased the 813C of deep waters they were mixed with. When combined with young Southern Ocean waters, Tethyan outflow might be expected to have produced relatively homogeneous deep water 813C in low latitude oceans. A consequence of this is that we should expect relatively low latitudinal 813C gradients when Tethyan waters are an important source of deep or intermediate watermasses. Pak & Miller (1992) suggest that such episodes of minimal latitudinal gradients occurred in the Paleocene (c. 60 Ma) and the late Early Eocene (53-50Ma). Note however, that an alternative explanation for low latitudinal 813C gradients is that oceanic nutrient levels and hence productivity were depressed. Hence, deep waters would apparently 'age' little as they flowed away from their source areas due to reduced surfaceto-deep 812C flux. These effects notwithstanding, the 813C signature of Tethyan outflow should be particularly pronounced adjacent to the ends of the Tethyan seaway in the NE Atlantic and the Northern Indian Ocean. If Tethyan outflow were an important source of deep waters than we would expect to see higher 813C in both these areas compared to the Pacific sites. Additionally, because the Pacific was a very large reservoir during the Paleocene, the effects of watermass aging (i.e. 12C enrichment) should be pronounced compared to waters near the Tethyan basins. Finally, even if Tethyan waters were never the primary source of deep water, their contribution to ocean circulation was likely to be substantially greater during the Paleocene than that currently supplied by the relatively small modem Mediterranean.

Methods We have assembled benthic 8)3C and 8180 data from sites of broadly equivalent palaeodepth

PALEOCENE DEEP WATER CIRCULATION Table 1. Palaeodepths of holes discussed in this paper Hole

Palaeodepth

DSDP 384 DSDP 525 DSDP 527 DSDP 577 ODP 690b ODP 690c ODP 738 ODP752

3000 1100 2700 2400 2040 1600-1700 1000-1200

Reference

Zachos & Arthur (1986) Zachos & Arthur (1986) Zachos & Arthur (1986) Zachos & Arthur (1986) Kennett & Stott (1990) Kennett & Stott (1990) Coffin et al. (1992) Driscoll et al. (1991)

(Table 1). All measurements are derived from

Nuttallides truempyi except for data from ODP 752 and some data from DSDP Hole 384 where Cibicidoides was used. Figure 1 shows the palaeopositions of the holes studied. Samples run in Oxford (some data from DSDP 527, DSDP 577 and ODP 690B) were analysed using a VG PRISM mass spectrometer with on-line common acid bath system. Reproducibility of the Oxford in-house NOCZ standard was better than 0.1%o. NOCZ standard is calibrated against NBS 19 on a regular basis and shared with other labs worldwide. Oxford is calibrated against the stable isotope labs at University of Cambridge (UK), University of Michigan (USA), University of Southern California (USA), University of Gothenburg (Sweden) and Woods Hole Oceanographic Institution (USA). Additional calibration against Woods Hole was achieved by reproducing some measurements on Miocene planktonic foraminifera. Data congruency on these samples also exceeded the 0.1%o requirement. Samples run in Woods Hole (DSDP 384) were analysed using a Finnigan MAT 251 with on-line individual acid bath system ('Kiel device'). Reproducibility of NBS 19 at Woods Hole is also better than 0.1%o. Data sources (including those from the literature) are listed in Table 2. We define the intermediate/deep water interface as occurring at about 2000-2500 m. Therefore most of the sites discussed herein can be considered upper deep water to lower intermediate water.

Results Figure 2 illustrates 813C data from DSDP Holes 384, 525, 527, 577 (Leg 74 sites), and ODP Holes 69013 and ODP 690C (Leg 113 sites) plotted using the timescale of Berggren et al. (1995). Figure 3 illustrates the 8180 data from the same sites. To avoid overly congesting these figures we have plotted only the more abundant Stensioina beccariformis data from DSDP Hole 384. This does not affect our interpretation, since the S. beccariformis

445

data and the N. truempyi data in this hole are entirely coincident. Figures 4 and 5 illustrate additional data from the Indian Ocean (discussed below), as well as the N. truempyi data from DSDP Hole 384. The 813C data can be summarized using eleven discrete stages to describe the evolution of Paleocene and early Eocene deep water circulation, similarly 8180 data can be summarized in seven stages. Although these 813C and 8180 stages are not synchronous the two data sets are complimentary and hence are discussed in tandem.

The late Cretaceous: 67-65 Ma (fil3c stage 1 and •180 stage I) DSDP 384 exhibits the most positive 813Cbenthic values. The high southem latitude holes on the Maud Rise (Leg 113 data) are more negative and the equatorial Pacific Hole DSDP 577 is the most negative. This implies that the youngest water occurs in the western sub-tropical Atlantic. Note also that the Atlantic-Pacific interbasinal isotopic gradient is very steep at this time (c. 2%0), implying that productivity was relatively high and sufficient to resolve deep water aging patterns in detail. The 8180 record (Fig. 3) suggests that the temperature of deep waters at DSDP 384 was 2~ higher than either high southern latitude or Pacific deep waters at this time (assuming that all deep water salinities were the same). Note however that it is likely that DSDP 384 experienced 8180 enrichment, if this site, as suggested, was adjacent to a locus of palaeo-NADW formation, hence the estimated palaeotemperatures for this site will be too low.

The K/T boundary interval: 65.6-64.6 Ma ( ~ 3 0 stage 2 and ~ 8 0 stage 2) Immediately prior to the K/T boundary, 813C values in DSDP Hole 577, the Maud Rise holes (Leg 113), the Walvis Ridge holes (Leg 74) and ODP Hole 738 increase and converge. This change is not tracked in DSDP 384 which suggests that the change reflects a transient reorganization of deep water circulation rather than a fractionation phenomenon involving the whole of the deep water 813C reservoir. The implication of this reorganization from the convergence in 813C values, is that the high southern latitudes became more important than the lower latitudes as a source for deep waters at this time. Coincident with this enrichment in 813C is a decrease in 8180 in the Maud Rise, Newfoundland Ridge, Walvis Ridge and Shatsky Rise sites. This event briefly eliminated the thermal contrast between the ocean basins.

446

R . M . CORFIELD & R. D. NORRIS

o,J .,..~

o .,..~

..= o

E

PALEOCENE DEEP WATER CIRCULATION Table 2. Data sourcesfor the isotope stratigraphies discussed in thispaper DSDP 384 DSDP 525 DSDP 527 DSDP 577 ODP 690B ODP 690C ODP 738 ODP 752

R.D. Norris, Woods Hole Shackleton et al. (1984a) Shackleton et al. (1984a) Corfield & Cartlidge (1992); Pak & Miller (1992) Stott et al. (1990); Corfield & Cartlidge (1992) Stott et al. (1990) Barrera & Huber (1991) Seto et al. (1991)

The aftermath of the K/T boundary: 65-64 Ma (t~3C stage 3 and ~ 8 0 stage 3) After the 813C increase near the K/T boundary values rapidly decrease in all sites except for DSDP 384 which also failed to register the 813C increase immediately prior to the K/T boundary. This nearglobal decrease is more rapid and of greater amplitude than the signal from DSDP 384 with the consequence that DSDP 384 has the heaviest values of all sites in the earliest Paleocene. This implies that once again the predominant source of deep waters was near the sub-tropical North Atlantic. 8180 values from DSDP 384 are once again more negative than all but the Indian Ocean sites, suggesting that palaeo-NADW was being produced in or near the sub-tropical North Atlantic.

The early to late Paleocene: 64-57.4 Ma (~3C stages 4-5, ~ 8 0 stages 4-5) During 813C stage 4 between 64.2 and 59.5 Ma interbasinal 813C gradients were relatively low. This implies that there was a comparatively weak south to north aging gradient in the Atlantic and hence that the Southern Ocean was a source area for deep waters and that the sub-tropical North Atlantic was at least adjacent to a deep water source area (such as Tethys). 8180 data suggest two main stages in the ocean thermal evolution during this interval; stage 4a during which time the North Atlantic was warmer than all other sites (consistent with its suggested position adjacent to a warm deep water production source area) and the equatorial Pacific, the South Atlantic and the southern high latitudes were coolest; and stage 4b when the high southern latitudes became cooler than the equatorial Pacific. 8180 stage 4b overlaps with 813C stage 5. 813C stage 5 (the interval between 59.6 and 57.4 Ma) is a transitional one in the history of benthic 813C change during the

447

Paleocene. Although data coverage is lacking in the Leg 113 holes, at this time they appear to have started becoming more positive than North Atlantic Hole DSDP 384. Additionally, the Atlantic and Pacific sites become more positive than the Indian Ocean sites. In general latitudinal 813C gradients increase at this time suggesting an increase in the importance of the Southern Ocean as the predominant source of deep waters at this time and the decrease in importance of other (i.e. low latitude) source areas.

The prelude to the end of the Paleocene: 57.4-55.4 Ma (~3C stage 6 and t~So stage 5) During this interval the Leg 113 holes show the most positive 813C values. The south Atlantic shows values are only slightly lighter. This implies that in the latest Paleocene the predominant source of deep water production has become the high southern latitudes. North Atlantic values are intermediate between Southern Ocean/South Atlantic values and Pacific values which is consistent with an area not producing deep waters but rather receiving waters generated elsewhere via a smaller ocean basin with less aging potential than the IndoPacific. Stage 5 of the 8180 record marks the beginning of the 8180 depletion so characteristic of the late Paleocene and early Eocene interval (e.g. Miller et al. 1987; Corfield & Cartlidge 1992); also noteworthy at this time is the convergence of the Leg 113 record with those of other sites. Hence the data suggest that all ocean basins were filled with waters of apparently similar temperature further supporting the concept os single deep water source during this interval. Superimposed on the Pacific signal are at least two short negative excursions in 813C and 8180 at 57.6 and 56.4 Ma. These may represent brief periods when warm, nutrient-rich waters entered the Pacific, perhaps through renewed deep water production in low latitude basins.

The Paleocene-Eocene boundary: 55.4-55 Ma (~3C stage 7 and ~ 8 0 stage 6) This event has been associated with a temporary change from thermohaline to the halothermal mode of deep water circulation which caused profound benthic foraminiferal extinctions in the marine realm (Thomas 1991, 1992). It may also have been coupled with a transient climate change (Stott 1992). Figure 2 shows that both the Leg 113 holes and DSDP 577 register approximately (within age control uncertainty) synchronous negative 8t3C

448

R.M. CORFIELD & R. D. NORRIS

,.,-:.,

e~0

9

o

o

,.t::

.e E o

PALEOCENE DEEP WATER CIRCULATION

449

,,.:,

"4 ca)

~e 0.)

o

t,Q ..g::

.e o

450

R.M. CORFIELD t~ R. D. NORRIS

excursions superimposed on the long term late Paleocene-early Eocene decline. These are contemporaneous with negative excursions in the 5180 record shown in Fig. 3.

The aftermath of the Paleocene-Eocene boundary: 54.8-53.4 Ma (~13C stage 8 and ~ 8 0 stage 7) Following the very negative 613C values that characterize the P/E boundary 'event' in the Leg 113 holes values increase and as in 513C stage 6 are again similar to values from the Leg 74 holes. Once again a steep latitudinal gradient exists between Atlantic sites and Indo-Pacific sites suggestive of a predominant Southern Ocean source for deep waters. However, the DSDP 384 record does not extend into this interval, precluding inferences concerning Tethyan deep water production. Note that during this stage 513C values temporarily plateau from the overall long term late Paleocene to early Eocene 513C decline. The 5180 record suggests a temperature increase followed at the end of 5180 stage 7 by a transient temperature decline (the trajectories of 5180 change are marked with arrows on Fig. 3). The temperature increase phase of 5180 stage 7 is characterized by shallow latitudinal 5180 gradients, suggesting homogenous ocean temperatures, consistent with the idea of a single deep water source. The temperature decline at the end of ~5180 stage 7 marks the beginning of an increase in latitudinal 6180 gradients with DSDP 577 values becoming progressively more negative than Leg 113 values.

The earliest Eocene, phase 1: 53.4-51 Ma (~3C stage 9 and Ffl80 stage 8) This interval is characterized by a resumption in the late Paleocene to early Eocene 5~3C decline. Although the latitudinal 5~3C gradient remains the same, all three areas where we have data show a pronounced negative trend. This suggests that the Southern Ocean was still the major producer of deep waters. 8180 values in DSDP 577 continue to diverge from those of the Leg 113 holes.

The earliest Eocene, phase 2: 53.4-51 Ma (~3C stage 10 and ~ 8 0 stage 8) 5~3C stage 10 shows the start of the post late Paleocene-early Eocene 513C recovery. 8180 values in DSDP 577 continue to decline, whereas those in the Leg 113 and Leg 74 holes do not.

Discussion

Sources of error and uncertainty in benthic ~3C comparisons Compilations of benthic 813C data such as that attempted here are limited by several factors. Chief among these is the uncertainty of the timescale for each site upon which comparisons of inter-site 613C compositions are completely reliant. Second is the choice of species of benthic foraminifera analysed, and the application (if any) of a nonequilibrium fractionation correction. Third is the relative palaeodepth of the sites and fourth is the precision with which the isotopic analysis is made. In this contribution, the most critical site yielding evidence for the occurrence of palaeo-NADW in the Paleocene ocean is DSDP 384. The magnetochronology for this site has been recently revised (M. Van Fossen, pers. comm.) and is currently at its most refined. The magnetobiochronologies of the other sites are derived from the relevant DSDP/ODP volumes and have not been reinterpreted here. We have however computed our sample ages using the recent revision of the Paleogene timescale of Berggren et al. (1995). This allows direct comparison with other papers in this volume and may also serve to delay the inevitable redundancy characteristic of all papers that use geological timescales. Note also that although absolute ages may vary, relative temporal differences are comparatively robust to different timescales. The benthic foraminifera Nuttallides and Cibicidoides have been exclusively used in this study, thereby minimizing artifacts owing to the analysis of different species. No correction for presumed non-equilibrium fractionation has been applied. Some authors (e~g. Shackleton et al. 1984a) have argued that certain species fractionate out of equilibrium with ambient seawater. While we do not dispute this, we have elected to make no correction in this case, since our compilation is entirely dependent on relatively small isotopic differences. It is self-evident that our argument hinges on relative differences rather than absolute values. Although the range of palaeodepths of the sites analysed is 2 km, we do not consider that this will affect our interpretation. Kroopnick (1985) has shown that the vertical 813C gradient below about 1 km is only 0.2 in the Atlantic Ocean and 0.4%0 in the Pacific. It follows, therefore, that the interval of our comparison most at risk from this potential variability is between about 63 and 59 Ma or the later part of carbon isotope stage 4 in our terminology. We have interpreted this shallow gradient to imply that both the Southern Ocean and the North Atlantic were producing deeper waters

PALEOCENE DEEP WATERCIRCULATION at this time. It may be that our conclusion for this discussed interval is more tentative than for other periods in our compilation. However the late Cretaceous and late Paleocene-early Eocene intervals in particular are characterized by relatively steep interbasinal gradients and hence are unlikely to be affected by this problem. The interlaboratory calibration between Woods Hole and Oxford has been discussed above and yields identical results in replicate analysis of foraminiferal samples. We therefore consider that the patterns of isotopic change and implied deep water circulation that we have discussed are defensible. We conclude that deep water formation in the North Atlantic was a palaeoceanographical feature of the late Cretaceous through early to late Paleocene interval. A counterpoint to the production of deep waters in the North Atlantic was the production of SOW near the Antarctic which assumed progressively more importance as the Paleocene progressed. Indian O c e a n intermediate w a t e r

Although benthic 8t3C and 8z80 data are available from ODP Holes 738C and 752 they have not featured in the above comparisons. The reasons for this are discussed below but for comparative purposes the 813Cand 5180 record of these sites are plotted on the same timescale as the holes already discussed and are shown in Figs 4 and 5. Indian Ocean hole ODP 752 has negative 813C compared to the North Atlantic and Southern Ocean throughout most of the Paleocene. Only during the interval between 59.5 and 61 Ma does the 813C of this hole converge with that of holes in the other oceans. As ODP 752 was comparatively shallow at an intermediate water depth of only 1200 mbsl during the Paleocene, its negative 813C probably reflects the presence of nutrient-rich intermediate waters in the southern Indian Ocean rather than genuinely depleted deeper waters. Indian Ocean Hole ODP 738C displays relatively negative 813C only between 57.2 and 59 Ma. Like ODP Hole 752, ODP Hole 738C was at intermediate water depths (c. 1800 m) during the Paleocene. Although the record from ODP 738A is of low resolution in the earlier part of the Paleocene, its 813C history is generally similar to results from the Southern Ocean. ODP Hole 738C also displays much more positive ~13C than ODP 752 between 57.6 and 56.4 Ma. Apparently, the nutrient-enriched waters covering ODP 752 reached as far south as the ODP Hole 738C at the southern Kerguelen Plateau only briefly during the middle late Paleocene (60-57.2 Ma). Nutrient-rich intermediate waters could either originate as outflow from Tethys as does strongly

451

13C-depleted water from the modern Arabian Sea (Kroopnick 1985), or as recirculated water from the Pacific. Hole 752 was located near 50~ latitude, and Hole 738C was adjacent to Antarctica before about 60 million years. Today, intermediate waters at these locations are derived almost entirely from Antarctic Intermediate Water, which is a nutrientdepleted, high oxygen watermass at these latitudes. The nutrient-enriched character of the Paleocene intermediate water in the southern Indian Ocean suggests that these waters were fed from a relatively distant source such as the Tethys or as intermediate water advected out of the Central Pacific. If these intermediate waters do represent waters that have transited the Pacific, then their 813C values reflect their passage across what was probably the strongest aging gradient present in the Paleocene ocean. The record of these sites further suggests that deep waters were not always formed adjacent to Antarctica in the Indian Ocean sector even though Hole 738C is at the same latitude as Hole 690B - the hole that provides the strongest evidence for a Southern Ocean source for deep water. W S D W in the N o r t h Atlantic ?

We can estimate the temperature and salinity variation of deep waters if we assume that all deep waters had similar density as they largely do today (Railsback et al. 1989; Mead et al. 1993). This is reasonable for the late Cretaceous and younger oceans because of the deep water connection between the North Atlantic and the Southern Ocean. Note however, that if the North and South Atlantic were not in communication, then our assumption of constant density of deep waters in the North Atlantic and the Southern Ocean will be wrong, and consequently our estimates of temperature and salinity gradients will be inaccurate. Our calculation of North Atlantic deep water temperature and salinity assumes a correction of 1.2%o for 5180(PDB) under ice free conditions (Mead et al. 1993), an average deep ocean salinity of 34%o for non-glacial deep water (Railsback et al. 1989), and average measured 5180 for the deep Pacific of 0.75%o taken from the Cretaceous record at DSDP 577. Assuming an average deep Pacific salinity of 34%o and a measured upper Cretaceous, deep Pacific Nuttallides 5180 value (in this case, since the absolute value is required, we have corrected for disequilibrium fractionation using a coefficient of-0.4%0) of 1.05%o, we suggest that deep Pacific and Southern Ocean waters had a density of 26.80 in the Cretaceous and temperatures of about 5.5-6~ The temperature and salinity of North Atlantic waters assuming a similar density can be estimated from the 5180 record at DSDP 384.

452

R . M . CORFIELD & R. D. NORRIS

c',i

9

9 '-6

9 ;-:q

E o

"5 ~ tt5

~D ,-&

o

~5 c~

co 6,o .,,.~

. ,...~

o

r,.)

453

PALEOCENE DEEP WATER CIRCULATION

e.i

"o

9 "6

9 "O

g: . ....~

o'x

%

O9

"~_,,

O

9 o~

..,.~

o

454

R.M. CORFIELD r R. D. NORRIS

Results of this calculation suggest that the Cretaceous deep North Atlantic was warmer by c. 2.7~ and more saline by c. 0.45%o than contemporaneous Southern Ocean waters. Hence a latitudinal temperature gradient of 2.5~ and a salinity gradient 0.55%0 can be calculated for the early Paleocene assuming that the deep Pacific had a salinity of 34%0 and 5180 corrected for disequilibrium of 0.55%o. These temperature and salinity estimates are robust to variation in the estimated mean ocean salinity. For example, correction of mean ocean salinity for the effects of the Messinian salinity crisis (Mead et aL 1993) suggests that mean ocean salinity may have been as much as 36.17%o. This is probably an overestimate since it assumes that ocean salinity is not in dynamic balance and that any salt removed from the system is not replaced on long timescales. Still, calculated temperature and salinity gradients between the North Atlantic and Southern Ocean assuming this somewhat higher mean ocean salinity range from 3.5-3.0~ and 0.7-0.85%o respectively for the late Cretaceous and early Paleocene. The estimated temperature and salinity gradients in the Cretaceous and early Paleocene deep oceans are similar or slightly higher than those between modem North Atlantic Deep Water and Antarctic Bottom Water (presently c. 3~ and 0.26%o). In contrast, modem source waters from marginal basins such as the Red Sea and Mediterranean have salinities of 36--40.5%o and temperatures of 1012.8~ (Thunell et al. 1987, 1988) Were the deep North Atlantic largely filled with such waters, the salinity and temperature gradient with the Southern Ocean would be > 1.3%o and > 10~ The modest calculated temperature and salinity differences between the North Atlantic and other ocean basins suggest therefore that waters filling the deep North Atlantic in the Cretaceous and early Paleocene cannot have been formed entirely from the overflow of marginal seas, because they would not readily have sunk. Hence the mode offormation of these apparent Cretaceous and Paleocene North Atlantic deep waters remains enigmatic. They may have been recharged by overflow of relatively warm, saline marginal seas which cooled and were diluted by

mixing with intermediate and surface waters flowing in from the South Atlantic. Modem North Atlantic Deep Water (NADW) is formed partly from the entrainment and cooling of Mediterranean outflow water. However, NADW is also strongly chilled by overflow of cold waters from the Norwegian--Greenland Sea. Polar temperatures were probably relatively modest in the Cretaceous and Paleocene based on Southern Ocean deep water temperatures of 5.57.5~ It is not clear that these high latitude temperatures were low enough to have led to extensive deep water production by surface ocean cooling. Still, the combination of surface cooling and saline outflow of the Tethys or other marginal seas around the Atlantic may have been sufficient to form deep water in the North Atlantic since all that was needed were the conditions necessary to create a density difference between surface and deep waters.

Summary The late Cretaceous and earliest Paleocene North Atlantic was a net exporter of deep water to the world ocean. The mechanism of the formation of such water remains enigmatic although simplistic interpretations of WSDW in Tethys are not supported by salinity/temperature comparisons of this palaeo-North Atlantic Deep Water (palaeoNADW) with Southern Ocean Water (SOW). Similarly, it is unlikely that late Cretaceous and earliest Paleocene polar temperatures were low enough to form deep waters by thermohaline processes. We therefore currently favour a mechanism for palaeo-NADW production that combines surface cooling and saline outflow from Tethys to produce slightly more dense waters in the subtropical north Atlantic. This outflow gradually decreased in importance through the Paleocene, being supplanted by the Southern Ocean as the major producer of deep waters. We thank Julie Cartlidge and Lu Ping Zou for technical assistance in Oxford and Woods Hole. We also thank W. A. Berggren, D. W. Oppo and two anonymous reviewers for comments on an earlier version of the manuscript. RDN's work on DSDP Site 384 supported by NSF Grant OCE-9115436.

References ARTHUR, M. A., HINGA, K. R., P1LSON, M. E. Q., WHITAKER, E. & ALLARD, D. 1991. Estimates of PCO2 for the last 120 Ma based on the 13C of marine phytoplankton organic matter. Eos, 72, 166. BARRERA, E. B. & HUBER, B. T. 1991. Paleogene and

early Neogene oceanography of the southern Indian Ocean: Leg 119 foraminifer stable isotope results. In: BARRON,J., LARSEN,B. Yr At. Proceedings of the Ocean Drilling Program, Scientific Results, 119, 693-717. - - , - - , SAVIN,S. M. & WEBB,P.-N. 1987. Antarctic

PALEOCENE DEEP WATER CIRCULATION marine temperatures: Late Campanian through early Paleocene. Paleoceanography, 2, 21-47. BERGGREN, W. A., KENT, D. V., SWISHER, C. C. Ill & AUBRu M.-P. 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A., KENT, D.V., AUBRY, M.-P. & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation; Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Publication, 54. BERNER, R. 1990. Atmospheric carbon dioxide over Phanerozoic time. Science, 249, 1382-1386. BRALOWER, T. J. & THIERSTEIN, H. R. 1984. Low productivity and slow deep-water circulation in mid-Cretaceous oceans. Geology, 12, 614-618. BRASS, G. W., SOIJTHAM,J. R. & PETERSON,W. H. 1982. Warm saline bottom water in the ancient ocean. Nature, 296, 620--623. CHAMBERLIN,T. C. 1906. On a possible reversal of deep sea circulation and its influence on geologic climates. Journal of Geology, 14, 363-373. COFFIN, M. E, GAHAGAN, L. M., LAWVER, L. A., LEE, T.-Y. & ROSENCRANa~,E. 1992. Atlas of Mesozoic/ Cenozoic reconstructions (200 Ma to present day). Plates Progress Report. University of Texas, Institute of Geophysics Technical Report, 122, 1-192. CORFIELD,R. M. & CARTLIDGE,J. E. 1992. Oceanographic and climatic implications of the Palaeocene carbon isotope maximum. Terra Nova, 4, 443-455. CURRY, W. B. & LOHMANN,G, P. 1982. Carbon isotopic changes in benthic foraminifera from the western South Atlantic: Reconstruction of glacial abyssal circulation patterns. Quaternary Research, 18, 218-235. & - 1983. Reduced advection into Atlantic Ocean deep eastern basins during last glaciation maximum. Nature, 306, 577-580. DRISCOLL,N. W., KARNER,G. D. & WEISSEL,J. K. 1991. Stratigraphic response of carbonate platform and terrigenous margins to relative sea level changes: are they really that different? In: WEISSEL,J., PEIRCE J., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 121,743-762. DUPLESSY, J-C., SHACKLETON,N. J., MATHtEWS, R. K., PRELL, W., RUDDIMAN,W. E, CARALP,M. & HENDY, C. H. 1984. 813C record of benthic foraminifera in the last interglacial ocean: implication for the carbon cycle and the global deep water circulation. Quaternary Research, 21,225-243. FREEMAN, K. H. & HAYES, J. M. 1992. Fractionation of carbon isotopes by phytoplankton and estimates of ancient CO 2 levels. Global Biogeochemical Cycles, 6, 185-198. KENNETr, J. E & STO't'r, L. D. 1990. Proteus and ProtoOceanus: Ancestral paleogene oceans as revealed from Antarctic stable isotope results; ODP Leg 130. In: BARKER,P. F. KENNETF,J. P. ETAL. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 829-848. -& -1991. Abrupt deep sea warming, paleoceanographic changes and benthic extinctions at the end of the Palaeocene. Nature, 353, 225-229.

455

KROOPNICK, P. 1985. The distribution of ~13C of TCO 2 in the world oceans, Deep Sea Research, 32, 57-84. MEAD, G. A., HODELL,D. A. & CIESELSKI,P. F. 1993. Late Eocene to Oligocene vertical oxygen isotope gradients in the S. Atlantic; implications for warm saline deep water. In: KENNEWT,J. P. & WARNKE, D. A. (eds) The Antarctic Paleoenvironments; a Perspective on Global Change. Part two. American Geophysical Union, Antarctic Research Series, 60, 27-48. MILLER, K. G. & FAIRBANKS, R. G. 1985. Oligocene to Miocene benthic foramininifera and abyssal circulation changes. In: SUNDQUIST, E. T. & BROECKER, W. S. (eds) The Carbon Cycle and Atmospheric C02: Natural Variations Archean to Present. American Geophysical Union, Washington DC, Geophysics Monograph Series, 32, 469-486. --, JANACEK, T. R., KATZ, M. E &. KErL D. J. 1987. Abyssal circulation and benthic foraminiferal changes near the Palaeocene/Eocene boundary. Paleoceanography, 2(6), 741-761. OPPO, D. & FAmBANKS,R. 1987. Variability in the deep and intermediate water circulation in the North Atlantic. Earth and Planetary Science Letters, 86, 1-15. PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene benthic foraminiferal isotopes and assemblages: Implications for deep-water circulation. Paleoceanography, 7, 405-422. RAILSBACK, L. B., ANDERSON, T. E, ACKERLY, S. C. & CISNE, J. L. 1989. Paleoceanographic modelling of temperature-salinity profiles from stable isotopic data. Paleoceanography, 4, 585-591. SETO,K., NOMURA,R. & NUTSUMA,N. 1991. Data Report: Oxygen and carbon isotope records of the upper Maastrichtian to lower Eocene benthic foraminifers at Site 752 in the eastern Indian Ocean. In: WEISSEL, J., PEIRCE, J., gr AL Proceedings of the Ocean Drilling Program, Scientific Results, 121, 885-890. SHACKLETON,N. J., HALL, M. A & BOERSMA,A. 1984a. Oxygen and carbon isotope data from Leg 74 foraminifers. In: MOORE, T. C. JR, RABINOWlTZ, P. D., ETAL. Initial Reports of the Deep Sea Drilling Project, 74, 599-612. --, IMBRIE, J. & HALL, M. A. 1984b. Oxygen and carbon isotope record of East Pacific core V19-30: implications for the formation of deep water in the late Pleistocene North Atlantic, Earth and Planetary Science Letters, 65, 233-244. Sxoa~r, L. D. 1992. Higher temperatures and lower oceanic PCO2: a climate enigma at the end of the Paleocene epoch. Paleoceanography, 7, 395-404. ~, KENNETr, J. P., SHACKLETON,N. J. & CORFmLD, R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: Inferences from the stable isotopic composition of planktonic foraminifers. In: BARKER,P. E, KENNEI'T,J. P. erAt.. Proceedings of the Ocean Drilling Program, Scientific Results, 113, 849-864. THOMAS, E. 1991. The latest Palaeocene mass extinction of deep sea benthic foraminifera: Result of global; climate change. Geological Society of America, Abstracts with Programs, 23, A141.

456

R.M. CORFIELD •

1992. Middle Eocene-Late Oligocene bathyal benthic foraminifera (Weddell Sea): faunal changes and implications for ocean circulation. In: PROTHERO,D. A. & BERGGREN,W. A. (eds) EoceneOligocene Climatic and Biotic Evolution. Princeton University Press, 245-271. THUNELL, R. C., WmUAMS, D. E & HOWELL, M. 1987. Atlantic-Mediterranean water exchange during the late Neogene. Paleoceanography, 2, 661678. , LOCKE, S. M. & WILLIAMS, O. F. 1988. Glacioeustatic sea-level control on Red Sea salinity. Nature, 334, 601-604. WOODRUFF, F. t~ SAVIN S. M. 1989. Miocene deepwater oceanography. PaIeoceanography, 4(1), 87140. WRIGHT, J. D., MILLER, K. G. & FAIRBANKS,R. G. 1991. Evolution of modern deepwater circulation:

R. D. NORRIS evidence from the late Miocene southern ocean.

Paleoceanography, 6(2), 275-290. , -& -1992. Early and middle Miocene stable isotopes: implications for deepwater circulation and climate. Paleoceanography, 7, (3), 357-389. ZACHOS, J. C. & ARTHUR,M. A. 1986. Paleoceanography of the Cretaceous/Tertiary boundary event: inferences from stable isotopic and other data. Paleoceanography, 1, 5-26. ~, LOHMANN,K. C., WALKER,J. C. G. & WISE, S. W. 1993. Abrupt climate change and transient climates during the Paleogene: A marine perspective. Journal of Geology, 101, 191-213. ~, STOTr, L. D. & LOHMANN,K. C. 1994. Evolution of early Cenozoic marine temperatures. Paleoceanography, 9(2), 353-387.

Early Eocene palaeoceanography and palaeoclimatology of the eastern North Atlantic: stable isotope results for DSDP Hole 550 S T E L L A D. C H A R I S I & B I R G E R S C H M I T Z

D e p a r t m e n t o f Marine Geology, Earth Sciences Centre, University o f G6teborg, S-413 81 GOteborg, Sweden Abstract: High-resolution oxygen and carbon isotope records for benthonic and planktonic

foraminifera have been established through the early Eocene at DSDP Site 550, Goban Spur in the North Atlantic. Benthonic stable isotopic records from Site 550 (estimated palaeodepth c. 4000 m) represent the first documentation of actual deep water properties during the early Eocene. At Site 550 in the earliest Eocene (end of NP10 Zone), a 0.5%o vertical 5180 (i.e. thermal) gradient between subsurface and bottom water was eliminated, probably reflecting upward displacement of subsurface water by intermediate and deep water. Thereafter vertical isothermal conditions prevailed for the rest of the early Eocene. This is consistent with earlier findings suggesting unusually low vertical and latitudinal temperature gradients in the early Eocene ocean. A comparison of the isotopic records from Site 550 and nearby DSDP Site 401 (palaeodepth 1800 m), reflecting intermediate water properties, confirms the isothermal nature of the water column in the eastern North Atlantic. This is in contrast to earlier findings for the Antarctic region (ODP Sites 689 and 690) in the early Eocene, where supposedly warm bottom waters originating in low latitudes were overlain by cooler intermediate waters originating in high latitudes. Both the oxygen and the carbon isotopic records for Sites 550 and 401 indicate that the eastern North Atlantic region was influenced by a single source of intermediate-to-deep water through most of the early Eocene. The benthonic 5180 and 813C values for Site 550 are generally low compared with coeval sections worldwide, indicating that during most of the early Eocene the North Atlantic was dominated by warm and nutrient-enriched deep water formed in a low latitude region. The earliest stages of the long-term cooling trend, which characterizes the second part of the Paleogene, can be traced by a 2~ decrease in bottom water temperatures across the NPI3-NP14 transition at both Sites 550 and 401.

Palaeoclimatic reconstructions based on stable isotopic studies show that the early Eocene was the warmest period in the Cenozoic (e.g. Savin 1977; Zachos et al. 1993; Corfield 1994). The period is characterized by reduced latitudinal thermal gradients (Wolfe 1978; Shackleton & Boersma 1981; Shackleton 1984; Boersma et al. 1987) and high polar ocean temperatures (e.g. M c K e n n a 1980; Stott et al. 1990; Barrera & Huber 1991). A significant decrease in wind intensity, related to low thermal gradients, has also been documented (Janecek & Rea 1983; Rea et al. 1990; Hovan & Rea 1992). The oceans provide an excellent vehicle for poleward heat transport, and a shift in the source region for global deep water from high to low latitudes may explain the warm and equable early Eocene climate (Miller et al. 1987; Barron & Peterson 1991; Kennett & Stott 1991; Pak & Miller 1992; Zachos et al. 1993; Corfield 1994). The earliest Eocene in the northeastern Atlantic region is characterized by intense volcanic activity associated with major tectonic reorganization,

including the opening of the Norwegian-Greenland Sea (Miller & Tucholke 1983; Knox & Morton 1988; Thiede et al. 1989; Eldholm & Thomas 1993). It is uncertain at what time incipient surface and/or deep water connections were established between the North Atlantic and the Arctic Ocean via the water passages east and west of Greenland. However, step-wise cooling events in bottom water temperatures (Kennett & Barker 1990; Zachos et al. 1993) occurred in the uppermost part of the early Eocene, and continued through the rest of the Paleogene, reflecting the successive transition from an early Paleogene ocean characterized by ice-free halothermal conditions to the glacial thermohaline Neogene ocean, ultimately leading to modern bi-polar deep water circulation patterns in the North Atlantic. Site 550 is the deepest site (4432 m present depth), recovered along the Goban Spur transect of DSDP Leg 80 in the North Atlantic (de Graciansky, Poag et al. 1985). Palaeodepth estimates, based on backtracking from an empirical subsidence curve for the North Atlantic (Tucholke & Vogt 1979),

From Knox, R. W. O'B., Corfield, R. M. & Dunay, R. E. (eds), 1996, Correlation of the Early Paleogene in Northwest Europe, Geological Society Special Publication No. 101, pp. 457--472.

457

458

s . D . CHARISI •

suggest that during the early Eocene water depths at Site 550 were 3800--4000 m (see also Masson et al. 1985). This indicates that Site 550 is one of the deepest recovered across the early Eocene, worldwide, and offers the opportunity to trace more directly deep water circulation patterns, in contrast to most earlier studies that have been limited to intermediate and upper deep water modes of circulation (e.g. Barrera & Huber 1991; Kennett & Stott 1991; Pak & Miller 1992; Corfield & Norris 1996). In this contribution we present high resolution oxygen and carbon isotope records for the early Eocene at Site 550. The vertical stratification of the water column in the eastern North Atlantic is portrayed by stable isotope records from this deepwater site as well as from nearby Site 401, with a palaeodepth of 1800 m, reflecting intermediate water properties (data from Pak & Miller 1992). Global intermediate-to-deep water circulation patterns between various oceanic basins can be determined by comparing bottom water isotopic records from Sites 550 and 401 with others from the world's oceans for the early Eocene.

Samples and methods Samples have been obtained at c. 1 m intervals, which corresponds to approximately 20 00060 000 years resolution, depending on sediment accumulation rates and core recovery. The samples were disaggregated in distilled water, sieved through a 63 ~tm screen and oven-dried (60~ Isotopic analyses have been performed on sizecontrolled monogeneric and monospecific foraminiferal assemblages in order to minimize influence from 'vital effect' disequilibrium isotopic fractionation (e.g. Hemleben et al. 1985; Shackleton et al. 1985). Isotopic analyses of planktonic S u b b o t i n a spp. (size fraction 185250 p.m) and benthonic Oridorsalis u m b o n a t u s (size fraction 125-250 ~tm) were conducted at the Department of Earth Sciences in Oxford using a VG Prism mass spectrometer (see Schmitz et al. 1996). Spinose Subbotinids have been excluded from the isotopic analyses because of their presumed deeper water habitat compared to respective non-spinose specimens (Boersma et al. 1987). Oridorsalis umbonatus has not been commonly employed in establishing bottom water conditions during the Paleocene-Eocene intervals. Therefore, in order to compare isotopic records from Site 550 with coeval sections worldwide, additional isotopic analyses were conducted on the benthonic Nuttallides truempyi at the Department of Marine Geology in G6teborg using a VG Prism Series II mass spectrometer. In both laboratories, analytical data have been normalized to NBS-19. In

B. SCHMITZ

Grteborg, replicate analyses of NBS-19 samples gave a standard deviation (2c~) of _+0.01 for 8~3C and _ 0.06 for fi180. The results are reported on a notation, which is defined as the per rail deviation from the Peedee Belemnite Standard (PDB). For age estimates we have used nannofossil (NP) zonations (MOiler 1985; Aubry 1995; Berggren & Aubry 1996), since foraminiferal biostratigraphic control is rather limited over the interval studied. Numerical ages have been calculated by linearly interpolating between nannofossil zonal boundaries using the time-scale of Berggren et al. (in press) as it was summarized in Hardenbol (1994). The same time-scale has been used when recalculating numerical ages for Sites 401,577 and 690, according to age-model parameters maintained by Pak & Miller (1992). In particular for Site 401, due to uncertainties in correlating planktonic foraminiferal datum levels (Pak & Miller 1992) to the new chronology based on Berggren et al. (in press), we have also considered, where necessary, calcareous nannofossil zonations by Mtiller (1979). In order to estimate palaeotemperature differences that correspond to the proportional shifts in oxygen isotope records, we have used the palaeotemperature equation of O'Neil et al. (1969) as it was recalculated by Shackleton (1974).

Foraminiferal isotope fractionation Recent benthonic foraminiferal species, do not commonly fractionate carbon or oxygen in isotopic equilibrium with the dissolved HCO 3- in ambient seawater (Woodruff et al. 1980; Belanger et al. 1981; Vincent et al. 1981; Grossman 1987). For well preserved monospecific assemblages of benthonic foraminifera, however, these departures from equilibrium have been inferred constant through time (Shackleton et al. 1984). Nuttallides truempyi, which survived the Paleocene-Eocene boundary benthonic faunal crisis (see Aubry et al. 1996), exhibits only minor species offsets and is therefore frequently employed in isotopic studies dealing with the oceans of the early Cenozoic. Based on data from earlier studies and new analyses, Pak & Miller (1992) estimated that N. truempyi is c. 0.12 lower for 5180 and c. 0.23 lower for ~13C, compared to Cibicidoides, which is considered as the genus recording the closest values to isotopic equilibrium with the dissolved bicarbonate (Graham et al. 1981; Duplessy et al. 1984). In all comparisons between worldwide deep sea sections, we have applied Pak & Miller's (1992) correction factors for all N. truempyi data (Figs 4-7). Oridorsalis umbonatus has been considered by most authors to fractionate oxygen isotopes close to isotopic equilibrium with the

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 550 ambient water (Graham et al. 1981; Vincent et al. 1981; Shackleton et al. 1984). As regards 813C, O. umbonatus shows values about 0.5-1.5%o lower than equilibrium values (Grossman 1987) 9 Shackleton et al. (1984) considered O. umbonatus erratic with respect to ~13C. Oxygen and carbon isotope values provide important means for determining quantitatively the temperature and water-depth habitat of planktonic foraminiferal species (e.g. Corfield & Cartlidge 1991). Oxygen isotope records for surface or deeper dwelling planktonic as well as benthonic foraminiferal species offer information on the temperatures at various depths in the water column.

459

In this way, the vertical thermal stratification of the water column can be established. The vertical 813C gradient in the water column (e.g. Kroopnick 1980) can be used to independently reconstruct the depth habitat of different planktonic foraminifera. Other processes such as biological productivity variations in surface water and upwelling also affect the planktonic 813C values. Foraminifera thriving within the euphotic zone may exhibit carbon disequilibrium fractionation because of symbionts (Spero & Lea 1992). The relatively positive 8180 and negative 813C values of Subbotina spp. reflect a subsurface, deep habitat with lower temperatures, possibly within the region

(~180 -1,5 I

Time (Ma)

-1 .,I

-0,5

0

I

I

0,5

50-

P8 51-

550 Subbotinaspp. 550 N. truempyi P7

j

550 O. umbonatus

52~ m ~d

53P6

54-

I

I

NP10!

Fig. 1. Oxygen isotope records of the deep-dwelling planktonic Subbotina spp. and the benthonics Nuttalides truempyi and Oridorsalis umbonatus for Site 550. Numerical ages have been calculated by linearly interpolating between nannofossil zonal boundaries in Hole 550 (see main text for references). The nannofossil boundaries reflect their measured position in Hole 550, whereas the magnetostratigraphy, planktonic foraminiferal zonation and chronology are inferred through their relationship to the NP zonation in the time-scale of Berggren et al. (in press) as illustrated in Hardenbol (1994).

460

s.D. CHARISI t~r B. SCHMITZ ~13 C

5180 0,2

I

I

!

yiO,59x +0,21 r =0,613

I

[

' I

....

I

y = 1,01x - 0,46 r = 0,866

0

'1

d

l

0,5 "

~ -0,2 ~t .,,..

.'~176 ~S

0

E

t -0,4

"~ -0,5

'i~ -0,6 i * -0,8 -1,25 (a)

I -1

I -0,75

I -0,5

-1,5 -0,25

Nuttallides truempyi

-1 (b)

1

I

I

-0,5

0

0,5

Nuttallides truempyi

Fig. 2. (a) Correlation plot for paired oxygen isotope measurements of Nuttallides truempyi and Oridorsalis umbonatus for Site 550. (b) Correlation plot for paired carbon isotope measurements of Nuttallides truempyi and Oridorsalis umbonatus for Site 550.

of dissolved oxygen minimum (Shackleton et al. 1985; Boersma et al. 1987; Corfield & Cartlidge 1991; Corfield 1993). Because of its deep planktonic habitat, Subbotina spp. shows less variable 8180 and ~513C values than surface-dwellers, for which isotopic variations may to a great extent reflect seasonal temperature changes.

If absolute 8180 values are interpreted in terms of temperature, the similarities between the Subbotina 8180 values and those of N. truempyi indicate isothermal conditions through, at least, a major part of the water column.

Oxygen isotopes (Table 1)

Benthonic and planktonic foraminiferal carbon isotope records exhibit in general parallel trends through the early Eocene (Fig. 3). A slightly decreasing trend can be observed in the earliest Eocene (nannofossil Zone NP10) for the 813C values of Subbotina and N. truempyi. Subbotina 813C values decrease from 1.5 to c. 0.7%~ and thereafter oscillate around this value through Zones NP11 and NP12. N. truempyi 813C values decrease moderately from 0.2 to c. -0.2%o in Zone NP10 and remain at this level in Zones N P l l and partly NP 12. This 813C decrease, which is expressed more prominently in the Subbotina than the N. truempyi trends, probably constitutes the final stages of the Paleocene-Eocene transitional long-term negative 813C shift (see also Stott et al. 1996). From the mid-NP12 to early NP13 interval, 813C values for N. truempyi increase gradually by 1%~. N. truempyi reach the highest early Eocene 813C values of c. 0.9%~ in the earliest NP13 Zone and thereafter remain this high for the rest of the early Eocene. Subbotina spp. analyses from the late early Eocene interval have given anomalous values reflecting diagenetic alteration and therefore have not been included in this study.

The 8~80 records of planktonic Subbotina and the benthonics N. truempyi and O. umbonatus exhibit minor fluctuations and covary through most of the early Eocene (Fig. 1). The only significant departures from an invariable trend is a convergence between Subbotina and the benthonic values in the upper NP 10 Zone and an approximate 0.5%o increase in 5180 for the two benthonic species across the NP 13/NP 14 transition. N. truempyi 8180 values are generally oscillating close to -1%o and they are constantly lower by 0.6 to 0.8%o compared to those for O. umbonatus (Fig. 2a). These isotopic differences are rather different from estimates of Shackleton et al. (1984), based on measurements from samples throughout the Cenozoic, that N. truempyi are 0.35%~ lower for 5180 compared to O. umbonatus. Subbotina are lower in 8180 by 0.5%o compared to N. truempyi within the middle part of the NP10 Zone. The two records progressively converge towards the NP10/NPll transition. Thereafter, through the N P l l and NP12 Zones, the Subbotina 5180 values are similar to those for N. truempyi.

Carbon isotopes (Table 1)

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 5 5 0

461

Table 1. Carbon and oxygen isotope results (%0 v. PDB) for planktonic and benthonic foraminiferal samples from

Site 550 Time (Ma)

Core section, interval (cm)

Depth (mbsf)

Subbotina spp. ~13C

49.05 49.18 49.48 49.50 49.87 49.99 50.22 50.35 50.48 50.52 50.59 50.89 50.93 50.97 51.01 51.77 51.85 51.94 52.02 52.10 52.19 52.27 52.36 52.44 52.45 52.48 52.54 52.61 52.67 52.74 52.79 52.87 52.93 52.96 53.09 53.16 53.29 53.35 53.42 53.48 53.54 53.58 53.68 53.68 53.71 53.74 53.77 53.79 53.83 53.86 53.89 53.92 53.97 54.00 54.03 54.06 54.09 54.12 54.15 54.19 54.25 54.28 54.34 54.39 54.43

24-1, 31-35 24-3, 140-144 24-5, 9-13 24-5, 106-110 25-1, 48-52 25-1, 142-146 25-3, 10-14 25-3, 108-112 25-4, 5-9 25-4, 90-94 25-4, 145-149 26-1, 6--10 26-1, 34-38 26-1, 83-87 26-1, 140-144 27-1, 90-94 27-2, 39-43 27-2, 137-141 27-3, 89-93 27-4, 35-39 27-4, 140-144 27-5, 89-93 27-6, 43-47 27-6, 140-144 27-7, 30-34 28-1, 36--40 28-1, 138-142 28-2, 90-94 28-3, 40--44 28-3, 142-146 28-4, 87-91 28-5, 38-42 28-5, 138-142 28-6, 27-31 29-1, 37-41 29-1, 144-148 29-2, 89-93 29-3, 41-45 29-3, 139-143 29-4, 85-89 29-5, 39-43 29-5 132-136 29-6 40--44 30-1 37-41 30-1 142-146 30-2 87-91 30-3 35-39 30-3 140-144 30-4 86-90 30-5 17-21 30-5 116-120 30-6 90-94 31-1 36--40 31-1 138-142 31-2 91-95 31-3 41-45 31-3 138-142 31-4 90-94 31-5 41-45 31-5 143-147 32-1 5-9 32-1 86-90 32-2 143-147 32-4 34-38 32-4 141-145

311.90 312.92 315.42 315.58 318.50 319.44 321.12 322.10 323.07 323.42 323.97 327.50 327.86 328.35 328.92 337.92 338.91 339.89 340.91 341.87 342.92 343.91 344.95 345.92 346.32 346.88 347.90 348.92 349.92 350.94 351.89 352.90 353.90 354.29 356.39 357.46 358.41 359.43 360.41 361.37 362.41 363.34 363.92 365.89 366.94 367.89 368.87 369.82 370.88 371.69 372.68 373.92 375.38 376.40 377.43 378.43 379.40 380.42 381.43 382.45 384.57 385.38 387.45 389.36 390.43

N. truempyi

O. umbonatus

(185-250 I.tm) 8180

0.795 0.589 0.451 0.808 0.705 0.603 0.413

-0.787 -0.781 -1.044 -0.811 -0.478 -0.738 -0.817

0.522 0.474 0.505 0.527

-1.077 -0.971 -0.869 -1.209

0.653 0.607 0.533 0.844 0.495

-0.770 -0.862 -0.923 -0.788 -0.752

0.836 0.597 0.000 0.598

-0.620 -0.943 -0.363 -0.854

0.232 0.797 0.812 0.723 0.785 0.854 0.859 0.714 0.842 0.634 0.718 0.479 0.832 0.603 0.888 0.900 1.097 1.256 0.935 0.949 0.900

-1.102 -0.798 -0.986 -1.045 -1.267 -1.198 -1.131 -1.383 -1.170 -1.485 -1.474 -1.724 -1.563 -1.400 -1.492 -1.365 -1.428 -1.460 -1.432 -1.476 -1.328

1.005 1.293 1.532

-1.664 -1.400 -1.729

513C

5180

0.619 0.755

-0.399 -0.300

0.837 0.851 0.291 0.850 0.791 0.515

0.352 -1.037 -0.848 -1.110 -0.798 --0.832 -1.004

0.302

-1.057

0.508 0.074 0.10l 0.067 -0.495 -0.149 -0.012 -0.061 -0.171 0.095

~13C

5180

-0.223 0.172 0.722 -0.118 0.437

1.047 0.032 -0.057

0.760 0.610 0.338 0.624 0.018 -0.023 0.199

-0.415 -0.243 --0.515 -0.225 -0.599 -0.579 -0.564

-0.978 -0.803 -0.886 -0.866 -1.228 -0.981 -0.971 -0.718 -0.875 -0.943

-0.106 -0.438 -0.524 -0.730 -0.539 -0.569 -0.566 -0.925 -0.565

-0.155 --0.219 -0.219 -0.630 -0.369 -0.295 -0.338 -0.534 -0.475

-0.318 -0.057

-1.042 -0.801

-0.789 -0.674

-0.475 -0.272

-0.217 0.145 0.043 -0.173 -0.272 -0.201

-0.521 -0.368 -0.318 -0.516 -0.563 -0.428 -0.523 -0.227 -0.583 -0.695 -0.603 -0.645 -1.159 -0.618 -0.842 -0.627 -0.579

-0.514 -0.090 -0.298 -0.385 -0.355 -0.248

0.108 0.006 -0.287 -0.146 -0.519 -0.614 -0.058 0.059 -0.072 -0.020 0.001 0.026 -0.330 -0.120 -0.255 -0.175 -0.329 -0.141 -0.042 -0.010 0.129 0.262 0.212 0.161 -0.183 0.056 -0.126 0.132

-0.999 -0.786 -0.845 -1.028 -1.043 -0.887 -0.891 -0.729 -0.782 -0.989 -0.929 -1.171 -1.171 -0.785 -0.772 -0.900 -0.871 -0.816 -0.716 -1.045 -0.801 -1.009 -0.964 -1.149 -0.880 -0.811 -0.812 -0.719 -0.645 -0.604 -0.626 -0.698 -0.876 -1.052 -0.922

-0.139 -0.358 -0.445 -0.277 -0.479 -0.625 -0.338 0.692 -0.197 -0.316

-0.857 -0.532 -0.861 -0.512

-0.312 -0.068 -0.302 -0.006

-0.954

-0.525

0.371

-0.759

-0.311

462

S.D. CHARISI • B. SCHMITZ indicates that Oridorsalis probably exhibits variable departures from isotopic equilibrium (cf. Shackleton et aL 1984). At the NP12/NP13 transition there is a convergence between the Oridorsalis and the Nuttallides values, that may reflect diagenetic artefacts. Diagenetic isotopic reequilibration or calcite infillings will in general erase original interspecific isotopic differences (Barrera et al. 1987).

Oridorsalis ~513C values follow a pattern similar to that exhibited by N. truempyi. Carbon isotope ratios of O. umbonatus are lower by 0.2-0.6%0 compared to those of N. truempyi and they are constantly oscillating around -0.6%~ through Zones NP10, NP11 and part of NP12. Within Zone NP12, O. umbonatus 513C rises by c. 1%o to reach values close to 0.5%e and oscillate around this level for the remaining early Eocene. On the basis of 15 reliable carbon isotope analyses from samples throughout the Cenozoic, Shackleton et al. (1984) estimated that, in general, 'vital-effect' isotopic fractionation leads to ~13C values lower by 1%o for Oridorsalis compared to Nuttallides. Our 39 paired analyses showed that on average, Oridorsalis ~p3C values are lower by 0.46%~ throughout the early Eocene (Fig. 2b), which

Early Eocene water mass stratification

The early Eocene was a period of equable oceanic conditions with the smallest vertical and latitudinal temperature gradients of the Cenozoic. This is

813C __

Time (Ma)

-1

-0,5

0

0,5

1

1,5

I

I

I

I

I

1

NP14 P9

C22

50-

NP13 P8

9

~

.

51gra ~ ) C 23

NP12

P7

52-

53-

NP11 P6 C 24

54-

NP10

Fig. 3. Carbon isotope records of the deep-dwelling planktonic Subbotina spp. and the benthonics Nuttallides truempyi and Oridorsalis umbonatus for Site 550. Stratigraphy as in Fig. 1.

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 550 apparent also at Site 550, where eliminated Subbotina-Nuttallides 5180 differences characterize most of the early Eocene, except for an interval in the NPI0 Zone, where the 8180 differences range between 0.5 and 1.0%o. Similarly eliminated thermal stratification of the water column for the early Eocene has been documented from other deep sea sites in the Atlantic as well (Boersma et al. 1987). The Subbotina-Nuttallides fi13C difference at Site 550 is approximately 0.5%0 through most of the early Eocene. This reflects a rather well developed vertical gradient, with respect to the deep-dwelling habitat of subbotinids (Corfield & Cartlidge 1991), which precludes reconstruction of the entire surface-to-deep vertical gradient. A well-developed 813C gradient is consistent with high productivity, which has also been registered in the southern Atlantic during the early Eocene (Boersma et al. 1987). Kennett & Stott (1990) compared the Paleogene benthonic 8180 records from ODP Site 689 (Eocene depth 1400 m) with those of the deeper Site 690 (Eocene depth 2250 m) drilled on the Maud Rise off Antarctica. They noticed that from the middle Eocene and continuing through much of the rest of the Paleogene the deeper site showed higher (warmer) benthonic oxygen isotopic values than the more shallow site. This was inferred to reflect a two-layered ocean consisting of warm saline deep waters formed at low latitudes, overlain by high latitude cool waters. This is in accordance with the idea that under the warmer and more equable climate of the early Paleogene, deep water production may have been primarily determined by salinity rather than temperature differences (Brass et al. 1982; Kennett & Stott 1990, 1991). Warm saline deep water (WSDW) formation under evaporative conditions, in shallow low latitude semi-restricted basins, has been suggested to be an important component of deep water circulation during the early Eocene (Kennett & Stott 1990, 1991; Pak & Miller 1992). By analogy with Kennett & Stott's (1990) rationalization, however, no apparent reversal in the benthonic 8180 records exists between deep and intermediate waters as it is reflected in our 5180 comparisons between Sites 550 and 401 (Figs 4-5). Benthonic isotopic data from Site 550 (estimated palaeodepth 38004000 m) represent the physicochemical properties of deep water masses, whereas corresponding data from Site 401 (palaeodepth 1800 m) reflect intermediate water-mass properties in the North Atlantic. The covariance between benthonic 5180 records, as well as the similarity in absolute 8180 values, for Sites 401 and 550, except perhaps from a brief interval at approximately 54 Ma, indicate that the two sites in the North Atlantic have been filled by a single source of deep and intermediate

463

waters through the early Eocene. This is also supported by the benthonic carbon isotope records for the two Sites 401 and 550. The 513C values are only slightly higher for Cibicidoides at Site 401 than N. truempyi at Site 550. On the basis of the similarity in 513C signatures of water masses bathing the two North Atlantic sites, we regard those water masses to have been of similar age and originating from the same source region, as it emerges from the assumption that deep waters, as they migrate from their source region, become progressively more enriched in nutrients and more depleted in 13C contents. Vertically isothermal conditions developed at Site 550 during the upper NP10 interval, as it is manifested by the Subbotina-N. truempyi 5180 convergence (Fig. 4). Diagenesis cannot account for this 8180 pattern since no corresponding convergence can be noted between the Oridorsalis and the Nuttallides records. The Subbotina 5180 increase occurs from approximately 54 to 53.5 Ma and suggests an abrupt cooling of subsurface water of about 2~ Thereafter, the vertical 5180 gradient was eliminated until at least 52 Ma in the mid early Eocene. Diminished 8180 gradients in the early Eocene have been observed also in southern high latitude Sites 738 in the Indian Ocean (Barrera & Huber 1991) and 690 in the Southern Ocean (Kennett & Stott 1990). For both these sites the reduced surface-to-deep 5180 gradient resulted from a decrease in benthonic 5180 values, coincident with an increase in the planktonic values (Kennett & Stott 1990; Barrera & Huber 1991). No similar decrease in benthonic 8180 values, however, occurs at Site 550. Furthermore, at Site 690, the Subbotina 5180 increase is more gradual than that at Site 550, with a duration of approximately one million years. The above suggest that decreased thermal stratification appears to be commonly the case for deepsea sections worldwide in early Eocene, whereas the rapid elimination of the Subbotina-N. truempyi 5180 gradient in the NP10 Zone at Site 550 has a regional rather than global character. The 5180 convergence probably reflects gradual upwards substitution of subsurface water by intermediate and deep water (see further next section). The 5180 increase for Subbotina does not correlate with any similar changes in the 813C record. The vertical 813C gradient reflects the removal of 12C in surface water due to photosynthesis and its input at depth by oxidation of the organic matter falling from the surface (e.g. Kroopnick 1980). No apparent change in the vertical surface-to-deep 813C gradient exists at Site 550 during the early early Eocene that could be linked to the changing thermal stratification of the water column at this time.

464

S.D. CHARISI (~ B. SCHMITZ 5180

Time

49

_

.

-1,5

-1

i

I

,

-0,5

0

n

i

.

(Ma)

NP14 P9 50-

22 NP13

P8

ubbotin ~pp. ~'0 ~ 550 N. true. pyi P\ (X~ ~_ "-"<)"'-- 401 Cibicido

51-

N 23

NP12 P 7

52-

I

53-

~ 24

NPll P6

II

54-

NP1C

Fig. 4. Comparison of oxygen isotope records of Subbotina spp. (indicative of subsurface water conditions) and Nuttallides truempyi (indicative of deep water conditions) from Site 550 with Cibicidoides spp. (indicative of intermediate water conditions) from Site 401 in the Bay of Biscay. N. truempyi isotope records have been adjusted for disequilibrium isotopic fractionations using correction factors from Pak & Miller (1992; see text). Stratigraphy as in Fig. 1.

E a r l y E o c e n e volcanism

and

deep

sea

c i r c u l a t i o n in t h e N o r t h A t l a n t i c The opening of the Norwegian--Greenland Sea and the North Atlantic connection to the Arctic have been extensively studied owing to the potentially important palaeoenvironmental consequences of these events. In the modem ocean, the connection between the polar basins is maintained via the Atlantic Ocean. An important component of deep sea circulation is the North Atlantic Deep Water (NADW), which is mainly formed in the Norwegian-Greenland Sea. These oceanographic conditions have evolved in response to plate tectonic rearrangement and changing basin con-

figuration in the region, which, in turn, had important consequences in the evolution of Cenozoic climate. In the earliest Eocene, North Atlantic volcanism is represented in and around the North Sea by tholeiitic ash layers indicative of unusually intense, large scale, volcanic eruptions that have been linked to the initial separation of Greenland from Eurasia in mid-NP10 Zone (Knox & Morton 1988; Morton & Knox 1990). A series of 50 bentonite layers at Site 550 in cores 32-34 (Knox 1985) have been considered as distal representatives of the volcanic activity in the Faeroe-Greenland region (Knox 1984; Knox & Morton 1988). The intense North Atlantic volcanism (Roberts et al. 1984) was

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 550 associated with globally enhanced hydrothermal activity that may have led, among others, to polar warming and consequently low latitudinal temperature gradients in the early Eocene (Owen & Rea 1984; Rea et al. 1990; Zachos et al. 1993). Knowledge about the development of marine connections between the North Atlantic and the Arctic is still rather fragmentary. Deep water circulation modes analogous to those of the present day probably did not prevail in the North Atlantic before the mid or late Miocene (e.g. Miller & Tucholke 1983; Miller et al. 1985; Thiede et al. 1989). However, a shallow marine, if not a deep water, connection existed between the Arctic and the Atlantic possibly since the early Eocene (e.g. Berggren & Olsson 1986). Many authors have

465

investigated the different possible alternative pathways linking the Norwegian-Greenland Sea with the North Atlantic during the early Eocene, such as the Iceland-Faeroe and/or the Faeroe-Shetland gateways along the Greenland-Scotland Ridge (Berggren & Schnitker 1983; Miller & Tucholke 1983; Thiede & Eldholm 1983; Berggren & Olsson 1986). More enigmatic are the northward connections of the Norwegian-Greenland Sea with the Arctic Ocean, such as the development of the Fram Strait and the separation of Greenland from Svalbard (Miller & Tucholke 1983; Eldholm et al. 1987; Myhre & Eldholm 1988). As discussed in Schmitz et al. (1996) there are indications in favour of extensive water exchange already in the early Eocene between the different, newly formed basins

613C Time

-0,5

0

0,5

1

1,5

I

I

I

I

I

(Ma)

mP9 50-

C22

~

es

51-.

550 Subbotina spp. 550 N. truempyi

,

P7

~,1 C23

~

401 Cibicidoides

52-

< 53P6

,C24

54

NP10

Fig. 5. Comparison of carbon isotope records of Subbotina spp. (indicative of subsurface water conditions) and Nuttalides truempyi (indicative of deep water conditions) from Site 550 with Cibicidoides spp. (indicative of intermediate water conditions) from Site 401 in the Bay of Biscay. N. truempyi records are corrected as in Fig. 4. Stratigraphy as in Fig. 1.

466

s.D. CHARISI & B. SCHMITZ

in the northeasternmost Atlantic and the North Atlantic south of the Greenland-Scotland Ridge. A connection with the Arctic Ocean may also have existed through the Labrador Sea in the earlymid Eocene (Gradstein & Srivastava 1980), which is supported by faunal data indicative of a warm climate at Ellesmere Island (McKenna 1980). Warm high latitude climatic conditions in the Norwegian-Greenland Sea during the early Eocene are also indicated by floral data at Spitsbergen (Schweitzer 1980). Considering the dramatic palaeogeographical changes taking place in the northeastern Atlantic region, there is a possibility that the pronounced ~180 increase in the Subbotina record at Site 550 in the NPI0 interval, is related to these changes. For example, cool water from high latitudes overflowing a submerging Greenland-Scotland Ridge may have infiltrated subsurface waters in the Goban Spur region. However, there are strong indications that the water masses in the sub-basins north of the Greenland-Scotland Ridge had reduced salinities related to the surrounding vast drainage areas (Schmitz et al. 1996). Admixture of such water in the upper water mass at Site 550 would have resulted in lower, rather than higher, ~180 values, therefore we consider this possibility unlikely. Also, it is noticeable that after the Subbotina-N.truempyi 8180 convergence, for a long period the benthonic and the planktonic values remain similar, reflecting that the water column had about the same properties throughout. It is unlikely that incoming northeasternmost Atlantic water had the same composition as the deep waters that dominated before the 8180 convergence in late Biochron NP10. We therefore instead consider that the 5180 convergence reflects gradual upward displacement of subsurface waters by intermediate and deep waters, perhaps emanating from an enhanced influence of the prime deep water source in the region. The development of Proto-Gulf Stream circulation in the region, which has been suggested to have commenced sometime between the late Paleocene and late Eocene (Pinet & Popenoe 1985; Tucholke & Mountain 1986; Barron & Peterson 1991), yet cannot account for the gradual 8180 increase at Subbotina depths towards the end of Biochron NP10. A Proto-Gulf Stream would have transported warmer water masses from low latitude western basins leading to lighter 5180 values at Site 550. A possibility, however, is that surfacewater circulation patterns analogous to modern Gulf Stream circulation were temporarily disrupted at the end of Biochron NP10 Zone, contemporaneously with the development of oxidized sediments above Reflector A b in the western North Atlantic, and with a weaker influence of deep

waters originating from the Southern Ocean (Mountain & Miller 1992). The benthic 8180 increase occurring through the NP13-NP 14 transition indicates important changes in palaeocirculation patterns in the North Atlantic. This late early Eocene 8180 increase of more than 5%0 has been observed at both Sites 401 and 550, indicating a c. 2~ cooling of both intermediate and deep waters. On the basis of faunal data, Kaminski et al. (1990) suggested that initiation of deep water connections between the North Atlantic and the Norwegian-Greenland Sea occured in the uppermost early Eocene Zone NP14. The benthonic 8180 increase coincides well with increased biosiliceous sedimentation mostly in the North Atlantic as well as in the equatorial Pacific in the middle Eocene (Berggren & Hollister 1974; McGowran 1989). This enhanced siliceous sedimentation has been considered to result from, among others, the opening of oceanic gateways (Baldauf & Barron 1990) and, vigorous bottom water circulation in the North Atlantic, which would in turn cause increased upwelling and possibly enhanced productivity (Berggren & Hollister 1974). The bottom-water cooling at the North Atlantic sites is an event of regional as well as global importance. It probably represents the initiation of the step-wise global climatic cooling continuing through the rest of the Eocene, the Oligocene and the early Miocene. This cooling was associated with the intensification of global deep water circulation and the opening of the Norwegian-Greenland Sea, ultimately leading to a North Atlantic circulation similar to that in the modern ocean.

Early Eocene global deep water circulation A change in the mode of deep water circulation from cold, highly oxygenated waters produced in the southern polar regions to warm, poorly ventilated waters produced in low latitude evaporative basins, has been proposed to have resulted in the Paleocene/Eocene boundary benthonic foraminiferal crisis (Miller et al. 1987; Kennett & Stott 1990; 1991; Katz & Miller 1991; Pak & Miller 1992). In the early Eocene the culmination of the long-term climatic warming that began in the Paleocene led to increased high latitude temperatures and decreased latitudinal thermal gradients (Savin 1977; Shackleton & Boersma 1981). These warm climatic conditions may have been important in reinforcing a low latitude halothermal mode of circulation during the early Eocene. Comparison of benthonic 513C records from different ocean basins have been used to trace deep waters back to their source regions. Deep waters close to their source regions are relatively enriched

EARLY EOCENE STABLEISOTOPES,DSDP HOLE 550

467

8180

Time (Ma)

A~

--

-1,5

-1

-0,5

0

I

I

I

I

0,5

50-

II +

550 N.

truempyi

401 Cibicidoides "4, , ~

51-

I

577 N. truempyi 690 Cibicidoides

Z O 52-

53I

54-

NPlC

Fig. 6. Comparison of benthonic oxygen isotope records from Site 550 (this study), Site 401 in the North Atlantic (Pak & Miller 1992), Site 577 in the equatorial Pacific (Pak & Miller 1992) and Site 690 on Maud Rise in the Southern Ocean (Kennett & Stott 1990). N. truempyi records from Sites 550 and 577 are corrected as in Fig. 4. Stratigraphy as in Fig. 1.

in 13C and oxygen, but poor in nutrients. As deep waters migrate they become progressively more depleted in oxygen and enriched in 12C due to extensive oxidation of the excess organic matter falling from the surface (e.g. Kroopnick 1980; Curry & Lohmann 1982, Broecker & Peng 1982). Here we compare the first actual deep water high resolution carbon and oxygen isotope records from Site 550 (palaeodepth c. 4000 m) with coeval records from intermediate to upper deep waters. The data are compared with benthonic isotopic records from Site 690 on Maud Rise in the Southern Ocean (2250 m palaeodepth), Site 577 in the equatorial Pacific (2000 m palaeodepth), and Site 401 in the Bay of Biscay, North Atlantic (1800 m palaeodepth) (Figs 6-7).

Carbon isotope records diverge from approximately 54.5 to 51 Ma in the early Eocene. The Antarctic 813C values are higher by 0.5%0 compared to the rest of the dataset, in accordance with the suggestion by Pak & Miller (1992) that the Southern Ocean has been bathed by 'younger' bottom waters, enriched in oxygen and poor in nutrients. These authors have proposed for the 54 to 51 Ma interval a two-component ocean, where deep waters with relatively high fi180 values formed in the Southern Ocean, whereas the lower latitudes basins represented the loci for formation of deep water with lower 8180 values. The question, however, is whether multiple sources of deep water can be plausible and/or detectable during the early Eocene when, owing to low

468

s . D . CHARISI & B. SCHMITZ

613C -0,5

0

0,5

1

I

I

I

1

Time (Ma)

m

1,5

P9

50-

,

51-

PJ C 23

~

"--"<>'--

550 N. truempyi

----<>--

401 Cibicidoides

r

577 N. truempyi

------4::)--

690 Cibicidoides

52_

c~ 53 C24

P6

'

54 84 ,

""~-....~

~

Fig. 7. Comparison of benthonic carbon isotope records from Site 550 (this study), Site 401 in the North Atlantic (Pak & Miller 1992), Site 577 in the equatorial Pacific (Pak & Miller 1992) and Site 690 on Maud Rise in the Southern Ocean (Kennett & Stott 1990). N. t r u e m p y i records from Sites 550 and 577 are corrected as in Fig. 4.

latitudinal thermal gradients, different potential source regions may have been characterized by similar water properties. Nevertheless, in our comparison we consider both Sites 401 (indicative of intermediate waters) and 550 (indicative of deep waters) to have been dominated by a single source with lower 5180 values and, as it emerges from the low ~13C values, at the North Atlantic sites the waters were depleted in oxygen and enriched in nutrients. The benthonic 513C records from all four sites studied, converge at approximately 51 Ma. Thereafter, ~13C records are ranging between 0.5-1.0%0. The oxygen isotope records converge at approximately 50 Ma and they remain at the same levels for the rest of the early Eocene. The convergences

in both carbon and oxygen isotope records suggest a development towards a single bottom water source in the late early Eocene. Both the 813C and 8180 records from the studied sites tend to become more positive in this interval, which may indicate an increasing influence of cool deep waters of relatively low nutrient and high oxygen content, originating at high latitudes. Summary

The palaeoceanography of the warm early Eocene time-interval has been characterized as a twocomponent ocean, with dual deep and intermediate water sources in high and in low latitudes, sometimes leading to an inverted temperature gradient,

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 5 5 0 because of saline low latitude water sinking below the cool, less saline water from high latitudes. In the eastern North Atlantic, however, the homogeneity in ~5180 signatures between subsurface, intermediate and deep waters indicates a single deep-intermediate water source for most of the early Eocene. In the earliest Eocene, at the end of the NP10 Biochron, a 0.5-1.0%o difference between Subbotina spp. and benthonic 8180 values from Site 550 was eliminated, probably caused by an enhanced influence of the prime deep water source, which resulted in upward displacement of subsurface water by colder and/or more saline intermediate and deep water. Low benthonic 5180 and ~13C values characterize most of the early Eocene (c. 54.5-51 Ma) at Sites 550 and 401, indicating a dominance of

469

warm and 12C-enriched water in the North Atlantic. This water probably sank from the sea surface at low latitudes. An abrupt 8180 increase at the end of the studied interval indicates the initial phases of a gradual change towards cooler and 'younger' deep waters originating at high latitude areas. We thank R. M. Corfield and J. E. Cartlidge for isotopic analyses performed in Oxford. We also thank O. Gustafsson for isotopic analyses in Gtiteborg, and T. Alavi and T. Andinsson for laboratory assistance. The Hellenic State Scholarships Foundation (S.S.E) provided grants to SDC. Additional funding was obtained from the Bank of Sweden Tercentenary Foundation and the Swedish Natural Science Research Council. Samples were kindly provided by the Deep Sea Drilling Project. We thank R. M. Corfield for helpful comments on an early version of the manuscript.

References AUBRY, M.-R 1995. From chronology to stratigraphy: assessing the continuity of the stratigraphic record. In: BERGGREN,W. A., KENT, D. V., AUBRY,M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Volume, 54.

1996. The upper Paleocene-lower Eocene stratigraphic record and the Paleocene-Eocene boundary carbon isotope excursion: implications for geochronology. This volume. BALDAUF, J. G. & BARRON, J. A. 1990. Evolution of biosiliceous sedimentation patterns-Eocene through Quaternary: paleoceanographic response to polar cooling. In: BLEIL,U. 8z THIEDE,J. (eds) Geological History of the Polar Oceans: Arctic versus Antarctic. Kluwer Academic Publishers, The Netherlands, 575-607. BARRERA, E. & HUBER, B.T. 1991. Paleogene and early Neogene oceanography of the southern Indian Ocean: Leg 119 foraminifer stable isotope results. In: BARRON,J., LARSEN,B., ET AL. Proceedings of the Ocean Drilling Program, Scientific Results, 119, 693-717. --, --, SAVIN, S. M. & WEBB, E-N. 1987. Antarctic marine temperatures: from Campanian through early Paleocene. Paleoceanography, 2, 21-48. BARRON, E. & PETERSON, W. H. 1991. The Cenozoic ocean circulation based on ocean General Circulation Model results. Palaeogeography, Palaeoclimatology, Palaeoecology, 83, 1-28. BELANGER,P. E., CURRY,W. B. & MATTHEWS,R. K. 1981. Core-top evaluation of benthic foraminiferal isotopic ratios for paleo-oceanographic interpretations. Palaeogeography, Palaeoclimatology, Palaeoecology, 33, 205-220. BERGGREN, W. A. & AUBRY, M.-P. 1996. A late Paleocene-early Eocene NW European and North , STOTT, L. t~z SINHA, A.

Sea magnetobiostratigraphic correlation network: a sequence stratigraphic approach. This volume. & HOLLISTER, C. D. 1974. Paleogeography, paleobiogeography, and the history of circulation in the Atlantic ocean. In: HAY,W. W. (ed.) Studies in Paleoceanography. Society of Economic Petrologists and Mineralogists, Special Publication, 20, 126-186. - & OLSSON,R. K. 1986. North Atlantic Mesozoic and Cenozoic paleobiogeography. In: VOGT, P. R. & TUCHOLKE, B. E. (eds) The Geology of North America, Volume M, The Western North Atlantic Region. Geological Society of America, 565-587. -• SCHNITKER,D. 1983. Cenozoic marine environments in the North Atlantic and NorwegianGreenland Sea. In: BOTT, M. H. P., SAXOV, J., TALWANI, M. & THIEDE, J. (eds) Structure and Development of the Greenland-Scotland Ridge. NATO Conference Series IV: Marine Sciences, Plenum Press, New York, 8, 495-548. - - . , KENT, D. V., SWISHER, C. C. III & AUBRY, M.-P. 1995. A revised Cenozoic geochronology and chronostratigraphy. In: BERGGREN, W. A., KENT, D. V., AUBRY, M.-E & HARDENBOL, J. (eds) Geochronology, Time Scales and Stratigraphic Correlation: Framework for an Historical Geology. Society of Economic Paleontologists and Mineralogists, Special Volume, 54. BOERSMA, A., PREMOLI-SILVA, I. & SHACKLETON,N. J. 1987. Atlantic Eocene planktonic foraminiferal paleohydrographic indicators and stable isotope paleoceanography. Paleoceanography, 3, 287-331. BRASS, G. W., SOUTHAM,J. R. & PETERSON,W. H. 1982. Warm saline bottom waters in the ancient ocean. Nature, 296, 620-623. BROECKER,W. S. & PENG,T.-H. 1982. Tracers in the Sea. Eldigio Press, New York. CORFIELD,R. M. 1993. Depth habitats and the Paleocene radiation of the planktonic foraminifera monitored

470

S . D . CHARISI & B. SCHMITZ

Paleoceanography and Taxonomy of Agglutinated Foraminifera. NATO Advanced Study Institutes

using oxygen and carbon isotopes. In: Evolutionary Patterns and Processes. Linnean Society, London, 59-70. 1994. Palaeocene ocean and climate: An isotopic perspective. Earth Science Reviews, 37, 225-252. - t~ CARTLIDGE, J. E. 1991. Isotopic evidence for the depth stratification of fossil and recent Globigerina: A review. Historical Biology, 5, 37-63. & NORRIS, R. D. 1996. Deep water circulation in the Paleocene Ocean. This volume. CURRY, W. B. & LOHMANN, G. P. 1982. Carbon isotopic changes in benthic foraminifera from the western South Atlantic: Reconstruction of glacial abyssal circulation patterns. Quateranary Research, 18, 218-235. DE GRACIANSKY,P. C., POAG, C. W., ET AL. 1985. Initial Reports of the Deep Sea Drilling Project, 80. DUPLESSY, J.-C., SHACKLETON,N. J., MATTHEWS, R. K., PRELL, W., RUDDIMAN,W. E, CARALP,M. & HENDY, C. H. 1984. 13C Record of benthic foraminifera in the last interglacial ocean: Implications for the carbon cycle and the global deep water circulation. Quaternary Research, 21, 225-243. ELDHOLM, O. & THOMAS, E. 1993. Environmental impact of volcanic margin formation. Earth and Planetary Science Letters, 117, 319-329. , FALEIDE, J. I. & MYHRE, A. M. 1987. Continentocean transition at the western Barents Sea/Svalbard continental margin. Geology, 15, 1118-1122. GRADSTEIN, E M. • SR1VASTAVA,S. P. 1980. Aspects of Cenozoic stratigraphy and paleoceanography of the Labrador Sea and Baffin Bay. Palaeogeography, Palaeoclimatology, Palaeoecology, 30, 261-295. GRAHAM, D. W., CORLISS, B. n., BENDER, M. L. & KEtGWlN, L. D. 1981. Carbon and oxygen isotopic disequilibria of recent deep-sea benthic foraminifera. Marine Micropaleontology, 6, 483497. GROSSMAN, E. L. 1987. Stable isotopes in modern benthic foraminifera: A study of vital effect. Journal of Foraminiferal Research, 17, 48-61. HARDENBOL, J. 1994. Sequence stratigraphic calibration of Paleocene and Lower Eocene continental margin deposits in NW Europe and the US Gulf Coast with the oceanic chronostratigraphic record. GFF, 116, 49-51. HEMLEBEN, C., SPINDLER,M., BREITINGER,I., t~; DEUSER, W. G. 1985. Field and laboratory studies on the ontogeny and ecology of some globorotaliid species from the Sargasso Sea off Bermuda. Journal of Foraminiferal Research, 15, 254-272. HOVAN, S. A. & REA, D. K. 1992. Paleocene/Eocene boundary changes in atmospheric and oceanic circulation: A Southern Hemisphere record. Geology, 20, 15-18. JANECEK, T. R. & REA, D. K. 1983. Eolian deposition in the northeast Pacific Ocean: Cenozoic history of atmospheric circulation. Geological Society of American Bulletin, 94, 730-738. KAMINSKI, M. A., GRADSTEIN, F., GOLL, R. M. & GREIG, D. 1990. Biostratigraphy and paleoecology of deep-water agglutinated foraminifera at ODP Site 643, Norwegian-Greenland Sea. In: HEMLEBEN, C., KAMINSKI, M. A., KUHNT, W. & SCOTT, D. B. (eds) Paleoecology, Biostratigraphy, -

Series C, 327, 345-386. KATZ, M. E. & MILLER, K. G. 1991. Early Paleogene benthic foraminiferal assemblages and stable isotopes in the Southern Ocean, ODP Leg 114. In: CIESIELSKI, P. F., KRISTOFFERSON, Y., ET AL.

Proceedings of the Ocean Drilling Project, Scientific Results, 114, 481-513.

-

KENNETT, J. P. & BARKER,P. E 1990. Latest Cretaceous to Cenozoic climate and oceanographic developments in the Wedell Sea, Antarctica: an ocean-drilling perspective. In: BARKER, P. E, KENYErr, J. P.,

eT AL. Proceedings of the Ocean Drilling Project, Scientific Results, 113, 937-960. & STOTT, L. D. 1990. Proteus and ProtoOceanus: Ancestral paleogene oceans as revealed from Antarctic stable isotope results. In: BARKER, P. E, KENNETT, J. P., ET AL. Proceedings of the Ocean Drilling Project, Scientific Results, 113, 865-880. & -1991. Abrupt deep sea warming, paleoceanographic changes and benthic extinctions at the end of the Paleocene. Nature, 353, 225-229. KNOX, R. W. O'B. 1984. Nannoplankton zonation and the Paleocene/Eocene boundary beds of NW Europe: An indirect correlation by means of volcanic ash layers. Journal of the Geological Society, London, 141, 993-999. 1985. Stratigraphic significance of volcanic ash in Paleocene and Eocene sediments at Sites 549 and 550. In: DE GRACIANSKY, P. C., POAG, C. W., ET AL.

-

-

-

-

Initial Reports of the Deep Sea Drilling Project, 845-850. & MORTON,A. C. 1988. The record of early Tertiary N. Atlantic volcanism in sediments of the North Sea Basin. In: MORTON, A. C. & PARSON, L. M. (eds) Early Tertiary Volcanism and the Opening of the NE Atlantic. Geological Society, London, Special Publication, 39, 407-419. KROOPNICK, P. M. 1980. The distribution of 13C in the Atlantic Ocean. Earth and Planetary Science Letters, 49, 469-484. McGOWRAN, B. 1989. Silica burp in the Eocene ocean. Geology, 17, 857-860. MCKENNA, M. C. 1980. Eocene paleolatitudes, climates and mammals of Ellesmere Island. Palaeogeography, Palaeoclimatology, Palaeoecology, 30, 349-362. MASSON, O. G., MONTADERT,L. & SCRUTTON,R. A. 1985. Regional geology of the Goban Spur continental margin. In: DE GRACIANSKY, P. C., POAG, C. W., 80,

--

ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 1115-1139. MILLER, K. G. & TUCHOLKE, B. E. 1983. Development of Cenozoic abyssal circulation south of the Greenland-Scotland Ridge. In: BOTT, M. H. P., SAXOV, J., TALWANI,M. & THIEDE, J. Structure and

Development of the Greenland-Scotland Ridge. --,

NATO Conference Series IV: Marine Sciences, Plenum Press, New York, 8, 549-589. FAIRBANKS, R. G. & THOMAS, E. 1985. Benthic foraminiferal carbon isotopic records and the development of abyssal circulation in the eastern North Atlantic. In: RUDDIMAN,W. F., KIDD, R. B.,

EARLY EOCENE STABLE ISOTOPES, DSDP HOLE 5 5 0 THOMAS, E., ET AL. Initial Reports of the Deep Sea Project, 94, 981-996. , JANECEK, T. R., KATZ, M. E. & KEIL, D. J. 1987. Abyssal circulation and benthic foraminiferal changes near the Paleocene/Eocene boundary. Paleoceanography, 2, 741-761. MORTON, A. C. & KNOX, R. W. O'B. 1990. Geochemistry of late Paleocene and early Eocene tephras from the North Sea basin. Journal of Geological Society, London, 147, 425--437. MOUNTAIN, G. S. & MILLER, K. G. 1992. Seismic and geologic evidence for early Paleogene deepwater circulation in the western North Atlantic. Paleoceanography, 7, 423-439. M~3LLER, C. 1979. Calcareous nannofossils from the North Atlantic (Leg 48). In: MONTADERT, L. & ROBERTS, D. G. ET AL. Initial Reports of the Deep Sea Drilling Project, 48, 589-620. 1985. Biostratigraphic and paleoenvironmental interpretation of the Goban Spur region based on a study of calcareous nannoplankton. In: GRACIANSKY, P. C. DE, POAG, C.W., ET AL. Initial Reports of the Deep Sea Drilling Project, 80, 573-599. MYHRE, A. M. & ELDHOLM, O. 1988. The western Svalbard margin (74-80~ Marine and Petroleum Geology, 5, 134--156. O'NEIL, J. R., CLAYTON, R. N. & MAYEDA, T. K. 1969. Oxygen isotope fractionation in divalent metal carbonates. Journal of Chemical Physics, 51, 55475558. OWEN, R. M. & PEA, D. K. 1984. Sea-floor hydrothermal activity links climate to tectonics: the Eocene carbon dioxide greenhouse. Science, 227, 166-169. PAK, D. K. & MILLER, K. G. 1992. Paleocene to Eocene benthic foraminiferal isotopes and assemblages: implications for deepwater circulation. Paleoceanography, 7, 405422. PINET, P. R. & POPENOE,P. 1985. A scenario of MesozoicCenozoic ocean circulation over the Blake Plateau and its environs. Geological Society of America Bulletin, 96, 618-626. PEA, D. K., ZACHOS, J. C., OWEN, R. M. & GINGERICH, P. D. 1990. Global change at the Paleocene-Eocene boundary: climatic and evolutionary consequences of tectonic events. Palaeogeography, Palaeoclimatology, Palaeoecology, 79, 117-128. ROBERTS, D. G., MORTON, A. C. & BACKMAN, J. 1984. Late Paleocene-Eocene volcanic events in the northern North Atlantic Ocean. In: ROBERTS,D. G., SCHNITKER, D., ET AL. Initial Reports of the Deep Sea Drilling Project, 81, 913-923. SAVIN, S. M. 1977. The history of the earth's surface temperature during the past 100 million years. Annual Reviews of Earth Planetary Sciences, 5, 319-355. SCHMITZ, B . , HEILMANN-CLAUSEN, C., KING, C., STEURBAUT,E., ANDREASSUN,F. P., CORFIELD,R. M. & CARTLIDGE,J. E. 1996. Stable isotope and biotic evolution in the North Sea during the early Eocene: the Alb~ek Hoved section, Denmark. This volume. SCHWEITZER, H.-J. 1980. Environment and climate in the Early Tertiary of Spitsbergen. Palaeogeography, Palaeoclimatology, Palaeoecology, 30, 297-311.

471

SHACKLETON, N. J. 1974. Attainment of isotopic equilibrium between ocean water and benthonic foraminifera genus Uvigerina: isotopic changes in the ocean during the last glacial. Les Mdthodes Quantitatives d'dtude des Variations du Climat au Cours du Pleistocene, Colloques Internationaux du CNRS, 219; 203-209. 1984. Oxygen isotope evidence for Cenozoic climatic change. In: BRENCHLEY,E (ed.) Fossils and Climate. John Wiley & Sons, 27-34. - & BOERSMA, A. 1981. The climate of the Eocene ocean. Journal of the Geological Society, London, 138, 153-157. , CORFIELD, R. M. • HALL, M. A. 1985. Stable isotope data and the ontogeny of Paleocene planktonic foraminifera. Journal of Foraminiferal Research, 15, 321-336. , HALL, M. A. & BOERSMA, A. 1984. Oxygen and carbon isotope data from Leg 74 foraminifera. In: MOORE, T. C. JR, RABINOWlTZ, P. D., ET AL. Initial Reports of the Deep Sea Drilling Project, 7 4 , 599-612. SPERO, H. J. & LEA, D. W. 1992. Intraspecific stable isotope variability in the planktonic foraminifera Globigerinoides sacculifer. In: SARNTHEIM, M., THIEDE, J. & RAINER, Z. (eds) Geomar Report 15. Berichte-Reports, Geologisch-Pal~iontologisches Institut, Universitiit Kiel, 57, ICP IV, Program and Abstracts, 268-269. STOTT, L. D., KENNETT, J. E, SHACKLETON, N. J. & CORFIELD, R. M. 1990. The evolution of Antarctic surface waters during the Paleogene: inferences from the stable isotopic composition of planktonic foraminifers, ODP Leg 113. In: BARKER, P. E, KENNETT, J. P., ET AL. (eds) Proceedings of the Ocean Drilling Program, Scientific Results, 113, 49-863. --, SINHA, A., THIRY, M., AUBRY,M.-P. & BERGGREN, W. A. 1996. GLOBAL ~13C changes across the Paleocene-Eocene boundary: criteria for terrestrial-marine correlations. This volume. THIEDE, J. & ELDHOLM, O. 1983. Speculations about the paleodepth of the Greenland-Scotland Ridge during Late Mesozoic and Cenozoic times. In: BOTT, M. H. P., SAXOV,J., TALWANI,M. & THIEDE, J. (eds) Structure and Development of the GreenlandScotland Ridge. NATO Conference Series IV: Marine Sciences, Plenum Press, New York, 8, 549-589. --, -& TAYLOR, E. 1989. Variability of Cenozoic Norwegian-Greenland Sea paleoceanography and northern hemisphere paleoclimate. In: ELDHOLM, O., THIEDE, J., TAYLOR, E., ET AL. Proceedings of the Ocean Drilling Project, Scientific Results, 104, 1067-1118. TUCHOLKE, B. E. & MOUNTAIN, G. S. 1986. Tertiary paleoceanography of the western North Atlantic Ocean. In: VOGT,P. R. & TUCHOLKE,B. E. (eds) The Geology of North America, Volume M, The Western North Atlantic Region. Geological Society of America, 631-650. , & VOGT, P. R. 1979. Western North Atlantic: sedimentary evolution and aspects of tectonic history. Initial Reports of the Deep Sea Drilling Project, 43, 791-825.

472

S.D. CHARISI 8z B. SCHMITZ

VINCENT, E., KILLINGLEY,J. S. & BERGER, W. H. 1981. Stable isotope composition of benthic foraminifera from the equatorial Pacific. Nature, 289, 639-643. WOLFE, J. A. 1978 . A paleobotanical interpretation of Tertiary climates in the northern hemisphere. American Scientist, 66, 694-703. WOODRUFF, E, SAVIN, S. M. t~ DOUGLAS, R. G. 1980.

Biological fractionation of oxygen and carbon isotopes by recent benthic foraminifera. Marine Micropaleontology, 5, 3-11. ZACHOS, J. C., LOHMANN, K. G., WALKER, J. C. G. & WISE, S. W. 1993. Abrupt climate change and transient climates during the Paleogene: A marine perspective. The Journal of Geology, 101, 191-213.

Index passim indicates scattered references between the pages. tables and figures are in italic Aalbeke Clay Member 32, 298-9 Aalterbrugge Formation 36-7 Abbey Wood, London Basin 207, 213 Airy backstripping 44-6 Airy isostasy 46, 55 Alba Formation 83 Alba sequence 109-11 Albmk Hoved section, Denmark 277, 279-80, 300, 301, 302 Alpha Ridge 267 Andrew sequence 23, 24 Andrew-Maureen formation boundary 23 Anton Dohrn Seamount 69, 70, 74 4~ dating, -17 and +19 ash beds 317 Arctic Ocean, diatomaceous sequences in 263 Arctic-Atlantic connection 464 Argiles ~ Lignites 342 Argiles d'Epernay 342-3 Argiles plastiques (badol6es and kaolinitiques) 4, 8, 342, 345 ash deposits/falls 317, 328-9, 376, 421,463-4 Denmark and North Sea 277, 277-8 Eocene 277, 377 North Atlantic ash series 326, 328-9, 331 North Sea 389 Site 550 317 see also tephra layers 'ash series', SE England 219 Avenay, Pads Basin 206, 207, 211 Balder Formation 7, 28, 82, 95, 100, 107, 238, 240, 249, 263, 266, 268, 328, 346 correlation with Wrabness Member 240 onlapping Erlend volcanic succession 72 onlapping Lower Lava Formation 65 Balder Sands 28 Baider sequence 28, 98, 106-8 Balder-Frigg sequence boundary 109 Balder/Sele boundary 246 Balmoral Tuffite 24-5 Barents Sea 301 Barents Shelf 301 Beauly Member/Formation 49, 51, 82, 100, 101,107-8, 249, 250 Belgium 213, 214, 344 lower Eocene palaeomagnetic data 139, 142 mammalian faunas 207, 213, 214, 343, 344 Beltinge Fish Bed 175 benthic foraminiferal extinction event, Paleocene/F_,ocene boundary 326, 339, 374, 381,403-4, 406, 447 Central North Sea 5 Denmark 279, 296 Georgia 323, 326 Israel 323, 326 Maud Rise and Walvis Ridge 408-13, 417-20, 422 Spain 323, 326 Sites 525 and 527 408, 413 Site 549 363,364, 365

Site 550 360, 362, 365 Site 577 323 Site 689 408, 413 Site 690 326, 339, 357, 359, 383, 408, 412 Beryl Embayment 105, 108, 111 Balder sequence 107 Dornoeh sequence 106 Maureen Formation 103-4 Bill Bailey Bank 70 Blackheath Beds 213, 219, 220, 244, 342 Blackstones Centre 70, 75 Borehole 79/07a 223,226, 233, 236, 237 Borehole 81/46a 233, 236, 237,238,239, 240, 246, 248, 249 Borehole 82/12 73 Borehole 85/2B 74 Borehole 90/3 249, 250 Borehole 404T (Jubilee Line) 185-92 Borehole 810 (Jubilee Line) 224, 227, 235, 237, 239 Borehole A4A (CTRL) 224, 227, 235, 238, 239 Bracklesham Bay, dinoflagellate stratigraphy 30, 31 Bracklesham Beds 30 Bracklesham Group 344, 346 Bradwell boreholes 25, 247 Borehole 205 226, 236, 238, 239, 240 Borehole 206 234 Borehole 209 138-9, 239 Borehole 217 138-9 magnetostratigraphy 138-9 Brendan Seamount 70, 72 Brioc sequence 34, 35 British Tertiary Igneous Province (BTIP) 1, 4, 347 Brito/Arctic Igneous Province 266 Brussels Sands 32, 33 Buchan Graben 46, 104, 105 Buchan Ridge 105 Bunkers Hill Borehole 230 calcareous nannoplankton biostratigraphy 2-3, 95, 287, 288, 335-6 Belgian Basin 35, 343-4, 346 Denmark 278-9, 298, 302-3, 460, 464 London Basin 25, 32, 34-5, 139, 141,160, 187-8, 249, 328, 340-3, 346, 389 North Atlantic 313-31 passim, 353-74 passim NP8/NP9 313, 346, 363 NP9/NP10 see existing entry NP10/NPll zonal boundary 355, 370 NPll/NP12 zonal boundary 317, 358, 361 Paris Basin 343-4 Site 549 313-15 Site 550 315-17, 333,334, 339, 341,359-62, 365, 367, 389, 462-6 Site 577 321-3 Site 690 319-20, 356-9, 365, 393 Site 690B 358 Sites 698 and 700 323 Upnor Formation 131, 186-8 473

474

INDEX

Caran Sandstone 34, 101,109, 110 Caran sequence 34, 35 Caravaca section 326, 419 carbon isotopes 53-4, 59, 215, 313-17, 319-47 passim, 363-74 passim, 381-97, 404-19 passim, 443-4, 447-50, 458-9, 460--2, 466-7 excursions 323, 327, 342, 345, 374, 381-2, 389, 393, 394, 395, 445, 447, 448-9, 450, 452-3 and the Paleocene/Eocene boundary 338, 345, 354 Cenodiscus claystone 23 Central Graben 104, 105, 106, 109, 346 Central London boreholes 139, 185 Chalk Group 100 Channel Tunnel Rail Link Borehole A4A 224, 227, 235, 238, 239 charophytes 211,213 Chichester Harbour section 228, 229, 235, 237, 238, 239 Chingford South Borehole 225, 227, 235, 237, 238, 239 Chron C23r/C24n boundary 319 Chron C24r/C25n boundary 331,359, 372 Chron C25n hiatus in London Basin 27, 134, 150 and the 'Oldhaven magnetozone' 195 Upnor Formation, London Basin 150, 185-92 climatic warming 7-8, 105, 119, 277 around the Paleocene/Eocene boundary 267 atmospheric CO 2 402-3, 443 long-term 466 Cond6--en-Brie, Pads Basin 207, 210 condensed sections, London Basin 169-70 primary 145-81 passim secondary 169, 178 ConglomErat de Cernay, age of 343 'conglom6rat de Coryphodon' 211 Conglom6rat de Meudon 342, 343 Crawford Ridge 88 Crawford Spur 108 Cretaceous, late, ~13C and 5180 data 445, 448-9 Cretaceous/Tertiary boundary see KIT boundary Cromarty Sandstone Member 177, 180 Cuisian (Ypresian)-Lutetian boundary 32 Danian 328 Danian/Selandian boundary 119 Darwin Complex 71, 72, 74-5 Denmark, late Paleocene to early Eocene sediments 277-8 depositional sequences, Paleogene, NW Europe 17 diatom biostratigraphy Fur Formation 259, 260-1,262 Fur-Olst formations 336 Harwich Formation 262 diatom blooms 268, 269 diatomaceous sequences 263, 265 diatoms 108, 230, 255-6, 257, 259-60, 355 ecology of 267-8 dinoflagellate cyst biostratigraphy graphic correlation, North Sea Basin 24, 26, 29, 31 late Paleocene-early Eocene, Denmark 278-9 NW Europe 421 Paleocene/Eocene boundary, correlation with mammal zones 214-15 Paleocene-Eocene

North Sea Basin 91-112, 122-7 NW Europe 328, 336, 342, 346 Paleocene--Oligocene, North Sea Basin 23-6 upper Paleocene to lower Eocene, SE England 145-80 correlation with North Sea Basin 169-70, 173-80 dinoflagellate cysts 3 palaeoecology and relative sea-level change 170-1, 243-8 dinoflagellates, Wursterheide borehole 28 Dormaal, Belgian Basin 207, 208, 210, 213 Dornoch Formation 49, 100, 101, 105-6 Domoch sequence 98, 105-6 Dowsett's Farm Borehole 224, 227, 234, 237, 238, 239 East Shetland Basin 104, 106 East Shetland Platform 59, 87, 107-8, 110, 111 ebridians 256, 267 Egem Sand (and Stone Band) 32, 33 Egem-Panisel Sands 346 Ekofisk Formation 20-2, 59, 328 Ekofisk sequence 101 English Channel 276, 300, 301,302 Eocene, base of 355 Erlend/West Erlend Complex 65, 71-2, 74-5 Erquelinnes 210, 213 eustasy 5 first-order sea-level variations 58 global, long-term 55-8 eustatic curve 17, 19-20, 22, 30, 36 extinctions mammal 205 see also benthic foraminiferal extinction event Faeroe Channel 301,302 Faeroe Plateau Lava Group (FPLG) 65, 69 Faeroe Trough 300 Faeroe-East Greenland (Blosseville Coast) basalts 347 Faeroe-Greenland igneous province 328 Faeroe-Shetland Escarpment (FSE) 65 Faeroe-Shetland Intrusive Complex (FSIC) 69, 73 Faeroes Intrusive Group 69 Fausses Glaises 342 Ferry Cliff, London Basin 224, 233, 236, 237, 240 Fladen Ground Spur 83, 87, 104, 110 foraminiferal biostratigraphy 109 benthic 7, 95, 104, 280 agglutinated 25, 104, 408 calcareous 296 planktonic 3, 95, 104, 280, 391 datum levels, Site 550 336-7,337 R0sn~es Clay, ~13C variations in 284-7 Formation de Varangeville 328 Forties Formation 346 Forties Salt Platform 105, 109 Forties Sandstone Member 7, 8, 25, 105, 117, 124-7, 175, 176, 179, 328 biostratigraphy 121-2 Forties sequence 105 Forties-Montrose High 108, 109 Frigg field, sand-rich depocentre 106 Frigg Formation 83 Frigg sequence 95, 108-9 Fulmer, M40 Motorway cuttings 225-6, 229, 234, 237

INDEX Fur Formation 219, 256, 268, 269, 278 ash beds in 328-9 diatoms from 259, 260-1,262 Geikie central volcano 69 Gelinden Marl 25 genetic stratigraphic sequences 99 Geomagnetic Polarity Time Scale (GPTS) 310, 328, 333, 335, 376-7 and the Paleocene/Eocene boundary 329, 331,337-8 Glen Bay granophyre 70 global warming 353, 381 atmospheric CO 2 402-3, 443 Goban Spur 385 graphic correlation 17, 18, 24, 26, 29, 31 Greenland-Scotland Ridge 301 as land bridge 4 proto-Greenland-Scotland Ridge 328 water mass salinity to north of 465-6 Grid Formation 111 Grid sequence 98, 111-12 Hainin Formation 23 Hales Borehole 141,221-50 passim magnetostratigraphy 134-6 Hales Clay 134, 135, 136, 141,143, 220--44passim, 310, 329, 337, 340 Halesworth Borehole 136-7 upper Paleocene/lower Eocene magnetostratigraphy 136-7 Halibut Horst 49, 59, 104 Hampshire Basin 111,131,132, 143, 219, 239, 244 Harre borehole 262 Harwich 200 mammal faunas 210-11 Harwich Formation 3, 131-2, 139, 143, 168, 176-7, 179, 180, 185, 220, 232-40, 331,340, 342, 389 base of 224-30, 337, 338, 345 depositional history 220-1,241, 243-8, 249 magnetostratigraphy 195, 201 palynology 230-40 sequence stratigraphy 245-8 stratotype at Wrabness 221 tephras of 132, 196 Harwich Member 220, 239 Harwich Stone Band 177, 196, 198, 200, 210, 213, 221 magnetostratigraphy 199, 20.1 Hebridean-Thulean volcanic province 328 Hebrides Terrace Seamount 69, 70, 75 Herne Bay, Kent 115, 146, 149-50, 153, 160-1, 179, 214, 225, 235, 237, 238, 239 aquatic palynomorph distribution 153, 160-1 dinoflagellate cyst biostratigraphy 154, 165-8 sequence stratigraphy 175-80 stratigraphic summary log 148, 228 Herne Bay Member 220 hiatuses, late-Paleocene-early Eocene 310 Holmehus Formation 162 Horda Formation 101,108, 109, 111 hydrocarbon shows, distribution of 88 hydrocarbons reservoirs 79, 269 Tertiary, exploration for t 12 Hyon Sand Formation 299

475

Iceland plume 60, 63-4, 74, 75 Iceland-Faeroe gateway 464 leper Clay Formation 139, 278, 337, 344 base of 345 IGCP Project 124 17, 36 IGCP Project 308 309-10, 312, 340, 354 Indian Ocean, intermediate water 451 Inner Moray Firth 59 Intra-Frigg terrace 32, 33 iridium spike, ROsnms Clay 279 isotope stratigraphy 4 Sites 690B, 550 and 549 367-70 isotopic excursions rapid, possible reasons for 420-1 see also carbon isotopes; oxygen isotopes Jaeren High 82, 105 Jubilee Line Borehole 404T 185-92 Jubilee Line Borehole 810 224, 227, 235, 237, 239 K/T boundary 101,445, 447, 448-9 kaolinite abundances 402 kaolinite influx 5, 7, 105 Khieu River section 326 Knokke Member 214 Knokke well 28, 343 Knowl Hill section, Twyford 228, 229, 234, 237, 238, 239, 247 Knudshoved Member 278, 279, 300 Kortemark Silt 32 Lambeth Group 25, 27, 131,139, 141,143, 146, 161, 166, 185-6, 249, 339, 340-1 see also Upnor Formation; Woolwich and Reading formations Landen deposits 27 Lark Formation 36 Leg 74 sites, C and O isotope 288, 448-9, 450, 452-3 Leg 113 sites, C and O isotopes 447, 448-9, 450, 452-3 Lellinge Greensand 24 Levington 198-9, 202, 236, 240 magnetostratigraphy of 200 Lilleb~elt Clay Formation 34, 278, 296, 300 Lista Formation 7, 25, 100, 101,104, 117, 173, 175, 178, 179, 328 green claystones 127 Lista sequence 104 Lofoten Basin 301 London Basin 131,132, 143, 196, 219, 244 Chron C25n in 185-92 mammalian fauna sites 207 London Clay basement bed 219 London Clay Formation 28, 32, 132, 139, 142, 146, 161, 166, 168, 178, 185, 191-2, 192, 214, 219, 220, 278,296, 328, 342, 344, 346 base of 354 London-Hampshire Basin 30 Blackheath-Hales Clay sequence 27 Priabonian deposits 35 upper Eocene depositional sequences 35 Lopra-1 borehole/well 65 Lousy Bank 70 Lower Balmoral Sandstone 173, 179 Lower Balmoral sequence 24-5

476 Lower Forties Sandstone .25, 27 Lower Forties sequence 25 Lower Frigg sequence 28, 32 Lower Lava Formation 65, 74 Lower Upnor 150 aquatic palynomorph distribution 155, 162 dinoflagellate cyst biostratigraphy 168 sequence stratigraphy 175-8, 179-80 stratigraphic summary log 149 magmatic underplating, northern North Sea 60 magnetobiochronology integrated, biochrons NP9 and NP10 327 NW Europe 309-47 magnetochronology Cenozoic 339 Chron C24n, duration 341 Chron C24r, chronology 372-5 magnetostratigraphy 2, 4 Borehole 404T 188-92 Bradwell, Essex 138-9 Faeroe Plateau Lava Group (FPLG) 65, 69, 72 London Basin, Paleocene 134-42 passim, 150, 173-7 Site 549 363, 365 Site 550 359, 365 Site 690 319, 357 Site 690B 356, 365 Thanetian/Ypresian stratotypes 129-30 mammals 3-5 biochronology, Europe 205-6 Europe-North America correlations 21 4-15 mammalian faunas, parsimony analysis 206-7 mammalian MP and PE zones 205-8, 210-14 Paris Basin 342-4 taxonomic turnover 210 madne-terrestrial correlation, NW Europe 339-45 Marnes de Gelinden 343 Marnes de Meudon 22 Maud Rise 356, 389, 405, 408-13,417-18,445 Maureen Formation 3-4, 100, 101, 117 Maureen sequence 23, 24, 101-4 maximum condensation surfaces 95, 100, 101, 106 top Balder sequence 108 top Dornoch 108 top Grid sequence 111 maximum flooding surfaces 98 Mediterranean, modem, deep waters of 444 Merelbeke Clay 32 Meudon, Paris Basin 206, 207, 210 microfaunal zonation, for the North Sea 95 microfossils 3 see also siliceous microfossils Middle Frigg sequence 30, 32 Middle Lava Formation 65, 73, 75 Middle Sele sequence 27 the Minch 73-4 Minch Fault 74 Mo Clay see Fur Formation Mons-en-P61~ve Sand Member 298 Mont Aim6-Vertus Formation 22 Montian stage boundary 22 Montrose Group 94, 100, 328 Moray Delta 87, 88 Paleocene bathymetric high 82

INDEX Moray Firth Basin 101,103, 104, 106, 107-8, 109, .111 Moray Group 94, 100, 101,328 Mousa Formation 101, 108, 109, 110, 111 Mutigny, Paris Basin 206, 207, 210, 211 Nacton 233, 236, 237, 240 Nahal Avdat section 327, 355 Nauchlan sequence 34, 35 Noah American Land Mammal Age (NALMA) 344-5, 354 North Atlantic 313-31,353-74 bottom waters in early Eocene circulation 466-8 Cretaceous and Paleogene deep waters 454 early Eocene 463-6 North Atlantic sites, bottom-water cooling 466 opening of 266 warm saline bottom water (WSBW) in 451,454 North Atlantic Deep Water (NADW) 463 North Atlantic Igneous Province (NAIP) 1, 63, 74 North Atlantic Ocean, initiation of 5, 73, 74 North Atlantic-Arctic rift system 276 North Rockall Trough 73 North Rockall Trough-Hebrides Lavas Group (RTHG) 69-70 North Sea anoxia, late Paleocene--early Eocene 269, 296 basin stratification, late Eocene-early Paleocene 268 carbon isotopes, early Eocene 285-6 circulation, latest Paleocene 268 northern, Paleocene uplift 60 oxygen isotopes 301-2 palaeocirculation 267 Paleocene-Eocene succession 94-101 restriction and eventual isolation of 7-8 semi-enclosed, early Eocene 299, 299-301 water exchange with ocean 303 North Sea Basin controls on Eocene sand deposition 88 earliest Eocene sediments 221 Paleocene-Eocene siliceous microfossils 255-70 palynomorph correlation with E England 248-9 sand-supply routes, Eocene 83-4, 90 structural elements 94 subsidence 346 water mass stratification, early Eocene 286, 300, 303, 462-3 North Sea sequences, correlation with NW Europe 346 North Viking Graben 60, 105, 106, 109 northern component water (NCW) 444 Norwegian--Greenland Sea 301,302, 464 connection with Arctic Ocean 465, 466 opening of 83, 463 NP9/NP10 chronozonal boundary 374 NP9/NP10 zonal boundary 318, 321,323, 326, 337, 338, 355, 358, 359, 363, 370, 373, 374, 375 estimated age of 376 hiatus at, Site 550 393 stratigraphic continuity, Site 690B 366-7 ocean-atmosphere 13CO2exchange 393-5 oceans decrease in productivity 403 heat transport, theory of increase in 403

INDEX long-term decrease in carbon isotopic composition 381,383 long-term warming of, transitional Paleocene-Eocene 402 mean salinity, Cretaceous 454 oxygen conditions, low, at extinction events 403 proto-Gulf Stream circulation, development of 466 Odin sandstones 107, 108 Oldhaven Beds 160, 161,219, 220, 225, 238, 244, 329, 337, 340, 346 ashes in 240 Oldhaven Formation 219 'Oldhaven magnetozone' 195-202 Oldhaven Member 146, 150, 168, 180, 237 basal pebble bed 177 ~Jlst Formation 27, 28, 162, 219, 279, 296 ash layers in 278 Orchies Clay Member 298 Orchies sequence, Belgium 32 Orkney-Shetland Platform 43 Ormesby Borehole 134, 141,221,223, 226, 233, 236, 237, 238, 239, 246 magnetostratigraphy 134, 135 Ormesby Clay Formation 130-1,134, 135, 136, 137, 141,142 Orwell Member 220, 236, 240, 243 Dsterrenden Clay 25 Outer Moray Firth Cretaceous succession 43 earliest Eocene, correlation with SE England 248 subsidence analysis 51-3 Tertiary succession 43 oxygen isotopes excursions 416-17, 420 Site 690 383 Paleocene Ocean 445,447,448-9 Site 550 459-60, 461 palaeoceanographic events, and isotopic shifts 290 palaeogeography, North Sea early Eocene 301-2 late Paleocene and early Eocene 266-7 Paleocene ocean, deep water circulation 419-20, 422, 443-54, 466-7 Paleocene/Eocene boundary 27, 28, 71,129, 143, 249, 255, 310 age of 406-7 C and O isotopes 393-5, 394, 447, 448-9, 450, 460 and a carbon isotopic excursion 338, 345, 354 current delineation 345 in Denmark 278-9 identity and geochronology, NW Europe 327-39 marine correlations around 354-78 N America-NW Europe relationship 344-5 North America 337 and the NP9/NP10 zonal boundary 311,345, 376 offshore diatom occurrences around 263 and the P5/P6 boundary 311 Pads Basin 384-5 Site 401 318 Site 690B 319, 320, 321 Site 698A 324 Site 702B 325

477

Paleocene/Eocene boundary events 5-9, 205 Site 690B 359 Paleocene/Eocene boundary interval 3 central North Sea lithostratigraphy 6-8 planktonic foraminiferal zonation 311-12, 406 Paleogene R/TF cycles 20, 22-7, 30-6 palynological zonation schemes, from oil industry studies 122-3 palynomorph association sequences 132, 250 palynomorph-palynofacies data, Paleogene Central North Sea 121-8 palynomorphs 3, 7 aquatic 171 Heine Bay 153, 160-1 Lower Upnor 155, 162 Pegwell Bay 151,159 Wrabness 157, 164-5 marine vs. non-marine, Well 22/10a-4 116, 117-19 Paris Basin 22, 27, 34, 35, 354, 395 mammalian faunas 206, 207, 342-4 Paleocene/Eocene boundary problem 384-5 preliminary carbon isotopic results from 397 Priabonian deposits 35 Sparnacian facies 342-3 Pegwell Bay, Kent 142, 149-50, 161-2 aquatic palynomorph distribution 151,159 dinoflagellate cyst biostratigraphy 163-5 stratigraphic summary log 147 Thanetian sequences 173, 175, 179 Pegwell Marls 25, 146, 163--4, 173, 179, 328, 343 planktonic foraminiferal biostratigraphy late Paleocene--early Eocene 310-21,326, 346, 354, 403 P4/5 zonal boundary 313, 321,354 P5/P6a zonal boundary 317, 321,323, 326, 337, 338, 354 P6b/P7 zonal boundary 354 Paleocene North Sea 230 Possagno section 355 Pourcy, Pads Basin 206, 207, 210, 211 Priabonian 35 radioisotopic ages, -17 and +19 ashes 333, 335 radiolada 109, 256, 257, 259, 326 radiometric dating 4 Reading Beds 8, 28 Reading Formation 27, 228,230, 340 see also Woolwich and Reading formations Reculver Silts 25, 115, 119, 146, 160, 164, 165, 173, 175, 179, 340 Reculver Tabular Band 175, 178 red mudstone 28, 131,137, 138 relative sea-level change 240-2, 243 and dinoflagellate cyst palaeoecology 170-1 rifting early Tertiary 59 Kimmeridgian 58 Mesozoic 43 North Rockall Trough 74 Rockall Complex 70-1 Rogaland Group 328 Rona Ridge 73 Rosemary Bank Seamount 70, 74

478

INDEX

RCsn~es Clay Formation 28, 277,278,279-80, 281,300, 301,303, 346 biostratigraphy 292-9 isotope data 284-90, 289 Roubaix Clay 298 Roubaix sequence, Belgium 32 Sables de Mons-en-P61~ve 28 Sables de Sinceny 342 Sables de TiIlet 25 Sables d'Erquelinnes 329 St Kilda central volcano 69 St Kilda Complex 70, 75 Scotland-Shetland landmass 94, 100, 109 Scottish Highlands 43, 346 sea-floor spreading, Norwegian-Greenland Sea 60, 74 Selandian 328 Selandian/Danian stage boundary 22-3 Sele Formation 3-4, 7, 8, 25, 100, 105, 106, 117, 124, 177, 180, 236, 240, 249, 263, 268, 328, 329, 346 black shales 126 Sele/Lista formation boundary 7, 25 sea-level fall at 8 sequence biostratigraphy North Sea 169-70 Upper Paleocene-Lower Eocene, SE England 169-71 sequence boundaries, late Paleocene-early Eocene Belgium 344 London Basin 180, 340, 341 sequence cycle curve, for NW Europe 36, 37 sequence stratigraphy 5, 15 Belgium Basin 22, 25, 27, 28, 30, 32, 35 Danish Basin 22, 23-4, 25, 28, 104 London-Hampshire Basin 27, 30, 32, 35 northern Germany 28, 30, 32, 33 NW European basins 339-40, 340, 346 Paleocene-Eocene, North Sea Basin 98-112 Paleocene/Eocene boundary events 5-9 Paris Basin 22, 25, 27, 28, 34, 35, 36 Rupelian and Eocene-Oligocene transition 36 Whitecliffe Bay 30, 31, 32-3, 34-5 S6zanne-Broyes, Paris Basin 207, 210 Shamblehurst Borehole 228, 229, 230, 235, 237, 238, 239 Shelford Pit 228, 235, 237, 239 Sheppey 1 Borehole (Harty Borehole) 225, 235, 238, 239, 329 Sheppey, Isle of 247 Shetland Platform 43 tectonic uplift 27 Shetland-Faroe Channel 276 Shotley Gate Borehole 221,223, 233,236, 237, 238, 240 siliceous microfossils, North Sea Basin 255-70 silicoflagellate biostratigraphy 256, 259, 263, 329 Sinceny, Paris Basin, mammal faunas 210, 211 Site 384 310, 450 C and O isotopes 445, 447, 448-9, 452-3 Site 401 317-19, 463, 466, 467 Site 403 355 ashes in 328, 329 Site 404 355 Site 525 405 C and O isotopes 407, 441 faunal data 411,436-7

Site 527 405 C and O isotopes 407, 441 faunal data 412, 438-9 Site 549 313-15, 354, 355, 359, 363-5, 372, 382, 385, 385, 391,419 benthic foraminiferal extinction event 326 carbonate dissolution 354 and the Chron C25n/C24r boundary 372 correlation with Sites 690B and 550 365-7, 371 stratigraphic gaps 363, 365 Site 550 243, 278, 281,310, 312, 315-17, 327, 333-4, 354, 355, 367, 370, 372, 382, 385, 386-7, 391, 465 ashes 328, 329 ash layers -17 and +19 311,317, 331,339, 354, 373, 376, 389 benthic foraminiferal extinction event 326 C and O isotopes 290-2, 293, 459-62, 461,463 calcareous nannoplankton zones/markers 341 carbonate dissolution 315, 317, 331,354, 373, 393 correlation with sites 549 and 690B 365-7, 371 development of vertically isothermal conditions, early Eocene 463 early Eocene water depths 457-8 magnetobiostratigraphic correlations 330, 359-62 NP9/NP10 zonal boundary 393 planktonic foraminiferal zones/markers 341 relative chronology in the NP10 zonal interval 372-4 sedimentation rates 333, 334--6, 339, 373--4 unconformities 315, 331,332, 333, 359, 365 Site 577 321-3,326, 419,443, 466, 467 C and O isotopes 445, 447, 448-9, 450, 452-3 Site 643 302 Site 689 405 C and O isotopes 407, 440 faunal data 409, 430-1 Hole 689B 405 Site 690 353, 354, 354-5, 356-9, 382, 405, 466, 477 C and O isotopes 354, 407, 440 faunal data 432-5 Hole 690B 319-21,355, 370, 374-5, 388 benthic foraminiferal extinction event 320-1,326 correlation with sites 549 and 550 365-7, 371 sedimentation rates 374-5 NP9/NP10 zonal boundary 393 Site 698 323, 326 Site 700 323, 326 Site 703 323, 326 Site 738 419, 445, 452-3 Hole 738C 451 Site 747 419 Site 752 419, 451,452-3 Site 762 419 Site 865, ~13C excursion 419 Sizewell Borehole 141,224, 233, 236, 237, 238, 239, 240 magnetostratigraphy 137-8 Skagerrak-Kattegat Platform, uplift of 35-6 Soissons, Paris Basin 207, 210, 211 Sole Pit structure 236 South Mimms, M25 Motorway cutting 225,229, 234, 237, 238,239 South Viking Graben 60, 86, 101,103, 105, 106, 108, 110, 111

INDEX South Viking Graben boundary fault 111 Southern Ocean deep waters source 447 'younger' bottom waters 467 southern ocean water (SOW) 444 Soviet Union (former), diatomaceous sequences 263 Sparnacian 27, 28, 139, 278, 337, 342-3, 354, 384-5 MP8-9 fauna in 342 stable isotope correlations regional, upper Paleocene-lower Eocene 377-8 see also carbon isotopes; oxygen isotopes stage boundaries, Paleogene 129 Stanford Rivers Borehole 224, 227, 234, 237, 238, 239, 247 Stourmouth Clays 146, 160, 163, 343 stratigraphic traps 79 Stronsay Group 94, 100, 101 Studd Hill, Herne Bay 214 submarine fans Paleocene, Central North Sea 121,123, 137 Paleogene, North Sea 23, 24, 34, 37, 178, 346 subsidence North Sea Basin 346 Outer Moray Firth Cretaceous-Tertiary, errors in analysis of 53-5 McKenzie subsidence 46, 49 Paleogene 46, 53, 59, 109 Paris Basin, controlling Upper Eocene gypsum deposition 35 see also thermal subsidence Suffolk Pebble Beds, London Basin 207, 208, 210, 221, 244, 342 mammal fauna possibly reworked 213 Swanscombe Member 219, 220, 221,224, 225 Tampen Spur 106 Tay Formation 83 Tay sand 34 tectonic cycles, Outer Moray Firth 94-5 tectonism earliest Eocene central North Sea 83, 84 NE Atlantic 457 and North Sea Basin evolution 276-7 and Paleocene/Eocene carbon shift 421 and restricted oceanic circulation 295-6 rifting of N Atlantic 249 and transfer of global silica 267 see also volcanism tephra layers, Harwich Formation 221,240, 246 tephrostratigraphy 5 and NP9/10 boundary 7 Tethys, home of Paleocene deep water? 444 Thames Group 131-2 see also Harwich Formation; London Clay Formation Thanet Sand Formation 3, 8, 25, 131, 138, 141,142, 160, 161,165-6, 168, 185, 186 at Pegwell Bay 161-2 Base Bed 146, 163, 173 sub-aerial weathering of 142 see also Pegwell Marls; Reculver Silts; Stourmouth Clays

479

Thanet Sands 328 Thanetian 24-5, 328 global stratotype point for base of 178 sequences, SE England 173-6, 179-80 type section of 149-50 thermal subsidence Cretaceous and Tertiary 49 post-rift 43, 58 reverse modelling of 46 Tilehurst Member 219, 220, 225 Tilehurst Quarry 226, 228, 234, 237, 238, 239 stratotype section, Tilehurst Member 228, 229 timescales see Geomagnetic Polarity Time Scale (GPTS) Tor Formation 20 transgressions 219, 221 early Eocene 267 London Clay 240 main Upnor Formation 141 Try, Paris Basin 210, 211,213 tuffaceous lacustrine sediments 65, 69 Tuffeau de Lincent 25, 344 tufts 27 Faeroes and Balder Formation 69 North Atlantic 328 Twyford Member 219, 220 unconformities 37, 101 base Harwich Formation 341-2 base Rupelian 33 and the Chron C24r interval 370-1 Ormesby Clay-Hales Clay 134, 136, 141 representing lowstands 173, 175, 176, 177, 178, 179, 180 Site 549 313 Site 550 315, 331,332, 333, 359 Site 690B 319, 358-9 stage boundaries as 23 Thanet Sand Formation-Lambeth Group 150 Whitecliffe Bay 33 uplift 15, 27, 43, 91 regional 58 and Iceland plume 60 Paleocene 51, 53, 55, 59, 60 Upnor Formation 8, 27, 34, 131,137, 142, 150, 160, 166, 168, 278, 329, 339, 340, 343 basal conglomerate 175, 176 calcareous nannoplankton biostratigraphy 186-8 Central London 186 Chron 25n in 139, 160 depositional environment 188 as highstand deposits 192 palynoflora 188 see also Woolwich Bottom Bed Upnor Pit 225, 228, 235, 237, 238,239 Upper Balmoral Sandstone 173, 175, 179 Upper Balmoral sequence 25, 27 Upper Chalk 146 Upper Forties sequence 27 Upper Frigg sequence 30, 34 Upper Lava Formation 65, 73, 75 Upper Sele sequence 27 Upper Tay sequence 34-5

480

INDEX

upwelling in a 'greenhouse world' 267 and isotopic composition 389 Maud Rise area 418 North Sea Basin 267, 268-9 Utsira High 110, 111 Varangeville 139, 142 Vestmanna-1 borehole/well 65 Viborg Borehole 23-4, 162, 178 Viborg Formation 36 Viking Graben 82, 84, 104, 105, 107, 109, 110 Vlierzele sequence, incisive 32, 33 volcaniclastic deposits 24-5, 72-3, 75 volcanism 5,266, 421-2 early Eocene and deep sea circulation in the North Atlantic 463-6 early Paleogene 328-9 East Greenland-Hebridean province, significance of 347 Thulean 276, 277-8 VOting Plateau Lava Lower Series 71 VOting Plateau volcanic margin 72 Walton Member 132, 143, 178, 196, 198, 199, 219, 221, 225, 228 magnetostratigraphy 200, 201 Walton-on-the-Naze 234, 239 magnetostratigraphy 200 summary log 224 Walvis Ridge 405, 408-13, 418-19, 445 Wardrecques Member type section 139, 142 warm saline bottom water (WSBW) 403 and early Eocene deep water circulation 463 in the North Atlantic 451,454 Waterschei Clay 25 Well 14/25-1 246, 248, 249 Well 15/28a-3 247, 248,249, 264 Well 21/9-1 264 Well 22/10a-4 6, 7, 115-19 Well 205/9-1 65 Well D.G.I. 83101 25, 27 West Central Graben 82 West Fladen High 49

West of Shetland 250 West Shetland Basin 73 Western Platform 83, 87 Whitecliffe Bay (loW) 139, 142, 196, 228, 230, 235, 238 Whitecliffe Bay section 34--5 sequence stratigraphy 30, 31, 32-3 wireline logs 98-9, 101, 108 Witch Ground Graben 46, 83, 103, 104, 105, 110, 111, 346 Witteting Formation 344, 346 Woolwich Beds 278 Woolwich Bottom Bed 146, 166, 176 Woolwich Formation 27, 137, 138, 168, 340 Woolwich and Reading Beds 131,328, 340 Woolwich and Reading formations 131,139, 142, 191, 192 interdigitated 136 Woolwich Shell Beds 176, 178, 180, 213, 329, 343 Woolwich Shell Beds/Oldhaven Beds hiatus 342, 346 Wrabness 146, 150, 179, 198, 233, 237, 238, 239 aquatic palynomorph distribution 157, 164-5 dinoflagellate cyst biostratigraphy 169 magnetostratigraphy of 200 stratigraphic summary log 150, 223 Ypresian sequences 177-8, 180 Wrabness Member 134, 136-7, 146, 169, 177, 196, 198, 199, 220, 221,223, 238, 246, 247, 310 magnetostratigraphy 202 Wrabness/Hales boundary (Norfolk) 246 Wursterheide borehole 28, 30, 32, 33 dinoflagellates 28 sequence stratigraphy 33 Wyville-Thompson Ridge 250 Wyville-Thomson Ridge 69,74 Ymir Ridge 74 Ypresian NW Europe 328, 345 base of 338 sequences 176-8, 180 Zin Valley section 326 Zumaya section 326

Correlation of the Early Paleogene in Northwest Europe edited by B. W. O'B. Knox (British Geological Survey, Nottingham, UK) R. M. Corfield (Department of Earth Sciences, University of Oxford, Oxford, UK)

and R. E. Dunay (Mobil North Sea Ltd, London, UK) The early Paleogene of northwest Europe has been the subject of intense investigation over the last 25 years, with important stimuli being provided by the search for oil and gas in the offshore basins and by lUGS-sponsored investigations of the onshore historical stage and stratotype sections. The book includes three categories of papers on the Paleocene and Eocene of NW Europe: detailed aspects of local stratigraphy in the North Sea, Denmark, Belgium, SE England and offshore NW Scotland; regional syntheses of the biostratigraphy and sequence stratigraphy in NW Europe; and papers placing the successions of NW Europe in a global context, primarily through correlation with oceanic sections of the eastern Atlantic. Topics covered include: biostratigraphy (foraminifera, calcareous nannofossils, dinoflagellate cysts, diatoms, mammals), isotope stratigraphy, tephrostratigraphy, igneous history, tectonic evolution, and sequence stratigraphy. The reader will have access to substantial amounts of new stratigraphic data and to compilations of existing data that are based on greatly improved regional and global stratigraphic frameworks. 9 488 pages 9 250 illustrations 9 22 papers 9 index Cover illustration: Paleocene/Eocene boundary section,Central North Sea

I ! !L! !I I! ! ISBN 1 - 8 9 7 7 9 9 - 4 7 - 0

9